jcr1de1

Epigene'c mechanisms in colorectal cancer Joaquín Custodio Rojo | Tesi doctoral 2014 Epigenetic mechanisms in colorectal
cancer
Memòria presentada per
Joaquín Custodio Rojo
per optar al grau de
Doctor en Biologia Cel·lular
Tesi realitzada sota la direcció del Dr. Miquel Àngel Peinado Morales
a l’Institut de Medicina Prediciva i Personalitzada del Càncer
Tesi adscrita al Departament de Biologia Cel·lular, de Fisiologia i
d’Immunologia de la Facultat de Medicina
Universitat Autònoma de Barcelona
Bienni 2013 – 2014
Director
Miquel A. Peinado Morales
Tutora
Rosa Miró Ametller
Barcelona, setembre de 2014
Doctorand
Joaquín Custodio Rojo
“La frase más excitante que se puede oír en ciencia, la que anuncia nuevos descubrimientos, no es ¡Eureka! (¡Lo encontré!) sino 'Es extraño...'.” “La ciencia se construye a partir de aproximaciones que gradualmente se acercan a la verdad.” ―Isaac Asimov― Als meus avis i iaios, Als meus pares i germà i a l’Anna, Gràcies INDEX ABBREVIATIONS INTRODUCTION COLORECTAL CANCER Colorectal cancer features Incidence and mortality Colorectal cancer risk factors Non-­‐modifiable risk factors Environmental risk factors Colorectal cancer as a stepwise progression model Screening and diagnosis Treatments EPIGENETICS DNA methylation, the fifth base DNA demethylation, the sixth, seventh and eighth bases of the genome Histone modifications Histone variants Polycomb and Trithorax complexes RNA epigenetics, brave new world Chromatin conformation, entering in the third-­‐dimension CANCER EPIGENETICS DNA methylation in cancer Histones in cancer 15 17 19 19 20 23 23 25 28 30 32 34 35 39 40 43 44 46 47 49 50 53 ALDOSE KETO REDUCTASES IN CANCER AKR1B1 in human diseases AKR1B10 and AKR1B15 in cancer RETINOIC ACID PATHWAY OBJECTIVES MATERIALS AND METHODS SAMPLES Patients’ samples Human stool samples and associated carcinoma tissues Mice samples Cell lines EXPRESSION VECTORS cDNA Vectors Vector amplification CELL CULTURE DNA transfections Lentiviral infection Dznep treatment All-­‐trans retinal treatment 5-­‐aza-­‐2'-­‐deoxycytidine treatment Luciferase enhancer assays Cell Proliferation Assay XTT Genome editing: CRISPR DNA METHYLATION ANALYSIS 56 56 58 62 67 71 73 73 73 74 74 76 76 77 77 77 78 78 78 78 79 81 82 84 Genomic DNA extraction Bisulfite treatment 84 84 Bisulfite sequencing 85 Infinium 450K methylation arrays 86 RNA EXPRESSION ANALYSIS 86 RNA extraction 86 Reverse transcriptase PCR 87 Quantitative Real-­‐Time PCR 87 CHIP AND CHIP-­‐SEQ EXPERIMENTS 88 PROTEIN ANALYSIS 90 Western blot 90 Immunohistochemistry 90 CHROMOSOME CONFORMATION CAPTURE (3C) 91 BIOINFORMATICS 92 The Cancer Genome Atlas project 92 TCGA: Gene expression 92 TCGA: DNA methylation 93 TCGA: Gene mutation and copy number alterations 94 Other bioinformatic tools 94 RESULTS 97 DNA METHYLATION PROFILE OF AKR1B GENES IN COLORECTAL CANCER 99 DNA methylation of AKR1B1 by bisulfite sequencing 99 DNA methylation of AKR1B10 by bisulfite sequencing 101 DNA methylation of AKR1B15 by bisulfite sequencing 102 AKR1B1 DNA methylation using TCGA data AKR1B10 DNA methylation using TCGA data AKR1B15 DNA methylation using TCGA data AKR1B1 DNA methylation in stools AKR1B1 CpG island methylation in other cancers Akr1b3 DNA methylation in mice mRNA EXPRESSION OF AKR1B GENES IN COLORECTAL CANCER mRNA expression of AKR1B1 in colorectal cancer samples mRNA expression analysis of AKR1B10 and AKR1B15 in colorectal cancer samples mRNA expression of AKR1B family in a second set of samples mRNA expression of AKR1B family in TCGA samples mRNA expression of AKR1B family genes in cancer cell lines Expression of genes near AKR1B locus Coexpression of AKR1B genes Expression of AKR1B family in mouse colon PROTEIN LEVEL OF AKR1B GENE FAMILY IN COLORECTAL CANCER Expression of AKR1B1 and AKR1B10 at the protein level AKR1B10 in colorectal cancer tissue microarrays AKR1B1 in colorectal cancer tissue microarrays EPIGENETIC GENE REACTIVATION THROUGH 5-­‐AZAC TREATMENT AKR1B genes DNA methylation and expression after 5-­‐AzaC treatment Chromatin profile of the AKR1B locus H3K27ac in normal tissues and in 5-­‐AzaC treated HCT116 cancer cell line 104 105 106 107 110 111 114 114 115 116 118 119 120 121 123 125 125 125 128 131 131 132 134 CHROMOSOME CONFORMATION CAPTURE (3C) LUCIFERASE ASSAYS Enhancer luciferase assays Methylation-­‐enhancer luciferase assays CRISPR ASSAYS CRISPRing the TFBS CRISPRing NF-­‐Y and C/EBP-­‐β genes HCT116 DZNEP TREATMENT FUNCTIONAL IMPLICATIONS OF AKR1B10 AND AKR1B15 RE-­‐EXPRESSION IN COLORECTAL CANCER Proliferation assay XTT assay RETINOIC ACID PATHWAY Expression profiling of the retinoic acid pathway in colorectal tissue through The Cancer Genome Atlas and RefExA databases Altered Expression of Genes involved in retinoic acid pathway is directly linked to survival in colorectal cancer patients Promoter DNA methylation profiling of the RA pathway in normal and colorectal tumor tissue Discovery of new genes related with the RA pathway by analysis of DNA methylation co-­‐regulation Demethylation of the retinoic acid pathways partially restored the altered gene expression Genomic profiling of the RA pathway in colorectal cancer DISCUSSION EPIGENETIC REGULATION OF AKR1B1 GENE FAMILY IN COLORECTAL CANCER Epigenetic regulation of AKR1B1 136 138 138 140 141 141 142 143 144 144 144 147 147 148 150 151 152 153 157 159 159 Epigenetic regulation of AKR1B10 and AKR1B15 ENHANCER ACTIVITY OF AKR1B1 CpG ISLAND Chromatin conformation in the AKR1B locus A distant enhancer regulates the expression of AKR1B10 and AKR1B15 H3K27ac and PRC2 dynamics in the activation/silencing of AKR1B10 and AKR1B15 Gene regulation by enhancer hypermethylation in cancer FUNCTION OF AKR1B10 IN CANCER, LOCATION, LOCATION AND LOCATION Consequences of AKR1B10 and AKR1B15 loss of function in colorectal cancer DEREGULATION OF THE RETINOIC ACID PATHWAY IN COLORECTAL CANCER RETINOIC ACID PATHWAY GENES AS DIAGNOSTIC MARKERS AKR1B1 CpG island DNA methylation as a non-­‐invasive biomarker in stools AKR1B10 protein downregulation in colorectal cancer RETINOIC ACID PATHWAY GENES AS PROGNOSTIC MARKERS CONCLUSIONS BIBLIOGRAPHY ANNEX ANNEX I. TRANSFECTION AND INFECTION VECTORS ANNEX II. BISULFITE PRIMERS ANNEX III. EXPRESSION PRIMERS ANNEX IV. LUCIFERASE PRIMERS CONSTRUCTS ANNEX V. CRISPR TARGET SEQUENCE 161 163 163 163 164 167 169 170 173 176 176 178 179 181 185 209 211 212 213 214 215 ANNEX VI. METHYLATION PROBES ANNEX VII. CHIP PRIMERS ANNEX VIII. 3C PRIMERS ANNEX IX. EXPRESSION PROBES ANNEX X. CHIP-­‐SEQ DATA ANNEX XI. P53 TMAS ANNEX XII. ARTICLE 216 217 218 219 220 221 222 3C
5-AzadC
5-FU
5caC
5fC
5hmC
5hmU
5mC
AKR
ATRA
ChIP
CNA
COAD
CpGi
CRC
CRISPR
CTCF
DNMT
DZNep
ecCEBPA
EGFR
EMT
eRNAs
FAP
FDA
FOBT
GO
HDAC
HNPCC
IBD
ISC
Kb
KEGG
LMR
lncRNA
miRNA
ncRNA
bp
PRC
PTMS
RA
RAR
RARE
RXR
TCGA
TDG
TET
TF
TFBS
TMAs
TNM
TrxG
TSS
VEL
chromosome conformation capture
5-aza-2’-deoxycytidine
5-Fluorouracil
5-carboxylcytosine
5-formylcytosine
5-hydroxymethylcytosine
5-hydroxymethyluracil
5-methylcytosine
aldo-Keto reductase
all-trans retinoic acid
chromatin immunoprecipitation
copy number alteration
colorectal adenocarcinoma
CpG island
colorectal cancer
clustered regularly interspaced short palindromic repeats
CCCTC-binding factor
DNA methyltransferase
3-Deazaneplanocin A hydrochloride
extra-coding CEBPA ncRNA
epidermal growth factor receptor
epithelial to mesenchymal transition
enhancer RNA
familial adenomatous polyposis
Food and Drug Administration
fecal occult blood test
gene ontology
histone deacetylase
hereditary non-polyposis colorectal cancer
inflammatory bowel disease
intestinal stem cells
Kilobase
Kyoto Encyclopedia of Genes and Genomes
low-methylated region
long non-coding RNA
micro-RNA
non-coding RNA
base pair
Polycomb repressive complex
post-translational modifications
retinoic acid
retinoic acid receptors
retinoic acid response elements
retinoic X receptors
The Cancer Genome Atlas
thymine DNA glycosylase
Ten-eleven translocation
transcription factor
transcription factor binding site
tissue microarrays
tumor node metastasis
Trithorax group
transcription start site
variant enhancer loci
c eCi a eti c
eoea n o r
a
nf a i di C e d c
Cx
%tl 1Ci a
s c ac i c eCi dd
I Ci R en i en
C eau
ddl x tI aC q%I c C e
s I c et i C 1tI c et
I ci u t
c dsl tl l ni R en e i di c
c
c
C e d eau i Cl
i di c i C
de C eti c l 0
ce
C I c et dds en l u
C x a n Cd 1 et F
n d CW L8qL%
0
l su 1ei u l
c
i di C e d
c
i c en
di
l tI c l i C
1 c
eti c i en
eau i CW tc I en
u c t l e eti c l
en
n
dllt d
di i
l ei i dl W
i c l et1 eti c W
l ei i d
tc
c ut W
C l
tc
l tp W di l l
i 11 ete Wdi l l i R tI ne
c
i ad
dl i tc da
c al
iC
( i u tetc I
x tre c
rau
W qóó: ,
C
f a p d r nel i e n oes a
eoer ma nf l r en a ea r h
Wqóóó%
0
ci
i u u i c es1 i i di C e d
0
ci
ci
tl en
CW i u 1Ctl tc I
C tc i u l i CtI tc e Ci u en
i c z u tdt C
en l CC e
c
C tc i u
n u i n l m es r n
c n ael s s s hr h euh
l e oc i R c
ci
C tc i u
tl
i CC e i ddi R za1 c eC eu c ei en 1 et c e0 CC e
C tc i u l en c eni l i c l i ddi R tc I en
x eetc I ei c
L8qZ%
0
l
c entl en l tl W R
tc
1 c
u tc i Ctes i i di C e d
i c ( c eti c d i C
tl 1d s
x tI aC L%
0
Ca t d ei
t
en
c i u l e c ei ( i d(
u i C C 1t ds ei
i CtI tc c tc da
u i le
1ten dtau i en tc e l etc 0
C tc i u l i ddi R eR i u tc 1 enR sl Wen
i en es1 i en
11Ci ) tu e ds ó: D i en ei e d
i c W 1 c tc I i c en u i d ad C de C eti c l
n 1Ci 1 C t I c i l
c
c
i c ( c eti c d1 enR s
Cl n l
c i c z 1ten dt d
C tc i t eau i Cl Wl C i u l c dsu 1ni t eau i Cl 0
R tdd C C ei
c eds t en s C tc da
c i c z u tdt C
tc en
i di C e d
i c ( c eti c di Cen l CC e
ci
C tc i u W
1 enR s0
qó
c eCi a eti c
380
M Bettington et al.
Figure 11. Putative pathways to colorectal cancer.
f a 2 d f n n u c n s i ne eoea n o r
ah
c ae f
ael
nn r ner
2. pPh
five features or with 10–20% necrosis are considered
fashion; epithelial serrations are typically lost (Figto be equivocal.110
ure 12F). Micropapillary structures are sometimes
The second pattern, accounting for 20% of cases, is
seen, and lymphatic invasion is common. Serrated
r mucinous
r adenocarcinoma
r l ean o(Figure
ni 12D).35,109 Extra- morphology carcinoma cytology remains apparent.
cellular mucin constitutes at least 50% of the
This pattern is often a minor component of the other
i cross-sectional
di C e d area,
c Cand
tl by definition
u ri C theyalare poorly
i u i Csubtypes,
t tes usually
c uati the
Ce advancing
dtes enCiedge
aI of
nithe
aetumour.
en
differentiated. Serrations can usually be identified, at
Serrated morphology carcinomas segregate as a
R ileast
Cd focally,
xR R Rand
0R typical
ni 0i CIserrated
%
0 i dcarcinoma
i C e d cytology
c C tl distinct
en lmolecular
i c usubset,
i l e strengthening
i u u i c thealassertion
i is retained. Cell balls and papillary rods floating in
that they represent a reproducible and distinctive
111
KRAS
the
are R
common,
are C
characteristic
subtype of cCRC.en
enmucin
R i Cd
t x and
tc pd
F i I dofl ethis
tc Wqóó/%
entCand uBRAF
i l emutations
i u u ihave
c
been identified in 45% and 33% of cases, respecpattern (Figure 12E).
112
in 16%,candteMSI-L
pattern
trabecular,
c TheC final
tc en
R iisCd
WR ten constituting
q0LE u tdapproxidti c c R tively,
l l MSS
t Iinc i50%,
l MSI-H
tc L887
n l in
30%.110 These figures suggest that serrated pathways
mately 7% of cases.35,109 These tumours are poorly differentiated,
cellsCgrowing
a trabecular
and
described
c l etu with
e tumour
c tc
l i in en
tc t c 1b W
u 2,otc
I a1 above,
ei earedoverrepresented.
i L0L u tddti c
l l t I ci l
i di C e d
c
1 Cs
C el u c c
ntI n C tc en tc al eCt dtp
R i u c du i l e 9a ddsW c
L8
l e Cc
aCi 1
en tc t c
i ac eCt l x i ae eR i zentC l %x i sd F
u i c I en u en ntI n l e tc t c
dc W
© 2012 Blackwell Publishing Ltd, Histopathology, 62, 367–386.
C s L8Z80
c
i i aen Cc
i di C e d
aCi 1 ,
c
c
C tl tc
tl
i c WL88L%
W
al eC dt W R
en di R l e tc t c
tc
c eCi a eti c
t
F
d c
l e Cc Ct
c
i aenz c eC d l t xnee12MMR R R 0R C 0i CI %x i sd
c I u c WL888%
0
otc I
0 0 l c ) u 1d i tc al eCt dtp
tc t c
c tc 1 Cet ad C i di C e d c
L8qZ%
0 c en
1 i 1d I
0 i di C e d c
i ac eCt l x i sd F
x tI aC Z%
0 tl
l
40 to 59
0.94 (1 in 106)
0.75 (1 in 134)
C
d q%x t I dW
c
C u i cI
C
e
60 to 69
1.40 (1 in 71)
0.98 (1 in 102)
h1 pz
Cdt C l e I l
i Fm r
0 c tc u i l ei en tc al eCt dtp
ce
WL8q8,
t I dWL8qZ%
a ei l C c tc I 1C et l c en
t t c e eC eu c e C I tu c l
en0
70 and older
4.19 (1 in 24)
3.80 (1 in 26)
r a s n r o n
r n au o ael ei o |
er m2. . 2h
C l tc I tc en
c I u c WL888,
e e
( td tdtes i u i C
c
c
i c WL88L%
0
c ae
o ni vb , e eoea n o
a
ael 2. . 3 ne 2. . Th
cn
i di C e du i Ce dtes C e tl
C a eti c tc
Ctc C l l R ten I x
C tl en u i l e C 9a c e i Cu i Cl c i d Cx i sd F
Birth to 39
0.08 (1 in 1212)
0.08 (1 in 1236)
o p d r
r
r n
n n e l
c i l C( en e
Cancer Statistics, 2013
U: s
Male
Female
i ac eCsWR
11 C ei n (
Ct( c entl
FIGURE 4. Trends in Death Rates Among Males for Selected Cancers, United States, 1930 to 2009.
Rates are age adjusted to the 2000 US standard population. Due to changes in International Classification of Diseases (ICD) coding, numerator information
has changed over time. Rates for cancers of the lung and bronchus, colorectum, and liver are affected by these changes.
Cancer Statistics, 2013
A
Cancer Statistics, 2013
Cancer Statistics, 2013
original article
Annals of Oncology
B
FIGURE 5. Trends in Death Rates Among Females for Selected Cancers, United States, 1930 to 2009.
FIGURE 4. Trends in Death Rates Among Males for Selected Cancers, United States, 1930 to 2009.
arenumerator
age adjusted
to the 2000 US standard population. Due to changes in International Classification of Diseases (ICD) coding, numerator information
Rates are age adjusted to the 2000 US standard population. Due to changes in International Classification of Diseases (ICD) Rates
coding,
information
has changed over time. Rates for cancers of the uterus, ovary, lung and bronchus, and colorectum are affected by these changes.
has changed over time. Rates for cancers of the lung and bronchus, colorectum, and liver are affected by these changes.
*Uterus includes uterine cervix and uterine corpus.
FIGURE 6. Total Number of Cancer Deaths Averted From 1991 to 2009 in Men and From 1992 to 2009
18 in Women.
CA: A Cancer Journal for Clinicians
The blue line represents the actual number of cancer deaths recorded
year,
and the red
represents
the number
of cancer
FIGURE in6.each
Total
Number
of line
Cancer
Deaths
Averted
Fromdeaths
1991that
to would
2009have
in Men and From 1992 to 2009 in Women.
been expected if cancer death rates had remained at their peak.
The blue line represents the actual number of cancer deaths recorded in each year, and the red line represents the number of cancer deaths that would have
been expected if cancer death rates had remained at their peak.
lung, and by more than 40% for prostate cancer. The 40 years, while lung cancer ranks first among men aged
years
and40%
older.forAmong
brain
decrease in lung cancer death rates—among lung,
men and
sinceby 40
years,
while lung cancer ranks first among men aged
more
than
prostatefemales,
cancer.tumors
The of40the
1990 and among women since 2002—is due to the and other nervous system are the leading cause of cancer
decrease in lung cancer death rates—among men since 40 years and older. Among females, tumors of the brain
20
reduction in tobacco use, while the decrease in death rates death among children and adolescents (aged younger than
1990 and among women since 2002—is due to the and other nervous system are the leading cause of cancer
for female breast, colorectal, and prostate cancers largely 20 years), 20breast cancer ranks first among women aged
deathcancer
among children and adolescents (aged younger than
reduction
in tobacco
whileand
the lung
decrease
in death
20 to use,
59 years,
cancer
causesrates
the most
reflects improvements in early detection
and/or
colorectal,
and 60
prostate
cancers
deaths
in those aged
years and
older.largely 20 years), breast cancer ranks first among women aged
treatment.17,21,22 Over the past 10 yearsfor offemale
data breast,
improvements in early detection and/or 20 to 59 years, and lung cancer causes the most cancer
(2000-2009), the largest annual declines in deathreflects
rates were
for chronic myeloid leukemia (8.4%), cancers
of the17,21,22
treatment.
Over Variations
the past in10Cancer
years Rates
of data deaths in those aged 60 years and older.
Regional
stomach (3.1%) and colorectum (3.0%), and non-Hodgkin
(2000-2009), the
largest
annual
declinescancer
in death
rates were
Tables
8 and
9 depict
incidence
and death rates
lymphoma (3.0%).
for selected cancers by state. Lung cancer shows the
Figure 1. Colorectalf cancer
incidence
mortality
(adjusted
age) among
men and women age
a P din ,Spain.
a r Trends
r linean
o ni a (adjusted
n ea for
o age
n and rregistry)
a iand
Fm
r r nrates n
n aelforpT3B
ne
>29 years, obtained
model.
The ,vertical
when
2. from
. Tm thec change-point
n ael
om2. pP
a r grey
r llines
eanshow
o ni athe
n time
ea interval
eoea n
o rthe achange
r c inr trend
v f occurred.
n
ea
, l er lfor chronic
r r myeloid
s elleukemia
r (8.4%),( cancers
2T i of the
a hRegional
c nVariations
aelin Cancer
1c Rates
gw r n
m2. p. h
geographic(3.0%),
variation
cancer occurrence by far,
stomach (3.1%)largest
and colorectum
andinnon-Hodgkin
8 and 9 depict cancer incidence and death rates
Recorded Number of Deaths From Cancer in 2009 reflecting the large historical and continuing Tables
differences
lymphoma (3.0%).
20
for selected
by state. Lung cancer shows the
in
smoking
prevalence
among
states.
A
total
of
2,437,163
deaths
were
recorded
in
the
United
For example,
lung cancers
Table 2.
Trends
in
incidence
(1975–2004)
and
mortality
(1975–2007)
in
Spain
among subjects aged >29 years, and identification
23
geographic variation in cancer occurrence by far,
States in 2009, 567,628 of these from cancer. Cancer is cancer incidence rates in Kentucky, which has largest
historically
Recorded
Number
of
Deaths
From
Cancer
in
2009
reflecting
4-foldthe large historical and continuing differences
the second leading cause of death, following heart disease, had the highest smoking prevalence, are almost
higher than
those
Utah, which
hasUnited
the lowest
accounting for 23% of all deaths. However, within
20-year
in smoking prevalence among states.20 For example, lung
A total
of 2,437,163
deaths
wereinrecorded
in the
23 100,000 men). In conprevalence
(128
vs
34
cases
per
age groups, cancer is the leading cause of death among
both
States in 2009, 567,628 of these from cancer. Cancer is cancer
incidence rates
in Kentucky, which
has historically
Change-point
Annual
percentage
change
(95% CI)
trast, state variations for other cancer sites are smaller in
men and women aged 40 to 79 years (Table 6).
the second leading
cause of death, following heart disease, had the highest smoking prevalence, are almost 4-fold
absolute
and CI)
proportionate terms. For example,
Table 7 presents the numbers
of deaths for all cancers both
P-values
Year
(95%
Overall
Below change-point
accounting for 23% of all deaths. However, within 20-year higher than those in Utah, which has the lowest smoking
combined and for the 5 most common sites for each the breast cancer incidence rate in Connecticut, which
prevalence
thehighest
leadingrate
cause(137
of death
has is
the
per among
100,000both
women),
is only (128 vs 34 cases per 100,000 men). In con20-year age group. Among males, leukemia is age
the groups,
leading cancer
Men cause of cancer death among those aged younger
state variations for other cancer sites are smaller in
men and
women
aged
40 to than
79 years
6).
28%
higher
that(Table
in Arizona,
which has trast,
the lowest
than
FIGURE 5. Trends in Death Rates Among Females
Selected
States,
to 2009. both absolute and proportionate terms. For example,
Table 7forpresents
theCancers,
numbersUnited
of 1997)
deaths
for1930
all cancers
Incidence
1995
(1993,
3.42information
(3.25, 3.59)
4.26 (3.92, 4.61)
Rates are age adjusted to <0.0001
the 2000 US standard population. Due to changes
in International
Classification
of Diseases (ICD) coding, numerator
20
CA: A Cancer Journal for Clinicians
combined
for theand5 colorectum
most common
forchanges.
each the breast cancer incidence rate in Connecticut, which
has changed over time. Rates for cancers of the uterus,
ovary, lungand
and bronchus,
are affectedsites
by these
and uterine corpus. 20-year age group.
Mortality *Uterus includes uterine cervix
<0.0001
1998
(1996,
2000)
1.86
(1.75,
1.97)
2.88
Among males, leukemia is the leading has the highest rate (137 per 100,000 women), is only (2.72, 3.05)
Lq
of changes in trend.
Above change-poin
2.53 (2.17, 2.89)
–0.39 (–0.70, –0.08
c eCi a eti c
i R ( CWen tc
) i tc t c
i ac eCsW i C ) u 1d tc en
tc
1 tc x c
x h1 pz
ce
u i l e i en
) nt tel
0 en tc t c
aCi 1 c
t C c e 1 ee Cc l
n l
C l
1 c tc I i c en
x t I dWL8qZ%
WR ntd
i ac eCt l %tl l etdd tc C l tc I x tI aC E%
WL8q8%
0
CA CANCER J CLIN 2013;63:11–30
A
original article
Annals of Oncology
FIGURE 3. Trends in Incidence Rates for Selected Cancers by Sex, United States, 1975 to 2009.
B
Rates are age adjusted to the 2000 US standard population and adjusted for delays in reporting.
*Liver includes intrahepatic bile duct.
According to data from the SEER 13 cancer registries, inciThe lifetime probability of being diagnosed with an dence rates in the most recent 5 years (2005-2009) decreased
invasive cancer is higher for men (45%) than for women in males by 0.6% per year and were stable in females (Table
(38%) (Table 4). However, because of the earlier median 5). Incidence rates are decreasing for all 4 major cancer sites
age at diagnosis for breast cancer compared with other except female breast, for which rates remained relatively
major cancers, women have a slightly higher probability of stable from 2005 to 2009 after decreasing by 2% per year
developing cancer before age 60 years. These estimates are from 1999 to 2005. Lung cancer incidence rates in women
f a R d , a r r r
r a n ea o n
r a i Fmr r n
n n ael pT3B ne
began declining in the late 1990s, more than a decade after
based2.on. Tm
the average
ofomthe
general
c n experience
ael
2. pP
, population
a r r r
r
a n ea eoea6 n o r a r c r
lungrcancer
incibegan
and may
v overf n oreaunderestimate
r a individual
nai , l erriskl because
r r sofel the
r decline
( 2T
i a inmmen.
c n Differences
ael 1c ingw
n
dence
patterns
between
men
and
women
reflect
historical
2. exposure
p. h
differencesmin
(eg, smoking history) and/or genetic
differences in tobacco use; cigarette smoking prevalence
susceptibility.
i C i ( CWen C C c i e i c ds t C c l peaked
u i cabout
I i20
acyears
eCt later
l W in0women
tc t than
c in men.
c 16 Recent
Trends in Cancer Incidence
rapid declines in colorectal cancer incidence rates have
u i Ce dtes C e l ( Cs i c l t C ds u i c I C t d c enc t I Ci a1l x
d L%
0
Figures 2 to 5 depict long-term trends in cancer incidence largely been attributed to increases in screening that can
17-19
andand
allow
the removal
of precancerous
polyps.men
and death
rates
for all Trends
cancersincombined
and for selected
Figure 1. Colorectal
cancer
in Spain.
incidence (adjusted
for age anddetect
registry)
mortality
rates (adjusted
for age) among
and women age
Although
trendthe
analysis
thatoccurred.
the incidence
cancerfrom
sitesthe
by change-point
sex. While incidence
ratesvertical
are declining
forshow
>29 years, obtained
model. The
grey lines
the time joinpoint
interval when
changeshows
in trend
most cancer sites, they are increasing among both men and rate for prostate cancer declined steadily by 1.9% per year
women for melanoma of the skin and cancers of the liver from 2000 to 2009, it is important to realize that annual
rates
fluctuate
widely
likely
variation
in
andinthyroid
(Fig.
3, Table and
5). Table
5 shows
incidence
Table 2. Trends
incidence
(1975–2004)
mortality
(1975–2007)
in Spain
among
subjects
aged(Fig.
>29 3),
years,
andreflecting
identification
of changes
in trend.
(delay-adjusted) and mortality trends for all cancers com- the prevalence of prostate-specific antigen testing for the
binedLL
and for selected cancer sites based on joinpoint regres- detection of prostate cancer. For example, in the SEER 13
Change-point
Annual percentage change (95% CI)
sion analysis. Joinpoint is a tool used to describe and quantify areas, the delay-adjusted prostate cancer incidence rate
P-values
Year (95% CI)
Overall
Below change-point
Above change-poin
trends by fitting observed rates to lines connected at ‘‘join- increased from 154 (per 100,000) to 164 from 2005 to 2006,
13,15
Men
Lifetime Probability of Developing Cancer
c eCi a eti c
White
African American
Male
Female
52.8
65.1
39.2
48.0
Male
Female
19.5
29.8
13.6
19.8
Assian American and
Pacific islanders
American Indian
and Alaska native
Hispanic
41.4
50.7
46.9
32.1
41.1
33.3
13.1
18.8
15.3
9.6
14.6
10.2
Incidence
Mortality
o 2d r
2. . Th
cn
n l
r
ael
r
l ean o ni a n om2. pPh
tl l tu td Ctet l tc
i Fma
i di C e d tc t c
i ac eCt l i C enc t dI Ci a1l u s
) 1i l aC ei Ctl o r
c
eeCt ae
ei Cl x1i ddaeti c tc en
nr
ni m r
r n
n n ael 2. . B ne
u i Ce dtes
eR
tc 1 Ce ei oc i R c c
t C c e
ei Cl l a n l
tCW t eW d i ni d tc e o T %
W
l l ei
ntI nz9a dtes l C c tc I Wetu ds t I c i l tl c en 9a dtes i en eC eu c e0
eoea n o r
( C dCtl o C
a a C
ei Cl C
nea lli t e
eR i es1 l i Ctl o c ( tCi c u c e dCtl o ei Cl 2 en
o a C
c
i i di C e d c
c i c zu i t t d Ctl o c
i di C e d
) u 1d i te0 i C en c ó8D i i di C e d c
C
l l i
C tl
c
I zC d e
WL887%
0 n tc t c
Cl i d 1 i 1d en c tc eni l
tc t c
tl I Ci R tc I tc si ac I 1 i 1d 0 i R
eR
en
q8 u i l e i u u i c ds
t I ci l
k
c i u ei al
Cl i c d tl ei Cs i I Ci R en i en
i di c
( di 1tc I
l su 1ei u et
i ( C en
ci u
c tI c
I i : 8 s
i di C e d
c
Cl
Ctl
c
tl
C tl
d C
:8
Cl 0 i R ( CWen
0 W i di C e d c
C tc R i u c
c
i ds1l 2 i ds1l C
Wqóó/%
0 n
) eC u ds ntI n, te n l
u c
i c l 9a c
i di c 1ten dtau etl l a x tI aC : % c
C tc i u l x tc R C
c
sl tc en
c
WL88/%
0
d l ti c l i en
aC tc 1 i 1d I
si ac I C en c E8 s
c L8 c Eó x tCd s
cn c
c
C e tl u i C en c : 8 etu l ntI n Ctc /8
ei Uó s
u i c I en
ei Cl
nea y
tddc l l
i Ci d Cx t l
C0 n C
ei Cl 0
er wl e
k I 2
R ten en tc t c
C tl
I
i C 1C aCl i C
dt etu
c l aI I l e
Ctl o i en e
c i u l C i ac tc u i C en c E8D i en tc t( t a dl
Cl x dt el ei l F
Ci u en l
i ado l WL88/%
0
ci u l x
II C F
Cds ó: D i l 1i C t
i al n sWL88ó%
0 n
LZ
c eCi a eti c
eC eu c e i c l tl el tc en C u i ( d i en 1i ds1l WR nt n C a l en Ctl o i ( di 1tc I i di C e d C tc i u l x tc R C
Wqóó/%
0
A
B
v ,
f a B dv ,
l o a
l n m neoe
o r l
ae ec a c a r n n er e
r o eoea n o c eoi c h
r el nef c eoi c e eoer c n r nh
c n ael s s s hs C c
hea 0
k c n Cte
I c et
Ctl oW u tdt d
n C te Cs c i c z1i ds1i l tl
i di C e d
c
c i u ei al 1i ds1i l tl x
C x
i c teti c tc R nt n c au Ci al 1i ds1l C i Cu
tc e l etc x tI aC : %
0
x Ct d
WL88q,
( di 1u c e i u adet1d
LE
tl
al
ddtl
%
2
tc en
tl
% c
c tc n Cte
1ten dtau i en d CI
s u ae eti c l tc en eau i C l a11C l l i C I c
Wqóóó%
0
c i u ei al 1i ds1l Wte n l
tl
n C e Ctp
c tc t c
s en
e tCen i Introduction about 1/8,300 and it manifests equally in both sexes. FAP just accounts for less than 1% of colorectal cancer cases (Galiatsatos & Foulkes, 2006). Most of the colorectal cancer patients (~70%) have a family history of colorectal polyps and cancer. Generally cancer starts to develop a decade after the appearance of the polyp, they normally are asymptomatic until the large number of them cause rectal bleeding or other complications. Without treatment, FAP individuals carry 100% of possibilities of developing colorectal cancer. Hereditary non-­‐polyposis colorectal cancer (HNPCC) or Lynch syndrome is another inherited disease caused by mutations on DNA mismatch repair genes, mainly MLH1 and MSH2. In addition to colorectal cancer, HNPCC patients show an increased relative risk in other cancers such as uterus, stomach, small bowel, pancreas, kidney and ureter. In contrast to FAP patients, HNPCC individuals develop few adenomas. Without treatment individuals affected with HNPCC have an approximate 80% lifetime risk of developing colon cancer (Lynch et al., 2009; Weissman et al., 2011). · Inflammatory bowel disease (IBD): IBD includes two diseases: ulcerative colitis and Crohn disease. Ulcerative colitis causes inflammation of the mucosa of the colon and rectum. Crohn disease causes inflammation of the full thickness of the bowel wall and may involve any part of the digestive tract from the mouth to the anus. These inflammatory conditions are known to increase the risk of developing colorectal cancer (Triantafillidis et al., 2009). In patients with inflammatory bowel disease the relative risk of developing colorectal cancer has been estimated to be increased 4 to 20 fold (Haggar & Boushey, 2009). Environmental risk factors: · Dietary and nutritional practices: Evidences from epidemiological studies highlight the strong influence of the diet in the risk of developing colorectal cancer (Boyle & Langman, 2000). It is believed that changing food habits might reduce up to 70% the possibilities of suffering colorectal cancer (Johnson & Lund, 2007). Diets enriched in animal fat are a major risk factor for colorectal cancer. The specific mechanisms that underlie the association between red meat and colorectal cancer are unclear. It has been proposed that red meat might stimulate secretion of endogenous insulin, which is a natural mitogen and 25 Introduction perhaps enhances the uncontrolled cellular growth (Chan & Giovannucci, 2010). Other relevant hypothesis propose red meat as a source of heme iron, which catalyzes the oxidation of polyunsaturated fats to some compounds that are known risk factors for several human diseases, such as Malondialdehyde or 4-­‐
Hydroxynonenal (Bastide et al., 2011). Fiber intake has also been linked with colorectal cancer where a reduced impact was observed in some African populations with high-­‐fiber diet more than 4 decades ago (O’Keefe et al., 1999). Since then, many studies have been testing this hypothesis, some of them successfully and others not (Chan & Giovannucci, 2010; Huxley et al., 2012). Anyway, it has been proposed that fiber could dilute or adsorb fecal carcinogens, modulate colonic transit time, alter bile acid metabolism, reduce colonic pH, or increase the production of short-­‐chain fatty acids (Huxley et al., 2012). It is well known that the composition of the individual’s flora can fluctuate under some circumstances and is deeply affected by dietary habits (Grölund et al., 1999; Smith, 1965). The human gut is the natural habitat for a large and dynamic bacterial community but a substantial part of these bacterial populations are still to be described. Out of curiosity, in the large intestine we can find bacterial concentrations up to 1011 or 1012 cells/g of luminal contents (Guarner & Malagelada, 2003), these concentrations are similar to those found in colonies growing under optimum conditions over the surface of a laboratory plate. Intestinal bacteria could play a role in initiation of colon cancer through production of carcinogens, co-­‐carcinogens or pro-­‐carcinogens. Furthermore, some intestinal microorganisms strongly increase damage to DNA in colon cells induced by heterocyclic amines, whereas other intestinal bacteria can uptake and detoxify such compounds (Wollowski et al., 2001). · Physical Activity and Obesity: Evidences from epidemiologic studies demonstrate that people with high occupational or recreational physical activity appear to be at a lower risk of developing colon cancer (Brown et al., 2012). Physical activity at a level equivalent to walking 4 h per week was associated with a decreased risk of colon cancer among women when compared to the sedentary group. 26 Introduction Obesity (commonly accompanied by diabetes type II) is also a cancer risk factor, although the molecular mechanisms are not well understood. There are some evidences that obese people have more of the mitogenic factor insulin-­‐like growth factor 1 in their bloodstream (Renehan et al., 2003), what could predispose cells to uncontrolled cellular growth. Adipose tissue is a complex, essential, highly active metabolic and endocrine organ (Kershaw & Flier, 2004). Polypeptide hormones derived from adipocytes are known as adipokines, among them, leptin and adiponectin are linked to cancer. Leptin concentration is directly correlated with the amount of body fat and is regulated by insulin. Leptin is adipogenic, antiapoptotic and mitogenic for various cell types, including hematopoietic progenitor cells, normal and transformed epithelial cells, and vascular endothelial cells (Bray, 2002; Rose et al., 2004; Vona-­‐Davis & Rose, 2007). Adiponectin is the most abundant of adipokines, it is an insulin-­‐sensing hormone and unlike leptin, its correlation is inverse with body mass index and appears in low levels in cancer patients (Roberts et al., 2010). · Alcohol intake: The risk factor that alcohol represents to some cancers like gastric or liver is clearly demonstrated, whereas in the case of colorectal cancer is not so well understood. Some studies point out that moderate and heavy drinking alcohol was associated with a 21% and 52% increased risk for colorectal cancer, respectively (Fedirko et al., 2011). The results for alcohol drinking and colorectal cancer risk appeared to be similar between men and women. · Cigarette Smoking: Colorectal cancer has not historically been linked to cigarette smoking although nowadays it is known that heavy cigarette smokers have a 2-­‐3 fold elevated risk of developing colorectal adenomas (Giovannucci, 2001). Overall, cumulated evidences within the past decade support the addition of colorectal cancer to the list of tobacco-­‐associated malignancies. It is known that tobacco’s carcinogens could reach the colorectal mucosa and produce cellular damage that could trigger the cancerous process (Fu et al., 2013; Wei et al., 2009). 27 c eCi a eti c
eoea n o r
a i di C e d c
s
Cl ei
n c s c ae a er l e
C tl
di c I etu
( di 1W( c
dd0 aCtc I entl etu
1Ci a
Ci c F
l en n ddu Col i c n c
c
i
l Wen u ri Ctes i eau i Cl e o
l i ad 1 l l ei
c i Cu d ddl u al e
eau i C x
oc i R c
u adetl e 1 tl
Ce
tc
o
( di 1
9atC
c
C tc i u
Ci al i I dl e tc Wqóó8%
0 n l
c
c
tc L888
C c
CI x
R C 1i l ead e
c n c F
tc
aeni Cl x tI aC /%x
en
l l l en e c i Cu d ddl ni ad i ( C1 l l ei
Ci al eaC l
s
tc
C ( tl
CI WL8qq%
0 n s
tc
tc
i u u dtI c c e0
Figure 3. Emerging Hallmarks an
Characteristics
B
n l
n ddu Col
i C Ci Ci u
An increasing body of research sugg
additional hallmarks of cancer are in
pathogenesis of some and perhaps
One involves the capability to mod
gram, cellular metabolism in order to
tively support neoplastic proliferation
allows cancer cells to evade im
destruction, in particular by T and B l
macrophages, and natural killer ce
neither capability is yet generalized a
dated, they are labeled as emergin
Additionally, two consequential char
neoplasia facilitate acquisition of bo
emerging hallmarks. Genomic instab
mutability endow cancer cells with g
ations that drive tumor progression.
by innate immune cells designed to fi
and heal wounds can instead result
vertent support of multiple hallmark
thereby manifesting the now widely
tumor-promoting consequences of
responses.
surface, rendering such cells hyperresponsive to otherwise-limiting amounts
of growth factor ligand; the same
outcome can result from structural alterations in the receptor molecules that
facilitate ligand-independent firing.
Growth factor independence may also
derive from the constitutive activation of
components of signaling pathways operating downstream of these receptors,
obviating the need to stimulate these
pathways by ligand-mediated receptor
activation. Given that a number of distinct downstream signaling
pathways
radiateefrom a ligand-stimulated
distinct receptor,
cancer cells have been
ability of genome maintenance s
ool aC rYet other
ah be
r functionally
n attributes
o important
ntheofi activaa thecdevelopment
ae of detect
fThe extraordinary
and resolve defects in the DNA ensures th
tion of one or anotherproposed
of theseto downstream
pathways, for
for
r n the oone
ni responding
rcancerl and
fton
n er
ue
rto the list
l ofl core
f rhallmarks naf
n er mutation are usually very low during
might
therefore
be added
spontaneous
example,
the
Rasm
signal
transducer,
al., the
Luo et al., instructions
may only
subset
r recapitulate
nn a f ra(Negrini
a etnof
ee2010;regulatory
r er2009;
Colotta
a et al.,af2009). Two
o eageneration.
nf l eaIn the course of acquiring the roster of mu
such attributes are particularly compelling. The first involves needed to orchestrate tumorigenesis, cancer c
transmitted by an activated
receptor.
major
reprogramming
of cellular energy metabolism in order to increase the rates of mutation (Negrini et al., 2010;
r
a
m
2.
.
.
r
2.
pph
Somatic Mutations Activate Additional Downstream
support continuous cell growth and proliferation, replacing the 2010). This mutability is achieved through increased
Pathways
metabolic program that operates in most normal tissues and to mutagenic agents, through a breakdown in one
High-throughput DNA fuels
sequencing
analyses
of cancer
the physiological
operations
of thecell
associated cells. The components of the genomic maintenance machiner
genomes have revealed
somatic
mutations
in certain
human
second
involves
active evasion
by cancer
cells from attack and In addition, the accumulation of mutations can be a
tumors that predict constitutive
signaling
circuits
eliminationactivation
by immuneof
cells;
this capability
highlights the dichot- by compromising the surveillance systems that norma
omous roles
of anfactor
immune
system that
both antagonizes and genomic integrity and force genetically damaged cells
usually triggered by activated
growth
receptors.
Thus,
enhances
development
and contain
progression. Both of these senescence or apoptosis (Jackson and Bartek, 200
we now know that !40%
of tumor
human
melanomas
capabilities
may well of
prove
to facilitate
the development and 2008; Sigal and Rotter, 2000). The role of TP53 is ce
activating mutations affecting
the structure
the B-Raf
protein,
progression
of
many
forms
of
human
cancer and therefore can leading to its being called the ‘‘guardian of the genom
resulting in constitutive signaling through the Raf to mitogenbe considered to be emerging hallmarks of cancer. These 1992).
activated protein (MAP)-kinase pathway (Davies and Samuels
enabling characteristics and emerging hallmarks, depicted in
A diverse array of defects affecting various compon
2010). Similarly, mutations
in the catalytic subunit of phosphoiFigure 3, are discussed individually below.
DNA-maintenance machinery—often referred to as
nositide 3-kinase (PI3-kinase) isoforms are being detected in
takers’’ of the genome (Kinzler and Vogelstein, 19
an array of tumor types,Anwhich
serve
to hyperactivate
the PI3Enabling
Characteristic:
Genome
Instability
been documented. The catalog of defects in these
kinase signaling circuitry,
including its key Akt/PKB signal
and Mutation
genes includes those whose products are involved in
Acquisition
of the multiple
hallmarks
enumerated
above depends ing DNA damage and activating the repair machinery,
transducer (Jiang and Liu,
2009; Yuan
and Cantley,
2008).
The
in
large
part
on
a
succession
of
alterations
in
the
genomes
of
repairing damaged DNA, and (3) inactivating or in
advantages to tumor cells of activating upstream (receptor)
neoplastic signaling
cells. Simply
depicted,
certain
versus downstream (transducer)
remain
obscure,
as mutant genotypes mutagenic molecules before they have damaged
conferofselective
advantage
on subclones
of cells, enabling their (Negrini et al., 2010; Ciccia and Elledge, 2010; Ja
does the functional impact
crosstalk
between
the multiple
outgrowth and eventual dominance in a local tissue environment. Bartek, 2009; Kastan, 2008; Harper and Elledge, 2007
pathways radiating from growth factor receptors.
Accordingly, multistep tumor progression can be portrayed as et al., 2006). From a genetic perspective, these caret
Disruptions of Negative-Feedback Mechanisms that
a succession of clonal expansions, each of which is triggered behave much like tumor suppressor genes, in that the
Attenuate Proliferative
by Signaling
the chance acquisition of an enabling mutant genotype. can be lost during the course of tumor progression,
Recent results have highlighted
the importance
of negativeBecause heritable
phenotypes,
e.g., inactivation of tumor losses being achieved either through inactivating m
feedback loops that normally
operate
to dampen
typesthrough epigenetic via epigenetic repression. Mutant copies of many of t
suppressor
genes,
can alsovarious
be acquired
of signaling and thereby
ensure homeostatic
the
mechanisms
such as DNAregulation
methylationof
and
histone modifications taker genes have been introduced into the mouse ge
andintracellular
Esteller, 2010;
Esteller,
2007; Jones and Baylin, result, predictably, in increased cancer incidence,
flux of signals coursing (Berdasco
through the
circuitry
(Wertz
2007),
some
clonal expansions
may well
and Dixit, 2010; Cabrita
and
Christofori,
2008; Amit
et be
al.,triggered by nonmu- their potential involvement in human cancer de
tationalDefects
changesin
affecting
the regulation
of gene expression.
(Barnes and Lindahl, 2004).
2007; Mosesson et al., 2008).
these feedback
mechanisms are capable of enhancing proliferative signaling. The
prototype of this type of 658
regulation
involves
the
Rasª2011
oncoprotein:
Cell 144,
March 4,
2011
Elsevier Inc.
the oncogenic effects of Ras do not result from a hyperactivation
of its signaling powers; instead, the oncogenic mutations
affecting ras genes compromise Ras GTPase activity, which
i c i e C C ei u d ac eti c i l 1 t t I c l Wen s i al tc
1 enR sl i C ti di I t d ac eti c l de C
9atl teti c 0
u i d ad C1i tc ei ( t R Wen
tc
c
C c
i c i e i ddi R
l 1 t t
de C eti c i en c i Cu d ddad C n o1i tc el
a ei n c I l tc en I c et l i C 1tI c et l i l 1 t t I c l 0 i
u ae eti c d l 1 eCau tc
1 C i Cu
1 tCl 0 n
UW
)i u
u ae eti c l
i di C e d
1eaC
c
EW
U LW
c
C
I c l R C
W
qW
et( etc I u ae eti c l )
C al a dds i c i I c t
x a n Cd 1 et F
CW n
ci u
l 9a c tc I i c LLE eau i C c
u i l e C 9a c eds u ae e
u ae eti c l R C tc
c
di
ed l 1Ci r e
c i Cu d u e n
W : ZW
LW
qLZ
1e i C
e
tc en
c
tc en
W
c
W
qW
W
W
I c l WR ni l
i i c qLWqZ i C /q
4, 2011 ª2011 Elsevier Inc. 647
n d CWL8qL%
0 c en l uCell 144,RMarch
i CoW
en s i u 1 C
etc I C d ( c e I c l l a n l
Z
ó0 dd
nCi u i l i u d
c l a z nCi u i l i u d n c I l tc L: U eau i Cl W e etc I L7 C aCC c e
L7
i aI d l
This illustration encompasses the six hallmark
capabilities originally proposed in our 2000 perspective. The past decade has witnessed
remarkable progress toward understanding the
mechanistic underpinnings of each hallmark.
number and thus maintenance of normal tissue architecture and
function.
f a Cancer
0 d cells,
o by deregulating
v , r thesea signals,
u become
v ,
masters of their own destinies. The enabling signals are
a f oin n
r part byoofgrowth
o a factors
r athatnbind m
r el
conveyed
large
cell-surface
receptors,
typically
containing
r nf l eaw
c ael
en r intracellular
r o l l tyrosine
n erkinaseu
domains. The latter proceed to emit signals via branched intrau
o
ec
l
r
n
h
c
n
ael
r
r|
cellular signaling pathways that regulate progression through
the cell cycle as well as cell growth (that is, increases in cell
size); often these signals influence yet other cell-biological properties, such as cell survival and energy metabolism.
Remarkably, the precise identities and sources of the proliferative signals operating within normal tissues were poorly understood a decade ago and in general remain so. Moreover, we still
know relatively little about the mechanisms controlling the
release of these mitogenic signals. In part, the understanding
of these mechanisms is complicated by the fact that the growth
factor signals controlling cell number and position within tissues
are thought to be transmitted in a temporally and spatially regulated fashion from one cell to its neighbors; such paracrine
signaling is difficult to access experimentally. In addition, the
bioavailability of growth factors is regulated by sequestration in
the pericellular space and extracellular matrix, and by the actions
of a complex network of proteases, sulfatases, and possibly
other enzymes that liberate and activate them, apparently in
a highly specific and localized fashion.
The mitogenic signaling in cancer cells is, in contrast, better
understood (Lemmon and Schlessinger, 2010; Witsch et al.,
2010; Hynes and MacDonald, 2009; Perona, 2006). Cancer cells
can acquire the capability to sustain proliferative signaling in
a number of alternative ways: They may produce growth factor
ligands themselves, to which they can respond via the expression of cognate receptors, resulting in autocrine proliferative
stimulation. Alternatively, cancer cells may send signals to stimulate normal cells within the supporting tumor-associated
stroma, which reciprocate by supplying the cancer cells with
various growth factors (Cheng et al., 2008; Bhowmick et al.,
2004). Receptor signaling can also be deregulated by elevating
the levels of receptor proteins displayed at the cancer cell
C
Figure 1. The Hallmarks of Cancer
A
tl
( c ea dds
CI WL888% c
c n c F
c i Cu d
eaC l ei
L8qq s en l u
ti di I t d1Ci
Ci u
EW
W
d eti c l W
c
nes in putative fusion events, of which 18 were predicted to code for instances at levels of 1003 normal. Altogether, we found 16 differe
-frame events (Supplementary Table 6). We found three separate altered WNT pathway genes, confirming the importance of t
ses in which the first two exons of the NAV2 gene on chromosome pathway in CRC. Interestingly, many of these alterations were fou
are joined with the 39 coding portion of TCF7L1 on chromosome 2 in tumours that harbour APC cmutations,
eCi a eti csuggesting that multi
upplementary Fig. 5). TCF7L1 encodes TCF3, a member of the lesions affecting the WNT signalling pathway confer selective advanta
CF/LEF class of transcription factors that heterodimerize with
Genetic alterations in the PI3K and RAS–MAPK pathways a
clear b-catenin to enable b-catenin-mediated
common
addition
Z0 c en i c eC transcriptional
CsWqU C I ti cregul n l tI
c t t c in
e iCRC.
d In
u 1d
t t eti to
c l IGF2
W and
etc IIRS2 overexpression,
ion. Intriguingly, in all three cases, the predicted structure of the found mutually exclusive mutations in PIK3R1 and PIK3CA as well
c l l alacks
n the
l TCF37W
:W
E Wdeletions
q inqPTEN
c in 2%, 15%
W uand
i c4%
I of
i en
Cl
AV2–TCF7L1 fusionI protein
b-catenin-binding
non-hypermutated
tumou
main. This translocation is similar to another recurrent transloca- respectively. We found that 55% of non-hypermutated tumours ha
x a n Cd 1 et F
n d CWL8qL%
0 ce I C e
c dsl tl i u ae eti c l Wu
n identified in CRC, a fusion in which the amino terminus of alterations in KRAS, NRAS or BRAF, with a significant pattern
TI1A is joined to TCF4,) 1C
which
mutual
6 and
l l tiis cencoded
c
iby1sTCF7L2,
c au aChomologue
de C eti c ntI
ndtI exclusivity
ne en e(Supplementary
en u i l e CFig.I ad
e Supplementary Table
TCF7L1 that is deleted or mutated in 12% of non-hypermutated We also evaluated mutations in the erythroblastic leukemia vi
1 enR 21
sl cases
tc iofdi Ctranslocation
e d c involving
C R C
W homolog
W Z (ERBB)
W
zfamily
S cof receptors
: Z because of the tran
mours4. We also observed
oncogene
TC28 located on chromosome 22 (Supplementary Table 6). In all lational relevance of such mutations. Mutations or amplifications
1 enR sl x tI aC U%
0
3
6
6
#
, + β
#
6
6
#
0- .
#
6
6
#
#
, + β
3
DKK1-4
. ;
LRP5
#
#
8
FZD10
TGFBR1 TGFBR2
#
( #
#
#
#
#
IGF1R
ACVR2A ACVR1B
#
#
#
#
6
6
#
#
( #
1
IGF2
#
#
ERBB2
ERBB3
#
#
#
#
#
IRS2
FAM123B
#
AXIN2
#
#
APC
#
#
#
#
#
FBXW7
#
#
#
#
#
!
ARID1A
#
#
#
BRAF
#
#
#
#
0
#
#
PIK3CA
0
!
KRAS
#
#
NRAS
#
PIK3R1
*
#
#
PTEN
#
#
1
TCF7L2
#
#
#
CTNNB1
#
#
SMAD3
SMAD2
6
5
&
!8
8 &
6
#
#
*
#
CTNNB1
#
SOX9
#
#
#
0
6
MYC
TCF7
+
"
7!
"
!
0 9
"
"
"
2
!
6!
0
-
!
!
&
'
/
"
5
"
#
0
)
0
6
#
#
$ #%
0
8
8
8
0
4
4
8
0 9
!
5
8
0 9
"
8
"
5
!8
8
)
:
!
!
!
WNT
TGF-
RTK/RAS
PI3K
TP53
gure 4 | Diversity and frequency
SMAD4). Alterat
f a 3 dof genetic
u a ni changes
r a t leading
f r i e to r n
r up o or downregulation
r ne a f o nof
ergene
e expression
r oo r c (IGF2,
n s iFZD10,
regulation of signallingrpathways
Red denotes activate
h onina CRC.
n er aNon-hypermutated
t f r a Fc(nHM;
a frequencies
c a r n are
e expressed
oo has a percentage
r en ofnall
u cases.
n
5 165) and hypermutated r(HM;
shows for each sam
rn 5o30)
f samples
r en with
r ncomplete
u n
rdata
h were
ennel c genes
r o and
es blue
ea denotes
linactivated
co
nogenes.
nerBottom
r panel
r
e ndefined
u by
c somatic
n s i mutations,
a
r n f a if at
on least
a hone gene
c n inael
n|
o am in this figure is alter
alysed separately. Alterations are
homozygous
each off the ao
fivec pathways
described
2. p2h
etions, high-level focal amplifications,
and, in some cases, by significant
i aCl Wdden l u ae eti c l i c i ei
aCtc en l u eau i Cc1 9ten
J U LCY een
2 0 1 2 l| VuO L
©2012 Macmillan Publishers Limited. All rights reserved
etu 0 au i Cl ( i d( Ci u
u ae eti c l u c eti c
c
c tI c ei u dtI c c e d l ti c l
i ( 0 n C tl
s
R dd l e dtl n
ui
C i C en I tc tc I i en l u ae eti c l x i I dl e tc
I e o 1tc I Wu ae eti c 1Ci ( t l
l d et(
1ten dt d ddW ddi R tc I en i ( CI Ci R en i en
di c x tI aC 7%
0 e o 1tc I u ae eti c l tc en
I c x CI F
tc pd CWqóóU%
0 n l u dd
I Ci R l l di R dsW ae l
i c u ae eti c tc
ci u
I Ci R en
d tc
i en
i di C e d
WL8qZ%
0 n tCl eWi C
( ce I
dd c
ei
i u tc I
i di c u i l ei e c i
c i Cu d
u t Ci l i 1t
aCtc en
en e C l adel Ci u entl u ae eti c
t C c e I c Wl a n l
l
i c Ci ac i di c dI Ci R en en e ddi R l c ) 1 c l ti c i F
i I dl e tc Wqóó8%0 tc I d
u ae c e
9atCtc I l i u
ddc au
WeCtI I Cl
Cx
Ci c
ddl u s 1 Cl tl e tc en eau i CWni R ( C
Ló
4 8 7 | N AT U R E | 3
c eCi a eti c
en s C 1C l c e
i u 1i l
s
l u dd 1Ci 1i Ceti c tc en R ni d 1i 1ad eti c WR nt n tl u tc ds
ddl en e n (
u ae eti c i ddi R
s
9atC
u ae eti c l tc i en I c l 0 ntl 1Ci
di c d ) 1 c l ti c
u ae eti c l tc I c l l a n l
u dtI c c e eau i C en e
Z
c
W
c tc (
l d eti c
E c
l l i i c etc a l R ten c R
: ZW ( c ea dds I c C etc I
1Ci ) tu d etl l a l
c
u e l e l tp ei
tl e c e
i CI c l 0
Downloaded from www.sciencemag.org on July 31, 2013
Fig. 2. Genetic alterations and the progression of colorectal cancer. nents of these pathways can be altered in any individual tumor. Patient age indicates
f a I d r n on a n er r n c ae a er e eoea n o r a n aef
i a v er u r n er o
The major signaling pathways that drive tumorigenesis are shown at the transi- the time intervals during which the driver genes are usually mutated. Note that
ctionsnbetween
s i ,each
h tumor
c aestage.
f One ofael
e genes
o n that
r encodem
2. pPhthis model may not apply to all tumor types. TGF-b, transforming growth factorñ b.
several driver
compo-
Ct( C u ae eti c l C eni l mutations,
en e ias cwell
Cas lthe dnumerous
et( epigenetic
I Ci R enthose mutations
( c ethatI truncate
l Wtethenencoded
l protein
c
and other genetic abnormalities. When a large
part of a chromosome is duplicated or deleted, it
is difficult to identify the specific “target” gene(s)
on the chromosome whose gain or loss confers a
growth advantage to the tumor cell. Target genes
are more easily identified in the case of chromosome translocations, homozygous deletions,
and gene amplifications. Translocations generally
fuse two genes to create an oncogene (such as
BCR-ABL in chronic myelogenous leukemia) but,
in a small number of cases, can inactivate a tumor
suppressor gene by truncating it or separating it
from its promoter. Homozygous deletions often
involve just one or a few genes, and the target is
always a tumor suppressor gene. Amplifications
contain an oncogene whose protein product is
abnormally active simply because the tumor
cell contains 10 to 100 copies of the gene per
cell, compared with the two copies present in
normal cells.
Most solid tumors have dozens of translocations; however, as with point mutations, the
majority of translocations appear to be passengers rather than drivers. The breakpoints of the
translocations are often in “gene deserts” devoid
of known genes, and many of the translocations
and homozygous deletions are adjacent to fragile sites that are prone to breakage. Cancer cells
can, perhaps, survive such chromosome breaks
more easily than normal cells because they contain mutations that incapacitate genes like TP53,
which would normally respond to DNA damage
by triggering cell death. Studies to date indicate
that there are roughly 10 times fewer genes affected by chromosomal changes than by point
mutations. Figure 3 shows the types and distribution of genetic alterations that affect proteincoding genes in five representative tumor types.
Protein-coding genes account for only ~1.5% of
the total genome, and the number of alterations
in noncoding regions is proportionately higher
than the number affecting coding regions. The
vast majority of the alterations in noncoding regions are presumably passengers. These noncoding
en x i pt
Cl C l adel tc
C tc i u
u ae eti c R l C ( tl
tc
i u 1 etet(
( ce I
le u
aCen Cu i C W : Z u ae eti c l
t I ci l
en
C
i C
i R c en 1i e c et dei C a
1548
iR
di i
Z8
sl en C
e l ex
ce
C ci e
ddl 0
C
a
de C eti c l
m
a
to
llo
bl
as
as
to
ei
:Z
( ce I W c
iu
C tc i u l
CC eti c l 0 ntl 1Ci
c i aI n etu
i di C e d c
m
G
lio
bl
tic
1tI c et
c s l su 1ei u
i ad
aeni Cl
s R td zes1 l e u
ed
u
c
Cl WR n e I t( l al
c
en e u ae eti c l
M
au ad eti c i i en I c et
q8 i C u i C s
ca
nc
er
c i u ei al 1i ds1l
ta
lc
i C W
re
c
l u c eti c
er
r e lo
r
Co
a r r
e Cu tc
i c teti c z 1 c
nc
( i C 0
n
et( eti c W c
tc e l etc l W di c l n C i Ctc I eni l
an
ce
r
e
i WL8qZ%
WR ni l
ddl C C 1d
tl 1d s
i ( C en
( ce I
di l l W
tc e l etc 0 n s
c u c s u ae e
l 1 t dds tc i dtetl z i re
a
u i al
ddl 0 n
c e l ea s x Cu ad c
ca
u ae eti c l tc en
e Cu tc tl et
C
tddti c l i nc
en
i c e tc tc I
Pa
9a c et t
WL8q8%
Wentl l dtI ne tc C l
s en
ll u s e o
11ds l C c tc I u eni l ei
11 Cl 0 i CC e l C c tc I 1i dt t l
C
en s /8D x
F
Ci c WL8qq%
0
C enC
u tc l C c tc I u eni l 2 n c i c ztc ( l t( 29 MARCH 2013 VOL 339 SCIENCE www.sciencemag.org
%
Wen tc ( l t( l tI u i t i l i 1s c en
i di c i l i 1s0
di
ade
i ad
CREDIT: FIG. 2, E. COOK
dd tCen c
the gene, as well as protein-truncating mutations in
the C-terminal 1200 amino acids, are passenger
gene mutations.
Numerous statistical methods to identify driver
genes have been described. Some are based on
the frequency of mutations in an individual gene
compared with the mutation frequency of other
genes in the same or related tumors after correction for sequence context and gene size (22, 23).
Other methods are based on the predicted effects
of mutation on the encoded protein, as inferred
from biophysical studies (24–26). All of these
methods are useful for prioritizing genes that are most likely
to promote a selective growth advantage when mutated. When
Translocations
80.0
the number of mutations in a gene
Deletions
is very high, as with TP53 or
Amplifications
70.0
Indels
KRAS, any reasonable statistic
60.0
SBS
will indicate that the gene is extremely likely to be a driver gene.
50.0
These highly mutated genes have
40.0
been termed “mountains” (1). Un30.0
fortunately, however, genes with
more than one, but still relatively
20.0
few mutations (so called “hills”)
10.0
numerically dominate cancer genome landscapes (1). In these
cases, methods based on mutation frequency and context alone
cannot reliably indicate which
genes are drivers, because the
background rates of mutation
Fig. 3. Total alterations affecting protein-coding genes in vary so much among different paselected tumors. Average number and types of genomic altera- tients and regions of the genome.
tions per tumor, including single-base substitutions (SBS), small Recent studies of normal cells
insertions and deletions (indels), amplifications, and homozygous have indicated that the rate of
deletions, as determined by genome-wide sequencing studies. For mutation varies by more than
colorectal, breast, and pancreatic ductal cancer, and medulloblastomas, 100-fold within the genome (27).
translocations are also included. The published data on which this In tumor cells, this variation can
be higher and may affect whole
figure is based are provided in table S1D.
Though it is easy to define a “driver gene mutation” in physiologic terms (as one conferring a
selective growth advantage), it is more difficult
to identify which somatic mutations are drivers
and which are passengers. Moreover, it is important to point out that there is a fundamental
difference between a driver gene and a driver
gene mutation. A driver gene is one that contains driver gene mutations. But driver genes
may also contain passenger gene mutations. For
example, APC is a large driver gene, but only
Number of alterations per tumor
s
c
within its N-terminal 1600 amino acids are driver
changes found in cancers, will be discussed later.
geneED
mutations.
throughout
n Ct( C uDrivers
ae Versus
eti c Passenger
tc C Mutations
l l i ae 80
en Missense
etu mutations
i u 1Ctl
ea
st
eR
en e
Br
l etu e
Introduction be guaiac based (gFOBT) or immunochemical (iFOBT), which is much more accurate and specific (Hol et al., 2010). FOBT screening has shown a decrease in mortality about 15-­‐20% (Faivre et al., 2004). Many other non-­‐invasive screening markers have been proposed, mainly through the stools or blood analysis. For example, detection of mutations in DNA obtained from stools has been suggested as screening method since early 90’s. Genetic tests that included TP53, KRAS, APC and BAT-­‐26 demonstrated sensitivities for cancer and adenomas between 52–91% and specificities of 93–100% (Ahlquist et al., 2000; Imperiale et al., 2004; Syngal et al., 2006; Tagore et al., 2003). Also enzymatic activity can be used as a diagnostic marker, for example M2-­‐PK has been identified as a key enzyme for colorectal cancer cells. Thereby, fecal M2-­‐PK tests are able to detect specific alterations in intestinal cells and it can detect carcinomas and even bleeding and non-­‐bleeding polyps with high sensitivity and specificity (Newton et al., 2012; Pox, 2011; Tonus et al., 2012). Abnormal DNA methylation of gene promoters has also been studied as a diagnostic marker in both blood and stools. Different markers such as SEP9, MLH1, EN1, VIM and WNT2 have been extensively studied, and the detection values range from 30 to 60% in most cases and the specificity between 80 to 100% (Azuara et al., 2010; Bosch et al., 2012; Herbst & Kolligs, 2012; Mayor et al., 2009). However, it has been shown that different cohorts exhibit important differences in specificities and sensitivities, maybe due to the different approaches used to analyze the samples (Herbst & Kolligs, 2012). Independently of the marker, any positive result requires additional examinations, normally colonoscopy. In this case it is possible to perform a biopsy of the putative tumor to obtain an accurate diagnosis. If the tumor is confirmed, next step would be the classification into TNM stage, which describes the spreading of primary tumor, also the lymph nodes affection and if there are metastasis and how many they are (Cunningham et al., 2010). In general, colonoscopic surveillance and polypectomy have had a tremendous impact on reducing CRC-­‐related deaths: colonoscopies have resulted in an overall reduction of CRC-­‐related deaths by 29% and deaths from distal CRC by 47% (Limketkai et al., 2013). Curiously, colonoscopies have resulted in no observed reduction in deaths caused by proximal CRC (Baxter & Goldwasser, 2009; Singh et al., 2010). 31 Introduction Treatments Surgery to remove the cancer and nearby lymph nodes is the most common treatment for early stage tumors (TNM stage I and II), whereas in more advanced cases, chemotherapy is used after or before surgery (TNM stage III and IV). Nowadays, in the era of personalized medicine, before the election of the treatment it is feasible to investigate some allele variation in some key genes in order to predict how the patient will respond to chemotherapeutic agents. The agents more frequently employed in the treatment of colorectal cancer are: ·5-­‐Fluorouracil (5-­‐FU): an antimetabolite, it is the mainstay of all current standard colorectal cancer chemotherapy therapies. It is known to cause side effects such as myelosuppression, mucositis, diarrhea, hand and foot syndrome and cardiac toxicity. Up to 30% of patients will experience serious side effects, with 0.5% of these being fatal. Some deaths by toxicity could be avoided by checking DDP (Dihydropyrimidine dehydrogenase) gene mutations. DDP is the enzyme responsible for metabolizing the 5-­‐FU in the liver (Meinsma et al., 1995), at least 17 known germ line mutations could affect the efficiency of the enzyme and produce the accumulation of the agent until systemic fail. ·Irinotecan is a topoisomerase inhibitor. It is known to be efficacious in the treatment of metastatic CRC when used in combination with 5-­‐FU. Side effects of Irinotecan are diarrhea and myelosuppression. Irinotecan is inactivated by UGTA1 enzyme, for which more than 50 less active variants or inactive ones have been described, the most studied is UGTA1*28 that has been investigated as a cause for increased irinotecan toxicity. Taking into account pharmacogenic trials (Ando et al., 2000; Innocenti et al., 2004; Marcuello et al., 2004; Rouits et al., 2004), the US Food and Drug Administration (FDA) has advised that patients homozygous for the UGTA1*28 allele should receive a lower starting dose of irinotecan. ·Oxaliplatin inhibits DNA synthesis by forming both inter-­‐ and intra-­‐ strand cross-­‐links in DNA. The main side effects are diarrhea, myelosuppression, hypokalemia, fatigue and neurotoxicity. Normally these drugs are combinated for the treatment of stage III and IV colorectal cancer. These combinations are 5-­‐fluorouracil, leucovorin and 32 Introduction oxaliplatin (FOLFOX); 5-­‐fluorouracil, leucovorin and irinotecan (FOLFIRI); and capecitabine (the oral form of 5-­‐fluorouracil) plus oxaliplatin (XELOX). Radiation in combination with chemotherapy is used in rectal cancer (Cunningham et al., 2010) and not routinely in colon cancer due to the sensitivity of the bowels to radiation (Shaib et al., 2013). Three targeted monoclonal antibody therapies approved by the FDA to treat patients with metastatic colorectal cancer are bevacizumab (inhibiting angiogenesis by inhibiting vascular endothelial growth factor A), cetuximab (epidermal growth factor receptor (EGFR) inhibitor) which is ineffective when KRAS is mutated, and panitumumab (EGFR inhibitor) (Newton et al., 2012). Besides, many others are in clinical trials or applied to other cancers (Scott et al., 2012; Vanneman & Dranoff, 2012). It is worth to note that aspirin treatment has multiple advantages and it has been demonstrated it antitumoral role in almost all stages of the cancer. In polyps, aspirin inhibits WNT signaling and reduces the size and frequency of sporadic colonic polyps and polyps in patients with familial adenomatous polyposis. Moreover, aspirin hampers epithelial to mesenchymal transition (EMT) in cancer by inhibiting COX-­‐2 cyclooxygenase (Oshima et al., 1996). In EMT and in metastasis, aspirin reduces lymphatic dilation and lymph node metastases in mouse models of tumor lymphatic metastases. 10% reduction in overall cancer incidence could substantially broaden the indications for prophylactic daily treatment with low-­‐dose aspirin (80mg) (Chia et al., 2012; Thun et al., 2012). Not only drug treatment is important; recent papers highlight the importance of gut microbiota as enhancer of the immune anticancer response, these papers demonstrate the importance of the bacterial environment in colorectal cancer (Iida et al., 2013; Viaud et al., 2013). 33 c eCi a eti c
c ao Csi et
1 o I
ddl WI c et tc i Cu eti c tl c i
Ci ac
tc en
c i e u Ci tI ne i C ntl ei c l W i Cu tc I
WR nt n tl etI neds
sc u t l eCa eaC
oc i R c l nCi u etc x aI CWL88/%x tI aC ó%
0
c
ao Csi et i CI c tl u l tc da tc I 1d c el W c tu dl
la r e
tu 1 Ce c
ei
c psu et u i t t eti c l e i en en
c
ac I t,
c 1Ci e tc d ( dl WR nt n
teti c d d s C i n Cte d tc i Cu eti c ei en
l l c et d i C C I ad etc I I c
n Cte d Ci u i c
) 1C l l ti c 0
I c C eti c ei en
i
c R i ad
n c l a n u i t t eti c l
c ) eWen s
u i t t eti c l c
c Ctc I n c I l tc I c
POLYGAMOUS DNA
L8qq, R F
i l c WL8q8%
0 fl
nCi u etc tl
C C CC
) 1C l l ti c x
ei
C
1tI c et
c toi F ntd et C W
Trojer and his colleagues
blocks
an enzyme
f a T d use
el ac number
n er eof epin
n aef
n active
aelin an
n epigenetic
r naf nf patha m r cards
n nand
s e stored
af indefinitely
o c r nby hospitals and
genetic techniques.
ChIP-seq
way but
not its nearest
l e
n er ais aflab
o nstandard
r n c ae
m
l n ineighbours,
o n er r he says.
nerIt is l ehealth-care
n ersystems.
h
cScientists
n
aelat Queen Mary,
in which antibodies
“the key” to the tech- a selectivity that has been difficult to come by in University of London are exploring how DNA
aFm2.are
p2h
nique, he says. But the company has also made the development of biotherapeutic drugs.
methylation patterns change between newborns
histone mass spectrometry a priority, because
Copeland believes that epigenetics drugs and in cells from the same children when they
c qóEL i c C
tc I ei c tc
1tI c et l l 5en C c n i ti di I s R nt n
it allows the scientists to query the chromatin fit into a trend of defining a cancer not by its are three years old. Differences in epigenetic
changes without using antibodies and to query anatomical location but by its molecular pro- marks could be clues to health.
ea t l en at once.
al The
d tccome C file,
eti which
c l includes
eR epigenetic
c I c lsignatures.
c enLike
tC 1Ci If the
a elsequencing
WR nt n companies
Ctc I l are betting
a number ofl modifications
pany set up an in-house high-throughput facil- many companies in this field, he and his col- right, then genome sequencing could become
1n c icompounds.
es1 tc ei
tc Ileagues
; 0 Cn
u s 1Cavailable
CC databases,
tc teticommonplace
c R i ad for many
en patients,
i c perhaps even
ity to screenen
for potential
mine1l
the publicly
Although other companies tend to out- noting that many genetic alterations in epige- part of an annual physical examination. An
u the company
c tl wants
Cdi toRintetc W Cinetic
u pathways
t c arecfound
t( Clintes
tc L88/2
c etread-out,
l n l updated
dR slat regular intersource these Ci
tasks,
human
cancers. 5 1tI
epigenetic
grate findings about chromatin biology into
Kouzarides believes that many cancer cells vals, might be an important companion file to
c anddin-house
en Rsuite
tCof tools
c Rwill
i cbe very
C ad
entc I l toenepigenetic
e c 6edrugs )that
1dgenome
tc
s I c et l 40
vulnerable
drug discovery with
sequence.
that includes mass spectrometry and biophys- because they rely on only one or two epigenetic
But this type of progress depends on deeper
n C explains.
l en u i l e
1epathways,
c i R whereas
sl normal
tl 25 1tI
c etonlsevtl enunderstanding
l ea s i ofuepigenetic
tei et dmechanisms
ds
ics analyses, Dionne
cells draw
and
Another Cambridge-based epigenetics com- eral pathways for their functions. At the same technology that has yet to evolve, Kouzarides
pany, Epizyme, focuses on a family of proteins time, he believes that epigenetics researchers says. Because epigenetic events change concalled histone methyltransferases. These epige- and technology developers will still want to stantly in the cell, “whatever you see in one
ZEact on histones, by catalysing the develop and refine experimental methods, for moment will change in the next”, he says. ■
netic modifiers
transfer of methyl groups onto specific positions example to explore the three-dimensional strucin the protein. The company has partnerships ture of epigenomic events, to see how chromatin Vivien Marx is technology editor at Nature and
with the pharmaceutical companies Glaxo- is changing throughout the genome. “It’s very Nature Methods.
c eCi a eti c
c Mi C u ti et dds n Cte d n c I l tc I c ac eti c en e
s n c I l tc
n
l 9a c
1tI c i u
ea dds te tl
tl
; x al l i
sc u t
e
c
c ( tCi c u c e c
t C c et eti c 1Ci
es1
i c eCt ae l ei en dtc
c
L8q8,
l
ac
d enCi aI ni ae dt W
CI i l 1C tl ds i c eCi dd
l l 0 ntl eC c l teti c tl t C c e
I
) 1d tc
WqóóU%
0
i l c i e 1 Cl tl e ac n c I
s en
i ( Cen
cci e
c etl l a
1 c tc I i c en
dd
ddad C i u u teu c e x i e F
tdW
toWL88U%
0
n i o n er mn
n
c nau c l R C C ei
u ensd I Ci a1 tl
u ensd eti c l en
ei en
R nt n tl oc i R c l en t en
ti n u t d 1Ci
sei l tc x % i Cu tc I
l i en I c i u
t(
en c u
: zu ensd sei l tc x: u %
W
x tI aC q8%x
u ensd eti c R l en tCl e 1tI c et u i t t eti c
i e notl l c tc tet dds C
ll Rn C s
C) WL8qL%
0
tl i ( C
tc qóE7
s
C0
i 1t sei l tc x i e notl l qóE7%
0
u ensd eti c tl l su u eCt tc
i en l eC c l
en
sei l tc
c
i
aCl tc
R tentc en
1
i c e ) e0 e tl R i Cen ei c i e
en e en C tl
u tc i Ctes i u ensd eti c en e i
tc
f a p. d
l n
c e n er e n l n i o aef c r n
eaa c er r i ne r i n
r gi l h
cn
ael n
eeC c a r c o e
r n m2. . Th
R n C te R l i ac
tc
et( eti c x tl e C
ci cz 1
ac
c Ct n
WL8qZ%
0
CC 1C l c e
u i c I en u /8A78D
tc a d i et
Cl ei i 0 1
e
tc
eW
tl l c CWL8qZ%
0 nC
i c l C(
C I di
dds
C L7 u tddti c 1 l tc nau c
c l tes tl en l e etl et dds ) 1 e
c
u d
u ensd eti c tc
tc a d i et l
z c l C I ti c l 0 n l
e en 1Ci u i e C C I ti c l i ni al o 1tc I
c etc c aCi c l
nCi u i l i u
C I c C dds u ensd e 0
x 1 t%x c e 9a C F tC WqóóZ%
0 1 t e c ei
x u ten F
ce
i c e ) eWc te tl
ac
1 i en
tc u u u dt c I c i u l 0 n C
1 l tc a d i et l C di
1
l 1 t dds
l 1te entl C
i c e ) e tl l etdd 1i i Cds ac
I ci u Wc
en
tc I c l en e l
ci cz 1
aCl
l l en c q8D i Ct n C I ti c l R n C
C e Cu
ac u ensd e
1 tl d c l
c i e c di
e
( di 1u c e d C I ad ei Cs I c l
c psu l W
u ensdeC c l C l q
Z:
c eCi a eti c
x
q%
W
Z
u tc e c c
x t
c
x
c
i Z
C C l 1i c l t d i C en
u ensd eti c x tI aC q8% c
WqóóL, o c i
C
l l c et d i C
u ensd eti c n c I l ral e i
tr
i c eC CsWen C
l a Cl c
WL8qZ%
0 c en
( di 1u c e0 u u
et(
( di 1u c e
aC tc di
dtp
C I ti c l
C l eCtotc I n c I l tc en I Cu dtc tc
t e ds e C Cetdtp eti c Wen 1 e Cc dI c i u
u ensd eti c c C u tc l tc en e l e e i C u adet1d Ci ac l i dd t( tl ti c l 0 aCtc I entl etu Wen
1 l l t(
c
Wqóóó%
0
dd t C c et eti c
Cds l e I l i 1i l teti c
u ensd eti c 0
u e Cc d I c i u
u ensd eti c 1 ee Cc l C
u ensdeC c l C l l
Z
c
Z
) 1 Ct c
l I C a dW
l e dtl n
aCtc I en
s en
( di 1u c e i en
REVIEWS
d l ei sl ex tI aC qq%x a F n c I WL8q8%
0
Female
Male
Fertilization
Zygote
ICM
De novo methylation
DNA methylation
a
2 cell
4 cell
8 cell
Morula
TE
Blastocyst
TE
ICM
Developmental stage
l
bf a pphd i r l e
r
ne n naec
ne
deni aI n u i l e i ExE
en
l n i o n er r
aoi n
al h
c n ael
f |
I ci u
Epi
l 9a c tc I i c tCuVEl en e l i u
E3.5
E6.5
tl
u
r nh
a
a ne r r a
u ensd e WR ni d I c i u
PGC
founders
I c i u t C I ti c l
E7.5
E8.5
u ensd
eti c 0 n lInitiation
I c i u t Specification
C I ti c l tc da
Blastocyst
BLIMP1
de CWqóóU%
Wl i
PRDM14
15meC
Ct c eCt n e Ci
e
u oec l
r m2. p. h
tl ad1nte
C C l tl e c e ei l a n R ( i E9.5
E10.5
E11.5
tu 1Ctc
etc I i c eCi dI c lGenital
x dridge
oF
Migration
C eCi eC c l 1i l i c l x C( tc
nCi u etc x i aI t C
oo
WL8qq% c
Wqóó7%
0 u 1Ctc e
I c l
c eCt
c
C eni l
Figure 2 | Dynamics of DNA methylation during development. a | Active demethylation in the zygotic paternal genome.
a sperm
active
demethylation
I Shortly
c l after
) 1C
l l fertilizes
tc an egg,1 the
C paternal
c ezi zigenome
CtI tcrapidly
zl 1 undergoes
t t u genome-wide
c c CW
Rn
e DNA
d| Molecular
l eiCellen
Nature
Reviews
Biology
and remains demethylated following multiple rounds of cell division. During this time, the maternal genome experiences
gradual,
methylation
by the
methyltransferases
DNMT3A
l td
c tcpassive
I i demethylation.
en c i De
c znovo
) 1C
l l tc I patterns
dd d are established
s
uDNA
ensd
eti c 0
teti c d
and DNMT3B during the development of the blastocyst. b | Active demethylation in primordial germ cells (PGCs). After
implantation
ectoderm i
(ExE)
(VE)
u ensd ofetithecblastocyst
C l tl at
e embryonic
c e C Iday
ti c7.5l (E7.5),Cthe extraembryonic
dl i C 9atC
C and
di visceral
c I ze endoderm
Cu
produce signals that specify a subset of epiblast cells (Epi) to become PGCs. This process requires two key transcription
PR
1 (PRDM1))
and PDRM14,
are expressed
thisc
ufactors,
ensdBLIMP1
eti c (also
l tdknown
c tcasI W
endomain
etl zinc
l alfinger
e tc protein
enCi
aI ni aeen
( which
di 1u
c e c during
dt l 1
stage of development. Following specification, PGC founder cells divide in the presence of the DNA methyltransferase
and migrate towards the genital ridge. During this migration and on arrival at the genital ridge, 5-methylcytosine
i DNMT1
c i CI
c tl u x u ten F
tl l c CWL8qZ, a F n c I WL8q8%
0
(5meC) is erased through an active mechanism. ICM, inner cell mass; TE, trophectoderm.
regions are resistant to this wave of DNA demethylation,
but one possibility is that methylation of these regions
Z/may be required to ensure transcriptional repression
and chromosomal stability. Additionally, the maternal
genome remains methylated during this time even though
it is exposed to the same cytoplasmic factors. Insight into
of transposable elements and repeat sequences results in
their reactivation and, if so, what the significance of their
reactivation is remains to be determined.
Genome-wide DNA demethylation of primordial germ
cells. After fertilization, the one-cell zygote undergoes
Introduction Two key papers in the 1970s proposed DNA methylation as a silencing epigenetic mark (Holliday & Pugh, 1975; Riggs, 1975). Nowadays this function has been confirmed in several papers and now this mechanism is widely accepted (Garinis et al., 2002). Today, thanks to the possibility of genome-­‐scale mapping of methylation, it is feasible to evaluate DNA methylation in different genomic compartments, such as gene promoters with or without CpG islands, gene bodies, regulatory elements and repetitive sequences. The silencing effect of DNA methylation into gene promoters with a CpG island has been extensively studied in cancer and other diseases and it is widely accepted. However, the scenario is not that clear in the genes that do not have CpG island in their promoters. Genomic analyses have identified low CpG promoters that are both transcriptionally active and DNA methylated. It has been shown that the CpG methylation of the CRE motive (TGACGTCA) enhances the DNA binding of the C/EBPα transcription factor, a critical protein for activation of differentiation in various cell types (Rishi et al., 2010). Furthermore, C/EBPβ has also been described to bind methylated DNA (Mann et al., 2013). CpG dinucleotides density in the promoter is a key factor to understand the role of DNA methylation in regulating gene expression. In this field, the work of Dr. Dirk Schübeler has shed some light on how some CG poor regulatory regions are protected from DNA methylation. By using whole genome bisulphite sequencing in mESC and differentiated neurons, he observed that low-­‐methylated regions (LMRs) are occupied by DNA-­‐binding factors, which attachment protect DNA from being methylated (Schübeler, 2012). He has also observed that cell-­‐type-­‐specific LMRs are occupied by cell-­‐type-­‐specific transcription factors, meaning that transcription factors are shaping the methylome (Bell et al., 2011; Stadler et al., 2011). In the same field, Dr. Daniel G. Tenen recently described a new mechanism by which DNA is protected from DNA methylation. He observed a new non-­‐coding transcript (ncRNA) transcribed some Kbs upstream the promoter of CEBPA gene that modulated the expression of a gene by inhibiting the methylation (Figure 12.A). The extra-­‐coding CEBPA ncRNA (ecCEBPA) binds at the same time to the DNA sequence (by sequence homology) and to the DNMT1, preventing by this way the DNA methylation of the DNA (Figure 12.B). In the same study the authors demonstrate that DNMT1, 37 , 8 5 7/
@
,(8 5?
)'
; :
0
54- 40;
?
54-
?
c eCi a eti c90
; 78
8<0 5 990/ =4 3
, 8 5 )'
;0 : 54- ?
78
=4 3
;0 :
enCi aI n tel
87/
7058. 4
!'
& +
-087/
2070 58. 4
e dset
i u tc W tc l
control.
In conclusion, we have generated the fir
cross-referencing DiRs to DNA methylation
These data demonstrate that RNA–DNMT
DNA m
R ten ntIspread
n C tcand
tes eimight
i d modulate
en c genomic
ei
u ensd eti 5c 5))*d ( d.l 8<0 C0/ tc ( Cl ds i CC d e
al ti
WL8qZ%
0
R ten
Discussion
2070 58. 4
d ( dl x t
This study explores the role of a new class of RN
I c C the
dWu iCEBPA
l e i en gene
I c as a model, we
RNAs.c Using
ecCEBPA
a A
i t is
l accompanied
C 1 1i i C C Iby
ti c lthe production
scription
CEBPA 5)&
)
species,
ecCEBPA.
studied, D
!
x deni
aI n te tl cIn
i e every
ac al a instance
d ei
CEBPA
are inversely
correlated
with ecCEBPA levels,
tc
1
tl d c l en C %
& 0 3?5 487 50<0 5
& 0 3?5 487 50<0 5
7 50<0 5
methylation is determined by the absence or
" 39 48 757 /
! " *
′ +)
?
c i Cu dds
) e c l t( ds
dB + 77;?. 49 487 55? 47 . 47<0 207864.
(Fig. 2).
7
;4 0;
ensd e
c that
i cecCEBPA
e tc tc I
5 582
60 3?5 7;1 -0 ;0 / 449. . 0;;4 - 50
ecCEBPA
b
c
Weudemonstrate
associates w
ecCEBPA
CEBPA 5)&
8ad
75et1d
7/
C 1 etet(
d u c el 0 ecCEBPA and
lish " a39u4functional
link between
!' & +
SP29
through RNA–DNMT1
36!'
u ensd eti c
association.
etc I
50<0 5
WeI show
of adopting
c ) i cthat
l tl RNAs
u ri C capable
al i + 7;. 49 487 55? . 4<0 207864. ;4 0
exhibit the potential to associate with DNM
u ae eti c l
al
s en
60 3?5 7;1 0 ;0 0 ; 4. 4<0
basis of this preferential interaction is recogni
USP29
tc eti 3).
c
i :u → 0
& +
0. 4fi. 9 0 ;
structureu (Fig.
d
e
/ 449 !'
& - 4/ )& 9
u′ tl+)uwee demonstrate
n l d
ei that this type
Importantly,
ecCEBPA
" 39 4
! " *n l
CEBPA
8 757 /
5)&
CEBPA
5)&
ecCEBPA
ciationtlis lnot
restricted
toc the
z al
tc I u ae eti
l tc CEBPA gene l
36!'
USP29
identified
associated
with DNM
en IRNA
Cu dtcspecies
c
c Cz
ship to DNA methylation and gene express
f a p2 d e o e and -unbound
p t f na n er h y
de alignment of DNMT1-bound
al tc I u) ae) eti c l tc l i u et
)
)
)
)
na r a c n a er) r) a o n er ne n l
h r defined
mnec m ) ) a large
set of expressed unmethylate
hylation and gene expression.
a, Two-way
p
r
ne
l Venn
l n ion
r el
d
d
l
x
t
i
ae
Wqóó8%
0 genes that co
f
g
/ 449
"3 9 48 7 57 set
/
mentary
of silent methylated
a overlapping
er h ennel m
r r en
ne
n
u
MT1-specific peaks
withp transcribed
l l n *i1
a 69<er
h763= / c. n ael
f 763
e= / .
<6o- n
2 763= / . r el
6 - 2
69<6eti
- 2 763
ensd
c = /tc. en of
I cthe iDiR.
s
CEBPA
expression
RNA-seq
libraries. * <6 - 2- following
HL-60transcripts.
total and poly(A)
- / 449
/ 449
- / 449
- / 449
m2. pPh -depleted
EBPA (small white
Our
tl findings
1i l tet( dssuggest
i CC d ea model
R ten of site-specific
nting
genesregion
within DNMT1-unbound, -bound and all
the coding
stratified
expression
transcripts by
by DNA methylation
act
asd ac shield,
) 1C l l ti c x i and
c lW
qóóó%
W ( levels.
c t en All
I c ini which
s i c eRNAs
tc l
1 tl
c tl halting DNMT
scripts
in
cellular
Supplementary Data 2. c, Examples of genes from the C DNA genomic methylation at their site of
t scales. qRT–PCR
) e cvisualized
l t( ds u ensd
e the
Wentl
R i ad
c i e tc nt te di c I eti c 0 1 Ce Ci u entl W
9)
clusters. Peaks are
using
SSIRs
(site
Indeed, the loss of CEBPA locus methylation fo
eti c uof sDNMT1
n (
Cisequestration.
d tc l 1dt tc I 0 e n l
c u i c l eC e en e
d, Model
rt sequence reads)u48.ensd
of ecCEBPA
does not support azmodel of trivial
Figure
demonstrate
transcriptionally
genomic
nsssCEBPA
tc inactive
tc I2 | Losseihemimethylated
Cand
x gain-of-function
% c 1Ci
ustudies
i e tc
dalofti DNMT1
c i that
R ecCEBPA
obut
a1lsuggests
eC u )ai ccis-regulatory
l s
role o
maintains
CEBPA
expression
by
regulating
methylation
of
the CEBPA
MT1 cannot access transcriptionally active
locus.
a, Diagram
the position
of1 target
shRNA
model
wherein
contain
t etc
I di d indicating
1i dsu
Cl
al tc Isequences
tc u for
uu
dt c uRNAs
i d l sl
e u i C a locus-selectiv
he regulation
mic
regions. of uconstructs
(sh1–sh3); the fragment derived from ecCEBPA used for
the ‘anchor’,
mooring
the DNMT
gain-of-function overexpression
de Cc et( l 1d
t tc Iregions
W denianalysed
aI n entlfor1Ci
l l in
c forming
l ei
n part,
( en (distal
ac u ensd
e
(R1);
changes
DNA
methylation
ell line (up to a promoter; coding sequence (CDS) and 39 untranslated
locus,
a 0DNMT1-interacting
part, the ‘b
region and
(UTR)).
c te tlstatus
tc nt tewas
R nobserved
c u ensd (Supe x naod
WL8qq%
n C i C Wte l u l
ne
locustargetmethylation
shRNAs)
Asterisks indicate number of base pairs away from the
CEBPA TSS.
b, c, The serving to lure the DNMT1 in
forming
sequence
crease of CEBPA dto
results
loss-of-function
ds of
enecCEBPA
e
u ensd eti c in
tc CEBPA-expressing
I c
i t l RU937
tdd n cells.
( iEffect
ae i of
l
si c en
byuRNA
does resemble a comp
ecCEBPA-targeting shRNAs on CEBPA mRNA levelssequestration
(qRT–PCR, bars indicate
,
suggesting
that
d genes within both
groups
according
to
levels
of
i I c6tps.d. (b))
ac and
eti cmethylation
tc en l tdofc the
tc ICEBPA
i tc eCpromoter
I cribed
c t C(c).
1for
etet(
l 9a c l 0RNAs, for example,
DNA
other regulatory
g of the CEBPA Cmean
6,7,23
methylation
changes
are
shown
as
the
ratios
of
methylated
(M)
to
hylation.
We
defined
genes
as
‘expressed’
or
‘low
15,16,41
promoter
. To
. i5However,
c n c Cl C
l tea
e in all
e (clones
Ct analysed
d tl e per
c RNAs
l Ciconstruct
u 1Ci u(n
e 14).
Cl c
C unlike
o s ei competing e
unmethylated
(UM)
CpGs
each
score was
above
or
below
4,
respectively;
the log
EBPA
and2 methys.c.,
scrambled
d, e, The
studies
i c eCi
ddtc I I control.
c ) 1C
l ti cresults
tc of( ecCEBPA
di 1u c gain-of-function
eecCEBPA
c
dd acmodel
eti c xinalso
c c introduces
nti
W the requirem
within
the distal K562
d’
or ‘methylated’
ifcells,
theinmean
of lall
scores
which CEBPA
is CpG
methylated
and silenced.
d,
Effect
of
ecCEBPA
and
physical
co-compartmentalization
of the
PA TSS; Fig. 2a). L8qZ%
0 e n on
l This
c l approach
nimRNA
R c enlevels.
e allowed
cUR,
n cunrelated
Cl
ns1 qRT–PCR,
Cu ensd bars
eti c i ad C l ade tc
upregulation
CEBPA
region.
0%,
respectively
(Methods).
increase in DNA indicate mean 6 s.d. (n 5 4). e, Effect of ecCEBPA upregulation
and DNMT1.
Given the ability of DiRs to bind
on methylation
usters
DNMT1-bound,
rol (Fig.within
2c andDNMT1-unbound,
of the CEBPA locus (DNA methylation changes weretoassessed
as described
in
suggest
that DiRs
represent a novel class of
c (nHypomethylated
5 14 for distal promoterand
and nexpressed
5 6 for CDS and 39 UTR)). f, g, The results
d groups (Fig. 5b).
Z7
Takenlevels
together,
these data support the hyp
of transcription inhibition in U937 cells. f, ecCEBPA expression
after
CEBPA
was
suffipredominant intreatment
the DNMT1-bound
group
(cluswith actinomycin D (actD) and ML-60218cipates
in synchronized
in theand
establishment of genomic methy
gion of ecCEBPA
unsynchronized
cells.
qRT–PCR,
bars
indicate
mean
6
s.d.
g,
DNA
or
56.64%,
whereas
hypermethylated
and
low
or
essing ecCEBPA
acting with DNMT1 and pave the way for the s
92
92
92
9-
! 8$ 5 / 2< 4 376 37
% ! 8$ 9 % ! 8$ 9
92
92
92
9-
>
>
>
) / 4 3: / / ; 8 / 99376
4/ : / 49
70- 76 74
67 5 43= / . 718S
92
92
92
4<
9-
>
>
>
) / 4 3: / / ; 8 / 99376
4/ : / 49
70- 76 74
67 5 43= / . 7 18S
?
0
-
37
6
37
6
-
4)
&
0
3-
(7
+7
95
4
"
-
%
" %
*'
% ! 8$ 9 % ! 8$ 9
37
5 / 2< 4 376
! 8$
"
-
%
" %
*'
>
>
"
-
%
" %
*'
) / 4 3: / / ; 8 / 99376
ecCEBPA 4/ : / 49
705 7-
! 8$ 5 / 2< 4 376 37
% ! 8$ 9 % ! 8$ 9
! 8$ 5 / 2< 4 376 37
% ! 8$ 9 % ! 8$ 9
074.
>
074.
>
>
>
"
-
%
" %
*'
) / 4 3: / / ; 8 / 99376
ecCEBPA 4/ : / 49
705 7-
) / 4 3: / / ; 8 / 99376
4/ : / 49
70- 76 74 )
67 5 43= / . 7 18S
!
<
78
92
%
92
92
$
Introduction reduced activity of the enhancer in reporter assays (Schmidl et al., 2009). The emerging picture is that DNA methylation has different functions and varies depending on the genomic context. DNA demethylation, the sixth, seventh and eighth bases of the genome In the last decades science has been committed to elucidate the mechanisms of DNA demethylation. There were some DNA demethylases proposed but the results obtained could never be reproduced in other labs (Jin et al., 2008). It was in 2009 when the enzymatic activity of TET proteins (Ten-­‐eleven translocation) was described (Iyer et al., 2009; Tahiliani et al., 2009) and related to a new mechanism of demethylation. TET proteins (TET1, TET2 and TET3 in humans) are Fe2+-­‐ and 2-­‐oxoglutarate-­‐dependent dioxygenases that successively oxidize 5-­‐methylcytosine (5mC) to 5-­‐hydroxymethylcytosine (5hmC), subsequently to 5-­‐
formylcytosine (5fC) and 5-­‐carboxylcytosine (5caC) in DNA (Figure 13) (He et al., 2011; Ito et al., 2011; Iyer et al., 2009; Tahiliani et al., 2009). These oxidative reactions have been observed to be modulated by vitamin C (Chen et al., 2013). All three forms of oxidized methylcytosine are now known to be present in different amounts in numerous mammalian tissues (Globisch et al., 2010; Kriaucionis & Heintz, 2009; Szwagierczak et al., 2010). For example, 5hmC is found at different levels in mammalian cells: it is present at 1% of the total level of 5mC in bone marrow populations (percentage that is decreased in some types of leukemia) (Ko et al., 2010), about 5 to 10% of the level of 5mC in embryonic stem cells (Iyer et al., 2009) and as high as 40% of 5mC in Purkinje neurons (Kriaucionis & Heintz, 2009). Until now, neurons are the cells with the higher content of 5hmC (Lister et al., 2013). In the case of 5fC and 5caC, they are present in mammalian cells at much lower levels than 5hmC (0.03% and 0.01%, respectively, of the level of 5mC in mouse ES cells) (He et al., 2011; Ito et al., 2011). This could be explained by the enzymatic mechanisms responsible for their removal. There are two different mechanisms proposed to remove DNA methylation, the first one and more accepted is thymine DNA glycosylase (TDG) (Figure 13, upper part) (Pastor et al.,2013). It is known that 5fC and 5caC can be excised by TDG and their replacement with cytosine results in demethylation. 39 c eCi a eti c
i u 1d ) C i I c tp l :
2 u tl u e n l R ten ntI n C tc tes en c
u tl u e n l W aCen Cu i C
d ( dl i : c
c :
dll
1d eti c
tc
L ei q8z i d tc C l
ddl x i ndt F n c I L8qZ%
0 n l
1e Wtc ( i d( l en
x: nu % s
al l
c
u tc eti c i : nu
c psu l W: nu
I ds i l sd l l c adetu e ds C 1d
2
tc en
ic u
n c tl u
ei : zns Ci ) su ensdaC td
tl en c C u i (
s
qiC
s sei l tc x tI aC qZWdi R C 1 Ce%x ai
WL8qq%
0
# #$* #
& $+ ,! '
#
#- ," '
NH2
NH2
%&$( #
N
N
O
OH
N
O
R
R
O
NH2
%&$( #
N
N
O
R
R
%
NH2 O
OH
N
N
O
N
O
NH2
%&$( #
N
N
R
$&
OH
O
HN
$&
#
N
O
R
"
f a pP d
l n i o n er
c aen r t f
eal i o i ne r
N
O
i C i ( CW: nu
N
O
'
n % lR
c 1Ci 1i lR
u ensd eti c 1Ci
li u
"
Ci d
"
O
u i tdtes
c
ei
"
l tc
"
"
"
"
WL8qL%
0
% !
ei C tc tc I l te l Wl R dd
"
"
|
*+ 3*
*!
, 10
0&2"
,
0%3
5/
+ #
!")"0%5(
0&+ *
!
" #"
."
0%
0
&*2+ (2"
+ 4&!&6"!
)"0%5( 50+ /&*"
&*0".
)"!&
0"/
"*
"("2"*
0.
*/(+ 0&+ *
,.
+ 0"&*/
/
-1
*0&
((!5
+ 4&!&6" %
)"0%5( 50+ /&*"
)
0+ %5!.
+ 45)"0%5( 50+ /&*"
%)
#+ .
)5( 50+ /&*"
#
*!
.
+ 45( 50+ /&*"
#
*!
*
"
.
")+ 2"!
5
0%5)&*"
$(5 + /5(
/"
*!
.
", ("!
550+ /&*"
2&
/"
"4 &/&+ *
.
", &.
(0%+ 1$%
0%"
"40"*0
0+ 3%&%
0%&/
)"%
*&/)
+ , ".
0"/
&*
/, "&#&"((
05, "/
!1.
&*$
!"2"(+ , )"*0
&/
1*'*+ 3*
0%".
,.
+ , + /"!
)"%
*&/)/
+ #
!")"0%5(
0&+ *
.
"
("//
3"((
"/0
(&/%"!
&* (1!&*$
!"
.
+ 45(
0&+ *
+ #
8
)"0%5(0.
*/#".
/"
)"!&
0"!
.
")+ 2
(
+ #
0%"
%5!.
+ 45)"0%5(
$.
+ 1, + #
%)
*!
!"
)&*
0&+ *
+ #
%)
*!
)
/""
)
&*
0"40
5
0%"50&!&*"
!"
)&*
/"/
0&2
0&+ *
&*!1 "!50&!&*"
!"
)&*
/"
*!
, + (&, + , .
+ 0"&*
) "!&0&*$
"*65)"
0
(50&
, + (5, ", 0&!"
"*65)"/
!"
)&*
0"50+ /&*"
/"/
&*
0+ *!
0%"
(
.
$".
#
)&(5
+ #
"*65)"/
%
2"
""*
,.
+ , + /"!
0+ "##"0
!")"0%5(
0&+ *
5
!"
)&*
0&*$
)
*!
%)
&*
0+ 5&"(!
0%5)&*"
*!
%)
.
"/, "0&2"(5
/
0%"/"
.
"
,.
"/"*0
&*
)&/)
0 %"!
*!
%)
/", &.
/
0%"5
%
2"
""*
,.
+ , + /"!
0+ "
"4 &/"!
5
/&*$("
/0.
*!
/"("0&2"
)+ *+ #1* 0&+ *
(
1.
&(
$(5 + /5(
/"
+.
%&/
)"%
*&/)
&/+ *0.
+ 2".
/&
(
%+ 3"2".
/""
)
&*
0"40
|
β
9
$(1 + /5(
%5!.
+ 45)"0%5(1.
&(
&+ /5*0%"/&/
%"
0%5)&!&*"
+ 4&!
0&+ *
/0", )"!&
0"!
5
&*!&*$
,.
+ 0"&*
+.
0+ ,.
+ !1 "
%5!.
+ 451.
&(
%)
&/
*
(+ $+ 1/
0+ 0%"
)
+ 4&!
0&+ *
)"!&
0"!
5
,.
+ 0"&*/
/
.
"
0%"
#+ 1*!&*$
)")".
/
+ #
0%"
7
/1, ".
#
)&(5
0%"
,.
"!&0"!
+ 45$"*
/"
!+ )
&*/
+ #
*!
3".
"
1/"!
/
0%"
/0
.
0&*$
, + &*0
#+ .
0%"
/"-1"* "
,.
+ #&("
/"
.%"/
0%
0
.
"+ 2".
"!
0%"
%+ )+ (+ $+ 1/
!+ )
&*/
+ #
0%"
0%.
""
)
))
(&
*
|
"%
*&/)
5
3%&%
%)+ 1(!
#&(&0
0"
.
", (&
0&+ *
!", "*!"*0
!")"0%5(
0&+ *
/5))"0.
&
((5
)"0%5(
0"
!
, /"-1"* "
&/+ *2".
0"!
!1.
&*$
.
", (&
0&+ *
&*0+ 03+ /5))"0.
&
((5
)"0%5(
0"!
/0.
*!/
("#08
,
*"(
")&)"0%5(
0"!
, /&0"/
.
"
.
"+ $*&6"!
5
0%"
+ (&$
0"
,
.
0*".
+ #
0%"
)
&*0"*
* "
)"0%5(0.
*/#".
/"
3%&%
.
"/0+ .
"/
/5))"0.
&
(
)"0%5(
0&+ *
,.
+ 0"&*/
0
0
)"0%5(
0"!
, /&0"/
0+ $"*".
0"
/5))"0.
&
((5
%5!.
+ 45)"0%5(
0"!
, /"-1"* "/
%)
*!
+ 0%".
+ 4&6&!"!
)"0%5( 50+ /&*"/
)
5
&), &.
)
&*0"*
* "
)"0%5(
0&+ *
5
&*%&&0&*$
&*!&*$
0&2&05
+.
+ 0%
.
&$%0
,
*"(
/
.
"/1(0
0%"
, /"-1"* "
,.
+ $.
"//&2"(5
(+ /"/
)"0%5(
0&+ *
0%.
+ 1$%
/1 "//&2"
.
", (&
0&+ *5 ("/
n er sc u t l eCa eaC en e u al e C l 1i c C 1t ds ei u adet1d l etu adt0
c i C C ei C I ad e
C
OH
N
l tc tl e dC I ad ei Cs dβu ! c%Rel% x$% a
ntl ei c l
c e i en
#tc
# c aCi c l ntI n d ( d
i u 1 CeuOH c el 0 i Ctc l e c" (#(,!W
l
( $#
O
nCi u etc tl
1 c
"% &
" #( # #
" ( ,! ( $#
O
OH c l Ct1eti c x c di R i : u %HNC i ac
tc eC
HO
ner l e
c tc
"
O
dtc o
ei 1d s
l l Wu tc ds β # !)a#$$*'#,!(ei& #'en& ' l 1 t d c Ct nu c e i entl
o s C I ad eti c I c i u t
i : nu
"
%&$( #
r es r v eo
aaes , r c f n n u v er n r f"
aaes , c n "s i e
( $#
$# v
n n r ueou eF g l n i o i ne r & r n al
n h %! r w
o u r na r oe%! n(er
,
O
r n ooi eF
g Bwl n i oOi ne OH r vBl , ne Bw i aeFi l n i o i ne r vB l , mBw
vB , HN
r Bw a eFi o i nHN
e r vB
, h c ae f
nea
2. pPh
"m
" ael
l l ei
u ri C u
W1i l ezeC c l d eti c d u i t t eti c l x
n c tl u
eaCc i ( C x c ec C F
l l c et dds
( Cs
s R nt n en
dd
c u i ad e c a d i l i u
c toi WL8qZ%
0 tl ei c
ddad C 1Ci
l l C 9atCtc I
l l Wtc da tc I
1C i u tc c eds I di ad C
)
ac l eCa eaC 0 l eCtotc I eaC i ntl ei c l W c 1 Cet ad Cds i en tC e tdl Wtl en
1e i C en tC
ll u d
C
c
E8
tl ei c l
u i t t eti c l
eC c l Ct1eti c WC 1dt eti c
| JUNE 2013 | VOLUME 14
C 1 tC0
%i tc en
iC
C
ze Cu tc d vve tdl W
jj R nt n
C
c eCi a eti c
Rt
( Ct es i 1i l eeC c l d eti c du i t t eti c l en e
t C c e ntl ei c
C l t a l 0 deni aI n u i l e i en
e Cu tc d i u tc Wl ( C d i en Cl n (
e Cu tc d i u tc x Ci 1
c C 1i Ce
ntl ei c
c CI CF
c i ac
c esCi l tc ns Ci ) sd eti c x c
u i t t eti c l
n CI
l i ad
c
C i tc en
z
tc en
z
ln (
t C l t a l i c ntl ei c l Wtc da tc I
esd eti c W 1Ci 1ti c sd eti c W aesCsd eti c W i Cu sd eti c W
Ct i l sd eti c c en C
s R nt n ntl ei c
e
tc en I di ad C c
1ni l 1ni Csd eti c W a t9atesd eti c W l au i sd eti c W
tl i u Ctp eti c W
C di
nc t CWL8qZ%
0 ed l eqZ es1 l i e i ( C q: 8 t C c e u tc i
u ensd eti c W
l
e d CI c au
c eds tl i ( C
WL8qq%
0 n C
i c eCt ae ei
teCaddtc eti c W 1Ci dtc
dsl tc
Ci ei c sd eti c
C eR i oc i R c u
n c tl u l
nCi u etc C I ad eti c 0 tCl eWntl ei c
tC eds u i ad e en 1 o I tc I i en
s de Ctc I c e
i en ntl ei c i C s de Ctc I tc e Czc a d i l i u d tc e C eti c l W tc I entl
d l e i c l 1 t dds tu 1i Ce c e tc I di ad C
c ec CF
CI C F
nc t CWL8qZ,
c toi WL8qZ%
0
i c dsWntl ei c
l 1 t t
l x Ci 1
l C I ad e
nCi u etc l eCa eaC
tc tc I i u tc l WR nt n C i I c tp u i t t
c ac eti c s C Catetc I
ntl ei c l ( t l 1 t dtp
l eCa eaC l x tI aC qE%x i ap Ct l WL88U%
0
Figure
f a 1.
pRRecruitment
dv , el of Proteins
r f to Histones
ea n a e r n er e el l e
a f v , F l c o e c aen r (A) Domains used for the recognition of methylated lysines, acetylated lysines, or phosphorylated serines. (B) Proteins found that associate preferef
r ne n a c a r n l e
a f r P r Rh
c n ael ef g a m2. . 3h
entially with modified versions of histone H3 and histone H4.
Ci e tc lpathways
en e Cneedi toI cbetpactivated
l 1 tot observe
d ntl eithec C that
l t aacetylation
l C of dH4K16
l l t thas atcnegative
ei enCeffect es1
signaling
on thel 2
modifications. In some cases, the specificity of enzymes
formation of a 30-nanometer fiber and the generation
ofs higher-order
et ial.,
enthat modify
i c l histones
en e can
C be
i Iinfluenced
c tp cby other
l a factors:
l 9a c ed
u i t s structures
en C l(Shogren-Knaak
t a
C oc
R2006;
c l
complexes in which enzymes are found may specify
also see Minireview by D. Trementhick, page 651 of this
preference
for nucleosomal
RaCte
Cl , en
C l 1i cverses
l t d free
l histones
i en(Lee t cissue).
et t Phosphorylation
ti c c isCanother
u i ( modification
u c e ithat
may
en
et al., 2005a); proteins that associate with the enzyme
well have important consequences for chromatin comaffect its selection of residue to modify (Metzger
via charge changes. The role of this modification
umay
i t t
eti c l C en C l Cl W tc dds en dpaction
l e I Ci
a1 C eni l 1Ci e tc l en e c
et al., 2005) or the degree of methylation (mono-, di-, or
has not been demonstrated rigorously in vitro but demontri-) at a specific site (Steward et al., 2006).
strations of it’s role in mitosis, apoptosis, and gametogenC i I c tp l 1 t t C l t a xu i t t i C cesis
i e%
Wsuggestive
tc ei oftesuchc a role)(Ahn
C ettl al., en
i Rc
are
2005;tC
Fischle
Mechanisms of Histone Modification Function
et al., 2005; Krishnamoorthy et al., 2006).
ac
l Wen
l C mechanisms
en l i z for
ddthe C
aC q:are%recruited
x R l to
i cmodifications
WL8qL%
0 via
Thereeti
arectwo
characterized
function Cl x tI
Proteins
and bind
of modifications. The first is the disruption of contacts
specific domains (Figure 1A). Methylation is recognized
between nucleosomes in order to ‘‘unravel’’ chromatin
by chromo-like domains of the Royal family (chromo,
and the second is the recruitment of nonhistone proteins.
tudor, MBT) and nonrelated PHD domains, acetylation is
The second function is the most characterized to date.
recognized by bromodomains, and phosphorylation is
Thus, depending on the composition of modifications
recognized by a domain within 14-3-3 proteins.
on a given histone, a set of proteins are encouraged to
A number of proteins have been identified that are rebind or are occluded from chromatin. These proteins
cruited to specific modifications (Figure 1B). The recent
isolation of several proteins that recognize H3K4me has
carry with them enzymatic activities (e.g., remodeling
ATPases) that further modify chromatin. The need to
highlighted the fact that their purpose is to tether enzyrecruit an ordered series of enzymatic activities comes
matic activities onto chromatin. BPTF, a component of
from the fact that the processes regulated by modificathe NURF chromatin-remodeling complex, recognizes
tions (transcription, replication, repair) have several steps.
H3K4me3 via a PHD domain. This recruitment tethers
Each one of these steps may require a distinct type of
the SNF2L ATPase to activate H0XC8 gene expression
Eq
chromatin-remodeling activity and a different set of
(Wysocka et al., 2006; Figure 3A). The PHD-finger protein
modifications to recruit them. Below is a more detailed
ING2 tethers the repressive mSin3a-HDAC1 histone dedescription of the different mechanisms by which modifiacetylases complex to highly active, proliferation-specific
c eCi a eti c
! ! &.&* ) * # % " ( & '
! ! &.&* ) * # % " ( &
( * ! & &# . & * )
( * ! & &# . & * )
+&$
" ")
+&$
'
. & 21 , &. " , 3
" )" . & 21 , &. " , 3
%
!
!
" ( * 0 ' * #
" ( * 0 ' * #
( * ! & #& . & * )
( * ! & #& . & * )
+&$
" ")
+&$
. & 2" , - " , 3
" )" . & 2" , - " , 3
!
#
!
" , / &.( " ) .
" , / &.( " ) .
+&$
" ")
+&$
. & 2, " ! " , 3
" )" . & 2, " ! " , 3
'." , .&* ) * #
'." , .&* ) * #
. " ( +' . " ! + , * " - . " ( +' . " ! + , * " - -
!
" !
!
" !
l
! f a & pB d
! $r ! c $ r n "
!
%
!
&
! $ !
$
"
!
& % m ean
!
a f ! o n! ea r !n a
& %
!
!
# aef! c y s a! n a m a! a !
r
#
!
!
!
!
a ! ! a h
c& % n& % ! ael!! ! ! % ! s& % ! &er! m # #
# m2.! p2h
!
!
!
#
!
!
!
!
!
$ !
&
!
!
!
$ !
&
! !
&
!
#
!
! !
&
!
#
!
!
!
!
!
!
!
&
!
'
!
!
!
&
!
'
!
!
!
!
! !
&
!
!$
!
!
! !
&
!
!$
"
! !
!
!
$ !
#
"
! !
!
!
$ !
%
#
" ! #
!
! ! !
%
#
" ! #
!
! ! !
"
! !
!
!
"
! !
!
!
!
!
!
!
"
! % !
%
$
!
"
! % !
%
$
!
$
!
! % !
%
!
$
!
! % !
%
%
!
$
%
!
$
! !
!
!
!"
!
!
! !
!
!
!"
!
!
!
!
" &
!
!
" &
. 3 .! 3+ 8! + 8
31 8/31 38/. / 3' . / -' : / - 1 : ./ + (1 . $ + (-,$ .,3(% . 3(% (& (&
6 ' 6 $ -' $ 3'- $ 3' 2$ ,2 $ , + 82($
1
$
2(#
4$
4#
$
1
&
.
$
2
$ + 82(- $ 1 $ 2(# 4$ 4- # $ 1 & . $ 2 . 3'- .$ 3'1 $ 1
, .,# (%. # (" (% 3(.
24"
' ' 2 "2$ 38+" $ 38+3(. -3(. (3
(" 3(. - 24"
- ,(3 ,8 3' 8$ 3'- $ / -1 ./ 1 .
5 (# 5 $ (# # $ . " # * . (-" * & (-2(3$
2
%
.
1
.
3'
$
1
/
1
.
3$
& 2(3$ 2 % . 1 . 3' $ 1 / 1 .(-3$ 2(-" 2. -" 3. -(-3 (-(-& (- &
iu
l C tc i u 1 et d R ten i en Cl Wi i aCl " $eni
l
en
e
u i t s en l 2u24"
38+
:
+
82($
!
' ' 2 ! 21 .! ,1 .., .
" $ 38+ : + 82(- $ (-! # (-(-# & (-# & . ,# . ,(- 2(- 24"
"
#
.
,
(2
(+
+
8
3.
#
#
3.
3'
$
"
3. 3' " $ . ," ./ ,+ $ / 7 + (38
$ 7 (38
c a d i et
C u aea " dds ) dal t( W" ae i en Cl
C , # -d.l8,i " ' (-1 .)2, dal
t(1 (-$ & 4++ + 8a3. 3.1 2 ei' # 5# $ 1aC
3("
"
, - 8 " ' 1 . , 3(- 1 $ & 4+ 3. 1 2 ' 5 ,$ .,1 $ . 1 3'$ 3'- .- - $ . - $
$ 3'(1 $ (1" ' 1 " .' ,1 . ,3(- 3(# $ # 1 l. i,# u
. ,(- (- -l# -l# 3' en
1nsl t d ddi l e" Ct
tc e C eti c l , l i u i en C
l 38/
i 38/$ ) tl.$ e% . 1 % $ c 1 # $ $ 1tc
"
!
(#
(&
"
!
$
%
41
3'
$
1
(%
+
4$
"
$
#
!
%
! (- # (- & " - ! $ % 41 3' $ 1 (- % + 4$ - " $ # 8 ! - 8$ (&- $ ' (&! .' 1 ! (-. 1 & (- &
%
'
(23.
-d
$ tl-, $nu
.,# (%. # ("c(% e("3(.i -3(.2 -c2.
" + "+ C
$ # + i+ $ ,c# 4+, 0
3(54+ 3(5+ $ - + 3$ -$ -3 $ 1i l teti c i l 1 t t u i t t eti c c n c l en l e' (23.
2i en
2.
& & & $ ,& $ $ ,- 3$ .- % 3' . (23
.
$
,
.
#
(%
("
3(.
2
% ' (23. - $ , . # (% (" 3(. - 2 ' $ 2$' $ 2$$ 7 $ ,7 ,
dd en l tc e C eti c l C oc i R c l ntl ei c u i t t/ + $ / 2eti
'+ $ 2(&c' ' (&+ (& 'Ci' + (&3l ' 3'l 3ze
$ 3' , $ do
4+, 3(%x4+ 3(%tI
" $ 3aC
$ ," ' $ " -' (2,- (2,
2 3'2 3'3 3
" $ $ # 3$,# q/%
"
'
1
.
,
3(1
$
#
$
1
2
42$
3
.
#
$
"
(/
'
$
1
3'
$
(31
npg " ' 10
. , 3(- 1 $ # $ 1 2 42$ 3. # $ " (/ ' $ 1 3' $ (-(" 31 ("3$ 3$
c
x
ctl e CF i ap Ct l WL8qq, aen c aCI
WL88U%
$
/
(&
$
' " 1 ' " 1 3$ "1 (93$ $ 1 (9# $ / # 1 ./ 1 .
$ / (& $ $ 3("- $ 3("+ - + # 2"- # 2"/ $ / $ ' $ ' ! $ $ 23
! $ " 23
388
3$ (-3$ ! (-(-! # (-(-# & (- / & . " / * . $ " 32
"
.
3
($
#
(- (-" ' 1 " .' ,1 . ,
* $ 32 " . - 3 (- $ 6 # (3'6 (3'
3(- 3(-22. 22.
" ( 3" $ (# 3$/ # 1 ./ 31$ .(-3$2(- 12$ 24,
,
1
(9
$
#
1 $ 24, , 1 (9 $ (-# (-(& (&
' $ ' (,$ /(,. 1 / 3. 1 -3" $ - ". $ % ". ' % 1 " .' ,1 . ,3(- 3(1 $ 1 # $ $ 1 # # $ 1 . ,# . ,(- 2(- 2
"
(- ,(- ,(- 3(-(-3(-(-& (-' & . ,' .$ ,. 23
2(2
6
2
(+ + 8 + 1 + $ 85 1 $ $ 5 + $ $ # + $ #
"
$ . 23 2(2 6 2(3(
(- (3(
(- (-3' $ 3' 2$$ ,2$ (-, (-+ . + ! 2$. ! 1 2$5 1 3(.
3'
3
,
43
3(. 3(.
- 2 -(-2 (5 3(. - 3' 3 , 43
3' $ 3' $
% (- % & (-$ 1 & $ . 1 % . %
! 1 .! & 1 . 3$& 3$3' $ 3' / $ 1 ./ 3$1 .(-3$ 2(- 2
$ 7 (232
! 8 ! 5 (18 34$
$ 1 $ $ 2(#1 $ 2(#
4$ 4$ ! (+ (38
3. 3.
! (-! # (-31# (,31 $(,3' $ 8+3' 8+3$ # 3$ #
1 $ # 1 4"$ # $ 4" $
$ 7 (232
5 (1 34$. % 3'. % $ 3' % $ " % 3 3'" 3 3'3 $ 3" $ ' " + 82(' + 82(! (+ (38
, ,8 ! 8$ ! 4-$ ,4-$ ,3' $ 8+3' 38+$ # 3$ ,# .,- .. ,- .$ ,3' $ 8+3' 38+$ # 3$ # # (, # $ (,3' $ 8+3' 8+ 1 $ " 1 .$ ," .! ,(-! (-3(. -3(. -(- (-6 ' 6 (" ' ' (" 5 ' 1 5 ( !1 (+ $ ! + $ # (5 # $ 1(5 2$$ 1 2$ - # - #
3$ # 3$ .# 1 .31 1 (,31 $ (,3' $ 8+3' 38+$ # 3$ # . 1 .(-1 23
- " $ - " 3'$ $ 3' / $ + / - + 3 -' 3. ' . ). (-).(-(-& (-& & $ -& $ $ -2$$ & 2$, & $ ,- 32
(- 23
$ - 321 $ 11 $ -1 # .- ,# .+ ,8 " + 8. ," .! ,(-! $ (-# $ #
, $ ,. # $ .. ,# . ,(- (% (- %& (-$ 1 & $ (2
1
$
&
(.
6
(3'
(#
1
$
24+
3
((,
,
4.
#
# 1 .# ,1 .$ ,2 $ 2 4 4
1 (2 1 $ & (. - 6 (3' (- # 1 $ 24+ 3 (- (, , 4- $ .% # (" $ ($% ("- ($" 8- "288 28/ 1 ./ 31$ .(-3$ (-3' 3'3 (23 (2
" / " ! / + $ ! + .$ % .# % $ 3# $ $ " 33(1 $ 1 # $ (-# & (-" & "
$ " &3(- ,& $ ,3' $ 8+3' 38+$ # 3$ # 3 3(.3 -3(.2 -3'2 3'3 3! 1 .! & 1 . 3$& 3$3' $ 3' " $ ' 1 " .' ,1 . ,3(- 3(' (23
.
$
2
'
$
%
(&
$
1
2
.
%
3'
$
/
1
.
3$
(2
/
"
(38
.
%
,
8
$
/
(&
$
$
3("
1
$
&
4+
3.
1
2
' (23. - $ 2 ' $
% (- & $ 1 2 . % 3' $ / 1 . 3$ (- 2
/ " (38 . % , - 8 $ / (& $ - $ 3(" 1 $ & 4+ 3. 1/ 2+ / 8+ 8- (-- (-# -#
/ 1 $ / % $1 $ 1 $% $ -1 3($ - + 3(+ 8 + + ! 8(-! # (- 4,
$
3'
8+
3
$
#
%
+
4$
. % .# % (2$# (2$2$ 22$ (# 4- , $ 3' 8+ 3$ # % + -4$3(- +3(1 .+ + 1$ . (+ $ (- 5 1 5 ($ 38
1 ($ 38
2 " (-+ 4#" + (4# & (- &
6
'
$
1
$
2
3'
$
%
(&
$
1
.
%
+ 82($
1
$
2(#
4$
2
2(&
3'
$
2$
"
'
1
.
,
3(1
$
#
$
1
,
.
#
4+
"
"
$
1
+ 82(- $ 1 $ 2(# 4$ 2
6 ' $ 1 $ 2 3' $
% (- & $ 1 . % " - " $ 1
2(- & 3' $ 2$ " ' 1 . , 3(- 1 $ # $ 1 , . # $ 4+2 $ 2
! (-! # (-2 # ,2.,23. 235 (# 5 + 8(# 3+ .8 3# . (, # $ (,3' $ 8+3' 38+$ # 3$ # - # -31# ( 31 ( 2 3' 2$ 3'1 $ / 1 $ 43("
/ $ 43("3 1 3& $ 1 32
& $ ,32 ,8 . 8% % .$ %1 % $ 1 4- (04$
4- (04$. / / . ./ 1 / . 1
3'
$
1
2(,
(+
1
,
$
3'
8+
:
+
8
, $ ,3' $ 8+3' 38+$ # 3$+ #82($
2
34(38
3.
3
(+
.
1
3'
$
1
/
($
2
3.
2/
$
"
(%
("
#
+ 82(- $ 2
3' $ 1 2(, (+ 1 , $ 3' 8+ : + 8 34- (38 3. 3 (+ . 1 3' $ 1 / ($ 2 3. 2/ $ " (% (" (2$# (2$2$ 22$ 2- Figure 2
2($
1
$
"
.
&
(3(.
,
.
3(%
2
$
7
(23
(-
" + 4#n
& "
1"
.
,1 .
.
$ 7
$ $
,7 / ,
+
.
% n.3'u% (2
/
1
.
"
$ 22
' 2 1 $ " $ - 3+ -83+! 8$ $ ! -$ $ / -1 ./ 1 .
2($ " .& -(3(.
- ,
.
3(% 2
(23
"
+ (4# (& e' '
,n
.er
/ 1 e + (2
22
f
a
p0
d
ner
l$ 1 e
er
ae n
o7 Ch
er
l
c
o3'i ea/ r1 . " $ n
u ' oi 2 1 $ " $ n
n
w
$ (
$ r
1 [53]
#
.
,
(2
34#
.
1
#
.
,
(2
#
3'
$
#
.
,
(5
(#
$
#
6
(3'
2,
+
+
,
.
+
$
"
4+
$
2
3'
en a l e
n er
n u. 1 # . , n (- 2r - # 3'n $
i r # aaes
$ r2 3'3 2/ 3
n$ 2/" (%$ ("" (% ("+ + 8 + + 8- # - #
# . ,h (-c e 2 34#
. , (- 5 (# r$ # 6 r(3' 2,n u + + , . + $ n" 4+
,
$
#
%
.
1
(32
"
'
1
"
3$
1
(23("
/
1
.
+
($
:
31
8/
3
.
/
'
:
5
(#
+
8
('
(!
(3
3'
$
3
#
$
,
!
1
.
,
.
#
i
on
h -cn
, $ # ael
% . 1 (32 " r' r 1 n" 3$a |1 (23("ef / g1 .a+ (- $ : m
312.8/ pph
3. / ' - :
5 (# + 8 (- ' (! (3 3' $ 3 - # $ , ! 1 . , .. ,# . ,(- 2(-. 2% .3' % $ 3' $
of binding of many chromatin prointeraction is stronger when H3K9 and H3K14 are also
n ac eti c i u i l entl ei c
l l analysing
etddC u thetcpattern
l ac oc
i R c Wdeni aI n en Ci d en e
teins [84]. However, given that most is known about the
acetylated on the same tail of H3 [60]. However, this staNewsimple
England
Journal of
Medicineabove, references below
domains
described
bilization of binding may be due to additional factors in Thetwo
The New England Journal of Medicine
l
i)
u(%
i from
en
u
1d
tc
d
d
jl%
ti diIDE
s DE
tl(
ee
C-ac
Cl
ei%
i26,!%
0
d
Z !
l au
Ctp
l enuses without permission.
Downloaded
nejm.org
atsCRAI
UNIVERSITAT
BARCELONA
on
November
2013.
For
personal
useuonly.
No other
%
!%
$
(
*"
$!
!
only.
$
!
"
#
%
#
%
#
%
!
Downloaded
from
nejm.org
at
CRAI
UNIVERSITAT
BARCELONA
on
November
26,
2013.
For personal
use
No other uses without permission.
Copyright © 2012 Massachusetts Medical Society. All rights reserved.
Copyright © 2012Both
Massachusetts
Medical Society.
All rights reserved.
heterochromatin
and euchromatin
are enriched,
itself.
Ci d i en
eoc i modiR c ntl ei
l 0depleted, of certain characteristic histone
andcindeed also
There ac
may eti
alsocbed
cooperation
betweenlhistone
-
%
!$
%
*
%
!
!
#
$%
%
modifications. However, there appears to be no simple
rules governing the localization of such modifications,
protein binds to nucleosomes bearing H3K9me3, but this
$
$ %
*
(%
&
!$
!
and there is a high degree of overlap between different
chromatin regions. Nevertheless, there are regions of
DNA isEL
CpG methylated [61]. Conversely, DNA methydemarcation between heterochromatin and euchromatin.
lation can inhibit protein binding to specific histone
$
+!&#
*
$%
,
#
!&*
$"
-
%
!#
$
!
-
%
! $
!!
)"
#
$
(
such as CTCF that play a role in maintaining the boundonly binds to nucleosomes bearing H3K9me3 when the
c eCi a eti c
Table 1. Different Classes of Modifications Identified on Histones
Chromatin Modifications
Residues Modified
Functions Regulated
Acetylation
K-ac
Transcription, Repair, Replication, Condensation
Methylation (lysines)
K-me1 K-me2 K-me3
Transcription, Repair
Methylation (arginines)
R-me1 R-me2a R-me2s
Transcription
Phosphorylation
S-ph T-ph
Transcription, Repair, Condensation
Ubiquitylation
K-ub
Transcription, Repair
Sumoylation
K-su
Transcription
ADP ribosylation
E-ar
Transcription
Deimination
R > Cit
Transcription
Proline Isomerization
P-cis > P-trans
Transcription
Overview of different classes of modification identified on histones. The functions that have been associated with each modification
are shown. Each modification is discussed in detail in the text under the heading of the function it regulates.
o Pd el
r r n a |
e n l ea nf
ef g a m2. pph
r n
a a f o neai f r n er e
n h
cn
ael
a protein has to be digested before such analysis can take
for the biological response. However we know that many
place limits its potential. New methodology that uses
of the histone-modifying enzymes have other nonhistone
ei c proteomics
l approach
C c i e(identify protein
l 9a
tc 1 cSo the
c eW
C mayc ebe 1a
eti c
atltop-down
first c
substrates.
response
goingdt through
and digest subsequently) gives promise that we may, in
another unidentified protein substrate. Furthermore, there
i c l look
eC eat the intact
en emodification
ntl ei c pattern
u i oft t
adsignaling redundancy
e such
s that
l i umoreIthan
c one
et
theufuture,
differ-eti c l mayi be
ent histones in a given nucleosome (Macek et al., 2006).
enzyme may be capable of modifying a specific site. In
( Once
Ct c el
x analysis
t o ofCall histone
WL8qZ%
WR nt
c i entheCeffects
) uof 1d
i enone enzyme
1 Cu may
c be
ce
global
modifications
is n tlthis case,
inactivating
done, a prediction would be that every single nucleosome
masked by an upregulation in the activity of a second
would
Ci l l be
zefound
do to be
eRmodified
c I incsome
et way.
l cThis picture
1tI isc et distinct
l 0 but related enzyme. Showing that mutation of
of course very static. The truth is that modifications on histhe modified residue gives the same output as mutating
tones are dynamic and rapidly changing. Acetylation,
the enzyme is a second stringent test. However, this is
methylation, phosphorylation, and deimination can
not possible in humans due to many histone genes
appear
present in the genome, but it is possible in yeast.
nerand udisappear
a r n on chromatin within minutes of
stimulus arriving at the cell surface. Thus examining bulk
So the truth is that we have ‘‘levels of confidence’’
histones under one specific set of conditions (with either
regarding the causative nature of different modifications
5antibodies
c i c t or d;mass
ntl ei
c I c l C i ac tc daldepending
e C Con1how el
WR n C l i en C ntl ei c
spectrometry) will identify only a
far the analysis has gone to prove the
proportion of the possible modifications.
issue. We also have to be realistic and accept that, howI There
c lW
enalsol iproblems
z dd of ntl
ei c that( are
Ctspecific
c el W C ever
es1tfar we
ddsgo iinac
I d modification
i 1s I cis
are
detection
proving lthat al tc
histone
for antibodies. Firstly, there are the obvious issues of
causative, we can never exclude the possibility that
xspecificity.
u o These
o Fare difficult
tI I tctol W
L88:
%
0 are
n no tl l tumodification
td Ctes of other
eR substrates
c en byu theri same
C ntlenzyme
ei c
avoid
as there
true controls for modifications in mammalian cells (unlike
will play a parallel role in the biological response being
l yeast)
a es1
l it cis impossible
en ( toCtmutate
c el thei ad
Ci u
Rmanyuother
tc i nonhistone
t l eisubstrates
d CIof
where
residueCtoc I
monitored.
The
make sure reactivity is lost. In addition, an adjacent
chromatin-modifying enzymes are not covered in this
modification
binding
or CsiReview.
t( CI c mayxedisrupt
d the
E%
0
i u of the
i antibody
en ao
et ntl ei c ( Ct c el Wl a n l
z W
a protein may occlude its recognition, both of which may
give
a false
Z0ZW
L reading.
0 i CSimilarly,
L 0 there
W Care problems
du i l e of
ac t( Histone-Modifying
Cl dds i c l C(Enzymes
enCi aI n ao Csi et
detection that are specific to mass spectrometry. Peptide
The identification of the enzymes that direct modification
coverage
for all parts
the L8q8%
histone and
has been the focus of intense activity over the last 10 years
( i daetiiscnotx equivalent
d CeF
c toiof W
0
this reduces the sensitivity of detection in these regions.
(Table 2). Enzymes have been identified for acetylation
These facts undoubtedly contribute to our underestima(Sterner and Berger, 2000), methylation (Zhang and Reinl
1te
i
c
l
C(
Ci
d
i
ntl
ei
c
l
l
I
c
Cd
1 o I tc I I c el Wte tl c i R
tion of the extent of modifications present on histones.
berg, 2006), phosphorylation (Nowak and Corces, 2004),
We assume that each individual modification on hisubiquitination (Shilatifard, 2006), sumoylation (Nathan
d Cleads
en toe a biological
c i en Cconsequence.
o s ac However
eti c iproof
en l et al.,
1Ci2006),
e tcADP-ribosylation
l tl ei i c(Hassa
C et
( al.,
Ct2006),
eti cdeimil tc
tones
of a consequence is not always easy to provide and is
nation (Cuthbert et al., 2004; Wang et al., 2004b), and pronCi ubasedetcon laeCa
eaC aeimodification
c l aC appears
sc uont 1 line
ee isomerization
Cc l i eC(Nelson
c l Ct1eti
c dC I ad eti c tc
often
correlation:
et al., 2006).
a gene under certain conditions (e.g., when it is tranMost modifications have been found to be dynamic,
ao Csiand
e ldisappears
x p when that state
W
L8qE%
0 n tcand
i C1i
C etithat
c remove
i ntl
c ( Cthave
c elbeentl
scribed)
is reversed
enzymes
theeimodification
(e.g., when the gene is silent). Proving causality for a
identified. One major exception is methylation of
involves
showing
catalytic
although
they
1modification
Cet ad Cd
s tu 1i
Ce c that
e ei the1Ci
l l activity
l l a n arginines:
l
C 1 tCW
u areti thought
et C toi ube dynamic,
tc eti ca W
of the enzyme that mediates the modification is necessary
demethylating activity has not yet been found. Instead
nCi u i l i u
l I C I eti c W eC c l Ct1eti c
694 Cell 128, 693–705, February 23, 2007 ª2007 Elsevier Inc.
nCi u i l i u
tc
ic
c l eti c
CI WL88ó%
0 c
c
l 1 Cu
d E li u
tc tet eti c
c
e Cu tc eti c W l )
nCi u etc 1 o I tc I x u 1i l F
i en l
ac eti c l
C C de
ei en
i CC l 1i c tc I ( Ct c e0
EZ
c eCi a eti c
%& /) <
"
!
"
,427+,2766,)+)120 )
24),-5621)
"
,427+,2766,)+)120 )
4)3%-4 0 )( -%6)( & :6,)
3,253,24:/%6)( *240 γ
"
%1( +)120 )-16)+4-6:
$
,427+,2766,)+)120 )
)1)%'6-8 %6-21 +)1)
5-/)1'-1+%1( ',420 2520 )
5)+4)+%6-21
%'42
255-& /:
&(
1%'6-8 )" ',420 2520 )
)1)5-/)1'-1+
'6-8 )" ',420 2520 )
%1( )7',420 %6-1
'6-8 )64%15'4-36-21
,427+,2766,)+)120 )
!
%/)-1*)46-/-6: 6,%6-5 ( )*)'65
-1 53)40 0 )-25-5 -1 0 -')
0 & 4:21-' /)6,%/-6:-1
$ .12'.2760 -')%6
=
4%-1 0 %/*240 %6-21 -1
;)& 4%*-5,
=
24),-5621)
255-& /:
,427+,2766,)53)40
,420 %6-1 62
+)120 )%1( -1 6)/20 )4)5 17'/)23426%0 -1)64%15-6-21
2*520 %6-' ')//5
255-& /:
4-0 %6)6)/20 )4)5-1
53)40
%56,)4)-512 +)1)
342( 7'6-1 0 -')
,427+,2766,)+)120 )
2*2/*%'624:1)74215
8 )4)934)55-21 2*2/*%'624:
4)')3624 -1 0 -')
%1(
,427+,2766,)+)120 )
24),-5621)
,427+,2766,)+)120 )
24),-5621)
,427+,2766,)+)120 )
)1)%'6-8 %6-21 5-/)1'-1+%1(
',420 2520 )5)+4)+%6-21
%( 7/6-1*)46-/-6:-1
+)1) 64%3 %1(
.12'.2760 -')
=
3)40 +)120 )%1(
17'/)2/752*520 %6-' ')//5
255-& /:
" %1(
#
2*)%',
7',420 %6-1 -1 ,20 -1-(
6)56-5
255-& /:
%56,)4)-512 +)1)
342( 7'6-1 0 -')
7',420 %6-1 -1 34-0 %6)5
)16420 )4)5
,420 2520 )5)+4)+%6-21
,427+,2766,)+)120 )
,-5621)
/-.)')16420 )4-' 3426)-1
4)3/-'%6-21 -1( )3)1( )16
o R d ea
m
2. pRh
ner
u a r n )0 & 4:21-' ( %:
nf a %56,)4)-512 +)1)
342( 7'6-1 0 -')
!
0 & 4:21-' /)6,%/-6:%6
=
-1 0 -')
24),-5621)
,-5621)
r u an a n
6:3)!
126( )6)40 -1)(
u oec l
r nh
4)3/-'%6-21 ( )3)1( )16
c ae f
ael
g
Therefore, enhanced dynamicity of nucleosomes that
In addition to testis-specific variants, somatic hiscontain testis-specific histone variants may be required tone variants have been shown to be involved in meiosis
for appropriate chromatin reorganization during sper- when the X and Y chromosomes condense together to
matogenesis. With the exception of recent evidence21 form XY bodies. These XY bodies are transcriptionthat shows a specific role for the H2B variant TSH2B in ally silenced through a process known as meiotic sex
eoi el
r
achromatin-to-nucleoprotamine
n ea F el c o F transitions, functional chromosome inactivation (MSCI) and are considered
roles for most variants during mammalian spermato- to be recombinantly inactive, thus protecting sex chroi ds i u
c
Cteni
) Cthose
1Ciof etheir
tc biophysical
i u 1d properties
) l tl 1d
stc I from
i 11i
l te
ac eti events,
c l 0 which are
genesisCbeyond
mosomes
illicit
recombination
remain unknown. However, an increasing amount hallmarks of active meiosis. The XY body is enriched in
i ds i u R l tClofed
s
l indicates
Ct
tc active incorporation of specific H2AW
R n C(including
i dsmacroH2A
iu
evidence
that
variants
(mH2A)24
25
these variants might directly affect transcription. For and H2A.X ) and H3.3 (REF. 26). The phosphorylated
depletion
u ae c e dt l ( dexample,
i 1 c i Cu
d iof H2A.Bbd
s l I uresults
c elin widespread
ae tc en form
R CiofcH2A.X
I 1d(γH2A.X)
W i Calso)hasua crucial
1d Wrole in male
reduction in basal gene expression and disruptions germ cells, as H2A.X-null and transgenic mice fail to
properly
l i u i en u ae ctoelalternative
( dsplicing
i 1 1n22. Furthermore,
c i es1 l lhuman
a n cells
l in ad
e c tuformdlXYRbodies
ten dand
I l undergo
tc en MSCI, which
which H2A.Bbd has been ectopically expressed show an results in male sterility 27. Another variant, H2A.Z, is also
Canonical histones
variant throughout active regions of enriched at heterochromatin domains of postmeiotic
1d
i histones
c eof a c c enrichment
x c c tlfori this
cW
qóó: %
0 n l u d i Cu eti c l C a ei
CC c e
Archetypical
the genome22,23, which collectively indicates a putative haploid spermatids, including at the sex chromosomes28,
given family to which all other
this testis-specific variant in the regulation of although its functional role during spermatogenesis
of that same
Ihistone
c proteinset(
eti c irole
c lfor
9a
c i en u ae eti c tc en remains
i ds unknown.
iu
i u 1d ) x
%
gene
expression.
family are compared.
I c l Wl
NATURE REVIEWS |
et(
eC c l Ct1eti c dds tc
i ds i u
l
Cteni C ) I Ci a1 x C) %
WR nt n i ac e C e en
et( etc I I c
EE
t c et t
nCi u etc tc
ADVANCE ONLINE PUBLICATION |
et( i Cvi j l e e x tddl WL8q8%
0
e Ci c WI c et l ea t l tc
oc i R c l
1Ci e tc l u s td ei o 1
i c I Ci a1 i 1Ci e tc l W
eti c i ) 1C l l ti c tc ni u i et I c l x tu i c Wqóó: %
0
i u 1d ) W
c eCi a eti c
$)*3
-
B
>1
5/C:
:)9
+ 64 )
*9
-):;
69
)3
# 0-7);
6+ -3
3
<3
)9
+ )5+ -9
79
6:;
);
-
+ )9
+1
564 )
4 -3
)564 ) + )5+ -9
)5,
-5,64 -;
9
1
)3
+ )5+ -9
!" #
&
9
-7-);
:
"-8<1
9
-,
.69
;
0-
$
)+ ;
1
=
1
;
@
6.
( (
( )5,
( #$
$
-:;
)*3
1
:0-:
4 - 4 )9
2
#%(
#%(
(1
5+ .1
5/-9
"-8<1
9
-,
.69
;
0-
$
)+ ;
1
=
1
;
@
6.
( %"
"!
)5,
"!
1
:;
65-
*1
5,1
5/
69
4 :
41
51
4 )3
5<+ 3
-6:64 -
*1
5,1
5/
4 6,<3
-
>1
;
0
#%(
!" !
'
'
)5,
'
09
64 6,64 )1
5
1
5,:
4 - 4 )9
2
-:670)/-)3
+ )5+ -9
4 -,<3
3
6*3
):;
64 )
3
-<2)-4 1
) 3
@
4 7064 )
5-<9
6*3
):;
64 )
0-7);
6+ -3
3
<3
)9
*3
),,-9
+ )9
+1
564 )
):;
9
6+ @
;
64 )
)5,
+ )5+ -9
79
6:;
);
-
+ )5+ -9
!
! ! )5,
! (1
5+ .1
5/-9
)5,
#!
"-8<1
9
-,
.69
:1
3
-5+ 1
5/
"
" )5,
"
" A
1
5+ .1
5/-9
%*1
8<1
;
1
5
3
1
/):-
!#
! )5,
(
(1
5+ .1
5/-9
!9
6;
-1
5D79
6;
-1
5
1
5;
-9
)+ ;
1
65
$"'
)5,
#$
)5,
$
0662
E
$
-:;
)*3
1
:0-:
-<2)-4 1
)D
4 - 4 )9
2
)5,
*1
5,1
5/
# # #$
$
-:;
)*3
1
:0-:
4 - 4 )9
2
# # & "
&
9
-7-);
:
::-5;
1
)3
.69
4 -
)5,
E
E
E
E
" "0)*,61
,
;
<4 6<9
:D
3
<5/
D
69
)3
+ )5+ -9
+ )5+ -9
#& D#
3
1
2-
0-3
1
+ ):- $!
):-
)+ ;
1
=
1
;
@
/):;
9
1
+
+ )5+ -9 # :21
5
+ )5+ -9
9
64 6,64 )1
5
1
5,:
)+ -;
@
3
);
-,
01
:;
65-: )5,
"
#" E
E
#"
#
" )5,
" E
65
+ );
)3
@
;
1
+
+ 69
-
:<*<51
;
E#& E#" #$
)5,
1
:;
65-
*1
5,1
5/
)5,
#& D#
3
1
2-
0-3
1
+ ):- $!
):-
)+ ;
1
=
1
;
@
%" !$
! A
1
5+ .1
5/-9
)5,
*9
64 6,64 )1
5
1
5,:
4 - )5,
*1
5,:
)+ -;
@
3
);
-,
01
:;
65-:
%"
"!
)5,
"!
E
E
E )5,
09
64 6,64 )1
5
)5, E$!
):-
)+ ;
1
=
1
;
@
#& D#
3
1
2-
0-3
1
+ ):-
)5,
09
64 6,64 )1
5
E$!
):-
)+ ;
1
=
1
;
@
#& D#
3
1
2-
0-3
1
+ ):-
)5,
! A
1
5+ .1
5/-9
#
09
64 6,64 )1
5
E$!
):-
)+ ;
1
=
1
;
@
)5,
#& D#
3
1
2-
0-3
1
+ ):-
)5,
! A
1
5+ .1
5/-9
# 9
-3
);
-,
#& D#
#
" )5,
" "
#& "
E
E
3
1
64 )
5-<9
6*3
):;
64 ) D
4 -3
)564 ) 4 -51
51
/1
64 )
D
70-6+ 09
64 6+ @
;
64 )
63
1
/6,-5,9
6/3
1
64 ) D
565
6,/21
5C:
$
;
0@
9
61
,
3
@
4 7064 )
+ 63
65
+ )5+ -9
+ )5+ -9
*9
-):;
+ -9
=
1
+ )3
+ )5+ -9 D
)5,
6=
)9
1
)5
+ )5+ -9
+ )5+ -9
" $91
+ 0
1
5;
-9
)+ ;
1
=
-
,64 )1
5
# )*:-5;
:4 )3
3
69
064 -6;
1
+ '
+ 09
64 6*6?
064 63
6/<-
+ 09
64 6,64 )1
5
0-3
1
+ ):-
*1
5,1
5/
+ 09
651
+
4@
-3
61
,
3
-<2)
-4 B
1
)
-4 *9@6
51
+
-+ a
;
6,-94
,
-3
674
-5;
(
-50)
5+ -9
6.
A
-:
064 63
6/<-
06:
;
+ea
-3
3
.)+F;
69
$
-;
;
#
0-),
o5-+
d)4
r
n
c-=
an
r
a e4 <3
n
eo
i ;-
el
r
a
n3
caen
r01:;65-
el4-6:
c0@
o3
F9)95:
.-9):-
u
c69
)5,
2
:8<
6<:
+ -3
3
+ )9
+1
564 )
#
21
:4 -;
;
1
73
-
-5,6+ 91
5-
5-673
):1
)
41
?
-,
1
5-)/-
3
-<2)-4 1
)
%"
5<+ 3
64 -
-4 6,-3
3
1
5/
.)+ ;
! c aec
n
e @+
n64 *
F @0
r64 -6;
l 1+
C n @0
f64n-6;1+o
r)5;
n
f r-6,
n64er)15
h!" c ae
ael
o !
m2.
763
@+6 4an
*
!+ 763
/el
9
6<7c
!o
763
!
763
064ea
63
6/<-
! a
73
064
763
@+6 4f*
9
-79
-::1
=
-
+ 64 73
-?
o
"
p. h
9
-;
1
56*3
):;
64 )
*1
5,1
5/
79
6;
-1
5
# :8<)4 6<:
+ -3
3
+ )9
+1
564 )
#%(
:<779
-::69
6.
A
-:;
-
064 63
6/<-
#!
$
$
+ -3
3
)+ <;
-
3
@
4 706*3
):;
1
+
3
-<2)-4 1
)
$9?
;
91
;
069
)?
/9
6 <7
& "
&
9
-7-);
,64 )1
5
tc
en c te R l
) e c l t( ds 1Ci (
en e i en i u 1d ) l R i Co ei I en C ei
| CNATURE
I adREVIEWS
e nCi
u etc sc u t l c I c
) 1C l l ti c Wl R dd l
© 2010 Macmillan Publishers Limited. All rights reserved
dd
ddad C 1Ci
l l l x Ci o F
c
Cteni C ) I Ci a1 C ntI nds sc u t
en e l l tI c l 1 t d eaC l ei en
i en I Ci a1l u i ad e
de Ctc I ntl ei c
c u ensd etc I
i natp c WL88q%
0 n
c
c
i u 1d ) x
nCi u etc x
tc e I C e
iCu
C
ddad Cu u i CsWtc
i ds i u
c
en
s t C c e 1Ci e tc l
d :%
0
l Wn c I tc I en 1i l teti c i C en
x na ee c I Ca
VOLUME 10 | O CTOBER 2010 | t e
lli t e
1Ci e tc l % s
i u 1i l teti c i c a d i l i u l
WL8qq, tu i c F tc I l ei c WL88ó%
0
E:
Introduction RNA epigenetics, brave new world Protein-­‐coding genes represent about 1% of the mammalian genome (Harrow et al., 2009), although up to 70 to 80% of the human genome may be transcribed at some point (Djebali et al., 2012). Those transcripts that do not codify for proteins are known as non-­‐coding RNAs (ncRNAs) and have been demonstrated to be extremely relevant in epigenetic regulation in a myriad of physiological and disease circumstances (Esteller, 2011). Within ncRNA family there is a number of different subfamilies such as transfer RNA (tRNA), ribosomal RNA (rRNA), small nucleolar RNA (snoRNA), micro RNA (miRNA), small nuclear RNA (snRNA), extracellular RNA (exRNA), piwi-­‐interacting RNA (piRNA) and long ncRNA (lncRNA). In the last decade, micro RNAs and long non-­‐coding RNAs have gained relevance as major epigenetic regulators. miRNAs are a large family of post-­‐transcriptional regulators of gene expression that are ~21 nucleotides in length and control many developmental and cellular processes in eukaryotic organisms. Drosha and Dicer complexes process pre-­‐miRNA to miRNA, which are exported to the cytoplasm, and form a protein-­‐miRNA complex called miRISC. The complex is guided to the target mRNA by miRNA sequence homology and induces its translational repression by deadenylation and degradation of the mRNA. In mammals, miRNAs are predicted to control the activity of ~50% of all protein-­‐
coding genes (Krol et al., 2010). Long non-­‐coding RNAs must have at least 100 nucleotides in length. The first lncRNA discovered in mammals was the X-­‐inactive-­‐specific transcript (XIST) (Brown et al., 1992). The Xist gene is expressed in the female X inactive chromosome (Xi), and produces a 17 to 20 kbs RNA that coats the Xi and in cooperation with the Polycomb Repressive Complex 2 (PRC2) complex triggers chromosome silencing (Lee, 2012; Wutz, 2011). Another well-­‐studied lncRNA is HOTAIR (HOX antisense intergenic RNA), localized in chromosome 12 within the homeobox C (HOXC) gene cluster. HOTAIR transcript is about 2.2 Kbs, it is co-­‐
expressed with HOXC genes and participates in the downregulation of various genes all along the genome, such as the HOXD gene cluster present in chromosome 2 (Tsai et al., 2010). In vitro studies demonstrated that HOTAIR 46 Introduction interacts with various chromatin-­‐modifying enzymes, like PRC2 (Greer and Shi, 2012; Tsai et al., 2010). These are just two examples of lncRNA regulation, but there are some other examples and presumably many more to be discovered. In fact, the ENCODE project revealed that in average there are approximately 10 transcripts overlapping any previously annotated gene (Djebali et al., 2012). Chromatin conformation, entering in the third-­dimension The spatial organization of the genome is closely linked to its biological function, yet the understanding of higher order genomic structure is still poorly understood. In the nucleus of eukaryotic cells, interphase chromosomes occupy distinct chromosome territories (Cremer & Cremer, 2010; Pennisi, 2011). Nowadays it is known that the location of these territories has functional implications. For example, it is known that the DNA bound to the nuclear lamina is completely silenced and this type of chromatin has been called “black chromatin” in D.melanogaster (Steensel, 2011). It is also described that these domains are highly dynamic during mammals cell differentiation (Peric-­‐Hupkes et al., 2010). In the last decade the field has gained important insights thanks to the development of chromosome conformation capture (3C) and derivate techniques (see materials and methods) to identify chromatin interactions. These techniques allowed the development of three-­‐dimensional (3D) chromosome structure models (Dekker et al., 2013; Dekker, 2006). This new technology confirmed what was previously observed using less resolutive techniques; that chromatin was organized into nuclear domains, perhaps with different functions and regulation events taking place on them. A recent study indicates that only a small fraction of the boundaries show clear differences between two cell types, suggesting that most structural domains are preserved among different cell types (Dixon et al., 2012). Noteworthy, most of the boundaries appear to be shared across evolution (53.8% of human boundaries are conserved in mouse and 75.9% of mouse boundaries are conserved in humans). CTCF creates boundaries between topologically associating domains in chromosomes and facilitates interactions between transcription regulatory sequences and promoters. Thus, CTCF links the architecture of the genome to its 47 Introduction function (Ong & Corces, 2014). Most of the topological boundaries are enriched for CTCF binding, whereas boundaries constitute just the 15% of the whole CTCF binding, which indicates that something else is needed to form a boundary. It has been shown that boundaries are enriched near house keeping genes and also in some repetitive sequences such as Alus (Dekker et al., 2013; Dixon et al., 2012). Single cell Hi-­‐C (whole genome based 3C technique) shows that individual chromosomes maintain domain organization at the megabase scale, but show variable cell-­‐to-­‐cell chromosome structures at larger scales. Despite this structural stochasticity, localization of active gene domains to boundaries of chromosome territories is a hallmark of chromosomal conformation (Nagano et al., 2013). There are many long range interactions within chromosomal territories and also between them. One of the most studied types of interaction are those associated with elements that modulate gene expression from a relative far away distance (in a DNA linear manner), known as enhancers (Pennacchio et al., 2013). The ENCODE project revealed more than 1,000 long-­‐range interactions between gene promoters and distal sites (Sanyal et al., 2012). In the same work they found that looping interactions and long-­‐range interactions were enriched for CTCF and also for histone modifications such as H3K4me1, H3K4me2, H3K4me3, H3K9ac and H3K27ac (typically in active enhancers), but were not enriched or significantly depleted for H3K27me3, a mark typically found at inactive or closed chromatin (Sanyal et al., 2012). Nowadays we know that there are different combinations of histone PTMs that define an active enhancer (Ernst et al., 2011), although the classical features are presence of H3K27ac and H3K4me1. Furthermore, active enhancers can express bidirectional transcripts known as enhancer RNAs (eRNAs) (Kim et al., 2010), nevertheless the biological function of eRNAs remains unclear and there is no consensus about whether they mediate specific biochemical functions (Shlyueva et al., 2014). 48 c eCi a eti c
n
ci u
ei
C tc i u
i I dl e tc tc qóó8 n
dtu te
Ct( c i Cu d i di c ei
en e en C
en l
1 enR sl tc ( i d(
ei
Cx
Ci c F
Ci c
i di C e d
l Ct
%c
c
ei en
CC l
c I t(
l etu adal 0 1tI c et
C n t d 2 ntl ei c 1i l ezeC c l d eti c d
u ensd eti c W nCi u etc C u i
Cteni C ) W u i c I i en Cl % c
l u
C c
1tI c et
de C eti c l ) 1d tc ni R eR i t c et dI c i es1 l
c
c
de C eti c l en e R i ad
tc I ei
i en I c et x l 1C ( ti al ds
C u tc ei 1t l i u i t t eti c l W
s
i I dl e tc Wqóó8%
0 e tl c i R R ddoc i R c
t C c e 1n c i es1 l tc C l 1i c l
de C eti c l
c
c
d 1Ci 1i l
C 1 Cei tC i I c et
C u adet1d u i d ad C 1 enR sl d
de C eti c l 0 1tI c et
Ctl
1Ci I C l l ti c u i
c i c i tc I
dtc I
i u 1d ) l x i ds i u
Cancer epigenomics Sandoval and Esteller 53
l x tI aC qU%x c i ( d F
l e dd CWL8qL%
0
Figure 1
NORMAL CELL
CANCER CELL
DNA methyl
modifying complexes
DNA methyl
modifying complexes
Tumor supressor genes
Promoter specific
hypermethylation
DNA methylation
DNA
methylation
Tissue specific genes
Onco-genes
repetitive regions
Global hypomethylation
CHROMATIN
STRUCTURE
Histone
modifications
NORMAL GENE
REGULATION
Aberrant histone
modification landscape
Histone Modifying
complexes
AcK4H3
AcK9H3
AcK18H3
AcK16H4
dimeK4H3
trimeK4H3
trimeK9H3
trimeK27H3
trimeK20H4
ABERRANT
CHROMATIN
STRUCTURE
Histone
modifications
Histone Modifying
complexes
GENE
MISREGULATION
epimiRNAs
Onco-miRNAs
miRNA
miRNA
Tumor suppresor miRNAs
miRNA
Machinery
Aberrant global
miRNA expression
miRNA
Machinery
Current Opinion in Genetics & Development
f a p3 d oe o c n er e c r el
on a n er f a r er e r h
a r
aa r n
Global
oncogenesis.
In conjunction
an aberrant
for the
c nndepiction
ar eaofnepigenomicaalterations
r n c during
r n
nea y
l withn accumulation
i o n er mof genetic
ner lesions,
l e there niser
r pattern
l
h
different epigenetic effectors: DNA methylation, histone modifications, and miRNAs. In normal cells, the interplay between the epigenetic factors and
c
ae
f
ael
r
eu
o
|
n
o
o
am
2.
p2h
the chromatin structure leads to a tuned gene regulation. However, in cancer cells tumor suppressor genes promoters become hypermethylated and
with an altered global pattern of histone modifications resulting in aberrant gene silencing. Moreover, global hypomethylation leads to chromosome
instability and fragility. Epigenetic changes, including DNA methylation and histone modifications are responsible for abnormal mRNA and miRNA
expression producing altered activation of oncogenes and silencing of tumor suppressor genes.
n Ci d i s
u ensd eti c tc i c i I c l tl n l
Cl Wi Ctc I
e td
1t eaC
i ae en
between
aberrant
and cancer.
en
C I adepigenomics
eti c i en
c InC the
I cfield
i uof
genomics, many laboratories around the world are working
together as part of the International Cancer Genome Conl ea t
d l l tc e c l ds ae en c ol ei
sortium, or collaborating in the Cancer Genome Project and
the Cancer Genome Atlas research program.
l a l e c et d ( c
ln (
cu
The combination and integration of epigenomics, genomics and all the other ‘omics’, including transcriptomic,
proteomic, and metabolomic aspects, will be essential to
maintain the increasingly rapid progress towards a full
understanding of the underlying molecular mechanisms
that govern the initiation and development of cancer
processes. Furthermore, these cancer signatures will help
us identify new potential prognostic and detection tools
and, eventually, to develop effective clinical therapies.
However, although the large-scale studies are invaluable,
c l ea t
i C u i C en c L8
t C c e Ci d l i u ensd eti c tc
0 near
n future
i en Cto analyze
1tI c the
et vast
u amount
Col nof(information
c
generated by powerful epigenomic platforms such as the
450 K DNA methylation microarray [65].
en
( di 1u c e i c R e nc t9a l W
Acknowledgements
aCtcSupported
I en bydGrants
l e SAF2011-22803,
0 the European Research Council
(ERC) Advanced Grant EPINORC and Fondo de Investigaciones Sanitarias
Grant PI08-1345. JS is funded by the Juan de la Cierva Research Program
(MICINN). ME is an ICREA Research Professor.
References and recommended reading
Papers of particular interest, published within the period of review,
have been highlighted as:
! of special interest
!! of outstanding interest
1.
Eó
Kinzler KW, Vogelstein B: Lessons from hereditary colorectal
cancer. Cell 1996, 87:159-170.
Introduction DNA methylation in cancer In breast cancer it has been demonstrated that hypermethylated genes are the effectors of the earliest steps in cancer formation. Aberrant methylation of genes is associated with immortalization as well as with subsequent steps in the malignant progression of these cells (Novak et al., 2009). It has been also proposed that the aberrant methylation of specific genes might lock stem cells in an undifferentiated state, enhancing them to malignant transformation (Widschwendter et al., 2007). This aberrant methylation is not occurring exclusively in the tumor, Shen et al. observed that the 50% of normal colon tissue adjacent to a tumor carried detectable hypermethylation within the promoter of the DNA repair gene MGMT (O6-­‐methylguanine methyltransferase) when the tumor was methylated. On the other hand, it only occured in 12% of normal tissue from cancer-­‐free control patients and only 6% of the tumor-­‐adjacent normal tissue carried methylated MGMT when the tumor was unmethylated (Shen et al., 2005). A similar phenomenon has also been described regarding hypomethylation of CDH3 and hypermethylation of EVL and p14 in normal colon epithelium adjacent to the tumor (Grady et al., 2008; Milicic et al., 2008; Shen et al., 2003). It is worth to note the genomic-­‐epigenomic cross-­‐talk observed in DNA repair methylation in colorectal cancer. It has been described that hypermethylation of MGMT and MLH1 (DNA repair genes) in the adjacent normal mucosa correlates with tumor alterations such as KRAS mutations or microsatellite instability, respectively (Hiraoka et al., 2010; Ramírez et al., 2008). The epigenetic silencing of repair genes in the normal tissue would create a scenario where mutation rate and other genetic alterations would be enhanced; these studies bring to light the importance of the genetic-­‐epigentic cross-­‐talk equilibrium. There are three main DNA methylation alterations in cancer: hypermethylation, hypomethylation and loss of imprinting. The most widely studied epigenetic deregulation in cancer is DNA hypermethylation. There are many studies that point out aberrant DNA methylation within the colorectal cancer genome, although only a subset of these genes are likely to be important in the development of the cancer. 50 c eCi a eti c
iu
i en u i l e C 9a c e I c
ns1 Cu ensd eti c l n (
tc tet eti c c 1Ci I C l l ti c i en 1i ds1 ei
Cds ns1 Cu ensd e
REVIEWS
x i F
i
I c l tc da
c dtc o
ei en
C tc i u
x tI aC q7%
0 ) u 1d l i en
EW
W
: 7W
L c
q
C sWL8qq%
0 i e dsWen u i l e l a l e c et dtc C l tc I c u ensd eti c
aCl tc en l e 1 Ci u c i Cu d i di c u a i l ei
Normal colonic
epithelium
Aberrant crypt focus
SLC5A8
MINT1*
MINT31*
SFRP1
SFRP2
CDH13
CRBP1
RUNX3
c i u x tu
Polyp/adenoma
p14 ARF
HLTF
ITGA4
CDKN2A/p16
CDH1
ESR1
WL88/%
0
Adenocarcinoma
TIMP3
CXCL12
ID4
IRF8
Methylated genes
Figure 2 | Commonly methylated genes (and loci) identified at the histological steps of the colorectal cancer polyp
f a pI d sequence.
el l er oiSpecific
l n gene
i o nand locirhypermethylation
v r oe ,has been
r n observed
nn
neoe steps
o of
nthe
c cancer
e n
adenocarcinoma
in the different
eoea n osequence,
r a which
c eoi cimplicates
r e these
a rgenes
el in either
t f the
r initiation
h c or progression
r
i ccolorectal
al n icancer.
o n erThe genes
r
progression
of
between
epithelium
as well
those
elisted
au
r normal
n
a r n nand
c aberrant
e n crypt
r focus
a c ae
a as
er
genes
t f r listed
h between
c ae f aberrant
aelcrypt focus
e | and
a im
polyp/adenoma
might be involved in the initiation of colorectal cancer. Those genes listed between polyp/adenoma and
2.
pph
adenocarcinoma could have a role in the progression and metastasis of colorectal cancer. Note that only the genes (and loci)
that have consistently been found to be hypermethylated at specific points in the progression sequence have been included.
I tl c tu 1i Ce c eCtl o ei Ctc en
( di 1u c ei i di C e d c C0 n I tc I
*MINT1 and MINT31 are not genes; rather they are loci that have been found to be ‘methylated in tumor’ (aka MINT).
i i di c t u a i l
tl 1d sl I zC d e
class of genes that seem to be more
lno
1 clear
t tfunctionalns1
Cu ensd eti c x l l F
commonly affected by DNA methylation in the adenoma
I di
dns1i u ensd eti c l R dd l di al z
DNA methylation and metastatic CRC
nar WL888%
0 nal Wtc I tc I c i en C
The progression of colonic adenomas to adenostep of colon cancer development compared with other
carcinomas is believed to be driven by genetic alteratC au l e c l WR
c tc
CC c e u ensd
e I c l tc en c i Cu d u a i l W
steps in the progression sequence.129 Despite the lack
tions, such as mutations in TP53, but is also probably
For0
definitive
proposed
Rofnt
n R i proof,
ad tcit has
C been
l en
Ctl othat
i aberrant
l a Ctc I a consequence
i di C e dof epigenetic
c Cx alterations
l nt (Figure 2).
WL88q%
example, the aberrant methylation of CXCL12, a chemomethylation of DNA repair genes, such as MGMT and
kine
ligand,
in
CRC
MLH1,
adenomas
n iin( colon
C ) 1C
l l ti cmight
i promote thel progresn l
c 1Ci
1i occurs
l
eihuman
) 1d
tc andencan promote
CC the
ce
metastatic behavior of colon cancer cell lines.135 In addision of adenomas to cancer by creating a permissive state
mostIcommon
ufor G>A
ensdmutations—the
eti c i aCCtc
tc enmutation
c seen
C I ction,
i u the0 frequency
i en tcof Ctumors
l carrying
) 1Ccertain
l l ti cmethylc
in KRAS in the case of methylated MGMT, and a perated genes, like TIMP3, ID4 and IRF8, is higher in colon
missive
satellite
in thelcase
metastases
adenomas,
thatc
tc
C state
l for micro
et( tes
i instability
en
W iofu 1 cancer
C and
R ten
c i Cu than
d etl
l a l Wsuggesting
n (
methylated MLH1.131,132
the inactivation of these genes by methylation provides
128,136
the detection
In addition
to specificW
C In
1i addition
Ce tctonau
c c ofClmethylated
tc da tcgenes
I i di ca clonal
c growth
C x l advantage.
l
WqóóZ,
i l ni
in adenomas, CIMP can be detected in advanced
aberrantly methylated genes being present in advanced
tubular 0
adenomas but is not commonly observed in
CRCs, hypomethylated LINE-1 elements are found to
L88ó%
early tubular adenomas.128,133 Interestingly, Kwon et al.
increase linearly in more progressed colon neoplasms
detected
of
e n l CIMP adenomas
c
u iincal collection
eC e
tcserrated
i enW (Figure 1).
dd dtc130,137,138
l However,
c
c in
tugeneral,
d u iti seems
dl thatenthee
adenomas but not in tubular adenomas, which suggests
aberrant DNA methylation of genes is predominantly
serrated
be theqprecursor
polyps Z involved
withathe early events
ithat
( C
) 1C adenomas
l l ti c i could
c
c tc
CC cofecolon
u cancer
ensd formation
eti c i for CIMP CRCs.133 It is not clear whether the detecand less with progression events.
ction
i Cuof CIMP
dds ac
u inensd
e phase
1 oftlCRC
d c formal Wl tu td C ei R n e tl i l C( tc nau c i di c
late
the polyp
tion reflects a technical issue regarding the number of
DNA methylation epigenotypes of CRC
C x ini the
l nistudies of methylated
WL88ó, genesnu
t e Epigenetic
WL88U%
0 in CRC
i R is(manifested
CWen Cin a variety
tl l etd
locic assessed
in adeinstability
ofd
nomas, or demonstrates that CIMP occurs late in the
ways, including hypermethylation of gene promoters that
i c eCi ( Cl sequence.
s tc en128 This
t dis W
i en
t lCpG
l niislands
R and
englobal
e tcDNA
C hypomethylation.
l
q
polyp→cancer
contain
an issue al
because
it is C l ea
Essentially all CRCs have at least some aberrantly methylwidely believed that CIMP tumors might arise from ser) 1Cadenomas,
l l ti c Win cwhichl acasel it would
9a cbe
e reasonable
ns1 Cu to ensd
etigenes,
c Wtlandcthere
i e lseems
tu 1d
l i t e distribuR ten
ated
tos be al unimodal
rated
tion of colon cancers when the proportion of methylated
assume that CIMP would be detected in early adenomas.
genes to cancer genome is assessed. In 1999, Toyota and
To date, methylated genes that distinguish sessile serIssa observed that some CRCs have a high frequency of
rated adenomas from hyperplastic polyps (CDX2, MLH1,
:q
methylated genes, and they put forth the concept that
and TLR2) have been identified, but there has yet to be
these tumors have a unique molecular patho genesis,
consistent reporting of a subset of early adenomas with
naming them CIMP cancers.93 This concept was met with
the CIMP signature of methylated genes.134
c eCi a eti c
c
C x nCdt n
WL88/%
0 c c i e c tc C l
l u c eti c
et( tes i en
en e en ac
qW
c i l sdu enti c tc
Z
ei
c
e tl
et( tes0
I c tl C I ad e
s
lW
Z WR nt n eC c l C u ensdI Ci a1l Ci u
tc
l nti c x
u ce d
l W aeC en C u tl tC e
1C ( ti al dsWen u ensd eti c l e e i tc da tc I
u tc e c c
eWte u tI ne
ten C
q%
0
x
tc e c c
Z
c
u ensd eti c i
C 1dt eti c WR n c n u tu ensd e
tl
i ac
s
z
Z %i C
aCl
aCtc I
q R ten ntI n
tc tes0
c c i Cu d i c teti c l u i l e i en
c
ic
C
ddl
1 tl d c l C u tc ac u ensd e WR n C l tc
CC c e ns1 Cu ensd eti c i l 1 t t
c l eti c i en
nCi u etc
c
1
tl d c l C l adel tc en
en l td c tc I i u adet1d I c l Wtc da tc I
eau i Cl a11C l l i CI c l x tI aC qó%
0
Normal colon epithelium
Exon 1
Exon 2
Exon 3
Repetitive sequences
i.e. LINE-1
Genomic stability
Open chromatin structure
Colorectal cancer
Exon 1
Exon 2
Exon 3
Repetitive sequences
i.e. LINE-1
Closed chromatin structure
Genomic instability
Methylated CpG site
Unmethylated CpG site
Figure 1 | CpG island DNA hypermethylation and global DNA hypomethylation in colorectal cancer as compared with
f a pT
d cepithelium.
o r Unmethylated
i c alCpGnislands
i o n er
r theopromoter
e o region
i cofelgenes
n iare
o ncorrelated
er r eo
eaannopen
o r a
normal
colonic
within
with
el
c a structure
s n r eal
o eoer whereas
c n methylated
o f l h r CpG
l nislands
i o n are correlated
c o r with
s a n
r n c ael
enchromatin
a a er e
chromatin
(euchromatin),
condensed,
closed
structure
silencing.
colonic
r a(heterochromatin)
eaa o n s and
n transcriptional
r ec r
ael n r Normal
naf n
f a vepithelium
f ael generally
n r ,ms hasaunmethylated
l n i oCpG
n
c
CpG islands
islands
o r in the
a promoter
eaa o regions
n
s ofngenes, whereas
er r aberrant
m oe hypermethylation
ael n of
r promoter
naf nf associated
a v n ae
ael nisra, r
hypermethylation
seen in colorectal cancers, global
nahallmark
r a cofnneoplasms.
er o o r In raddition
h c to
aethef aberrant
aellocal e
| a i m2. pph
hypomethylation at LINE-1 sequences is also observed, which has been shown to be associated with genomic instability.
Interestingly, an inverse association exists between local CpG island hypermethylation and global LINE-1 hypomethylation
td c tc Iprogress.
s
u ensd eti c
c
i u 1dtl n
s t C c e
as ccolonicl neoplasms
urepression
n c tlinvolves
u l 2 proteinsu commonly
ensd etioverexpressed
c
i ad
tu 1 tC en tc tc I i eC c l Ct1eti c
DNA methylation and cancer
cancer, such as EZH2, and cooperates with aberrant
in
ei Cl Wl a n l
zLW
WL W
c
DNA methylation to silence gene expression by regulat-
Although the importance of gene mutations in the
z3 x u 1 c Ci
WL888, i u
pathogenesis of cancer is widely accepted, the role of
ing chromatin compaction.49 A close association exists
in cancer has been controversial
Fbetween
i i methylated
u c Wqóó8,
cI
WL88q, epi
i genetic
F Calterations
sWL8qq%
0 aCen Cu i C W
until the past decade. The first epigenetic alteration
CpG islands and histones containnearly three decades ago by
repressive post-translational modifications (such
uing
ensd
eti c u s C Cate c psu l en e in
c cancer
i u 1was ereported
en nCi
u etc c C 1C l l
Feinberg and Vogelstein, who showed extensive global
as H3K9me3 and H3K27me3). The establishment of
loss of 5-methylcytosine content in colon cancers comI these
c histone
) 1C modifications
l l ti c x (which
CF cause
CI uchromatin
c WL88ó%
0 ensdz tc tc I 1Ci e tc l C tc ( i d(
pared with the normal colon.54 The global loss of DNA
condensation) is thought to be mediated by polycomb50,51
predominantly affects CpG dinucleotides
tcrepressive
en l tdcomplexes.
c tc I i enThis association
nCi u etchas sraised
tc tcmethylation
I R ten ntI
n tc tes ei u ensd e
found in repetitive sequences of DNA, such as LINE-1
the possibility that abnormal activity of PRC2 might be the
and satellite repeats. This global hypomethylation can
mechanism for inducing aberrant DNA methylation
be found in the colon in an age-dependent fashion as
in cancer.31
steps of the polyp→cancer progres: LFeinberg and colleagues have extended the concept well as in the early
sion sequence.55–57 Thus, it seems that the global hypoof CpG islands to ‘CpG island shores’, which are also
abnormally methylated in cancer. CpG island shores
methylation of DNA is an early event in the formation of
refer to areas of less dense CpG dinucleotides within
CRC. A study in Dnmt1+/– mice demonstrated that DNA
52
REVIEWS
Introduction and start a cascade of events altering chromatin structure by recruiting other proteins. These proteins include a wide variety of epigenetic remodelers such as histone deacetylases, histone methylases and chromatin remodeling complexes, thereby condensing the chromatin and blocking access of transcription factors to the promoter region (Bird, 2002). Some examples of these methyl binding proteins are MeCP2, MBD1, MBD2 and MBD4; the zinc finger proteins Kaiso, ZBTB4 and ZBTB38; and the SET-­‐and RING-­‐finger associated proteins, UHRF1 and UHRF2, among others (Tsai & Baylin, 2011; Engeland et al., 2011). The first epigenetic alteration in cancer was observed in 1983 by two independent studies showing extensive global loss of 5-­‐methyl cytosine content in colon cancer (Feinberg & Vogelstein, 1983; Gama-­‐Sosa & Slagel, 1983). This global demethylation affects predominantly repetitive elements, such as LINEs and satellite repeats (Figure 19). DNA hypomethylation is an early event in the development of colorectal cancer and is age-­‐related (as hypermethylation) (Issa & Ahuja, 2000; Suzuki et al., 2006). A study in Dnmt1+/– mice demonstrated that DNA hypomethylation might contribute to cancer formation by allowing or inducing genomic instability (Laird et al., 1995). Later on, other studies have added further insights into the role of DNA hypomethylation in tumorigenesis (Eden et al., 2003; Gaudet et al., 2003). Loss of imprinting is defined as the loss of parental allele-­‐specific monoallelic expression of genes due to aberrant hypomethylation profiles at one of the two parental alleles. For example, loss of imprinting of IGF2 has been associated with an increased risk of colorectal cancer among other cancers (Lim & Maher, 2010). Histones in cancer Aberrant patterns of histone posttranslational modifications (PTM) are a hallmark of cancer, although it is still poorly understood whether they are cause or consequence of the malignant transformation. There are common alterations in the histone modifications patterns in cancer, for instance a global reduction of H4K20me3 and H4K14ac in repetitive sequences (Fraga et al., 2005), H3K27me3 targeting loci for de novo DNA methylation in cancer cells (Vire et al., 2006; Widschwendter et al., 2007) and loss of the active enhancer mark H3K27ac in cancer (Aran & Hellman, 2013; Ferrari et al., 2014). Some specific histone PTM 53 c eCi a eti c
de C eti c l n (
c
c i l C(
CR n C I di
Z Eu L c
tl Ca1eti c i en
x dtI l i c
1tI c et u
c
C es1 Wl a n l tc 1Ci l e e
d
de C
1i l teti c i WL88: %
0
ntc Ct l W ten C s u ae eti c W d eti c i C de C
c s i en tC i u 1i c c el Wtl oc i R c ei 1Ci ( i o
) 1C l l ti c 1 ee Cc l en e I t(
de C eti c i en
1 t t
1 Cet ad C
dntl ei c u i t t eti c 1 ee Cc l C (
L q7
) 1C l l ti c i tc
l
Ctl
ei
dd es1t d
Ci l l en I c i u
1tu ae eti c l n (
c dtc o
c
C n C e Ctl et l Wl a n l
x i CgI a pz C
ei
CC c e I c
l F
Ce tc es1 l i c
l e dd CWL8qq%
0
Cx p y
W
L88/%
0
c et d l ti c l tc nCi u etc u i t t Cl
dc l
1 c i e i c ds tu 1ds
1Ci ( t
n (
c
I di
d de C eti c l tc en
al et( Ci d i C en l 1Ci e tc l tc
1i e c et d e CI el i C en C 1 aet tc e C( c eti c 0
dC
s
c
( di 1
c au
c
1tI c et
C ae dl i
C i tc nt tei Cl
I tc l e nCi u etc C I ad ei Cl x tI aC L8%x
Rli c
F i ap Ct l WL8qL%0
Figure
f a 1.2.Epigenetic
d
Inhibitors
l as Cancer
c n Therapies
n c ae ea c r n af
u oec l r n r n
f aa r n n nf This schematic depicts the process for epigenetic drug development and the current status of various epigenetic therapies. Candidate small molecules are first
etested
u ain ef
c r cell
n lines
n foraspecificity
c hand phenotypic
c n ael
s ermay,
| in the
effirst
g instance,
a m
2. p2h
vitro in malignant
response. These
assess
the inhibition of proliferation, induction of
apoptosis, or cell-cycle arrest. These phenotypic assays are often coupled to genomic and proteomic methods to identify potential molecular mechanisms for the
observed response. Inhibitors that demonstrate potential in vitro are then tested in vivo in animal models of cancer to ascertain whether they may provide
therapeutic benefit in terms of survival. Animal studies also provide valuable information regarding the toxicity and pharmacokinetic properties of the drug. Based
on these preclinical studies, candidate molecules may be taken forward into the clinical setting. When new drugs prove beneficial in well-conducted clinical trials,
they are approved for routine clinical use by regulatory authorities such as the FDA. KAT, histone lysine acetyltransferase; KMT, histone lysine methyltransferase;
RMT, histone arginine methyltransferase; and PARP, poly ADP ribose polymerase.
iu
i en
n (
dC
CaI l
s
1t e
tc tI aC L8 xe CI etc I
c 11Ci (
s en
0 ntl l a
lW
l
c
L%
W
l l tl en C d ) i en tc e C l e
en
e 1tIin acvastetarray CaI
l n has
( provided
1Ci aacatalogtcof enregulators
t dhave
i (raised
C enintriguing
d l ecorrelations between
0 i Rcancer( CW
sequencing
of cancers
recurrent somatic mutations in numerous epigenetic regulators
(Forbes
al., 2011;
Stratton
en
C etdtes
tl en
e enet al.,
t 2009).
d i A central
1tI tenet
c etin
analyzing these cancer genomes is the identification of ‘‘driver’’
mutations (causally implicated in the process of oncogenesis). A
key feature of driver mutations is that they are recurrently found
in a variety of cancers, and/or they are often present at a high
prevalence in a specific tumor type. We will mostly concentrate
our discussions on suspected or proven driver mutations in
epigenetic regulators.
For instance, malignancies such as follicular lymphoma
contain recurrent mutations of the histone methyltransferase
MLL2 in close to 90% of cases (Morin et al., 2011). Similarly,
UTX, a histone demethylase, is mutated in up to 12 histologically distinct cancers (van Haaften et al., 2009). Compilation of
the epigenetic regulators mutated in cancer highlights histone
acetylation and methylation as the most widely affected epigenetic pathways (Figures 3 and 4). These and other pathways
that are affected to a lesser extent will be described in the
following sections.
Deep sequencing technologies aimed at mapping chromatin
modifications have also begun to shed some light on the origins
of epigenetic abnormalities in cancer. Cross-referencing of
DNA methylation profiles in human cancers with ChIP-Seq
data for histone modifications and the binding of chromatin
tc l e c
l tc nt tei Cl WR nt n n ( ( Cs dteed c psu
C t
CaI x tc a t F
etc I 1tI c et
CCnsenu t
c
el tc en c i Cu d etl l a 0 i C
c
c psu l tc
14 Cell 150, July 6, 2012 ª2012 Elsevier Inc.
c
l 1 t t tesWi ad n (
( i u tetc I ei
i c u CCi R
c aCi di I t d ei ) t tesW 1 c tc I i c
dt tWL88/%
0 c tI aC Lq C C d e
c eteau i Cen C 1t l 0
:E
el C c I tc I Ci u etI a Wc al
ei ) t tesWC ( Cl t d
en
CaI l t
tl i ( Cs tl l etddtc tel tc c s c
C c
(i t
CaI
u i C l ea t l
l ( C lt
ei
associated DNA hypermethylation and genes marked with
‘‘bivalent’’ histone modifications in multipotent cells (Easwaran
et al., 2012; Ohm et al., 2007). These bivalent genes are marked
by active (H3K4me3) and repressive (H3K27me3) histone modifications (Bernstein et al., 2006) and appear to identify transcriptionally poised genes that are integral to development and
lineage commitment. Interestingly, many of these genes are
targeted for DNA methylation in cancer. Equally intriguing are
recent comparisons between malignant and normal tissues
from the same individuals. These data demonstrate broad
domains within the malignant cells that contain significant alterations in DNA methylation. These regions appear to correlate
with late-replicating regions of the genome associated with the
nuclear lamina (Berman et al., 2012). Although there remains little
mechanistic insight into how and why these regions of the
genome are vulnerable to epigenetic alterations in cancer, these
studies highlight the means by which global sequencing platforms have started to uncover avenues for further investigation.
Genetic lesions in chromatin modifiers and global alterations in
the epigenetic landscape not only imply a causative role for
these proteins in cancer but also provide potential targets for
therapeutic intervention. A number of small-molecule inhibitors
have already been developed against chromatin regulators
(Figure 1). These are at various stages of development, and three
en u tc u ae eti c l
CW c 1 Cn 1l CaI e CI e d 1Ci e tc l i C
and JAK2) have already
ood and Drug Administrat that the interest in epigediscovery had been high
reality is that the field of
held back due to concerns
he drugs and their targets.
(against HDACs) have little
anism of action remains
6).
netic cancer therapies may
w level of support following
hibitors against BRD4, an
ein (Dawson et al., 2011;
s et al., 2010; Mertz et al.,
lar mechanisms governing
have also been largely
This process is pivotal for
e inhibitors
methylated
onthe
theclinic.
side
into
inelistresidues.
Methylation,
g
of genetic
lesions in
tion,
not have
alter now
the
fact does
that we
nes
may be mono-, di-, or
herapies.
es may be symmetrically
est-characterized
sites of
to
Cancer
occur on lysine residues
s of this residues
section. Although
cytosine
(5mC) in
histones
are methylated,
cribed
covalent
modifica3K27,
H3K36,characterized
H3K79, and
extensively
36,
and
H3K79)
are noted
often
hylation is primarily
hromatin,
whereas others
ive X-chromosomes,
and
with
heterochromatic
,ated
2011;
Robertson,
2005).
007).
Differentobserved
methylation
commonly
in
so
localize
differently.
For
enetic
alterations
in cancer
sccur
the within
transcriptional
start
CpG islands,
i et al., promoters.
2007), whereas
ammalian
CpG
ed with
active enhancers
role
in transcriptional
reguwhereas
monomethylation
ing malignant transformaes, trimethylation
of H3K9
son,
2005). NGS platforms
arski of
et al.,
2007).
maps
CpG
methylation.
sine
methylation
contain
a
%–10%
of normally
unmesesses
methyltransferase
me
abnormally
methylated
hDOT1L,
the enzyme
that
so
demonstrate
that CpG
enly
KATs,
thethe
histone
lysine
affects
expression
ehehighly
specificofenzymes
expression
various
ne
residues.
Cytogenetic
e a role
in malignant
transs cancer genomes,
have
mportantly,
these genomesalso
and/or
coding mutations
uncovered
intriguing
MMSET,
hin
gene EZH2,
bodiesand
andMLL
at
sequences upstream and
erted by
the MLLoffusions
ctional
relevance
these
viewed
(Krivtsov
and Armyet to be
fully deciphered,
nterest
is the
dichotomous
hey have
challenged
the
es.
EZH2 is equates
the catalytic
n invariably
with
ich
is
primarily
responsible
e studies have established
studies
s gene-expression
have high levels of
DNA
H2 as a progressive
event
uggesting
that the context
rostate isand
breast
cancer
ylation
vital
in transcripThese
011). initial studies sug. However, NGS and tares have recently identified
ous lymphoid and myeloid
muddied the waters by
mor-suppressive roles for
utations resulting in the
within the SET domain of
s with diffuse large B-cell
ctional characterization of
conferred increased cataconverting H3K27me1 to
ontention that EZH2 is an
contrast, loss-of-function
g a poor prognosis, have
alignancies (Ernst et al.,
-ALL (Ntziachristos et al.,
ting a tumor-suppressive
c eCi a eti c
enzymes. In addition to their catalytic function, many chromatin
modifiers also possess ‘‘reader’’ domains allowing them to bind
to specific regions of the genome and respond to information
conveyed by upstream signaling cascades. This is important,
as it provides two avenues for therapeutically targeting these
epigenetic regulators. The residues that line the binding pocket
of reader domains can dictate a particular preference for
specific modification states, whereas residues outside the
binding pocket contribute to determining the histone sequence
specificity. This combination allows similar reader domains to
dock at different modified residues or at the same amino acid
displaying different modification states. For example, some
methyl-lysine readers engage most efficiently with di/tri-methylated lysine (Kme2/3), whereas others prefer mono- or unmethylated lysines. Alternatively, when the same lysines are now acetylated, they bind to proteins containing bromodomains (Taverna
et al., 2007). The main modification binding pockets contained
within chromatin-associated proteins is summarized in Table 1.
Many of the proteins that modify or bind these histone modifiFigure 2. Cancer Mutations Affecting Epigenetic Regulators of DNA
cations are misregulated in cancer, and in the ensuing sections,
Methylation
weThe
will5-carbon
discussof the
mostnucleotides
extensively
histone
modificacytosine
are studied
methylated
(5mC) by
a family of
DNMTs.
One oftothese,
DNMT3A, and
is mutated
in AML, myeloproliferative
tions
in relation
oncogenesis
novel therapeutics.
diseases
and myelodysplastic
syndromes
addition to
of(MDS).
lysineInresidues
is its
Histone (MPD),
Acetylation.
The Nε-acetylation
catalytic activity, DNMT3A has a chromatin-reader motif, the PWWP domain,
a major
histone
transcription,
chromatin
which may
aid inmodification
localizing this involved
enzyme toinchromatin.
Somatically
acquired
structure,
DNA may
repair.
neutralizes
lysine’s
mutationsand
in cancer
alsoAcetylation
affect this domain.
The TET
familyposiof DNA
hydroxylases
metabolizes
5mC into several weaken
oxidative intermediates,
including
tive
charge and
may consequently
the electrostatic
5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxinteraction
between histones and negatively charged DNA. For
ylcytosine (5caC). These intermediates are likely involved in the process of
this
reason,
histone acetylation
isthree
oftenTET
associated
withare
a more
active
DNA demethylation.
Two of the
family members
mutated
in cancers,
including conformation.
AML, MPD, MDS,Consistent
and CMML. Mutation
types
are as
‘‘open’’
chromatin
with this,
ChIPfollows:
M,
missense;
F,
frameshift;
N,
nonsense;
S,
splice
site
mutation;
Seq analyses have confirmed the distribution of histone acetyla-and
T, translocation.
tion at promoters and enhancers and, in some cases, throughout
the transcribed region of active genes (Heintzman et al., 2007;
Wang et al., 2008). Importantly, lysine acetylation also serves
Three
active
DNA
methyltransferases
(DNMTs)
have been
as the
nidus
for the
binding
of various proteins
with bromodoidentified
higher eukaryotes.
DNMT1 is(PHD)
a maintenance
methylmains
and in
tandem
plant homeodomain
fingers, which
transferase
that recognizes
hemimethylated
recognize
this modification
(Taverna
et al., 2007). DNA generated
during
DNA replication
then methylates
newly synthesized
Acetylation
is highly and
dynamic
and is regulated
by the
CpG dinucleotides,
whose
on families,
the parental
competing
activities of
two partners
enzymatic
the strand
histoneare
already
methylated (Li et
al., 1992).
Conversely,
DNMT3a and
lysine
acetyltransferases
(KATs)
and the
histone deacetylases
DNMT3b,
although
also
capable
of methylating
(HDACs).
There
are two
major
classes
of KATs: (1) hemimethylated
type-B, which
function primarily
as de and
novomodify
methyltransferases
estabareDNA,
predominantly
cytoplasmic
free histones, to
and
(2)
lish DNA
methylation
during
embryogenesis
al., 1999).
type-A,
which
are primarily
nuclear
and can be (Okano
broadlyet
classified
DNA
provides
a platformfamilies.
for several methyl-binding
into
themethylation
GNAT, MYST,
and CBP/p300
proteins.
These
include
MBD1,
MBD2,
MBD3,
and MeCP2.
KATs were
the first
enzymes
shown
to modify
histones.
The
These in turn
function
to recruit
histone-modifying
enzymes
importance
of these
findings
to cancer
was immediately
to coordinate
chromatin-templated
(Klose
and
apparent,
as onethe
of these
enzymes, CBP, processes
was identified
by its
Bird,to
2006).
ability
bind the transforming portion of the viral oncoprotein
in DNA methyltransferases
and that
MBD
E1A Although
(Bannistermutations
and Kouzarides,
1996). It is now clear
proteins
been
known
contribute
to developmental
many,
if nothave
most,long
of the
KATs
have to
been
implicated
in neoplastic
abnormalities and
(Robertson,
2005),
we have
only recently
become
transformation,
a number
of viral
oncoproteins
are known
of somatic
mutations
of these
key genes
in human
maligto aware
associate
with them.
There are
numerous
examples
of recurnancies
(Figure 2).
Recent sequencing
of cancer[Wang
genomes
has
rent
chromosomal
translocations
(e.g., MLL-CBP
et al.,
identified
recurrent [Huntly
mutations
in 2004])
DNMT3A
in up mutations
to 25% of
2005]
and MOZ-TIF2
et al.,
or coding
patients
with acute
leukemia
(AML) (Ley
2010).
(e.g.,
p300/CBP
[Iyer myeloid
et al., 2004;
Pasqualucci
et et
al.,al.,
2011])
Importantly,
these
invariably
and
involving
various
KATsmutations
in a broadare
range
of solid heterozygous
and hematologaremalignancies
predicted to(Figure
disrupt3).
theFurthermore,
catalytic activity
of expression
the enzyme.
ical
altered
Moreover,
theirofpresence
to been
impact
prognosis
(Patel
levels
of several
the KATsappears
have also
noted
in a range
al., 2012).
However,and
at present,
the mechanisms
by which
of etcancers
(Avvakumov
Côté, 2007;
Iyer et al., 2004).
In
A
C
B
D
Figure 3. Cancer Mutations Affecting Epigenetic Regulators
Involved in Histone Acetylation
These tables provide somatic cancer-associated mutations identified in
histone acetyltransferases and proteins that contain bromodomains (which
recognize and bind acetylated histones). Several histone acetyltransferases
possess chromatin-reader motifs and, thus, mutations in the proteins may
alter both their catalytic activities as well as the ability of these proteins to
scaffold multiprotein complexes to chromatin. Interestingly, sequencing of
cancer genomes to date has not identified any recurrent somatic mutations in
histone deacetylase enzymes. Abbreviations for the cancers are as follows:
AML, acute myeloid leukemia; ALL, acute lymphoid leukemia; B-NHL, B-cell
non-Hodgkin’s lymphoma; DLBCL, diffuse large B-cell lymphoma; and TCC,
transitional cell carcinoma of the urinary bladder. Mutation types are as
follows: M, missense; F, frameshift; N, nonsense; S, splice site mutation; T,
translocation; and D, deletion.
some cases, such as the leukemia-associated fusion gene
MOZ-TIF2, we know a great deal about the cellular consequences of this translocation involving a MYST family member.
MOZ-TIF2 is sufficient to recapitulate an aggressive leukemia
in murine models; it can confer stem cell properties and reactivate a self-renewal program when introduced into committed
Figure 5. Cancer
Mutations
Epigenetic
Regulators
hematopoietic
progenitors,
andAffecting
much of this
oncogenic
potential
Involved
in Histone
is
dependent
on itsPhosphorylation
inherent and recruited KAT activity as well as
Recurrent mutations in signaling kinases are one of the most frequent oncoits
ability to bind to nucleosomes (Deguchi et al., 2003; Huntly
genic events found in cancer. Some of these kinases signal directly to chroet
al., Activating
2004). and inactivating mutations of these have been noted in
matin.
a range
of malignancies.
Thus far,
contains with
a BRCT
domain,to
is
Despite
these insights,
theBRCA1,
great which
conundrum
regards
the only potential phosphochromatin reader recurrently mutated in cancer. It
unraveling
the molecular mechanisms by which histone acetylshould be noted, however, that BRCA1 binding to modified histones via its
transferases
contribute
to firmly
malignant
transformation
has been
BRCT domain has
not yet been
established.
As our knowledge
about
histone phosphatases
and phosphohistone
increases,
we are likelyon
to
dissecting
the contribution
of alteredbinders
patterns
in acetylation
find mutations
in many of these
proteins
that contribute
to oncogenesis.
histone
and nonhistone
proteins.
Although
it is clear
that global
Abbreviations for the cancers are as follows: AML, acute myeloid leukemia;
histone
patterns
arechronic
perturbed
cancers
(Fraga
ALL, acuteacetylation
lymphoid leukemia;
CML,
myeloidinleukemia;
NHL,
non-
Hodgkin’s lymphoma; MPD, myeloproliferative diseases; and T-PLL, T cell
prolymphocytic leukemia. Mutation types are as follows: M, missense; F,
S, 150,
spliceJuly
site6,mutation;
T, translocation;
and 17
D,
Cell 150, July
6, 2012 ª2012
Elsevier
Inc. 15 frameshift; N, nonsense;Cell
Figure 4. Cancer Mutations
Affecting
Epigenetic
Regulators
2012 ª2012
Elsevier Inc.
Involved
in Histone
f a 2p
d f nMethylation
n er er c r n s a n a m a a deletion.
ea a
a mn oe n er e n l f n n er r n
methyltransferases, demethylases, and
nfRecurrent
l ea smutations
a in histone
r ef
r h
c n ael
s er |
methyllysine binders have been identified in a large number of cancers. These
mutations may significantly alter the catalytic activity of the methyltransferases
or demethylases. In addition, as many of these enzymes also contain chromatin-reader motifs, they may also affect the ability of these proteins to survey
and bind epigenetic modifications. Abbreviations for the cancers are as
follows: AML, acute myeloid leukemia; ALL, acute lymphoid leukemia; B-NHL,
B-cell non-Hodgkin’s lymphoma; DLBCL, diffuse large B-cell lymphoma;
HNSCC, head and neck squamous cell carcinoma; FL, follicular lymphoma;
MDS, myelodysplastic syndromes; MPD, myeloproliferative diseases; and
TCC, transitional cell carcinoma of the urinary bladder. Mutation types are as
follows: M, missense; F, frameshift; N, nonsense; S, splice site mutation; T,
translocation; D, deletion; and PTD, partial tandem duplication.
The precise mechanisms by which gain and loss of EZH2
activity culminate in cancers are an area of active investigation.
In light of the varied roles that polycomb proteins play in
Cell 150, July 6, 2012 ª2012 Elsevier Inc. 19
ef g a
m2. p2h
replication, and transcription. Generally speaking, the specific
histone phosphorylation sites on core histones can be divided
into two broad categories: (1) those involved in transcription
regulation, and (2) those involved in chromatin condensation.
Notably, several of these histone modifications, such as
H3S10, are associated with both categories (Baek, 2011).
Kinases are the main orchestrators of signal transduction
pathways conveying extracellular cues within the cell. Altered
expression, coding mutations, and recurrent translocations
involving signaling kinases are some of the most frequent oncogenic phenomena described in cancer (Hanahan and Weinberg,
2011). Many of these kinases have established roles as signal
transducers in the cytoplasm; however, it has recently been
recognized that some kinases may also have nuclear functions,
which include the phosphorylation of histones (Baek, 2011; Bungard et al., 2010; Dawson et al., 2009) (Figure 5). One such
enzyme is the nonreceptor tyrosine kinase, JAK2, which is
frequently amplified or mutated in the hematological malignan-
::
been given a broader a
as Hodgkin’s disease an
in which this mechanism
genesis (Rui et al., 20
inhibitors against kinase
pies, it is interesting to n
Aurora inhibitors) result i
fications laid down by th
fore be considered as p
Histone phosphorylat
modification, which is re
activities of protein kina
phatases, like protein kin
serine/threonine residue
dual specificity; they a
requirement for a metalli
there is little doubt tha
chromatin biology, outsi
lation of mitosis, little is
these enzymes at chrom
in cancer (Xhemalce et a
The phosphorylation
residues may serve as
proteins. Proteins such
breaks by tethering to
(Stucki et al., 2005). Fur
of which there are seve
conserved phosphoseri
as 14-3-3z, use to bind
proteins, including 14-3human malignancies an
ing them may prove ben
Cancer Mutations in H
Two recent studies have
tions in genes encoding
variant H3.3 (H3F3A) and
in up to one-third of ped
et al., 2012; Wu et al.,
heterozygous and are c
in amino acid substitut
of histone H3 (K27M, G
they disrupt, these mut
influence on chromatin
mutation alters the abilit
ylated and acetylated. T
H3K27 have different ge
and heterochromatin; th
netic readers and are u
scriptional outcomes. S
mutations, due to their
transcription. In support
carrying the K27M and G
expression profiles, and
strated a global increase
Introduction ALDO-­‐KETO REDUCTASES IN CANCER Aldo-­‐Keto reductases (AKRs) are multifunctional enzymes that catalyze the reduction of several endogenous and xenobiotic aldehydes and ketones (Penning & Drury, 2007a). The AKR superfamily comprises 15 families containing over 150 members, which are further classified into subfamilies (Salabei et al., 2011). These enzymes utilize a wide range of substrates, among them steroid hormones, prostaglandins, retinals, lipids and sugars. Most of them prefer the use of NADPH over NADH as the reducing cofactor (Barski et al., 2008; Penning & Drury, 2007b). The AKRs are about 35–39kDA in weight and generally cytosolic and monomeric, although some, such as AKR7 can exist as a dimer and associated to membrane (Kelly et al., 2002; Kozma et al., 2002). Some of AKR proteins do not have or have a very poor enzymatic activity for typical substrates (such as MVDP, AKR1C12-­‐C13, Kvβ), and some of them have been shown to have structural roles, like Rho in the crystalline (Fujii et al., 1990; Weng et al., 2006). Mammalian AKRs have been classified into 3 families: AKR1, AKR6 and AKR7. Among them, the AKR1 family is the largest and has been further divided into subfamilies, from A to E. To date, 13 human AKRs have been described, and they participate in xenobiotic detoxification, biosynthesis and metabolism. Increasing evidence suggests the involvement of human AKR proteins in cancer development, progression and treatment. Some of them demonstrate multiple functional features in addition to being a carbonyl groups reductase. In this section we are going to review the most studied AKRs in cancer: AKR1B1 and AKR1B10. AKR1B1 in human diseases The most widely expressed member of the AKR1B subfamily is AKR1B1. Historically, it is the one that has received most of the attention, because of its role in diabetic complications and development discovered many years ago (Pastel et al., 2012). AKR1B1 is considered a causative factor for diabetic complications by converting glucose to sorbitol under hyperglycemia. Several studies have shown that inhibition of AKR1B1 could prevent, delay or even reverse tissue injury induced by hyperglycemia (Gabbay, 2004). Moreover, 56 c eCi a eti c
q q l ni R l ntI n l 1 t t tes tc C a tc I i en C Ca t d1nsl ti di I t dl a l eC e l
tc da tc I
d ns
( c
I ds
eti c
c
l xl a tc t l u t d ns
Ci u et
d ns
l x C i )s
o ei c l xc teCi
c p d ns
c psu
c
ei 1n c i c %
W-­‐ W
Szac l eaC e
Ct( l e (
x
c
c teCi
o ei c l x
q q u s
%
W Ci u et
ei c
c
%
W
1ni l 1ni dt1t l x
e l ac t( Cl d 1Ci e et(
I tc l e C i c sdl 0
a1C I ad e
tc u c s
C( t dW l i 1n I
1Ci l e e
c
c
dWn
d ao u t W nCi c t
tc nau c etl l a l W c
Cl 2 d
c
c
te n l
c
i R c C I ad e
&
!
!
&
&
#%
f a 22 d aec e nf l eah
C
er e
( Cti al
r
u
c n s i s
n c tl u l
&
CaI C l tl e c
s C a tc I en
i ) i Ca t tc x
t tc
a
p p 1Ci 1i l
ei
i C W
dtu tc etc I
) 1d tc
r
r
r
R ns
q q
ae n
q q tl
c tc a
c eteau i C CaI l Wl a n l
c c tc I WL88U%
0 aCen Cu i C W
sei ei ) t
q q 1Ci a l ns1 CI ds
&
!
r ueou
C i c sdI Ci a1 i l i u
WL8qL, tc F
dd l aC( t( d s
&
'$
( c e I i al i C en eau i C0 i C ) u 1d Wte tl oc i R c en e
1Ci u i e l
tc
eC l nWL8qL%
0
"
&
l Ct
oWot c sW z dd ae d ao u t W z dd ae
d ao u t Wdsu 1ni u Wu d c i u ,
Cx tc F
c
CW l eCi sei u l WI dti d l ei u l Wdt( CW
&
l Ct
q/zo ei l eCi c %
W
c p d ns
ti c % c i ) t tp
WL88E%
0
q q tl a t9atei al ds ) 1C l l
n C
% 1C aCl i Cl W dt1n et
WqW
Lzc 1neni 9atc i c
t C i sdxu ensdI dsi ) d c n ) c
WL88ó,
1Ci a el
q q
C i c sd i u 1i ac l 0 l
u t en ed
l ei i ) t et( l eC l l W
:U
c eCi a eti c
Rn e
et( e l
1Ci u i e l
l
i dd I Ci R en
u u dt
( c el
c
c tc I tc en
et( eti c i 1Ci dt C eti c x tI aC LL% x
WL88/%
0 iC i( C
q q
c
u c
e l
L-­‐
l sc en l WR nt n tl c tc d u u ei Cs u i d ad en e eCtI I Cl
c C de
R ten
c n c1- +
1ten dt d
z3
R nt n
WL88/,
1Ci l e I d c tc
u " -+ n
. 0( c
1- tl
, -u
%2' $ en
. 0- 2$e( , n1$ / l3$ , " $
dd 1Ci dt C eti c tc
-%
5 ( 2'
c i u eCt d
( , ' $ 02 1. * $ $ , , # * 3,
* * - " * ( 8 2( - , + ( & ' 2! $ # 3$ 2- 11- " ( 2( - , 5 ( 2' , - 2' $ 0 1 7$ 2
3( 2- 31* 7 $ 6. 0$ 11$ #
3, ) , - 5 , + $ + ! 0 , $ . 0- 2$ ( ,
c i C tc i u x d l
WL88ó%
0 aCen Cu i C Wtc nt teti c i q q *i ad 5 ( 2' 2' $ * - 5 $ 12* $ 4$ * 1%- 3
2 * 72( " " 2( 4( 27 5 1 + $ 130$ # ( , 2' $ 1- * 3! * $ , # + ( " 0- 1- +
2' $ * $ 4$ * 1 - %2' $
%0 " 2( - , 1- %2' $ . 0- 2$ ( , $ 620 " 21%0- + 20 , 1%$ " 2$ # , # 3, 20 , 1%$ " 2$ #
1C ( c e tc d u u ei Cs tl l 9l " $ l * a* 1 31(
n , & l. 9,l( 20-1l! $ tl, 8 * # c$ ' 7# $ l enu
tc u t u i1 dl* - 5 ( , 2' $ ' $ 02 , # 2' $
, # dl9& * 7" $ 0 * # $ ' 7# $
( , 2' $ * ( 4$ 0 " - * - , , # * 3
1' - 5 , ( ,
! *$
2' $ 1- * 3! * $ %0 " 2( - , - %2' $
920 , 1%$ " 2$ #
x u c
WL88/,
W
" $ * * (1 $ 6' ( ! ( 2$ # WL88ó%
9%- * # ' 0( & ' $ 0l t" 2( 4(l 2
7 " -C0
+ . 0$ # ( tc
5 ( 2' C3,c 20en
, 19 c %$ 0 1- + $ 5 ' 2! $ 25 $ $ , 1
2' $ $ , 87+
2( " " 2( 4( 27 #
%$ " 2$ # " $ * * 1 5 ( 2' . 9, ( 20- ! $ , 8 * # $ ' 7# $ 1 13! 120 2$
- 5 $ 0 ! 32
i dd I a l n ( dl i l ni R c + en$ 130
eu ! t* $ i" 2(( 4( 2C7 )51C
c! $ 25 $ $ ,
,# ,
1 # l$ l2$tc
" 2$I # nau
( , ! - 2'c 2' $ . $ *q
* $ 2q, l# ni
13.R$ 09
, 2 , 2 -%
920 , 1%$ " 2$ # " $ * * 1 ( , # ( " 2( , & 2' 2 2' ( 1 , - 4$ *
cn c
tc d u u ei Cs C . l0-1i2$ ( c, l+ ( & ' tc2 ( , d# a$ $ # tc. I- 11$tc11 C
- 6( # l- 0$ # 3" 2sei
1$ $ otc
, 87+ l 2( " d (
" 2( 4(dl27 c
- " 2( 4( 27 5 1 %- 3, # ( , 2' $
920 , 1%$ " 2$ # " $ * * 1 # $ 1. ( 2$ ( 21
c aeCi 1ntd au ad eti c tc en
dac
WL88ó%
0
" 72- 1* ( " I* - " c
* ( 8 i2( en
-, ! d
7t d$ 12$ d
0,dl ! x
* - 2 ( tc C c en
, 2' ( 1 0$ . - 02 5 $ # $ 1
aCC c eds c i d i l C a e l x %tc nt tei C tl ac C dtc t d ( da eti c tc25 - 17$ 23, ) , - 5 , + $
- + . 32 2( - , * , * 71( 1! ! #
! "
# "
2
'
nau c l i C en 1C ( c eti c i c C0 i R ( CWl i u i q q I c tc nt tei Cl 2 2' $ 1$ 25 - & $ , $ 1 $ 6(
- , . 0$ # ( " 2$ # 1$ / 3$ , " $ 1 5
( 1 # $ 1( & , 2$ # 1 + 30( , $ - 02' - * - & - %
( , 2' $
( 1 ( + t . l- 02 i
, 2C! $ " t31$
n (
dC s I i c enCi aI n # 2 ! jl1$ 1 n' $ l/ 3$z12( - , -d%tc- 02
t ' -d* - & l7ea
et + . * ( %7 %0- + 2' $ 2- 2 *
1. * ( " $ # 20 , 1" 0( . 21 5 ( 2'
( 2 " - 3* # * $ # 2" * $ 0 ( # $ , 2( " 2( - , - % 2' $ + 30( , $ - 02' - * - & - %
! 7 " - + . 32 2( - , * , * 71
# 2' $ 0$ ! c
7 al
$ , ! *x
$ 2' $u 31$
) , - " ) - 32
i u 1dt eti c l c i ac ei
l i ,Cnau
u -dt% + - 31$ WL8qq%
0 ,#
" 2( 4$ & $ , $ 1 , # 2' 22' $
20 , 1& $ , ( " 2$ " ' , - * - & ( $ 1 2- ( , 4$ 12( & 2$ 2' $ . ' 71( - * - & ( " * %3, " 2( - ,
6. 0$ 11( - , - %
-%
- " - + . 0$ 2( 113$ # ( 120( ! 32( - , - %2' $ 25 - 20 , 1" 0( . 21
$ 6. 0$ 11( - , 1712$ + 1 . 0- #
5 $ . $ 0%- 0+ $ #
-,
. , $ * - %2( 113$ 1 31( , & . 0( + $ 01 1. $ " ( " %- 0
* ( 22* $ 2- , - " 2 * 72( " " 2(
,# !
0$ 1. $ " 2( 4$ * 7 ' $ . 0( + $ 01 5 $ 0$ # $ 1( & , $ #
p p. r
p pB r r a !
. 9, ( 20- ! $ , 8 * # $ ' 7# $ +
- , # ( %%$ 0$ , 2$ 6- , 12- 4$ 0( %7 2' 22' $ " - 00$ " 2
1. * ( " ( , & - " " 301 1
1'
5
,
(
,
(
&
(
13!
(
/
3(
2
31*
7
$
6.
0$
11$
#
(
,
+
30(
,
$
2
(
113$
1
n i en C eR i u u Cl i en
q u tds tc nau c l C
q q8 c 1- , 1 %- 0 2' ( 1 0$ 3, " * $ 0
5 ( 2' 2' $
. 0- 2$
5 ( 2' 2' $ $ 6" $ . 2( - , - %2$ 12$ 1 ( & ' $ 1220 , 1" 0( . 2* $ 4$ * 1 5 $ 0$ %- 3, #
q q: 0 i C u c s s
q q8
c eds
d
q q8
q q: WR l eni aI ne ei
0- 2$ ( , 1$ / 3$ , " $ ( # $ , 2( 2( $ 1 %- 0' 3+
t
x d
l Ct
l
1Ci e tc
s
C0
ac eti c d
i tc I
t
l 9a c
I c
) 0 !
) 0 !
) 0o!
) 0 !
ael
WL8qq%
0
Wc
0 d aen r t f r
o
m2. pph
q q8
U8D i C /UD
c
eR
C I ad eti c c ac eti c i i c eC CsWen C
C0 l R
c
C
l 1 t t tet l W i C
r n n ea f l
r
h
R ten
R tdd l n
1Ci e tc
q qWC l 1 et( ds x
q q8 c
cn
d /%
0
q q: W( Cs dteed tl oc i R c
li u
dtI ne i c en
1tI c et
q q: 0
u sCt
ral e u c eti c
q qWR n C l en l
:7
1
q q: l n C óqD i en
iu1 C
i ae en d l e i c 0 c entl R i Co R
c
, # + - 31$
) 0 !
l 1te en ntI n ni u i di I s
c en
,
1l a i I c Woc i R c l
W te R l
i R ( CW
dto %
0
C
x
Cl
i 1 1 Cl e dotc I
W
t C c
i ae
q q8Wu tc ds tc
q q8 tl l tu td C tc 1Ci e tc l 9a c
l
ddi R en u ei n (
) u 1d tc l aI C u e i dtl u W
ei
t C c e l a l eC e l
q q8
c ci e C a
)
Introduction efficiently glucose to sorbitol (as AKR1B1 does). AKR1B10 has similar reducing efficiency with AKR1B1 in a wide range of substrates, such as lipid peroxides (Aldini et al., 2007; Ravindranath et al., 2009), cigarette and environmental pro-­‐
carcinogen polycyclic aromatic hydrocarbons (PAHs) (Quinn et al., 2008) and cytostatic anticancer agents such as daunorubicin (Heibein et al., 2012; Plebuch et al., 2007). However, AKR1B10 has a special affinity to reduce All-­‐trans-­‐
retinaldehyde (and other retinaldehydes) to retinol, about 50-­‐fold higher than AKR1B1, what makes AKR1B10 a crucial enzyme in the retinoic acid biosynthesis pathway in some tissues (Ruiz et al., 2009). It has been recently observed the ability of AKR1B10 to do protein prenylation, which is a lipid modification, involving the covalent addition of either farnesyl (15-­‐carbon) or geranyl (20-­‐carbon) isoprenoids to C-­‐terminal cysteines of the target protein (Chung et al., 2012). This modification allows the proteins to be located in cellular membranes. Proteins that undergo prenylation include Ras and Ras-­‐
related GTP-­‐binding proteins (G proteins) and kinases, which are involved in cell growth, differentiation, maintenance of the cellular cytoskeleton and vesicle trafficking (Casey, 1994; Gibbs et al., 1996; Zhang & Casey, 1996). In contrast with AKR1B1, AKR1B10 shows restricted tissue distribution, predominately expressed in the small intestine, colon, stomach and adrenal gland (Cao, 1998; Hyndman & Flynn, 1998). But this restricted expression changes drastically in different cancer types where it is significantly upregulated. For example, in B-­‐cell and T-­‐cell acute leukemia as well as in chronic leukemia (Laffin & Petrash, 2012), liver carcinoma (Schmitz et al., 2011), lung adenocarcinoma, squamous carcinoma and smoking-­‐related Non-­‐small-­‐cell Lung Cancer (NSCLC) (Kang et al., 2011; Laffin & Petrash, 2012), pancreas carcinoma (Chung et al., 2012), breast carcinoma (Ma et al., 2012) and oral cancers (Nagaraj et al., 2006). The fact that AKR1B10 is upregulated in so many cancers is not a coincidence and actually the literature is plenty of examples about the benefits that overexpression of AKR1B10 have for cancer cells (Figure 23). As it occurs with AKR1B1, AKR1B10 can reduce damaging carbonyl radicals and anti-­‐tumor drugs (doxorubicin and others) promoting cancer cell survival. Also, AKR1B10 blocks ubiquitin-­‐dependent degradation of the Acetyl-­‐CoA carboxylase-­‐α (ACCA) and thus enhances fatty acid/lipid synthesis, which affects 59 oncogenic
volved
inin
protein
cell
proliferation
oncogenic
function.
nvolved
proteinprenylation,
prenylation,
cell
proliferation
oncogenicfunction.
function.
nvolved
in
prenylation,
cell
proliferation
oncogenic
function.
nvolved
in protein
protein
prenylation,
cell
proliferation
E-cadherin
isis
d
apoptosis
(Figure
6).
Farnesol
and
geranylgerE-cadherin
glycoprotein
present
th
nd
apoptosis
(Figure
6).
Farnesol
and
geranylgerE-cadherin
is a aaa glycoprotein
glycoprotein present
present inin
in the
th
and
apoptosis
(Figure
6).
Farnesol
and
geranylgerE-cadherin
is
glycoprotein
present
in
th
and
apoptosis
(Figure
6).
Farnesol
and
geranylgeradherens
junctions
of
epithelial
cells.
Disruption
iol,
the
reduced
products
of
AKR1B10
from
adherens
junctions
of
epithelial
cells.
Disruptio
niol,
the
reduced
products
of
AKR1B10
from
adherens junctions
junctions of
of epithelial
epithelial cells.
cells. Disruptio
Disruptio
aniol,
the
reduced
products
of
AKR1B10
from
adherens
aniol,
the
reduced
productsare
of phosphorylated
AKR1B10
from ofof
the
proteins
comprising
these
junctions
leads
rnesyl
and
geranylgeranyl,
the
proteins
comprising
these
junctions
lead
arnesyl
and
geranylgeranyl,
are
phosphorylated
of
the
proteins
comprising
these
junctions
lead
arnesyl
and
geranylgeranyl,
are
phosphorylated
c
eCi
a
eti
c
ofchanges
the proteins
comprising
these
junctions lead
arnesyl
andand
geranylgeranyl,
arepyrophosphates,
phosphorylated toto
in
cell
adhesion
and
dedifferentiation.
to
farnesyl
geranylgeranyl
changes
in
cell
adhesion
and
dedifferentiation
nto
farnesyl
and
geranylgeranyl
pyrophosphates,
to
changes
in
cell
adhesion
and
dedifferentiation
nto
farnesyl
and
geranylgeranyl
pyrophosphates,
to changesexpression
in cell adhesion
and dedifferentiation
nto
farnesyl
and are
geranylgeranyl
pyrophosphates,
E-cadherin
isis
frequently
downregulated
spectively,
which
the
intermediates
ofof
cholesE-cadherin
expression
frequently
downregulate
espectively,
which
are
the
intermediates
cholesE-cadherin
expression
is
frequently
downregulate
espectively,
which
are
the
intermediates
of
cholesE-cadherin
expression
isinvasive,
frequently
downregulate
espectively,
which
are the
intermediates
of choles1212
or
absent
in
highly
invasive,
poorly
differenrol
synthesis
involved
in
protein
prenylation.
12
or
absent
in
highly
poorly
differen
erol
synthesis
involved
in
protein
prenylation.
u
tei
ni
c
Ct
d
ac
eti
c
W
i
)
t
et(
l
e
eal
c
d
t1t
1
Ci
)
t
eti
c
i
d
d
l
c
12
or
absent
in
highly
invasive,
poorly
differen
erol
synthesis
involved
in
protein
prenylation.
or absent in highly invasive, poorly differen
erol synthesis involved in protein prenylation.
( c ea dds c n c tc I ddl aC( t( dx F i WL8qq%
0
ure
66
The
proposed
interactive
role
ofof
AKR1B10
inin
carbonals,
farnesyl
and
lipid
metabolism
asas
inin
maintaining
the
homeostasis
of
igure
The
proposed
interactive
role
AKR1B10
carbonals,
farnesyl
and
lipid
metabolism
maintaining
the
homeostasis
Figure
6
The
interactive
role
of
AKR1B10
in
carbonals,
farnesyl
and
lipid
metabolism
as
in
maintaining
the
homeostasis
Figure
6retinal
The proposed
proposed
ofapoptosis,
AKR1B10
carbonals,
the
homeostasis
inol
to
that
is
involved
in
cellular
proliferation
to
pancreatic
carcinogenesis.
finteractive
a 2P d
crole
aec e r n a nprotein
uin
aeo
eprenylation
p farnesyl
p. and
r and
aand
er
ilipid
o m armetabolism
and
i oand
rcontributes
o c asl innmaintaining
eo
lpancreatic
etinol
to
retinal
that
is
involved
in
cellular
apoptosis,
protein
prenylation
proliferation
contributes
to
carcinogenesi
etinol
in cellular
apoptosis, protein
prenylation and
and proliferation and
and contributes to
to pancreatic
pancreatic carcinogenesi
carcinogenesi
etinol to
to retinal
retinal that
that is
isr involved
involved
l r n r in
r cellular
n
el apoptosis,
e n e protein
a n r eomprenylation
r n a r ueou lproliferation
r n r a r e rcontributes
h
cn
ael
fr
logy (2012)
25,25,
758–766
hology
(2012)
758–766
hology
(2012)
25,
758–766
hology (2012) 25, 758–766
m2. p2h
i C en c ó: D i nau c 1 c C et
u ae eti c W c
nt (
en
Cl
CCs
et( etc I I c
1Ci e tc C 9atC l 1C c sd eti c i C tel
s i ( C ) 1C l l tc I
q q8 x nac I
c 1 Cn 1l i en C 1Ci e tc l W
en tc C l
c
d ( d i WL8qL%
0 s 1C c sd etc I
q q8 tl 1Ci u i etc I
c
C 1Ci dt C eti c 0 tc ddsW
q q8 ) 1C l l ti c tc c i 1d l et
c psu et C a eti c i C etc dei C etc i dW i c i u te c e
c
l nt e tc l tI c dtc I Ci u
x atp
WL88ó,
q q8 C
i l C(
c eds 1Ci 1i l
en e en
q q8 tc
q q8 tc
c
/8
t W
ddad C 1Ci dt C eti c
Ws
n l nt c
i dd I a l
Ct u s tc c n c
u c c CWen e R i ad
l
1: Z e CI e0 n s
en
) 1C l l ti c i ) 1d tc en a1C I ad eti c i Cl R ten 1: Z i ( C ) 1C l l ti c x C 9a c eds u ae e %
0 i R ( CW
en C tl i ( C ) 1C l l ti c i x n l nt
c
WL8qL%
0 e tl R i Cen ei c i e en c R Ci d i et( eti c i 1: Z s
il z 1 c
ddl C l adel tc
C l tc C etc i t
t C c et eti c ei ac i 11i l
CCbl
et( tesWR nt n tl
q q8 tc
c
Cl R ten i C R teni ae 1: Z ) 1C l l ti c
WL8qZ%
Wl i 1: Z tl c i e ) dal t( ds C l 1i c l t d i C en C I ad eti c i Introduction AKR1B10 and the implications of the protein in such crucial pathways makes the regulation much more complicated. However, gathering all the data about AKR1B10 in cancer, appears to be obvious the advantages of upregulating this gene, whereas in biology there is always a case that breaks the rule, and in this case is colorectal cancer, where AKR1B10 is downregulated in most of the patients (Laffin & Petrash, 2012; Ohashi et al., 2013). We are interested in this dual role of AKR1B10 in cancer, and mainly in the special relation between AKR1B1 genes and colorectal cancer. 61 Introduction RETINOIC ACID PATHWAY Vitamin A and its natural or synthetic derivatives (retinoids) have been shown to act as cancer chemopreventive agents (Niles, 2004). The relationship between Vitamin A and cancer was firstly described by Wolback and Howe in 1925; they observed higher proliferation rate in vitamin A deficient epithelial cells (Wolbach & Howe, 1925). Later on it was observed that Vitamin A deficiency elevated the spontaneous and chemically induced tumors in animals (Lasnitzki, 1955). Underpinning these reports, it was also seen that pharmacological concentration of Vitamin A added to the diet decreased the incidence of chemically induced tumors in animals (Bollag, 1972; Harisiadis et al., 1978). Vitamin A or its derivatives (retinoids) cannot be synthesized de novo by any animal species and are only obtained through diet mainly in the form of retinol, retinyl ester or β-­‐carotene (depicted as a carrot in Figure 24), the latter two would be converted into retinol in the intestinal cells. Vitamin A or retinoic acid pathway (head and tail molecules of the pathway) begins with the absorption of the vitamin, which is modified through a number of reductive reactions to the bioactive compound, all-­‐trans retinoic acid (ATRA). In fact, diet contains very little all-­‐trans retinoic acid and intestinal cells are mainly exposed to retinol and retinal. Retinoids are a family of numerous compounds (over 4000) chemically related to vitamin A. The basic structure of retinoid molecules is composed by a cyclic end group, a polyene side chain and a polar end group (Figure 24). Although many retinoids can be found physiologically, just few of them are known to be bioactives per se. Retinoids biological active forms can be summarized under three main structures: retinol and retinoic acid (or RA). · Retinol has six biologically active forms (all-­‐trans, 11-­‐cis, 13-­‐cis, 13-­‐di-­‐
cis, 9-­‐cis and 11, 13-­‐di-­‐cis retinol), being retinol all-­‐trans the predominant physiological form (Baron et al., 2005). · Retinoic acid (RA) active forms include all-­‐trans retinoic acid (ATRA), 9-­‐
cis retinoic acid, 11-­‐cis retinaldehyde, 3,4-­‐didehydro retinoic acid, and less commonly found 14-­‐hydroxy-­‐4, 14-­‐retro retinol, 4-­‐oxo retinoic acid, and 4-­‐oxo retinol (Baron et al., 2005). 62 Introduction It is known that retinol is acting in a non-­‐retinoic acid receptors way, whereas RA and oxo-­‐RA work activating the classical pathway. The starting metabolite can be different, but all of them converge to retinol, which is reversibly oxidized by retinol dehydrogenases producing retinal. Subsequently retinal could be irreversibly oxidized to all-­‐trans retinoic acid (ATRA) by retinal dehydrogenases. Retinoic acid is oxidized to more polar metabolites (such as 4-­‐oxo-­‐RA, which can also activate RARs) by cytochrome P450 family members such as CYP26A1, -­‐B1, and -­‐C1 (Tang & Gudas, 2011) (Figure 24). Retinoic acid exerts its functions through two distinct nuclear receptors, retinoic acid receptors (RAR) and retinoic X receptors (RXR). Both of them have three different subtypes: α, β and γ. Each of the receptors possess different affinities for their ligands, but once bound RAR and RXR form heterodimers and activate downstream factors by binding to the retinoic acid response elements (RARE), which are normally located in the gene promoter (Bushue & Wan, 2010; Tang & Gudas, 2011). More than 500 genes have been reported to be regulated by retinoic acid (Theodosiou et al., 2010). Retinoic acid receptor pathway controls many important and diverse functions such as cell proliferation, cell differentiation, neural and vision functions, immune processes and the proper establishment of the body plan during early development. In the last years retinol has been proposed to inhibit the growth of colorectal tumor cells by an independent RA-­‐receptor mechanism (Park et al., 2005). Although it is unclear how this inhibition is achieved, several labs have found out different antitumor pathways activated by retinol treatment such as the inhibition of β-­‐Catenin (Dillard & Lane, 2007) or the loss of Phosphatidylinositol 3-­‐Kinase Activity (Park et al., 2008). In summary, retinol, ATRA and oxo-­‐RA (and close derivates) are the main bioactive retinoids with antitumorigenic activity described. 63 c eCi a eti c
Cyclic
Polyene side chain
Retinil
etinil Ester
Polar
All-trans
tran
trans
ns retinol
ns
reti
ns retinaldehyde
re
etina
et
All-trans
reti
et
acid
All-transs retinoic
All-trans 4-oxo
o retin
retinoic acid and other
polar retinoic acid metabolites
r
f a 2R d a r c o l n eo n e n a n r e
n n c a r e n el c ef r r n
1t u ti di I t d c
c tu d u i
d l ea t l n (
1Ct( eti c tl tC eds i CC d e
i e c WL88L,
d
t C c et eti c
c
n u i 1C ( c et(
1C u dtI c c e c
d1 Cc
l
ic
R ten
n l
1Ci z 1i 1ei et
I c el x c ddt
ntI n C Ctl o i WL8q8,
C tc F
eau i C ( c e x i c I
Wqóó8,
WqóóZ%
0 n l
C etc i t z l
C x ac F
C
tc I al
l
s d Cl otWL8q8%ei eC e
WL888% c
bc C
c
c etz1Ci dt C et( W1Ci z
Wqóó/,
Wqó7/, aCt
aaen
en e ( te u tc
( di 1tc I
el i C etc i t l Wen l
u dtI c c e d l ti c l x d c l
i cI
o naf nf a h
u i c l eC e
Wqó78%
0 aCen Cu i C W a ei en
WqóóE,
l ei Ctc i
c ns i r
nh
WL88q,
eC eu c el n (
( tc o
Wqóó: ,
( c ei 1C ( c e
aei
Wqóó/,
c
11dt
ei
du i l e ( Cs etl l a 0
n tu 1i Ce c
i C etc i t l tc
c
C 1C ( c eti c c eC eu c e n l C
t(
di e i ee c eti c WR n C l en
C I ad eti c i c i I c i al C etc i t l 1 enR s tl
l etddc i eR ddac
C li u
C etc i t
/E
Cl ei i 0 n C
t 1 enR s x
da l
i ae
C I ad eti c i l i u i en
1 enR s%I c l W i C ) u 1d W
S R nt n tl 1d sl
Introduction hypermethylation in some cancers (Hayashi et al., 2001; Narayan et al., 2003; Wang et al., 2003); ALDH1A2 which is downregulated in prostate cancer (Kim et al., 2005); CYP26A1, which is found highly expressed in breast cancer (Osanai et al., 2010) and LRAT which expression is also decreased in some cancers (Mongan & Gudas, 2007). To date, no attempt has been made to fully characterize the epigenetic regulation of the retinoic acid pathway in cancer. In this study, we perform a comprehensive analysis of genetic and epigenetic alterations in this pathway and we evaluate the prognostic and diagnostic applications of gene expression and DNA methylation markers in human colorectal cancer. 65 Objectives OBJECTIVES AKR1B1 enzymes have an important role in multiple biological processes, such as inflammation, differentiation and sugar metabolism. AKR1B1 genes are included in the retinoic acid pathway; these enzymes reduce retinal to retinol, which finally will end up in retinoic acid. We hypothesize that retinoic acid pathway genes, especially AKR1B1 subfamily, display an altered expression pattern through the specific DNA hypermethylation during colorectal tumorigenesis. This work is focused on understanding the epigenetic alterations of retinoic acid pathway genes in colorectal cancer and the applications that those changes may have in terms of diagnostic, prognostic and therapeutic strategies. The general objective of this thesis is to bring light into the mechanisms regulating AKR1B10 and AKR1B15 genes and what is their role in colorectal cancer. To address this goal we propose the following specific objectives: · To elucidate the epigenetic mechanisms regulating the expression of AKR1B10 and AKR1B15 genes in normal colon and in cancer cells. · To determine the functional implications of AKR1B10 and AKR1B15 deregulation in colorectal cancer. · To evaluate the utility of AKR1B1 CpG island hypermethylation as a non-­‐
invasive screening biomarker for colorectal cancer. · To characterize the gene expression and changes in the DNA methylation profiles of retinoic acid pathway genes in normal colon and in tumor samples. · To evaluate the use of molecular alterations of the retinoic acid pathway genes in the management of colorectal cancer patients (prognosis and therapeutic strategy). 69 Materials and Methods SAMPLES Patients’ samples Normal and tumor tissues were available from 25 colorectal cancer patients. DNA methylation and mRNA expression of the AKR1B1, AKR1B10 and AKR1B15 genes were analyzed. A set of 16 patients samples were collected from the Institut Català d’Oncologia (ICO, Barcelona, Spain) and 9 from the Hospital Universitari Germans Trias i Pujol (Badalona, Spain). The mean age of these patients at the time of collection was 69 years, in a range between 48 and 86 years (Table 7). Both genders were well represented, with 12 women and 13 men (Table 7). Tumors were evaluated according to the tumor node metastasis (TNM) staging system (Hu et al., 2011). Samples were fresh-­‐frozen at -­‐
80ºC within the two hours of removal. !"#$%&'()*$
+,$
-$%*$. '/01'2&",$
67
8"%
99
(3454
(345:
;<
8"%
999
(345=
>;
?)8"%
999
(345<
;7
8"%
999
(345;
<;
?)8"%
999
(345@
@;
?)8"%
999
(345>
@<
?)8"%
999
(3457
@:
?)8"%
999
(34:4
;=
8"%
999
<64>5:45
@=
?)8"%
999
47A6
>4
8"%
999
<64>54<@
@4
8"%
999
<64>54;4
<7
?)8"%
9B
<64>54;6
6>
8"%
999
47A;=
;=
?)8"%
999
47A5;4
>6
?)8"%
9B
47A6<6
>4
?)8"%
999
47A@=;
;@
8"%
999
47A>5@
@=
8"%
9B
47A>5=
<5
?)8"%
99
47A6=@
;6
8"%
999
47A<=4
;>
8"%
9B
47A5:
@=
8"%
9B
<64>47>4
@=
?)8"%
9B
47A;7@
>5
8"%
9B
Table 7 | Relevant clinical data for the twenty five patients examined, the two sources of samples are represented in different colors, lime for Hospital Universitari Germans Trias i Pujol and green for Institut Català d’Oncologia. Human stool samples and associated carcinoma tissues A total of 143 stool samples were collected at the Institut Català d’Oncologia (Barcelona, Spain); 45 of them from patients with colorectal carcinomas, 35 from patients with colorectal adenomas, 5 from patients with colonic inflammatory bowel disease (IBD) which could be either ulcerative colitis or Crohn’s disease, finally 58 control subjects who underwent surgery for diverticular disease or an endoscopy that revealed no lesions. Stools DNA obtained from healthy individuals paired by age and sex were used as control group. For 7 carcinoma patients (Carcinoma 27, 28, 29, 33, 34, 38 and 41) we also had the DNA from the tumor, what enabled us to compare among the DNA methylation of the tumor and the stool of the same patient. 73 3 ot i7 s
3an 7
P pe’i p
t
7 l di 7 u o n 3 ts
ts
nii
o ins R d ts I u an donmt
ins R d ts I donmt
UG8R
donmt
UTHR
no
a 7 l di
+ - $,
$.
(
$.
)
( *(+ &
$
L751
L752
LL397
LL349
LL392
LL352
LL393
Pn
3a
Tj
l ts gº
7 l di 7
U8j 1 a
o
ts
7 ot
m l mPplP ps l
s
HGPR
U8U s
o1 s s
l t R o1
gP l t
'
(
)
'#)
'#*
*
( #'
( #(
( #)
'
(
)
*
(,
+
,
! '
! (
! )
! *
( *,
i 9x
U8HR
3a
3a
o ins R d ts IR donmt
no 3tns u t3a
o3
o1
tM 7 6
tm 7 6
No
HGj R
R
R
3ii 6
R
HT) R j T9R j THR
a o 3 ot73t 7 s
l
3 ot i
i 41
$
$
$
$
!
!
!
!
e s Hp p e ’ i p n icr
Pn l PpplH cf
ii iPn p
ii
ii its 7 N7
nii 3tns 6
onl
jG
ts 3at7 u nor u o n 3 ts
IR w d3 no
6 n7dt3 i 3
onl 3a
Ru at a u 7 donmt
in o M 3R d ts I1
l
ot s
v
o1
vd
Ni3No
s i 73 ii o
3 ot i7 s
ii
ii7 u o Monu s 3 Uj S
n t t
its 6
Mi
tNl 6
s
T-­‐
I u t3a 7nl
7d t t t3t 7
Ni
d s ts Mns
n07
a
ii
ii iPn p Hp Pn l Pp plH cf
E
Hi
s gp s
PP
E h iHl e Pn R
E l i s uPn
mHe f ii ’ ms H lp m
Uf
®
ii7 u o
jT
Ni3No
l
3ovd7ts tb
R
3l n7da o Rts nl di 3
i 8I1
i q x ii HilHm m HPm e nlp s ml
i
PHe R cmE s PHe cmHu l R
ms e
, nuPlms n| mip
| |
ts
H
3an 7
H
inu 4) -­‐ n Ni3No
U 3tl
i 7r 71 niinu ts M
7d ou
IR s
3a s
s 3t tn3t 7 s
a r
s 7t3v s
3 inu d 77 M 7Rs n l no 3a s HTR
o nl l
r RN7ts M ovd7ts B) 1HMg
tiN3
G 3n 4 ni 1
d otn t iiv no
s
3tns 7R
ii7 u o
6 s mt3onM s R oi7
ii7 u o
Ni3No
ts
7s
R
n ns 3 l ts 3tns 6 nNs M
R
H) 9) I1
no
a wd otl
iN dn7t3tm
7d t t
wd otl
s 3R ii7 u o
ii7 u o
w iN
nNs 3
IR3a s
N7ts M 3a a l n v3nl
ii7 u o di 3
3 o 6 ovd s
7 ddondot 3 no 3a
s 31
jT
3 ot i7 s
3an 7
ls mp
9 9) Bd
ao
B
s
B9 m 3no7 6
donmt
v
9 9TBd
i 9) I u o r ts iv
o1 Nl
N3cs nl
oo57 6 s tm o7t3 3
o ins R
ii27R d ts I
6 s tm o7t3v n B
s
o1
o s vni
i M
nNt7mtii R
1
i
o7r v
s 3N r vR
I
9 9TBd
B
o 7d 3tm iv1
9 9) Bd
ao
B
s
B9 ts 7 o37 u o
U19 º
ins
ts 3n d
vMon m 3no 6 s mt3onM s R
R
IRu at a iinu
l dt tiits o 7t73 s
tM 73
3a
3a
P Hm àMx’
donl n3 o
6 tMNo HTI1 ii m 3no
9 9) Bd
l d3v m 3no d
R
3a avMonl t ts
7 i 3tns 1 ii m 3no7 ns 3 ts
s
oi7
B
s
OfCy c ms D u ls me ’ àMf
d7 o
dt 3
9 9TBd
Bao
ts s s w 1
B9 m 3no7 s
U19g vMon 6º I u t3a 3a o 73ot 3tns s bvl
7 ts t 3
ts
i 9) 1
i CI x
lHm ps l
9 9) Bd
’ i pe P p e ’ is c
B
s
M on7 M iR 3 9H)
7
s
dNot t
tni
9 9TBd
Bao
no U) l ts 1 a s 3a
N7ts M 3a
M iR po s R ol
Not t
Pn l PpplH cf
N i n dts O
l ts N3 7R 7 o nl l
tM 73tns u 7 oNs ts 3n 9-­‐
ddondot 3 7tb
i s
s u 7 N3 ts
i s BNd r t3 6
R
u t3a
s 3 o 3t
an7da 3 7 6 u
I 3 Uj S no 9 anNoR s 3a s ts
s
v 3a l
s N 3tm 3
jP
a o vB
s Mi s
3 PTS no T
3No o1 at7 don 77 o l nm 7 3a T0
dan7da 3 7 mnt ts M3a m 3no o B to Ni otb 3tns s ts o 7ts M3a
itM 3tns 1
a
s vI1
m 3no7 u o ts N 3
7R m oivR
B9
t t s v n Materials and Methods Vector and insert were ligated using T4 DNA Ligase (New England Biolabs, Beverly, MA, USA) in proportion 1:3 respectively. The reaction took place at 16ºC during 4 hours. Vector amplification All vectors were transformed in DH5-­‐α Escherichia coli strain by heat-­‐shock procedure. Transformed bacteria were seeded into LB agar, at 50 µg/ml of ampicillin. Plates were cultured overnight at 37ºC. Individual clones were selected and amplified in liquid LB at 100 µg/ml of ampicillin, in a volume of 5ml or 250ml according to the experiment. For small amounts of plasmids, and starting from 5ml of grown LB media, we used the GenElute™ Plasmid Miniprep Kit (Sigma-­‐Aldrich, St Lois, MO, USA), eluting in 100 µl of pure grade water. In larger scale experiments, we started with 250ml of grown LB media and purified the vector with the GenElute™ HP Plasmid Maxiprep Kit (Sigma-­‐Aldrich, St Lois, MO, USA), eluting the purification in a final volume of 3ml of pure grade water. CELL CULTURE DNA transfections All transfections were carried out in 6-­‐well plates, seeding 150,000 cells per well at day 0. Cells were transfected with Lipofectamine® 2000 (Invitrogen, Carlsbad, CA, USA). Transfection (day 1) was performed when cells were 70% confluent, as recommended by the manufacturer. Ten µl of Lipofectamine® 2000 were diluted in 480 µl of Opti-­‐MEM® (Invitrogen, Carlsbad, CA, USA) and incubated for 5 minutes at room temperature. Then we diluted 3 µg of the plasmid in 495 µl of Opti-­‐MEM® and mixed together with the Lipofectamine® 2000/Opti-­‐MEM® solution, incubated for 10 minutes at room temperature. After, we added 1 ml of the mix to each well in a drop wise manner. Cells were incubated at 37ºC for 5h, after this time the transfection media was removed and substituted by fresh medium. 77 Materials and Methods In the case of vectors carrying resistance for hygromycin, selection of transfected HCT116 cells was done by adding 100 µM of hygromycin B (Invitrogen, Carlsbad, CA, USA) to the media for 10 days (changing the media every two days); afterwards the antibiotic concentration was decreased to 50 µM. Untransfected HCT116 cells were used as a control and treated with hygromycin B in the same conditions until death. Lentiviral infection Infectious lentivirus particles were generated and packed using the 293T cell line. We used a second-­‐generation system with co-­‐transfection of three plasmids: (1) the lentiviral vector pLX302 that contains the gene of interest (AKR1B10 or AKR1B15) and LTRs (Long Terminal Repeats), (2) the pCMV-­‐VSVG plasmid for pseudo-­‐typing with VSVG, allowing the lentivirus to infect a broad range of cells, and (3) the psPAX2 plasmid with gag, pol, and rev genes (Naldini et al., 1996). As a transfection and infection control we used a GFP lentiviral plasmid instead of AKR1B10 or AKR1B15. The virus generation and the infection process were performed following the standard protocol (Tiscornia et al., 2006). Dznep treatment In order to know the relevance of the EZH2 protein in the inactivation of the enhancer, we treated HCT116 cells with the EZH2 inhibitor 3-­‐Deazaneplanocin A hydrochloride (DZNep) (Sigma-­‐Aldrich, St Lois, MO, USA). Cells were treated for 72 hours with 5µM of DZNep, replacing the medium every 24h. All-­trans retinal treatment HCT116, Sw480 and HT29 colorectal cancer cell lines were treated with All-­‐trans retinal (Sigma-­‐Aldrich, St Lois, MO, USA) for 24h at 0.5 and 2µM. 5-­aza-­2'-­deoxycytidine treatment To study the importance of the methylation in the regulation of the genes studied, we treated cells with the demethylating drug 5-­‐aza-­‐2’-­‐deoxycytidine (5-­‐
AzadC) (Sigma-­‐Aldrich, St Lois, MO, USA). 5-­‐AzadC is a nucleoside analogue that is converted into nucleotide and incorporated into the DNA. There, 5-­‐AzadC can 78 Materials and Methods trap DNMTs by forming covalent complexes (Sheikhnejad et al., 1999), resulting in global DNA demethylation in a replication-­‐dependent manner. We seeded a total of 1.5x106 HCT116 cells (106 in the case of control cells) in a 10 cm2 plate and treated for 48 hours with 0.5 µM of 5-­‐AzadC, changing the medium every 24 hours. After 48 hours we removed the 5-­‐AzadC medium and let the cells recover with normal medium for 24 hours. Then cells were collected. To verify the success of the treatment we analyzed the re-­‐expression of the silenced genes EN1 and INHBB, by qPCR and the demethylation of their CpG island associated promoter by bisulfite sequencing (primers are in the Annex II and III). Luciferase enhancer assays To test the enhancer capability of some sequences of the AKR1B1 locus we used the pGL4.23[luc2/minP] vector (Promega, Madison, WI, USA), which is specially designed to test luciferase enhancer activity of the fragments inserted. We first amplified the selected sequences using the Phusion® High-­‐Fidelity DNA Polymerase (Promega, Madison, WI, USA) to avoid any PCR error. Then PCR amplicons were inserted in the multicloning site (MCS) of the pGL4.23 vector (Figure 26A). pGL4.23 vector is mainly composed by a MCS, a minimal promoter and the firefly luciferase gene (Figure 26B). The different constructs and the empty vector were transfected in different wells, each of them in quintuplicate (in 5 wells) and in at least three independent experiments. The assay was done using the Dual-­‐
Luciferase Reporter Assay kit (Promega, Madison, Wi, USA) in accordance to the manufacturer’s instructions. Luciferase activity was assayed using 10 µl of the cell supernatant and the Dual-­‐
luciferase reporter plasmid system (Promega, Madison, WI, USA). Luciferase measurements of the different fragments were normalized against Renilla luciferase activity (cotransfected with the pGL4.23) to determine the relative luciferase activity and the fold activation. Luciferase/Renilla signal was analyzed with a Centro LB960 luminometer (Berthold technologies, Bad Wildbad, Germany). Empty pGL4.23 vector was included as control. Next, we detected 79 RVprimer3
Multiple cloning region
Minimal promoter
luc2 reporter gene
3SV40
ot i7
s
3an 7
late poly(A)
region
Reporter Vector primer 4 (RVprimer4) binding region
ColE1-derived plasmid replication origin
(Ampr)
3a toSynthetic
iN t β-lactamase
o7
3tm
t3t 7coding
ts region
ii7 u t3a
Synthetic poly(A) signal/transcriptional pause signal
wd Reporter
otl sVector
371 primer 3 (RVprimer3) binding region
t o
1–70
78–108
141–1793
1828–2049
2117–2136
2374
3165–4025
s 3 m ot
s 37 ts
4130–4283
4232–4251
5´...ACAAAACAAACTAGCAAAATAGGCTGTCCCCAGTGCAAGTGCA
Synthetic poly(A) signal/transcriptional pa
SfiI
Acc65I
KpnI
BglI
EcolCRI
SacI
NheI
XhoI
EcoR
GGCCTAACTGGCCGGTACCTGAGCTCGCTAGCCTCGAGGATAT
3 i 73 U ts
d s
SfiI
BglI
s3
HindIII
GGCCTCGGCGGCCAAGCTTGGCAATCCGGTACTGTTGGTAAA
Synthetic poly(A) signal/
transcriptional pause site
(for background reduction)
A
AGACACTAGAGGGTATATAATGGAAGCTCGACTTCCAGCTT
minP
Ampr
ori
Acc65I
KpnI
EcoICRI
SacI
NheI
BmtI
XhoI
EcoRV
BglII
HindIII
pGL4.23[luc2/minP]
Vector
2067 SalI
2061 BamHI
(4283bp)
SV40 late
poly(A) signal
Figure 2. Multiple cloning region for the pGL4.23[luc2/
15
19
24
26
28
32
34
42
47
66
Sequence information and restriction enzyme tables for the pG
available online at: www.promega.com/vectors
For more information see the pGL4 Luciferase Reporter Vecto
#TM259, online at: www.promega.com/protocols
Minimal promoter
B
5948MA
luc2
Figure 1. pGL4.23[luc2/minP] Vector map.
pGL4.23 + Fragment
Empty pGL4.23
P Hm à8 x U ’
plH cf
. fàOG
n 3 73 anu
l
o Ml s 3R u
ddit
3o s 7 l
3
3a
d
ye Pn j u ls me ’
3avi 3tns
G1HU m 3no u t3a 3a
33a
to73 l
l
sa s
3avi 3
3avi3o s 7 o 7 6
ts
s bvl
3n o
M on7 M iRu a o 3a
3 ol ts
o
pP p
37 3a
m ot 3tns n 3a
3avi 3 R ts no o 3n
nNi
U
e
s
sa s
o
73 s
P m nl ’
d
o l
ts 7 o3
. fàO Hp
tit3v n 3an
7i 3
6 tMNo Hj I1
o Ml s 3 Ns l
u a 3a o t o s
Pn l
7 ns
3avi 3
l
no
3avi 3tns
3tmt3v1
3a
d
G1HU º
R d7u at aR
3a l
3avi 3
7to
i s
i s B d r t3 6
3a dNot t
don N 3 u 7 N7
s
o Ml s 3 m 3noR N7ts M 3a
R
IR 3 ou o 7 u
o Ml s 31 a
u 7 N3 s
u t3a
tM 73tns u 7oNs ts 3n 9-­‐
dNot t
a o vB M iR po s R ol
3n 77 77 3a l
tM 73
77
N7ts M3a
s vI1 7l
N i n7dts
ii l nNs 3 n 3avi 3tns 73 3N7 n 3a o Ml s 3R
P
P
the USA 800-356-9526 · Telephone 608-274-4330 · Intern
v 3Promega
u n tsCorporation
d s · 2800
s 3Woods
tMHollow
73tnsRoad·Madison,
7R u t3a WI l53711-5399s U.S.A. l· Toll Free
1 inn3
a s bvl 7 o
t7n7 atbnl
3a
4)
o7 3a 3 N3 3a 7 hN s
7t3 u t3ats 3a 3 oM 3 t7 l
1 nu m oR l
3avi 3 Ru ati
l
t7 in r
u tii N3 n3aRl
ua s
3avi 3
3 ot i7 s
s
Ns l
l
3avi 3 1 3a
l N73
in r
ns 3oni no l
77 l
3avi 3tns u 7 noo 3 3a
1 s Ns l
3avi 3
3an 7
tM 73tns u t3a 3a
u 7 i7n
7
dn7t3tm
1
#)
!
!
!
!
& "$ %
!
P Hm àz x
e
s l
3 ou o 73a l
3o s 7 3
ts iivRu N7
m e nl Pn n Hne l ci l
7 Ml s 3u 7itM 3
99P
77 s bvl
o 7Ni37 n 3a
Hm ls Pnp ml e l ci l
3avi 3
3n
u t3anN33a
3a
’ ms
ii71 a
ts 3a l
Ns l
3n 3a
3avi 3
3avi 3tns o
ns 3oni u 7 ns
ts d o ii i
3tns 6 tMNo Hj I1 a s nol
3avi 3
3a
ti nwns 3 73ts no o 3n 77 77 3a 73 3t73t i 7tMs t t s
3u
l
G1HU m 3no s
Ns l
t o s
s
l d3v d
u ls mf
3avi 3
3o s 7 3tns 7 u o
s 3a 3u n 7 l di 71 nno ts 3 7 n itb
nl d o 1
n a o Ml s 3 o ts
ss w 1
ii ms iP m lPs n pp c
N7
3a
s t s R
9 9T1
ii donit o 3tns r t3
6
I 6 n a
I 3n 73N v 3a mt tit3v n 3a
t7
ninotl
3ot
onit o 3tns
t3
l t3n ans ot
s bvl
no s M nino
u 3 oB7niN i
7 i3 6
7 3a 3 o
u t3a
3 ol ts 3a s Nl
N7 7 HRUB t7B6HBl
3 3o bnitNl BTB o nw s tit
3
9 9)
s
77 v no 3a s ns Bo tn 3tm hN s 3t t 3tns n ii donit o 3tns s mt tit3v1 s no o 3n
ii
ii7 3o s 7 t Ms n73t 7 nodno 3tns R
d
I1
o n mt i
ii7R
3anwvBGBs t3onBTB7Ni nda s viIBH B
s iv
i 3n o N
itmts M
3n nol
ii7 a m
3tm
b s Ru at a t7 s
v 6 tMNo H4I1
49
Yh
development of methods, such as incorporation of radioactively labeled 3H-thymidine into DNA or the use of 5-Bromo-2‘deoxyuridine (BrdU) as a substitute for radioactive thymidine to label DNA in living cells.
The above methods have a number of disadvantages, including: use of radioactive materials and relatively complex
3 ot i7Thesuse of tetrazolium
3an 7 salts, such as MTT, commenced in the 1950s, is based on the fact that living cells reduce
techniques.
tetrazolium salts into colored formazan compounds.
Vn
X^iniZhi^c\
dVXi^kZVhhVn
i^c\VcYk^VW^a^in
d[XZaah
EBH
b^idX]dcYg^Va
YZ]nYgd\ZcVhZ
MIIiZigVoda^jb
MII[dgbVoVc
I]ZiZigVoda^jbhVaid[MII'!("7^h"'"bZi]dmn")"c^igd"*"hja[de]ZcnaR"'="iZigVoda^jb"*"XVgWdmnVc^a^YZhVai^hVc^ccZghVai!i]Vi^hXaZVkZY
P
Hm à9 x P i
iipm H lPs n s
ls l o l mhps iH i s me r n p ilf
id[dgbVoVcWni]ZhjXX^cViZYZ]nYgd\ZcVhZhnhiZbd[i]Zb^idX]dcYg^VagZhe^gVidgnX]V^c#Dcana^k^c\XZaah!edhhZhh^c\Vc^ciVXi
b^idX]dcYg^Va
bZbWgVcZVcYVahdVc^ciVXiXZaabZbWgVcZ!Yd]VkZVXi^kZYZ]nYgd\ZcVhZ#6\Zcihi]ViY^hgjeii]ZbZbWgVcZhVcYYZhigdni]ZgZhe^gVidgn
X]V^cl^aa^cVXi^kViZi]ZZconbZVcYi]ZgZ[dgZi]Z[dgbVi^dcd[i]ZhdajWaZdgVc\Z[dgbVoVcWngZYjXi^dcd[i]ZnZaadliZigVoda^jbhVai#
I]ZgZVXi^dcgZfj^gZhi]ZegZhZcXZd[VcZaZXigdcXdjea^c\gZV\Zci!l]^X]^he]ZcVo^cZbZi]dhja[ViZ!hZgk^c\VhVc^ciZgbZY^ViZZaZXigdc
a o no R 3a
ns s 3o 3tns n 3a
v t7 dondno3tns i 3n 3a s Nl
o n VXXZeidg#
l
3 nit iiv
l
s N a
77 v u 7 d o nol
t7
MNi oiv s 3 o7d
Noo s 3iv 3a
3 as ninMv no M s nl
ano3
77n t 3
s nM s nN7
s Mts
ots M 6
sN i 7 R
3 ot i
I ts 3n
it
o
7ts Mi
781 a
s
atl
iiR 3a
6 s s t7tRH) 9UI1
a M
g
anl ninMv
7 hN s
78 nl di wt7 o oNt3
3u
s 3a
ts 3a M s nl t
M
7 hN s
tl l
4H
78
I 6 tMNo H8I1
s
ts 3tns n 3a
3o s 7B 3tm 3ts M
u t3a 3a 7 ni ts Mdond o3t 7 n s
o77 hN s
s
s 3a
78 o
l n t t
no d ol
7 o in 3
3a
ts s s w 1
7 Bd tots M
s 3 3n 3a
e
s3
1 a
N7ts M
3aonNMa ns
3 oM 3
n3t 6
nN i
I
ts ts M n 3a
78 3n 3a M s nl t 3 oM 3 7 hN s
o d to
s s 3iv
78R3a M s nl t 3 oM 3
o
3 oM 3 7 hN s
wdo 77
v 3a
nl di l
noo 3 on3n7d
s N3 n3a 73o s 7 n 7 3a
I 3o s 7 otd31 a M
77 Ni ts ts Mn t 3 iv niinu ts M 3a
78 nl di w in itb 7 3a
3a u ti B3vd
6
6M
3n 3a 3 oM 3 7 hN s
1 no 7N
o 3u n t73ts 3
nl
6 3tm 3ts M s
o
7 hN s
l N73 i7n ns 3 ts 3a
g
t7
ot MNt
7 hN s
M
MNt
3o o
dotl
s ns N i 7
s s ns N i 7 Rts 3at7
M s nl t 3 oM 3 7 hN s
t7oNd3
I vd
B Nt
ts 3n 7ts Mi 3o s 7 otd31 a s 3a M
ts 3a
d 376
RH) 9UI1 a o
s
ts 7 3a 3 oM 3ts M7d t t t3v n 3a
3a 3o o
its onl t
l n73 nl l ns iv N7
nl dns s 37 3n 3at7 7v73 l K MNt
nl
no ts M 3n 3a
PlPn v
iN73 o
7v73 l
ii71
3No o07ts 73oN 3tns 71
ns e
a
3tm
3o s
7n 3a 3
o r
n 3u n M s o i o d to
3 ot i7 s
d 3au v7K a
ns B nl ninMnN7 s
nl ninMv to 3
P Hm àq x
e
il mPn Pn e s pl s l
a
d to 6
nts ts M6
I
nPpe s
lPs nf pq o s Hi ’ ms H
p pl PnPlP i p H n f
’l
ms e
s i
3n o l
t7oNd3ts M3a nd s o
o hNto 7 3a do 7 s
t3a Niiv ndt 7 3a 7 hN s
d t t s N i n3t
u t3a
7at 37 s gno do l
ts M o l
n a sM 7
o d to d 3au v no 3a
I d 3au v1
o d to d 3au v n 3 s o 7Ni37 ts ts 7 o37g
7t3 3a 3
3an 7
6
s H i plm n
nnPpP| àI COf
m t Pn l
i 3tns 7 6 s
i7I 3 3a
3No 73nd n ns 7R I n 3a 3 oM 3
Ms 1 a
o d to 3 l di 3 Ru at a t7 N7
s
ts 3on N
ts 3n
o d to 3 l di 3 1 n d o nol
ts t 3tns 7n a s Ms 3No don3n ni d d o 6 s
1
N3 3 oM 3 7 hN s
3 oM 3
wd otl
3tm iv
d 3au v
3n tw 3a
n 3a o d to 3 l di 3 3n 3a
|
Ms
s 37 u
1
v 3a N7 n niinu
3a
RH) 9UI1
4U
Materials and Methods DNA METHYLATION ANALYSIS Genomic DNA extraction Genomic DNA extractions were done using the PureLinkTM Genomic DNA Mini Kit (Invitrogen, Carlsbad, CA, USA), according to manufacturer’s instructions. DNA was eluted using elution buffer and quantified using the NanoDrop (Thermo Scientific, Rockford, IL, USA). Bisulfite treatment There are different methods to study DNA methylation at specific genomic loci (Laird, 2010; Lister & Ecker, 2009). They range from methylation-­‐sensitive restriction enzymes to Methylated DNA immunoprecipitation (MeDIP) (Down et al., 2008) and bisulfite modification of DNA (Frommer et al., 1992). Sodium bisulfite modification followed by sequencing is currently the gold standard in DNA methylation analysis as it provides single base resolution and the highest coverage (Lister & Ecker, 2009). Sodium bisulfite treatment deaminates unmethylated cytosines (C) to produce uracil (Figure 30A), while methylated cytosines (mC) remain unaltered. Uracils are read as thymines by DNA polymerase. In that way, umethylated cytosines will appear as thymines in sequencing of PCR amplicons obtained after sodium bisulfite treatment of DNA. Alternatively, methylated cytosines will be preserved as cytosines in the electropherogram (Figure 30B and 30C). By comparing the modified DNA with the original sequence, the methylation state of the original DNA can be inferred. Bisulfite conversion was performed using the EZ DNA Methylation™ kit (ZymoResearch, Orange, CA, USA), using 250 ng of DNA as starting material and eluting in 50 µl of elution buffer. 84 3 ot i7 s
3an 7
A
B
5’- CATTCTAGCGATCTCGAATCGC -3’ !
5’- UATTUTAGCGATUTUGAATCGU -3’ !
5’- TATTTTAGCGATTTTGAATCGT -3’ !
C
G G G CG
TTTG
CGG
TTG G CG
CGT
CG
TTG!
G G G TG
TTTG
TGG
TTG G TG
TGT
TG
TTG!
P Hm OI x U
e Pn lPs n s
U s PHe PpHi Pl lm np s me
e l ci l
cls pPn p , Pn iH
s n Pp Hiic e l ci l
n l
l
cls pPn i
Pn ls m Pi HmPn l ps PHe PpHi Pl lm le nlf
lPs n| Hne l ci l
cls pPn p ,Pn m U m s nu ml ls Hm Pip o Pi
U m e Pn Hn
l f U i lms ’ ms m e s lo s p e ’ i p| l
H’ ’ m
is o ms n PpHne l ci l f
PpHi Pl p H n Pn
t7Ni t3 3o 3l
s 3 t7 a ol Ni 3n
s o 7Ni37 ts atMaiv o Ml s 3
Rt3 ts 3on N 7 m otnN7
7ts Mi B73o s
1 3a 7
73o s
s
o r7
l ns 73o 3
4T
Materials and Methods that this degradation affects between 84–96% of the total DNA (Grunau et al., 2001). Because of the poor quality of the DNA recovered from bisulfite treatment, nested PCRs were performed to improve the performance. Two µl of bisulfite converted DNA were used in the first PCR (external). The second PCR (internal) was performed using one µl of a dilution of the first one. Dilution ranged from 1:1 to 1:20. Bisulfite primers are listed in Annex II. Infinium 450K methylation arrays To obtain a genome-­‐wide profile of DNA methylation in HCT116 control cells and treated with 5-­‐AzadC, we performed 450K Infinium methylation array, which has a good coverage of the whole genome, specially CpG islands (97% are covered) and other interesting regions, such as shore CpGi islands and promoters. One µg of DNA was quantified using the Qubit® 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA), and transformed using the EZ DNA Methylation ™ kit (ZymoResearch, Orange, CA, USA). Biological duplicates of each experiment were performed. Methylation probes used in this retinoic acid pathway experiments are listed in the Annex VI. RNA EXPRESSION ANALYSIS RNA extraction Total RNA extractions were done using TRIzol® (Invitrogen, Carlsbad, CA, USA), according to manufacturer’s instructions and following the single-­‐step RNA isolation method (Chomczynski & Sacchi, 2006). All processes were carried out on ice and centrifugations steps at 4ºC in order to preserve RNA integrity. Cultured cells were trypsinized and washed with cold PBS 1X twice before RNA extraction. Then cells were resuspended using one ml of TRIzol® per 2.106 cells. In the case of tissue samples, mechanical homogenization with the razor blade was crucial to obtain high recovery yields. At that point samples could be stored at -­‐80ºC for at least one year. During the extraction protocol, we applied a recombinant DNAse I treatment (Applied Biosystems, Foster City, CA, USA) to 86 Materials and Methods avoid DNA contamination mainly from those samples coming from transfection experiments. RNAs were quantified in the NanoDrop (Thermo Scientific, Rockford, IL, USA) and were run in a 1% agarose gel with ethidium bromide staining in order to assess the integrity of the RNA. We also performed RNA expression microarrays and RNA-­‐seq of HCT116 control cells and treated with 5-­‐AzadC. Each experiment was performed in biological duplicates. For these experiments, total RNA was extracted with the miRNeasy® Mini kit (Qiagen, Venlo, Netherlands) which allows the recovery of all sizes RNA. The quality of these RNAs was checked using Agilent Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) and a minimum score of 9 was required to be included in the experiment. Two µg of RNA were used for each RNA-­‐seq experiment. For RNA expression microarray analysis, we used Agilent SurePrint G3 8x60K array (Agilent Technologies, Santa Clara, CA, USA) and the data obtained were normalized and analyzed statistically using Limma (R package) (Smyth et al., 2005). Reverse transcriptase PCR For each reaction we used 500 ng per sample that were reverse transcribed by M-­‐MLV (Invitrogen, Carlsbad, CA, USA), in the presence of pd(N)6 Random Hexamers (GE Healthcare, Piscataway, NJ, USA) and RNasin® ribonuclease inhibitor (Promega, Madison, WI, USA). The reaction was performed according to manufacturer’s recommendations in a final volume of 25 µl. Two negative controls were added, one without the M-­‐MLV enzyme and the other without the RNA. Quantitative Real-­Time PCR The expression levels were analyzed using the quantitative Real-­‐Time PCR technique in the LightCycler® 480 (Roche Diagnostics Corporation, IN, Indiana, USA) platform with Fast Start DNA Master SYBR®Green I mix (Roche Diagnostics Corporation, IN, Indiana, USA) in a final volume of 10 µl. Samples were diluted 4 to 20 fold (except for 18S control that was diluted 1000 fold), depending on the expression levels of each gene. Every sample was 87 Materials and Methods analyzed in triplicate. When possible, forward and reverse primers were designed in different exons to avoid genomic DNA interference. Amplified products were run in an agarose 2% gel and sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA). We used 3 to 6 housekeeping genes (PUM1, MRPL19, PSMC4, 18S, B2M (β2-­‐
microglobulin) and PPIA) in order to normalize the expression levels. Efficiency was calculated for each PCR reaction as described (Schefe et al., 2006). Standard deviation was calculated on PCR triplicates. Expression primers are listed in Annex III. CHIP AND CHIP-­‐SEQ EXPERIMENTS We used chromatin immunoprecipitation (ChIP) against some post-­‐translational histone modifications, histone variants and transcription factors (TF) (summarized in Table 11), in order to characterize the epigenetic landscape of the genes of interest. In this study we used two different applications of ChIP: the locus specific ChIP (analyzed by quantitative PCR, ChIP-­‐qPCR) and the genome-­‐wide ChIP (analyzed by next generation sequencing, ChIP-­‐seq). Both applications were performed using the Millipore EZ-­‐Magna ChIPTM A/G kit (Billerica, Massachusets, USA) according to the manufacturer’s instructions. First we collected cells by trypsinization (as explained previously). Then we proceeded to the cross-­‐linking step (1% formaldehyde, 37ºC, 15 min) followed by a 5 min inactivation step with 125mM glycine. 20x106 cells in 1mL of cell lysis buffer were sonicated using Bioruptor sonicator (Diagenode, Liège, Belgium) with 60 s ‘on’ and 60 s ‘off’ for 34 min. Fragmented chromatin was aliquoted in 50 µl (106 cells), which is the recommended starting material for the EZ-­‐Magna ChIPTM A/G kit. In this study we carried out whole genome ChIP-­‐seq in HCT116 control cells and HCT116 treated with 5-­‐AzadC (as described previously). Besides, ChIP-­‐qPCR was performed for specific regions in different cell lines and also to validate ChIP-­‐seq data. 88 3 ot i7 s
no a Bh
Ru
N7
3 l di 3 no 3a h
s nol
itb 3a
ns
1
wd otl
7 n in N7 7d t t
Fi n a aonl
i7n N7
ns
s 31 m ov h
3ts tl l Ns ndo tdt3 3tns
Fi n ts dN3 7
u 7 d o nol
in
ts 3otdit 3 R s
a B7 h wd otl
7
ts M ns 3oni 3n
a R3a tl l Ns ndo tdt3 3tns u 7 d o nol
3otdit 3 7Ru at a u 7 s n3 dn77t i ts 3a
3an 7
ts 3a
ts tninMt i
s 3 N 3n 3a atMa
n73n 3a 3 as thN 1
Antibody
Source
Catalogue ID
Ab/IP
H3
abcam
ab1791
5 !g/IP
H3K4me3
Millipore
NG171579 0.6 !g/IP
H3K4me
Millipore
07-436
3 !g/IP
H3K27me3
Millipore
07-449
5 !g/IP
H3K27ac
Millipore
05-1334
3 !g/IP
H3K9me3
Diagenode
pAB-060-050 1 !g/IP
H3K9me2
Diagenode
pAB-056-050 3 !g/IP
H3K36me3
abcam
ab9050
5 !g/IP
H4K20me3
abcam
ab9051
5 !g/IP
H2A.Z
abcam
ab4174
5 !g/IP
mH2A
Dr. Marcus Buschbeck lab
5 !g/IP
RNApol II
Millipore
05-623
1 !g/IP
RNApol III
Santa Cruz
sc-21754
2 !g/IP
CTCF
Millipore
17-10044
5 !g/IP
IgG
Jackson Immunoresearch 011-000-120 5 !g/IP
ChIP-seq
Control cells
1
1
3
1
2
1
1
1
5
2
4
3
5
2
i CC x He e mc s l
nlP s P pHp Pn
h n
hp nHe ms P m nl
pn
ls m
l Càn m HPm s ml
s n PlPs np| s nlms i
CC8 iip n
CC8 lm l o Pl Mh r
f
s 3a
7 n a B7 hRu
3a 9H s Ms
u 7 s nol
itb
3n n 3a
3n M 3a o
a B7 h s iv7t7 6
RH) ) 4x n otMN b
t o s 3 s Nl
7 3n o
a
i 99I1 a B7 h wd otl
s3
7u o d o nol
3n 7 3 3a inu o itl t37 6 otMni
RH) ) 4I1 a
ChIP-seq
Treated cells2
1
1
2
1
2
1
1
1
9
5
4
4
9
5
d’ mPe nlpf C b à m mls l
hp d’ mPe nlp Pn s l
o n N7ts M3a ts dN31 n7t3tm s s M 3tm ns 3oni7 no
l n t t 3tns u o N7
1h
s
1
l dit t 3tns dotl
a at73ns
RH) ) Px n otMN b
o7 o it73
ts
ss w
ts 3otdit 3 1
48
Materials and Methods PROTEIN ANALYSIS Western blot Tissues were snap-­‐frozen in liquid nitrogen, then smashed and homogenized in lysis buffer, containing 10 mM Tris-­‐HCl (pH 7.5), 1 mM MgCl2, 1 mM EGTA, 10% glycerol, 0.5% 3-­‐[(3-­‐cholamidopropyl) dimethylammonio]-­‐1-­‐propanesul-­‐fonate, 1 mM β -­‐mercaptoethanol, and 0.1 mM phenylmethyl sulfonyl fluoride. Extracts were vortexed for 30 minutes at 4°C and, after centrifuging for 20 minutes at 15,000 g, the supernatants were stored at -­‐20°C. Before doing the western blot experiment, extracts were boiled for 5 minutes in Laemmli sample buffer and equal amounts of protein (20-­‐30 µg) were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-­‐PAGE) and transferred electrophoretically to a immobilon-­‐P membrane (Millipore, Billerica, Massachusets, USA). The membrane was blocked in TBS 0.5% plus 5% nonfat dried milk for 90 minutes at room temperature and probed with a primary antibody AKR1B1 (1:3000; Dr. Jaume Farrés Lab) or AKR1B10 (1:3000; Sigma-­‐
Aldrich) at 4°C overnight. The membrane was washed thrice for 10 minutes in TBS 0.5% and incubated with anti-­‐rabbit secondary antibody (1:3000; Dako, Agilent Technologies, Santa Clara, CA, USA) for 90 minutes at 4ºC. Targeted proteins were visualized using SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific, Rockford, Illinois, USA). We used GAPDH (Abcam, Cambridge, Massachusetts, USA) as a loading control. Immunohistochemistry A tissue microarray with 196 tumors from patients diagnosed of advanced colorectal cancer and homogenously treated with combinations of 5-­‐FU and irinotecan was constructed. Sections were stained using a fully automated staining system for immunohistochemistry (Ventana Medical Systems, Tucson, Arizona, USA). The primary antibody used was anti-­‐AKR1B10 (Sigma-­‐Aldrich) and anti-­‐AKR1B1 (Dr. Jaume Farrés Lab) at dilution 1/100. Immunostaining of HepG2, HCT116 and SW480 cell lines were used as a controls. The staining was semiquantitative evaluated as diffuse positive, focal and negative. 90 3 ot i7 s
Nots M3a d 73 u v o7 s u 3 as thN 7 a m
3n
3 3 dav7t i aonl
3 as thN 7 u o
3u n 7d t t
s
l
3n t s 3t v s
7 hN s
7
otm 3 7 6G R T
m ind
u at a iinu N7
3ts ts 3 o 3tns 7 n Noots M u t3ats 3a s N i N71 a 7
7tMs
N a s iv7 7 o
s
3an 7
73a 3 o
hN s 3t v 7d t t itM 3tns don N 37 n in7 iv in itb
ns aonl n7nl
s
tB I1
ns nol
a
ts 3a s N i N7
3tns
1
d3No 6U I 6 tMNo U9I
3 as thN a 7 7nl
7d t t t3v 3a 3
r 7t3l no no i 777Nt3 i no t o s 3 ddon a 71
s 3at7 73N v u
N7
3a
U
3 as thN 6 r r oR H) ) PI u at a
t s 3t t 3tns n dav7t i ts 3 o 3tns 7
3u
s do mtnN7iv 7 i 3
7 Ml s 371U 3 as ninMv t7 d o3t Ni oiv N7 Ni 3n t s 3t v aonl
nol
ts M s nl t o Mtns 7o s Mts M onl
Cross-linking!
aonl
3ts
3ts ts 3 o 3tns 7
u r 7Nd 3n 7 m o i aNs o 7n r 71
Digestion!
PCR specific!
amplification!
iinu 7 3a
Ligation!
Reverse cross-linking!
3C library!
P Hm OC x ms e lPn s nl lp m ms pphiPnt
o Pl s me i
c | l mo m p l
ms e lPn Pp
P pl
o Pl
m plmP lPs n nrce
n iP l
o Pl
.
iP p f
’ HmPP iP lPs n ’ ms H lp
mPu
ms e l O iP m mc m ’ m p nl ii ’ s ppP i
ms e lPn Pnl m lPs npf s nu nlPs n i n icpPp s
O iP m mP p Pnus iu p Pn PuP H i e ’ iPP lPs n s O iP lPs n aHn lPs np c p e Ph H nlPl lPu
n
ms p
i l lPs nf
a to73 73 d ts s v
3a
nm i s 3 its r
its r
ts
tM 73
aonl
3u
3 as thN t7 nol
s ts 3 o 3ts M aonl
3ts t7 7niN titb
u 7 3a
ns
i
N7
s
tM 73
av
on77Bits r ts MRu at a don N 7
3ts 7 Ml s 37 u t3ats 3a
u t3a
ii1 on77B
7Nt3 i o 73ot 3tns s bvl
ts 3at7 73N v1 3 o dNot t 3tns R on77Bits r
R
s
o Ml s 37 o itM 3 1 a o 7Ni3ts M 7 l di t7 3a U it o ovRu at a
ts iN 7 i oM s
nl di wo do 7 s 3 3tns n 3a
aonl
3ts ts 3 o 3tns 71
a
89
Materials and Methods ligation product is the result of a unique interaction, the quantification and comparison of different interactions was done by quantitative PCR using specific primers for the given interaction. The quantification of these ligation products in a given population of cells leads to relative frequency maps where physical interactions can be inferred. The specific primers used in the AKR1B1 locus are described in the Annex VIII table. BIOINFORMATICS The Cancer Genome Atlas project Most of the genomic data used in this project were downloaded from The Cancer Genome Atlas (TCGA) project. This huge project attempts to get a better knowledge about the most common cancers (17 cancers in the last update, 09/10/2013), by applying genome-­‐wide techniques that provide information on gene expression, DNA sequence and DNA methylation to large series of cases. We have used the TCGA database to investigate DNA methylation, gene expression and gene mutation in the regions or genes of interest, using publicly available bioinformatic tools or developing our own scripts. TCGA: Gene expression Colon cancers expression array data was downloaded from The Cancer Genome Atlas database (tcga-­‐data.nci.nih.gov). Gene expression data from the Agilent G4502A microarray (Agilent Technologies, Santa Clara, CA, USA) was available for 155 colon cancer patient samples. There were 12 tumor-­‐matched normal tissues and 7 normal unmatched tissues available that were also included in the analysis. Expression results were obtained using the TCGA dataset on Oncomine database (www.oncomine.com) and using in-­‐house made R scripts. TCGA scripts were constructed with R (http://www.r-­‐project.org/) using the Bioconductor open source R package (http://www.bioconductor.org) (Gentleman et al., 2004). Expression probes are depicted in Annex IX. 92 3 ot i7 s
v
3an 7
e l ci lPs n
nino 3 i
a
s
no
a
s
o
o GT)
s nl
s
3t77N 67Nl l
s ts tNl l
3i 7
3
ou a
otb
3avi 3tns oo v
nu s in
onl
71
t o s 3 s Nl
ts
3 u o
o n 3Nl no7 s 3a
77n t 3
s nol
i
i 9HI1
Cancer Type
Code
Normal Samples
Tumor samples
Bladder Urothelial Carcinoma
Brain Lower Grade Glioma
Breast invasive carcinoma
Colon adenocarcinoma
Glioblastoma multiforme
Kidnery renal papillary cell carcinoma
Kidney renal clear cell carcinoma
Liver hepatocellular carcinoma
Lung adenocarcinoma
Lung squamous cell carcinoma
Pancreatic adenocarcinoma
Prostate adenocarcinoma
Rectum adenocarcinoma
Stomach adenocarcinoma
Thyroid carcinoma
Uterine Corpus Endometrioid Carcinoma
BLCA
LGG
BRCA
COAD
GBM
KIRC
KIRP
LIHC
LUAD
LUSC
PAAD
PRAD
READ
STAD
THCA
UCEC
78
2
91
38
1
160
44
49
3
40
6
49
7
2
28
25
13
140
548
254
112
283
81
98
220
149
24
153
95
69
230
334
i Cà x He e mc s l
n lHe s mp e ’ i p s m
m ind
P m nl
n mf
n mlc’ p u Pi
7 otd3 3n s ivb
w l di n 3a nN3dN3R
niNl s 7K ns 3oniRs nol
i Pn l
s mt7N itb 3a
a Mo da o do 7 s 37 s ts tmt N i don
l
3avi 3tns m iN 7 n s nol
3a
l
3avi 3tns m iN 7 n 3a 3Nl no 7 l di 71
3Nl no 7 l di 7 onl 3a 7 l
d to n don
l
i 3t77N 7 s
3avi 3tns 73 3N7R3a
3avi 3
don
ts 3 s 7t3t 7 n 3a l
3avi 3
s
l
s Ns l
Ns l
3avi 3
71 n
3 R3u n l
3
3an
s u t iv N7
i
niNl s t7 o oo
3n 3a
otMa3 niNl s tiiN73o 3 7 3a
7a
its
3n l
Noo s 3iv o nl l
iiNl ts
ts 7 d to
s nol
iB
don
ns 3a l
I 3n l
7No
s dondn7
7No 3a
3n l
d
ts 3 s 7t3t 7
7No 3a
3 Bm iN Ro s Mts M onl ) 3n 9R
7No 3a d o s 3 M n s
77 v N3titb 7
ii i 7 3 3a ts 3 oonM 3
7
ii
s ts tNl
3avi 3
3an 7 a m
3avi 3tns i m i1 a to73 ns t7
u at a a 7
t73a l
s
3avi 3tns i m i t7 3a s 73tl
n 3at7 d to n don
R ns 3 ts ts MU
d 3t s 31
76 l
7t3 1 a l
ms ns me i
i s 3Nl no1
Rl t
3 3a
nHe
3 1 s tMNo UH 3a o t7 s
3 niNl s t7 s ts 3 os i ns 3oni n 3a
n 73tl
n l
v iiNl ts 6 t tr nm
l
3avi 3tns 1 at7
RH) ) PI1
8U
Materials and Methods Figure 32 | Example of the output generated by the script developed to visualize DNA methylation data from the TCGA. The second method is the log2 ratio of the intensities of methylated probe versus unmethylated probe (Irizarry et al., 2008). We have used the beta value to represent DNA methylation data as it offers a biological meaningful information. For DKO DNA methylation we used already published data from Dr. Peter Jones, stored at Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo), under the code GSM896399 (De Carvalho et al., 2012). TCGA: Gene mutation and copy number alterations We also used the TCGA data to investigate genetic alterations in our genes of interest. We used genetic TCGA data through the cBioportal webtool (http://www.cbioportal.org) (Gao et al., 2013). Other bioinformatic tools We have used DAVID (http://david.abcc.ncifcrf.gov) webtool for Gene ontology (GO) and KEGG pathways analysis (Kyoto Encyclopedia of Genes and Genomes) (Huang et al., 2009). In order to define the RA pathway genes expression in normal colon mucosa we used 94 the Refexa standarized database (http://sbmdb.genome.rcast.u-­‐
Materials and Methods tokyo.ac.jp/refexa). High and moderate expression was assessed using RefExa criteria. For the composition of the survival test, we used 4 gene expression datasets of human colorectal cancers (GEO: GSE14333, GSE12945, GSE17536, GSE17537)(Bailey et al., 2012; Jorissen et al., 2009; Staub et al., 2009) with follow-­‐up information available to analyze the survival rate depending on gene expression. Using PrognoScan algorithm (Mizuno et al., 2009) patients were divided into two groups depending on gene expression levels. High or low expression groups were made using the minimum p-­‐value approach. Survival curves were constructed using the Kaplan-­‐Meier method. 95 Results DNA METHYLATION PROFILE OF AKR1B GENES IN COLORECTAL CANCER DNA methylation of AKR1B1 by bisulfite sequencing Silencing of multiple cancer-­‐related genes is associated with de novo DNA methylation of promoter CpG islands. In this study we have focused on AKR1B1 gene, which has a CpG island (CpGi) associated to its promoter (Figure 33A). We analyzed the DNA methylation status of AKR1B1 CpGi island in a panel of 7 cancer cell lines and in 25 colon tumor samples, as well as in their associated normal mucosa by bisulfite sequencing (see materials and methods for more detail). To cover the 93 CpG dinucleotides that comprise the CpG island, we needed to perform 3 independent bisulfite PCRs, named A, B and C (Figure 33A). The results obtained from the three amplicons within the AKR1B1 CpG island have provided an overview of the methylation landscape of this gene in colorectal cancer (Figure 33). All 25 normal colon mucosa samples displayed a fully unmethylated pattern all along the CpG island (Figure 33B). On the contrary, almost all tumors analyzed (24 out of 25) exhibited hypermethylation. It is important to note that hypermethylation is more prone to occur in the C segment of the CpG island (Figure 33C), where the AKR1B1 transcription start site (TSS) is located (Figure 33A). Clonal analysis of tumor samples with partial DNA methylation showed the coexistence of densely methylated with poorly methylated molecules in all cases, confirming the presence of cell populations with heterogeneous DNA methylation profiles. Most of the colorectal cancer cell lines presented a high grade of DNA methylation in the AKR1B1 CpG island (HCT116, CaCo2, HT29, DLD1 and SkCo1), except for DKO (deficient for DNA methylation) (Rhee et al., 2002) which exhibited a fully unmethylated pattern. Hepatocellular carcinoma cell line HepG2 was also fully unmethylated. In the cell lines analyzed, there was no enrichment in hypermethylation regarding the C amplicon. In most cases the whole CpG island displayed a homogeneous DNA methylation profile. 99 colgc
Scale
chr7:
A
AKR1B1
10 kb
hg19
134,130,000
134,135,000
134,140,000
134,145,000
UCSC Genes (RefSeq, UniProt, CCDS, Rfam, tRNAs & Comparative Genomics)
RefSeq Genes
AKR1B1
CpG Islands (Islands < 300 Bases are Light Green)
CpG: 93
Repeating Elements by RepeatMasker
SINE
LINE
LTR
DNA
Simple
Low Complexity
Satellite
RNA
Other
Unknown
Scale
chr7:
AKR1B1
CpG: 93
134,143,100
134,143,200
134,143,300
500 bases
hg19
134,143,400
134,143,500
134,143,600
134,143,700
134,143,800
134,143,900
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
134,144,000
134,144,100
134,144,200
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
Transcription Factor Binding Sites by ChIP-seq from ENCODE/HAIB
B
C
H1-hESC CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
H1-hESC CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
K562 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
K562 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
A549 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
A549 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
D
IMR90 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
IMR90 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
RNA-seq from ENCODE/Caltech
Classified Translation Initiation Sites
miRcode Predicted MicroRNA Target Sites, Highly Conserved Families
Colonic_Mucosa.Donor_32 Chromatin State Segmentation by HMM from Roadmap
Adult_Liver Chromatin State Segmentation by HMM from Roadmap
Duodenum_Mucosa.Donor_61 Chromatin State Segmentation by HMM from Roadmap
Gastric Chromatin State Segmentation by HMM from Roadmap
Rectal_Mucosa.Donor_29 Chromatin State Segmentation by HMM from Roadmap
Rectal_Mucosa.Donor_31 Chromatin State Segmentation by HMM from Roadmap
ah GG g 3
C C A Dm l mA A Stomach_Mucosa
e yDChromatin
A StateyaD
l byNHMMeD
mAyb 3
C C e yD A
Segmentation
from Roadmap
N l f D l mA yl lay A p2 AmhN D mDmA N a my y 3 l N l
laN mh y N eD b 3
N l f D l mA
Chromatin State Segmentation by HMM from Roadmap
m
C C A e A Dm A h hESC_Derived_CD184+_Endoderm_Cultured_Cells
DDDA yb
hESC_Derived_CD56+_Mesoderm_Cultured_Cells Chromatin State Segmentation by HMM from Roadmap
iPS-20b_Cell_Line Chromatin State Segmentation by HMM from Roadmap
(, ,
iPS_DF_19.11_Cell_Line Chromatin State Segmentation by HMM from Roadmap
iPS_DF_6.9_Cell_Line Chromatin State Segmentation by HMM from Roadmap
Results DNA methylation of AKR1B10 by bisulfite sequencing The DNA methylation level of the promoter region of the AKR1B10 gene was analyzed although it did not have an associated CpG island. The analyzed amplicon was composed by seven CpG dinucleotides (depicted in Figure 34A). We have studied AKR1B10 promoter DNA methylation in the 25 colorectal tumors and in the associated normal mucosa. Seven cancer cell lines were also analyzed. The DNA methylation pattern of AKR1B10 was heterogeneous among the different normal samples. For example, samples N.09.817 and N.CR017 were almost unmethylated, whereas N.09.437 was heavily methylated. In contrast to AKR1B1 promoter, AKR1B10 promoter tended to be methylated in normal colorectal tissue (Figure 34B). DNA methylation of AKR1B10 in colorectal tumor samples showed a great variability among the different samples. Some of them were almost unmethylated such as T.CR010, while T.09.454 and T.09.530 were almost fully methylated (Figure 34C). As it occurred in normal colon mucosa, AKR1B10 promoter tended to be methylated in colorectal tumor samples, but slightly less methylated than in the normal tissue. Regarding cell lines methylation, HT29, DKO and HepG2 displayed a completely unmethylated profile, while SKCo1, DLD1, CaCo2 and HCT116 presented different degrees of DNA methylation (Figure 34D). 101 colgc
Scale
chr7:
10 kb
134,215,000
A
134,210,000
hg19
134,220,000
RefSeq Genes
AKR1B10
134,225,000
134,230,000
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
Repeating Elements by RepeatMasker
Basic Gene Annotation Set from ENCODE/GENCODE Version 7
Your Sequence from Blat Search
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
Restriction Enzymes from REBASE
Scale
chr7:
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
500Raw
bases
HepG2 CEBPB v041610.1 ChIP-seq
Signal Rep 1 from ENCODE/HAIB
134,212,200
134,212,300
134,212,400
134,212,500
134,212,600
AKR1B10
134,212,700
hg1
134,212,800
134,212,900
134,213,000
134,213,100
134,213,200
134,213,300
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
Your Sequence from Blat Search
HepG2 CEBPB v041610.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
Basic Gene Annotation Set from ENCODE/GENCODE Version 7
HepG2 CEBPD v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
B
RefSeq Genes
Repeating Elements by RepeatMasker
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
C
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
Restriction Enzymes from REBASE
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 CEBPB v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
CEBPB v041610.1
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal RepHepG2
2 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 CEBPD v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 2 from
ENCODE/HAIB
HepG2
CREB1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
D
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
ah G. g 3
C Cd A Dm l mA A l
mhh yemA A
yaD l N eD mA yla
3
C Cd
N l f D l mA yl lay A p2 AmhN D mDmA N a my y A A 3 l
N l
laN mh y N eD b 3
N l f D l mA m
C Cd A e A Dm A h DDDA yb
N l f D l mA m
R
( (
C C2 f
t ela R c n9m e gR l cg a nch 9gR e Rot gR uc g9 gR e ge e ge9 9
e (, )
( ( W ( ( WR c 9 S ecl hm
9 gR W S
yaD l y t a A A
v l
e
e
g9 f lo g gR
e o l 9ge c 9 gR Sn9t 9g nv e on FW KA
t 9oc h
h) , ( ( KA
t gRal ge9 colgc
Scale
chr7:
A
134,230,000
AKR1B10
20 kb
134,240,000
134,235,000
hg19
134,245,000
134,250,000
134,255,000
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
AKR1B15
134,260,000
134,265,000
134,270,000
134,275,000
Your Sequence from Blat Search
Basic Gene Annotation Set from ENCODE/GENCODE Version 7
RefSeq Genes
Repeating Elements by RepeatMasker
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
Scale
chr7:
134,233,600
134,233,700
134,233,800
134,233,900
AKR1B15
500 bases
Restriction
Enzymes from REBASE
134,234,000
134,234,200
134,234,300
134,234,400
134,234,500
134,234,600
HepG2 CEBPB134,234,100
v041610.1 ChIP-seq Peaks
Rep 1 from ENCODE/HAIB
HepG2 CEBPB v041610.1 ChIP-seq Raw Signal
Rep 1 Genes
from ENCODE/HAIB
UCSC
(RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
Gene
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 2 from Basic
ENCODE/HAIB
Your Sequence from Blat Search
Annotation Set from ENCODE/GENCODE Version 7
HepG2 CEBPB v041610.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HepG2 CEBPD v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
B
C
RefSeq Genes
Repeating Elements by RepeatMasker
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
Restriction Enzymes from REBASE
CEBPB v041610.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 2 HepG2
from ENCODE/HAIB
HepG2
v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep
2 from CEBPB
ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CEBPB v041610.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 CEBPB v041610.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HepG2 CEBPD v041610.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 2 HepG2
from ENCODE/HAIB
CREB1 v042211.1
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep HepG2
1 from ENCODE/HAIB
CREB1 v042211.1
ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 CREB1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 CEBPB v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
D
HepG2 SP1 PCR1x ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HepG2 SP1 PCR1x ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 1 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Peaks Rep 2 from ENCODE/HAIB
HCT-116 SP1 v042211.1 ChIP-seq Raw Signal Rep 2 from ENCODE/HAIB
ah G2 g 3
C C2
A Dm l mA A l
mhh yemA A
yaD l
N eD mA yla
b 3
C C2
N l f D l mA yl lay A p2 AmhN D mDmA N a my y A A 3l N l
laN mhy N eD yb
3
N l f D l mA m
C C2 A e A Dm A h DDDA yb
c eg ec cR9m e t gRal g
e on FW
h
e 9nt l 9l9 c t Sl c ( ( W Sn9t 9g n g c g9
ecSl ac
got 9nc t Sl chcet el nla g9 mR gR SS n e gR
o t gRal g h mRel
n
9) h
enRaS9t gRal ge9 e 9l9 e ll le c S lh
S ) h u 9( h
Re Rla
( ( , Sn9t 9g nA
)5
y(
m n
9t Sl g la
( ( z m n f ne la
t gRal g A
(, F
hg19
134,234
colgc
C C
N l f D l mA ay A
l
9n ng9 9 ent 9onSn f e9oc 9 c nf ge9 chm oc
9l9n g l
n
9 n e 9t
9t
gl c v
S ge gc v
K vt g ne lc ) W3 9l9 got 9n c t Sl c cgo e
gR
K
F2 t g R
9l9 t gRal ge9 e S t gRal ge9 g 9R9ng9
g c g n9t
t gR9 cKA R
9R9ng 9 cecgc 9
9nt l 9l9 t o 9c cA
g oce R
c neSg vc
R f
t g ne lc t gR9 cKA
lar
gR
t gRal ge9 e gR
Sn9t 9g nh gR c t
9f n
a ( ( Sn9
S ecl n e9 Sn f e9ocla cgo e
c e gR llot e 3W,
cc9 e g
megR gR
a ecol eg
t gRal ge9 ( (
A Rec n e9 ec
nn ac v e on FzKA
"#
Methylation status for region (AKR1B1): 7:134143116−134144063
Number of probes: 11 | Number of samples: 373
0.8
1.0
Methylation levels (Control Analyte)
beta.value
0.6
!
0.4
#
$
"
'
#
$
$
&
&
#
'
0.0
0.2
&
"
"
!
#
!
&
"
$
#
$
&
$
"
&
#
!
%
'
#
!
#
'
"
$
'
#
'
%
'
!
"
#
#
%
'
'
#
%
!
&
$
ah G5 g
ml G
CC DDaN A . 2d N l f D l mA ehm y Dm l
Al
yaD l N eD mAy r R A 3 y Dym e l b
9 ent e 9on Sn f e9oc 9 c nf ge9 ch
( (
C C e
S
ecl yD A b
Dm l mA
m c 9t Sl g la
o t gRal g
v e on F7KA
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Genomic Position
(, 3
e 9nt l 9l9 llc RaS nt gRal g
e 9l9n g l
n e 9t
colgc
Methylation for probe cg13801416
CONTROL
NORMAL
0.2
1.0
0.6
0.4
0.2
1.0
0.8
0.4
Methylation for probe cg06864853
1.0
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.4
0.4
0.4
0.0
beta.value
0.6
0.8
1.0
1.0
Methylationfor
forprobe
probe cg02215070
cg13801416
Methylation
0.8
0.8
beta.value
beta.value
0.6
beta.value
NORMAL
0.0
0.0
0.0
0.2
CONTROL
0.2
0.2
0.4
0.2
TUMOR
NORMAL
NORMAL
TUMOR
TUMOR
CONTROL
NORMAL
TUMOR
SAMPLES
SAMPLES
SAMPLES
Methylation for probe cg10795359
Methylation for probe cg14629509
1.0
0.8
beta.value
0.0
0.8
0.2
0.4
0.6
beta.value
0.4
0.2
0.0
NORMAL
CONTROL
TUMOR
NORMAL
TUMOR
CONTROL
NORMAL
SAMPLES
SAMPLES
N l f D l mA h yaDly mh l
Am h AmN
l y lb
TUMOR
SAMPLES
CC ehm y mAl A
Al
C C e
yD A
Al
9n 9f n m m n
l g9 f ne a gR
S ecl A n9
0.4
e gR
beta.value
0.6
ah G0 g
mDmA
Methylation for probe cg09957386
0.6
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
1.0
0.8
0.6
beta.value
0.4
0.2
0.0
CONTROL
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.8
1.0
SAMPLES
Methylation for probe cg09957386
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
) Khm n RaS nt gRal g
c l9 g
( ( K c Sn f e9ocla 9 c nf
e N l f D l mA ay A
c 9
( (,
9 gR
e S ecl t Sle 9 v( h
vFh3hWhzh
v e on F7KA
l
hgR n m n socg gm9 t gRal ge9 Sn9
CONTROL
gR Sn9t 9g nA R cego ge9 9 gR gm9 Sn9
e t Sle 9 a ecol eg
0.0
C Cd
e n ge l RaS nt gRal ge9 S gg n c gR g
og l cc gR gR 9 c Sl
0.2
NORMAL
0.2
NORMAL
SAMPLES
0.6
0.6
0.8
1.0
CONTROL
TUMOR
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
CONTROL
CONTROL
TUMOR
CONTROL
0.0
0.0
NORMAL
SAMPLES
NORMAL
NORMAL
0.6
0.8
0.6
0.2
0.4
0.4
CONTROL
0.0
CONTROL
CONTROL
beta.value
0.6
0.0
TUMOR
Methylation for probe cg21079345
n Sn c g
TUMOR
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
1.0
0.6
0.8
1.0
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.2
0.4
beta.value
0.6
beta.value
0.4
0.2
0.0
NORMAL
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
gR
NORMAL
SAMPLES
Methylation for probe cg10795359
SAMPLES
1.0
CONTROL
TUMOR
SAMPLES
Methylation for probe cg16132520
CONTROL
7h2h5h( ,
0.0
0.0
0.0
TUMOR
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
g99u Sl
0.8
1.0
0.6
0.4
0.2
0.8
0.2
0.4
beta.value
beta.value
0.6
0.6
beta.value
0.4
0.2
0.0
NORMAL
SAMPLES
Methylation for probe cg18416881
0.8
1.0
CONTROL
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.8
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.8
1.0
1.0
Methylation for probe cg24059115
0.8
1.0
Methylation for probe cg13056495
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
c c Sl
NORMAL
gR e e ge9 9 gRSAMPLES
e n
e on F2 A
(, W
colgc
0.6
0.8
1.0
Methylation status for region (AKR1B10) 7:134212304−134212575
Number of probes: 2 | Number of samples: 373
Methylation for probe cg25171118
0.8
1.0
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.6
beta.value
0.4
0.2
0.0
0.0
0.2
0.2
0.4
beta.value
0.6
0.4
0.8
beta.value
1.0
Methylation for probe cg12001930
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
CONTROL
NORMAL
TUMOR
CONTROL
NORMAL
0.0
SAMPLES
TUMOR
SAMPLES
ah G4 g 3
lo m DDaN A . 2d
N l f D l mA ehm y Dm l
Al
C Cd ehmN ml hb 3
mDmA
Am h AmN
N l f D l mA h yaDly mh l lo mehm y mAl A
Al
C Cd
ehmN ml hb
Sample
ce gR
t gRal ge9 g m 9 c nf
c t Sl c v) , yW, b t gRal ge9 K cSeg gRec R g n9 ega m
t gRal ge9 one gR
me
f ne ge9 e 9nt l
f t 9n e got 9n c t Sl c v e on F2 KA
9ol
9g 9 c nf
n9oc Sn9 ccA R c
l n g a e g t g R megR 9on Sn f e9oc
ecol eg n colgcA
C C2
N l f D l mA ay A
R n m n gm9 t gRal ge9 Sn9
l
c megRe gR Sn9pet l Sn9t 9g n 9 gR
( ( W A R cego ge9 9 gR gm9 Sn9
Se g
e 9l9 t o 9c
c t Sl c ecSl a
9 gR c t Sl
(, z
gR e e ge9 9 gR
n
e on F5 A
( ( WSn9t 9g n pRe eg
9nt l
c Re R 9nt 9
v S e n g f lo c 9
9 gR
Sn9
t gRal ge9 e gR
9 ce n Kh mRel
t gRal ge9 gR g m c t 9n 9n l cc c f n
Sn9
v e on F5 KA
got 9n
S e colgc
0.6
0.8
1.0
Methylation status for region (B15promoterlong) 7:134233618−134234063
Number of probes: 2 | Number of samples: 373
!
Methylation for probe cg09322278
1.0
0.8
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
0.6
beta.value
0.4
0.2
0.0
0.0
0.2
0.2
0.4
beta.value
0.6
0.4
0.8
beta.value 1.0
!
Methylation for probe cg03515829
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
CONTROL
NORMAL
TUMOR
CONTROL
NORMAL
TUMOR
SAMPLES
0.0
SAMPLES
ah GT g 3
lo m DDaN A . 2d
N l f D l mA ehm y Dm l
Al
C C2 ehmN ml hb 3
mDmA
Am h AmN
N l f D l mA h yaDly mh l lo mehm y mAl A
Al
C C2
ehmN ml hb
Sample
C C
N l f D l mA A ylmmDy
t gRal ge9 e9t nu nc 9n 9 ye f cef
Sn onc9nl ce9 c R f
e ne gR
oc
n
t gRal ge9 9 gR
megR gR t g R
ll gR cg99lc n9t
lar
A gR
S
g ne lc ecl 9 9nc lar
e t gR9 cKA
( (
n i o g pRe eg
S
S
n9t gR cg99l
m n
l A
olla o t gRal g
g gR g m R f
9g 9 c nf
a 9 gR 9nt l 9l9 t o 9c c t Sl c
c 9 e l t t g9na 9m l ec c c t Sl chm
t gRal ge9 h c egR SS gR
c t g megRe gR
gR g ecSl ac t 9n
S gg n v e on 3, KA Rec 9e e c megR gR
t gRal ge9 9 gRec
f lo g
got 9ngecco mR egm c f el
R lgRa
oc
c m e 9cge t nu n 9n 9l9n g l
RaS nt gRal ge9 v e on FF KA R n colgc 9 g e 9t S n
m
a ( 3F cg99l c t Sl c vc
gRec ec gR 9 n le e l Sn ge h lgR9o R 9
cg lecR A
( ( S ecl 9R9ng 9t S9c
lar
ecl onn g Sn9g9 9lc e e9t nu nc R f
t gRal ge9 9 gR
pg cef la cgo e A e n gS lc R f
gg t Sg g9 et Sn9f
e eg
e 9cec 9 9l9n g l
e 9g9 c nf
megR R lgRa 9 9nc v e on 3, KA
(, 7
colgc
A
Scale
chr7:
134,143,100
134,143,200
AKR1B1
134,143,300
500 bases
hg19
134,143,400
134,143,500
134,143,600
134,143,700
134,143,800
134,143,900
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
134,144,000
134,144,100
CpG Islands (Islands < 300 Bases are Light Green)
CpG: 93
Human mRNAs from GenBank
Human mRNAs
Transcription Factor Binding Sites by ChIP-seq from ENCODE/HAIB
B
Normal 58
H1-hESC CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
Normal 1
H1-hESC CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
K562 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
K562 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
C
A549 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
A549 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
IMR90 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
IMR90 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
RNA-seq from ENCODE/Caltech
ah . d g 3
AmN
Dm l mA m l
N eD mA o l A l
C C e
yD A b 3
N l f D l mA h yaDly m 24 ylmmD
y N eD y hmNClassified Translation
Dl f Initiation
mAmSites
hyb 3
N l f D l mA h yaDly m i
miRcode Predicted MicroRNA Target Sites, Highly Conserved Families
ylmmDy N eD y hmN A D N N lmhf mo Colonic_Mucosa.Donor_32
D y y r Chromatin
3e State
l Segmentation
Alyb by HMM from Roadmap
n e cg99l
c t Sl c n9tAdult_LiverS Chromatin
ge StategcSegmentation
megRby HMM from Roadmap 9t hm
9ol
g g
t gRal ge9 e ( F 9og 9 FWc Duodenum_Mucosa.Donor_61
t Sl chmReChromatin
R nStateSn
c gc
g ge9 n g 9 F7A( 3
Segmentation by HMM from Roadmap
Gastric
Chromating
State
b v e on 3( KhmRe R ec cet el n g9
mR
mSegmentation
9o by HMM from
e Roadmap
gR cg99l
n e 9t
S ge gcN( z 9og 9 3W
c t Sl chmRe R n Sn c gc FWAWWb
Rectal_Mucosa.Donor_29 Chromatin State Segmentation by HMM from Roadmap
n g v e on 3( KA
Rectal_Mucosa.Donor_31 Chromatin State Segmentation by HMM from Roadmap
Stomach_Mucosa Chromatin State Segmentation by HMM from Roadmap
hESC_Derived_CD184+_Endoderm_Cultured_Cells Chromatin State Segmentation by HMM from Roadmap
hESC_Derived_CD56+_Mesoderm_Cultured_Cells Chromatin State Segmentation by HMM from Roadmap
iPS-20b_Cell_Line Chromatin State Segmentation by HMM from Roadmap
iPS_DF_19.11_Cell_Line Chromatin State Segmentation by HMM from Roadmap
(, 2
iPS_DF_6.9_Cell_Line Chromatin State Segmentation by HMM from Roadmap
c t Sl c 9
g ge9 134,144,200
colgc
A
Scale
chr7:
AKR1B1
CpG: 93
Human mRNAs
134,143,100
134,143,200
134,143,300
500 bases
hg19
134,143,400
134,143,500
134,143,600
134,143,700
134,143,800
134,143,900
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
134,144,000
134,144,100
134,144,200
CpG Islands (Islands < 300 Bases are Light Green)
Human mRNAs from GenBank
Transcription Factor Binding Sites by ChIP-seq from ENCODE/HAIB
B
H1-hESC CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
H1-hESC CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
C
K562 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
K562 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
A549 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
A549 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HeLa-S3 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
HEPG2 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
IMR90 CEBPB IgG-rab ChIP-seq Peaks from ENCODE/SYDH
IMR90 CEBPB IgG-rab ChIP-seq Signal from ENCODE/SYDH
RNA-seq from ENCODE/Caltech
Classified Translation Initiation Sites
miRcode Predicted MicroRNA Target Sites, Highly Conserved Families
Colonic_Mucosa.Donor_32 Chromatin State Segmentation by HMM from Roadmap
Adult_Liver Chromatin State Segmentation by HMM from Roadmap
Duodenum_Mucosa.Donor_61 Chromatin State Segmentation by HMM from Roadmap
Gastric Chromatin State Segmentation by HMM from Roadmap
Rectal_Mucosa.Donor_29 Chromatin State Segmentation by HMM from Roadmap
Rectal_Mucosa.Donor_31 Chromatin State Segmentation by HMM from Roadmap
Stomach_Mucosa Chromatin State Segmentation by HMM from Roadmap
hESC_Derived_CD184+_Endoderm_Cultured_Cells Chromatin State Segmentation by HMM from Roadmap
ah . C g 3
AmN
Dm l mA m l
N eD mA o l A l
C C e
yD A b 3
State Segmentation by HMM from Roadmap
N l f D l mA h yaDly m G2 ylmmDy N eD iPS-20b_Cell_Line
y hmN Chromatin
Am
N e l Alyb 3
N l f D l mA h yaDly m . 2
ylmmDy N eD y hmN h AmN e l Alyb
hESC_Derived_CD56+_Mesoderm_Cultured_Cells Chromatin State Segmentation by HMM from Roadmap
iPS_DF_19.11_Cell_Line Chromatin State Segmentation by HMM from Roadmap
iPS_DF_6.9_Cell_Line Chromatin State Segmentation by HMM from Roadmap
(, 5
colgc
lc9 R u
c t Sl c S en
c t Sl chgR
gR
t gRal ge9 cg goc 9 7 gecco
megR gR
9nn cS9 e t gR
3( KA gR 9gR n gm9
t gRal ge9 e gR cg99lc
cSeg gR
n e 9t
v e on 3) KA W 9og 9 7
t gRal ge9 cg goc 9 gR cg99lc
vS ge gc ) 7h) 2h) 5hF2 FF cg99lc
9l9n g l
c cm
n e 9t
gR g9 gR gecco
9ol 9g
g g
m c t gRal g
vS ge gc
F3KA
ah . p g
N l f D l mA A Df y y m ylmmD
y N eD y A e h
SSle
e 9cge t nu n
e ge9 megR 9gR nch ongR n cgo e c t ocg
S n 9nt
g9 olla cc cc gR e 9nt gef cc 9 gRec
C C e
9 ce ne 9t
yD A N l f D l mA A ml
gR g Snet na got 9nc n9t
h
t gRal ge9 S gg n c 9o n Sn9 o
e 9gR n
Sn9 el c e g9g l 9 ( z got 9ngaS c f el
((,
t gRal ge9 c
g nt e ge9 A
A hy
l9c la n l g
9t t 9 t 9l ol n ce gon c v o R nl S ge d
gR g gR
( (
l yya b
lgR9o R gR c n colgc coSS9nggR
9n e elega 9
h AmN
e n gaS cA R n 9n m
gecco c g g9 cR n
R l nh ) , ( ) Kh m S9cgol g
9l9 lar
l gRn9o R gR
n 9ol
gR
cet el nla
t gRal ge9 S9ng lA
colgc
D CG g h Al
m aN A A hb
m
y yo l
f e hN l f D l mA A l
h 9g9 la 9l9 h og lc9 n got
RaS nt gRal ge9 9
nc 9 gR
C C e
cg9t
yD A
h Al lf e y
R got 9nc Sn c gRe R n g c 9
( ( hmRe R S9e gc 9oggR n l f e cgef gn gv
A C5
9 gRec lg n ge9 e l ( FKA
N l f D l mA A N
R
ecgne oge9 S nge lla 9 c nf
gRn
l9 ge9 9
(
gm Rot c mRel e t e gR n
S ecl 9t
c 9 Rot chgR n
n 9onA cSeg gRec e n cS e ch c n cet el nla 9ne g g
h
t ela megRe gR
t 9oc A gR
9 cet el nege cNgR l9 oc 9 g e c socg 9 gR c t
hgR n
n
n Sl ga
e gR mR9l n e9 e f t 9n hgR e R 9ne ec
9gR
c n
v e on 3FKA
ah . G g AmN
yt a h yb
Dm l mA m
C
A y A aN A A N may b e
yD A y h h eh y Al
y
(((
colgc
oc 9 gR
9 c nf ge9 9 gRec n e9 gRn9o R f 9loge9 hm m9 RaS nt gRal ge9 SR 9t 9 m9ol
e
g9 lar
v
t gRal ge9 9 gR
t Sle 9 cgo e
ecol eg Snet nc c i o A
e gR
lc9 9 on e t 9oc A 9 f lo g eg m
S
ecl ( ( t 9oc 9ngR9l9 K e 9nt l 9l9 h
t e c t Sl cA R
33A
gR
n
9t
cc9 e g
a ecol eg c i o e ec
c n n l g
p A
e gR
Se g
megR
n e 9t
e e on
mm10
34,300,000
NAs & Comparative Genomics)
34,305,000
34,310,000
34,315,000
Akr1b3
Bases are Light Green)
y RepeatMasker
enes
m GenBank
efSeq, GenBank, tRNAs & Comparative Genomics)
nds (Islands < 300 Bases are Light Green)
epeating Elements by RepeatMasker
RefSeq Genes
Mouse mRNAs from GenBank
B
Restriction Enzymes from REBASE
C
Non-Mouse RefSeq Genes
ne Predictions - archive Ensembl 70 - jan2013
Mammal Basewise Conservation by PhyloP
Multiz Alignments of 60 Vertebrates
D
from REBASE
Seq Genes
hive Ensembl 70 - jan2013
Conservation by PhyloP
60 Vertebrates
ah . . g 3 ah A
A Dm l mA A l
mhh yemA A N eD mAy yla
f yaD l
y ta A A b 3
N l f D l mA yl lay A . AmhN D mDmA N a my y1 3 2
AmN y N eD y
A A 3.
h AmN y N eD y A l
p5 DDDA rN may mDmA A h DDDA 3b
lar
9on 9nt l 9l9 c t Sl ch ef
t Sl c m n Sn9f e
(()
a gR l
9n g9na 9
9t c o n
gll v
9on
n e 9t cA
h n l9 K Results brief description of the characteristics of the samples is presented in Table 14 in the materials and methods chapter. As occured in humans, the CpG island was not methylated in normal colon samples (Figure 44B). There was also a complete unmethylated pattern in the adenomas studied (Figure 44C), unlike the human counterpart that were frequently hypermethylated (Figure 33). In the case of the CT26 carcinoma cell line and “carcinoma4” sample, we could not detect any DNA methylation in the region analyzed; whereas in “carcinoma1” and “carcinoma3” we detected partial methylation of a few CpG sites; and finally “carcinoma2” was fully methylated (Figure 44D). These results confirm that the hypermethylation of AKR1B1 CpG island in colorectal cancer is also observed in its murine ortholog Akr1b3, however it seems to occur less frequently than in human, at least in regard to the murine cancer model used. 113 colgc
S
ecl Sn9t 9g n
cc9 e g
megR
RaS nt gRal ge9 ec
cel e A 9
RaS nt gRal ge9 e (
i
( (h
y
lacec 9
c c lar
Snet nc oc
N
9n
g9 gR
c m S n 9nt
e gRec cgo a SS
n
gR
( (,
ne gR
p
9t t 9 la
( ( S ecl pSl9n g9na g n gy en g
( ( We gR ) W 9nt lI got 9n
t gRal ge9 v e on c FFh F3 neh yy mA m
c p t e g nt e SR 9t 9 FWKA R
pSn cce9 A
C C A mDmh l D A hy N eD y
9 ce n
I ec Re R n9nl9m ngR ( oSn ol g
9n 9m n ol g
mR gR l9 )
y( hn cS gef laA
A
B
"
"
"
"
ah . 2 g neh yy mA m
C C A l 2d e h y N eD y A Df K b 3 h D l i
neh yy mA h e
o h AmhN D mDmA y N eD y h mDmh
A D l h A A laN mh y N eD y A h b 3 m p rc 3R
o h A
l i i Da y N A mo Ah aD l mA A emy l i mA y aeh aD l mA m l
neh yy mA A
laN mh mN e h o l l AmhN D mDmA l yya b
N l f D l mA i Da y h l
i h
m l o mD
e yD A
N l f D l mAb
((3
colgc
ue gRec gRn cR9l e g9
v33b KS ge gc 9o gh
oSn ol g
( ( m c 9m n ol g
e F 9og 9 ) Wv( ) b KS ge gcA c
n colgc 9 ent gR g gR c t Sl c gR g ecSl a Re R n
lc9 gR 9 c gR g n t 9n
c t Sl c n
R n l gef
3W A
Rel
n la
pSn cce9 f lo c 9
e e on 3W gR
9nt lI got 9nn ge9 N
9n f oSn ol g
( ( 9n
n Sn c g
gm S en
mR9l hgR c
t gRal ge9 f lo c n
9m n ol g N9 gR
g
e ( ( 9og 9 ) W
9 gn nahS99nla t gRal g
v
, ( F S ge gK v e on 3WKA
R c t Sl
f lo c
n
Se g
9nn cS9 e g9 gR
e on
l9 )
c t Sl cA
neh yy mA A Df y y m
C Cd A
C C2 A mDmh l D A h
y N eD y
9 gn cg megR
t 9n
Sla
( ( hgR
g
v) F 9og 9 ) WK pRe eg
one pSn cce9 9
gR
( (,
( ( W c m c
n9oc Sn9 ccA5) b S n
9m n ol ge9 9
3b v( 9og9 ) WK9 gR t m n oSn ol g
( (,
v e on 3z
A
B
C
D
g 9 gR
c c
pSn cce9 hmR n c socg
KA
E
"
"
"
"
ah . 5 g N
neh yy mA m
C Cd A
C C2 A l p2 e h y N eD y A Df K b A 3
h Dl i
neh yy mA h e o h AmhN D mDmA y N eD y h mDmh
A D l h A A laN mhy N eD y
A h b
A
3 m p r c 3R o h A
l i i Da y N A mo Ah aD l mA A emy l i mA y
aeh aD l mA m l
neh yy mA A laN mh mN e h o l l AmhN D mDmA l yya b yaN N h K y l
e yD A
C C
N l f D l mA yl lay mh
y N eD b
((W
Results The scenario is very similar for AKR1B15, which was downregulated in 88% of the samples (22 out of 25) and there is just 1 case out of 25 where the gene was upregulated (Figure 46B and D). The behavior of AKR1B10 and AKR1B15 was almost identical in all samples analyzed, which may indicate a co-­‐regulation of both genes. As we have previously reported, promoter DNA methylation of AKR1B10 and AKR1B15 is almost unaltered during cancer development despite both genes display a dramatic downregulation. However, we could observe a link between the DNA methylation of AKR1B1 CpG island and the expression of AKR1B10 and AKR1B15 (Figure 46E): hypermethylation of AKR1B1 CpG island was associated with strong downregulation of both genes (Figure 46C and D). An exception is CR010 patient, in which, despite the low expression of both genes in the tumor tissue, there was an apparent upregulation explained by the lack of expression of AKR1B10 and AKR1B15 in the normal colon. mRNA expression of AKR1B family in a second set of samples Next, we extended our study by applying microarray expression analysis to a cohort of 100 patients from the Institut Català d’Oncologia (ICO, Barcelona), with the corresponding normal and tumor sample for each patient. Additionally, to discard possible sample cross-­‐contamination and in order to identify possible disease-­‐related features in the normal tissues of patients, this analysis included the colonic tissues from a control cohort comprising 100 healthy individuals. We then analyzed the expression of AKR1B1, AKR1B10 and AKR1B15 in the expression microarrays. Although AKR1B1 displayed a modest tendency to be downregulated in colorectal cancer, this was not a general fact and many tumors did not change the expression levels and in some of them were even upregulated (Figure 47). On the contrary, AKR1B10 and AKR1B15 showed high expression values in normal colon that suffered a dramatic downregulation in almost all tumor samples analyzed (Figure 47). 116 colgc
ah . 0 g
C CR
C Cd A
h ABAmhN D
Dl f N a my hmN
laN mh A A h B mDmA laN mhy N eD b
g ec m9ngR g9 9g gR g e ( ( W pRe eg
t o 9c hc9t
C C2 A neh yy mA A G
h Al y ly m y N eD yb A
Dl f mAmhyR Da BAmhN D mDmA N a my yym l
lml
ll c t Sl c n9t R lgRa
Re R l f lc 9 t
c t Sl c ln
a ecSl a
9 9nch
( (,
pSn cce9 mR n c e 9nt l t g R
l9m pSn cce9 l f lc e gR n 9
got 9ngecco cA
((7
colgc
N
neh yy mA m
C
N Df A
9 cgn gR 9on 9 loce9 chm
( (,
( ( We t ge9 y N eD y
e
g9 lar gR
pSn cce9 9
gRen c g 9 c t Sl c 9nn cS9 e ( (h
g9 gR
ln
a
Sn9s gv e on 32KA
AKR1B1
Rectum
Colon
Cec.Ad
Col.Ad
Normal tissue
Col.Mu
Rec.Ad
Rec.Mu
AKR1B10
Rectum
Colon
Rec.Sg
Col.Ad
Cec.Ad
Normal tissue
Cancerous tissue
Col.Mu
Rec.Ad
Rec.Sg
Rec.Mu
Cancerous tissue
AKR1B15
Normal colon (19)
Normal rectum (3)
Cec.Ad: Cecum adenocarcinoma (22)
Col.Ad: Colon adenocarcinoma (101)
Col.Mu: Colon mucinous adenocarcinoma (22)
Rec.Ad: Rectal adenocarcinoma (60)
Rec.Mu: Rectal mucinous adenocarcinoma (6)
Rec.Sg: Rectosigmoid adenocarcinoma (3)
Rectum
Colon
Col.Ad
Cec.Ad
Normal tissue
l
ah . 4 g neh yy mA m
l
y 3b aN
C
A
N Df A
hy A e h Al y y A
c Sn f e9ocla 9 c nf
h
n e 9t c mR n c
t 9cg 9l9n g l
( ( Wm n
n c t Sl c v e on 32KA R n
n e ( (
n cle Rgla t 9n
pSn cc
e gR l n
le hn got got 9nc R f
e gR ot
((2
n
e ct ll co c g 9
Sla
9m n ol g
e n cA 9n p t Sl
e g cge 9gR
got 9nc v e on 32KA lgR9o R gR n
pSn cce9 9m n ol g
e 9nt l gecco n
gR c t
c
n c9t
R n9oSA
c e ( (,
e pSn cce9 ( (W
n gR n gR e gR n got A e gla l cc pSn cc
e n got 9n l9 ge9 heg ec et S9ng g g9 g u e g9
n9 c t Sl c 9n
Rec.Sg
Rec.Mu
Rec.Ad
h Al AmhN D A mDmh l DlaN mh l yya y r hmN
l l AaN hm y N eD y e h hmaeb
( ( m c
( (,
Col.Mu
Cancerous tissue
gR 9l9 ( (
c e 9o g gR
ecS nega
colgc
ce A
AKR1B10
gR
gm9 e S pSn cce9 g cgo e c v cS n
h ) , , 2D unraS r u
m
e ge e
( (,
v e on
e 9l9n g l
Colorectal adenoma
(45)
9og gR
et S9ng 9
n f l9St gA
Colorectal carcinoma
(36)
C Cd A neh yy mA A AmhN DR
h AmN mDmA l yya b
el
bRpdd41 hKfe K bRpdCdb
neh yy mA m
C
N Df
R pg cg S m c g9 lar gR t
e Rg 9l9n g l
v
9 gRec lg n ge9 ( ( , e gR got 9n SS
AKR1B10
N
9t c
35Kh mRe R e e g c gR
S9e gc
ah . T g
AmN
A
hmN
ye h
9
Colorectal adenoma(56)
B
Colon mucosa
(24)
h ) , ( , Kh
9m n ol ge9 nla e e Colon mucosa(22)
g 9
n
A y A
A h
DDDA y
pSn cce9 9
ll le c 9 a 9 c nf
e (
c e R S g9 llol n
n e 9t
S l 9
ll le S ) KA
et el nla g9 mR g m
ln
cR9m R g n9 9oc
pSn cce9 9m f n gR c ne9 R mRe R ec t 9
pSn cc
e gR 9gR n
9l9n g l
l f lc n
c mR m g lu
e ll le cA
S ) R S g9 n e 9t
c v e on W, K nch 9l9n g l
n e S )
c
S9cegef
ll le c
W, KA
S99nla 9n 9g
t Sle e
g g gR
ll le A S ) pSn cc
m c g u n
( ( , v e on W, Kh
( ( W ec f na l g
socg
( ( v e on
9og
) 5h Re Rla e ll le c v e on W, KD go llahm
( ( We (
n g la pSn cc
a i
e pSn cce9 9
Re R l f lc 9
9 gn9l 9n
ll
( (,
pSn cce9 e ongR n pS net gcA
((5
colgc
A
B
# #
%("
##'
C
#
$)
#
$
# #"
$
%("
D
# #&
##'
#
$)
#
$
m g
A yA
cel e ge 9l9n g l
pSn cce9 h mRe R n e R 9n
cn l g
e e 9l9n g l
9t Sl g la
nA
##'
#
$)
#
$
$
A y A
e A Dm
9
a9n
(
l9 c ec l9 oc
n Se ge
9t t 9 SR 9t 9 h ) , , 5KA 9
n cc gRec
t gRal ge9 9
9og ( ) W FW 3 zWu ch n cS gef lah
v e on W( KA
FW 3
n mRel
c g e ch l9 g
e on W( hgR
9l9n g l
9 c i o h ) , , zD
gR
(
$
C C2 r 3
c e gR n e9 hmRe R ec
n v ne 9l
n cg
$
9m n ol ge9 9
9 gn nah eg ec
lar
#
C Dm ay
t olgeSl
i o cge9 m
n9t gR C Cd r 3 A
g9 lo e g mR gR n gR
cS e e 9nh 9 gR
e h
$)
$
ah 2d g N
neh yy mA m
C C r 3R
aN A mDmh l D A h DDDA y A A e pb
neh yy mA m
#
#
%("
%("
##'
pSn cce9 9
pSn cce9 ecSl a
n t e FW 3 m c 9g lg n
e g oSn ol ge9 e o lg n A
t gRal ge9 9gR Sn9t 9g nc egR n e 9nt l m c
n9oc gecco
v e on W( KA
9 loce9 hgR
gR
(
S 9m n ol ge9 9
c l9 och
( (,
( ( Wec 9 e oc gR n ec 9g 9m n ol ge9 9 gR e R 9ne c 9n RaS nt gRal ge9 A Rec et Sle c gR g 9m n ol ge9 9
ec cS e e (),
9g
g9
9 c i o 9
l9 n Se ge cel e A
(
c
colgc
SLC35B4
1.0
0.6
0.4
0.2
NORMAL
TUMOR
CONTROL
NORMAL
TUMOR
0.8
0.6
0.4
0.0
0.2
0.0
0.2
0.4
beta.value
0.4
0.2
0.0
TUMOR
CONTROL
NORMAL
SAMPLES
NORMAL
SAMPLES
SLC35B4
1.0
TUMOR
TUMOR
SAMPLES
C
CONTROL
beta.value
0.6
0.6
0.8
0.8
1.0
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
CONTROL
NORMAL
TUMOR
Methylation for
probe cg25920630
Methylation
probe
cg25920630
CONTROL (N=81)
NORMAL (N=38)
TUMOR PAIRED (N=38)
TUMOR UNPAIRED (N=216)
1.0
NORMAL
SAMPLES
Methylation
probe
cg23601994
Methylation
for probe cg23601994
Methylation for probe cg00033643
BPGM
CONTROL
1.0
B
0.0
0.0
0.2
0.4
beta.value
0.6
0.8
A
0.8
1.0
Methylation for probe cg01627553
BPGM
Expression
0.0
0.2
0.4
0.6
0.8
Expression
NORMAL
TUMOR
CONTROL
NORMAL
TUMOR
ah 2C g 3 AmN Dm l mA A
G2 . b
yt a h y A
l
e
N l f D l mA l m
G2 . A
r h A3R mDmh l DlaN mhrh 3 A l
m neh yy mA m
C
A y
9m egec l ngR ggR
t e la
g
t
ge yl A y m
C Dm ay A l
A
mh A y
A
yD A
yym l
lm ehmN ml hyb 3
mDmh l D
A yb 3 neh yy mA l m ml
A y A
Dl f mAmhy
uAmhN DqN a my yym l r Da 3b
nn g 9m n ol ge9 9 onne ( (,
cet olg 9ocla 9gR
R ect
9n 9gR
( ( Wh og m m9 cA 9
g
c S9ng l vc
9 gR ( , , got 9nc lar
( ( W ecSl a
v, A5F2K n ol ge9 9
9nn l g
megR
pSn cce9 t g ne lc g gR
A c
Cc v, A5FzK
gm9 e g 9 g e e gR
S gR
pSn cce9 g
e on W) h
( (,
lt 9cg e ge l f lo che ( ( WA g ec m9ngR g9 9g gR g
g
n9t
g v e on W) KA R c n colgc co
( ( , egR n
nec
9t t 9 n ol g9na
t gR9 cK Se g
e gRec cel e n cc gR g i o cge9 m oc
f na Re R 9nn l ge9 h ( (,
9l9n g l
n
chmR g m9ol e e g
9l9n g l c gc 9 c t Sl ch gR
9t e e cg
9gR
9y
( ( m c 9g
( ( Wh mRe R 9 ent c e S g
()(
colgc
n ol ge9 9
( ( f ncoc
( (,
( ( W e 9nt l got 9n
9l9n g l gecco A
A
B
C
ah 2p g mhh D l mA m neh yy mA D i Dy lo A
C Cd r 3 A
C C2 r 3 o l
A y
A Da
Al
mDmh l D neh yy mA l br 3 msh aD l mA m
C Cd A
C C2 ay A
l y lyR A mDmA N a my hmN
Dl f mAmhyR A AmhN DN a my hmN mDmh l De l Aly A
Al N l
laN mhb
())
colgc
neh yy mA m
c
c ne
C
S ecl t 9oc A
( ( pSn cce9 9 gR
c t Sl chS9laSc l9 oc ec S nge lla 9 c nf
n 9ngR9l9 c 9gR R f
c e c 9
e n g t 9oc 9nt l 9l9 gR n m c lt 9cg 9 R g9 mR g9 onn
S gm gR
pSn cce9 9 gR
gla 9 gR t gRal ge9 cg goc v e on WFKhcet el nla
e gR Rot e 9gt
( (, A
n e 9t A
e n g c t Sl che n e9 (
mR n c gR n ec 9 t 9oc 9ngR9l9 u 9m 9n Rot gR
gR
mDmA
Sn f e9oclah gR cgno gon 9
gm Rot lar
N Df A N may
( ( mR n
t gRal ge9 9 gR Sn9t 9g n
cel e A
!
!
"
"
"
"
ah 2G g
h Dl i
neh yy mA A
h Al N may y N eD yb
h mDmh A
N l f D l mA yl lay m l
e yD A b D 6 aDDf N l f D l R h f 6 e hl DDf N l f D l
aAN l f D l
A Da 6 aA l hN A b
ec Re Rla
9m n ol g
cel e pSn cc
e f e 9nt l
n e 9t ch e e n c
l yl
Ro l 6
og gecco ch R9m f n eg ec cle Rgla
9 gn cg megR S9laSc mR n gR n ec 9
pSn cce9 e c t Sl c
F5) F5F v e on
W3 KA
n e 9m n ol ge9 e t gRal ge9 cg goc 9
9
h
pSn cce9 h pRe eg
nh cS e lla 9n
S ecl t 9
v e on W3
9 c 9g SS
n g9
cg g a 9
KA R
ggR
pSn cce9 A
()F
colgc
A
B
C
ah 2. g
A
h Dl i
neh yy mA A
h Al N may y N eD yb
h
mDmh A
l yl
N l f D l mA yl lay m l
e yD A b D 6 aDDf N l f D l R h f 6 e hl DDf
N l f D l Ro l 6 aAN l f D l
A Da 6 aA l hN A b
e 9 loce9 ht 9oc
9l9 ()3
c ecSl a e n g
nh c eg 9 onc e Rot cA 9m f nhgR
S ecl t ela
(
9 c 9g cR9m
e gR t 9oc A
n c9
9m n ol ge9 t gRal ge9 cg goc 9
l n cc9 e ge9 megR gR
pSn cce9 9 gR
colgc
neh yy mA m
C C A
9n n g9 9 ent gR
9l9n g l
C Cd l l
pSn cce9 lg n ge9 c 9 c nf
n S ge gch m S n 9nt
c t Sl c 9nt lI got 9nA
lar
ehml A D i D
m cg n e (
c e l9g cc ac 9 ( ) S en
( ( ( ( , Sn9g e pSn cce9 e cep S enc 9 9nt lygot 9ngecco c t Sl c v e on WWKA
A
B
ah 22 g
yl hA Dml A Df y y m
C Cr 3 A
AmhN DclaN mhb
y ay
y may s e A
A b
cg n n t ge
l9g o 9m n ol g
ue 9m n ol g
pSn cce9 v7h 2h ( , h ( (
lacech g9 gR n gR c n colgc m
ec cle Rgla 9m n ol g
e DDDA y A Cp e hy
S9e g
( ) K v e on WW KA
n e 9 9n g megR gR Sn9g e e t 9cg9 gR
A y A.
( ( , hmRe R ec l nla 9m n ol g
e c t Sl ( v e on WW KA
l f lc m n
WW KA
t
9m n ol ge9 9
z c t Sl c lar
t
nSe C Cd r 3
9l9n g l
9 ent
gR
9og gR
e W9og9
( ( m c l nla
ll le c lar
pSn cce9 9 c nf
gR g
hgR
v e on
( ( , ec n t ge lla
nc t Sl c lar
hmRel
( (
m Sn9S9nge9 9 gR t A
C Cd A mDmh l D A hl yya N hm hh f y
pg m e f cge g
oce gR S9cce l
gecco t e n9 nn ac v
o ge9 9
S ) c
S9cegef
c
Sn9 9cge t nu nh
cKe ( 5z 9l9n g l got 9ncA 9n n g9 c g oS gR
9 ege9 c 9ngR et t o 9Recg9 R t ecgnahm oc
( (,
9 gn9l 9
( ( z c pSn cce9 v e on WzK gef
9 gn9l
gR R t g9pale y
()W
colgc
9ce cg e e e 9n n g9 R u gR i o lega 9 gR c t Sl vc
t g ne lc t gR9 cKA
ah 25 g
A
3 l y h eh y Al
3
l
N lmnf DA A my A yl A A r
N N aAm ylm N ylhf o l
e on W7 m
9 c nf
9nt l 9l9 t o 9c
ge 9 aA c Sn f e9ocla 9 c nf
ecSl a
ce Re R 9n
9t Sl g la R9t 9 9oc pSn cce9 9
( (,
gef
FAzb Sn c g
gR ne R f
ot
cel n 9
ll gR c
gef
( (,
9 gn9lc m n
pSn cce9 v e on W2 K
( ( , v e on W2 KA
gR
9l9 got 9nc lar
pSn cce9 9
got 9nc Sn lo
e f cge g 9nn l ge9 c megR n l f g le e l
()z
e oc
9 ent gR g5zA3b 9 gR
9n cgn9 la 9m n ol g
( ( , A
9 gn nah z) AWb 9 gR got 9nc m n
pSn cce9 9
megR gR
l i Df b
( (, A
v e on W2 KhFFA5b R
g m
a
gR g ll 9nt l 9l9 t o 9c
pSn cce9 A gR
S9cegef
CC5Rh ye
a 9gR n t gR9 chgR 9nt l 9l9 t o 9c
gRec g R 9l9 ahm 9 c nf
S9cegef
3m
e p A
C Cd Al m f b
c
gA
( ( , A R Re R
a cg gecge l
lacec g9
colgc
A!
B!
ah 20 g mhN D mDmA N a my
A
yl A A
A
A 3 f o l l
B
ah 24 g n N eD y m l
hl h
mDDmo
lm ymhl l
yl A A b 3A
l i y N eD o l ymN AmhN D mDmA l yya
yl A A b 3emy l i laN mhb
Al m f Al
C Cdb
C
Al
h Al mDmA laN mhy mh
C Cd
mllmN m l
h N Alb 3
ay
()7
colgc
C C A mDmh l D A hl yya N hm hh f y
ce gR c t
9ll ge9 9
( ( A gRec
c m oc
m32,
c R u
c pS g h
n e S9cegef
((z c
yl A A mh
CC5
C C Al m f b
9n gR
pSn cce9 9
( ( m c pSn cc
gR n colgchm 9 c nf
9n
pSn cce9 9
( (
S9cegef
gR Sn9g e pSn cce9 9
9 gn9l 9n pSn cce9 A
o . 4dR h ye
l i Df b 3 A
ag9Sl ct 9 9nt l 9l9 gR g ll 9nt l 9l9 t o 9c
pSn cce9 A gR
3 y l
( ( e gR 9nt l 9l9 v e on z, KA
e gR
( ( v e on z( K ( ( v e on z( KA
()2
lc9 R u
gef v e on W5KA
ah 2T g 3 A
3
N N aAm ylm N ylhf o l
pghm
chm
llcA
9 gn9lc m n
c 9 gR got 9nch 52b n g e socg ) b 9 gR t l9cg gR
gR
pSn cce9 9
colgc
A!
ah 5d g 3 mhN D mDmA N a my
B!
yl A A
A!
ah 5C g n N eD y m l
h l h mDDmo
3emy l i laN mhb 3A
l i laN mhb
3 f o l l
Al m f Al
C Cb
B!
lmymhl l
h Al mDmA laN mhy mh
C C yl A A b
()5
Medias y medianas del tiempo de
supervivencia
Mediana
Intervalo de confianza al 95%
AKR1B1_DEF Límite inferior Límite superior
colgc
,00
1,00
2,00
Global
,000
7,742
7,763
7,875
11,047
9,495
9,540
9,164
cSeg gR l9m ot
n 9 9l9n g l
ce e e glaComparaciones
cc9 eglobales
g megR
n
m9nc 9f n ll Chi-cuadrado
gl
Sig.
Log Rank (Mantel-Cox)
12,151
2
,002
Prueba de igualdad de distribuciones de supervivencia para
diferentes niveles de AKR1B1_DEF.
ec c y n
conf ef l megR
Sy
, A, , , ( hn cS gef la v e on z) KA
Free survival
Time (months)
ah 5p g h
neh yy mAb
( ( 9m n ol ge9 m c
Overall survival
f lo 9 , A, ) c ch
Time (months)
A mi h DDyahi i D A mDmh l D A Kaplan-Meier:
he l Aly Associació
e A A mAOS agrupant
C C ehm
0l+ 1Avs 2
[Conjunto_de_datos1] /Users/Eva/Documents/docs
TTD/CPT.ES1.604/PROJECTE TUMORS/CPT.ES1.604 TUMORS DEFINITIVA.sav
Resumen del procesamiento de los casos
Censurado
AKR1B1 low vs high
Nº total
Nº de eventos
Nº
Porcentaje
1,00
115
78
37
32,2%
2,00
54
33
21
38,9%
Global
169
111
58
34,3%
( F,
colgc
N
C
9
A y
N l f D l mA A
neh yy mA
e e ce Rgc e g9 R9m 9n
9
(
R n g ner
t 9 e e ge9 c v
S n 9nt
lh lN Al
t gRal ge9 m9ol
l9 ochm gn g
R h m
l h2s K
((z
gR
llc megR gR
n e9 a
g gR n ol ge9 t gRal ge lare Recg9 g Wy r
A
S9cggn cl ge9 l
K a Rn9t ge et t o 9Sn eSeg ge9 v R KA 9n 9f nhm
pSn cce9 t gRal ge9 pS net gc 9 gR
(
l9 ocA
R n colgc S9e g
l9 gR
(
mRel
9og gR g gR
l9 ocA
( (,
t gRal ge ( (
S
g 9 Wy r
ecl ecSl a
( ( W Sn9t 9g nc
t
t 9
ec 9g o e 9nt
cg
lt 9cg olla
t gRal ge9 h
t gRal g
g ngR gn gt gv e on zFKA
A. AKR1B1 CpG island
B. AKR1B10 promoter
C. AKR1B15 promoter
ah 5G g hmN ml h
o l 2s K
b
g nt c 9
gn gt gh
N l f D l mA yl lay m l
pSn cce9 h lgR9o R gR
( (
pSn cce9 n t e coSS9ngc 9on Sn f e9oc 9 loce9 gR g
9g t s9n
( ( Wco
z3
ge n ol ge n
Sn9t 9g n
C
A y A
S ecl o lg n
ec l cc t gRal g
t gRal ge9 9
g n gR
v e on z3 KA Rec n colg
( ( pSn cce9 A gR
t gRal ge9 CC5 mAlhmD A lh l
( ( Sn9t 9g n ec
9 gn nah
gR a m n n y pSn cc
( (,
v e on
KA
( F(
colgc
A
B
C
D
ah 5. g
C CR
o l 2s K
b
C Cd A
C C2 h D l i
hmN l A ehm D m l
ecg9 lar
e e R
n colgc S9e g
Sn9t 9g n e vt g ne lc 9og
S9l h
) ( ( zA gn g
t nu e gR
cR9m
((z
ne Rt g 9
llc m
9t e n e9 e lo e l f lc m n t 9
A et el nlah m
nc gn gt g l nla e o
n e9 c n
gef g
ucA R
l9
l
9ol 9g
l9 gR
gef
l9 oc
cS e l n l f n g la ne R
gen n e9 A Rec
((z
( (
e 9 gR t nu
g g F 3t (
llch lgR9o R oS9 gR gn gt gh n t gc 3 Sn9t 9g nA Rec n colg R c
R ( ( , Sn9t 9g nhmRe R ec 9 cecg gmegR
g n gn gt gh ll
gef ge9 9 gR
((z
F 3t F e gR
9ol 9 c nf
gR
ne Rt g 9n F 3t ( S
e Km n
t gR9 cKv e on zWKA
ct ll
gR n y pSn cce9 9 gR cel g
( F)
Sn9g e c v
a Rn9t ge et t o 9Sn eSeg ge9 v R K e megR Wy r
n
CC5 mAlhmD A lh l
C Dm ay
9gR n
gn g
m c
neh yy mA A
t 9 e e ge9 c v F 3t ( h F 3t Fh F ) 7 h F ) 7t FKh Recg9 f ne g ) A e gR
A
l9
ec occ
e gR
(
cS e lla e gR
F )7
( (
megR gR
l g nA R n c F ) 7t F
l9 oc e e 9 gRec Recg9 g m c Sn f e9ocla 9 c nf
megR Wy r
n t nu
oc eg ec cc9 e g
eg mell
l
R z
((z
llchWy r
t 9 e e ge9 e gR l
gn gt gv 9 ne o r
n9cc gR
e 9gR n cel g
h) , , 2KA
colgc
7p21.3
chr7 (q33)
21.1 15.3
$
7p14.3
p14.1 13
11.21 11.22 11.23 7q21.11
($
&$
q21.3 q22.1
7q33 7q34 7q35 36.1
,$
*$
% %
7q31.1
%$$
%&$
% %$
% %)
%
&
F.3
%
(
)
-
%$
*
,
+
"
#
$
"
%
!
'!
' &+
#
$
"
%
$
"
"
!
' (
'!
!
!
' (
!
' &+ !
!
!
$
"
!
!
$
"
#
$
"
%
$
#
"
!
!
%
"
$
#
!
%
"
$
#
#
$
"
%
!
&
$
"
%
!
#
$
"
%
!
!
!
$
!
%
"
$
#
%
"
$
#
"
!
&!
$
"
$
"
!
$
"
ah 52 g hm D y m
emD R p b R
pR
A .
ylmA N h y p p0N GR p p0 R
G .N G A
G p0
A mAlhmD A 2s K
lh l
CC5 DDyb
i Da y h eh y Al l
A
l i mAlhmDb hmN l A N N aAmeh e l l mA i Da y h y mo A y l
Ah N Al h l mA mi h
Aealb hhmh hy A
l yl A h
i l mAb
( FF
colgc
S9 gn gt gh gR n m c
( (
)
v n t g ) K S9l
n t ge l9cc 9 ce l e n t g 2 v
( ( W Sn9t 9g nK
no gn gt gh gn c neSge9 l n
m n
9t S e
) hmR n gR S
S
a gR l9cc 9
e n t g F e gn g
e g cegaA g n cge lah
gef ge9 9
l9 gR
I
e A 9n
( (,
( (W
gen l9 och p Sg 9n n t g
u ec lt 9cgo lg n A
lc9 lar
( (
e t 9cg 9 gR n e9 h p Sg
( ( WSn9t 9g nhmR n gR n m c
) A h gR n m c
llch mRel
oS9 l9cc 9
gR
ecl e e e 9
y
( ( zh
y1 gn c neSge9 ( ( z gn g
g9nc g9 gR
megR Wy r
S )
v e on zzKA
K
ah 55 g
A
R
e e s A
e pb
c
s| ehm D A l
9
9gR
( ( , h c eg mell
e e 9
9gR
g9 gR
( ( , A S ) pSn cc
A 2s K
A AmhN Dl yya y A
9on R yc i
cn l g
a t e c 9 gR
e e o gn g
gef
R o l 2s
pSn cce9 9
gR g gR
c e n c
R
gR
gR Re R nf lo c 9
S eA
lh l
So le la f el
l
n Recg9 Wy r
e on z7 hgR n ec S
socge gR9c gecco c mR n
( F3
g
CC5 lh l
n m 9 c nf
megR Wy r
( (,
G p0
9nt l gecco c CC5R
ec no e l 9n gR
mR m gn g
9 gR
g el gR
A
cR9m e pg pS net gcA
e e h cS e lla e gR ne Rgce
oc
yD A
( ( S ecl m c e n c
pSn cce9 9
C C e
u9
CC5
R yc i
t nu
gn g
F )7
( ( , ec pSn cc
N 9l9 DDDA
g g9 lar e F )7
((z
e gR
A h
e e n g
llcA
( ( S ecl ct ll e g cge A
(
$
*
& "'
(
$
*
(
$
*
& "'
(
$
*
&
colgc
&
A
+
+ +
Scale
chr7:
!
"
Pol2A
Pol2C
k27me3A
k27me3C
k9me3A
k9me3C
CTCFA
CTCFC
H3K4me3A
!
#
H3K4me3C
*
RNAseqA1
. **
+***
!
RNAseqA2
+. **
, ***
"
RNAseqC1
RNAseqC2
k36me3A
B
k36me3C
+ +
AKR1B1
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
+
CpG Islands (Islands < 300 Bases are Light Green)
!
"
k27acA
+ +.
++/ . ' $
+ +*
+ +
k27acC
*
++/
*&,
*&-
*&/
!
*&0
+
"
Basic Gene Annotation Set from ENCODE/GENCODE Version 19
AKR1B1
AKR1B1
k4me1A
ah 50 g 3 D i Dy m G p0
h
e l
A
h Al AmhN D l yya y rD l e A D3 A l
neh yy mA m
C C A
C Cd A l my l yya y rh l e A D3b 3 G p0
A mAlhmD
CC5
mDmh l D A h DD DA
A lh l
o l 2s K
rD l e A D3b A
neh yy mA m
C CR
C Cd A
C C2 A mAlhmD A
CC5 2s K
lh l
DDy rh l e A D3b
k4me1C
t 9 cgn g
((z
v e on
9l9n g l
F )7
pSn cce9 9
R c n colgc l
Sn9t 9g nn e9 9 t nu e gR
Methylation450K
Methylation450K
n
ll le z7KA g n cge lah gR
gn gt gmR gl
gR
l9cc 9 gR
9t S n
F )7
S
g9 gR n y pSn cce9 9
( (
SS
nc g9
oc g9 RaS9gR cer
( (,
o Sog gef
oce megR gR 9nt l 9l9 t o 9c
u ec n 9f n
( (,
g
R g n Wy r
( ( WA a gR
n
Se ge
gef ega 9 gR
9 gn cgh
R cA
( (
( ( W c n ol ge9 A gR nt nuc m n
lc9 lar
( ( Sn9t 9g n n e9 e R yc i
g v p KA
( FW
colgc
D
9n n g9 lo e g gR
9 9nt ge9 l cgno gon 9 gR
megRe gR o l och m
pS net gc vc
c
t g ne lc ( ( , ec gR
t
S n 9nt
Rn9t 9c9t
9 9nt ge9 gR g R c egc pSn cce9 t 9n
m c g gR
9 cg g Snet n 9 egc Sn9t 9g nA m9
( (,
( ( , Km n
lar
n i o a t S 9
encg l 9gR
( ( z n a gR
g co
cg
R m c pSn cc
S ) vcR9me S ) e gR
9ol 9 c nf
lahgR nn9m S
z2KA R c
hm
n t ge llah gRec
c e Rn9t ge 9n 9gA 9n gRec n c9 h
ll le c
( ( z vmegR oll
Re R gn c neSge9 l f lc 9
pSn cce9 9
e g n ge9 9 g g
t ela l9 ocA
gR g gR S gg n c m n
u 9 9 g gec l9 g
cgn9 (
lt 9cg gR c t
e ( ( , m c 9t Sl g la e n gA
n9o gR
gm gR
S ecl v e on
( ( , Sn9t 9g n
( ( S ecl A
ah 54 g G
l ml
C C Dm ay A
CC5 A
e p DDDA yb
Dm Cd h t a A f m l
Al h l mAy o h y l
ny yl
yl A m
eh N hRDm l
Al
C Cd ehmN ml hb
( Fz
t ela
Sgon vF K
a gR F g R ei o A e on z2 n Sn c gc gR
ll le c f gR9o R gR
t nu
K n c9t
S e g
9 mR gR n gR
co
cgno gon
9
t gR9 cKA
oc gRe u gR g egc Sn9t 9g n 9ol
cel e ( (
n y h eh y Aly l
emAl lml
mAyl Al
colgc
e on z5 m
9 c nf
l9c
R f
n Sn c g
A R t 9cge g n cge g9 gR
( (
e g n ge9 megR gR
S
R
h Al h t a A f mAl
pSn cce9 e S t
R ect
nl g
( (,
gla 9 gR
ge ge c
l emAly o l A l
co
lt 9cg gR c t
cgc gR g gR
c 9 I 9 cmeg RA
c 9 onne e gR
S ecl ( ( , Sn9t 9g nhn e 9n e gR
S ecl g9 m9nu c R ( (,
e g n ge9 S gla 9 gR
ec t e g e gR g gR n t ocg
9gR n
R ect ec
( ( S eA
g S9e g
gR
9og gR
e RaS9gR cer gR g gRec t
ongR nt 9n hgR F
en g
C C Dm ayb
pSn cce9 cg g h e e ge g9 Se ge R ecg l n ol g9n 9
f na
9yn ol g A
9 9nt ge9 m c
9
n i o a S9e gc
ecl h c gRec n colg ec 9 cecg g megR
S e gR g 9ol
g gR g gR
e n g c9ngc 9
9noc m c gR Re R n i o a S9e gc l9 g
( W pSn cce9 hmRe R n
ah 5T g
gR
l9c n l ge9 cReS
nh c Sn9S9c
gm ( (
f t 9n gR leu la n9l 9
9f A
( F7
colgc
A
A hDa
9 cgo a gR
e
gR
hy
R n
g9 oc gR S
o l o e nc
R 9 gR t 9
gn c
gef ega 9 gR
3A) F f
( (
g9n vc
S e t g ne lc S9ng n egA
9og ( u
v e on 7, KA
yy f y
encg c l g
l9 g
9 t gR9 cKh 9t
e S ) ((z
e megR
F n t gc v ( h ) FK
pg g9 gR 9gR n e gR
R n t g m c en ge9 lla co
g
e R 9n n e9 ch m
n
l9 e g9 S
9t
3A) F f
g9n
ll le cA
A
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
AKR1B1
CpG Islands (Islands < 300 Bases are Light Green)
CpG: 93
B
F3
F2
F1
Empty
0
0.5
1
1.5
2
2.5
3
3.5
5
6
C
F3
F2
F1
Empty
0
A
1
2
ah 0d g 3 AmN Dm l mA m l
h N Aly A Df K
Al
A h yy f h yaDly A e p A 3 A
CC5 DDDA y mN e h
R
)IS
3A) F f
g9ncR9m
vS f lo q, A, ( K e 9gR
v e on 7,
KA gR
cR9m
egR n 9 9 ege9 c e 9gR
( ( S e 9ol
( F2
9n 4
A A h Da
o l N elf i
h y yy f b 3
lmhb
ce e e gl f l 9 gn c neSge9 l oSn ol ge9 ll le cN
S ) vT) AW 9l K 9 gn nah f
g9nc
(IS
le e l l f lc 9 n S9ng n
ll le c v e on 7,
g c 3
R ne S ) ( ( z vT3AW 9l K
3A) F
gef ega o FI S
3A) F
n e ge l g cg
KA R c n colgc 9 ent gR g
((z
ll le cA
colgc
ongR nt 9n h m
n cS9 ce l
ef e
9ngR
gR
cge m g
R S ecl gR c
g9
n
lo e g
mRe R Sn9g e elega 9
( ( S ecl A ongR nt 9n hm encg
e gm9 S ngch t
n t gc ) A( n t gc megR lo e n c
R nA pgm c l g
gn c neSge9 ( ( S ecl N
g cg
I
9n cS e e
t 9 c9t
R n
gR g
n t g ) h
gef gR n t g ) h og cgell l nla
ceg c Sn e g9nch g9
y
n gR gR mR9l
gR n t g) AFhmRe R m c lar
g9nc e e 9 gR t m n c l g
) AF v e on 7( KA
cc a v e on 7( Kh m n ler
n t g ) AF ec f t 9n S9m n ol c R R9m f n n t g ) A( m c l cc
9n Sn9g e c m n
oce ge c
e n g
e g chgm9
gef ega megRe gR
y1A
A
Scale
chr7:
134,143,200
134,143,300
500 bases
134,143,400
AKR1B1
134,143,500
134,143,600
134,143,700
134,143,800
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
hg19
134,143,900
134,144,000
134,144,100
CpG Islands (Islands < 300 Bases are Light Green)
CpG: 93
B
" #!
" #"
" #!
" #
" !
"
$
!
!$
"
"$
!!%
m
ah 0C g A A h l i lf m
h Al h N Aly o l A
C C e yD A b 3 AmN emy l mA
i Da l
h N Alb 3 D l i A A h l i lf mN e h o l l
N elf mAylha lb
y
I
S nge eS ge9 m
gR
mRe R gR
y1
e f lo g
v ) AFA( Kh p lo e e e ceg 9
g9 9f nl SSe gR
R gR
n
c i o c
g9
9 ent
gR en
gef ega 9 cR9ng n n t gc e lo e v ) AFA) K h e llah 9gR n 9 cgno g e y
I
y1 m c t og g
g9 n t 9f
gR
n 9 ege9 t 9ge v ) AFA( t ogKv e on 7( KA
( F5
colgc
ll 9 cgno gc e lo e gef egahmRel
gR e g g
R n
A hDa
g nm n ch m m g
gef ega 9
( (
m S n 9nt
n c
R n
9 gR c gm9
( ( S ecl A
hy
S eh c 9on
yy f y
t gRal ge9 o t gRal g
vc
g
gR
R n
g e got 9nc e e g co R l9c n l ge9 cReSA
t gRal ge9 m gn c
R e 9og gR et S9ng g9 R u R9m
9 f lo g gR n9l 9
cc ac S9e g
gef ega 9 gR
l f D l mAs A
t gRal g
gR Re R cg l f lc 9 lo e n c
l ge9 9n t og ge9 9 gR ceg n colg
gef ega v e on 7( KA R c
9ngR
cR9m
t g ne l nlo e n c
g
f
g9n ) I S
t gR9 cK e 3A) Fh 9gR
((z
ll le cc aA
F2 Unmethylated
F2 Methylated
0
ah 0p g A
mn3b
A h yy f y h yaDly lh Ay
t gRal g
)IS
3A) F f
( ( S e ( 3,
p aAN l f D l
ro
1
l
3A) F ce e e gla oSn ol g
pSn cce9 9 gR lo e n c
)IS
lA
0.5
9og FAW 9l c 9t S n
g9n v e on 7) KA R c
g coSS9ng gR
t 9 cgn g egc t 9 ol ge9 a
1.5
2
2.5
3
3.5
mn3 A N l f D l
4
r D vS f lo q, A, ( K gR
megR gR t gRal g
R t gRal ge9 A
n
gef ega 9
colgc
A l
9n n g9 g cg
gR
( (
t gR9 cKA
R
S
encg 9 c
ecl hm
9 cecg
9 lo e n c
ecl A
I
y
I
y1 e e lar
gR
t Sga f
g9n t o R l cc
e ge9 9 gR
( ( , gR gR
pS net g m
c i o cgn9ac gR
e e t g ne lc
gef ega 9 gR
n9m c i o g
( v e on 7FKA g n cge lah
9 gn9l
S ) A ongR n R
I
cS e e c i o c n l9 g
e p A
S
ll le h mRe R Re Rla
S ) gn c
C Cd neh yy mA mA
C C e
y e l
gef ega 9
l9 c y1 e e m
lg n ge9 e A 9ngo g lah l9 y
ah 0G g
s| A l
n
l9 c gR g 9 g e n c 9
ATTGGTTG!
n
9gR Sn9g e ch mRe R
R S )
g9 9 ent gR
TFBS:!
9
no e l 9n gR
( ( , e gR
S )
R pS net gc vc
pS net g oce l ge9 gR g
pSn cce9 9
y1 e gR
gm9
gR t e 9n n g9 l99u 9n gR9c
( v ( K Sn c g
c
f l9S
pS net gc n
gRec
( ( , A g n gR
c i o y
e gR gno S n 9nt
pSn cc
gR
gR n9l 9
ot
n
v e on 7FKA
megR gR
S )
(
pSn cc
pS net gc n
g9 gR c i o A
ATT*****!
e p N elf A
e p Cb
aA hl
hyb yl h y y A
l l
Dl
mh s
y yb
A
( 3(
colgc
A
s
A
c
9 f ne a gR et S9ng Re f
u 9 u9og9
9 l9 c
gR gno c
9
ge9 9 gR
C d neh yy mA A
og e 9g
nSe I
y1
cS e e c i o y1h m oc
9n cS e e
g R ei o g9
c m
e 9g g9
lg n ge9 chm socg e p N lef R
e e pSn cce9 9
g9 gR c i o e e p A
s
gR
e p
9m n ol ge9 ( ( h lgR9o R gR
v e on 73KA
mR gm 9 c nf oce pS net gc n c n l9 g
e p
g9nc Sn9 o
9 gR c Sn9g e c ec no e l 9n gR
S ecl A ongR n R
( 3)
ggR
n colgc e e I
9gR gn c neSge9 c ec e gR Sn9t 9g n 9 gRec
gR g gR
A
C C A
( (,
o y
9gR Sn9g e c S nc S n g A gRec
g n cge lahgR l9cc 9
9gR
A y
oc m m n 9g l99ue ah 0. g
s| b
9
s|
9
pS net gc
gR lo e n c
R 9 cgno gcD
n n9l 9 gR
g9 9 ent gR
R
e e c pS net gA
( (
9
y
colgc
A
S ec e Re eg9n 9
9la 9t
n t 9f
t )
Sn ccef
9t Sl p ) v
p R gR
t FK ) KA
cc9 e g
((z
llchmR n
R h ) , , 2D et 9 d
gR g gRec cel e ( (,
9t Sl p c
9nt c v F ) 7t h
c g9 gR n e9 v 9
gef ge9 9 gR
9 g cg eg m gn g
)
g lage co o eg 9 gR
t nu a gR t gRal g
RaS9gR cer
9n gR e ) e Re eg9nv
F )7
ggn g
e cg9 h ) , , 5KA
n cS9 ce l
) Sn9g e h mRe R ec gR
t
n e R ect
ec gR
9l9n g l
( ( , ec 9g pSn cc
nA
hmegR SKv e on 7WKA
! !
! !
#
"%
"
!%
!
%
)
#
(
#
'
#
&
#
%
#
$
#
#
#
"
#
!
#
*
!!&
ah 02 g
!!&
C C A
C Cd neh yy mA A Df y y m
g n cge lahm 9 c nf
mR g 9ol e e g gR g gR
n Sn cce9 t
R R !!&
R nm c
R ect A gR 9gR n R h gR
F )7
DDy lh l
co cg ge l e n c e gR
g n gR gn gt gA ongR n R
c e CC5
e e gef g
!!&
o l
eb
pSn cce9 9
o g9 gR l9cc 9
pSn cce9 9
pS net gc n ( (
( (,
)
e 9g
g9 cc cc gR
g ngR gn gt gA
( 3F
colgc
N
hmD h l mA yy f
9n n g9 l ne a gR
9l9n g l
o ge9 9 gR cel nh m e ) 5 9l9n g l
t gR9 cKA n
gR
g nc
gn c
g
e c S n g la
ll le c oce ll le c m n
Sn9le n ge9 n g c vc
R9onc
g
t g ne lc c
i o l ot
9gR
( (,
c e l gef en l f
( ( We ( ( zh m32,
g9nc vc
t g ne lc cg lecR h m Sn9
t gR9 cK a
n9
g9
ll 9o ge 9 gn9lh t Sgah
g nt e g) 3h32 ( ( , 9n
7)
( (W
llcA
"&
$
&
$
"
'
&
%
$
#
"
!
&
$
"
"
"&
"
%
$
"&
%
"&
#
#
'
"#
"
!#
!
#
#
#
"
"&
%
'
#
ah 05 g hmD h l mA yy f A
CC5R
DDyR N elf A A l o l
C Cd A
c cR9m e R t S n
m32,
e on 7zh gn c
pT mDmh l D A h
ge9 9
llol n Sn9le n ge9 e llc gR
. 4d A
C C2b
( (,
( ( z )5
DDDA y A mAlhmDy
cS e lla
ll le cA n c e Sn9le n ge9 m c 9 la 9 c nf
( (W
( ( z g n7) RA
yy f
lc9 S n 9nt
nl g
( 33
cc achmRe R e e g gR
megR Sn9le n ge9 vc
t g ne lc
llol n
t gR9 cKA
gef ega gR gec en gla
S n 9nt
colgc
cc ac g ) 3h32 megR gR
7) Rhe ( (, h
( ( zh m32,
( ( W9n t Sga f
) 5 mel gaS
g9n c
''"
# (&
# &
%(&
''"
g
%)"
$
%(&
e 9 gn9lA
##'
$ $&
# $&
#
" (&
" &
" $&
"
$%
%)
##'
$"""
# )" "
# '" "
# %" "
# $" "
# """
" )" "
" '" "
" %" "
" $" "
" """
($
# #"
$%
# #&
%)
%)"
($
# #"
# #&
%(&
''"
$*
$"""
# )" "
# '" "
# %" "
# $" "
# """
" )" "
" '" "
" %" "
" $" "
" """
$%
%)
%)"
ah 00 g
lh Ay l A
($
# #"
yy f A
C Cd A
# #&
CC5R o . 4d A
C C2b
et el nla g9 mR g 9 onn
t 9 cgn g
gn c
g
l9m n
le c
pghm m g
cc a
g ) 3h 32 9t
e gR
cS e lla e llol n
ge l m9ol
megR
ll gR c Sn9le n ge9 ( (,
t 9n
l nmR m
( ( W e 9l9n g l
) 5 mel
t Sga f
g9 n ge 9l a
Sla
n ge l g9 gR t
g hco
cg
9 gn9l
( ( Wh
n
llcA
e 9g Sn9le n g D
lt 9cg o one gaS
g9n c
( (,
g
llc m n
gR
t gR9 cKmegR
( ( Wm n 9g ef e e cc ac
gef egaK e llc mR n ge l m c
geSn9le n gef n9l e 9l9n g l
( ( , 9n
cc a
( ( Wv e on 77KA
( ( W 9n gR
R9m f nh g , AWB I t l 9 n ge lh 9 gn9l
g
llol n
( ( zh m32,
llc m n
mAlhmDy A
cc ah gR
l cc
gef ega 9 gR
n o
g n ge l 9c c 9 ) B I t l ll
gn c
DDDA yRo l
n ge l gn gt gvt g ne l ( (, h
mRe R ec u 9m g9 R f
Sn9le n ge9 ve e ge 7) R9onc e megR gR
v e on 72KA
g9 R u gR
eot A
g
c9n ( (,
g9 gR t
gn c
e pT mDmh l D A h
g
llc
gR gn gt gA
geygot 9n
nh f ngR l cc gRec
og
gef ega 9
g ec t o R
eot A
( 3W
colgc
ah 04 g
C C2b
( 3z
l A Dlh lN Al A
CC5R o . 4d A
pTRlh Ay l
h Al mA Alh l mAy m h l A Dr) cN D3o h l yl
Al
mhAml o l
h Al
C Cd A
DDDA yb
colgc
neh yy mA ehm DA
l hma
m l
h l Am
AmN
lD y A
A h
R e gn
ot
llol n 9 n 9
rat c e f 9lf
g e n g cg
n cS9 ce l
n
e gR
g
c
9n
e
gl c v
gR
pSn cce9 e S gRm a v
Retinoloxidizing
ADH1A
ADH1B
ADH1C
ADH4
ADH5
ADH6
ADH7
DHRS3
DHRS4
RDH5
RDH8
RDH10
RDH11
RDH12
RDH13
RDH14
mDmh l D l yya
y y
9 gn9ll
n ge 9e
ll
a gR
e9ca gR cec e pSn cce9 g nt e p K vRggSNI I c t
g c gcA R
c megR
gR
c
9l9n g l gecco h m oc
n A 9t An cgAoy
encg cg S
Sog gef
R
I Kh
S gRm a e 9nt l 9l9 h 9n gRec SonS9c m
9l9 gecco 9 W,
S gRm aKA
e n g gaS c 9
y g A eA eRA 9f I g
lacec v
9t e Cc
gef ege c 9
e S gRm a v
e n ge ge9 A 9
K vRggScNI I g
g9ua9A AsSI K e ge ae l
A
rat c f na n gla t 9 c 9
9n n ge 9e
9t
n
gn ge9 c 9 n ge 9e c n
t S9ng glahgR l f lc 9 gR c
llc
e l o f
9 cecg
lar
gR
n9l e gR n ge 9e
Retinalreducing
AKR1B1
AKR1B10
AKR1C1
AKR1C2
AKR1C3
AKR1C4
SDR16C5
HSD17B6
9 R n g ner gR
n le l Sn9
9l9 v
Retinaloxidizing
ALDH1A1
ALDH1A2
ALDH1A3
ALDH8A1
c e f 9lf
RA-oxidizing
Nuclear TF
CYP26A1
CYP26B1
CYP2E1
RARA
RARB
RARG
RXRA
RXRB
RXRG
n o e 9pe ere e gR n ge 9e
c vS f lo q , A, WK gecco A R
( h
e
l ( 3KA
Other
functions
ABCG5
ABCG8
AWAT1
AWAT2
BCO2
BCMO1
CRABP1
CRABP2
DGAT1
DGAT2
LRAT
RBP1
RBP4
RBP7
RETSAT
SULT1A1
D C. g A y Ai mDi
A l h l Am
e l o f A mDmAb
D l
A y h neh yy
AmhN D mDmAb A f DDmo h l my
Df neh yy
A A h f l my N m h l Df neh yy
A
AmhN D mDmh l Dl yya b
oc
e ( h
c v
rat c v
rat c
Wh
9 ce n
Fh
( (h
( (h
gR
2 n ge 9ly9pe ere (, h
( (h
( ) Kh )
9l9 hm
p e g cega f lo c e Rot e lo
3h
( (, h
e ca gR cec e gR
A
((h
( )h
y9pe ere c
( ) Kh W n ge ly
( FKh ) n ge ly
rat c v
)z (h
( 37
colgc
) z ( Kh3 o l n gn c neSge9 9gR n
c e f 9lf
( ( KA gef
Dl h
e gR
h
S gRm a v
c n Re Rle Rg
neh yy mA m
g9nc v
A y Ai mDi
e (h
h
h
K (h
W
l ( 3A
A h l Am
e l o f y yym l
o l yahi i D A mDmh l D A he l Aly
9
g e ce Rgc e g9 gR
9l9n g l
)z
nhm
c e f 9lf
9l9n g l
e n
le e l et Sle ge9 c 9
lar
lg n ge9 c e gR
cg gecge l cc9 e ge9 c 9
S gRm a e g c g gR g e lo
9l9 A
S gRm a e pSn cce9 l f lc 9 gR
g m n 9 g e ( 5 9nt l 9l9 n9t gR
( WW 9l9n g l
got 9nc t Sl cA
ah 0T g A
neh yy mA D i Dy m l
h l Am
e l o f
A y y A
AlDf Dl h
mDmh l D A hbCT AmhN D mDmA y N eD y r A h f 3 A C22 mDmA laN mhy r A h 3o h ay b
( 32
A
colgc
on n colgc l nla cR9m
S gRm a m c lg n
) zK m n
one 9l9n g l got 9ne cecA g9g l 9 ( W c v9og 9
lg n Dt 9cg 9 gR t v( FK m n
( Fh
( (h
)z (h
K pSn cce9 Sn9
9
gR g gR 9f n ll gn c neSge9 l Sn9 el 9 gR
(h
gm9 oSn ol g
c oc
e g c gc 9 Rot gR c
g m
S gRm a
9l9n g l
9o e n gR g e lo
lar
9 gR
(h
pSn cce9 9ol
g
pSn cce9 conf ef l e 9nt ge9 A egR
e n conf ef l onf c 9n
R lg n
t gR9 cKA
ec c
lg n
n
conf ef l m c
pSn cce9 S gg n 9n
c v e on 2, KA
0.059672
0.017763
0.018734
0.029271
0.010574
0.049359
0.003597
0.012905
eD As
h yahi i D ahi y mh
e l o f
A yo l
mn e i Da I dbd2 r n el mh
h hy lm
neh yy mA A Da lmDmo neh yy mAb ml DA y h D l l
mA
A
( (, h
gR g pRe eg
Sn9 cch h
A
F e n g
gR g 9f n ll conf ef l 9n
0.014312
ah 4d g
C Cd3b
Al hi Db
( (, h
h
p
ce e e gla l9m ne gR9c S ge gc gR g ecSl a
c9t
( (h
( ) Kv e on 75KA lg n
nhm
v
h
(,
Sl y
t g ne lc g n cge lah m
v
e S gRm a
9l9n g l
9 cgno g
vc
3h
e gRec lacec n lecg
n cc mR gR n n ge 9e
conf ef l 9og 9t
9m n ol g
( )h
)z (h
pSn cce9 l9m l f lc 9 gR c
v e on 2, KA gR
9 gn nah
(h
3h
h
9m n ol ge9 one c n
( , ec
cc9 e g
gR
megR
n
9l9n g l
c
n
m9nc conf ef l n g
gR g g c g9
oSn ol g
e ( 35
colgc
9l9n g l
nh e gRec
c h gR oSn ol ge9 ec en gla le u
conf ef l v e on 2, KA R c n colgc e e g gR g gR
c ec l9c la n l g
c9t
u a
S ge gc megR 9l9n g l
hmN ml h
lg n ge9 9
megR gR conf ef l 9 gR S ge g c t a R f
SSle ge9 c
g9 S99n
Sn9 9cge
gR
S gRm a
n ol ge9 9
g9n 9 S99n conf ef l 9n
ncA
N l f D l mA ehm DA m l
e l o f A AmhN D A
mDmh l DlaN mhl yya
R n c t 9cg 9 gR
nhgR n ec 9 l n t
t gRal ge9 e gR cel e t gRal ge9 9 gR
9l9 h 9t S n
Color Key
and Histogram
megR gR
400
Count
600
t gRal ge9 h ( (
)z (h
c v
(h
h
( (h
h
9l9n g l
lar
cA
g e 9nt l
9nt l
( (h
gR
e cg got 9n
( )h
)z (h
COX meth genes
( Km n RaS nt gRal g
e 9l9 1000
200
0
0
3000
2000
nv e on 2(
KAgenes T
RAP
0.2
0.6
Color
Value
A.
RAKey
pathway
and Histogram
1000
3000
Count
( Fh
Color Key
and Histogram
Color Key
and Histogram
e e S gRm a
9l9 got 9n gecco A 9t S ne COX meth genes
n m
9l9 t gRal ge9 4000
gR
9 gR n ge 9e
3000
lar
9m n ol g
9og R9m eg 9 oncA
2000
n9l 9
R ect
c n
Count
S gRm a u a
0.2
0.6
Value in colon tumor samples!
B. RA pathway
in normal colon samples!
RAP genes T
TCGA−AA−3713−01A−21D−1721−05
TCGA−D5−5537−01A−21D−1926−05
TCGA−A6−6649−01A−11D−1772−05
TCGA−D5−6932−01A−11D−1926−05
TCGA−CM−6168−01A−11D−1651−05
TCGA−CA−6717−01A−11D−1837−05
TCGA−G4−6297−01A−11D−1721−05
TCGA−D5−5539−01A−01D−1651−05
TCGA−A6−2671−01A−01D−1407−05
TCGA−AA−3496−01A−21D−1837−05
TCGA−G4−6298−01A−11D−1721−05
TCGA−AA−3495−01A−01D−1407−05
TCGA−CK−4947−01B−11D−1651−05
TCGA−AA−3660−01A−01D−1721−05
TCGA−D5−6926−01A−11D−1926−05
TCGA−G4−6627−01A−11D−1772−05
TCGA−AZ−5407−01A−01D−1721−05
TCGA−AD−6899−01A−11D−1926−05
TCGA−DM−A1DB−01A−11D−A153−05
TCGA−CM−5864−01A−01D−1651−05
TCGA−A6−2680−01A−01D−1407−05
TCGA−A6−6654−01A−21D−1837−05
TCGA−CK−6747−01A−11D−1837−05
TCGA−D5−6541−01A−11D−1721−05
TCGA−D5−6535−01A−11D−1721−05
TCGA−AY−6197−01A−11D−1721−05
TCGA−AZ−4616−01A−21D−1837−05
TCGA−D5−6530−01A−11D−1721−05
TCGA−A6−5657−01A−01D−1651−05
TCGA−AZ−6605−01A−11D−1837−05
TCGA−A6−2681−01A−01D−1407−05
TCGA−AD−6901−01A−11D−1926−05
TCGA−AA−3506−01A−01D−1407−05
TCGA−A6−6138−01A−11D−1772−05
TCGA−G4−6306−01A−11D−1772−05
TCGA−D5−6529−01A−11D−1772−05
TCGA−AD−6965−01A−11D−1926−05
TCGA−AZ−6607−01A−11D−1837−05
TCGA−AD−6964−01A−11D−1926−05
TCGA−G4−6294−01A−11D−1772−05
TCGA−AA−3488−01A−01D−1407−05
TCGA−AM−5820−01A−01D−1651−05
TCGA−A6−5659−01A−01D−1651−05
TCGA−D5−6532−01A−11D−1721−05
TCGA−DM−A288−01A−11D−A16X−05
TCGA−D5−6923−01A−11D−1926−05
TCGA−AD−6548−01A−11D−1837−05
TCGA−G4−6303−01A−11D−1772−05
TCGA−G4−6302−01A−11D−1721−05
TCGA−D5−6898−01A−11D−1926−05
TCGA−AY−5543−01A−01D−1651−05
TCGA−CM−5344−01A−21D−1721−05
TCGA−CM−6169−01A−11D−1651−05
TCGA−DM−A1D6−01A−21D−A153−05
TCGA−F4−6807−01A−11D−1837−05
TCGA−CA−5256−01A−01D−1407−05
TCGA−AD−6895−01A−11D−1926−05
TCGA−D5−6533−01A−11D−1721−05
TCGA−CM−5348−01A−21D−1721−05
TCGA−F4−6855−01A−11D−1926−05
TCGA−CM−6164−01A−11D−1651−05
TCGA−A6−5656−01A−21D−1837−05
TCGA−DM−A1D7−01A−11D−A153−05
TCGA−G4−6315−01A−11D−1721−05
TCGA−AA−3655−01A−02D−1721−05
TCGA−DM−A28K−01A−21D−A16X−05
TCGA−CM−6166−01A−11D−1651−05
TCGA−CM−6163−01A−11D−1651−05
TCGA−CM−5341−01A−01D−1407−05
TCGA−A6−6780−01A−11D−1837−05
TCGA−CM−6172−01A−11D−1651−05
TCGA−A6−5666−01A−01D−1651−05
TCGA−D5−6537−01A−11D−1721−05
TCGA−CA−5796−01A−01D−1651−05
TCGA−DM−A0XF−01A−11D−A153−05
TCGA−CM−5349−01A−21D−1721−05
TCGA−CM−4746−01A−01D−1407−05
TCGA−CM−4747−01A−01D−1407−05
TCGA−DM−A28A−01A−21D−A16X−05
TCGA−DM−A1HA−01A−11D−A153−05
TCGA−CK−5914−01A−11D−1651−05
TCGA−D5−5540−01A−01D−1651−05
TCGA−DM−A1D9−01A−11D−A153−05
TCGA−A6−2682−01A−01D−1407−05
TCGA−AA−3509−01A−01D−1407−05
TCGA−F4−6805−01A−11D−1837−05
TCGA−AZ−4315−01A−01D−1407−05
TCGA−G4−6310−01A−11D−1721−05
TCGA−AZ−4682−01B−01D−1407−05
TCGA−AZ−6603−01A−11D−1837−05
TCGA−G4−6295−01A−11D−1721−05
TCGA−G4−6307−01A−11D−1721−05
TCGA−AA−3494−01A−01D−1407−05
TCGA−F4−6569−01A−11D−1772−05
TCGA−F4−6704−01A−11D−1837−05
TCGA−CM−6678−01A−11D−1837−05
TCGA−CM−6679−01A−11D−1837−05
TCGA−A6−6782−01A−11D−1837−05
TCGA−A6−5667−01A−21D−1721−05
TCGA−DM−A28F−01A−11D−A16X−05
TCGA−G4−6311−01A−11D−1721−05
TCGA−CM−6680−01A−11D−1837−05
TCGA−D5−5538−01A−01D−1651−05
TCGA−D5−6929−01A−31D−1926−05
TCGA−AZ−6600−01A−11D−1772−05
TCGA−CM−5862−01A−01D−1651−05
TCGA−CM−4752−01A−01D−1407−05
TCGA−D5−6922−01A−11D−1926−05
TCGA−AA−3662−01A−01D−1721−05
TCGA−CM−5868−01A−01D−1651−05
TCGA−AZ−4308−01A−01D−1407−05
TCGA−F4−6703−01A−11D−1837−05
TCGA−CM−6162−01A−11D−1651−05
TCGA−G4−6304−01A−11D−1926−05
TCGA−F4−6854−01A−11D−1926−05
TCGA−F4−6808−01A−11D−1837−05
TCGA−G4−6625−01A−21D−1772−05
TCGA−AA−3697−01A−01D−1721−05
TCGA−A6−2685−01A−01D−1407−05
TCGA−AZ−4681−01A−01D−1407−05
TCGA−G4−6309−01A−21D−1837−05
TCGA−CK−5912−01A−11D−1651−05
TCGA−AZ−5403−01A−01D−1651−05
TCGA−A6−6648−01A−11D−1772−05
TCGA−A6−6140−01A−11D−1772−05
TCGA−DM−A1D0−01A−11D−A153−05
TCGA−DM−A28E−01A−11D−A16X−05
TCGA−A6−5662−01A−01D−1651−05
TCGA−D5−6536−01A−11D−1721−05
TCGA−D5−6924−01A−11D−1926−05
TCGA−AZ−4313−01A−01D−1407−05
TCGA−AZ−6608−01A−11D−1837−05
TCGA−D5−6928−01A−11D−1926−05
TCGA−D5−6534−01A−21D−1926−05
TCGA−D5−5541−01A−01D−1651−05
TCGA−AA−3712−01A−21D−1721−05
TCGA−F4−6459−01A−11D−1772−05
TCGA−G4−6317−01A−11D−1721−05
TCGA−CM−6170−01A−11D−1651−05
TCGA−A6−2684−01A−01D−1407−05
TCGA−A6−6141−01A−11D−1772−05
TCGA−G4−6626−01A−11D−1772−05
TCGA−G4−6314−01A−11D−1721−05
TCGA−AD−6888−01A−11D−1926−05
TCGA−A6−6142−01A−11D−1772−05
TCGA−CM−6674−01A−11D−1837−05
TCGA−AD−6963−01A−11D−1926−05
TCGA−CM−4748−01A−01D−1407−05
TCGA−CA−5255−01A−11D−1837−05
TCGA−AD−6889−01A−11D−1926−05
TCGA−AZ−4323−01A−21D−1837−05
TCGA−F4−6809−01A−11D−1837−05
TCGA−F4−6856−01A−11D−1926−05
TCGA−CM−6165−01A−11D−1651−05
TCGA−A6−6652−01A−11D−1772−05
TCGA−D5−6930−01A−11D−1926−05
TCGA−F4−6460−01A−11D−1772−05
TCGA−CM−5863−01A−21D−1837−05
TCGA−DM−A280−01A−12D−A16X−05
TCGA−CM−6167−01A−11D−1651−05
TCGA−AZ−4684−01A−01D−1407−05
TCGA−CA−5797−01A−01D−1651−05
TCGA−D5−6539−01A−11D−1721−05
TCGA−AY−6196−01A−11D−1721−05
TCGA−AA−3511−01A−21D−1837−05
TCGA−AA−3510−01A−01D−1407−05
TCGA−CK−6748−01A−11D−1837−05
TCGA−CA−6715−01A−21D−1837−05
TCGA−A6−6651−01A−21D−1837−05
TCGA−AU−3779−01A−01D−1721−05
TCGA−A6−6137−01A−11D−1772−05
TCGA−AA−3489−01A−21D−1837−05
TCGA−CK−4948−01B−11D−1651−05
TCGA−A6−2675−01A−02D−1721−05
TCGA−A6−6650−01A−11D−1772−05
TCGA−CM−6676−01A−11D−1837−05
TCGA−DM−A28C−01A−11D−A16X−05
TCGA−AD−6890−01A−11D−1926−05
TCGA−G4−6321−01A−11D−1721−05
TCGA−DM−A285−01A−11D−A16X−05
TCGA−CM−4750−01A−01D−1407−05
TCGA−CM−4751−01A−02D−1837−05
TCGA−DM−A0XD−01A−12D−A153−05
TCGA−AM−5821−01A−01D−1651−05
TCGA−G4−6323−01A−11D−1721−05
TCGA−D5−7000−01A−11D−1926−05
TCGA−F4−6806−01A−11D−1837−05
TCGA−G4−6299−01A−11D−1772−05
TCGA−G4−6586−01A−11D−1772−05
TCGA−CA−6719−01A−11D−1837−05
TCGA−A6−4107−01A−02D−1407−05
TCGA−CM−6171−01A−11D−1651−05
TCGA−A6−4105−01A−02D−1772−05
TCGA−AZ−4614−01A−01D−1407−05
TCGA−CA−5254−01A−21D−1837−05
TCGA−CM−6675−01A−11D−1837−05
TCGA−DM−A0X9−01A−11D−A153−05
TCGA−D5−6531−01A−11D−1721−05
TCGA−D5−6540−01A−11D−1721−05
TCGA−CK−4952−01A−01D−1721−05
TCGA−A6−6653−01A−11D−1772−05
TCGA−AZ−6598−01A−11D−1772−05
TCGA−DM−A1HB−01A−22D−A17Z−05
TCGA−DM−A1DA−01A−11D−A153−05
TCGA−AZ−6606−01A−11D−1837−05
TCGA−AA−3492−01A−01D−1407−05
TCGA−CM−5861−01A−01D−1651−05
TCGA−A6−5661−01A−01D−1651−05
TCGA−G4−6588−01A−11D−1772−05
TCGA−CM−4744−01A−01D−1407−05
TCGA−G4−6320−01A−11D−1721−05
TCGA−CA−6718−01A−11D−1837−05
TCGA−F4−6570−01A−11D−1772−05
TCGA−AZ−4615−01A−01D−1407−05
TCGA−CM−4743−01A−01D−1721−05
TCGA−F4−6461−01A−11D−1772−05
TCGA−CK−5916−01A−11D−1651−05
TCGA−G4−6628−01A−11D−1837−05
TCGA−A6−5665−01A−01D−1651−05
TCGA−DM−A1D8−01A−11D−A153−05
TCGA−AD−5900−01A−11D−1651−05
TCGA−CK−6746−01A−11D−1837−05
TCGA−AA−3663−01A−01D−1721−05
TCGA−D5−6920−01A−11D−1926−05
TCGA−CA−6716−01A−11D−1837−05
TCGA−A6−5664−01A−21D−1837−05
TCGA−F4−6463−01A−11D−1721−05
TCGA−AZ−6601−01A−11D−1772−05
TCGA−G4−6293−01A−11D−1721−05
TCGA−CM−6677−01A−11D−1837−05
TCGA−CK−4950−01A−01D−1721−05
TCGA−A6−6781−01A−22D−1926−05
TCGA−AU−6004−01A−11D−1721−05
TCGA−CM−6161−01A−11D−1651−05
TCGA−DM−A1D4−01A−21D−A153−05
TCGA−DM−A28M−01A−12D−A16X−05
TCGA−CK−4951−01A−01D−1407−05
TCGA−AZ−6599−01A−11D−1772−05
TCGA−A6−2679−01A−02D−1407−05
TCGA−AY−6386−01A−21D−1721−05
TCGA−DM−A28G−01A−11D−A16X−05
TCGA−DM−A282−01A−12D−A16X−05
TCGA−CK−6751−01A−11D−1837−05
TCGA−CK−5913−01A−11D−1651−05
TCGA−DM−A28H−01A−11D−A16X−05
TCGA−CM−5860−01A−01D−1651−05
TCGA−CK−5915−01A−11D−1651−05
TCGA−A6−5660−01A−01D−1651−05
TCGA−D5−6538−01A−11D−1721−05
TCGA−D5−6931−01A−11D−1926−05
TCGA−AA−3502−01A−01D−1407−05
TCGA−D5−6927−01A−21D−1926−05
TCGA−A6−2686−01A−01D−1407−05
TCGA−G4−6322−01A−11D−1721−05
TCGA−AA−3713−11A−01D−1721−05
0
2000
TCGA−AZ−6600−11A−01D−1772−05
0.2
1000
Count
TCGA−A6−4107−11A−01D−1551−05
TCGA−AA−3502−11A−01D−1407−05
0.6
TCGA−AA−3510−11A−01D−1407−05
Value
TCGA−AA−3494−11A−01D−1407−05
0.6
ADH1C
Aldh1a1
AKR1B1
Aldh1a2
DHRS4
RDH13
Sult1a1
Adh5
Rxrb
Dhrs3
RxRa
AKR1C1
Rbp1
RaRb
AKR1C3
RaRg
ADH1B
AKR1C2
CYP26B1
RDH11
Retsat
RaRa
Crabp1
CYP16A1
RxRg
RDH10
AKR1B10
LRAT
ALDH1A1
AKR1B1
RDH11
ALDH1A2
DHRS4
SULT1A1
RXRB
DHRS3
RXRA
AKR1C1
RBP1
AKR1B10
RARB
AKR1C3
DGAT1
ADH1B
ADH1C
RETSAT
AKR1C2
CYP26B1
RDH10
RDH12
CYP26A1
ADH5
LRAT
RARA
Value
ALDH1A1
AKR1B1
RDH11
ALDH1A2
DHRS4
SULT1A1
RXRB
DHRS3
RXRA
AKR1C1
RBP1
AKR1B10
RARB
AKR1C3
DGAT1
ADH1B
ADH1C
RETSAT
AKR1C2
CYP26B1
RDH10
RDH12
CYP26A1
ADH5
LRAT
RARA
0
TCGA−AA−3660−11A−01D−1721−05
0.2
TCGA−A6−4107−01A−02D−1407−05
TCGA−CM−4744−01A−01D−1407−05
TCGA−CM−6171−01A−11D−1651−05
TCGA−CK−6746−01A−11D−1837−05
TCGA−CK−4952−01A−01D−1721−05
TCGA−D5−6920−01A−11D−1926−05
TCGA−AZ−6598−11A−01D−1772−05
TCGA−G4−6320−01A−11D−1721−05
TCGA−A6−5665−01A−01D−1651−05
TCGA−DM−A1HB−01A−22D−A17Z−05
TCGA−AZ−6606−01A−11D−1837−05
TCGA−A6−6653−01A−11D−1772−05
TCGA−AZ−6598−01A−11D−1772−05
TCGA−DM−A1DA−01A−11D−A153−05
TCGA−G4−6625−11A−01D−1772−05
TCGA−CK−5915−01A−11D−1651−05
TCGA−CA−6716−01A−11D−1837−05
TCGA−DM−A28M−01A−12D−A16X−05
TCGA−F4−6806−01A−11D−1837−05
TCGA−DM−A1D8−01A−11D−A153−05
TCGA−AA−3663−01A−01D−1721−05
TCGA−CM−4746−01A−01D−1407−05
TCGA−AZ−6601−11A−01D−1772−05
TCGA−A6−5660−01A−01D−1651−05
TCGA−CK−4951−01A−01D−1407−05
TCGA−DM−A1D4−01A−21D−A153−05
TCGA−AD−6890−01A−11D−1926−05
TCGA−G4−6321−01A−11D−1721−05
TCGA−AM−5821−01A−01D−1651−05
TCGA−CM−4751−01A−02D−1837−05
TCGA−CA−6719−01A−11D−1837−05
TCGA−AA−3506−11A−01D−1407−05
TCGA−G4−6299−01A−11D−1772−05
TCGA−G4−6586−01A−11D−1772−05
TCGA−G4−6588−01A−11D−1772−05
TCGA−D5−6540−01A−11D−1721−05
TCGA−CK−6751−01A−11D−1837−05
TCGA−CK−5913−01A−11D−1651−05
TCGA−A6−5661−01A−01D−1651−05
TCGA−AA−3488−11A−01D−1407−05
TCGA−A6−4105−01A−02D−1772−05
TCGA−CA−5254−01A−21D−1837−05
TCGA−AZ−4614−01A−01D−1407−05
TCGA−A6−4107−01A−02D−1407−05
TCGA−AA−3492−01A−01D−1407−05
TCGA−CM−4744−01A−01D−1407−05
TCGA−CM−5861−01A−01D−1651−05
TCGA−CM−6171−01A−11D−1651−05
TCGA−D5−6531−01A−11D−1721−05
TCGA−CK−6746−01A−11D−1837−05
TCGA−DM−A0X9−01A−11D−A153−05
TCGA−CK−4952−01A−01D−1721−05
TCGA−AA−3495−11A−01D−1407−05
TCGA−CM−6675−01A−11D−1837−05
TCGA−D5−6920−01A−11D−1926−05
TCGA−AU−6004−01A−11D−1721−05
TCGA−G4−6320−01A−11D−1721−05
TCGA−G4−6323−01A−11D−1721−05
TCGA−A6−5665−01A−01D−1651−05
TCGA−D5−7000−01A−11D−1926−05
TCGA−DM−A1HB−01A−22D−A17Z−05
TCGA−G4−6628−01A−11D−1837−05
TCGA−AZ−6606−01A−11D−1837−05
TCGA−A6−6781−01A−22D−1926−05
TCGA−A6−6653−01A−11D−1772−05
TCGA−CK−4950−01A−01D−1721−05
TCGA−AZ−6598−01A−11D−1772−05
TCGA−AA−3697−11A−01D−1721−05
TCGA−CM−6161−01A−11D−1651−05
TCGA−DM−A1DA−01A−11D−A153−05
TCGA−CM−4743−01A−01D−1721−05
TCGA−CK−5915−01A−11D−1651−05
TCGA−AD−6548−01A−11D−1837−05
TCGA−CA−6716−01A−11D−1837−05
TCGA−D5−6923−01A−11D−1926−05
TCGA−DM−A28M−01A−12D−A16X−05
TCGA−CM−6677−01A−11D−1837−05
TCGA−F4−6806−01A−11D−1837−05
TCGA−G4−6293−01A−11D−1721−05
TCGA−DM−A1D8−01A−11D−A153−05
TCGA−DM−A285−01A−11D−A16X−05
TCGA−AA−3663−01A−01D−1721−05
TCGA−AZ−4615−01A−01D−1407−05
TCGA−CM−4746−01A−01D−1407−05
TCGA−AA−3712−11A−01D−1721−05
TCGA−CM−6166−01A−11D−1651−05
TCGA−A6−5660−01A−01D−1651−05
TCGA−DM−A0XD−01A−12D−A153−05
TCGA−CK−4951−01A−01D−1407−05
TCGA−CM−4750−01A−01D−1407−05
TCGA−DM−A1D4−01A−21D−A153−05
TCGA−F4−6570−01A−11D−1772−05
TCGA−AD−6890−01A−11D−1926−05
TCGA−CK−5916−01A−11D−1651−05
TCGA−G4−6321−01A−11D−1721−05
TCGA−CA−6718−01A−11D−1837−05
TCGA−AM−5821−01A−01D−1651−05
TCGA−F4−6461−01A−11D−1772−05
TCGA−CM−4751−01A−02D−1837−05
TCGA−G4−6297−11A−01D−1721−05
TCGA−A6−2686−01A−01D−1407−05
TCGA−CA−6719−01A−11D−1837−05
TCGA−AD−5900−01A−11D−1651−05
TCGA−G4−6299−01A−11D−1772−05
TCGA−G4−6322−01A−11D−1721−05
TCGA−G4−6586−01A−11D−1772−05
TCGA−CK−5914−01A−11D−1651−05
TCGA−G4−6588−01A−11D−1772−05
TCGA−CM−6163−01A−11D−1651−05
TCGA−D5−6540−01A−11D−1721−05
TCGA−CM−5341−01A−01D−1407−05
TCGA−CK−6751−01A−11D−1837−05
TCGA−A6−5666−01A−01D−1651−05
TCGA−CK−5913−01A−11D−1651−05
TCGA−AA−3663−11A−01D−1721−05
TCGA−D5−6537−01A−11D−1721−05
TCGA−A6−5661−01A−01D−1651−05
TCGA−CM−6172−01A−11D−1651−05
TCGA−A6−4105−01A−02D−1772−05
TCGA−DM−A1D7−01A−11D−A153−05
TCGA−CA−5254−01A−21D−1837−05
TCGA−G4−6315−01A−11D−1721−05
TCGA−AZ−4614−01A−01D−1407−05
TCGA−AA−3655−01A−02D−1721−05
TCGA−AA−3492−01A−01D−1407−05
TCGA−DM−A28K−01A−21D−A16X−05
TCGA−CM−5861−01A−01D−1651−05
TCGA−AY−6386−01A−21D−1721−05
TCGA−D5−6531−01A−11D−1721−05
TCGA−A6−2679−01A−02D−1407−05
TCGA−DM−A0X9−01A−11D−A153−05
TCGA−AA−3509−11A−01D−1407−05
TCGA−AZ−6599−01A−11D−1772−05
TCGA−CM−6675−01A−11D−1837−05
TCGA−DM−A28H−01A−11D−A16X−05
TCGA−AU−6004−01A−11D−1721−05
TCGA−D5−6927−01A−21D−1926−05
TCGA−G4−6323−01A−11D−1721−05
TCGA−AA−3502−01A−01D−1407−05
TCGA−D5−7000−01A−11D−1926−05
TCGA−D5−6538−01A−11D−1721−05
TCGA−G4−6628−01A−11D−1837−05
TCGA−D5−6931−01A−11D−1926−05
TCGA−A6−6781−01A−22D−1926−05
TCGA−AZ−4616−01A−21D−1837−05
TCGA−CK−4950−01A−01D−1721−05
TCGA−AA−3655−11A−01D−1721−05
TCGA−D5−6541−01A−11D−1721−05
TCGA−CM−6161−01A−11D−1651−05
TCGA−CM−5860−01A−01D−1651−05
TCGA−CM−4743−01A−01D−1721−05
TCGA−AY−6197−01A−11D−1721−05
TCGA−AD−6548−01A−11D−1837−05
TCGA−D5−6535−01A−11D−1721−05
TCGA−D5−6923−01A−11D−1926−05
TCGA−AZ−5407−01A−01D−1721−05
TCGA−CM−6677−01A−11D−1837−05
TCGA−AZ−6600−01A−11D−1772−05
TCGA−G4−6293−01A−11D−1721−05
TCGA−DM−A288−01A−11D−A16X−05
TCGA−DM−A285−01A−11D−A16X−05
TCGA−A6−2675−11A−01D−1721−05
TCGA−A6−5659−01A−01D−1651−05
TCGA−AZ−4615−01A−01D−1407−05
TCGA−D5−6532−01A−11D−1721−05
TCGA−CM−6166−01A−11D−1651−05
TCGA−AA−3495−01A−01D−1407−05
TCGA−DM−A0XD−01A−12D−A153−05
TCGA−CK−4947−01B−11D−1651−05
TCGA−CM−4750−01A−01D−1407−05
TCGA−AA−3496−01A−21D−1837−05
TCGA−F4−6570−01A−11D−1772−05
TCGA−A6−2671−01A−01D−1407−05
TCGA−CK−5916−01A−11D−1651−05
TCGA−DM−A28A−01A−21D−A16X−05
TCGA−CA−6718−01A−11D−1837−05
TCGA−CM−5864−01A−01D−1651−05
TCGA−F4−6461−01A−11D−1772−05
TCGA−G4−6320−11A−01D−1721−05
TCGA−A6−2680−01A−01D−1407−05
TCGA−A6−2686−01A−01D−1407−05
TCGA−DM−A1HA−01A−11D−A153−05
TCGA−AD−5900−01A−11D−1651−05
TCGA−CM−6679−01A−11D−1837−05
TCGA−G4−6322−01A−11D−1721−05
TCGA−AZ−4315−01A−01D−1407−05
TCGA−CK−5914−01A−11D−1651−05
TCGA−A6−6782−01A−11D−1837−05
TCGA−CM−6163−01A−11D−1651−05
TCGA−F4−6805−01A−11D−1837−05
TCGA−CM−5341−01A−01D−1407−05
TCGA−CM−6169−01A−11D−1651−05
TCGA−A6−5666−01A−01D−1651−05
TCGA−A6−2682−11A−01D−1551−05
TCGA−AZ−4684−01A−01D−1407−05
TCGA−D5−6537−01A−11D−1721−05
TCGA−DM−A280−01A−12D−A16X−05
TCGA−CM−6172−01A−11D−1651−05
TCGA−A6−2682−01A−01D−1407−05
TCGA−DM−A1D7−01A−11D−A153−05
TCGA−G4−6295−01A−11D−1721−05
TCGA−G4−6315−01A−11D−1721−05
TCGA−DM−A1D9−01A−11D−A153−05
TCGA−AA−3655−01A−02D−1721−05
TCGA−AZ−4682−01B−01D−1407−05
TCGA−DM−A28K−01A−21D−A16X−05
TCGA−D5−6929−01A−31D−1926−05
TCGA−AY−6386−01A−21D−1721−05
TCGA−G4−6314−11A−01D−1721−05
TCGA−AA−3509−01A−01D−1407−05
TCGA−A6−2679−01A−02D−1407−05
TCGA−G4−6307−01A−11D−1721−05
TCGA−AZ−6599−01A−11D−1772−05
TCGA−F4−6569−01A−11D−1772−05
TCGA−DM−A28H−01A−11D−A16X−05
TCGA−G4−6311−01A−11D−1721−05
TCGA−D5−6927−01A−21D−1926−05
TCGA−F4−6704−01A−11D−1837−05
TCGA−AA−3502−01A−01D−1407−05
TCGA−CM−6678−01A−11D−1837−05
TCGA−D5−6538−01A−11D−1721−05
TCGA−DM−A28F−01A−11D−A16X−05
TCGA−D5−6931−01A−11D−1926−05
TCGA−CM−6680−01A−11D−1837−05
TCGA−AZ−4616−01A−21D−1837−05
TCGA−A6−2686−11A−01D−1551−05
TCGA−CM−4752−01A−01D−1407−05
TCGA−D5−6541−01A−11D−1721−05
TCGA−CM−5862−01A−01D−1651−05
TCGA−CM−5860−01A−01D−1651−05
TCGA−D5−6922−01A−11D−1926−05
TCGA−AY−6197−01A−11D−1721−05
TCGA−CM−6165−01A−11D−1651−05
TCGA−D5−6535−01A−11D−1721−05
TCGA−AA−3660−01A−01D−1721−05
TCGA−AZ−5407−01A−01D−1721−05
TCGA−A6−6652−01A−11D−1772−05
TCGA−AZ−6600−01A−11D−1772−05
TCGA−AD−6889−01A−11D−1926−05
TCGA−DM−A288−01A−11D−A16X−05
TCGA−AA−3492−11A−01D−1407−05
TCGA−AZ−4323−01A−21D−1837−05
TCGA−A6−5659−01A−01D−1651−05
TCGA−D5−6930−01A−11D−1926−05
TCGA−D5−6532−01A−11D−1721−05
TCGA−F4−6460−01A−11D−1772−05
TCGA−AA−3495−01A−01D−1407−05
TCGA−G4−6627−01A−11D−1772−05
TCGA−CK−4947−01B−11D−1651−05
TCGA−A6−5657−01A−01D−1651−05
TCGA−AA−3496−01A−21D−1837−05
TCGA−AD−6901−01A−11D−1926−05
TCGA−A6−2671−01A−01D−1407−05
TCGA−D5−6530−01A−11D−1721−05
TCGA−DM−A28A−01A−21D−A16X−05
TCGA−A6−2671−11A−01D−1551−05
TCGA−G4−6303−01A−11D−1772−05
TCGA−CM−5864−01A−01D−1651−05
TCGA−AA−3506−01A−01D−1407−05
TCGA−A6−2680−01A−01D−1407−05
TCGA−CM−6168−01A−11D−1651−05
TCGA−DM−A1HA−01A−11D−A153−05
TCGA−AY−6196−01A−11D−1721−05
TCGA−CM−6679−01A−11D−1837−05
TCGA−D5−6926−01A−11D−1926−05
TCGA−AZ−4315−01A−01D−1407−05
TCGA−AZ−6605−01A−11D−1837−05
TCGA−A6−6782−01A−11D−1837−05
TCGA−A6−2681−01A−01D−1407−05
TCGA−F4−6805−01A−11D−1837−05
TCGA−G4−6311−11A−01D−1721−05
TCGA−D5−6898−01A−11D−1926−05
TCGA−CM−6169−01A−11D−1651−05
TCGA−D5−6932−01A−11D−1926−05
TCGA−AZ−4684−01A−01D−1407−05
TCGA−CM−5863−01A−21D−1837−05
TCGA−DM−A280−01A−12D−A16X−05
TCGA−CA−6717−01A−11D−1837−05
TCGA−A6−2682−01A−01D−1407−05
TCGA−AD−6965−01A−11D−1926−05
TCGA−G4−6295−01A−11D−1721−05
TCGA−AZ−6607−01A−11D−1837−05
TCGA−DM−A1D9−01A−11D−A153−05
TCGA−A6−6649−01A−11D−1772−05
TCGA−AZ−4682−01B−01D−1407−05
TCGA−AA−3713−01A−21D−1721−05
TCGA−D5−6929−01A−31D−1926−05
TCGA−A6−5667−11A−01D−1721−05
TCGA−A6−5664−01A−21D−1837−05
TCGA−AA−3509−01A−01D−1407−05
TCGA−A6−6654−01A−21D−1837−05
TCGA−G4−6307−01A−11D−1721−05
TCGA−A6−6651−01A−21D−1837−05
TCGA−F4−6569−01A−11D−1772−05
TCGA−CM−5348−01A−21D−1721−05
TCGA−G4−6311−01A−11D−1721−05
TCGA−G4−6302−01A−11D−1721−05
TCGA−F4−6704−01A−11D−1837−05
TCGA−D5−5541−01A−01D−1651−05
TCGA−CM−6678−01A−11D−1837−05
TCGA−AA−3712−01A−21D−1721−05
TCGA−DM−A28F−01A−11D−A16X−05
TCGA−A6−2681−11A−01D−1551−05
TCGA−AZ−6608−01A−11D−1837−05
TCGA−CM−6680−01A−11D−1837−05
TCGA−F4−6808−01A−11D−1837−05
TCGA−CM−4752−01A−01D−1407−05
TCGA−F4−6854−01A−11D−1926−05
TCGA−CM−5862−01A−01D−1651−05
TCGA−CK−5912−01A−11D−1651−05
TCGA−D5−6922−01A−11D−1926−05
TCGA−A6−6648−01A−11D−1772−05
TCGA−CM−6165−01A−11D−1651−05
TCGA−AA−3697−01A−01D−1721−05
TCGA−AA−3660−01A−01D−1721−05
TCGA−A6−6140−01A−11D−1772−05
TCGA−A6−6652−01A−11D−1772−05
TCGA−AZ−6599−11A−01D−1772−05
TCGA−DM−A1D0−01A−11D−A153−05
TCGA−AD−6889−01A−11D−1926−05
TCGA−AA−3494−01A−01D−1407−05
TCGA−AZ−4323−01A−21D−1837−05
TCGA−G4−6309−01A−21D−1837−05
TCGA−D5−6930−01A−11D−1926−05
TCGA−AZ−4681−01A−01D−1407−05
TCGA−F4−6460−01A−11D−1772−05
TCGA−DM−A28E−01A−11D−A16X−05
TCGA−G4−6627−01A−11D−1772−05
TCGA−D5−6928−01A−11D−1926−05
TCGA−A6−5657−01A−01D−1651−05
TCGA−AZ−4313−01A−01D−1407−05
TCGA−AD−6901−01A−11D−1926−05
TCGA−CA−5797−01A−01D−1651−05
TCGA−D5−6530−01A−11D−1721−05
TCGA−G4−6322−11A−01D−1721−05
TCGA−D5−6534−01A−21D−1926−05
TCGA−G4−6303−01A−11D−1772−05
TCGA−AZ−4308−01A−01D−1407−05
TCGA−AA−3506−01A−01D−1407−05
TCGA−G4−6304−01A−11D−1926−05
TCGA−CM−6168−01A−11D−1651−05
TCGA−CM−6162−01A−11D−1651−05
TCGA−AY−6196−01A−11D−1721−05
TCGA−AZ−5403−01A−01D−1651−05
TCGA−D5−6926−01A−11D−1926−05
TCGA−A6−2685−01A−01D−1407−05
TCGA−AZ−6605−01A−11D−1837−05
TCGA−F4−6703−01A−11D−1837−05
TCGA−A6−2681−01A−01D−1407−05
TCGA−A6−2685−11A−01D−1551−05
TCGA−D5−6924−01A−11D−1926−05
TCGA−D5−6898−01A−11D−1926−05
TCGA−AA−3489−01A−21D−1837−05
TCGA−D5−6932−01A−11D−1926−05
TCGA−CM−6674−01A−11D−1837−05
TCGA−CM−5863−01A−21D−1837−05
TCGA−CK−4948−01B−11D−1651−05
TCGA−CA−6717−01A−11D−1837−05
TCGA−G4−6625−01A−21D−1772−05
TCGA−AD−6965−01A−11D−1926−05
TCGA−A6−6142−01A−11D−1772−05
TCGA−AZ−6607−01A−11D−1837−05
TCGA−CA−6715−01A−21D−1837−05
TCGA−A6−6649−01A−11D−1772−05
TCGA−A6−2684−11A−01D−1551−05
TCGA−AA−3511−01A−21D−1837−05
TCGA−AA−3713−01A−21D−1721−05
TCGA−A6−2675−01A−02D−1721−05
TCGA−A6−5664−01A−21D−1837−05
TCGA−AD−6963−01A−11D−1926−05
TCGA−A6−6654−01A−21D−1837−05
TCGA−CM−4748−01A−01D−1407−05
TCGA−A6−6651−01A−21D−1837−05
TCGA−CA−5255−01A−11D−1837−05
TCGA−CM−5348−01A−21D−1721−05
TCGA−A6−6141−01A−11D−1772−05
TCGA−G4−6302−01A−11D−1721−05
TCGA−A6−2684−01A−01D−1407−05
TCGA−D5−5541−01A−01D−1651−05
TCGA−G4−6626−01A−11D−1772−05
TCGA−AA−3712−01A−21D−1721−05
TCGA−G4−6295−11A−01D−1721−05
TCGA−D5−6536−01A−11D−1721−05
TCGA−AZ−6608−01A−11D−1837−05
TCGA−A6−5662−01A−01D−1651−05
TCGA−F4−6808−01A−11D−1837−05
TCGA−CM−6676−01A−11D−1837−05
TCGA−F4−6854−01A−11D−1926−05
TCGA−F4−6459−01A−11D−1772−05
TCGA−CK−5912−01A−11D−1651−05
TCGA−G4−6297−01A−11D−1721−05
TCGA−A6−6648−01A−11D−1772−05
TCGA−D5−5539−01A−01D−1651−05
TCGA−AA−3697−01A−01D−1721−05
TCGA−D5−5537−01A−21D−1926−05
TCGA−A6−6140−01A−11D−1772−05
TCGA−A6−2679−11A−01D−1551−05
TCGA−G4−6298−01A−11D−1721−05
TCGA−DM−A1D0−01A−11D−A153−05
TCGA−AD−6895−01A−11D−1926−05
TCGA−AA−3494−01A−01D−1407−05
TCGA−D5−6539−01A−11D−1721−05
TCGA−G4−6309−01A−21D−1837−05
TCGA−CA−5796−01A−01D−1651−05
TCGA−AZ−4681−01A−01D−1407−05
TCGA−CM−4747−01A−01D−1407−05
TCGA−DM−A28E−01A−11D−A16X−05
TCGA−CM−5349−01A−21D−1721−05
TCGA−D5−6928−01A−11D−1926−05
TCGA−DM−A0XF−01A−11D−A153−05
TCGA−AZ−4313−01A−01D−1407−05
TCGA−A6−2680−11A−01D−1551−05
TCGA−CM−6167−01A−11D−1651−05
TCGA−CA−5797−01A−01D−1651−05
TCGA−G4−6314−01A−11D−1721−05
TCGA−D5−6534−01A−21D−1926−05
TCGA−AD−6888−01A−11D−1926−05
TCGA−AZ−4308−01A−01D−1407−05
TCGA−AZ−6601−01A−11D−1772−05
TCGA−G4−6304−01A−11D−1926−05
TCGA−F4−6463−01A−11D−1721−05
TCGA−CM−6162−01A−11D−1651−05
TCGA−F4−6809−01A−11D−1837−05
TCGA−AZ−5403−01A−01D−1651−05
TCGA−F4−6856−01A−11D−1926−05
TCGA−A6−2685−01A−01D−1407−05
TCGA−AD−6899−01A−11D−1926−05
TCGA−G4−6302−11A−01D−1721−05
TCGA−F4−6703−01A−11D−1837−05
TCGA−AA−3662−01A−01D−1721−05
TCGA−D5−6924−01A−11D−1926−05
TCGA−DM−A1DB−01A−11D−A153−05
TCGA−AA−3489−01A−21D−1837−05
TCGA−CM−5868−01A−01D−1651−05
TCGA−CM−6674−01A−11D−1837−05
TCGA−AZ−6603−01A−11D−1837−05
TCGA−CK−4948−01B−11D−1651−05
TCGA−D5−5540−01A−01D−1651−05
TCGA−G4−6625−01A−21D−1772−05
TCGA−D5−6533−01A−11D−1721−05
TCGA−A6−6142−01A−11D−1772−05
TCGA−AU−3779−01A−01D−1721−05
TCGA−CA−6715−01A−21D−1837−05
TCGA−G4−6298−11A−01D−1721−05
TCGA−A6−6137−01A−11D−1772−05
TCGA−AA−3511−01A−21D−1837−05
TCGA−CK−6748−01A−11D−1837−05
TCGA−A6−2675−01A−02D−1721−05
TCGA−AY−5543−01A−01D−1651−05
TCGA−AD−6963−01A−11D−1926−05
TCGA−CM−5344−01A−21D−1721−05
TCGA−CM−4748−01A−01D−1407−05
TCGA−DM−A1D6−01A−21D−A153−05
TCGA−CA−5255−01A−11D−1837−05
TCGA−DM−A28C−01A−11D−A16X−05
TCGA−A6−6141−01A−11D−1772−05
TCGA−AM−5820−01A−01D−1651−05
TCGA−A6−2684−01A−01D−1407−05
TCGA−AA−3488−01A−01D−1407−05
TCGA−G4−6626−01A−11D−1772−05
TCGA−G4−6294−01A−11D−1772−05
TCGA−D5−6536−01A−11D−1721−05
TCGA−AA−3510−01A−01D−1407−05
TCGA−A6−5662−01A−01D−1651−05
TCGA−A6−6650−01A−11D−1772−05
TCGA−CM−6676−01A−11D−1837−05
TCGA−F4−6807−01A−11D−1837−05
TCGA−F4−6459−01A−11D−1772−05
TCGA−CA−5256−01A−01D−1407−05
TCGA−G4−6297−01A−11D−1721−05
TCGA−D5−5538−01A−01D−1651−05
TCGA−D5−5539−01A−01D−1651−05
TCGA−G4−6310−01A−11D−1721−05
TCGA−D5−5537−01A−21D−1926−05
TCGA−CK−6747−01A−11D−1837−05
TCGA−G4−6298−01A−11D−1721−05
TCGA−A6−6138−01A−11D−1772−05
TCGA−AD−6895−01A−11D−1926−05
TCGA−G4−6306−01A−11D−1772−05
TCGA−D5−6539−01A−11D−1721−05
TCGA−D5−6529−01A−11D−1772−05
TCGA−CA−5796−01A−01D−1651−05
TCGA−AD−6964−01A−11D−1926−05
TCGA−CM−4747−01A−01D−1407−05
TCGA−A6−5667−01A−21D−1721−05
TCGA−CM−5349−01A−21D−1721−05
TCGA−CM−6164−01A−11D−1651−05
TCGA−DM−A0XF−01A−11D−A153−05
TCGA−F4−6855−01A−11D−1926−05
TCGA−CM−6167−01A−11D−1651−05
TCGA−A6−6780−01A−11D−1837−05
TCGA−G4−6314−01A−11D−1721−05
TCGA−A6−5656−01A−21D−1837−05
TCGA−AD−6888−01A−11D−1926−05
TCGA−DM−A28G−01A−11D−A16X−05
TCGA−AZ−6601−01A−11D−1772−05
TCGA−DM−A282−01A−12D−A16X−05
TCGA−F4−6463−01A−11D−1721−05
TCGA−CM−6170−01A−11D−1651−05
TCGA−F4−6809−01A−11D−1837−05
TCGA−G4−6317−01A−11D−1721−05
TCGA−F4−6856−01A−11D−1926−05
TCGA−AD−6899−01A−11D−1926−05
TCGA−AA−3662−01A−01D−1721−05
TCGA−DM−A1DB−01A−11D−A153−05
TCGA−CM−5868−01A−01D−1651−05
TCGA−AZ−6603−01A−11D−1837−05
TCGA−D5−5540−01A−01D−1651−05
TCGA−D5−6533−01A−11D−1721−05
TCGA−AU−3779−01A−01D−1721−05
TCGA−A6−6137−01A−11D−1772−05
TCGA−CK−6748−01A−11D−1837−05
TCGA−AY−5543−01A−01D−1651−05
TCGA−CM−5344−01A−21D−1721−05
TCGA−DM−A1D6−01A−21D−A153−05
TCGA−DM−A28C−01A−11D−A16X−05
TCGA−AM−5820−01A−01D−1651−05
TCGA−AA−3488−01A−01D−1407−05
TCGA−G4−6294−01A−11D−1772−05
TCGA−AA−3510−01A−01D−1407−05
TCGA−A6−6650−01A−11D−1772−05
TCGA−F4−6807−01A−11D−1837−05
TCGA−CA−5256−01A−01D−1407−05
TCGA−D5−5538−01A−01D−1651−05
TCGA−G4−6310−01A−11D−1721−05
TCGA−CK−6747−01A−11D−1837−05
TCGA−A6−6138−01A−11D−1772−05
TCGA−G4−6306−01A−11D−1772−05
TCGA−D5−6529−01A−11D−1772−05
TCGA−AD−6964−01A−11D−1926−05
TCGA−A6−5667−01A−21D−1721−05
TCGA−CM−6164−01A−11D−1651−05
TCGA−F4−6855−01A−11D−1926−05
TCGA−A6−6780−01A−11D−1837−05
TCGA−A6−5656−01A−21D−1837−05
TCGA−DM−A28G−01A−11D−A16X−05
TCGA−DM−A282−01A−12D−A16X−05
TCGA−CM−6170−01A−11D−1651−05
TCGA−G4−6317−01A−11D−1721−05
ADH1C
Aldh1a1
AKR1B1
Aldh1a2
DHRS4
RDH13
Sult1a1
Adh5
Rxrb
Dhrs3
RxRa
AKR1C1
Rbp1
RaRb
AKR1C3
RaRg
ADH1B
AKR1C2
CYP26B1
RDH11
Retsat
RaRa
Crabp1
CYP16A1
RxRg
RDH10
AKR1B10
LRAT
ah 4C g
N l f D l mA
l N ey m
e l o f
A y A mDmh l D A hb l o h
ml A
mhN l
l
y b l f D l mA ehm D y m AmhN D mDmA y N eD y r 3 A laN mh
y N eD y r 3o l f e hN l f D l
A y
D l b
9n t 9cg 9 gR t hgRec ec gR
h
( W,
encg cgo a n S9nge nA g n cge lah 9 e n m c 9 c nf
e egc RaS nt gRal ge9 e t gRal ge9 9t S ne cg
v g 9g cR9m KhmR g Re Rle Rgc gR Sn f l 9l9 c h
9
colgc
gRec lg n ge9 e Sn9
c oc
c
c ne
75K
nla cg
c 9 got 9ne cecA llot e e gRec lacec n e p
9f hgR n m n ( F
2
9
gR t
c 9m n ol g
vz( AWFb K
9nt l 9l9 gR9c
lc9
Hypermethylation cases (%)
D C2 g l l yl
l yya b
e pRe eg
S gRm a v e on
ce e e g Sn9t 9g n
l ( W e on 2( KA
c ce e e gla RaS nt gRal g
S f lo q, A, WK9n
Norm mean
Norm sd
Tum mean
Tum sd
Wilcoxon pval
adj pval
t gRal ge9 A
RaS nt gRal ge9 e gR got 9n c t Sl v
c RaS nt gRal g
3W,
ge 9 ce n
v 9t S n
t 9n gR ) Wb 9 gR
megR
c cA
ALDH1A1
0.290
0.063
0.420
0.150
3.17E-08
2.06E-07
AKR1B1
0.019
0.019
0.397
0.234
7.91E-15
2.06E-13
ALDH1A2
0.080
0.025
0.343
0.230
4.37E-09
3.78E-08
RBP1
0.094
0.045
0.385
0.257
2.30E-07
9.96E-07
RARB
0.036
0.013
0.188
0.211
0.130
0.198
AKR1C3
0.084
0.041
0.069
0.092
7.64E-08
3.97E-07
DGAT1
0.014
0.002
0.081
0.137
8.33E-06
2.71E-05
CYP26B1
0.091
0.133
0.323
0.270
9.63E-07
3.58E-06
CYP26A1
0.128
0.042
0.359
0.249
0.000
0.000
LRAT
0.060
0.048
0.330
0.299
0.002
0.005
RARA
0.052
0.021
0.084
0.106
0.363
0.429
45.669
79.921
71.654
66.142
45.669
7.480
39.370
57.874
61.811
55.512
27.559
DyaN N hf m l
y mi hf m A o
h Al DDf N l f D l
A y h Dl
o l
l
A y A AmhN D A laN mh mDmA
e l o f
f
A Df y y m
gR
e n g
N l f D l mA msh aD l mA
pgh m
9 ce n
9n
pSl9n
t gRal ge9 R n A gR c
Re Rle Rg
9yt gRal g
c
n
9nn l ge9 e t 9 ot
c
n9 cgo e c gR g 9y pSn cc
o ge9 lla n l g
SRace lla e gR o l oc cgno gon v uol u9 d
S nR Sc e g n ge lt ch) , ( FD 9og f
h
) , , ) D n c nd e ut 9n h) , , 7KA
m9 n
S gRm a
mRe R 9gR n
9t SlecR egh m
S gRm a
vRggSNI I
e gR
lar
m n
fe A
v a9g9 a l9S
g n cge lahe llot e lar
9gR
ne Rt g 9 n ge 9e
9t t 9 S gRm a 9n e9l9 e l
9yt gRal ge9 9
e 9l9n g l
R 9 9 gR ( (
n megR gR n cg 9
c
3W,
nn a vt g ne lc t gR9 cKA 9yt gRal g
megR
gR
9g ge9 A e n A 9f I Kh 9n
e 9
megR gR RaS nt gRal g
gRn9o R gR c RaS nt gRal ge9 Sn9 cc cA 9n n
c RaS nt gRal g
n Sn c g
c
9yt gRal g
c g9 lo e g mR gR ngR n ec
o ge9 cS e e lla lg n
g9
c n
c v g99l
g9l9 aK S gRm a
9t cK lacecA
o ge9 l e nl g
o ge9 l
S gRm a lacec m 9 c nf
o ge9 chco R c
on9 l e g n ge9 ch e n ge ge9 h
ll t e n ge9 f l9St g l Sn9 cc ch
n ge 9e
e
( W(
colgc
f eg t e ce le S gRm ac v e on 2) KA 9 gR nhgR c
Sn9 cc 9 9yt gRal ge9 e f 9lf e e n gn ge 9e
A. KEGG pathway analysis!
&
g coSS9ng cS e e
e nl g
cA
B. GO functional analysis!
"
#
! #
! #
! #
#
"
! #
&
! #
)%(( &()
)%(( &(-
)%(( &(0
)%(( &)+
)%(( &).
ah 4p g msN l f D l mA A Df y y m l
y A
Al msN l f D l ehm y o h y D l
N l f D l mA m l
A
e l o f y e hl DDf h ylmh
t 9 cgn g
a gn ge h) , ( , KA 9
cel e gR
g
llc megR gR
9 gR n ge 9e
( ( zh
)%(( &*-
Df
Dl h
e S gRm ah m
( (h
( (h
oS9 ( (h
pSn cce9 Wy r
cc9 e g
9 9g R f
( )h
3W, K Sn9 el c e llc vu 9 u 9og 9n
t gRal ge9 A
( )
t e et l R megR gR l9cc 9
gm ) z ( Kh mRe R m n
c
S nge lla
llc cR9m
ll le v
c 9n 9 ) z ( K ecSl a
S n
e n c 9
t gRal ge9 v e on 2FKA
t gRal ge9 S e cc9 e g
((z
S gRm a
Sn9t 9g nc v e on 2FKA 9on 9 gR RaS nt gRal g
c ec 9g l nA c megR e g nt
gR S n g l
v Ro pSn cce9 v el g
((z
gn gt gA
( Fh
9 cc9 e ge9 pRe eg
R f a t gRal ge9 v g f lo T, AWKe ef
t gRal ge9 9 gR c t
cR9m
lar
t gRal ge9 e gR
t gRal ge9 v llot e megR Wy r
( Fh
t gRal g
c
g Wy r
F KhmRe R R f f na l9m l f lc 9
( ( z ecSl a
c v
t gRal ge 9og gR n9l 9
nn aK ( ( z gn g
( ( W)
l
gR g Se ge lla cel gg n Se gon
on ne g F 2pz,
)%(( &)/
l Am
e l o f f e hN l f D l
A yb
lme h mhN
r 3 A
r 3 A Df y yb
h l Am
l nla
gef g
v
)%(( &))
neh yy mA
gR c
n
)%(( &(,
pSn cce9 h og gRec
gR n9l 9 t gRal ge9 e gR9c ue e g 9n l9m l f lc 9
)z (h
g ll e (h
h
pSn cce9 A
( (
9
t gRal ge9 e h
( K
colgc
ah 4G g mN e h ymA m AmhN DK
neh yy mA l i y N l f D l mA | Fi Da mh
ml
f e hN l f D l
A y o l A h l Am
e l o f A A l l h N l f D l mA N m DyB
CC5
r Da h mN ay3R
CC5 lh l o l 2s K
rh yt a h y3 A
DDDA y r h A lh A D y3b
AmN
9
m
ehm DA m l
e l o f A mDmh l D A h
ge ce Rgc e g9 gR e f 9lf t g9 gR
ongR n e g nn9 g
n i o e c ge 9e
gR
9Sa ot
vgRn9o R
n lg n ge9 v
e S gRm a ec S99nla
c n f
l
c e 9l9n g l
e9S9ng l m
n
g99lK 9n t og ge9 Ke 9l9n g l got 9nc v e on 23KA
g
a
S gRm a t og ge9 ec l9mhn Sn c ge 9 gR
S gRm a
9t e
lg n ge9 cA R 9f n ll
3A( b v) 3 9og 9 W2WK9 gR
c cD 9 Re R n i o a 9 t og ge9 c v e on 23 KA 9cg 9 gR t
n t ecc c t og ge9 c v g 9gcR9m KA
gR 9gR n R h
e c pRe eg
SS
l9cc c
lg n ge9 c nec oS g9 7A5b v3z 9og 9 W2WKhmegR 9gR
S e n i o g 9Sa ot
9 gR
n
v e on 23 KA g n cge lah
e c nc g9 pSl e gR oSn ol ge9 9 c nf
f e gR
(,
t Sle e ge9 c v) A5b Kh gR g
pSn cce9 lacec
v e on 23 KA
( WF
colgc
A. Mutation analysis!
B. CNA analysis!
C. RDH10 expression vs CNA!
ah 4. g h Al
e l o f
A A
Dl h l mAb
( W3
m N al l mAy r 3 A mef AaN h Dl h l mAy r 3
lA
mDmh l D A h y N eD yb 3
Cd neh yy mA i y l
Dl h
mef AaN
h
Discussion EPIGENETIC REGULATION OF AKR1B1 GENE FAMILY IN COLORECTAL CANCER Epigenetic regulation of AKR1B1 The DNA methylation pattern in the genome is basically bimodal, with the large majority of CpG sites methylated at high levels (>85%) and CpG islands largely unmethylated (<10%) (Bergman & Cedar, 2013). This apparently stable DNA methylation scenario is modified in a number of diseases and it has been extensively studied in cancer. DNA methylation alterations in cancer are mainly due to global hypomethylation and focal hypermethylation (Bergman & Cedar, 2013). Many gene promoter hypermethylations have been described in cancers, and in most of them a clear association with a decrease in gene expression has been reported (Table 16). We started with the observation that AKR1B1 promoter CpG island was hypermethylated in 24 out of 25 cases of colorectal cancer patients. The high incidence of this alteration was confirmed using the TCGA database, where we detected hypermethylation in about 85% of the colorectal cancer cases analyzed (Figures 33 and 37). The interesting fact is that AKR1B1 changes in DNA methylation do not appear to affect its expression, at least in a consistent manner. On one hand, we know that the gene promoter is highly hypermethylated in cancer, but we observed a very faint downregulation (2% of the cases) in the expression of AKR1B1 in colorectal cancer, at both mRNA (results from three different sources: our own data, TCGA and ICO data) and protein level (western blot and TMAs performed in our laboratory). On the other hand, we have checked the expression of this gene after the 5-­‐AzadC treatment and, despite that the CpG island is hypomethylated, gene expression does not change at all. Gathering all these data we could confirm that the expression of AKR1B1 gene is not affected by the DNA methylation status of its promoter. 159 sh Nhhs1 REVIEWS
Table 1 | Representative genes commonly methylated and silenced in CRC
Gene
Protein
Function*
Effect of loss of function
APC
Adenomatous polyposis coli
Wnt signaling pathway inhibition
Increased Wnt/β-catenin signaling
MLH1
MutL homolog 1
DNA mismatch repair
Microsatellite instability
MGMT
O-6-methylguanine-DNA
methyltransferase
Repair of alkylation DNA damage
Increased G>A mutation frequency
RASSF1A
Ras association domain family 1
(isoform A)
Negative RAS effector, proapoptotic,
microtubule stabilization
Increased RAS/RAF/MAP kinase
signaling, death-receptor-dependent
apoptosis
SLC5A8
Sodium solute symporter family 5
member 8
Sodium and short chain fatty acid
transporter, suppresses colony
formation
Not known
RUNX3
Runt-related transcription factor 3
Transcription factor
Decreased TGF-β/BMP signaling
MINT1‡
Methylated in tumor locus 1
NA
NA
Methylated in tumor locus 31
NA
NA
SFRP1
Secreted frizzled-related protein 1
Wnt antagonist
Increased Wnt/β-catenin signaling
SFRP2
Secreted frizzled-related protein 2
Wnt antagonist
Increased Wnt/β-catenin signaling
CDH1
E-cadherin
Calcium dependent cell–cell
adhesion glycoprotein
Loss of cell adhesion, possible
increased Wnt/β-catenin signaling
CDH13
Cadherin 13
Selective cell recognition and
adhesion, antiapoptotic
Increased PI3K/Akt/mTOR
signaling, MAPK signaling
CRABP1
Retinol-binding protein 1
Carrier protein for transport of
retinol, promotes apoptosis
Not known
CDKN2A/p16
Cyclin-dependent kinase inhibitor 2A
Regulates cell cycle G1 progression
Increased cell proliferation
HLTF
Helicase-like transcription factor
dsDNA translocase, fork remodeling
activity, ubiquitin ligase
Impaired DNA repair
CDKN2A
(P14, ARF)
p14(ARF)
Inhibits E3 ubiquitin ligase
Decreased p53 stabilization and
activation
MINT31
‡
ESR1
Estrogen receptor 1
Ligand-activated transcription factor
Loss of estrogen receptor signaling
TIMP3
Tissue inhibitor of
metalloproteinase 3
Inhibition of MMPs and ADAMs
Increased EGFR signaling, TNF
signaling
CXCL12
Chemokine (CXC motif) ligand 12
Alpha chemokine
Increased tumor cell metastases
ID4
Inhibitor of DNA binding 4
Transcription factor
Not known
IRF8
Interferon regulatory factor 8
Transcription factor
Interferon signaling
THBS1/TSP1
Thrombospondin 1
Cell-to-cell and cell-to-matrix
adhesive glycoprotein
Decreased TGF-β1 signaling
DAPK
Death associated protein kinase
Induction of cell death
Interferon gamma signaling, TNF
alpha signaling, Fas/APO1 signaling
VIM
Vimentin
Stablizing cytoskeleton
No known biological effect
SEPT9
Septin 9
GTPase, formation of filaments
Impaired cytokinesis and loss of
cell cycle control
*Many of these gene products have multiple functions. The listed function in this table is the one most commonly cited as the one responsible for CRC
formation. ‡MINTs are ‘methylated in tumor’ loci, and are not specific genes. Abbreviations: ADAM, A Disintegrin and Metalloproteinase; CRC, colorectal cancer;
EGFR, epidermal growth factor receptor; MAPK, mitogen-active protein kinase; MMP, matrix metalloproteinase; TNF, tumor necrosis factor.
B| A
E p i ie
p
c
i c i
p
l
E g
E
r
E cuT1BBl
progression sequence.128 The proportion of neoplasms
methylated
genes have in the initiation and progression
of the polyp→cancer sequence continues to evolve as
that carry methylated genes is not markedly different
T emethylated
heN y i genes
h hi are
1m ei e (Figure 2).
D h msei between
5 Tn shhsA
1e Tand
hi adenocarcinomas,
1m 1 e additional
identified
advanced5T1n
adenomas
As might be expected given the detection of methylbut the frequency of methylation of specific genes does
D11
1TT
o es1 n eiatyo
es1 he vary
e 1between
i early
adenomas,
T T Ds1 advanced
h > T adenomas
w
ated genes
in the
normalmsei
mucosaeiof people
increased
and
risk of CRC, methylated genes can be detected in
adenocarcinomas.128,129 LINE-1 DNA hypomethylation
0)
<B8wcrypt
i shfoci
sh and
1eadenomas
ei
h well
1 as in adeno< <b h
sehpolyp→cancer
5T1n 1e Tsequence
5 sho
a similar
1 h pattern
1e
aberrant
as
in the
shares
carcinomas.123 Examples of genes methylated early in the
with local promoter hypermethylation in that there is
55 T e1
s sequence
sei include
T ITGA4,
esAMGMT,
1T 5 a Tn
shhsA difference
he e w in the
levels
b mof LINE-1
i A hypoe1
substantial
adenoma→carcinoma
123–128
methylation and local hypermethylation occurring at the
SLC5A8, SFRP2 and MINT1.
Notably, the most
i substantial
y51ei hsl
ei e ei n s 5sD es n i polyp
shnto adenoma
1 T DNo
es1 1
< < n sDi e
step of the sequence, but no difference
increase in gene methylation is seen in the
between levels at the adenoma to cancer step.130 There is
normal mucosa to adenoma step of the polyp→cancer
1e
s ei
5 sho w
NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY
<3)
VOLUME 8 | DECEMBER 2011 | 693
sh Nhhs1 h s i Nn bei
5T1n 1e T 5 sho 1o1 eshhN b i 1m A T m
h n 5o h oyl
N oe T
1No
e e
w h5se ei sh
> sDNT 2B8w 1Nehs
sh 1n 5o e oy N n ei yo e
i y5 Tn ei yo es1 s h1n
ebei
e ei
T1n ei
1
t 5T hhs1 1
T DNo e1Ty T Ds1 1
5 sho w ei
m1Tu 1
1Ne W<
n 1 No es D
1e i A
Nh ei s u ei e
< < D t 5T hhs1 n c1T T1o s ei
1e 1TT o e msei
ei
D n 1 heT e
hh1 s e
5 sho n soy,
< < s ei
i T1n 1h1n
s 1Tn o 1o1 e Tn h 1 n
sh5o y
s T h
1
T 0 sh
sn 51Te 1 ei sh
T h51 hs o
1T
1 h
5 sho 1 h
h hw
B B6
em i y5 Tn ei yo es1 1
< <bm s A hesD e
< <2 D hbmi s i
. 9BBw ei sh
T n es
1ei T n n
T
oh1 o1 e
Th 1 ei
t e e1
h b 1ei D h T i sDi oy t 5T hh
1m T DNo es1 s < <2 T
Tw i sh shbs ei h em1 D h
1T 1A Tb
5T1n 1e T 5 sho 5T1e s o A ow NT5Tshs Doyb s h5se
< <)
ei
< <b h m i A i y51ei hsl
ei o u 1 1TT o es1 < <) y
< <
mi i y5 Tn ei yo es1 1 ei
hh y
t 5T hhs1 5 Ti 5h ei
B B1
m i A
5T1n 1e Th 1
oh1 1 h TA
1m T DNo es1 s n 1he1 ei
E g i
< hN
n 1 heT e
x28w
s s D hse h T ei
T DNo es1 1
1A w i sh msoo t 5o s mi y ei
i
> sDNT
< <
T0
sho 1 ei sh T Ds1 1T
1A T t 5T hh w i
T T Ds1 n
5
oN s T h
t 5T hhs1 i e ic
T1n ei
ei
T 0w i y
sn 51Te i E
h Ts
j shi sn NT b 0) ) 28w
E
sh
1 e s h B s s D hse h
1Tei
i
8
1Ne 1
i sh T Ds1 gE v6 A E T
B Bl
< < > < < t 5T hhs1 b
> shi s u r
yB
T
on 1he
1T
mi s i sh o1 e
T n s shi s u
u y T Ds1 ei y
1 ei
s 1o1T e o
T s h A T hso s Db ei
s eoy i y51n ei yo e
5T1n 1e T n ei yo es1 s 1o1T e o
1 h 1e n e i
<3<
sh Nhhs1 sei Tmsei ei
t 5T hhs1 we sh m1Tei e1 n es1 ei e
T 1j t 5T hh
s 1Tn o gE v| A
i
E
i
ig
E p
i c i
p
eTsus Doybei T sh os u
ei
1m T DNo es1 1
i y5 Tn ei yo es1 1m T DNo es1 i y51ei hsl
t 5T hhs1 1
<30
ei
1
B B
i c i
p
t Rl
K
p
KE
em ei
< <)
ei
1
< <
< <)
i c i
E p i
n i
B
ic
p t E pp
tR
n ei yo es1 1 ei
< < D < <2 D hw n 1he
5
sho < <2w
1TT o e h msei
< <2 > sDNT
51hhs o T DNo e1Ty T1o 1
< <) < <2
eNn 1TeshhN w
in
p
< <) < <
h h ei
hN he es o
x38w i T 1T
5
l
E i
sho s m
ei
Discussion ENHANCER ACTIVITY OF AKR1B1 CpG ISLAND Chromatin conformation in the AKR1B locus Due to the clear evidences that pointed out a relationship between AKR1B1 CpG island and the regulation of the expression in AKR1B10 and AKR1B15, we decided to investigate the chromatin conformation of AKR1B genes. It is known that enhancers need a close contact with the promoter in order to regulate gene expression (Carter et al.,2002; Tolhuis et al., 2014), for this reason we decided to analyzed the structural conformation between AKR1B1 CpG island and AKR1B10, which is more downregulated than AKR1B15. The Chromosome Conformation Capture (3C) results revealed a close contact between AKR1B1 CpG island surroundings and the AKR1B10 promoter in two cell lines that highly express and do not express AKR1B10 (HepG2 and HCT116, respectively). At that time we wondered if the chromatin interactions of the AKR1B genes promoters would change in carcinogenesis and if this alteration could lead to the observed silencing of AKR1B10 and AKR1B15 genes. Surprisingly, although the expression of AKR1B10 was very different between HepG2 and HCT116 cell lines, the chromatin conformation of the area was almost identical, what suggested that the silencing of AKR1B10 and AKR1B15 is produced by deregulating the enhancer function of AKR1B1 CpG island without altering the chromatin structure. This result suggests that DNA methylation, and perhaps other epigenetic mechanisms, are the main factors that regulate the expression of AKR1B10 and AKR1B15. A distant enhancer regulates the expression of AKR1B10 and AKR1B15 To get more insights about the role of AKR1B1 CpG island as distal regulator we decided to use enhancer luciferase approach. The results indicated that whole CpG island already had enhancer activity, however we reduced the size of the fragment to elucidate the TFs responsible for the expression of AKR1B10 and AKR1B15 genes. Finally, we found a single transcription factor binding site (TFBS), used by NF-­‐Y and C/EBP-­‐β, which was associated with an important increase of the luciferase signal. It is worth to mention that within the island, 163 Discussion other fragments also present enhancer activity (F2.1 or F2.3.2), but the mentioned binding site was the one that produced more luciferase signal. Luciferase experiments provide important clues about the activity of the region, but it is also a limited approach, as the genomic context (nucleosome occupancy, histone marks, DNA methylation, distal interactions and more) is not maintained in this assay. For these reasons we decided to manipulate the original NF-­‐Y and C/EBP-­‐β TFBS, using CRISPR technique (materials and methods) to alter the binding site and study the consequences in the expression of AKR1B10. Finally, we obtained a clone from HepG2 cell line that had a deletion in the TFBS resulting in a downregulation of AKR1B10 gene compared with the control cells. This is a direct evidence that the TFBS identified by enhancer luciferase experiments indeed plays a role in the regulation of AKR1B10 and AKR1B15. Next step was to find which TF is the responsible for the enhancer activity: NF-­‐y or C/EBP-­‐β. To answer this question we performed CRISPR experiments to truncate both TF separately. Unexpectedly, the truncation of both TF produces the downregulation of AKR1B10. This means that both proteins are required to activate the expression of AKR1B10 and AKR1B15. Intriguingly, AKR1B1 did not display changes in gene expression, although the two TFs bind to their promoter region. Gathering all the data we could confirm that AKR1B1 CpGi works as an enhancer of AKR1B10 and AKR1B15. Furthermore, the activator role of the enhancer is due to both NF-­‐Y and C/EBP-­‐β binding. These results are in agreement with a recent study showing that in cancer cell lines, a large number of genes are regulated by DNA methylation changes in enhancer regions rather than in its promoter (Aran et al., 2013). H3K27ac and PRC2 dynamics in the activation/silencing of AKR1B10 and AKR1B15 Back to 2012, Dr. Batool Akhtar-­‐Zaidi et al. analyzed the enhancer histone marks H3K4me1 and H3K27ac to determine gain and loss of enhancer activity at genome-­‐wide level in primary colon cancer cell lines relative to normal colon crypts. They identified thousands of variant enhancer loci (VEL) that comprise a signature that is robustly predictive of the in vivo colon cancer transcriptome. 164 sh Nhhs1 i
i
jh 9 oyhsh 1 ei
0) <) 8bT A o
ei e
Tw ei
Ty5eh e e
0) &. o1he
g
Ts i
i
E
B 0.
s 1o1T e o
1T B 0.
s 1Tn o
w Nei 1Th
h 1n n 1 e1 sA 1T n 1T
oo os hb
i
B 0.
b
s h 1n n 1 e1 hst 1Tn 1T
p E
Nei 1Th
s eseyb T
os hw
E iE
pi
n 1 heT e ei eo1he i T n 1T
jTshu
e ei D s i 1 1T e
hei D s Tw ui e T mi 1o jD 1n
1Tn o 1o1 m 1 h TA
i
T 1A Ty 1
B 0.
T j t 5T hhs1 1
s ei
Th s B B1
E
h> ui e Tj
1n 5 s h ei
< <) < < 5
sho B 0.
1 b0) <08w
s 1NT T Ds1 w sThe s ei esho1hes <<3 T eT e
< <2
e Tei
s eNn 1T
Ty5e
n 1T T1 Nheoy hh1 s e
s s
esA es1 1 ei
Ti A
h ey5s y 1o1 1 h TA es1 h m T T 5T1 N
5 u1
1o1T e o Thw 1he
T1hh eNn 1Th e T hes Doybei sh 5 u sh T 1A T
B 0.
h T
> T yDi e1 9NsT
e e o o A oh 1
03) & D s DT e T N es1 o
ei
h
B 0.
El
1T 1A Tbei
msei
esA n Tu
1Ne &) 6 1 D s hi 1m AsTeN ooy 1
gE vs A
E i
T
1ei T hse b. ) 6 1 o1he
h5 s s ooy
i i <<3
msei 2j l
wi
Tbmi s i
t5 o s h
eT en ew i
ooh n sDi e
T oohw
T n 1A
h 1
y ei
<32
sh Nhhs1 1oy 1n
5T hhsA
eyo h h >
ei
1n 5o t
8e1 ei
s T h s o1 De Tn s 08w
0 m1No
T Ds1 h mi e m1No 5T1 N
B 0. n
5T1 hh m1No s oN
0 >
n Tu y
oh1 ei
esA es1 1 ei
0>
ei
eeT e i she1 T hs D1
B0 .
0 hN N se8w i sh hso s D
n ei yoeT h T h h ei e m1No T hNoe s ei
i T> sDNT x. 8w
"
#
"
!
gE vv A
ie i
i
B B1
i sh i y51ei hsh sh N ooos B 0.
E i E 6a T5s y 1NT i
sh heT1 Doy i y5 Tn ei yo e
5 u s ei
B 0.
5 u
T hNoehw
sh5o yh B 0. n B < <)
y ei
il
<<3 1o1T e o
0 s ei
< < 5
ooh msei 2j l
T
0 s ei
T Ds1 sho b m 1 h TA
< <2b mi s i sh N e1 ei
T n 1A o 1
< < 5 sho b 1 h 1ei A
> sDNT 328w 1m A Tb mi m eT e ei
t 5T hhs1 1
iE i
Tj
T 1 heseNes1 1
ei
1 h TA
n ei yo es1 > sDNT xx8w
A n 1T bmi 1 h TA
ei
<<3
ooh m T eT e
T j t 5T hhs1 1
msei
5>
0 s i s se1T8bm
< <) b Ne 1 i D s < < t 5T hhs1 > sDNT . 28w
ei Ts D oo ei
1 sTn ei e ei
< < 5 sho <33
5sD es
t 5T hhs1 e 1
5sD es he e b h5 s ooy ei
sh u y s s s D ei
< <) N osey
esAsey he e 1 ei
< <2bm
B 0.
i (
0b 1
Tw ei
Discussion contrary, the epigenetic state of AKR1B10 and AKR1B15 promoters does not seem to play an important role in the regulation of the genes. Gene regulation by enhancer hypermethylation in cancer The DNA methylation status of the enhancer has important consequences in gene expression. Dr. Dvir Aran et al. analyzed DNA methylation and gene expression data in 58 cell types and developed a machine-­‐learning algorithm for methylation-­‐expression relationships in gene promoters and enhancers. They mapped numerous sites at which DNA methylation was associated with expression of distal genes. Strikingly, some of these identified sites were drastically altered in cancers: hypomethylated enhancer sites associated with upregulation of cancer-­‐related genes and hypermethylated sites with downregulation (Aran et al., 2013). Moreover, changes in the DNA methylation state of the enhancer were better predictors of gene expression than changes in the promoter. Additionally, there was much more hypermethylation in enhancers than in promoters . Methylation of distal regulatory sites is closely related to gene expression levels across the genome. Single enhancers may modulate ranges of cell-­‐specific transcription levels, from constantly open promoters (Aran & Hellman, 2013). DNA methylation itself does not produce gene silencing; there are two proposed mechanisms by which DNA methylation hampers gene expression: by avoiding the binding of TFs to its targets or by recruiting DNA methylation binding domain proteins (MBD) that in general do not activate gene expression. Our observations showing an increase of C/EBP-­‐β binding in the AKR1B1 CpG island once demethylated by treatment with 5-­‐AzadC are concordant with Aran et al. data showing a depletion of TF in methylated enhancers. We also tested the importance of DNA methylation in the activity of the AKR1B10 enhancer. Using luciferase enhancer assay, we transfected the CpG island methylated and unmethylated, the result was that when the CpGi was unmethylated has an increased enhancer activity compared with the methylated construct. Our results are explained by two distinct mechanisms, the first one is based on H3K27ac. This mark is the responsible for the activation of the enhancer and is 167 Discussion fundamental for AKR1B10 and AKR1B15 expression. In those cases where there is no H3K27ac, the enhancer activity is completely lost. Secondly, the enhancer needs to attach some TF to be active. According to our results, NF-­‐Y and C/EBP-­‐β are the most important TFs, but as mentioned before, there are other regions and perhaps other TF that also contribute to express AKR1B10 and AKR1B15. Probably, DNA methylation plays a role in silencing the enhancer hampering the binding of TFs. Although we just describe H3K27ac and DNA methylation mechanisms separately, they are in fact steps of the same silencing process and it is difficult to elucidate the specific contribution of each one, as we always found H3K27ac loss and DNA hypermethylation of the enhancer to occur concomitantly in colorectal cancer. 168 Discussion FUNCTION OF AKR1B10 IN CANCER, LOCATION, LOCATION AND LOCATION The literature is plenty of studies reporting deregulation of AKR1B genes in cancer. According to previous works, AKR1B10 is most prominently upregulated in liver and lung cancer among others (Fukumoto et al., 2005; Heringlake et al., 2010; Kang et al., 2011; Schmitz et al., 2011; Woenckhaus et al., 2006). AKR1B1 over-­‐expression is more common amongst different tumor types than that of AKR1B10 (Laffin & Petrash, 2012). Alternatively, few studies have reported AKR1B1 and AKR1B10 under-­‐expression in human cancers. Regarding AKR1B15, there is not much literature of this gene and nothing about its alterations in cancer to date. In many papers, it has been described the advantages of AKR1B10 upregulation (Figure 23). However, AKR1B10 is deeply downregulated in colorectal cancer. Then, what makes colorectal cancer cells downregulate AKR1B10 while other cancers pursuit its upregulation? One possible explanation would be the one recently proposed by Ohashi et al.; they identified AKR1B10 as a direct transcriptional target of the tumor suppressor TF p53. Ohashi et al treated HCT116, LoVo and RKO colon cancer cell lines with adriamycin, which activates endogenous p53 protein by inducing DNA damage. As a response to the treatment, they observed an increase in the expression of AKR1B10 protein together with p53 in a dose-­‐dependent manner (Ohashi et al., 2013). Authors claim that p53 activates AKR1B10 by binding to a p53 response element (p53 RE) near the promoter (it is worth to mention that there are also p53 RE in the AKR1B1 CpG island). They proposed that the inactivating mutations of p53 in colorectal cancer could be the responsible for the AKR1B10 downregulation. However, the mechanisms must be more complex, because AKR1B10 is downregulated in most of the colorectal cancers cases, whereas the rate of p53 mutation in colorectal cancer is about 40% (Petitjean et al., 2007). We have also analyzed the relationship between the downregulation of AKR1B10 together with the mutation of p53 using tissue microarrays approach, and we could not observe any significant relationship (Annex XI). Moreover, AKR1B10 was not expressed when they used p53-­‐induced 169 Discussion transactivation directly (Ohashi et al., 2013). Moreover, what it is should be noted is that adriamycin is a substrate of AKR1B10 (Loeffler-­‐Ragg et al., 2009). For this reason, we think that adriamycin triggers AKR1B10 expression as a response of a cellular mechanism that pursuits the inactivation of the harmful agent. Activation of AKR1B10 upon adriamycin treatment needs to be very quick in order to protect the cell against DNA damage. We have seen that adriamycin activates more than 100 fold the expression of AKR1B10 in 24 hours and it is restored to normal levels in about 8 days (data not shown). This prompt activation may differ from the mechanism that regulates the expression of AKR1B10 in normal colon and does not explain the silencing produced in cancer; further experiments are needed to better understand how adriamycin induces the expression of AKR1B10. Consequences of AKR1B10 and AKR1B15 loss of function in colorectal cancer In contrast to the ubiquitously expressed AKR1B1, AKR1B10 is mainly expressed in normal small intestine and colon, which suggests that the physiological function of AKR1B10 is specifically required in intestinal tissues (Cao, 1998). Interestingly, AKR1B10 deregulation in cancer shows an inverted pattern to its condition in normal tissue: silencing in colorectal cancer (where the normal tissue expresses it), and upregulation in but cancers where AKR1B10 is lowly or not expressed in the normal tissue (i.e. pancreas) (Chung et al., 2012; Ma et al., 2012). Actually, AKR1B10 is more than just another tissue-­‐specific gene. Related to this, Dr. Batlle’s group described, in mouse and human, the genes responsible for the differentiation of intestinal stem cells (ISC) to well-­‐differentiated colonocytes (Merlos-­‐Suárez et al., 2011). ISCs are selected by high expression of two genes: the ISC marker EphB2 (Merlos-­‐Suárez et al., 2011) and the classical ISC marker Lgr5 (Barker et al., 2007). Both genes become gradually silenced when ISC differentiate. Using the gradient expression of both genes, they identified the panel of genes activated specifically at each point of differentiation: ISC, transient amplifying cells, late terminal amplifying cells and terminal differentiation. Terminal amplifying cells (TA) result from the expansion of ISC (through several 170 Discussion rounds of mitosis) while they migrate upwards along the crypt axis. Close to the intestinal lumen, TA cells undergo cell-­‐cycle arrest and become terminal differentiated cells. Among all these colon differentiation steps, AKR1B10 appeared in the panel of genes defining the signature for the late-­‐TA step (supplementary table 5 in Merlos-­‐Suárez et al., 2011 paper). However, due to the impact of AKR1B10 silencing in colorectal cancer, our hypothesis is that this gene is acting as a tumor suppressor gene, hampering the development of the tumor. A recent paper underpinned this hypothesis; Dr. Yao et al. described that AKR1B10 is downregulated in gastric cancer and observed that the expression of AKR1B10 was inversely correlated with tumor size, lymph node metastasis, distant metastasis and TNM stage. They concluded that the presence of AKR1B10 in the tumor is a good prognostic indicator in gastric cancer patients (Yao et al., 2013). AKR1B10 is expressed in gastric, colon and rectum normal tissues and in the cancers of these tissues is frequently downregulated. At this point, a question arises: what makes gastric, colon and rectum tissues different from others like pancreas, liver, lung, blood or breast, where this gene is upregulated in cancer? AKR1B10 is a metabolic enzyme that catalyzes the reduction of a wide range of substrates; the cellular implications of expressing AKR1B10 could vary depending on the available substrate. In other words, each tissue has different accessibility to different compounds, and the reduction of these distinct compounds could be beneficial or detrimental for the cancer cell growth. This may be why AKR1B10 is upregulated in some cancer cells and downregulated in others. At that moment, we realized that those cancers where AKR1B10 is downregulated (gastric, colon and rectum) are tissues that are in direct contact with the diet. Among all nutrients in the diet, we can find vitamin A ester form, also known as retinal. This compound is efficiently reduced by AKR1B10 to retinol (but not by AKR1B1), which is described as an inhibitor of the cellular growth in colorectal cancer cell lines, through an independent retinoic acid pathway (Dillard & Lane, 2007; Park et al., 2005; Park et al., 2008). Almost all the retinal coming from the diet is converted to retinol in the lumen of the gastrointestinal tissue, specially in the intestines (Fidge et al., 1968). Retinol is 171 Discussion then specifically conjugated to different proteins, distributed across the body and stored in the liver. To summarize, gastrointestinal tissues are distinctively exposed to high amounts of free retinal and retinol. Our hypothesis is that gastrointestinal cancers specifically silence the activity of AKR1B10 and AKR1B15 to avoid the reduction of the dietary retinal into retinol, which inhibits the growth of the tumor. The inhibition of AKR1B10 is an early alteration already present in adenoma, which would allow uncontrolled growth of tumor cells. In the line of this hypothesis, normal ISC do not express AKR1B10 until the last state of differentiation, when cells do not need to divide anymore. AKR1B10 (and perhaps AKR1B15) is expressed to perform its own functions but also to maintain the non-­‐dividing state of differentiated cells. Actually, this hypothesis is underpinned by our retinal treatment results. When we performed proliferation assays with colon cancer cell lines transfected with AKR1B10 or AKR1B15 and control, we observed a slight decrease in the proliferation of AKR1B10 and AKR1B15 transfected cells. However, when we do the same experiment adding retinal to the medium, cells that are transfected with AKR1B10 or AKR1B15 are deeply affected and are less proliferative compared with the controls cells. This effect is due to the increased amount of retinol that was produced by retinal treatment in the AKR1B10 and AKR1B15 transfected cells. The increased amount of retinol produced by AKR1B10 and AKR1B15 transfected cells was confirmed by studies performed in Jaume Farrés lab (UAB) (data not shown). These experiments pointed out the crucial antiproliferative role of AKR1B10 and AKR1B15 per se but also, and more important, in combination with retinal. Non-­‐gastrointestinal tissues, which are not exposed to free retinal, tend to overexpress AKR1B10 to acquire the pro-­‐tumor activities described in the literature, without the risk of growth inhibition by retinol production. 172 Discussion DEREGULATION OF THE RETINOIC ACID PATHWAY IN COLORECTAL CANCER As we have discussed above, gene silencing is a well-­‐known cellular process that commonly occurs in cancer (Khare & Verma, 2012; Nephew & Huang, 2003). The literature is plenty of examples of specific genes silenced in a given cancer type; however, in most of the studies the gene is analyzed overlooking its role in the pathway. Understanding the impact of one single gene alteration in the whole pathway is extremely important, because other genes could easily overcome the function of one silenced gene within the whole pathway. In the present study, we have screened for aberrant gene expression in genes directly involved in the retinoic acid pathway in normal colon, and we have focused our attention on the production of the key metabolites responsible for the adequate function of the pathway (Retinol, All-­‐trans RA and All-­‐trans 4 oxo-­‐
RA). Previous investigations have reported epigenetic silencing of several genes of the retinoic acid pathway. The most well studied is the promoter hypermethylation of RARβ gene, which affects several cancer types (Hayashi et al., 2001). Many other members of the pathway have also been described to undergo silencing through DNA hypermethylation of the promoter (Chim et al., 2005; Chou et al., 2012; Okudela et al., 2013; Tsunoda et al., 2009; Wang et al., 2003). Despite these works, there is not a comprehensive study of the retinoic acid pathway deregulation in cancer. Here we report a deep screening of the genetic and epigenetic alterations of the retinoic acid pathway genes in human colorectal cancer. We show that 13 genes of the RA pathway were downregulated in colorectal cancer (AKR1B1, AKR1B10, AKR1C3, ALDH1A1, CYP26B1, DGAT1, DHRS4, LRAT, RARA, RARB, RBP1, RETSAT and RXRA) and among them, nine displayed promoter DNA hypermethylation (AKR1B1, AKR1C3, ALDH1A1, CYP26B1, DGAT1, LRAT, RARA, RARB and RBP1) during the cancerous process. We have checked for the importance of promoter DNA methylation and expression on RA pathway genes and, in most cases, the results showed a strict correlation between methylation levels and changes in expression. Taken together, these 173 sh Nhhs1 e
n 1 heT e ei e
n ei yo es1 sh 5o ys D n c1TT1o s ei hso s D1
5 ei m y D hw 1T 1A TbD es n Ne es1 h 1ei A h5 s oT o A <) s oe T
T
1 ei hso 1o1T e o o1 u
D hd1 ei
Tsh n s oy N e1 D D h s ei mi 1o
5 ei m ybm
s 1o1T e o e sTheDo bm
1o1T e o
o1 u
ALDH1A1
i n c
n E g i
i
p
s T
1No 1 h TA ei eT es 1s
Tw i
< < 5e1Th,
5 1 s o T es 1s
5T1 N es1 1 ei
03 <w
b
e n
T
h
> 1n n 1 oy T
n 5os s es1 8 511Toy N 0) <) 8bmi s i n
1e 5T1 N
< h T es o8 y ei
e1 s i s se ei
o T s i s ses1 1 T es 1s
The11
T es 1s
s
b 0) ) . d
T
5e1T
Tu
b
1T T es 1o 5T1 N es1 b sei T
oojeT h T es o i y
1A T t 5T hh
sT eoy 1 e s b T
o1 u
DT1mei 1 ei
1ei b T es yo he Th T es 1ob
Nh 1 ei
1TT es yo he Th l yn h
s <) > y D s e > i 1 1hs1N
t 51h
ooj
b
e1 ei n b Ne cNhe T es 1o
eNn 1Tw 1Tei sh T h1 ei
oojeT h T es o i y
s i s ses1 1 ei
< <b
y
e1 oojeT h T es o i y
y ei
h ei e 1o1T e oeNn 1Th T
n 1 heT e
s 5 ei m y sh
1m T DNo es1 1 ei
1 T es 1o > soo T r
es1 h T h51 hs o
p
p
o <0w e sh m1Tei e1 1e ei e T es 1obT es yo he Th eT h T es o i y
o y ei
Tw 1T 1A TbT es 1osh T 5s oy 5Nn 5
TT
p
i
w
1t s sl s D T es yo he Th 1T T N s D
1o1T e o E i
i E
s1 esA n e 1ose > oojeT h
1 bei T sh
i shn
0) ) 2b 0) ) x8w i
<. &
E i ig
El
E
i EEg i
EE n p
ci
i E i
pl
s 5 ei m y sh o Toy N esA es1 1 ei
1n ei s D hsn so T 1 NTh e1 ei
>
l
ei sT 51o T n e 1ose h8 sh i n 5 T
l yn h
sh N RARA
RARB
RXRA
CYP26B1
e em1 o A ohw sThebei
sh
1Neei 5 ei h ei e
T> sDNT xM8w
gE vmA i E i
p E p i E i
i p cg E g i
u
E i p
i E
EE n pi p E
s n 5os s es1 w y T 5T h es D
AKR1B1
AKR1B10
AKR1C3
DHRS4
DGAT1
LRAT
s T oe T es1 h 1
1 eT TybN5T DNo es1 1
1No D e s
RDH10
RDH12
i
i
15y Nn
5 ei m y
1n 51N h
T h51 hs o
< <) b
< B l yn h
&
Discussion for all-­‐trans retinaldehyde). Even more, if some retinol is produced or absorbed by the tumor through the diet, it would rapidly become All-­‐trans retinaldehyde because of the upregulation of the responsible enzymes (RDH10 and RDH12). In summary, we demonstrate the inhibition of the biosynthesis of active metabolites of the retinoic acid pathway in colorectal cancer, which results in the impairment of canonical and non-­‐canonical anti-­‐tumorigenic effects of these signals. Moreover, at the genome scale, other genes are also co-­‐methylated with genes of the retinol acid pathway. A deeper study of these events revealed that processes such as embryo development, neuronal differentiation or cell migration are “co-­‐
silenced” with retinoic acid pathway genes. Interestingly these processes are known to be driven by retinoic acid. These results indicate that the production of bioactive metabolites from the retinoic acid pathway is prevented in colorectal cancer cells. 175 sh Nhhs1 B B
p
i c i
T s bm T 51Te ei
h1n
D h 1 ei
T es 1s
< < 5 sho >x26 1
i sDi 5T A o sh o Toy ei
11T s e
1n h E
E
pi
p
i y5 Tn ei yo es1 s on 1he oo 1o1T e o
n 1he i y5 Tn ei yo e
ei e ei sh i y5 Tn ei yo es1 sh 305
1
a e pe
s 5 ei m y s oo 1o1T e oeNn 1Th8 1n 5 T
oTumor
T Biol.
y 5T
h e s (2012) 33:297ñ
p
T s 1n hw
o1 Nh s ei
msei 1Tn o 1o1 w
i A
5 ei m y
oh1 hi 1m Toy A e NTs DeNn 1NTsD hshb
se sh 9N ooy 5T h e s ei
s T e
Nh se sh
T s 1n
301
he D h> e 1ehi 1m 8w
Table 3 Diagnostic stool markers
Gene
Name
Sensitivity/specificity
Method
Reference
CRC (75%, 15/20)/AD (68%, 17/25)/SP (90%, 3/30)
MSP
[49]
[45]
APC, ATM,
Misc.
hMLH1, SFRP2,
HLTF, MGMT,
GSTP1
CDKN2A, MGMT Misc.
or hMLH1
EN1
Homeobox protein engrailed-1
HP (55%, 16/29), SP (63%, 7/19)
MSP
CRC (27%, 8/30)/SP (97%, 1/30)
GATA4
HIC1
Transcription factor GATA-4
Hypermethylated in cancer 1 protein
CRC (59%, 44/75)/SP (88%, 9/75)
CRC (42%, 11/26)/AD (31%, 4/13)/SP (100%, 0/32)
Melting curve [54]
analysis
qMSP
[24]
MSP
[44]
ITGA4
NDRG4
Integrin alpha-4
N-Myc downstream-regulated gene 4
HP (69%, 9/13)/SP (78%, 6/28)
CRC (56%, 42/75)/SP (96%, 3/75)
qMSP
qMSP
[58]
[57]
OSMR
RASSF2
CRC (38%, 26/69)/AD (13%, 2/16)/SP (95%, 4/81)
CRC (45%, 38/84)/AD (13%, 7/56)/SP (95%, 6/113)
qMSP
COBRA
[59]
[52]
SFRP1
SFRP2
Oncostatin M receptor-β
Ras association domain-containing
protein 2
Secreted frizzled-related protein 1
Secreted frizzled-related protein 2
CRC (84%, 16/19)/AD (100%, 7/7)/SP (86%, 2/14)
CRC (83%, 19/23)/SP (77%, 6/26)
MSP
MethyLight
[53]
[43]
SFRP2
SFRP2
SFRP2
SFRP2
Secreted
Secreted
Secreted
Secreted
MSP
MSP
MSP
MSP
[48]
[50]
[51]
[37]
SFRP2
TFPI2
TFPI2
Secreted frizzled-related protein 2
Tissue factor pathway inhibitor 2
Tissue factor pathway inhibitor 2
CRC (94%, 49/52)/AD (52%, 11/21)/SP (96%, 1/24)
CRC (60%, 18/30)/AD (44%, 11/25) SP (97%, 1/31)
CRC (87%, 60/69)/AD (62%, 21/34)/SP (93%, 2/30)
CRC (84%, 142/169)/AD (46%, 29/63)/SP
(93%, 2/30)
CRC (63%, 53/84)/AD (32%, 18/56)/SP (92%, 9/113)
CRC (81%, 59/73)/AD (21%, 4/19)/SP (85%, 11/75)
CRC (68%, 41/60)/AD (55%, 11/20)/SP (100%, 0/30)
COBRA
qMSP
qMSP
[52]
[55]
[56]
TMEFF2
(HPP1/TPEF)
Vim
Vim
Transmembrane protein with EGF-like CRC (71%, 37/52)/AD (57%, 12/21)/SP (100%, 0/24) MSP
and two follistatin-like domains
Vimentin
CRC (46%, 43/94)/SP (89%, 20/178)
MSP
Vimentin
CRC (41%, 9/22)/AD (45%, 9/20)/SP (95%, 2/38)
MethylBEAMing
Vimentin
CRC (78%, 9/40)/SP (84%, 19/122)
MSP
Vim
frizzled-related
frizzled-related
frizzled-related
frizzled-related
protein
protein
protein
protein
2
2
2
2
[48]
[46]
[39]
[47]
Sensitivities in hyperplastic polyps (HP), adenomas (AD), and colorectal carcinomas (CRC) and the specificity (SP)
Bs A was pi
pi in almost
E Eplall primary
i
Eadenoma
E pi r marker pu
T1BTl
methylation
detected
[58].
To identify new methylation markers, three
samples, the detection rate of methylated TFPI2 in fecal
colon cancer cell lines were treated with the demethylation
h was
1 only
ei 21%.
h In ea second
b m study,
i A detection
A oN rates
e for
ei reagent,
e es1 5-aza-2′-deoxycytidine,
1
< < 5 sho
and differentially
DNA
methylated TFPI2 DNA in stool samples of patients with
expressed genes identified by microarray analysis. From a
n colorectal
ei yo es1 1 lower
js A(sensitivity,
hsA h n68%),
5o hb
i h list
he11o
h candidate
b 1genes,
1o1T
e al.
o chose
the
T gene Oncoscancers were
buthN
speciof 52
Kim et
tatin M receptor-β (OSM) for further analysis and were able
ficity of this assay was higher (100%) [56]. GATA4 repre5 sents
es another
eh tumor
s isuppressor
oei y gene
1 eT1o
hwtranscription
y
yo es1 s1n of
TutheThOSM
i A gene in 26 out of 69
whose
is n toeidetect
methylation
reduced in colorectal carcinomas versus normal colon tispatients with colorectal cancers (sensitivity, 38%) and in 4
5T151h
s ei o hey Th msei h1n 1 ei n 55
Ts D h A Ty 5T1n shs D N e1
out 81 healthy controls (specificity, 95%) [59].
sue. Analysis of GATA4 DNA methylation in fecal DNA
showed an overall detection rate of 59% and a specificity of
The number of publications focusing on the prognostic
88% [24]. Similar results were published by the same group
value of DNA hypermethylation in stool is limited: Tang et
al. published data that detection of methylated SFRP2 DNA
for the N-Myc downstream-regulated gene 4 (NDRG4), a
<.novel
3 tumor suppressor [57]. With a sensitivity of 56% and in stool samples is associated with a reduced survival of
a specificity of 96%, NDRG4 has similar qualities a bioaffected patients (P00.0219). However, detection of methmarker for asymptomatic patients with colorectal cancers as
ylated SFRP2 DNA in stool samples has no prognostic
sh Nhhs1 ei
i sDi
e es1 T e
0) <Bd
oo T u Th
b0) ) Md
s e
y1T
D11 h5 s s sey > l N T
ei
b0) ) Md
Dhu
b0) ) M8>
o <. 8w 1n
. 26 1
T s 1n Ch
e es1 msei M) 6 1
i ei
s o hesoo e i s9N Nh
n ei yo es1 h5 s s seyb
e1
s1n Tu T 1No
b0) <08>
h e1
i Nc b0) <Bd sn
heTs e T s e Tn h 1
e e n ei yo es1 b
5 s D 1 ei
> 1h i
b
T her
s1n Tu Th T
1oosDhb0) <0d 1 Dr
Tn 1 b0) ) Md
h5 s s seyw 1m A Tbei
b0) <) d
Nh s h1n
5 oh 1
1i 1Te heN y
h h ei
hn
sh5o y sn 51Te eA Ts es1 h s h hsesAsey h e 1 h n 5o h oyl
ei
e i s9N h Nh
o <. 8w
A
B
C
i
p
gE m1 A R
E pig
pl
ic
B B p p i e ic
R
B B
B B
E n i
1n 5 Ts D
1e ei
he 1 < <
5
> sDNT M) 8w i
heT1 D he 51s e 1
sho sh N n ei yo e
Th ei
e es1 T e
n ei 1 1o1Ds h> i o9Nshe
1n
Ep E
Ep E
1T
1n 5o n e ei
b0) ) ) d
D1T
R
T s 1n hsh
< <
s e h
o u 1
s <) ) 6 1 ei
< < n ei yo es1 p
p
e es1 b
< < sh ei
i sDi he h5 s s sey s ei
T hNoeh s s e ei e
55T1 i ei e
E
E
p
sh n 1T h hsesA ei 1ei T 5T151h
oei y 5 15o b mi e 1 hs o
u
E n i
E
n i
i E
E Epl
i
E
> sDNT M) 8b Ne T D T s D
> sDNT M) 8w 1m A Tb ei
i
i
< < msei 1ei T s1n Tu Thbseh
5 T 1Tn h T ooy m oo 51hsesA h,
E
op i
op p p i e ic
i E E
p
oh
he11oh T1n
h n 5o h oyl
e es1 n y
NTT e s D 1hes
h T s D
b0) ) B8w
<. .
Discussion AKR1B10 protein downregulation in colorectal cancer Tissue microarray experiments revealed loss of AKR1B10 protein (downregulation or complete absence) in the cancer tissue of 96.4% of the tumors analyzed (162 out of 168). Due to its almost universal occurrence, it is unlikely that the identification of this alteration could have prognostic utility. Nevertheless, it remains to be investigated whether this analysis could be applied to distinguish between a benign lesion and cancer in cases where there is a lesion in the epithelium but it is difficult to determine its malignancy by other methods. Further experiments are required until this application could be used, and different types of lesions including adenomas and bowel disease samples are required to have a complete profile of AKR1B10 in the different possible scenarios involving the intestinal tissue. 178 Discussion RETINOIC ACID PATHWAY GENES AS PROGNOSTIC MARKERS The discovery of new biomarkers is crucial for the future of medicine. Biomarkers could be related with the tumor biology and be essential for the election of the treatment. Biomarkers could also predict the outcome of the disease and help to improve the management of the disease. In this work we described some gene expression alterations of retinoic acid pathway genes that could also provide several confident prognosis markers. The alteration (downregulation or upregulation) of the normal gene expression of ALDH1A2, CYP26B1, DGAT1, DHRS4, LRAT, RARA, RDH10 and RXRA, in colorectal cancer is significantly related with worse overall survival. These results suggest that alterations in RA pathway genes expression could be used for prognostic prediction in colorectal cancers. Regarding AKR1B1 protein expression, we obtain some interesting results with the tissue microarrays analysis, the few cases that have no expression of AKR1B1 display a worse overall and disease free survival than the cases where it is not altered. This result is very motivating but we need to validate it with an independent cohort. In summary, we have shown that transcriptional disruption of the retinoic acid pathway (partly due to epigenetic alterations) is an important feature of most colorectal cancers, and its analysis might have important applications in the clinical management of patients. 179 Conclusions CONCLUSIONS 1. AKR1B1 CpG island is deeply and frequently hypermethylated in colorectal cancer, but with minimal or no effect on the expression of the associated gene. 2. Hypermethylation of AKR1B1 CpG island is strongly associated with downregulation of AKR1B10 and AKR1B15, which promoter region is 60Kb upstream of the CpG island. 3. AKR1B1 CpG island presents enhancer properties in regard to AKR1B10 and AKR1B15 genes, which explains the functional association. 4. AKR1B10 and AKR1B15 are tumor suppressor genes in colorectal cancer, especially in the presence of retinal, a form of vitamin A and precursor of retinoic acid. 5. Akr1b enzymes are also dowregulated in a murine colorectal cancer, suggesting that the impairment of its function in colorectal cancer is not an epiphenomenon. 6. The retinoic acid pathway is deregulated in most colorectal cancers by epigenetic mechanisms resulting in an overall decrease of the antitumor compounds of the retinoic acid pathway. 7. Detection of AKR1B1 CpG island hypermethylation (resulting in downregulation of AKR1B10) in stools DNA is a non-­‐invasive, sensitive and specific marker of colorectal adenomas and carcinomas, rendering it as a proper candidate for colorectal cancer screenings. 8. Low expression levels of AKR1B1, AKR1B10, ALDH1A2, CYP26B1, DGAT1, DHRS4, LRAT, RARA and RXRA; and high expression levels of RDH10 are associated with tumor aggressiveness, which might be useful as a prognostic factor. 183 Bibliography BIBLIOGRAPHY Ahlquist, D. A., Skoletsky, J. E., Boynton, K. A., Harrington, J. J., Mahoney, D. W., Pierceall, W. E., … Shuber, A. P. (2000). Colorectal cancer screening by detection of altered human DNA in stool: Feasibility of a multitarget assay panel. Gastroenterology,(119) 1219-­1227. Akhtar-­‐Zaidi, B., Cowper-­‐Sal-­‐lari, R., Corradin, O., Saiakhova, A., Bartels, C. F., Balasubramanian, D., … Scacheri, P. C. (2012). Epigenomic enhancer profiling defines a signature of colon cancer. Science, (336) 736–739. Akulenko, R., & Helms, V. (2013). DNA co-­‐methylation analysis suggests novel functional associations between gene pairs in breast cancer samples. Human Molecular Genetics, (22) 3016–3022. Albanes, D., Heinonen, O. P., Taylor, P. R., Virtamo, J., Edwards, B. K., Rautalahti, M., … Huttunen, J. K. (1996). Alpha-­‐Tocopherol and beta-­‐carotene supplements and lung cancer incidence in the alpha-­‐tocopherol, beta-­‐
carotene cancer prevention study: effects of base-­‐line characteristics and study compliance. Journal of the National Cancer Institute, (88) 1560–1570. Aldini, G., Dalle-­‐Donne, I., Facino, R. M., Milzani, A., & Carini, M. (2007). Intervention strategies to inhibit protein carbonylation by lipoxidation-­‐
derived reactive carbonyls. Medicinal Research Reviews, (27) 817–868. Ando, Y., Saka, H., Ando, M., Sawa, T., Muro, K., Ueoka, H., … Hasegawa, Y. (2000). Polymorphisms of UDP-­‐glucuronosyltransferase gene and irinotecan toxicity: a pharmacogenetic analysis. Cancer Research, (60) 6921–6926. Antequera, F., & Bird, a. (1993). Number of CpG islands and genes in human and mouse. Proceedings of the National Academy of Sciences of the United States of America, (90), 11995–11999. Aran, D., & Hellman, A. (2013). DNA methylation of transcriptional enhancers and cancer predisposition. Cell, (154) 11-­‐13. Aran, D., Sabato, S., & Hellman, A. (2013). DNA methylation of distal regulatory sites characterizes dysregulation of cancer genes. Genome Biology, (14), R21. Azuara, D., Rodriguez-­‐Moranta, F., de Oca, J., Soriano-­‐Izquierdo, A., Mora, J., Guardiola, J., … Capellá, G. (2010). Novel methylation panel for the early detection of colorectal tumors in stool DNA. Clinical Colorectal Cancer, (9), 168–76. Baba, S. P., Barski, O. A., Ahmed, Y., O’Toole, T. E., Conklin, D. J., Bhatnagar, A., & Srivastava, S. (2009). Reductive Metabolism of AGE Precursors: A Metabolic Route for Preventing AGE Accumulation in Cardiovascular Tissue. Diabetes, (58). Bailey, C. E., Eschrich, S., Kis, C., Levy, S., & Kay, M. (2012). Predicts Recurrence and Death in Patients With Colon Cancer, (138), 958–968. Bannister, A. J., & Kouzarides, T. (2011). Regulation of chromatin by histone modifications. Cell Research, (21) 381–95. Barker, N., van Es, J. H., Kuipers, J., Kujala, P., van den Born, M., Cozijnsen, M., … Clevers, H. (2007). Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature, (449) 1003–1007. Baron, J. M., Heise, R., Blaner, W. S., Neis, M., Joussen, S., Dreuw, A., … Jugert, F. K. (2005). Retinoic acid and its 4-­‐oxo metabolites are functionally active in 187 Bibliography human skin cells in vitro. The Journal of Investigative Dermatology, (125), 143–53. Barski, O. A., Tipparaju, S. M., & Bhatnagar, A. (2008). The Aldo-­‐Keto Reductase Superfamily and its Role in Drug Metabolism and Detoxification. Drug Metabolism Reviews, (40) 553–624. Bastide, N. M., Pierre, F. H. F., & Corpet, D. E. (2011). Heme iron from meat and risk of colorectal cancer: a meta-­‐analysis and a review of the mechanisms involved. Cancer Prevention Research. (4) 177–184. Bavinck, J. N., Tieben, L. M., Van der Woude, F. J., Tegzess, A. M., Hermans, J., ter Schegget, J., & Vermeer, B. J. (1995). Prevention of skin cancer and reduction of keratotic skin lesions during acitretin therapy in renal transplant recipients: a double-­‐blind, placebo-­‐controlled study. Journal of Clinical Oncology, (13) 1933–1938. Baxter, N., & Goldwasser, M. (2009). Association of Colonoscopy and Death From Colorectal Cancer. Annals of Internal medicine,(150) 1–9. Bell, O., Tiwari, V. K., Thomä, N. H., & Schübeler, D. (2011). Determinants and dynamics of genome accessibility. Nature Reviews. Genetics, (12) 554–564. Benelli, R., Monteghirfo, S., Venè, R., Tosetti, F., & Ferrari, N. (2010). The chemopreventive retinoid 4HPR impairs prostate cancer cell migration and invasion by interfering with FAK/AKT/GSK3beta pathway and beta-­‐catenin stability. Molecular Cancer, (9) 142. Berg, V., & Kinzler, K. W. (1997). Gatekeepers and caretakers. Nature, (386) 761-­
763. Bergman, Y., & Cedar, H. (2013). DNA methylation dynamics in health and disease. Nature Structural & Molecular Biology, (20) 274–281. Bettington, M., Walker, N., Clouston, A., Brown, I., Leggett, B., & Whitehall, V. (2013). The serrated pathway to colorectal carcinoma: current concepts and challenges. Histopathology, (62), 367–386. Bibikova, M., Lin, Z., Zhou, L., Chudin, E., Garcia, E. W., Wu, B., … Fan, J.-­‐B. (2006). High-­‐throughput DNA methylation profiling using universal bead arrays. Genome Research, (16), 383–393. Bird, A. (2002). DNA methylation patterns and epigenetic memory. Genes & Development, (16) 6–21. Bollag, W. (1972). Prophylaxis of chemically induced benign and malignant epithelial tumors by vitamin A acid (retinoic acid). European Journal of Cancer, (8) 689–693. Bosch, L. J. W., Mongera, S., Terhaar Sive Droste, J. S., Oort, F. a, van Turenhout, S. T., Penning, M. T., … Meijer, G. a. (2012). Analytical sensitivity and stability of DNA methylation testing in stool samples for colorectal cancer detection. Cellular Oncology (Dordrecht), (35) 309–315. Boutanaev, A. M., Kalmykova, A. I., Shevelyov, Y. Y., & Nurminsky, D. I. (2002). Large clusters of co-­‐expressed genes in the Drosophila genome. Nature, (420) 666–669. Boyle, P., & Langman, J. (2000). ABC of colorectal cancer Epidemiology. BMJ: British Medical Journal, (321) 805–808. Boyle, P., & Leon, M. E. (2002). Epidemiology of colorectal cancer. British Medical Bulletin, (64) 1–25. Bozic, I., Antal, T., Ohtsuki, H., Carter, H., Kim, D., Chen, S., … Nowak, M. a. (2010). Accumulation of driver and passenger mutations during tumor progression. 188 Bibliography Proceedings of the National Academy of Sciences of the United States of America, (107), 18545–50. Bray, G. A. (2002). The Underlying Basis for Obesity: Relationship to Cancer. The Journal of Nutrition, (132) 3451–3455. Brock, H. W., & van Lohuizen, M. (2001). The Polycomb group — no longer an exclusive club? Current Opinion in Genetics & Development, (11) 175–181. Brown, C. J., Hendrich, B. D., Rupert, J. L., Lafrenière, R. G., Xing, Y., Lawrence, J., & Willard, H. F. (1992). The human XIST gene: Analysis of a 17 kb inactive X-­‐
specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell (71) 527-­542. Brown, J. C., Winters-­‐Stone, K., Lee, A., & Schmitz, K. H. (2012). Cancer, Physical Activity, and Exercise. Comprehensive Physiology, (4) 2775-­27809. Bushue, N., & Wan, Y.-­‐J. Y. (2010). Retinoid pathway and cancer therapeutics. Advanced Drug Delivery Reviews, (62) 1285–1298. Campanero, M. R., Armstrong, M. I., & Flemington, E. K. (2000). CpG methylation as a mechanism for the regulation of E2F activity. Proceedings of the National Academy of Sciences , (97) 6481–6486. Campos, E. I., & Reinberg, D. (2009). Histones: annotating chromatin. Annual Review of Genetics, (43) 559–99. Cao, D. (1998). Identification and Characterization of a Novel Human Aldose Reductase-­‐like Gene. Journal of Biological Chemistry, (273) 11429–11435. Carmona, F. J., Azuara, D., Berenguer-­‐Llergo, A., Fernández, A. F., Biondo, S., de Oca, J., … Moreno, V. (2013). DNA methylation biomarkers for noninvasive diagnosis of colorectal cancer. Cancer Prevention Research, (6) 656–665. Carter, D., Chakalova, L., Osborne, C. S., Dai, Y., & Fraser, P. (2002). Long-­‐range chromatin regulatory interactions in vivo. Nature Genetics, (32) 623–626. Casey, P. J. (1994). Lipid modifications of G proteins. Current Opinion in Cell Biology, (6) 219–225. Cedar, H., & Bergman, Y. (2009). Linking DNA methylation and histone modification: patterns and paradigms. Nature Reviews. Genetics, (10) 295–
304. Chan, A. T., & Giovannucci, E. L. (2010). Primary prevention of colorectal cancer. Gastroenterology, (136) 2029–2043 Chen, J., Guo, L., Zhang, L., Wu, H., Yang, J., Liu, H., … Pei, D. (2013). Vitamin C modulates TET1 function during somatic cell reprogramming. Nature Genetics, (45) 1504–1509. Chia, W. K., Ali, R., & Toh, H. C. (2012). Aspirin as adjuvant therapy for colorectal cancer-­‐-­‐reinterpreting paradigms. Nature Reviews. Clinical Oncology, (9) 561–570. Chim, C.-­‐S., Wong, S.-­‐Y., Pang, A., Chu, P., Lau, J. S., Wong, K.-­‐F., & Kwong, Y.-­‐L. (2005). Aberrant promoter methylation of the retinoic acid receptor alpha gene in acute promyelocytic leukemia. Leukemia, (19) 2241–2246. Chomczynski, P., & Sacchi, N. (2006). The single-­‐step method of RNA isolation by acid guanidinium thiocyanate-­‐phenol-­‐chloroform extraction: twenty-­‐
something years on. Nature Protocols, (1) 581–585. Chou, A. P., Chowdhury, R., Li, S., Chen, W., Kim, A. J., Piccioni, D. E., … Lai, A. (2012). Identification of retinol binding protein 1 promoter hypermethylation in isocitrate dehydrogenase 1 and 2 mutant gliomas. Journal of the National Cancer Institute, (104) 1458–1469. 189 Bibliography Chuang, J. C., Warner, S. L., Vollmer, D., Vankayalapati, H., Redkar, S., Bearss, D. J., … Jones, P. a. (2010). S110, a 5-­‐Aza-­‐2’-­‐deoxycytidine-­‐containing dinucleotide, is an effective DNA methylation inhibitor in vivo and can reduce tumor growth. Molecular Cancer Therapeutics, (9) 1443–1450. Chung, Y. T., Matkowskyj, K. a, Li, H., Bai, H., Zhang, W., Tsao, M.-­‐S., … Yang, G.-­‐Y. (2012). Overexpression and oncogenic function of aldo-­‐keto reductase family 1B10 (AKR1B10) in pancreatic carcinoma. Modern Pathology  : An Official Journal of the United States and Canadian Academy of Pathology, (25) 758–766. Comb, M., & Goodman, H. M. (1990). CpG methylation inhibits proenkephalin gene expression and binding of the transcription factor AP-­‐2. Nucleic Acids Research , (13), 3975–3982. Cremer, T., & Cremer, M. (2010). Chromosome territories. Cold Spring Harbor Perspectives in Biology, (3). Creyghton, M. P., Cheng, A. W., Welstead, G. G., Kooistra, T., Carey, B. W., Steine, E. J., … Jaenisch, R. (2010). Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proceedings of the National Academy of Sciences of the United States of America, (107) 21931–21936. Cunningham, D., Atkin, W., Lenz, H.-­‐J., Lynch, H. T., Minsky, B., Nordlinger, B., & Starling, N. (2010). Colorectal cancer. The Lancet, (375) 1030–1047. Dawson, M. a, & Kouzarides, T. (2012). Cancer epigenetics: from mechanism to therapy. Cell, (150) 12–27. De Carvalho, D. D., Sharma, S., You, J. S., Su, S.-­‐F., Taberlay, P. C., Kelly, T. K., … Jones, P. A. (2012). DNA Methylation Screening Identifies Driver Epigenetic Events of Cancer Cell Survival. Cancer Cell, (21) 655–667. Dekker, J. (2006). The three “ C ” s of chromosome conformation capture  : controls , controls , controls, (3) 17–21. Dekker, J., Marti-­‐Renom, M. a, & Mirny, L. a. (2013). Exploring the three-­‐
dimensional organization of genomes: interpreting chromatin interaction data. Nature Reviews. Genetics, (14) 390–403. Deng, G., Chen, A., Pong, E., & Kim, Y. S. (2001). Methylation in hMLH1 promoter interferes with its binding to transcription factor CBF and inhibits gene expression. Oncogene, (48) 7120–7127. Di Ruscio, A., Ebralidze, A. K., Benoukraf, T., Amabile, G., Goff, L. a, Terragni, J., … Tenen, D. G. (2013). DNMT1-­‐interacting RNAs block gene-­‐specific DNA methylation. Nature, (503) 371-­‐376. Dillard, A., & Lane, M. (2007). Retinol Decreases β-­‐Catenin Protein Levels in Retinoic Acid-­‐Resistant Colon Cancer Cell Lines. Molecular Carcinogenesis, (329), 315–329. Dixon, J. R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., … Ren, B. (2012). Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature, (485) 1–5. Djebali, S., Davis, C. a, Merkel, A., Dobin, A., Lassmann, T., Mortazavi, A., … Gingeras, T. R. (2012). Landscape of transcription in human cells. Nature, (489) 101–108. Down, T. a, Rakyan, V. K., Turner, D. J., Flicek, P., Li, H., Kulesha, E., … Beck, S. (2008). A Bayesian deconvolution strategy for immunoprecipitation-­‐based DNA methylome analysis. Nature Biotechnology, (26) 779–785. 190 Bibliography Eden, A., Gaudet, F., Waghmare, A., & Jaenisch, R. (2003). Chromosomal instability and tumors promoted by DNA hypomethylation. Science, (300) 405. Ehrlich, M., Woods, C. B., Yu, M. C., Dubeau, L., Yang, F., Campan, M., … Laird, P. W. (2006). Quantitative analysis of associations between DNA hypermethylation, hypomethylation, and DNMT RNA levels in ovarian tumors. Oncogene, (18) 2636–2645. Ernst, J., Kheradpour, P., Mikkelsen, T. S., Shoresh, N., Ward, L. D., Epstein, C. B., … Bernstein, B. E. (2011). Mapping and analysis of chromatin state dynamics in nine human cell types. Nature, (473) 43–49. Esteller, M. (2011). Non-­‐coding RNAs in human disease. Nature Reviews. Genetics, (12), 861–74. Fairley, T. L., Cardinez, C. J., Martin, J., Alley, L., Friedman, C., Edwards, B., & Jamison, P. (2006). Colorectal cancer in U.S. adults younger than 50 years of age, 1998–2001. Cancer, (107) 1153–1161. Faivre, J., Dancourt, V., Lejeune, C., Tazi, M. a., Lamour, J., Gerard, D., … Bonithon-­‐
Kopp, C. (2004). Reduction in colorectal cancer mortality by fecal occult blood screening in a French controlled study1. Gastroenterology, (126) 1674–1680. Fearon, E. R., & Vogelstein, B. (1990). A genetic model for colorectal tumorigenesis. Cell, (61) 759–767. Fedirko, V., Tramacere, I., Bagnardi, V., Rota, M., Scotti, L., Islami, F., … Jenab, M. (2011). Alcohol drinking and colorectal cancer risk: an overall and dose-­‐
response meta-­‐analysis of published studies. Annals of Oncology  : Official Journal of the European Society for Medical Oncology, (22) 1958–1972. Feinberg, A. P., & Vogelstein, B. (1983). Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature, (301) 89–92. Ferrari, K. J., Scelfo, A., Jammula, S., Cuomo, A., Barozzi, I., Stützer, A., … Pasini, D. (2014). Polycomb-­‐dependent H3K27me1 and H3K27me2 regulate active transcription and enhancer fidelity. Molecular Cell (53) 49–62. Fidge, N. H., Shiratori, T., Ganguly, J., & Goodman, D. S. (1968). Pathways of absorption of retinal and retinoic acid in the rat. Journal of Lipid Research, (9) 103–109. Fijten, G. H., Starmans, R., Muris, J. W. M., Schouten, H. J. A., Blijham, G. H., & Knottnerus, J. A. (1995). Predictive value of signs and symptoms for colorectal cancer in patients with rectal bleeding in general practice. Family Practice, (12) 279–286. Fraga, M. F., Ballestar, E., Villar-­‐Garea, A., Boix-­‐Chornet, M., Espada, J., Schotta, G., … Esteller, M. (2005). Loss of acetylation at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human cancer. Nature Genetics, (37) 391–400. Fraser, P., & Bickmore, W. (2007). Nuclear organization of the genome and the potential for gene regulation. Nature, (447) 413–417. Friedl, W., Caspari, R., Sengteller, M., Uhlhaas, S., Lamberti, C., Jungck, M., … Propping, P. (2001). Can APC mutation analysis contribute to therapeutic decisions in familial adenomatous polyposis. Experience from 680 FAP families. Gut , (48) 515–521. Frigola, J., Song, J., Stirzaker, C., Hinshelwood, R. a, Peinado, M. a, & Clark, S. J. (2006). Epigenetic remodeling in colorectal cancer results in coordinate 191 Bibliography gene suppression across an entire chromosome band. Nature Genetics, (38) 540–539. Frommer, M., McDonald, L. E., Millar, D. S., Collis, C. M., Watt, F., Grigg, G. W., … Paul, C. L. (1992). A genomic sequencing protocol that yields a positive display of 5-­‐methylcytosine residues in individual DNA strands. Proceedings of the National Academy of Sciences of the United States of America, (89) 1827–1831. Fu, Z., Shrubsole, M. J., Li, G., Smalley, W. E., Hein, D. W., Cai, Q., … Zheng, W. (2013). Interaction of cigarette smoking and carcinogen-­‐metabolizing polymorphisms in the risk of colorectal polyps. Carcinogenesis, (34) 779–
786. Fujii, Y., Watanabe, K., Hayashi, H., Urade, Y., Kuramitsu, S., Kagamiyama, H., & Hayaishi, O. (1990). Purification and characterization of rho-­‐crystallin from Japanese common bullfrog lens. Journal of Biological Chemistry, (265) 9914–
9923. Fukumoto, S., Yamauchi, N., Moriguchi, H., Hironaka, M., Ishikawa, Y., Kodama, T., & Nishimura, M. (2005). Overexpression of the Aldo-­‐Keto Reductase Family Protein AKR1B10 Is Highly Correlated with Smokers ’ Non − Small Cell Lung Carcinomas Overexpression of the Aldo-­‐Keto Reductase Family Protein AKR1B10 Is Highly Correlated with Smokers ’ Non – Small Cell Lung. Clinical Cancer Research, (11) 1776–1785. Gabbay, K. (2004). Aldose reductase inhibition in the treatment of diabetic neuropathy: Where are we in 2004? Current Diabetes Reports, (6) 405–408. Galiatsatos, P., & Foulkes, W. (2006). Familial adenomatous polyposis. The American Journal of Gastroenterology, (18) 7–30. Gama-­‐Sosa, M., & Slagel, V. (1983). The 5-­‐methylcytosine content of DNA from human tumors. Nucleic Acids Research, (19) 6883–6894. Gao, J., Aksoy, B. a., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S. O., … Schultz, N. (2013). Integrative Analysis of Complex Cancer Genomics and Clinical Profiles Using the cBioPortal. Science Signaling, (269) pl1. Garinis, G. a, Patrinos, G. P., Spanakis, N. E., & Menounos, P. G. (2002). DNA hypermethylation: when tumour suppressor genes go silent. Human Genetics, (111) 115–127. Gaspar, C., Cardoso, J., Franken, P., Molenaar, L., Morreau, H., Möslein, G., … Fodde, R. (2008). Cross-­‐species comparison of human and mouse intestinal polyps reveals conserved mechanisms in adenomatous polyposis coli (APC)-­‐
driven tumorigenesis. The American Journal of Pathology, (172) 1363–1380. Gaudet, F., Hodgson, J. G., Eden, A., Jackson-­‐Grusby, L., Dausman, J., Gray, J. W., … Jaenisch, R. (2003). Induction of tumors in mice by genomic hypomethylation. Science, (300) 489–492. Gentleman, R. C., Carey, V. J., Bates, D. M., Bolstad, B., Dettling, M., Dudoit, S., … Zhang, J. (2004). Bioconductor: open software development for computational biology and bioinformatics. Genome Biology, (5) R80. Gervin, K., Hammero, M., Akselsen, H. E., Moe, R., Nygard, H., Brandt, I., … Lyle, R. (2011). Extensive variation and low heritability of DNA methylation identified in a twin study. Genome Research, (11) 1813–1821. Gibbs, J., Kohl, N., Koblan, K., Omer, C., Sepp-­‐Lorenzino, L., Rosen, N., … Oliff, A. (1996). Farnesyltransferase inhibitors and anti-­‐Ras therapy. Breast Cancer Research and Treatment, (38) 75–83. 192 Bibliography Giovannucci, E. (2001). An Updated Review of the Epidemiological Evidence that Cigarette Smoking Increases Risk of Colorectal Cancer. Cancer Epidemiology, Biomarkers & Prevention, (7) 725–731. Globisch, D., Münzel, M., Müller, M., Michalakis, S., Wagner, M., Koch, S., … Carell, T. (2010). Tissue Distribution of 5-­‐Hydroxymethylcytosine and Search for Active Demethylation Intermediates. PLoS ONE, (12) e15367. Grady, W. M., Parkin, R. K., Mitchell, P. S., Lee, J. H., Kim, Y.-­‐H., Tsuchiya, K. D., … Tewari, M. (2008). Epigenetic silencing of the intronic microRNA hsa-­‐miR-­‐
342 and its host gene EVL in colorectal cancer. Oncogene, (27) 3880–3888. Greer, E. L., & Shi, Y. (2012). Histone methylation: a dynamic mark in health, disease and inheritance. Nature Reviews Genetics, (13) 343-­‐357. Grölund, M.-­‐M., Lehtonen, O.-­‐P., Eerola, E., & Kero, P. (1999). Fecal microflora in healthy infants born by different methods of delivery: permanent changes in intestinal flora after cesarean delivery. Journal of Pediatric Gastroenterology and Nutrition, (28) 19-­‐25. Grunau, C., Clark, S. J., & Rosenthal, A. (2001). Bisulfite genomic sequencing: systematic investigation of critical experimental parameters. Nucleic Acids Research, (29) E65–5. Guarner, F., & Malagelada, J.-­‐R. (2003). Gut flora in health and disease. Lancet, (361) 512–519. Guo, J. U., Su, Y., Zhong, C., Ming, G., & Song, H. (2011, April 29). Hydroxylation of 5-­‐methylcytosine by TET1 promotes active DNA demethylation in the adult brain. Cell, (145) 423-­‐434. Haggar, F., & Boushey, R. (2009). Colorectal cancer epidemiology: incidence, mortality, survival, and risk factors. Clinics in Colon and Rectal Surgery, (212) 191–197. Halpern, A. C., Schuchter, L. M., Elder, D. E., Guerry, D., Elenitsas, R., Trock, B., & Matozzo, I. (1994). Effects of topical tretinoin on dysplastic nevi. Journal of Clinical Oncology , (5) 1028–1035. Hanahan, D., & Weinberg, R. a. (2011). Hallmarks of cancer: the next generation. Cell, (144) 646–674. Hanahan, D., & Weinberg, R. A. (2000). The Hallmarks of Cancer. Cell, (100) 57–
70. Hardin, J., & Mydlarski, P. R. (2010). Systemic retinoids: chemoprevention of skin cancer in transplant recipients. Skin Therapy Letter, (15) 1–4. Harisiadis, L., Miller, R. C., Hall, E. J., & Borek, C. (1978). A vitamin A analogue inhibits radiation-­‐induced oncogenic transformation. Nature, (274) 486–
487. Harrow, J., Nagy, A., Reymond, A., Alioto, T., Patthy, L., Antonarakis, S. E., & Guigó, R. (2009). Identifying protein-­‐coding genes in genomic sequences. Genome Biology, (10) 201. Hayashi, K., Yokozaki, H., Goodison, S., Oue, N., Suzuki, T., Lotan, R., … Tahara, E. (2001). Inactivation of retinoic acid receptor beta by promoter CpG hypermethylation in gastric cancer. Differentiation; Research in Biological Diversity, (68) 13–21. He, J., & Efron, J. E. (2011). Screening for Colorectal Cancer. Advances in Surgery. Mosby Year Book. 193 Bibliography He, Y.-­‐F., Li, B.-­‐Z., Li, Z., Liu, P., Wang, Y., Tang, Q., … Xu, G.-­‐L. (2011). Tet-­‐mediated formation of 5-­‐carboxylcytosine and its excision by TDG in mammalian DNA. Science , (333) 1303–1307. Heibein, A. D., Guo, B., Sprowl, J. a, Maclean, D. a, & Parissenti, A. M. (2012). Role of aldo-­‐keto reductases and other doxorubicin pharmacokinetic genes in doxorubicin resistance, DNA binding, and subcellular localization. BMC Cancer, (12) 381. Hellebrekers, D. M. E. I., Lentjes, M. H. F. M., van den Bosch, S. M., Melotte, V., Wouters, K. a D., Daenen, K. L. J., … van Engeland, M. (2009). GATA4 and GATA5 are potential tumor suppressors and biomarkers in colorectal cancer. Clinical Cancer Research  : An Official Journal of the American Association for Cancer Research, (15) 3990–3997. Henikoff, S., & Shilatifard, A. (2011). Histone modification: cause or cog? Trends in Genetics, (27) 389–396. Herbst, A., & Kolligs, F. T. (2012). Detection of DNA hypermethylation in remote media of patients with colorectal cancer: new biomarkers for colorectal carcinoma. Tumour Biology  : The Journal of the International Society for Oncodevelopmental Biology and Medicine, (33) 297–305. Heringlake, S., Hofdmann, M., Fiebeler, A., Manns, M. P., Schmiegel, W., & Tannapfel, A. (2010). Identification and expression analysis of the aldo–
ketoreductase1-­‐B10 gene in primary malignant liver tumours. Journal of Hepatology, (2) 220-­‐227. Hiraoka, S., Kato, J., Horii, J., Saito, S., Harada, K., Fujita, H., … Yamamoto, K. (2010, January 1). Methylation status of normal background mucosa is correlated with occurrence and development of neoplasia in the distal colon. Human Pathology, (1) 38-­‐47. Hol, L., van Leerdam, M. E., van Ballegooijen, M., van Vuuren, A. J., van Dekken, H., Reijerink, J. C. I. Y., … Kuipers, E. J. (2010). Screening for colorectal cancer: randomised trial comparing guaiac-­‐based and immunochemical faecal occult blood testing and flexible sigmoidoscopy. Gut, (59) 62–68. Holliday, R., & Pugh, J. E. (1975). DNA modification mechanisms and gene activity during development. Science, (187) 226–232. Hong, L., & Ahuja, N. (2013). DNA methylation biomarkers of stool and blood for early detection of colon cancer. Genetic Testing and Molecular Biomarkers, (17) 401–406. Hong, W. K., Endicott, J., Itri, L. M., Doos, W., Batsakis, J. G., Bell, R., … Strong, S. (1986). 13-­‐cis-­‐Retinoic Acid in the Treatment of Oral Leukoplakia. New England Journal of Medicine, (24) 1501–1505. Hong, W. K. I., Lippman, S. M., Itri, L. M., Karp, D. D., Lee, J. S., Byers, R. M., … Goepfert, H. (1990). Prevention of Second Primary Tumors with Isotretinoin in Squamous-­‐Cell Carcinoma of the Head and Neck. New England Journal of Medicine, (323) 795–801. Hotchkiss, R. (1948). The quantitative separation of purines, pyrimidiines, and nucleosides by paper chromatography. Journal of Biological Chemistry, (175) 315-­‐332. Hu, H., Krasinskas, A., & Willis, J. (2011). Perspectives on current tumor-­‐node-­‐
metastasis (TNM) staging of cancers of the colon and rectum. Seminars in Oncology, (38) 500–510. 194 Bibliography Huang, D. W., Sherman, B. T., & Lempicki, R. a. (2009). Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Research, (37) 1–13. Huang, D. W., Sherman, B. T., & Lempicki, R. A. (2008). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols, (4) 44–57. Huxley, R. R., Woodward, M., & Clifton, P. (2012). The Epidemiologic Evidence and Potential Biological Mechanisms for a Protective Effect of Dietary Fiber on the Risk of Colorectal Cancer. Current Nutrition Reports, (2) 63–70. Hyndman, D. J., & Flynn, T. G. (1998). Sequence and expression levels in human tissues of a new member of the aldo-­‐keto reductase family. Biochimica et Biophysica Acta (BBA) -­ Gene Structure and Expression (2–3), 198–202. Iida, N., Dzutsev, A., Stewart, C. a., Smith, L., Bouladoux, N., Weingarten, R. a., … Goldszmid, R. S. (2013). Commensal Bacteria Control Cancer Response to Therapy by Modulating the Tumor Microenvironment. Science, (342) 967–
970. Imperiale, T. F., Ransohoff, D. F., Itzkowitz, S. H., Turnbull, B. A., & Ross, M. E. (2004). Fecal DNA versus fecal occult blood for colorectal-­‐cancer screening in an average-­‐risk population. New England Journal of Medicine, (26) 2704–
2714. Innocenti, F., Undevia, S. D., Ramirez, J., Mani, S., Schilsky, R. L., Vogelzang, N. J., … Ratain, M. J. (2004). A phase I trial of pharmacologic modulation of irinotecan with cyclosporine and phenobarbital. Clin Pharmacol Ther, (76) 490–502. Irizarry, R. A., Ladd-­‐acosta, C., Carvalho, B., Wu, H., Brandenburg, S. A., Jeddeloh, J. A., … Feinberg, A. P. (2008). Comprehensive high-­‐throughput arrays for relative methylation (CHAR). Genome research, (18) 780–790. Issa, J.-­‐P. J., & Ahuja, N. (2000). Aging , methylation and cancer. Histology and Histopathology, (15) 835–842. Issa, J.-­‐P. J., Vertino, P. M., Wu, J., Sazawal, S., Celano, P., Nelkin, B. D., … Baylin, S. B. (1993). Increased Cytosine DNA-­‐Methyltransferase Activity During Colon Cancer Progression. Journal of the National Cancer Institute, (15) 1235–
1240. Ito, S., Shen, L., Dai, Q., Wu, S. C., Collins, L. B., Swenberg, J. A., … Zhang, Y. (2011). Tet Proteins Can Convert 5-­‐Methylcytosine to 5-­‐Formylcytosine and 5-­‐
Carboxylcytosine. Science , (333) 1300–1303. Iyer, L. M., Tahiliani, M., Rao, A., & Aravind, L. (2009). Prediction of novel families of enzymes involved in oxidative and other complex modifications of bases in nucleic acids. Cell Cycle, (11) 1698–1710. Jin, S.-­‐G., Cai Guo,& Pfeifer G.P. (2008). GADD45A does not promote DNA demethylation. PLoS Genetics, (3) p1. Jin, Y., & Penning, T. M. (2007). Aldo-­‐keto reductases and bioactivation/detoxication. Annual Review of Pharmacology and Toxicology, (47) 263–292. Johnson, I. T., & Lund, E. K. (2007). Review article: nutrition, obesity and colorectal cancer. Alimentary Pharmacology & Therapeutics, (26) 161–181. Jones, P. A. (1999). The DNA methylation paradox. Trends in Genetics, (15) 34–37. Jorissen, R. N., Gibbs, P., Christie, M., Prakash, S., Lipton, L., Desai, J., … Sieber, O. M. (2009). Metastasis-­‐Associated Gene Expression Changes Predict Poor 195 Bibliography Outcomes in Patients with Dukes Stage B and C Colorectal Cancer. Clinical Cancer Research  : An Official Journal of the American Association for Cancer Research, (24) 7642–7651. Kaaij, L. T., van de Wetering, M., Fang, F., Decato, B., Molaro, A., van de Werken, H. J., … Ketting, R. F. (2013). DNA methylation dynamics during intestinal stem cell differentiation reveals enhancers driving gene expression in the villus. Genome Biology, (14) R50. Kamakaka, R. T., & Biggins, S. (2005). Histone variants: deviants? Genes & Development, (19) 295–310. Kang, M.-­‐W., Lee, E.-­‐S., Yoon, S. Y., Jo, J., Lee, J., Kim, H. K., … Kim, H. (2011). AKR1B10 is associated with smoking and smoking-­‐related non-­‐small-­‐cell lung cancer. The Journal of International Medical Research, (39) 78–85. Kelly, V., Sherratt, P., Crouch, D., & Hayes, J. (2002). Novel homodimeric and heterodimeric rat γ-­‐hydroxybutyrate synthases that associate with the Golgi apparatus define a distinct subclass of aldo-­‐keto reductase 7. Biochem. J., (366) 847-­‐861. Kennison, J. a. (1995). The Polycomb and trithorax group proteins of Drosophila: trans-­‐regulators of homeotic gene function. Annual Review of Genetics, (29) 289–303. Kershaw, E. E., & Flier, J. S. (2004). Adipose Tissue as an Endocrine Organ. The Journal of Clinical Endocrinology & Metabolism, (89) 2548–2556. Khare, S., & Verma, M. (2012). Cancer Epigenetics, (863) 177–185. Kim, H., Lapointe, J., Kaygusuz, G., Ong, D. E., Li, C., van de Rijn, M., … Pollack, J. R. (2005). The retinoic acid synthesis gene ALDH1a2 is a candidate tumor suppressor in prostate cancer. Cancer Research, (65) 8118–8124. Kim, M. S., Louwagie, J., Carvalho, B., Terhaar Sive Droste, J. S., Park, H. L., Chae, Y. K., … Sidransky, D. (2009). Promoter DNA methylation of oncostatin m receptor-­‐beta as a novel diagnostic and therapeutic marker in colon cancer. PloS One, (8) e6555. Kim, T.-­‐K., Hemberg, M., Gray, J. M., Costa, A. M., Bear, D. M., Wu, J., … Greenberg, M. E. (2010). Widespread transcription at neuronal activity-­‐regulated enhancers. Nature, (465) 182–187. Kim, Y.-­‐H., Petko, Z., Dzieciatkowski, S., Lin, L., Ghiassi, M., Stain, S., … Grady, W. M. (2006). CpG island methylation of genes accumulates during the adenoma progression step of the multistep pathogenesis of colorectal cancer. Genes, Chromosomes and Cancer, (45) 781–789. Kinzler, K. W., & Vogelstein, B. (1996). Lessons from hereditary colorectal cancer. Cell, (87) 159–170. Ko, C.-­‐Y., Hsu, H.-­‐C., Shen, M.-­‐R., Chang, W.-­‐C., & Wang, J.-­‐M. (2008). Epigenetic silencing of CCAAT/enhancer-­‐binding protein delta activity by YY1/polycomb group/DNA methyltransferase complex. The Journal of Biological Chemistry, (283) 30919–30932. Ko, M., Huang, Y., Jankowska, A. M., Pape, U. J., Tahiliani, M., Bandukwala, H. S., … Rao, A. (2010). Impaired hydroxylation of 5-­‐methylcytosine in myeloid cancers with mutant TET2. Nature, (468) 839–843. Kohli, R. M., & Zhang, Y. (2013). TET enzymes, TDG and the dynamics of DNA demethylation. Nature, (502) 472–479. Kota, S. K., & Feil, R. (2010). Epigenetic transitions in germ cell development and meiosis. Developmental Cell, (19) 675–86. 196 Bibliography Kouzarides, T. (2007). Chromatin modifications and their function. Cell (128) 693–705. Kozma, E., Brown, E., Ellis, E. M., & Lapthorn, A. J. (2002). The Crystal Structure of Rat Liver AKR7A1: a dimeric member of the aldo-­‐keto reductase superfamily. Journal of Biological Chemistry , (18) 16285–16293. Kriaucionis, S., & Heintz, N. (2009). The Nuclear DNA Base 5-­‐
Hydroxymethylcytosine Is Present in Purkinje Neurons and the Brain. Science, (324) 929–930. Krol, J., Loedige, I., & Filipowicz, W. (2010). The widespread regulation of microRNA biogenesis, function and decay. Nature Reviews. Genetics, (11) 597–610. Kucherlapati, R., & Wheeler, D. A. (2012). Comprehensive molecular characterization of human colon and rectal cancer. Nature (487) 330–337. Kurie, J. M., Lee, J. S., Khuri, F. R., Mao, L., Morice, R. C., Lee, J. J., … Hong, W. K. (2000). N-­‐(4-­‐Hydroxyphenyl) Retinamide in the Chemoprevention of Squamous Metaplasia and Dysplasia of the Bronchial Epithelium, (8) 2973–
2979. Laffin, B., & Petrash, J. M. (2012). Expression of the Aldo-­‐Ketoreductases AKR1B1 and AKR1B10 in Human Cancers. Frontiers in Pharmacology, (3) 104. Laird, P. W. (2010). Principles and challenges of genomewide DNA methylation analysis. Nature Reviews. Genetics, (11) 191–203. Laird, P. W., Jackson-­‐Grusby, L., Fazeli, A., Dickinson, S. L., Edward Jung, W., Li, E., … Jaenisch, R. (1995). Suppression of intestinal neoplasia by DNA hypomethylation. Cell, (81) 197-­‐205. Lao, V. V., & Grady, W. M. (2011). Epigenetics and colorectal cancer. Nature Reviews. Gastroenterology & Hepatology, (12) 686–700. Lasnitzki, I. (1955). The influence of A hypervitaminosis on the effect of 20-­‐
methylcholanthrene on mouse prostate glands grown in vitro. British Journal of Cancer, (3) 434–41. Law, J. a, & Jacobsen, S. E. (2010). Establishing, maintaining and modifying DNA methylation patterns in plants and animals. Nature Reviews. Genetics, (3) 204–220. Lee, J. T. (2012). Epigenetic Regulation by Long Noncoding RNAs. Science, (338) 1435–1439. Li, E., Bestor, T. H., & Jaenisch, R. (1992). Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell, (69) 915-­‐926 Lim, D. H. K., & Maher, E. R. (2010). Genomic Imprinting Syndromes and Cancer. Advances in Genetics, (70) 145–175. Limketkai, B. N., Lam-­‐Himlin, D., Arnold, M. a, & Arnold, C. a. (2013). The cutting edge of serrated polyps: a practical guide to approaching and managing serrated colon polyps. Gastrointestinal Endoscopy, (77) 360–375. Lister, R., & Ecker, J. R. (2009). Finding the fifth base: genome-­‐wide sequencing of cytosine methylation. Genome Research, (19) 959–66. Lister, R., Mukamel, E. a, Nery, J. R., Urich, M., Puddifoot, C. a, Johnson, N. D., … Ecker, J. R. (2013). Global epigenomic reconfiguration during mammalian brain development. Science, (341). Loeffler-­‐Ragg, J., Mueller, D., Gamerith, G., Auer, T., Skvortsov, S., Sarg, B., … Zwierzina, H. (2009). Proteomic identification of aldo-­‐keto reductase AKR1B10 induction after treatment of colorectal cancer cells with the 197 Bibliography proteasome inhibitor bortezomib. Molecular Cancer Therapeutics, (7) 1995–
2006. López-­‐Abente, G., Ardanaz, E., Torrella-­‐Ramos, a, Mateos, a, Delgado-­‐Sanz, C., & Chirlaque, M. D. (2010). Changes in colorectal cancer incidence and mortality trends in Spain. Annals of Oncology  : Official Journal of the European Society for Medical Oncology, (21) iii76–82. Luger, K. (2006). Dynamic nucleosomes. Chromosome Research, (14) 5–16. Lynch, H. T., Lynch, P. M., Lanspa, S. J., Snyder, C. L., Lynch, J. F., & Boland, C. R. (2009). Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clinical Genetics, (76) 1–18. Ma, J., & Cao, D. (2011). Human aldo-­‐keto reductases: structure, substrate specificity and roles in tumorigenesis. BioMolecular Concepts, (2) 115–126. Ma, J., Luo, D.-­‐X., Huang, C., Shen, Y., Bu, Y., Markwell, S., … Cao, D. (2012). AKR1B10 overexpression in breast cancer: association with tumor size, lymph node metastasis and patient survival and its potential as a novel serum marker. International Journal of Cancer. Journal International Du Cancer, (6). Majumdar, S. R., Fletcher, R. H., & Evans, A. T. (1999). How does colorectal cancer present[quest] symptoms, duration, and clues to location. Am J Gastroenterol, (10) 3039–3045. Mali, P., Yang, L., Esvelt, K., Aach, J., & Guell, M. (2013). RNA-­‐guided human genome engineering via Cas9. Science, (339) 823–826. Mann, I. K., Chatterjee, R., Zhao, J., He, X., Weirauch, M. T., Hughes, T. R., & Vinson, C. (2013). CG methylated microarrays identify a novel methylated sequence bound by the CEBPB|ATF4 heterodimer that is active in vivo. Genome, (23) 988-­‐997. Marcuello, E., Altes, A., Menoyo, A., del Rio, E., Gomez-­‐Pardo, M., & Baiget, M. (2004). UGT1A1 gene variations and irinotecan treatment in patients with metastatic colorectal cancer. Br J Cancer, (91) 678–682. Marx, V. (2012). Epigenetics: Reading the second genomic code. Nature, (491) 143–147. Mayor, R., Casadomé, L., Azuara, D., Moreno, V., Clark, S. J., Capellà, G. & Peinado, M. A. (2009). Long-­‐range epigenetic silencing at 2q14.2 affects most human colorectal cancers and may have application as a non-­‐invasive biomarker of disease. British Journal of Cancer, (100) 1534–1539. Maze, I., Noh, K.-­‐M., Soshnev, A. a., & Allis, C. D. (2014). Every amino acid matters: essential contributions of histone variants to mammalian development and disease. Nature Reviews Genetics, (15) 259-­‐271. McVicker, G., Geijn, B. van de, & Degner, J. (2013). Identification of genetic variants that affect histone modifications in human cells. Science, (342) 747–
749. Meinsma, R., Fernandez-­‐Salguero, P., Van Kuilenburg, a B., Van Gennip, a H., & Gonzalez, F. J. (1995). Human polymorphism in drug metabolism: mutation in the dihydropyrimidine dehydrogenase gene results in exon skipping and thymine uracilurea. DNA and Cell Biology, (14) 1–6. Ménard, S., Camerini, T., Mariani, L., Tomasic, G., Pilotti, S., Costa, A., … Veronesi, U. (2001). Re: Randomized Trial of Fenretinide to Prevent Second Breast 198 Bibliography Malignancy in Women With Early Breast Cancer. Journal of the National Cancer Institute , (93) 240–241. Merlos-­‐Suárez, A., Barriga, F. M., Jung, P., Iglesias, M., Céspedes, M. V., Rossell, D., … Batlle, E. (2011). The intestinal stem cell signature identifies colorectal cancer stem cells and predicts disease relapse. Cell Stem Cell, (8) 511–24. Milicic, A., Harrison, L.-­‐A., Goodlad, R. A., Hardy, R. G., Nicholson, A. M., Presz, M., … Jankowski, J. A. Z. (2008). Ectopic Expression of P-­‐Cadherin Correlates with Promoter Hypomethylation Early in Colorectal Carcinogenesis and Enhanced Intestinal Crypt Fission In vivo. Cancer Research, (68) 7760–7768. Mills, A. a. (2010). Throwing the cancer switch: reciprocal roles of polycomb and trithorax proteins. Nature Reviews. Cancer, (10), 669–82. Minucci, S., & Pelicci, P. G. (2006). Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer, (6) 38–51. Mizuno, H., Kitada, K., Nakai, K., & Sarai, A. (2009). PrognoScan: a new database for meta-­‐analysis of the prognostic value of genes. BMC Medical Genomics, (2) 18. Moise, A. R. (2010). Pharmacology of Retinoid Receptors. Mongan, N. P., & Gudas, L. J. (2007). Diverse actions of retinoid receptors in cancer prevention and treatment. Differentiation; Research in Biological Diversity, (75) 853–70. Muto, Y., Moriwaki, H., Ninomiya, M., Adachi, S., Saito, A., Takasaki, K. T., … Kojima, T. (1996). Prevention of Second Primary Tumors by an Acyclic Retinoid, Polyprenoic Acid, in Patients with Hepatocellular Carcinoma. New England Journal of Medicine, (334) 1561–1568. Nagano, T., Lubling, Y., Stevens, T. J., Schoenfelder, S., Yaffe, E., Dean, W., … Fraser, P. (2013). Single-­‐cell Hi-­‐C reveals cell-­‐to-­‐cell variability in chromosome structure. Nature, (502) 59–64. Nagaraj, N. S., Beckers, S., Mensah, J. K., Waigel, S., Vigneswaran, N., & Zacharias, W. (2006). Cigarette smoke condensate induces cytochromes P450 and aldo-­‐keto reductases in oral cancer cells. Toxicology Letters, (165) 182–194. Nagasaka, T., Tanaka, N., Cullings, H. M., Sun, D.-­‐S., Sasamoto, H., Uchida, T., … Goel, A. (2009). Analysis of fecal DNA methylation to detect gastrointestinal neoplasia. Journal of the National Cancer Institute, (101) 1244–1258. Naldini, L., Blömer, U., Gallay, P., Ory, D., Mulligan, R., Gage, F. H., … Trono, D. (1996). In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science, (272) 263–267. Narayan, G., Arias-­‐Pulido, H., & Koul, S. (2003). Frequent promoter methylation of CDH1, DAPK, RARB, and HIC1 genes in carcinoma of cervix uteri: its relationship to clinical outcome. Molecular Cancer, (12) 1–12. Nephew, K. P., & Huang, T. H.-­‐M. (2003). Epigenetic gene silencing in cancer initiation and progression. Cancer Letters, (190) 125–133. Newton, K. F., Newman, W., & Hill, J. (2012). Review of biomarkers in colorectal cancer. Colorectal Disease  : The Official Journal of the Association of Coloproctology of Great Britain and Ireland, (14) 3–17. Niles, R. M. (2004). Signaling pathways in retinoid chemoprevention and treatment of cancer. Mutation Research, (555) 81–96. Nishinaka, T., & Yabe-­‐Nishimura, C. (2005). Transcription factor Nrf2 regulates promoter activity of mouse aldose reductase (AKR1B3) gene. Journal of Pharmacological Sciences, (51) 43–51. 199 Bibliography Nosho, K., Shima, K., Irahara, N., Kure, S., Baba, Y., Kirkner, G. J., … Ogino, S. (2009). DNMT3B Expression Might Contribute to CpG Island Methylator Phenotype in Colorectal Cancer. Clinical Cancer Research , (15) 3663–3671. Novak, P., Jensen, T. J., Garbe, J. C., Stampfer, M. R., & Futscher, B. W. (2009). Stepwise DNA methylation changes are linked to escape from defined proliferation barriers and mammary epithelial cell immortalization. Cancer Research, (69) 5251–5258. O’Keefe, S. J. D., Kidd, M., Espitalier-­‐Noel, G., & Owira, P. (1999). Rarity of colon cancer in Africans is associated with low animal product consumption, not fiber. Am J Gastroenterol, (94) 1373–1380. Ohashi, T., Idogawa, M., Sasaki, Y., Suzuki, H., & Tokino, T. (2013). AKR1B10, a transcriptional target of p53, is Downregulated in Colorectal Cancers associated with poor prognosis. Molecular Cancer Research, (11) 1554-­‐1563. Okano, M., Bell, D. W., Haber, D. A., & Li, E. (1999). DNA Methyltransferases Dnmt3a and Dnmt3b Are Essential for De Novo Methylation and Mammalian Development. Cell, (99) 247-­‐257. Okudela, K., Woo, T., Mitsui, H., Suzuki, T., Tajiri, M., Sakuma, Y., … Ohashi, K. (2013). Downregulation of ALDH1A1 expression in non-­‐small cell lung carcinomas-­‐-­‐its clinicopathologic and biological significance. International Journal of Clinical and Experimental Pathology, (6) 1–12. Olek, A., & Walter, J. (1997). The pre-­‐implantation ontogeny of the H19 methylation imprint. Nature Genetics, (17) 275-­‐276. Ong, C.-­‐T., & Corces, V. G. (2014). CTCF: an architectural protein bridging genome topology and function. Nature Reviews Genetics, (15) 234–246. Osanai, M., Sawada, N., & Lee, G.-­‐H. (2010). Oncogenic and cell survival properties of the retinoic acid metabolizing enzyme, CYP26A1. Oncogene, (29) 1135–1144. Oshima, M., Dinchuk, J. E., Kargman, S. L., Oshima, H., Hancock, B., Kwong, E., … Taketo, M. M. (1996). Suppression of intestinal polyposis in Apc delta716 knockout mice by inhibition of cyclooxygenase 2 (COX-­‐2). Cell, (87) 803–
809. Ozdağ, H., Teschendorff, A. E., Ahmed, A. A., Hyland, S. J., Blenkiron, C., Bobrow, L., … Caldas, C. (2006). Differential expression of selected histone modifier genes in human solid cancers. BMC Genomics, (7). Park, E. Y., Dillard, A., Williams, E. a, Wilder, E. T., Pepper, M. R., & Lane, M. a. (2005). Retinol inhibits the growth of all-­‐trans-­‐retinoic acid-­‐sensitive and all-­‐trans-­‐retinoic acid-­‐resistant colon cancer cells through a retinoic acid receptor-­‐independent mechanism. Cancer Research, (65) 9923–9933. Park, E. Y., Wilder, E. T., Chipuk, J. E., & Lane, M. a. (2008). Retinol decreases phosphatidylinositol 3-­‐kinase activity in colon cancer cells. Molecular Carcinogenesis, (47) 264–274. Pastel, E., Pointud, J.-­‐C., Volat, F., Martinez, A., & Lefrançois-­‐Martinez, A.-­‐M. (2012). Aldo-­‐Keto Reductases 1B in Endocrinology and Metabolism. Frontiers in Pharmacology, (3) 148. Pastor, W. a, Aravind, L., & Rao, A. (2013). TETonic shift: biological roles of TET proteins in DNA demethylation and transcription. Nature Reviews. Molecular Cell Biology, (14) 341–356. 200 Bibliography Pastorino, U., Infante, M., Maioli, M., Chiesa, G., Buyse, M., Firket, P., … Valente, M. (1993). Adjuvant treatment of stage I lung cancer with high-­‐dose vitamin A. Journal of Clinical Oncology , (11) 1216–1222. Pennacchio, L. a., Bickmore, W., Dean, A., Nobrega, M. a., & Bejerano, G. (2013). Enhancers: five essential questions. Nature Reviews Genetics, (14) 288–295. Penning, T. M., & Drury, J. E. (2007). Human aldo-­‐keto reductases: Function, gene regulation, and single nucleotide polymorphisms. Archives of Biochemistry and Biophysics, (464) 241–250. Pennisi, E. (2011). Does a Gene’s Location in the Nucleus Matter? Science, (334) 1050–1051. Pennisi, E. (2013). The CRISPR craze. Science, (341) 833–836. Peric-­‐Hupkes, D., Meuleman, W., Pagie, L., Bruggeman, S. W. M., Solovei, I., Brugman, W., … van Steensel, B. (2010). Molecular maps of the reorganization of genome-­‐nuclear lamina interactions during differentiation. Molecular Cell, (38) 603–613. Petitjean, A., Mathe, E., Kato, S., Ishioka, C., Tavtigian, S. V, Hainaut, P., & Olivier, M. (2007). Impact of mutant p53 functional properties on TP53 mutation patterns and tumor phenotype: lessons from recent developments in the IARC TP53 database. Human Mutation, (28) 622–629. Plebuch, M., Soldan, M., Hungerer, C., Koch, L., & Maser, E. (2007). Increased resistance of tumor cells to daunorubicin after transfection of cDNAs coding for anthracycline inactivating enzymes. Cancer Letters, (255) 49–56. Pox, C. (2011). Colon cancer screening: which non-­‐invasive filter tests? Digestive Diseases (29), 56–59. Quinn, A., Harvey, R., & Penning, T. (2008). Oxidation of PAH trans-­‐dihydrodiols by human aldo-­‐keto reductase AKR1B10. Chemical Research in Toxicology, (21) 2207–2215. Ramana, K. V, Bhatnagar, A., Srivastava, S., Yadav, U. C., Awasthi, S., Awasthi, Y. C., & Srivastava, S. K. (2006). Mitogenic Responses of Vascular Smooth Muscle Cells to Lipid Peroxidation-­‐derived Aldehyde 4-­‐Hydroxy-­‐trans-­‐2-­‐nonenal (HNE): Role of aldose reductase-­‐catalized reduction of the HNE-­‐glutathione conjugates in regulation cell growth. Journal of Biological Chemistry , (281) 17652–17660. Ramana, K. V, Fadl, A. a, Tammali, R., Reddy, A. B. M., Chopra, A. K., & Srivastava, S. K. (2006). Aldose reductase mediates the lipopolysaccharide-­‐induced release of inflammatory mediators in RAW264.7 murine macrophages. The Journal of Biological Chemistry, (281) 33019–33029. Ramírez, N., Bandrés, E., Navarro, A., Pons, A., Jansa, S., Moreno, I., … García-­‐
Foncillas, J. (2008). Epigenetic events in normal colonic mucosa surrounding colorectal cancer lesions. European Journal of Cancer, (26) 2689-­‐2695. Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. a, & Zhang, F. (2013). Genome engineering using the CRISPR-­‐Cas9 system. Nature Protocols, (8) 2281–2308. Rashid, A., Shen, L., Morris, J. S., Issa, J.-­‐P. J., & Hamilton, S. R. (2001). CpG Island Methylation in Colorectal Adenomas. The American Journal of Pathology. American Society for Investigative Pathology. Ravindranath, T. M., Mong, P. Y., Ananthakrishnan, R., Li, Q., Quadri, N., Schmidt, A. M., … Wang, Q. (2009). Novel role for aldose reductase in mediating acute 201 Bibliography inflammatory responses in the lung. Journal of Immunology, (183) 8128–
8137. Reik, W. (2007). Stability and flexibility of epigenetic gene regulation in mammalian development. Nature, (447) 425–432. Renehan, A. G., O’Connell, J., O’Halloran, D., Shanahan, F., Potten, C. S., O’Dwyer, S. T., & Shalet, S. M. (2003). Acromegaly and Colorectal Cancer: A Comprehensive Review of Epidemiology, Biological Mechanisms, and Clinical Implications. Horm Metab Res, (35) 712–725. Rhee, I., Bachman, K. E., Park, B. H., Jair, K.-­‐W., Yen, R.-­‐W. C., Schuebel, K. E., … Vogelstein, B. (2002). DNMT1 and DNMT3b cooperate to silence genes in human cancer cells. Nature, (416) 552–556. Rideout, W. M., Coetzee, G. A., Olumi, A. F., & Jones, P. A. (1990). 5-­‐Methylcytosine as an endogenous mutagen in the human LDL receptor and p53 genes. Science, (249) 1288–1290. Ries, L. A. G., Melbert, D., Krapcho, M., Stinchcomb, D. G., Howlader, N., Horner, M. J., … Edwards, B. K. (2008). SEER cancer statistics review, 1975-­‐2005 . Bethesda, MD : U.S. National Institutes of Health, National Cancer Institute . Riggs, A. (1975). X inactivation, differentiation, and DNA methylation. Cytogenetic and Genome Research, (14) 9-­‐25. Rishi, V., Bhattacharya, P., Chatterjee, R., Rozenberg, J., Zhao, J., Glass, K., … Vinson, C. (2010). CpG methylation of half-­‐CRE sequences creates C/EBPalpha binding sites that activate some tissue-­‐specific genes. Proceedings of the National Academy of Sciences of the United States of America, (107) 20311–20316. Roberts, D. L., Dive, C., & Renehan, A. G. (2010). Biological mechanisms linking obesity and cancer risk: new perspectives. Annual Review of Medicine, (61) 301–316. Rodriguez, J., Muñoz, M., Vives, L., Frangou, C. G., Groudine, M., & Peinado, M. a. (2008). Bivalent domains enforce transcriptional memory of DNA methylated genes in cancer cells. Proceedings of the National Academy of Sciences of the United States of America, (105) 19809–19814. Rodriguez, J., Vives, L., Jordà, M., Morales, C., Muñoz, M., Vendrell, E., & Peinado, M. a. (2008). Genome-­‐wide tracking of unmethylated DNA Alu repeats in normal and cancer cells. Nucleic Acids Research, (36) 770–784. Rodríguez-­‐Paredes, M., & Esteller, M. (2011). Cancer epigenetics reaches mainstream oncology. Nature Medicine, (17) 330–339. Rose, D. P., Komninou, D., & Stephenson, G. D. (2004). Obesity , adipocytokines , and insulin resistance in breast cancer. Obesity Reviews, (7) 153–165. Rougier, N., Bourc’his, D., Gomes, D. M., Niveleau, a., Plachot, M., Paldi, a., & Viegas-­‐Pequignot, E. (1998). Chromosome methylation patterns during mammalian preimplantation development. Genes & Development, (12) 2108–2113. Rouits, E., Boisdron-­‐Celle, M., Dumont, A., Guérin, O., Morel, A., & Gamelin, E. (2004). Relevance of different UGT1A1 polymorphisms in Irinotecan-­‐
induced toxicity a molecular and clinical study of 75 patients. Clinical Cancer Research, (10) 5151–5159. Ruiz, F. X., Gallego, O., Ardèvol, A., Moro, A., Domínguez, M., Alvarez, S., … Farrés, J. (2009). Aldo-­‐keto reductases from the AKR1B subfamily: retinoid specificity 202 Bibliography and control of cellular retinoic acid levels. Chemico-­Biological Interactions, (178) 171–177. Ruiz, F. X., Porté, S., Parés, X., & Farrés, J. (2012). Biological role of aldo-­‐keto reductases in retinoic Acid biosynthesis and signaling. Frontiers in Pharmacology, (3) 58. Russo, V. E. A., Martienssen, R. A., & Riggs, A. D. (1997). Epigenetic Mechanisms of Gene Regulation. Edited. Genetics Research, (69) 159–162. Ruthenburg, A. J., Li, H., Patel, D. J., & Allis, C. D. (2007). Multivalent engagement of chromatin modifications by linked binding modules. Nature Reviews. Molecular Cell Biology, (8) 983–994. Salabei, J. K., Li, X.-­‐P., Petrash, J. M., Bhatnagar, A., & Barski, O. a. (2011). Functional expression of novel human and murine AKR1B genes. Chemico-­
Biological Interactions, (191) 177–184. Salabei, J., Li, X., & Petrash, J. (2011). Functional expression of novel human and murine AKR1B genes. Chemico Biological Interactions, (191) 177–184. Sales, K. J., Maldonado-­‐Pérez, D., Grant, V., Catalano, R. D., Wilson, M. R., Brown, P., … Jabbour, H. N. (2009). Prostaglandin F2α-­‐F-­‐prostanoid receptor regulates CXCL8 expression in endometrial adenocarcinoma cells via the calcium–
calcineurin–NFAT pathway. Biochimica et Biophysica Acta (BBA) -­ Molecular Cell Research, (12) 1917–1928. Sandoval, J., & Esteller, M. (2012). Cancer epigenomics: beyond genomics. Current Opinion in Genetics & Development, (22) 50–55. Sanyal, A., Lajoie, B. R., Jain, G., & Dekker, J. (2012). The long-­‐range interaction landscape of gene promoters. Nature, (489) 109–113. Schefe, J. H., Lehmann, K. E., Buschmann, I. R., Unger, T., & Funke-­‐Kaiser, H. (2006). Quantitative real-­‐time RT-­‐PCR data analysis: current concepts and the novel “gene expression’s CT difference” formula. Journal of Molecular Medicine, (84) 901–910. Schmidl, C., Klug, M., Boeld, T. J., Andreesen, R., Hoffmann, P., Edinger, M., & Rehli, M. (2009). Lineage-­‐specific DNA methylation in T cells correlates with histone methylation and enhancer activity. Genome Research, (19) 1165–
1174. Schmidt, W. M., Sedivy, R., Forstner, B., Steger, G. G., Zöchbauer-­‐Müller, S., & Mader, R. M. (2007). Progressive up-­‐regulation of genes encoding DNA methyltransferases in the colorectal adenoma-­‐carcinoma sequence. Molecular Carcinogenesis, (46) 766–772. Schmitz, K. J., Sotiropoulos, G. C., Baba, H. a, Schmid, K. W., Müller, D., Paul, A., … Loeffler-­‐Ragg, J. (2011). AKR1B10 expression is associated with less aggressive hepatocellular carcinoma: a clinicopathological study of 168 cases. Liver International  : Official Journal of the International Association for the Study of the Liver, (31) 810–816. Schübeler, D. (2012). Molecular biology. Epigenetic islands in a genetic ocean. Science, (338) 756–757. Schuettengruber, B., Martinez, A.-­‐M., Iovino, N., & Cavalli, G. (2011). Trithorax group proteins: switching genes on and keeping them active. Nature Reviews. Molecular Cell Biology, (12) 799–814. Scott, A. M., Wolchok, J. D., & Old, L. J. (2012). Antibody therapy of cancer. Nature Reviews. Cancer, (12) 278–87. 203 Bibliography Seligson, D. B., Horvath, S., Shi, T., Yu, H., Tze, S., Grunstein, M., & Kurdistani, S. K. (2005). Global histone modification patterns predict risk of prostate cancer recurrence. Nature, (435) 1262–1266. Shaib, W., Mahajan, R., & El-­‐Rayes, B. (2013). Markers of resistance to anti-­‐EGFR therapy in colorectal cancer. Journal of Gastrointestinal Oncology, (4) 308-­‐
318. Sheikhnejad, G., Brank, a, Christman, J. K., Goddard, a, Alvarez, E., Ford, H., … Cheng, X. (1999). Mechanism of inhibition of DNA (cytosine C5)-­‐
methyltransferases by oligodeoxyribonucleotides containing 5,6-­‐dihydro-­‐5-­‐
azacytosine. Journal of Molecular Biology, (285) 2021–2034. Shen, L., Kondo, Y., Hamilton, S. R., Rashid, A., & Issa, J. J. (2003, March 1). p14 methylation in human colon cancer is associated with microsatellite instability and wild-­‐type p53. Gastroenterology, (124) 626-­‐633. Shen, L., Kondo, Y., Rosner, G. L., Xiao, L., Hernandez, N. S., Vilaythong, J., … Issa, J.-­‐
P. J. (2005). MGMT promoter methylation and field defect in sporadic colorectal cancer. Journal of the National Cancer Institute, (97) 1330–1338. Shlyueva, D., Stampfel, G., & Stark, A. (2014). Transcriptional enhancers: from properties to genome-­‐wide predictions. Nature Reviews Genetics (15), 272–
286. Shukla, S., Kavak, E., Gregory, M., Imashimizu, M., Shutinoski, B., Kashlev, M., … Oberdoerffer, S. (2011). CTCF-­‐promoted RNA polymerase II pausing links DNA methylation to splicing. Nature (479) 74-­‐79. Siegel, R. (2013). Cancer statistics, 2013. … Cancer Journal for Clinicians, (63) 11–
30. Simon, J. (1995). Locking in stable states of gene expression: transcriptional control during Drosophila development. Current Opinion in Cell Biology, (7) 376–385. Simon, J. a, & Kingston, R. E. (2009). Mechanisms of polycomb gene silencing: knowns and unknowns. Nature Reviews. Molecular Cell Biology, (10) 697–
708. Singh, H., Nugent, Z., Demers, A. a, Kliewer, E. V, Mahmud, S. M., & Bernstein, C. N. (2010). The reduction in colorectal cancer mortality after colonoscopy varies by site of the cancer. Gastroenterology, (139) 1128–1137. Skrzypczak, M., Goryca, K., Rubel, T., Paziewska, A., Mikula, M., Jarosz, D., … Ostrowsk, J. (2010). Modeling oncogenic signaling in colon tumors by multidirectional analyses of microarray data directed for maximization of analytical reliability. PloS One, (5). Smith, H. W. (1965). Observations on the flora of the alimentary tract of animals and factors affecting its composition. The Journal of Pathology and Bacteriology, (89) 95–122. Smith, Z. D., & Meissner, A. (2013). DNA methylation: roles in mammalian development. Nature Reviews. Genetics, (14) 204–220. Smyth, G. K., Michaud, J., & Scott, H. S. (2005). Use of within-­‐array replicate spots for assessing differential expression in microarray experiments. Bioinformatics, (21) 2067–2075. Srivastava, S., Spite, M., Trent, J. O., West, M. B., Ahmed, Y., & Bhatnagar, A. (2004). Aldose Reductase-­‐catalyzed Reduction of Aldehyde Phospholipids. Journal of Biological Chemistry , (279) 53395–53406. 204 Bibliography Stadler, M. B., Murr, R., Burger, L., Ivanek, R., Lienert, F., Schöler, A., … Schübeler, D. (2011). DNA-­‐binding factors shape the mouse methylome at distal regulatory regions. Nature (480) 490-­‐495. Staub, E., Groene, J., Heinze, M., Mennerich, D., Roepcke, S., Klaman, I., … Rosenthal, A. (2009). An expression module of WIPF1-­‐coexpressed genes identifies patients with favorable prognosis in three tumor types. Journal of Molecular Medicine (Berlin, Germany), (87) 633–644. Sun, S.-­‐Y., & Lotan, R. (2002). Retinoids and their receptors in cancer development and chemoprevention. Critical Reviews in Oncology/hematology, (41) 41–55. Suzuki, K., Suzuki, I., Leodolter, A., Alonso, S., Horiuchi, S., Yamashita, K., & Perucho, M. (2006). Global DNA demethylation in gastrointestinal cancer is age dependent and precedes genomic damage. Cancer Cell, (9) 199–207. Syngal, S., Stoffel, E., Chung, D., Willett, C., Schoetz, D., Schroy, P., … Ross, M. (2006). Detection of stool DNA mutations before and after treatment of colorectal neoplasia. Cancer, (106) 277–283. Szwagierczak, A., Bultmann, S., Schmidt, C. S., Spada, F., & Leonhardt, H. (2010). Sensitive enzymatic quantification of 5-­‐hydroxymethylcytosine in genomic DNA. Nucleic Acids Research, (38) e181. Tagore, K. S., Lawson, M. J., Yucaitis, J. A., Gage, R., Orr, T., Shuber, A. P., & Ross, M. E. (2003). Sensitivity and Specificity of a Stool DNA Multitarget Assay Panel for the Detection of Advanced Colorectal Neoplasia. Clinical Colorectal Cancer, (3) 47–53. Tahiliani, M., Koh, K. P., Shen, Y., Pastor, W. A., Bandukwala, H., Brudno, Y., … Rao, A. (2009). Conversion of 5-­‐Methylcytosine to 5-­‐Hydroxymethylcytosine in Mammalian DNA by MLL Partner TET1. Science, (324) 930–935. Talbert, P. B., & Henikoff, S. (2010). Histone variants-­‐-­‐ancient wrap artists of the epigenome. Nature Reviews. Molecular Cell Biology, (11) 264–275. Tammali, R., Ramana, K. V, Singhal, S. S., Awasthi, S., & Srivastava, S. K. (2006). Aldose Reductase Regulates Growth Factor-­‐Induced Cyclooxygenase-­‐2 Expression and Prostaglandin E2 Production in Human Colon Cancer Cells. Cancer Research , (66) 9705–9713. Tammali, R., Srivastava, S. K., & Ramana, K. V. (2011). Targeting aldose reductase for the treatment of cancer. Current Cancer Drug Targets, (11) 560–571. Tan, M., Luo, H., Lee, S., Jin, F., Yang, J. S., Montellier, E., … Zhao, Y. (2011). Identification of 67 histone marks and histone lysine crotonylation as a new type of histone modification. Cell, (146) 1016–1028. Tang, X.-­‐H., & Gudas, L. J. (2011). Retinoids, retinoic acid receptors, and cancer. Annual Review of Pathology, (6) 345–364. Theodosiou, M., Laudet, V., & Schubert, M. (2010). From carrot to clinic: an overview of the retinoic acid signaling pathway. Cellular and Molecular Life Sciences, (67) 1423–1445. Thun, M. J., Jacobs, E. J., & Patrono, C. (2012). The role of aspirin in cancer prevention. Nature Reviews. Clinical Oncology, (9) 259–267. Tiscornia, G., Singer, O., & Verma, I. M. (2006). Production and purification of lentiviral vectors. Nature Protocols, (1) 241–245. Tolhuis, B., Palstra, R.-­‐J., Splinter, E., Grosveld, F., & de Laat, W. (2014). Looping and Interaction between Hypersensitive Sites in the Active β-­‐globin Locus. Molecular Cell, (10) 1453–1465. 205 Bibliography Tonus, C., Sellinger, M., Koss, K., & Neupert, G. (2012). Faecal pyruvate kinase isoenzyme type M2 for colorectal cancer screening: a meta-­‐analysis. World Journal of Gastroenterology, (18) 4004–4011. Triantafillidis, J. K., Nasioulas, G., & Kosmidis, P. a. (2009). Colorectal cancer and inflammatory bowel disease: epidemiology, risk factors, mechanisms of carcinogenesis and prevention strategies. Anticancer Research, (29) 2727–
2737. Tropberger, P., & Schneider, R. (2013). Scratching the (lateral) surface of chromatin regulation by histone modifications. Nature Structural & Molecular Biology, (20) 657–661. Tsai, H.-­‐C., & Baylin, S. B. (2011). Cancer epigenetics: linking basic biology to clinical medicine. Cell Research, (21) 502–517. Tsai, M.-­‐C., Manor, O., Wan, Y., Mosammaparast, N., Wang, J. K., Lan, F., … Chang, H. Y. (2010). Long Noncoding RNA as Modular Scaffold of Histone Modification Complexes. Science, (329) 689–693. Tsunoda, S., Smith, E., Young, N. J. D. E., Wang, X., Tian, Z., Liu, J., … Drew, P. A. (2009). TMEFF2 in esophageal squamous cell carcinoma, (21) 1067–1073. Van Engeland, M., Derks, S., Smits, K. M., Meijer, G. A., & Herman, J. G. (2011). Colorectal Cancer Epigenetics: Complex Simplicity. Journal of Clinical Oncology , (29) 1382–1391. Van Steensel, B. (2011). Chromatin: constructing the big picture. The EMBO Journal, (30) 1885–1895. Vanneman, M., & Dranoff, G. (2012). Combining immunotherapy and targeted therapies in cancer treatment. Nature Reviews. Cancer, (12) 237–251. Vermeulen, L., Morrissey, E., van der Heijden, M., Nicholson, a. M., Sottoriva, a., Buczacki, S., … Winton, D. J. (2013). Defining Stem Cell Dynamics in Models of Intestinal Tumor Initiation. Science, (342) 995–998. Viaud, S., Saccheri, F., Mignot, G., Yamazaki, T., Daillere, R., Hannani, D., … Zitvogel, L. (2013). The Intestinal Microbiota Modulates the Anticancer Immune Effects of Cyclophosphamide. Science, (342) 971–976. Vire, E., Brenner, C., Deplus, R., Blanchon, L., Fraga, M., Didelot, C., … Fuks, F. (2006). The Polycomb group protein EZH2 directly controls DNA methylation. Nature, (439) 871–874. Vogelstein, B., Papadopoulos, N., Velculescu, V. E., Zhou, S., Diaz, L. a, & Kinzler, K. W. (2013). Cancer genome landscapes. Science, (339) 1546–1558. Vona-­‐Davis, L., & Rose, D. P. (2007). Adipokines as endocrine, paracrine, and autocrine factors in breast cancer risk and progression. Endocrine-­Related Cancer, (14) 189–206. Wald, N., Idle, M., Boreham, J., & Bailey, A. (1980). Low serum-­‐vitamin-­‐A and subsequent risk of cancer: Preliminary results of a prospective study. The Lancet, (316) 813–815. Walker, J. M., & Trygve O.F. (2004). DNA Methylation. Protocols Epigenetics. Human Press. Wallis, Y. L., Morton, D. G., McKeown, C. M., & Macdonald, F. (1999). Molecular analysis of the APC gene in 205 families: extended genotype-­‐phenotype correlations in FAP and evidence for the role of APC amino acid changes in colorectal cancer predisposition. Journal of Medical Genetics , (36) 14–20. 206 Bibliography Wang, Y., Fang, M. Z., & Liao, J. (2003). Hypermethylation-­‐Associated Inactivation of Retinoic Acid Receptor β in Human Esophageal Squamous Cell Carcinoma, Clinical Cancer Research, (9) 5257–5263. Wei, P.-­‐L., Chang, Y.-­‐J., Ho, Y.-­‐S., Lee, C.-­‐H., Yang, Y.-­‐Y., An, J., & Lin, S.-­‐Y. (2009). Tobacco-­‐specific carcinogen enhances colon cancer cell migration through alpha7-­‐nicotinic acetylcholine receptor. Annals of Surgery, (249) 978–985. Weissman, S., Bellcross, C., Bittner, C., Freivogel, M., Haidle, J., Kaurah, P., … Daniels, M. (2011). Genetic Counseling Considerations in the Evaluation of Families for Lynch Syndrome—A Review. Journal of Genetic Counseling, (20) 5–19. Weng, J., Cao, Y., Moss, N., & Zhou, M. (2006). Modulation of Voltage-­‐dependent Shaker Family Potassium Channels by an Aldo-­‐Keto Reductase. Journal of Biological Chemistry, (281) 15194–15200. Widschwendter, M., Fiegl, H., Egle, D., Mueller-­‐Holzner, E., Spizzo, G., Marth, C., … Laird, P. W. (2007). Epigenetic stem cell signature in cancer. Nature Genetics, (39) 157–158. Winawer, S. J., Zauber, A. G., Gerdes, H., O’Brien, M. J., Gottlieb, L. S., Sternberg, S. S., … Bishop, D. T. (1996). Risk of Colorectal Cancer in the Families of Patients with Adenomatous Polyps. New England Journal of Medicine, (334) 82–87. Woenckhaus, M., Klein-­‐Hitpass, L., Grepmeier, U., Merk, J., Pfeifer, M., Wild, P. J., … Dietmaier, W. (2006). Smoking and cancer-­‐related gene expression in bronchial epithelium and non-­‐small-­‐cell lung cancers. The Journal of Pathology, (210) 192–204. Wolbach, S., & Howe, P. (1925). Tissue changes following deprivation of fat-­‐
soluble A vitamin. The Journal of Experimental Medicine (42) 753-­‐777. Wollowski, I., Rechkemmer, G., & Pool-­‐zobel, B. L. (2001). Protective role of probiotics and prebiotics in colon cancer. The American Journal of Clinical Nutrition, (73) 451S-­‐455S. Wu, S. C., & Zhang, Y. (2010). Active DNA demethylation: many roads lead to Rome. Nature Reviews. Molecular Cell Biology, (11) 607–20. Wutz, A. (2011). Gene silencing in X-­‐chromosome inactivation: advances in understanding facultative heterochromatin formation. Nat Rev Genet, (12) 542–553. Yadav, U. C. S., Ramana, K. V, Aguilera-­‐Aguirre, L., Boldogh, I., Boulares, H. a, & Srivastava, S. K. (2009). Inhibition of aldose reductase prevents experimental allergic airway inflammation in mice. PloS One, (4) e6535. Yao, H.-­‐B., Xu, Y., Chen, L.-­‐G., Guan, T.-­‐P., Ma, Y.-­‐Y., He, X.-­‐J., … Shao, Q.-­‐S. (2013). AKR1B10, a good prognostic indicator in gastric cancer. European Journal of Surgical Oncology, (40) 318-­‐324. Young, L., Sung, J., Stacey, G., & Masters, J. R. (2010). Detection of Mycoplasma in cell cultures. Nature Protocols, (5) 929–934. Yu, M., Hon, G. C., Szulwach, K. E., Song, C.-­‐X., Zhang, L., Kim, A., … He, C. (2012). Base-­‐resolution analysis of 5-­‐hydroxymethylcytosine in the mammalian genome. Cell, (149) 1368–1380. Zentner, G. E., & Henikoff, S. (2013). Regulation of nucleosome dynamics by histone modifications. Nature Structural & Molecular Biology, (20) 259–66. Zhang, F. L., & Casey, P. J. (1996). Protein Prenylation: Molecular Mechanisms and Functional Consequences. Annual Review of Biochemistry, (65) 241–269. 207 Bibliography 208 Vector Map
pCMV-HA-C
Catalog No(s).
635690
pCMV-HA-C Vector Information
A
B
CMV promoter 1–602
multiple cloning site 651–715
3× FLAG tag 716–787
internal ribosome entry site 823–1397
hrGFP ORF 1407–2123
SV40 polyA 2188–2571
f1 origin 2709–3015
LoxP sequence 3178–3211
ampicillin resistance (bla) ORF 3256–4113
pUC origin 4260–4927
P CMV
pUC ori
MCS
3x FLAG
IRES
pIRES-hrGFP-1a
5.0 kb
ampicillin
hrGFP
LoxP
f1 ori
SV40 pA
pIRES-hrGFP-1a Multiple Cloning Site Region
Appendix
(sequence shown 651–727)
™
Sac I*
™
Map of pcDNA 3.1/Hygro (+) and pcDNA 3.1/Hygro (–)
Vectors
Sac II*
Not I*
Srf I
Sma I/Xma I
BamH I
EcoR I
GA GCT CCA CCG CGG TGG CGG CCG CTC TAG CCC GGG CGG ATC CGA ATT C ...
STOP*
Sph I
Sal I
Xho I
start of FLAG tag
...GC ATG CGT CGA CTC GAG GAC TAC AAG GAT
The fi ure below summarizes the features of the pcDNA™3.1/H ro (+) and
Map of
pcDNA™3.1/Hygro (–) vectors. The complete sequences for
pcDNA™3.1/
pcDNA™3.1/Hygro (+) and pcDNA™3.1/Hygro (–) are available for
Hygro (+/–)
downloading
our web
site atsite.
www.lifetechnologies.com or from
pCMV-HA-C vector map andfrom
multiple
cloning
K=
!$#
K=
Clontech Laboratories, Inc.
A Takara Bio Company
1290 Terra Bella Avenue, Mountain View, CA 94043, USA
U.S. Technical Support: [email protected]
United States/Canada
800.662.2566
Last update: (121511)
Asia Pacific
+1.650.919.7300
Europe
+33.(0)1.3904.6880
!"#$N
$%#$N
&-0$N
560$N
5"4$N
!41$N
/,1O$N
234P$#
/,1O$N
/0%C$N
&,-7@3$N
.-+$N
)*+F$NNN
&'($NN
$%#$N
!"#
!"#$N
$%#$N
&'($NN
)*+F$NNN
&,-7@3$N
.-+$N
/0%C$N
/,1O$N
234P$#
/,1O$N
!41$N
5"4$N
560$N
&-0$N
$%#$N
Technical Support (see page 13).
To create a fusion of your protein of interest and the HA tag, insert your gene of interest in-frame with the HA coding sequence.
C
Japan
D
+81.(0)77.543.6116
!C
IJ4 +/
B
)2
Page 1 of 2
"*
@
*@ 456*"#
5,@
FG "
%
ED
/ # + @ , @? ? @
!"#$%&'()*+,-.
/'0123
%
AA b u
nA
) tys o I n d-­ it
c
BnI | a|
+/
e
E
!"#$%&'(')*&+$,-.*.$/012345
67$%&'(')*&8%&9(9:;$.9)*+$,-.*.$345233/
+H !
"<=)9%=*$>=':9:;$.9)*+$,-.*.$31?2@0@0
ABC$&*D*&.*$%&9(9:;$.9)*+$,-.*.$@0//2@051
ABC$%'=E-F*:E=-)9':$.9;:-=+$,-.*.$@0/@2@/5?
G@$'&9;9:+$,-.*.$@/132@7@@
H#I0$%&'(')*&$-:F$'&9;9:+$,-.*.$@7742/@00
CE;&'(E>9:$&*.9.)-:>*$;*:*+$,-.*.$/@@325@I@
H#I0$*-&=E$%'=E-F*:E=-)9':$.9;:-=+$,-.*.$5@?I25?/4
%J!$'&9;9:+$,-.*.$57342II?4$K>'(%=*(*:)-&E$.)&-:FL
M(%9>9==9:$&*.9.)-:>*$;*:*+$,-.*.$I40@2?I4@$K>'(%=*(*:)-&E$.)&-:FL
G
DF
Comments for
pcDNA™3.1/Hygro (+):
!"##$%&'()"*(+,-./0123456*"(789:
5597
nucleotides
((((;;<=(%>,?$"&@A$'
A B | | Ba ov i e
r
N ie
r I
ry
A
i
xee
Primer name
AKR1B1 CpGi C F1
AKR1B1 CpGi C R1
AKR1B1 CpGi C F2
AKR1B1 CpGi C R2
AKR1B1 CpG A F1
AKR1B1 CpG A R unique
AKR1B1 CpG A F2
AKR1B1 CpGi B F1
AKR1B1 CpGi B R1
AKR1B1 CpGi B F2
AKR1B1 CpGi B R2
AKR1B10 F1
AKR1B10 R1
AKR1B10 F2
AKR1B10 R2
AKR1B15 F1
AKR1B15 R unique
AKR1B15 F2
Mouse AKR1B3 CpGi F1
Mouse AKR1B3 CpGi R1
Mouse AKR1B3 CpGi F2
Mouse AKR1B3 CpGi R2
EN1 CpGi C F1
EN1 CpGi C R1
EN1 CpGi C F2
EN1 CpGi C R2
INHBB F1
INHBB R1
INHBB F2
INHBB R2
AA b
xex
u a Sequence
GAGTATTTATTTTTTAGGTATT!
TTTCCCACCAAATACAACA!
AATTTTAGGATGGGTATTTTG!
AACTAAAAAACTCCTTTCTAC!
ATTACCTCACTTATAAATAATTAA
TTTTTTGAATTATTAATTAGTTA
!
GGGTGTTTAGATTTTTTTTTT!
AGGGAAAGGAGGTTGGGTTT
!
CCATCCTAAAATTAAATACCTAA!
GGAAAGGAGGTTGGGTTTG
!
ATTAAATACCTAAAAAATAAATACT
!
TTATGTTGGTTAGGTTGGTTTT!
CTTACTTTCCCTATAATAATTTCCT!
ATTATAGGTGTGAGATATTATGTT
!
CTTACTTTCCCTATAATAATTTCCT
!
GGTTTATATTATTTTTAAGAGTTAT
!
ACCTTATACCCAATTAAATATCCT
!
TTGTAATATATATTGTGAAGGTTTATG!
AGAGGATAGAGAAATAAGTTGG
!
AAACCAAAACTTCTCCCCAATCTC
!
TGGGTGTGTTTTTATTTGGGG
!
TATCTCTAATCTACTAATTATTTTC
!
AGAATAATAAAGATAAGAGAT!
ACTATCCTACTTATAAACTC!
GTTTTAGGGATTTAGAGTTT!
CTACTTATAAACTCAACCAA!
AAGTTTTGAGGTTTAGTGGTTTT!
AAACCCTACCTCTATCCCAA!
GGTTAGGTTTTTAGTGGTTATT!
CCTCCCCAACCAACAAAAA!
A A aI A
|a
B eI I|t
!
Primer name
AKR1B1 Fw
AKR1B1 Rv
AKR1B10 Fw
AKR1B10 Rv
AKR1B15 Fw
AKR1B15 Rv
EN1 Fw
EN1 Rv
INHBB Fw
INHBB Rv
CPLX2 Fw
CPLX2 Rv
18S Fw
18S Rv
PP1A Fw
PP1A Rv
Beta-2-Micro Fw
Beta-2-Micro Rv
PSMC4 Fw
PSMC4 Rv
PUM1 Fw
PUM1 Rv
MRPL19 Fw
MRPL19 Rv
Akr1b3 mice Fw
Sequence
CCGTCTCCTGCTCAACAAC!
TACACATGGGCACAGTCGAT!
GCTTCTCGATCTGGAAGTGG!
GTAATGCCATCGGTGGAAAA!
CCCTTTGACTGGCCTAAAGAG!
AATGTGGCGATATTCTGCATCA!
TGGGTGTACTGCACACGTTATTC!
CTTGTCCTCCTTCTCGTTCTTCTT!
CGCGTTTCCGAAATCATCA!
GGACCACAAACAGGTTCTGGTT!
GAGGCGGAGCGGGAGAAGGTC!
GCCCGGGCAGGTATTTGAGCA!
GCGAAAGCATTTGCCAAGAA!
CATCACAGACCTGTTATTGC!
CTCCTTTGAGCTGTTTGCAG!
CACCACATGCTTGCCATCC!
CCAGCAGAGAATGGAAAGTC!
GATGCTGCTTACATGTCTCG!
TGTTGGCAAAGGCGGTGGCA!
TCTCTTGGTGGCGATGGCAT!
CGGTCGTCCTGAGGATAAAA!
CGTACGTGAGGCGTGAGTAA!
CAGTTTCTGGGGATTTGCAT!
TATTCAGGAAGGGCATCTCG!
AGGCCGTGAAAGTTGCTATTG
Akr1b3 mice Rv
Akr1b8 mice Fw
ATGCTCTTGTCATGGAACGTG
TCTGATTCGGTTTCACATCCAG
Akr1b8 mice Rv
Akr1b10 mice Fw
CCAGTTTCTGTTGAAGCTAAGGA
CTAGTGCCAAACCAGAGGACC
Akr1b10 mice Rv
Akr1b7 mice Fw
TCCTGTATTCGAGAAGGTGTCA
AGGCTGGGAATGCGTTATTAC
Akr1b7 mice Rv
mGAPDH Fw
GGGTGAGATAAGGGTGGCTCT
TGCACCACCAACTGCTTAG
mGAPDH Rv
mB2M Fw
GATGCAGGGATGATGTTC
GCTATCCAGAAAACCCCTCAA
mB2M Rv
m18S Fw
CATGTCTCGATCCCAGTAGACGGT
TTGACGGAAGGGCACCACCAG
m18S Rv
GCACCACCACCCACGGAATCG
AA b
u beI | | nA eI Genome
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Human
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
I|t
xe2
Primer name
Sequence
GCGTTGCATCATGCTTAGAA !
TCTGGAACTCGGGAGAAAAA!
GCGTTGCATCATGCTTAGAA !
TCTGGAACTCGGGAGAAAAA!
CAACCCCAGTTTGATGCTCT!
GCAACCTCCAAGTCAAGAGG!
ATTTTTCTCCCGAGTTCCAGA!
AATCCTGCGAGTGGGGTTTG!
TTCGCTTTCCCACCAGATAC!
CACGAGCACCTACCTTCCAG!
CCAAGCTTTTCGCTTTCCCACCAGATAC!
CCAAGCTTGGAGCCTTCTGATTGGTTGC!
CCAAGCTTTTCGCTTTCCCACCAGATAC!
CCAAGCTTGGGATGCTTCTTCCGCCT!
CCAAGCTTTTCGCTTTCCCACCAGATAC!
CCAAGCTTGGAGCCTTCTGAATCGTTGC!
AA b u a
I | eI I | NI
A a
I | be I AB| t
xe1
Genome coordinates
Amplicon size
chr7:134142082+134143103
1022 bp
Fw. Fragment 1 Luciferase
Rv. Fragment 1 Luciferase
Fw. Fragment 2 Luciferase
chr7:134142082+134143103
1022 bp
chr7:134144202+134145228
1027 bp
chr7:134143083+134143531
449 bp
chr7:134143736-134144109
374 bp
chr7:134143927-134144117
191 bp
chr7:134143985-134144117
133 bp
chr7:134143927-134144117
191 bp
Rv. Fragment 2 Luciferase
Fw. Fragment 3 Luciferase
Rv. Fragment 3 Luciferase
Fw. Fragment 2.1 Luciferase
Rv. Fragment 2.1 Luciferase
Fw. Fragment 2.2 Luciferase
Rv. Fragment 2.2 Luciferase
Fw. Fragment 2.3.1 Luciferase
Rv. Fragment 2.3.1 Luciferase
Fw. Fragment 2.3.2 Luciferase
Rv. Fragment 2.3.2 Luciferase
Fw. Fragment 2.3.1 Mut Luciferase
Rv. Fragment 2.3.1 Mut Luciferase
Bc
An nnI
A B
A
e
nA | D n
I
ABa|
Primer name
Sequence
Rv.CRISPR_AKR1B1.CpGi.GUIDE A
Rv.CRISPR_AKR1B1.CpGi.GUIDE A
Fw.CRISPR_AKR1B1.CpGi.GUIDE B
Rv.CRISPR_AKR1B1.CpGi.GUIDE B
Fw.CRISPR_NF-Y.Gene.GUIDE A
Rv.CRISPR_NF-Y.Gene.GUIDE B
Fw.CRISPR_NF-Y.Gene.GUIDE B
Rv.CRISPR_NF-Y.Gene.GUIDE B
Fw.CRISPR_CEBPB.Gene.GUIDE A
Rv.CRISPR_CEBPB.Gene.GUIDE A
Fw.CRISPR_CEBPB.Gene.GUIDE B
Rv.CRISPR_CEBPB.Gene.GUIDE B
AA b
u
BI
B| xa A | n
CACCGAGCCTTCTGATTGGTTGCGCTGG!
AAACCCAGCGCAACCAATCAGAAGGCTC!
CACCGCCGCTGCGCGAAGGAGCCTTCTG!
AAACCAGAAGGCTCCTTCGCGCAGCGGC!
CACCGCCACTGACCTGCACCATTAATGG!
AACCCATTAATGGTGCAGGTCAGTGGC!
CACCGTGGCCAACCCATCATGGTCCAGG!
AACCCTGGACCATGATGGGTTGGCCAC!
CACCGTCGGCACCGCCTGGTAGCCGAGG!
AACCCTCGGCTACCAGGCGGTGCCGAC!
CACCGTCCTCCTCGTCCAGCCCGCCCGG!
AACCCGGGCGGGCTGGACGAGGAGGAC!
n| t
xe3
Gene
450k Methylation Probe
ABCG5&8
cg26706238
ADH1
cg23949936
ADH4
cg12011299
AKR1B1
cg02215070
AKR1B10
cg11693019
AKR1B15
cg27018005
AKR1C1
cg10568634
AKR1C2
cg18841653
AKR1C3
cg12530994
AKR1C4
cg09272256
ALDH1A1
cg01812894
ALDH1A2
cg02888131
ALDH1A3
cg21359747
AWAT1
cg20220865
AWAT2
cg23904115
BCDO2
cg10503334
BCMO1
cg01678878
CD36C
cg06719671
CRABP1
cg22496102
CRABP2
cg03734660
CYP26A1
cg22536554
CYP26B1
cg20097440
CYP26E1
cg14250048
DHRS3
cg01083689
LPLC
cg07964216
LRAT
cg26084529
NPC1L1
cg06907626
RARA
cg24957657
RARB
cg12479047
RARG
cg08104462
RBP1
cg11027570
RBP4
cg18585903
RBP7
cg18086187
RDH10
cg20179892
RDH11
cg02218134
RDH12
cg21238221
RDH5
cg02192520
RDH8
cg25661884
RETSAT
cg02301651
RXRA
cg14654324
RXRB
cg08949668
RXRG
cg22675486
SCARB1
cg00894440
SULT1A1
cg15812873
AA b
xe4
u a A
B o BnA eI n
| a|
AB
A o| | n B
I BAn
e B k ot
Primer name
Sequence
Fw.Fragment1.AKR1B1.ChIP GCTCTTTCTGCTCCATCACC!
Rv.Fragment1.AKR1B1.ChIP TCCAGGAGGGAGCACTTTTA!
Fw.Fragment2.AKR1B1.ChIP CGGCGCGTACCTTTAAATAG!
Rv.Fragment2.AKR1B1.ChIP GAAAGAATCCGCTGCCACTA!
Fw.Fragment3.AKR1B1.ChIP TCTACCGGCTTGAGATGCTT!
Rv.Fragment3.AKR1B1.ChIP GGAAAGTGGGCTTCACAGAC!
Fw.Fragment4.AKR1B1.ChIP TTAAGGGCCTGGTGTTTCAC!
Rv.Fragment4.AKR1B1.ChIP AATGGAAGAATTGCCCACAC!
Fw.Fragment5.AKR1B1.ChIP TTTTCTGAGCCCCAACAATC!
Rv.Fragment5.AKR1B1.ChIP CCTTCACAAGAGGGAAGCAC!
Fw.Fragment6.AKR1B1.ChIP CCTCCCTCCTCCAACTCTTC!
Rv.Fragment6.AKR1B1.ChIP ATCCATTCCAGGCTTTTGTG!
Fw.Fragment7.AKR1B1.ChIP TTGCTCGGTTACGACATTTG!
Rv.Fragment7.AKR1B1.ChIP TTCCCCTCCAGAAGACAGTG!
Fw.Fragment8.AKR1B1.ChIP CATGCTGCCATCTGTCAACT!
Rv.Fragment8.AKR1B1.ChIP CCAACCCCTCCTGTGAGATT!
Fw.Fragment9.AKR1B1.ChIP CCTGCCAGAATGAACCCTAA!
Rv.Fragment9.AKR1B1.ChIP TGGAATTAAGCCTGCCAATC!
Fw.Fragment10.AKR1B1.ChIPCATTCCCATTTCCAGGTTTG!
Rv.Fragment10.AKR1B1.ChIP TGGGCTAGAAATCGTTGACC!
TTCTTCTGGGGTTGTTCCGTAATC!
THOC3 Fw ChIP
CACACCCGCTAGCCCTTTTCAT!
THOC3 Rv ChIP
CCTCCAGACCCCCACCCCATCC!
CPLX2 Fw ChIP
AGGCTTCCCCGCGGCTTCTCAG!
CPLX2 Rv ChIP
TTGCAACCGGGAAGGAAA!
GADPH Fw ChIP
TAGCCTCCGTCCAGCTGACTT!
GAPDH Rv ChIP
CAGTTTGCCAGATAGTCTCTTT!
16CEN Fw ChIP
GAGACATTTGGGAAGGTCACTGAAT!
16CEN Rv ChIP
AA b
u
eI I|t
xe5
Primer name
Fw8.akr1b.3C
Fw17.akr1b.3C
Fw16.akr1b.3C
Fw15.akr1b.3C
Fw13.akr1b.3C
Fw2.akr1b.3C
Fw4.akr1b.3C
Fw5.akr1b.3C
Fw19.akr1b.3C
Fw20.akr1b.3C
Fw22.akr1b.3C
Fw31.akr1b.3C
Sequence
CAGGGTCTCACTGTCTGTCG!
CCATGATGCTGTGAATGGTT!
GTAGGCTGCATGGACACGTA!
TTGGGAGGCTGAGGTATGAG!
TTGAGCCCAGGAATTGGA!
GCCTGTAGTCCCAGCTACCC!
TTTCCCAAGAGCTTGAAACTG!
TCCTGGCACACATAGCAAGA!
TCCTTCCAGTTTCCGTCTTAAA!
GTAAAAGGCTTTGGGTTCCAC!
AGAGGATGCAGAGCCAGAAG!
AGAAACAAAGCCAACGCACT!
AA b
u I n n| n d n i |B
A nI eI eI n nB I
xe6
nA nI BnA
I A | n B
eBaI eI I | t k V
AB
y yh
Gene
Expression probe
AKR1B1
A_23_P258190
AKR1B10
A_24_P129341
AKR1C3
A_23_P138541
ALDH1A1
A_23_P83098
ALDH1A2
A_24_P73577
CRABP1
A_33_P3326483
CYP26A1
A_23_P138655
CYP26B1
A_23_P210101
DGAT1
A_23_P112162
DHRS4
A_23_P14376
LRAT
A_32_P113066
RARA
A_32_P5251
RARB
A_24_P3243
RBP1
A_23_P257649
RETSAT
A_23_P209944
RXRA
A_23_P423197
RXRG
A_23_P23292
SULT1A1
A_24_P262201
AA b u
AB pCh;
beI | | nA eI n | t
In II o
xe7
Annex ANNEX X. CHIP-­‐SEQ DATA Scale
chr7:
134,142,900
134,143,000
134,143,100
AKR1B1
53 _
CpG: 93
134,143,200
134,143,300
500 bases
134,143,400
hg19
134,143,500
134,143,600
134,143,700
134,143,800
134,143,900
UCSC Genes (RefSeq, GenBank, CCDS, Rfam, tRNAs & Comparative Genomics)
134,144,000
134,144,100
134,144,200
134,144,300
134,144,400
CpG Islands (Islands < 300 Bases are Light Green)
CTCF_Control.profile
CTCF_Control.sig
0_
53 _
CTCF_Aza.profile
0_
62 _
H2AZ_Control.profile
CTCF_Aza.sig
H2AZ_Control.sig
0-18 _
62 _
H2AZ_Aza.profile
H2AZ_Aza.sig
0-18 _
38 _
H4K20me3_Control.profile
H4K20me3_Control.sig
0-18 _
38 _
H4K20me3_Aza.profile
H4K20me3_Aza.sig
0-18 _
95 _
H3K27me3_Control.profile
H3K27me3_Control.sig
0-36 _
95 _
H3K27me3_Aza.profile
H3K27me3_Aza.sig
0-36 _
260 _
H3K4me3_Control.profile
H3K4me3_Control.sig
-15 _
260 _
H3K4me3_Aza.profile
H3K4me3_Aza.sig
-15 _
62 _
H3K9me2_Control.profile
H3K9me2_Control.sig
0-
-35 _
62 _
H3K9me2_Aza.profile
H3K9me2_Aza.sig
0-
-35 _
33 _
H3K9me3_Control.profile
H3K9me3_Control.sig
0-
-23 _
33 _
H3K9me3_Aza.profile
H3K9me3_Aza.sig
0-
-23 _
Annex X | ChIP-­seq data on HCT116 control and treated with 5-­AzadC. 220 A
AA b
C
B
u
C) | B A A
AB
| A oD
nI
y yh A
y y eI nB A beI | | nAt
xxe
Annex ANNEX XII. ARTICLE Regina Mayor, Mar Muñoz, Marcel W. Coolen, Joaquin Custodio, Manel Esteller, Susan J.Clark and Miguel A. Peinado. Dynamics of bivalent chromatin domains upon drug induced reactivation and resilencing in cancer cells. Epigenetics 2011, 9: 1138-­‐1148. 222 RESEARCH PAPER
Epigenetics 6:9, 1138-1148; September 2011; © 2011 Landes Bioscience
Dynamics of bivalent chromatin domains
upon drug induced reactivation and resilencing
in cancer cells
Regina Mayor,1 Mar Muñoz,1 Marcel W. Coolen,2,3 Joaquin Custodio,1 Manel Esteller,4 Susan J. Clark2,5 and Miguel A. Peinado1,*
Institut de Medicina Predictiva i Personalitzada del Càncer (IMPPC); Badalona; Barcelona, Spain; 2 Epigenetics Group, Cancer Program; Garvan Institute of Medical Research;
Sydney, Australia; 3Department of Human Genetics; Nijmegen Centre for Molecular Life Sciences (NCMLS); Radboud University Nijmegen Medical Centre; Nijmegen, The
Netherlands; 4Cancer Epigenetics and Biology Program (PEBC); Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), L’Hospitalet; Barcelona, Spain; 5St. Vincent’s Clinical
School; University of NSW; Sydney, New South Wales, Australia
1
Key words: chromatin remodeling, DNA methylation, epigenetic drugs, colorectal cancer, epigenetic silencing
Epigenetic deregulation revealed by altered profiles of DNA methylation and histone modifications is a frequent event
in cancer cells and results in abnormal patterns of gene expression. Cancer silenced genes constitute prime therapeutic
targets and considerable progress has been made in the epigenetic characterization of the chromatin scenarios
associated with their inactivation and drug induced reactivation. Despite these advances, the mechanisms involved
in the maintenance or resetting of epigenetic states in both physiological and pharmacological situations are poorly
known. To get insights into the dynamics of chromatin regulation upon drug-induced reactivation, we have investigated
the epigenetic profiles of two chromosomal regions undergoing long range epigenetic silencing in colon cancer cells
in time-course settings after exposure of cells to chromatin reactivating agents. The DNA methylation states and the
balance between histone H3K4 methylation and H3K27 methylation marks clearly define groups of genes with alternative
responses to therapy. We show that the expected epigenetic remodeling induced by the reactivating drugs, just achieves
a transient disruption of the bivalent states, which overcome the treatment and restore the transcriptional silencing
approximately four weeks after drug exposure. The interplay between DNA methylation and bivalent histone marks
appears to configure a plastic but stable chromatin scenario that is fully restored in silenced genes after drug withdrawal.
These data suggest that improvement of epigenetic therapies may be achieved by designing strategies with long lasting
effects.
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Introduction
The combination of genetic and epigenetic lesions in cancer
cells results in altered gene expression profiles. DNA methylation and different modifications on histone tails are the two
principal forms of epigenetic regulation and both are largely
disturbed in cancer cells.1,2 Many unmethylated CpG islands
often become hypermethylated during cancer progression
resulting in epigenetic inactivation of the associated gene.
This appears to be a principal mechanism of tumor suppressor inactivation, but not all silenced genes are considered to
have antitumor activities and heterogeneous profiles have been
identified for different tumor types.1,2 Most CpG island hypermethylations appear as isolated and independent events, but
the concurrent hypermethylation of neighboring CpG islands
accompanied by global remodeling of the chromatin in large
chromosomal regions may also occur.3,4 This is a phenomenon
known as Long Range Epigenetic Silencing (LRES) and has
been reported to affect multiple chromosomal regions in different tumor types, including colorectal, prostate and breast
cancer.3-10
Histone modifications can lead to either activation or repression depending upon which residues are modified and the type of
modification.11 For instance, in histone 3, dimethylaton and trimethylation of lysine 4 (H3K4me3) are associated with transcriptional activity, while trimethylation of lysine 27 (H3K27me3) is
characteristic of silenced promoters. This later mark is driven by
the presence of the polycomb repressor complex and is believed
to direct de novo methylation in cancer-related silenced genes.12,13
Different studies have reported that in genes that become silenced
in cancer cells, the repressive mark H3K27me3 co-exists with
active marks (H3K4me3 and H3K4me2).6,14,15 These two opposite modifications participate in the mitotic inheritance of lineagespecific gene expression programs and have key developmental
functions. Its co-localization is considered characteristic of stem
cells and is believed to keep developmental regulator genes poised
for induction.16-21 More recently, it has been shown that bivalent
chromatin domains are also prone to DNA hypermethylation
*Correspondence to: Miguel A. Peinado; Email: [email protected]
Submitted: 04/18/11; Accepted: 07/15/11
DOI: 10.4161/epi.6.9.16066
1138
Epigenetics
Volume 6 Issue 9
RESEARCH PAPER
RESEARCH PAPER
in aging, providing an epigenetic link between the processes of
aging and cancer.22,23
Stem-cell like signatures are mimicked by cancer cells and
contribute to define their properties.24-26 Resetting of epigenetic
profiles appears as a novel and promising therapeutic strategy in
cancer,1,27 but a better understanding of the mechanisms setting
epigenetic memory and how the chromatin landscapes are either
maintained or modified is required to make it a broad reality.
This is especially important, since most of the critical genes in
cell programming and targets of epigenetic therapies show bivalent chromatin signatures in both stem and cancer cells.
Genome-scale approaches just illustrate the ubiquitous and
complex reorganization of epigenomic profiles upon treatment
with gene-reactivating drugs, as shown in a recent study.28 Due
to the heterogeneity of chromatin landscapes before and after the
treatment, insights into the regulatory mechanisms can be only
obtained from a detailed analysis of specific loci. An interesting example is a recent investigation in which a clear picture of
chromatin and gene expression dynamics has been achieved by
analysis of green fluorescent protein expression under the control of a methylated cytomegalovirus promoter after treatment
with 5-AzadC.29 This artificial setting illustrates the different
epigenetic events associated with changes in gene activity, but
it remains unsolved which mechanism or driving signal retains
the “switch-off” memento and is able to reset the original state of
the chromatin. Si et al. suggest that residual DNA methylation
near the CMV-GFP locus could play a role. It is also unknown
if the dynamics of chromatin remodeling and DNA methylation
observed in this system apply to endogenous genes.29
Although important insights into the mechanisms of epigenetic control have been made, the interaction between different coexisting epigenetic marks and the dynamics of the events
responsible for the repression/activation of chromatin are still
poorly known.11 Here we have characterized the epigenetic profiles of two chromosomal regions undergoing long range epigenetic silencing in most colorectal cancers, how these profiles are
affected by epigenetic drugs and how they are reset upon drug
withdrawal and maintenance in in vitro culture. Our results
illustrate the dominant nature of DNA methylation and bivalent
histone marks, which endure drug-induced changes in transcriptional activity.
chromosome maps and the molecular profiles of both regions are
depicted in Figure 1 and Supplemental Figure 1.
According to the expression levels and the epigenetic state of
their promoter, the genes could be classified into three groups
(Fig. 2). First group consists of silenced genes exhibiting methylated CpG island-promoter; this group includes EN1, SCTR
and INHBB (2q14.2) and HRH2, CPLX2 and SNCB (5q35.2)
that express at low levels in normal colon cells.3,6 A second group
consisted of genes with unmethylated CpG island-promoter,
and includes DDX18, INSIG2, PTPN4, RALB, TSN (2q14.2),
SFXN1 and THOC3 (5q35.2). These genes tend to be downregulated in colorectal tumors (as compared with the normal colon
cells) but still retain high expression levels.3,6 Finally, a third
group includes MARCO (2q14.2) and PCLKC (5q35.2) genes
that are low expressed and do not contain a CpG island in the
promoter region (Fig. 1 and Sup. Fig. 1).
Histone modification profiles were also analyzed by chromatin
immunoprecipitation (ChIP) and each one of the groups exhibited characteristic profiles (Fig. 2). A good agreement between
H3K4me2 and H3K4me3 profiles was observed (Sup. Fig. 2)
and for simplification, H3K4me2 data have been represented in
figures and Tables throughout the manuscript.
The concurrent presence of repressive H3K27me3 and active
H3K4me2/me3 chromatin marks was observed in genes of
the first group containing a DNA hypermethylated promoter
(Fig. 2). This is consistent with a bivalent state, as it has been
reported previously for some of these genes6 and in other genes
that become hypermethylated in cancer.14,15,31 Genes with these
features (N1, SCTR and INHBB, HRH2, CPLX2 and SNCB)
are referred as MBV (methylated and bivalent). Noteworthy, the
same genes that showed higher levels of H3K4me2/me3 within
this group, presented the highest levels of H3K27me3, as it is the
case of CPLX2 and INHBB (Fig. 2). H3K9Ac was absent from
MBV genes and some of them (i.e., EN1, SCTR) exhibited low
levels of H3K9me2 (Sup. Fig. 3). The H3K27me3/H3K4me2
ratio was above 0.5 in all MBV genes.
The activated state of genes of the second group was associated with high levels of H3K4me2/me3 (Fig. 2) and H3K9Ac
marks (Sup. Fig. 3). However, there was no correlation between
the gene expression levels and the amount of active marks when
comparing genes among them. As expected, H3K27me3 mark
was absent on the promoters of these genes. The H3K27me3/
H3K4me2 ratio was at least ten fold lower than the minimum
value observed in MBV genes. Genes of this group are referred as
ACTIVE, because they were expressed in all experimental situations, although the levels were not maintained.
MARCO and PcLKC (also known as PCDH24) genes displayed histone modification profiles compatible with bivalent
chromatin and accordingly their gene expression levels were very
low. MARCO exhibited the highest levels of repressive marks
H3K27me3 and H3K9me2 among all analyzed genes (Fig. 2 and
Sup. Fig. 3). Genes of this group do not contain a CpG island in
the promoter and are referred as NoCpGi.
Dynamics of epigenetic profiles in LRES regions upon
5-AzadC/TSA treatment and drug withdrawal. It has been
repeatedly shown that epigenetically silenced genes can be
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Results
DNA methylation, expression and histone patterns define
three types of genes in long range epigenetic silencing (LRES)
regions. In previous studies we have shown that LRES affects
chromosomal regions 2q14.2 and 5q35.2 in most colorectal cancers.3,6,30 Although genetic activity is downregulated all along the
LRES region, uneven epigenetic profiles are likely to define different chromatin domains. To better characterize the epigenetic
regulation of the diverse chromatin domains in LRES we have
investigated gene expression and epigenetic profiles in nine and
six genes embedded in chromosomal regions 2q14.2 and 5q35.2,
respectively, in HCT116 colon cancer cells. An overview of the
www.landesbioscience.com
Epigenetics
1139
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 1. Scheme of the chromosomal regions analyzed at epigenetic level. Genes are depicted as arrowheads (indicating the transcription direction)
and coded according to its epigenetic signature in HCT116 cells. Closed black arrowhead: DNA methylated and bivalent chromatin marks (MBV), Open
arrowhead: Active genes, Grey arrowhead: No CpG island.
reactivated by treating the cells with the demethylating agent
5-Aza-2'-deoxycytidine (5AzadC) and the inhibitor of histone
deacetylases Trichostatin A (TSA).27,32 More recently it has
also been shown that bivalent domains in silenced genes are
not resolved in spite of the reactivation upon drug treatment.6
To get a better picture of the events associated with gene activation and silencing in these chromosomal regions we performed
a time course experiment after treating the cells separately with
5AzadC or TSA, and with a combination of both. Cells were
analyzed at days 1 (the same day the drugs were removed), 16,
22 and 29 after treatment. As expected, 5AzadC alone or in combination with TSA resulted in global demethylation of the seven
hypermethylated genes as determined by direct bisulfite sequencing and melting curve analysis (data not shown). For a subset of
genes (EN1, SCTR and INHBB), a quantitative determination
was performed using the MassCLEAVETM DNA methylation
assay system. A 60% of loss of methylation was observed just
after the treatment (day 1) in 5AzadC treatment alone (data not
shown) and after the co-treatment that affected most CpG sites
(Fig. 3). Some CpG sites (SCTR 6_7 and SCTR 40) exhibiting
lower methylation levels (about 50%) in untreated cells did not
display changes upon treatment.
1140
Upon drug withdrawal, a remethylation to >90% was reached
in 5AzadC treated cells after 16 days (data not shown), while the
co-treatment appeared to delay the full remethylation, that was
achieved two weeks later (day 29) (Fig. 3). TSA treatment alone
did not affect DNA methylation (data not shown).
5AzadC treatment restored the expression of genes with a hypermethylated promoter (MBV group) (data not shown), although
the co-treatment with TSA resulted in a more substantial reactivation, that was maintained for about two weeks but returned
to original levels 4 weeks later (Fig. 4), consistent with the DNA
methylation profiles. H3K4me2/3 and H3K27me3 marks were
also increased in response to the drugs, which is suggestive of
balanced dynamics. After drug withdrawal, the active chromatin
mark H3K4me2/me3 exhibited a time dependent decrease that,
after four weeks, reached levels similar to the untreated cells for
genes of the chromosomal band 5q35.2 and even lower for those
of the 2q14.2 region. On the other hand, the repressed chromatin
mark H3K27me3 displayed a heterogeneous profile and in some
genes retained higher levels all along the time course study (Fig.
4). This result could be interpreted as a partial disruption of the
balance between active and inactive marks during the resilencing process. Nevertheless, the alternation of different cell populations with distinct chromatin signatures might also contribute
Epigenetics
Volume 6 Issue 9
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 2. Gene expression and chromatin modification patterns of genes classified into three groups according to their epigenetic signatures in
HCT116 cells. MBV corresponds to silenced genes exhibiting a DNA methylated promoter embedded in bivalent chromatin (presence of methylation
in H3K27 and H3K4); ACTIVE corresponds to unmethylated and expressed genes; NoCpGi corresponds to genes without CpG island. Gene expression levels were normalized to 18S. H3K4me2 and H3K27me3 ChIP levels were normalized to input and the ratios between both marks are shown. All
quantifications were performed by real-time PCR in triplicate. Error bars indicate standard deviation. Note that gene expression and H3K4me2 values
are represented in a different scale for ACTIVE genes due to their high values as compared with the other two groups. The regional profiles of genes
arranged by chromosomal position in shown in Supplemental Figure 1.
to the observed changes. H3K4me2 and H3K4me3 active marks
showed parallel changes in most of the experiments in which both
www.landesbioscience.com
were analyzed, although H3K4me2 changes appeared to occur at
a slower pace than H3K4me3 (Sup. Fig. 2).
Epigenetics
1141
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 3. Relative methylation levels of the named CpG sites contained in the EN1, SCTR and INHBB amplicons as analyzed by MassCLEAVETM after
5AzadC/TSA co-treatment. A pool of three direct bisulfite PCR products was used for analysis. Represented values correspond to the ratio of treated
versus untreated HCT116 cells (Y axis). The gradient scale represents the different time points corresponding to days 1, 16, 22 and 29 after treatment
(from dark to light). Mean methylation values of all the CpG sites are represented in the right part.
On the other hand, ACTIVE genes exhibited a completely
different behavior in response to the various treatments. Most
of them suffered a clear decrease in the transcription rates by
day 1, either with 5AzadC or TSA alone (data not shown), or
with the combination of both (Fig. 4). Reduction of expression
was paralleled by an initial decrease in H3K4me2/me3 mark.
The initial expression levels were recovered 16 days after the
drug withdrawal as the H3K4me2/me3 mark did. H3K27me3
mark remained low or undetectable in these genes along the time
course, although its sporadic presence could be detected in some
genes (Fig. 4).
Finally, MARCO and PcLKC gene expression was slightly
induced by the treatment. This reactivated status was maintained
over 16 days. Afterwards, transcription rates started to decrease to
the low or undetectable levels of untreated cells (Fig. 4). Changes
1142
in chromatin modifications were also minor, with slight increases
of H3K4me2/3 and H3K27me3 in PcLKC along the time course
(Fig. 4).
Notably, for all three groups of genes, an enrichment of
H3K4me2/me3 levels was found after treating the cells with TSA
alone (Sup. Fig. 2B and data not shown), but this effect had disappeared by day 16 for most of the genes, and was not reflected
on the expression patterns. Regarding other histone modifications analyzed, H3K9Ac was increased in most genes after TSA
or the combined treatment but not when treated with 5dAzaC
alone as expected (data not shown). H3K9me2 and total H3 levels displayed small variations without a consistent pattern.
Maintenance of chromatin states in DNA methylation deficient cells. Since DNA demethylation appeared to reactivate
silenced genes but was not able to resolve bivalent chromatin,
Epigenetics
Volume 6 Issue 9
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 4. Relative expression and histone mark (H3K4me2 and H3K27me3) levels in the three classes of genes after 5-AzadC and TSA co-treatment.
Relative values were calculated as the log2 of the treated/untreated ratio. The gradient scale indicates the different time points corresponding to days
1, 16, 22 and 29 after treatment (from dark to light).
we wondered if stable demethylation will be sufficient to override the epigenetic memory. We analyzed the DKO cells, a variant of HCT116 colon cancer cells which is deficient for DNA
methyltransferases,33 although retaining some residual DNMT1
activity.34 As expected, hypermethylated genes (MBV) in the
wild-type HCT116 were fully demethylated in DKO cells (data
not shown), and their expression was restored to higher levels than in the parental cells, although two of the genes, EN1
and CPLX2, exhibited a limited reactivation (Fig. 5A). As in
drug induced gene reactivation, DKO cells exhibited an overall
increase of the H3K4me2/me3 and H3K27me3 marks in the
demethylated MBV genes (Fig. 5A), which suggests a dynamic
rebalancing of these two histone modifications enduring the
retention of the bivalent states. These results are in agreement
with previous observations in other genes.15 Genes with unmethylated CpG islands (ACTIVE group) in the HCT116 cells and
those without CpG island (noCpGi group) exhibited null or
minimal changes in gene expression and histone modification
marks (Fig. 5A and data not shown).
The treatment of DKO cells with TSA resulted in an
increased expression of all the MBV genes (Fig. 5B) that was paralleled by a partial disruption of the balance between H3K4me2
Figure 5 (See opposite page). (A) Relative gene expression and histone modification profiles in DKO cells in regard to parental HCT116. Relative
values were calculated as log2 of DKO/HCT116 ratio. Gene expression was normalized to 18S. ChIP results were normalized to the respective input
fractions. Major differences are seen in the methylated bivalent genes (MBV) that exhibit DNA demethylation, higher expression and increased levels
of histone marks H3K4me2 and H3K27me3. (B) Relative gene expression and modification profiles in DKO cells treated with TSA in regard to untreated
DKO cells. Relative values were calculated as log2 of treated/untreated cells ratio. Gene expression was normalized to 18S. ChIP results were normalized to the respective input fractions.
www.landesbioscience.com
Epigenetics
1143
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 5. For figure legend, see page 6.
1144
Epigenetics
Volume 6 Issue 9
only a fraction of the cells and the fraction of cells exhibiting changes may be different for every assessed parameter.
To determine whether drug induced DNA demethylation is
directly associated with the changes detected in active and
repressive histone modifications, we analyzed DNA methylation in H3K4me2 and H3K27me3 chromatin fractions in
untreated HCT116 and DKO cells, 5-AzadC and 5-AzadC/
TSA treated HCT116 cells and TSA treated DKO cells.
DNA methylation profiles of five MBV genes (EN1, SCTR,
INHBB, CPLX2 and HRH2) were analyzed in the Input,
H3K4me2 and H3K27me3 ChIP fractions by bisulfite
sequencing and MassCLEAVETM.
High levels of DNA methylation were observed in all
chromatin fractions analyzed in HCT116 cells, suggesting
that DNA methylation was homogeneous and ubiquitous
in the five genes analyzed. Illustrative results of HRH2 and
EN1 genes have been represented (Fig. 6A). Upon treatment
(5AzadC alone or in combination with TSA), marked differences between H3K4me2/me3 and H3K27me3 fractions
were appreciated. A dramatic decrease in DNA methylation
was observed in the input and the H3K4me2/me3 fraction of treated cells (20–30% and 10–30% of that found
in untreated cells), but the H3K27me3 fraction retained
a higher level of DNA methylation (40 to 100% of that
found in untreated cells). This discordance suggests that
drug treatment ensues in at least two different cell populations. The responsive cell population would be characterized by the loss of DNA methylation, gene re-expression,
increased levels of H3K4me2, and probably by loss of the
H3K27me3 mark. A second population of resistant cells
would retain DNA methylation and probably increased levels of H3K27me3.
Next we wondered if cells retaining DNA methylation
underwent replication. We analyzed the DNA methylation
and histone marks in replicating and non-replicating cells
after fluorescence assisted cell sorting (FACS) of cells cultured in medium containing bromodeoxyuridine (BrdU).
All fractions exhibited DNA demethylation, although it was
deeper in cells that have incorporated high levels of BrdU.
After 1 week, the demethylation was maintained and the differences between BrdU positive and negative cells were still apparent (Fig. 6B). Combined analysis of histone modifications and
DNA methylation was performed in cells labeled with propidium
iodide. Cells were fractioned by FACS into two groups (G0 /G1 and
S/G2). Both groups retained balanced levels of H3K27me3 and
H3K4me3, although DNA demethylation tended to be deeper in
H3K4me3 fractions and cells in S/G2 phases. Equivalent results
were obtained 1 day and 7 days after treatment (data not shown).
We do not know if gene reactivation of silenced genes takes
place in all cells, but a deeper DNA demethylation is consistently
associated with enrichment of the H3K4me2/me3 marks and, as
expected, DNA replication. H3K27me3 mark was present in all
cell fractions. DKO cells exhibited low methylation levels similar
to drug treated cells and no differences among treatments and
chromatin fractions.
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Figure 6. DNA methylation analysis of cell fractions after 5AzadC/TSA treatment. (A) MassCLEAVETM DNA methylation analysis of input, H3K4me2 and
H3K27me3 immunoprecipitated DNA fractions in EN1 and HRH2 gene CpG
islands. Each bar represents the mean methylation (1.0 is full methylation)
of all CpG sites contained in the amplicon analyzed in untreated HCT116,
5AzadC and 5AzadC/TSA treated cells and DKO and TSA DKO treated cells.
(B) Direct bisulfite sequencing of CpG island regions corresponding to three
MBV genes in HCT116 cells after BrdU labeling and FACS.
and H3K27me3 marks (higher H3K4me2 levels but lower
H3K27me3) (Fig. 5B). The rest of genes were not affected by
TSA, with the exception of DDX18 and PcLKC, which were also
upregulated. A global increase of H3K4me2 levels was also seen
in most genes (Fig. 5B).
These results suggest that retention of the bivalent nature of
chromatin in the MBV genes is preserved even in the absence
of an efficient DNA methylation system, although the balance
between H3K4me2/me3 and H3K27me3 marks is partially disrupted after inhibition of histone deacetylases with TSA.
Epigenetic profiles of different chromatin and cell fractions. In vitro treatments with epigenetic drugs clearly show
that silenced genes can be reactivated but do not illustrate the
direct associations among the different components of the epigenetic code and gene activity in response to the drug. This is
because changes observed in response to drugs may be driven by
www.landesbioscience.com
Epigenetics
1145
Discussion
Reactivation of tumor suppressor genes that have undergone epigenetic silencing appears as a promising therapeutic strategy.32,35
It has been shown that reactivated promoters are often slowly
remethylated and the gene is resilenced after withdrawal of the
DNA methylation inhibitor.36-40 Recent discoveries have revealed
chromatin scenarios linking stem cell and cancer biology 6,12,14,15,25
and have underscored the impact of neighboring epigenetic states
on the regulation of chromatin stability.3,6,9 To get insights into
the dynamics of chromatin reactivation and resilencing in cancer
cells submitted to epigenetic treatments, we have investigated the
epigenetic profiles of two of these chromosome regions, 2q14.2
and 5q35.2, containing genes silenced in colorectal cancer. We
have defined three different genetic compartments based on the
epigenetic signatures of the genes.
The first group is represented by six genes undergoing epigenetic silencing in cancer cells (as denoted by DNA hypermethylation) and the retention of bivalent chromatin epigenetic marks
(MBV genes). It is of note the bivalent nature of these genes in
murine colon cells, that has been demonstrated by co-immunoprecipitation of H3K4me3 and H3K27me3 (unpublished data).
When reactivated, the expression of these genes is restored but at
low levels, similar to those detected in non-tumorigenic cells,3,6
and histone modifications profiles are also changed to a more pronounced bivalent state (elevated H3K4me2/3 and H3K27me3
signals) as it has been previously reported in reference 6 and
40. Upon drug withdrawal, the bivalent marks are maintained,
but the balance is partially disrupted by a predominance of the
repressive mark H3K27me3, suggesting its contribution to the
recovery of the silenced state. The overall data are consistent with
a coordinated dynamics of bivalent chromatin signatures among
the MBV genes Nevertheless, it should be noted that our analysis
is limited to specific regions inside the CpG island. Regional profiles along and outside the promoter region may be also affected
by the treatments and may exhibit different dynamics during
the resilencing process. Future studies a genome scale should
address this issue. The effects of the treatment are reversed at
mid-term and a complete resilencing of the reactivated genes is
accomplished three weeks after drug withdrawal, while the DNA
remethylation is fully restored one week later for most of the CpG
islands analyzed. The combined treatment resulted in an extended
recovery period as compared with 5-AzadC alone. These results
are consistent with a secondary nature of DNA methylation in
regard to gene activity.39,41 The percentages of demethylation and
the rate of remethylation were of the same order as previously
described for other genes.36,38,39 While treatment with the histone
deacetylases inhibitor agent TSA alone did neither induce DNA
demethylation nor gene reactivation, it had a synergistic effect
with 5-AzadC, in agreement with other studies.42,43
The use of cells deficient in DNA methyltransferase activity
(DKO) has allowed us to contrast the results presented above
in a more stable system. As it has been described before, most
hypermethylated tumor suppressor genes in HCT116 cells,
such as p16INK4a or TIMP3, are found demethylated and reexpressed in DKO cells as compared with the parental HCT116
cells.33,44 Most HCT116 hypermethylated genes are reexpressed
and exhibit higher levels of H3K4me2 and H3K27me3 marks
in the DKO cells (Fig. 5A). These results are similar to what
we obtain when HCT116 cells are treated with 5-AzadC alone
(our data and ref. 40). As expected, DKO cells treated with
TSA exhibited increased expression of genes hypermethylated in
HCT116. Interestingly, TSA had deep effects on DKO cells by
disrupting the balance between bivalent marks (Fig. 5B). While
H3K4me2 is increased in most genes, the repressive H3K27me3
mark is lost from most of the genes in which promoter hypermethylation was present (MBV). Our results suggest that DNA
methylation also contributes to maintain the repressive histone
code in genes poised for silencing and confirm the role of DNA
methylation as the foremost player in preserving the inactivation
of silenced genes.38 Besides that, retention of bivalent signatures
probably acts as a memento and is likely to participate in the restoration of the silencing.6
Another interesting finding arises from the analysis of DNA
methylation in different chromatin fractions and in various
experimental conditions. At global level, all the MBV genes analyzed in HCT116 untreated cells show coexistence of histone
marks H3K4me2 and H3K27me3 with fully methylated DNA
(Fig. 6A). Nevertheless, upon DNA demethylation by treatment
with 5AzaC, H3K27me3 chromatin fraction retains higher levels of DNA methylation, while the H3K4me2/me3 fractions are
extensively DNA demethylated. This result could be interpreted
as a partial disruption of bivalent domains that become active
in a fraction of the cells (unmethylated DNA and H3K4me2),
while reaching a new balance in “drug-resistant cells” (methylated DNA and bivalent marks).
It remains to be elucidated if the recovery of the silenced
state after drug removal is due to chromatin re-remodeling or to
a cell population renewal in which a reservoir of cells resistant
to the drug (denoted by the presence of high levels of methylation associated with the H3K27me3 chromatin fraction) become
prevalent in the cell population. While it seems clear the homogeneity of the parental cells, it could be hypothesized that the
treatment induces the presence of mixed cell populations with
different chromatin landscapes (active, silenced, bivalent). We
tried to resolve this conundrum by coimmunoprecipitation of
both histone marks H3K4me2 and H3K27me3, but it did not
work in our hands, although we have been able to do it in murine
cells (data not shown). Our data indicate that the retention of
repressive marks is unlikely to be driven by a resting cell population, since epigenetic profiles of bivalent chromatin domains
are similarly affected in replicating and non-replicating cells. In
agreement with previous studies,6,15 H3K27me3 is retained and
even increased in the reactivated loci. Hence, this repressing histone modification could trigger the resilencing and the DNA
remethylation in the affected regions. Additional studies using
inhibitors of H3K27 methyltransferase, as the DZNep,45 may
contribute to better understand the role of H3K27me3 mark in
this process. Residual DNA methylation in the locus or near the
locus is also likely to play a role.29
As we have shown, treated cells recover the original chromatin
states after a few weeks of drug withdrawal. Moreover, they show
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
1146
Epigenetics
Volume 6 Issue 9
similar sensitivity and response to the drugs upon retreatment
as never-treated cells (data not shown), suggesting that selection of drug-resistant cells is not sufficient to explain our results.
Epigenetic drugs used here appear to induce plastic effects on
chromatin. The transient disruption of the silenced states
achieved by these treatments is overcome by a dominant repressive chromatin landscape determined by DNA methylation and
bivalent chromatin marks. Design of new drugs with long lasting
effects should improve the efficiency of epigenetic therapies.
In summary, our work illustrates the complexity of the epigenetic changes occurring upon drug-induced activation of silenced
genes. While DNA methylation appears to play a dominant role
in long lasting silencing, other players, characterized here as bivalent histone marks, contribute to maintain the epigenetic memory
of silenced genes. The histone modification H3K27me3 is likely
to play an important role in maintaining the memory of silenced
genes that are in a bivalent context. Novel epigenetic strategies in
cancer should target this repressive mark in order to get a more
efficient effect of current DNA demethylating therapies.
Materials and Methods
Cell culture and 5-Aza-2'-deoxycytidine and trichostatin A
treatments. The HCT116 colorectal carcinoma cell line was
obtained from the American Type Culture Collection (ATCC).
Cell culture and treatment with 5-Aza-2'-deoxycytidine
(5AzadC) and Trichostatin A (TSA) were performed as described
in reference 3, with minor modifications. Briefly, 0.65 x 106 cells
were seeded in 10-cm cell culture dishes and 24 h later treated
with 0.5 μM 5-Aza-2'-deoxycytidine (5AzadC; Sigma) for 48 h.
Cells were treated with Trichostatin A (TSA) (Sigma) at 0.3 μM
for 16 h. For co-treatment of cells with 5AzadC and TSA, 0.8 x
106 cells were seeded and treated initially with 5AzadC for 48
h, afterwards the medium was removed and cells were treated
with TSA for an additional 16 h; then, the medium was changed
and cells were harvested at the indicated time points. For control
samples, half a million cells were seeded and cultured at the same
time in which the drugs were omitted. DKO cells (HCT116 deficient in both DNA methyltransferases),46 were cultured under
the same conditions as wild type HCT116 cells. A complete new
treatment of HCT116 cells with 5AzadC and TSA was done in
parallel as control of drug efficiency.
Fluorescence-activated cell sorting (FACS). Cell sorting
of replicating and non-replicating cells was performed using a
modification of the method described previously in reference 47.
Briefly, untreated and treated (5AzadC + TSA) cells were incubated for 24 h with 30 μM BrdU (Sigma) immediately and 7 days
after the treatment. Cells were fixed with 3.7% formaldehyde at
room temperature for 10 min, washed 3 times with cold PBS.
The blocking solution (15% goat serum in 0.1% TritonX100)
was added to the cells and incubated for 30 minutes at room
temperature. After removal of blocking solution, fixed cells were
incubated for 1 h at room temperature with anti-BrdU mouse
monoclonal antibody 1:250 (Becton Dickinson, Ref. 555627),
washed 3 times with PBS, and incubated again with anti-mouse
FITC-conjugated antibody 1:250 (Invitrogen, Ref. A10543) for
45 minutes at room temperature in a dark chamber. Cells were
washed twice with PBS and Tween (0.02%) and 2 more times
with PBS. Flow cytometry and sorting was performed on an
Influx flow cytometer (BD FACSAriaTM II). Standard negative
controls were used to set up the threshold and to fractionate cells
positive and negative for BrdU.
Cell cycle analysis. Cell cycle analysis using PI was performed
in 5dAzadC+TSA treated cells 1 and 7 days after the treatment.
Cells were resuspended in 0.9 ml PBS, permeabilized with 2.1
ml 100% ethanol and kept at -20°C for 30 min. Then, cells were
washed twice to eliminate the ethanol and 1 ml of the analysis solution compound (Propidium iodide 500 μg/ml (Sigma),
sodium citrate 38 mM and Ribonuclease A (Sigma) was added.
Flow cytometry and sorting was performed as described above.
Fractions for G0 /G1 and S/G2 were collected.
Bisulfite sequencing and massCLEAVETM analysis. Genomic
DNA was obtained using standard protocols. Bisulfite treatment
was performed as previously described in reference 41, or using
the EZ DNA methylation kitTM (Zymo Research). Three independent PCR reactions were carried out and products were pooled to
ensure a representative methylation profile. The primers used for
the bisulfite PCR amplifications are listed in Supplemental Table
1. MassCLEAVETM methylation analyses were performed as previously described in reference 48. Data was analyzed using the
MassCLEAVETM technology as previously reported in reference 48.
RNA extraction and quantitative real-time RT-PCR. RNA
and the corresponding cDNA were obtained using standard protocols. Expression was quantified using the ABI PRISM 7900
HT sequence detection system (Applied Biosystems) or the Light
Cycler 2.0 real time PCR system (Roche Diagnostics). The primers used for RT-PCR amplification are listed in Supplemental
Table 2. The reactions were performed in triplicate. Gene expression levels were normalized using 18S determinations.
Chromatin immunoprecipitation (ChIP) assays. ChIP assays
were carried out using the Chromatin Immunoprecipitation
Assay Kit (Upstate Biotechnology) according to the manufacturer’s instructions. The complexes were immunoprecipitated with
antibodies specific for total histone H3 from Abcam (ab1791)
and acetylation of histone H3 (Lys 9) (no. 07-352), dimethylhistone H3 (Lys 4) (no. 07-030), dimethyl-histone H3 (Lys 9)
(no. 07-441), trimethyl-histone H3 (Lys 27) (no. 07-449) from
Upstate Biotechnology. As negative control we used rabbit IgG
serum (Jackson Immunoresearch). The amount of immunoprecipitated target was measured by real-time PCR as described
above. Positive and negative controls for each histone modification were used to set the lower limits.3,6,49 Amplification primers
for gene promoters are listed in Supplemental Table 3. PCRs
were performed in triplicate.
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
www.landesbioscience.com
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgments
R.M. was supported by a FPI fellowship from the Spanish
Ministry of Science and Innovation. This work was supported
by grants from the Spanish Ministry of Science and Innovation
Epigenetics
1147
(SAF2008/1409, CSD2006/49), National Health and Medical
Research Council (NHMRC) and Cancer Institute NSW,
Australia. M.W.C. is supported by the Netherlands Organization
for Scientific Research NWO/ZonMW
(916-10-108) and a Marie Curie grant
(IRG: 248397).
Note
Supplemental materials can be found at:
w w w.landesbioscience.com/journals /
epigenetics/article/16066
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Esteller M. Cancer epigenomics: DNA methylomes
and histone-modification maps. Nat Rev Genet 2007;
8:286-98.
Jones PA, Baylin SB. The epigenomics of cancer. Cell
2007; 128:683-92.
Frigola J, Song J, Stirzaker C, Hinshelwood RA,
Peinado MA, Clark S. Epigenetic remodeling in
colorectal cancer results in coordinate gene suppression
across an entire chromosome band. Nat Genet 2006;
38:540-9.
Clark SJ. Action at a distance: epigenetic silencing of
large chromosomal regions in carcinogenesis. Hum
Mol Genet 2007; 16:88-95.
Hitchins MP, Lin VA, Buckle A, Cheong K, Halani N,
Ku S, et al. Epigenetic inactivation of a cluster of genes
flanking MLH1 in microsatellite-unstable colorectal
cancer. Cancer Res 2007; 67:9107-16.
Rodriguez J, Muñoz M, Vives L, Frangou CG, Groudine
M, Peinado MA. Bivalent domains enforce transcriptional memory of DNA methylated genes in cancer cells.
Proc Natl Acad Sci USA 2008; 105:19809-14.
Novak P, Jensen T, Oshiro MM, Watts GS, Kim CJ,
Futscher BW. Agglomerative epigenetic aberrations are
a common event in human breast cancer. Cancer Res
2008; 68:8616-25.
Dallosso AR, Hancock AL, Szemes M, Moorwood K,
Chilukamarri L, Tsai HH, et al. Frequent long-range
epigenetic silencing of protocadherin gene clusters
on chromosome 5q31 in Wilms’ tumor. PLoS Genet
2009; 5:1000745.
Coolen MW, Stirzaker C, Song JZ, Statham AL, Kassir
Z, Moreno CS, et al. Consolidation of the cancer
genome into domains of repressive chromatin by longrange epigenetic silencing (LRES) reduces transcriptional plasticity. Nat Cell Biol 2010; 12:235-46.
Devaney J, Stirzaker C, Qu W, Song JZ, Statham AL,
Patterson KI, et al. Epigenetic deregulation across
2q14.2 differentiates normal from prostate cancer and
provides a regional part of novel DNA methylation
cancer biomarkers. Cancer Epidemiol Biomarkers Prev
2010; 10.1158/055-9965.EPI-10-0719.
Kouzarides T. Chromatin modifications and their function. Cell 2007; 128:693-705.
Schlesinger Y, Straussman R, Keshet I, Farkash S,
Hecht M, Zimmerman J, et al. Polycomb-mediated
methylation on Lys27 of histone H3 pre-marks genes
for de novo methylation in cancer. Nat Genet 2007;
39:232-6.
Vire E, Brenner C, Deplus R, Blanchon L, Fraga M,
Didelot C, et al. The Polycomb group protein EZH2
directly controls DNA methylation. Nature 2006;
439:871-4.
Ohm JE, McGarvey KM, Yu X, Cheng L, Schuebel
KE, Cope L, et al. A stem cell-like chromatin pattern
may predispose tumor suppressor genes to DNA hypermethylation and heritable silencing. Nat Genet 2007;
39:237-42.
1148
15. McGarvey KM, Van Neste L, Cope L, Ohm JE,
Herman JG, Van Criekinge W, et al. Defining a chromatin pattern that characterizes DNA-hypermethylated
genes in colon cancer cells. Cancer Res 2008; 68:5753-9.
16. Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert
DJ, Cuff J, et al. A bivalent chromatin structure marks
key developmental genes in embryonic stem cells. Cell
2006; 125:315-26.
17. Mikkelsen TS, Ku M, Jaffe DB, Issac B, Lieberman E,
Giannoukos G, et al. Genome-wide maps of chromatin state in pluripotent and lineage-committed cells.
Nature 2007; 448:553-60.
18. Pan G, Tian S, Nie J, Yang C, Ruotti V, Wei H, et al.
Whole-genome analysis of histone H3 lysine 4 and
lysine 27 methylation in human embryonic stem cells.
Cell Stem Cell 2007; 1:299-312.
19. Zhao XD, Han X, Chew JL, Liu J, Chiu KP, Choo A, et
al. Whole-genome mapping of histone H3 Lys4 and 27
trimethylations reveals distinct genomic compartments
in human embryonic stem cells. Cell Stem Cell 2007;
1:286-98.
20. Barski A, Cuddapah S, Cui K, Roh TY, Schones
DE, Wang Z, et al. High-resolution profiling of histone methylations in the human genome. Cell 2007;
129:823-37.
21. Ku M, Koche RP, Rheinbay E, Mendenhall EM, Endoh
M, Mikkelsen TS, et al. Genomewide analysis of PRC1
and PRC2 occupancy identifies two classes of bivalent
domains. PLoS Genet 2008; 4:1000242.
22. Rakyan VK, Down TA, Maslau S, Andrew T, Yang TP,
Beyan H, et al. Human aging-associated DNA hypermethylation occurs preferentially at bivalent chromatin
domains. Genome Res 2010; 20:434-9.
23. Teschendorff AE, Menon U, Gentry-Maharaj A,
Ramus SJ, Weisenberger DJ, Shen H, et al. Agedependent DNA methylation of genes that are suppressed in stem cells is a hallmark of cancer. Genome
Res 2010; 20:440-6.
24. Clarke MF, Fuller M. Stem cells and cancer: two faces
of eve. Cell 2006; 124:1111-5.
25. Ben-Porath I, Thomson MW, Carey VJ, Ge R, Bell
GW, Regev A, et al. An embryonic stem cell-like gene
expression signature in poorly differentiated aggressive
human tumors. Nat Genet 2008; 40:499-507.
26. Spivakov M, Fisher AG. Epigenetic signatures of stemcell identity. Nat Rev Genet 2007; 8:263-71.
27. Sharma S, Kelly TK, Jones PA. Epigenetics in Cancer.
Carcinogenesis 2010; 31:27-36.
28. Komashko VM, Farnham PJ. 5-azacytidine treatment
reorganizes genomic histone modification patterns.
Epigenetics 2010; 5:229-40.
29. Si J, Boumber YA, Shu J, Qin T, Ahmed S, He R, et
al. Chromatin remodeling is required for gene reactivation after decitabine-mediated DNA hypomethylation.
Cancer Res 2010; 70:6968-77.
30. Mayor R, Casadome L, Azuara D, Moreno V, Clark
SJ, Capella G, et al. Long-range epigenetic silencing at
2q14.2 affects most human colorectal cancers and may
have application as a non-invasive biomarker of disease.
Br J Cancer 2009; 100:1534-9.
31. Widschwendter M, Fiegl H, Egle D, Mueller-Holzner
E, Spizzo G, Marth C, et al. Epigenetic stem cell signature in cancer. Nat Genet 2007; 39:157-8.
32. Egger G, Liang G, Aparicio A, Jones PA. Epigenetics
in human disease and prospects for epigenetic therapy.
Nature 2004; 429:457-63.
33. Rhee I, Bachman KE, Park BH, Jair KW, Yen RW,
Schuebel KE, et al. DNMT1 and DNMT3b cooperate
to silence genes in human cancer cells. Nature 2002;
416:552-6.
34. Egger G, Jeong S, Escobar SG, Cortez CC, Li TW,
Saito Y, et al. Identification of DNMT1 (DNA methyltransferase 1) hypomorphs in somatic knockouts
suggests an essential role for DNMT1 in cell survival.
Proc Natl Acad Sci USA 2006; 103:14080-5.
35. Das PM, Singal R. DNA methylation and cancer. J
Clin Oncol 2004; 22:4632-42.
36. Bender CM, Gonzalgo ML, Gonzales FA, Nguyen CT,
Robertson KD, Jones PA. Roles of cell division and
gene transcription in the methylation of CpG islands.
Mol Cell Biol 1999; 19:6690-8.
37. Liang G, Gonzalgo ML, Salem C, Jones PA.
Identification of DNA methylation differences during tumorigenesis by methylation-sensitive arbitrarily
primed polymerase chain reaction. Methods 2002;
27:150-5.
38. Egger G, Aparicio AM, Escobar SG, Jones PA.
Inhibition of histone deacetylation does not block resilencing of p16 after 5-aza-2'-deoxycytidine treatment.
Cancer Res 2007; 67:346-53.
39. Kagey JD, Kapoor-Vazirani P, McCabe MT, Powell
DR, Vertino PM. Long-term stability of demethylation
after transient exposure to 5-aza-2'-deoxycytidine correlates with sustained RNA polymerase II occupancy.
Mol Cancer Res 2010; 8:1048-59.
40. McGarvey KM, Fahrner JA, Greene E, Martens J,
Jenuwein T, Baylin SB. Silenced tumor suppressor
genes reactivated by DNA demethylation do not return
to a fully euchromatic chromatin state. Cancer Res
2006; 66:3541-9.
41. Stirzaker C, Song JZ, Davidson B, Clark SJ.
Transcriptional gene silencing promotes DNA hypermethylation through a sequential change in chromatin modifications in cancer cells. Cancer Res 2004;
64:3871-7.
42. Cameron EE, Bachman KE, Myohanen S, Herman
JG, Baylin SB. Synergy of demethylation and histone
deacetylase inhibition in the re-expression of genes
silenced in cancer. Nat Genet 1999; 21:103-7.
43. Zhu B, Benjamin D, Zheng Y, Angliker H, Thiry S,
Siegmann M, et al. Overexpression of 5-methylcytosine
DNA glycosylase in human embryonic kidney cells
EcR293 demethylates the promoter of a hormoneregulated reporter gene. Proc Natl Acad Sci USA 2001;
98:5031-6.
44. Paz MF, Wei S, Cigudosa JC, Rodriguez-Perales S,
Peinado MA, Huang TH, et al. Genetic unmasking of
epigenetically silenced tumor suppressor genes in colon
cancer cells deficient in DNA methyltransferases. Hum
Mol Genet 2003; 12:2209-19.
45. Miranda TB, Cortez CC, Yoo CB, Liang G, Abe M,
Kelly TK, et al. DZNep is a global histone methylation inhibitor that reactivates developmental genes not
silenced by DNA methylation. Mol Cancer Ther 2009;
8:1579-88.
46. Rhee I, Jair KW, Yen RW, Lengauer C, Herman JG,
Kinzler KW, et al. CpG methylation is maintained in
human cancer cells lacking DNMT1. Nature 2000;
404:1003-7.
47. Ormerod MG. Analysis of cell proliferation using
the bromodeoxyuridine/Hoechst-ethidium bromide
method. Methods Mol Biol 1997; 75:357-65.
48. Coolen MW, Statham AL, Gardiner-Garden M, Clark
SJ. Genomic profiling of CpG methylation and allelic
specificity using quantitative high-throughput mass
spectrometry: critical evaluation and improvements.
Nucleic Acids Res 2007; 35:119.
49. Rodriguez J, Vives L, Jorda M, Morales C, Munoz M,
Vendrell E, et al. Genome-wide tracking of unmethylated DNA Alu repeats in normal and cancer cells.
Nucleic Acids Res 2008; 36:770-84.
©20
1
1L
andesBi
os
c
i
enc
e.
Dono
tdi
s
t
r
i
but
e.
Epigenetics
Volume 6 Issue 9