-
Research Article
DNA Methylation of Telomere-Related Genesand Cancer RiskBrian T.
Joyce1, Yinan Zheng1, Drew Nannini1, Zhou Zhang1, Lei Liu2,Tao
Gao1, Masha Kocherginsky3, Robert Murphy4, Hushan Yang5,Chad J.
Achenbach6, Lewis R. Roberts7, Mirjam Hoxha8, Jincheng Shen9,Pantel
Vokonas10,11, Joel Schwartz12, Andrea Baccarelli13, and Lifang
Hou1
Abstract
Researchers hypothesized that telomere shorteningfacilitates
carcinogenesis. Previous studies found incon-sistent associations
between blood leukocyte telomerelength (LTL) and cancer. Epigenetic
reprogramming oftelomere maintenance mechanisms may help
explainthis inconsistency. We examined associations
betweenDNAmethylation in telomere-related genes (TRG) andcancer. We
analyzed 475 participants providing 889samples 1 to 3 times (median
follow-up, 10.1 years)from 1999 to 2013 in the Normative Aging
Study. Allparticipants were cancer-free at each visit and
bloodleukocytes profiled using the Illumina 450K array. Of121
participants who developed cancer, 34 had prostatecancer, 10
melanoma, 34 unknown skin malignancies,and 43 another cancer. We
examined 2,651 CpGs from80 TRGs and applied a combination of Cox
and mixedmodels to identify CpGs prospectively associated with
cancer (at FDR < 0.05). We also explored trajectories
ofDNAmethylation, logistic regression stratified by time
todiagnosis/censoring, and cross-sectional models of LTLat first
blood draw. We identified 30 CpGs on 23 TRGswhosemethylationwas
positively associatedwith cancerincidence (b ¼ 1.0–6.93) and one
protective CpG inMAD1L1 (b¼�0.65),ofwhich87%were located
inTRGpromoters.Methylation trajectoriesof21CpGs increasedin cancer
cases relative to controls; at 4 to 8 years
pre-diagnosis/censoring, 17 CpGs were positively associatedwith
cancer. ThreeCpGswere cross-sectionally associatedwithLTL.
TRGmethylationmaybeamechanismthroughwhich LTL dynamics reflect
cancer risk. Future researchshould confirm these findings and
explore potentialmechanisms underlying these findings,
includingtelomere maintenance and DNA repair dysfunction.Cancer
Prev Res; 11(8); 511–22. �2018 AACR.
IntroductionTelomeres are tandem TTAGGG nucleotide repeats
that
"cap" the ends of eukaryotic chromosomes and serve tomaintain
genomic stability and limit cellular proliferation(1). Blood
leukocyte telomere length (LTL) shortens withage, and this process
can be accelerated by exposureto environmental risk factors (in
particular those knownto cause oxidative stress and/or chronic
inflammation,two major carcinogenic pathways; ref. 2). Prior
studies
demonstrated that LTL shortening may reflect in situchanges in
telomere length among precancerous andcancerous cells (2) and that
cellular senescence inducedby critical telomere shortening and the
Hayflick limit isgenerally thought tobe a tumor-suppressive
process,whichcancer cells must overcome early in carcinogenesis
(3).However, the exact role of LTL in cancer developmentremains
uncertain. There are numerous studies reportingassociations between
LTL and cancer risk (2), with largelyinconsistent results. These
inconsistencies may be due to
1Center for Population Epigenetics, Robert H. Lurie
Comprehensive CancerCenter and Department of Preventive Medicine,
Northwestern UniversityFeinberg School of Medicine, Chicago,
Illinois. 2Division of Biostatistics,Washington University in St.
Louis, St. Louis, Missouri. 3Department ofPreventive Medicine,
Northwestern University Feinberg School of Medicine,Chicago,
Illinois. 4Center for Global Health, Feinberg School of
Medicine,Northwestern University, Chicago, Illinois. 5Division of
Population Science,Department of Medical Oncology, Sidney Kimmel
Cancer Center, ThomasJefferson University, Philadelphia,
Pennsylvania. 6Department of Medicine,Northwestern University
Feinberg School of Medicine, Chicago, Illinois. 7Divisionof
Gastroenterology and Hepatology, Department of Medicine, Mayo
Clinic,Rochester, Minnesota. 8Molecular Epidemiology and
Environmental EpigeneticsLaboratory, Department of Clinical
Sciences and Community Health, Universit�adegli Studi di Milano,
Milan, Italy. 9Department of Population Health Sciences,University
of Utah School of Medicine, Salt Lake City, Utah. 10VA
Normative
Aging Study, VA Boston Healthcare System, Boston,
Massachusetts.11Department of Medicine, Boston University School of
Medicine, Boston,Massachusetts. 12Department of Environmental
Health, Harvard School ofPublic Health, Boston, Massachusetts.
13Department of Environmental HealthScience, Mailman School of
Public Health, Columbia University, New York,New York.
Note: Supplementary data for this article are available at
Cancer PreventionResearch Online
(http://cancerprevres.aacrjournals.org/).
Corresponding Author: Brian T. Joyce, Northwestern University,
680 N. LakeShore Drive, Suite 1400, Chicago, IL 60611. Phone:
312-503-5407; Fax: 312-908-9588; E-mail:
[email protected]
doi: 10.1158/1940-6207.CAPR-17-0413
�2018 American Association for Cancer Research.
CancerPreventionResearch
www.aacrjournals.org 511
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://crossmark.crossref.org/dialog/?doi=10.1158/1940-6207.CAPR-17-0413&domain=pdf&date_stamp=2018-7-19http://cancerpreventionresearch.aacrjournals.org/
-
differences in study design (e.g., variations in time betweenLTL
measurement and cancer diagnosis) and relativelysparse data from
prospective observational studies. Ourrecent prospective study
found that incident cancer casesexperienced accelerated LTL
shortening until around 4years prior to diagnosis, at which point
their LTL stabilizedrelative to controls (4), suggesting a dynamic
relationshipbetween LTL and cancer development.The underlying
regulatory mechanisms responsible for
the telomere shortening-lengthening balance and its relat-ed
cancer risk are only partially understood at present. Aprevious
study of genetic mutations in telomere-relatedgenes (TRG) found
limited associations with LTL (5). Thismay be because of the low
genetic variability of these genesin human populations (6).
Conversely, a genome-widemeta-analysis identified loci at TRGs
associated with bothLTL and cancer (7). One possible alternative to
a geneticmechanism is epigenetic control of TRGs. In human
stud-ies, LTL has been associated with DNA methylation
insubtelomeric regions and selected loci within TRGs (8)and
repetitive elements Alu and LINE-1 (surrogates
forglobalmethylation; ref. 9). The rate of telomere shorteningover
time was also associated with LINE-1 methylation,suggesting a
time-dependent association between DNAmethylation and telomere
length (9).However, to our knowledge no prior population-based
studies have examined DNA methylation of TRGs inrelation to LTL
dynamics and cancer risk, particularlyin a prospective,
longitudinal setting. In light of ourprior finding of the shift
from accelerated telomereshortening to telomere stabilization prior
to cancer diag-nosis (4), a prospective examination of
epigeneticchanges in TRGs may shed light on the involvement ofLTL
dynamics in carcinogenesis. Thus, our primaryobjective is to assess
whether blood DNA methylationin TRGs is prospectively associated
with cancer risk.Our secondary objective is to explore whether
DNAmethylation of cancer-associated CpG sites on TRGs isassociated
with LTL.
Materials and MethodsStudy populationThe Normative Aging Study
(NAS) was established in
1963 by the U.S. Department of Veterans Affairs to assessthe
determinants of healthy aging in an initial cohort of2,280 men.
Eligibility criteria included being between theages of 21 and 80,
veteran status, living in the Boston area,and having no history of
chronic health conditions (car-diovascular disease, cancer, etc.).
Participants returned forclinical examinations every 3 to 5 years,
and starting in1999, these examinations included a 7-mL blood draw
forgenetic and epigenetic analysis. From enrollment to 1999,981
participants died and 470 were lost to follow-up(primarily by
moving away from the Boston area);descriptive analysis previously
found no differences in
characteristics between either of these subgroups and the829
participants remaining as of 1999 (4).Between January 1, 1999, and
December 31, 2013, 802
of 829 (96.7%) active participants consented to blooddonation
(median follow-up time, 10.1 years). Of these,686 were randomly
selected for whole-epigenome profil-ing using the Illumina Infinium
HumanMethylation450BeadChip array, and 491 were cancer-free at the
time oftheir first methylation measurement. To minimize
con-founding due to genetic ancestry, we excluded 16 partici-pants
of non-white race, leaving 889 observations of 475participants for
analysis. In total, 157 (33%) participantshad data from one blood
draw, 222 (47%) participantsfrom two blood draws, and 96 (20%)
subjects from threeblood draws. Among this final set, 121 cases
developedcancer (34 prostate, 34 unspecified skin malignancies,
10melanomas, 8 lung, 5 bladder, 4 colorectal, 26 others) and354
participants remained cancer-free for our entire fol-low-up.
Information on medical history obtained fromquestionnaires was
confirmed via blinded medical recordreview and included cancer
diagnoses and comorbidities.We identified TRGs using a PubMed
literature search for
genes linked to telomere maintenance, elongation, andrepair
(5–7, 10–29). This resulted in 80 TRGs (Table 1)containing 2,651
CpG sites available in our dataset, whichwe list with accompanying
annotation information(and mean/SD methylation at the first blood
draw) in
Table 1. Number of CpGs in each gene of interest by pathway
Helicase Repair OtherBLM 17 ATM 59 ACYP2 26DDX1 10 BTBD12 28
BHMT 15DDX11 19 DCLRE1C 20 BICD1 29PIF1 15 DDB1 17 C17orf68 21RECQL
14 FEN1 25 CLPTM1L 53RECQL4 19 HMBOX1 24 CXCR4 26RECQL5 54 MRE11A
21 DCAF4 24WRN 41 MSH2 14 DCLRE1B 16
Shelterin NBN 10 EHMT2 177ACD 31 PARP3 19 MAD1L1 731POT1 15 PCNA
26 MCM4 14RAP1A 17 PML 31 MEN1 24TERF1 12 RAD50 14 MPHOSPH6 15TERF2
15 RAD51 18 MTR 22TERF2IP 17 RAD51AP1 16 MTRR 20TINF2 15 RAD51C 15
MYC 37
Telomerase RAD51L1 78 NAF1 18DKC1 22 RAD51L3 15 OBFC1 18GAR1 14
RAD54L 13 PARP1 19NHP2 19 SIRT1 17 PARP2 9NOP10 11 SIRT6 17 PIK3C3
11TEP1 13 SMC5 9 PINX1 30TERC 9 SMC6 16 PRKDC 37TERT 100 TP53BP1 30
PRMT8 35WRAP53 34 XRCC6 20 PXK 21
RTEL1 33SIP1 9TNKS 28TNKS2 16UCP2 13ZNF208 9ZNF676 1
Joyce et al.
Cancer Prev Res; 11(8) August 2018 Cancer Prevention
Research512
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
Supplementary Table S1. For ease of presentation, we
alsoclassified genes [based on Mirabello and colleagues' work(5) or
literature review and GeneCard search] into one offive
telomere-related pathways: Helicase, Shelterin, Telo-merase,
Repair, or Other.
Telomere measurementLaboratory methods for measuring LTL in the
NAS have
been described previously (4). In brief, LTL was measuredusing
quantitative qPCR. Relative LTL was measured bytaking the ratio of
the telomere (T) repeat copy number tosingle-copy gene (S) copy
number (T:S ratio) in a givensample and reported as relative units
expressing the ratiobetween test DNA LTL and reference pooled DNA
LTL. Thelatter was created using DNA from 475 participants
(400ng/sample) and used to generate a fresh standard curvefrom 0.25
to 20 ng/mL in every T and S qPCR run. Allsampleswere run in
triplicate, and the average of the three Tmeasurements was divided
by the average of the three Smeasurements to calculate the average
T:S ratio. The intra-assay coefficient of variation for the T/S
ratiowas 8.1%. Theaverage coefficient of variation for the T
reaction was 8%,and for the S reaction 5.6%. When the coefficient
ofvariation for the T or S reactions was higher than 15%,the
measurement was repeated.
DNA methylation measurementBuffy coat DNA was isolated from each
sample via the
QIAamp DNA Blood Kit (QIAGEN) and a 0.5 mg aliquotwas bisulfite
converted with the EZ-96 DNA MethylationKit (Zymo Research). In the
NAS, this was done on bloodcollected between 1999 and 2007. DNA
methylation wassubsequently detected by the Infinium
HumanMethyla-tion450 BeadChip platform (Northwestern
University,Feinberg School of Medicine, Center for Genetic
Medicine,Chicago, IL). Technical effects due to the plate/chip
wereminimized by utilizing a two-stage age-stratified algorithmto
randomize the samples, thereby ensuring comparableage distribution
across plates/chips.Quality control samples consistedof replicate
pairs anda
single sample that was run within and between plates/chips to
help detect batch effects. Analytic plates were runconsecutively,
by the same technician, and processed andread on the same scanner.
Quality control approaches alsoincluded the detection and removal
of 15 DNA samplesand 949 probes via the pfilter command in the
Biocon-ductor wateRmelon package, which excluded DNA sam-ples
containing >1% of probes with detection P values>0.05 and
probes having >1% of samples with detection Pvalue >0.05
(after omitting samples excluded above).Furthermore, we also
excluded probes with specific designand/or annotation, namely 65
with genotyping function,3,091 used for detecting CpH methylation,
and 3,688containing an SNP in the last 10 bases with a minor
allelefrequency greater than 0.01 in the CEU reference set. Anumber
of these probeswere already excludedby thepfilter
command, so after these steps,wefinally obtained 477,927probes
(i.e., �98.4% out of 485,512), which were used toobtain DNA
methylation. Finally, we applied a 3-part,preprocessing pipeline to
our data: (i) background correc-tion via the out-of-band (noob)
method by Triche andcolleagues (30); (ii) dye-bias adjustment by
the Biocon-ductor methylumi package (31); and (iii) probe-type
cor-rection with BMIQ according to Teschendorff and collea-gues
(2013; ref. 32), as provided by wateRmelon (33).
Statistical analysisFor descriptive analyses, we performed c2 or
Kruskal–
Wallis tests to assess differences in participant
character-istics at the first methylationmeasurement by cancer
status(patients who would later develop cancer during the
studyperiod vs. those who remained cancer-free throughout).We next
used a joint model under the shared randomeffects model framework
(reduced method by Liu andHang; ref. 34) to combine our repeated
methylation mea-sures (linear mixed model) and time to cancer
diagnosisdata (Cox proportional hazards model) and to
examineassociations between cancer incidence and DNA methyl-ation
of all 2,651 CpG sites of interest.This method was designed as an
extension of the shared
random effects model and uses a Gaussian quadraturetechnique
with a piecewise constant baseline hazard toapproximate the
baseline hazard in a Cox model, whileincorporating repeatedmeasures
as with amixedmodel. Atraditional approach to evaluating
longitudinal biomar-kers with time to event data is to use observed
values as atime-varying covariate in a Cox proportional
hazardsmodel. However, this requires a complete set of
repeatedmeasures in a time-continuous process, whereas in
reality,our biomarkers of interest are measured only at
discretetime points, generally not including the time of
eventoccurrence (35). Although the value of the biomarker atevent
time can be obtained by, for example, last observa-tion carried
forward (LOCF), this practice could be crudeand lead to
inappropriate inferences, especially when thetime interval
betweenbiomarkermeasurement anddiseaseoutcome is long (35).
Furthermore patient survival toevent occurrence might depend on the
"underlying true"(or expected) values of biomarkers, rather than
theobserved valueswithmeasurement errors; in this situation,a
traditional model would be biased toward the null (35).Thus, we
used a joint model of longitudinal biomarkers
and survival. Our model accounts for selection bias by therandom
effects shared between the mixed model of meth-ylation markers and
survival model for time to event.Rather than LOCF, the missing
methylation measures atthe event time can be imputed by empirical
Bayes estimate(posterior expected value of random effects
conditional onthe observed data) from the mixed model, based on
theobserved history of individuals who did not have an eventup to
that time. Also, the "underlying true" (expected)biomarkers, rather
than the observed values accompanying
Telomere Gene Methylation and Cancer
www.aacrjournals.org Cancer Prev Res; 11(8) August 2018 513
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
measurement errors, are incorporated in the survival mod-el,
which address the "biased toward the null" concern.This model is
designed to maximize statistical power andminimize bias in the
analysis of correlated repeated mea-sures (e.g., DNAmethylation
data) with time to an event asthe outcome, without making
assumptions regarding thedata structure ormissingness. Themodel
failed to convergefor 37 of the 2,651 CpG sites (1.4%), which were
excludedfrom analysis. We used the Benjamini–Hochberg FDR tocorrect
for all of the remaining 2,614 tests and report CpGsites with FDR 8
years).All methylation values were standardized to have a stan-dard
deviation equal to 1 for this analysis. For participantswith
multiple observations within the same stratum, weused the first
observation from each subject only. We alsoexplored cross-sectional
associations betweenmethylationat each of these significant CpG
sites and LTL, both mea-sured at the first blood draw only and
restricted to subjectswho were cancer-free for the entire follow-up
to minimizepotential confounding by age- and cancer-related
factors.All of the above analyses were conducted using SAS v.
9.4(SAS Institute) and adjusted for age, BMI, education,smoking
status and pack-years, alcohol consumption,blood cell type
abundances (CD8, CD4, natural killer, Bcells, and monocytes; ref.
36), and five principal compo-nents (previously calculated to
represent 95% of DNAprocessing batch effects), all based on our
prior workstudying DNA methylation in this cohort (37).
Bioinformatic analysisFinally, we performed a regulatory
enrichment analysis
of the 31 cancer-associated CpG sites using R v. 3.4.0. Weused
DNase I hypersensitivity sites (DNase), transcriptionfactor–binding
sites (TFBS), and annotations of histonemodification ChIP peaks
pooled across cell lines (dataavailable in the ENCODE Analysis Hub
at the EuropeanBioinformatics Institute). For each regulatory
element, we
then calculated the number of overlapping CpGs amongthe 31
significant CpGs (observed) and 10,000 sets ofrandomly selected
CpGs across the genome (expected).We calculated the ratio of
observed tomean expected as theenrichment fold and obtained an
empirical P value fromthe distribution of the expected in the
background.
ResultsTable 2 shows the characteristics of all participants at
the
first blood draw by cancer status. Briefly, participants whowere
cancer-free for the full follow-up were slightly olderthan those
who later developed cancer. Our descriptiveanalysis identified no
other significant differences in par-ticipant characteristics
across cancer status. Table 3 showsthe results of the
jointmodel,with31CpGsites on23TRGsassociated with cancer incidence
at FDR
-
available upon request). Generally speaking, CpGs on thesame
gene as our primary findings tended to be associatedwith cancer
incidence in the same direction as the primaryfinding, albeit not
to the same degree of statistical signif-icance. In the unadjusted
correlation analysis, methylationof significantCpG sites tended
tobe significantly correlatedwith one another (0.3–0.6 for almost
all CpG sites; dataavailable upon request) despite their disparate
locations inthe genome.For the trajectory analyses, overall we
found significant
differences in methylation over time by cancer status in 21of 31
cancer-associated CpG sites including both of theCpGs in each of
the Helicase, Shelterin, and Telomerasepathways (Supplementary
Table S3). Figure 1 plots thetrajectory analyses of DNA methylation
over time at selectnoteworthy CpG sites by cancer status. CpGs on
TINF2 (animportant telomere-regulating gene) as well as PIF1 (in
theHelicase pathway) and the DNA repair genes DDB1 andPARP3 (Fig.
1A–C and E, respectively) showed strongtrends with higher
methylation in subjects developingcancer, generally beginning
around 4 to 6 years prediag-nosis/censoring. Conversely, CpGs on
DKC1 in the Telo-merase pathway as well as MYC (Fig. 1D and F,
respec-tively) showed few differences between cancer cases
andcancer-free subjects, and no clear temporal trend. FormanyCpG
sites, methylation trajectories between cancer casesand cancer-free
participants began to diverge as early as 6 to
8 years prior to diagnosis/censoring, with clear trendsvisible
for most cancer-CpG sites beginning 4 years pre-diagnosis (see
Supplementary Fig. S2 for correspondingfigures with 95% CIs added;
Supplementary Fig. S3 con-tains figures for the remaining 25 CpG
sites).Table 4 shows the results of the logistic regression
analysis of DNA methylation and later cancer status at 0to 4 and
4 to 8 years prediagnosis/censoring. In the stratumof 0 to 4 years
prediagnosis/censoring, we found 11 CpGsites associated with cancer
incidence: one CpG in each ofthe Shelterin (TINF2) and Telomerase
(WRAP53) path-ways, one CpG on each of three TRGs in the DNA
Repairpathway (PARP3, FEN1, and SIRT6) and four CpGs on afourth
(DDB1), and one CpG on each of CLPTM1L andMAD1L1. In the stratum of
4 to 8 years prediagnosis/censoring, we found 17 CpG sites
associated with cancerincidence: one CpG in each of the Helicase
(RECQL4),Shelterin (ACD), and Telomerase (DKC1) pathways; eightCpGs
on six TRGs in the DNA repair pathway (DCLRE1C,DDB1,MSH2, PARP3,
RAD51L3, and SIRT6); andoneCpGon each of BICD1,
CLPTM1L,MAD1L1,MTRR, RTEL1, andSIP1. Methylation at four CpG sites
(on CLPTM1L, DDB1,PARP3, and SIRT6) was associated with incident
cancer inboth time strata. Supplementary Table S4 shows the
logis-tic regression results in samples collectedmore than 8
yearsprediagnosis/censoring; oneCpGonPARP3was associatedwith cancer
incidence.
Table 3. Cancer-associated CpGs in TRGs by pathway at FDR <
0.05Pathway Gene CpG Region Island ba 95% CI FDR
Helicase PIF1 cg11013726 50UTR Island 1.00 0.57–1.43 0.02RECQL4
cg17368874 TSS200 Island 6.67 3.13–10.20 0.04
Shelterin ACD cg04265926 TSS1500 Island 6.67 3.09–10.25
0.04TINF2 cg02271180 1stExon Island 1.99 0.88–3.10 0.05
Telomerase DKC1 cg19944582 TSS200 Island 5.80 2.74–8.85
0.04WRAP53 cg25053252 TSS1500 Island 5.47 2.82–8.12 0.03
Repair BTBD12 cg04157159 TSS200 Island 4.25 1.92–6.58
0.04DCLRE1C cg14369264 TSS1500 Island 1.11 0.53–1.69 0.04DCLRE1C
cg24866702 TSS200 Island 6.53 3.12–9.95 0.04DCLRE1C cg04785461
TSS200 Island 5.38 2.37–8.40 0.05DDB1 cg23053918 1stExon Island
5.45 2.75–8.15 0.03DDB1 cg20772347 TSS200 Island 5.68 2.65–8.72
0.04DDB1 cg24840365 TSS200 Island 5.49 2.55–8.43 0.04DDB1
cg25530631 1stExon Island 6.63 3.04–10.22 0.04DDB1 cg08724919
1stExon Island 1.45 0.64–2.26 0.05FEN1 cg25628257 TSS200 Island
3.95 2.03–5.87 0.03HMBOX1 cg14143435 TSS200 N_Shore 1.41 0.73–2.10
0.03MSH2 cg00547758 50UTR Island 6.23 2.97–9.48 0.04PARP3
cg14974841 TSS1500 Island 5.22 2.49–7.95 0.04PARP3 cg14262432
TSS200 Island 6.93 3.00–10.86 0.05RAD51L3 cg19223675 TSS200 S_Shore
5.16 2.27–8.05 0.05RAD54L cg24955114 TSS1500 OpenSea 6.16 2.67–9.66
0.05SIRT6 cg15034464 50UTR Island 5.11 2.78–7.44 0.03
Other BICD1 cg21587861 TSS200 Island 6.93 3.12–10.75 0.04CLPTM1L
cg19739264 1stExon Island 4.89 2.37–7.41 0.04MAD1L1 cg09776772 Body
OpenSea -0.65 �0.97 to �0.33 0.03MAD1L1 cg13247668 TSS200 Island
5.00 2.30–7.71 0.04MTRR cg26627933 1stExon Island 5.45 2.73–8.16
0.03MYC cg07871324 TSS1500 Island 5.81 2.91–8.71 0.03RTEL1
cg27236539 TSS200 Island 1.25 0.56–1.94 0.04SIP1 cg15533434 TSS200
Island 5.28 2.48–8.09 0.04
aBeta coefficients represent the average difference in
methylation (M-value) between cases and controls.
Telomere Gene Methylation and Cancer
www.aacrjournals.org Cancer Prev Res; 11(8) August 2018 515
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
A B C
D E F
13
-5.9
Met
hyla
tion
leve
l (M
)cg02271180 in TINF2 (1stExon, Shelterin) cg11013726 in PIF1
(5’UTR, Helicase)
cg19944582 in DKC1 (TSS200, Telomerase)
-5.8
-5.7
-5.6
-5.5
-5.4
-5.3
12 11 10 9 8 7 6 5 4 3 2 1
13
-5.9
Met
hyla
tion
leve
l (M
)
Time to cancer diagnosis or censoring
Cancer-free
Note: Supplementary Table S2 shows corresponding tabular
results; Supplementary fig. S2 shows results with 95% confidence
intervals.
Incident cancer : P < 0.05 : P < 0.01
-5.8
-5.7
-5.6
-5.5
-5.4
12 11 10 9 8 7 6 5 4 3 2 1 13
-6.3
Met
hyla
tion
leve
l (M
)
Time to cancer diagnosis or censoring
-6.2
-6.1
-6.0
-5.9
12 11 10 9 8 7 6 5 4 3 2 1 13
-6.1
Met
hyla
tion
leve
l (M
)
Time to cancer diagnosis or censoring
-6.0
-5.9
-5.8
12 11 10 9 8 7 6 5 4 3 2 1
13
-5.0
Met
hyla
tion
leve
l (M
)
cg14974841 in PARP3 (TSS1500, Repair)
-4.5
-4.0
-3.5
-3.0
12 11 10 9 8 7 6 5 4 3 2 1
cg08724919 in DDB1 (1stExon, Repair)
13
-6.4
Met
hyla
tion
leve
l (M
)
cg07871324 in MYC (TSS1500, Other)
-6.2
-6.0
-5.8
-5.6
12 11 10 9 8 7 6 5 4 3 2 1
Figure 1.
DNA methylation by years to cancer diagnosis/censoring and
cancer status for select CpG sites.
Table 4. Logistic regression results stratified by time interval
between blood draw and diagnosis/censoring
0–
-
Table 5 shows the results of our cross-sectionalmodels ofLTL on
DNAmethylation. DNA methylation of three CpGsites, all of them on
DNA repair genes, was positivelyassociated with LTL at the first
blood draw: cg24866702onDCLRE1C, cg00547758 onMSH2, and cg25530631
onDDB1.We found no other significant associations betweenDNA
methylation of TRGs and LTL. Finally, Supplemen-tary Fig. S4 shows
the results of our regulatory elementenrichment analysis. Five
histone modifications (notablyH3K27ac, H3K4me2, H3K4me3, H3K79me2,
andH3K9ac) were significantly enriched at the CpG sites
sig-nificantly associated with cancer (all P < 0.001).
Supple-mentary Table S5 contains more detailed tabular
findings.
DiscussionTo our knowledge, this is the first study to identify
DNA
methylation changes in TRGs that are prospectively asso-ciated
with cancer. In this cohort, we identified positiveassociations
between cancer incidence and methylation at30 CpG sites (and one
negative association), most in genepromoter regions, on 23 genes
related to telomere main-tenance and regulation. Over time,
methylation of 21 CpGsites began to diverge by later cancer status
several yearsprior to diagnosis/censoring. In general, cancer cases
expe-rienced increased static methylation and cancer-free
parti-
cipants experienced decreased methylation. Furthermore,our
logistic regression identified 11 and17CpG siteswheremethylation at
0 to 4 years and 4 to 8 years prediagnosis/censoring, respectively,
was associated with cancer inci-dence (including four CpGs in both
strata). Finally, inparticipants who remained cancer free, at the
first blooddraw, DNA methylation at three CpG sites was
associatedwith telomere length. Few studies have examined
thesegenes as potential blood-based cancer biomarkers; thus,our
findings should be validated in other populations.Nonetheless,
these findings suggest mechanisms throughwhich cancer cells may be
able to alter telomere homeo-stasis, possibly as a precursor to
clinical disease, thusindicating DNA methylation of TRGs as a
potentiallyuseful biomarker of cancer.We identified methylation of
CpG sites (cg19944582
and cg25053252) in the promoters of two genes (DKC1and WRAP53)
involved with telomerase, a well-character-ized telomere
maintenance pathway, as positively associ-ated with cancer. The two
genes involved in the telomerasepathway,DKC1 andWRAP53, jointly
promote telomeraseexpression and telomeremaintenance.Mutations
ofDKC1were identified in cancer cells (38), as was
promoterhypermethylation ofDKC1 (39). Further evidence suggeststhat
reductions in DKC1 expression may increase cancersusceptibility
through nontelomere mechanisms, such as
Table 5. Associations between cancer-associated CpG sites and
telomere length at first blood draw (N ¼ 346)CpG Gene Pathway b 95%
CI P
cg11013726 PIF1 Helicase 0.03 �0.07–0.12 0.56cg17368874 RECQL4
Helicase 0.06 �0.02–0.14 0.14cg04265926 ACD Shelterin 0.05
�0.10–0.19 0.52cg02271180 TINF2 Shelterin �0.02 �0.20–0.17
0.87cg19944582 DKC1 Telomerase �0.13 �0.33–0.08 0.23cg25053252
WRAP53 Telomerase 0.07 �0.20–0.34 0.60cg04157159 BTBD12 Repair 0.02
�0.13–0.18 0.76cg04785461 DCLRE1C Repair �0.01 �0.12–0.10
0.88cg14369264 DCLRE1C Repair 0.04 �0.04–0.13 0.31cg24866702
DCLRE1C Repair 0.25 0.08–0.41
-
reduced p53 expression (40), which may partially explainthe lack
of association with telomere length in our cross-sectional
analysis. DKC1 downregulation has also beenassociated with exposure
to arsenic, a known carcinogen(41). Similarly, reduced WRAP53
expression was associ-ated with cancer prognosis (42). Methylation
of thesegenes may thus be involved with cancer risk and/or
pro-gression independent of telomere length.Our study also
identified methylation of two CpGs
(cg04265926 and cg02271180) in the promoters of twogenes (ACD
and TINF2) in the shelterin pathway, anotherwell-characterized
telomere maintenance pathway, as pos-itively associated with
cancer. Changes in shelterin com-plex expression have been
implicated in a variety of cancertypes, including germlinemutations
in bothACD (43) andTINF2 (44). However, limited evidence exists to
supportthis hypothesis as previous DNA methylation studies
oftelomerase-associated genes tended to focus on TERT.However,
studies of TERTmethylation in blood leukocytesfound no associations
with cancer (45), concordant withourfindings.One possible
explanation is that the normallystrict regulatory control of
TERTmay be preserved in bloodleukocytes even in participants
experiencing carcinogene-sis, suggesting that future studies of
blood DNA methyl-ation should focus on other shelterin complex
genes.Among other DNA repair genes, we identified methyl-
ation at multiple loci within the promoters of three
genes(PARP3, DCLRE1C, and DDB1) as positively associatedwith
cancer. A prior study of cancer samples found down-regulation of
bothPARP3 andDCLRE1C in cancer cells andwas additionally associated
with telomerase reactivation(13). Downregulation of DCLRE1C has
also been associ-ated with chronic exposure to ionizing radiation
(46).Furthermore, methylation at one of the significant loci oneach
ofDCLRE1C andDDB1was also positively associatedwith telomere length
at the first blood draw. Thus, epige-netic repression of these DNA
repair genes may be onemechanism through which cancer cells can
activate telo-mere maintenance mechanisms.In our prior examination
of LTL and cancer incidence in
this same cohort, we identified cancer-associated acceler-ated
LTL shortening that stabilized starting approximately4 years prior
to diagnosis (4). We observed that higherDNA methylation at CpG
sites on 15 TRGs of interest inthis study was associated with
cancer status 4 to 8 yearsprior to diagnosis (Table 4). In
addition, we observedsignificantly different methylation
trajectories betweencancer cases and cancer-free participants in 21
sites on17 genes. Examples of this divergence can be seenwith
fourCpGs (cg02271180 in TINF2, cg11013726 in PIF1,cg08724919 in
DDB1, and cg14974841 in PARP3)in Fig. 1. In all of these cases, DNA
methylation began tosignificantly differ between cases and controls
beginning atleast 4 years prediagnosis/censoring. Finally, three
CpGs(cg24866702 on DCLRE1C, cg25530631 on DDB1, and
cg00547758 on MSH2) were positively associated withtelomere
length in our cross-sectional analysis. These find-ings all
occurred prior to (or at the same time as) the shift inLTL change
that our previous study observed. Our trajec-tory analyses suggest
that increased DNA methylation atthese and other sites may be an
early event in the devel-opment of cancer, either reflecting
constitutive exposuresthat also increase cancer risk or correlating
with DNAmethylation changes occurring in cancer cells, that
remainsdetectable for years. Also of note, methylation of
bothcg24866702 and cg00547758 was associated with bothtelomere
length at the first blood draw andwith cancer risk4 to 8 years
prior to diagnosis (but not 0–4 years prior).This suggests that
thesemethylation changes occur prior toour previously observed
change in telomere length andmay be involved in driving this change
via a DNA repair-related mechanism. As studies of the relationship
betweenLTLmeasured at a single time point and cancer risk
remainunclear (47), alterations in DNA methylation of theseTRGs may
help explain the between-study differences(e.g., differences in the
timing of LTL measurement relativeto cancer diagnosis). Together,
our results suggest thatstudying DNA methylation in blood
leukocytes is promis-ing for future research into the role of
dynamic changes intelomere length during cancer development, and
that incor-porating epigenetic data may help improve the utility
oftelomere length in blood leukocytes as a cancer
biomarker.Finally, although we lacked gene expression data, we
were able to identify enrichment of numerous importantregulatory
elements in the set of CpGs associated withcancer. These
includeH2A.Z, TFBS, andDNase, whichmayall point toward a role of
methylation of these CpGs in cis-regulatory changes and potential
transcriptional activation(consistent with most of the CpGs being
located in genepromoter regions). We also identified five
activating his-tones andone repressive histone in associationwith
our setof CpGs at P < 0.001. The repressive histone, H3K9me1,has
been previously found to have altered levels in somecancers (48).
Similarly, the activating histone markerH3K27ac has been found to
be dysregulated in cancer(49). Expression levels of some of these
histones havepreviously been associated with DNA methylation
(50).Taken together, these findings bolster our conclusion thatthe
identifiedCpG sites in these important TRGsmay affectgene
expression.This study is subject to limitations. Although the
longi-
tudinal nature of our study design allowed us to exploreaspects
of the temporal associations between DNA meth-ylation of TRGs, LTL,
and current cancer risk, it remainschallenging to accommodate a
formal mediation analysisof longitudinal mechanisms. Our
conclusions regardingthe interplay of DNA methylation and LTL in
carcinogen-esis over time thus require confirmation in
additionalprospective studies. In addition, the study population
ofthe NAS is not representative, and thus, more diverse
Joyce et al.
Cancer Prev Res; 11(8) August 2018 Cancer Prevention
Research518
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
populations should be studied to validate our findings,although
there is little reason to believe these mechanismswould
substantially vary by gender or by race. Further-more, the sample
sizes of most specific cancer types in theNAS were too small to
permit a statistically rigorousexploration of these associations
for individual cancers.Although we hypothesized that dysregulation
of telomeremaintenance mechanisms are a general mechanism
affect-ing many different cancer types, this should also be
testedin larger studies. Similarly, the relatively small number
ofcases coupled with our time stratification limited thesample size
for each time stratum analyzed. This may haveresulted in false
negatives, which may explain some asso-ciations that were
significant 4 to 8 years prediagnosis butnot 0 to 4 years.
Alternatively, this discrepancymay reflect asubset of cancer-free
subjects who developed cancer afterthe end of our study period (and
were thus effectivelymisclassified). Validation with a longer
follow-up and/orlarger study population would be necessary to test
thesepossible explanations. The dearth of significant associa-tions
between DNA methylation and LTL in our cross-sectional model may
also be a consequence of reducedsample size; the dynamic natures of
LTL and methylation(and their potential relationships with one
another andwith cancer) limited us to a cross-sectional model.
Thus,our findings may represent a lack of a biological effect or
alack of statistical power and should be interpreted withcaution
until they can be validated. Similarly, the NASdataset lacked gene
expression data to provide functionalverification of our findings
posited above. Future researchshould verify the relationship
between DNA methylationand expression of these specific
genes.Nonetheless, this study provides an important, poten-
tiallymechanistic explanation for thedynamic relationshipbetween
LTL and cancer that we previously observed.Future studies should
confirm and explore these CpG sitesand genes as potential early
detection biomarkers andtherapeutic targets; the strong
correlations between mostCpG sites in our analysis (despite their
disparate locationson the genome) further bolster the possibility
of theseCpGs collectively making a biomarker in the future.
Futureresearch in larger, more diverse populations should focuson
examining changes in the DNA methylation of theseTRGs in termsof
gene expression, LTL, and cancer to furtherelucidate the temporal
sequence of these events and theirpotential role in mechanisms of
carcinogenesis. DNAmethylation of TRGs could be an important early
eventin carcinogenesis and, with appropriate confirmation,could
have extremely valuable clinical applications forcancer. These DNA
methylation changes in blood leuko-cytes may have been induced by
environmental exposures(pollutants, nutrients, etc.) acting
constitutionally; thus,our findings may provide important
information on onepossible mechanism of action for previously
identifiedcarcinogenic exposures. Future research should
explore
this possibility by examining potential exposure–methyl-ation
relationships in the geneswe identified. Furthermore,if epigenetic
changes in these genes do influence the lengthof cancer cells'
telomeres, therapeutically targeting thesechanges could
theoretically induce cellular senescence incancer cells and thus
provide a new effective, safe, andtargeted therapy for cancer.
However, for this to happen,future studies will need to validate
the epigenetic changeswe have identified in blood both in cancer
and in normalhealthy tissue. Nonetheless, these findings provide
impor-tant information for future cancer early detection,
preven-tion, and treatment. This may be particularly true in
popu-lations with underlying immune dysfunction or
chronicinflammation (e.g., chronic HIV infection,
autoimmunedisorders). Exploring these pathwaysmay also facilitate
theuse of cancer immunotherapies to correct immune dys-function and
cancer-specific immune responses.
Disclosure of Potential Conflicts of InterestM. Kocherginsky has
provided expert testimony for The University
of Chicago. No potential conflicts of interest were disclosed by
theother authors.
Authors' ContributionsConception and design: B.T. Joyce, Y.
Zheng, J. Schwartz, L. HouDevelopment of methodology: B.T. Joyce,
L. Liu, M. Hoxha, L. HouAcquisition of data (provided animals,
acquired and managedpatients, provided facilities, etc.):M.Hoxha,
P. Vokonas, J. Schwartz,A. BaccarelliAnalysis and interpretation of
data (e.g., statistical analysis, bio-statistics, computational
analysis): B.T. Joyce, Y. Zheng, Z. Zhang,L. Liu, M. Kocherginsky,
R. Murphy, J. ShenWriting, review, and/or revisionof
themanuscript:B.T. Joyce, Y. Zheng,D. Nannini, Z. Zhang, L. Liu, T.
Gao, M. Kocherginsky, R. Murphy,H. Yang, C.J. Achenbach, L.R.
Roberts, J. Shen, J. Schwartz, L. HouAdministrative, technical, or
material support (i.e., reporting ororganizing data, constructing
databases): B.T. Joyce, R. Murphy,J. Shen, P. VokonasStudy
supervision: B.T. Joyce, L. Liu, P. Vokonas, L. Hou
AcknowledgmentsThe Normative Aging Study is supported by the
Epidemiology
Research and Information Center of the U.S. Department of
VeteransAffairs (NIEHS R01- ES015172) and is a research component
of theMassachusetts Veterans EpidemiologyResearch and
InformationCen-ter (MAVERIC). L. Hou received additional support
from the North-western University Robert H. Lurie Comprehensive
Cancer CenterRosenberg Research Fund. L. Hou, R. Murphy, and L.R.
Roberts alsoreceived support from the NCI: 1U54CA221205-01 and
D43TW009575. A. Baccarelli and J. Schwartz received additional
supportfrom the National Institute of Environmental Health
Sciences;
NIEHSR01-ES021733,NIEHSR01-ES015172,NIEHS-R01ES025225,NIEHSP30-ES009089,
and NIEHS P30-ES00002.
The costs of publication of this article were defrayed in part
by thepayment of page charges. This articlemust therefore be
herebymarkedadvertisement in accordance with 18 U.S.C. Section 1734
solely toindicate this fact.
Received December 15, 2017; revised April 3, 2018; accepted
May22, 2018; published first June 12, 2018.
Telomere Gene Methylation and Cancer
www.aacrjournals.org Cancer Prev Res; 11(8) August 2018 519
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
References1. Frias C, Pampalona J, Genesca A, Tusell L. Telomere
dysfunction
and genome instability. Front Biosci (Landmark Ed)
2012;17:2181–96.
2. Hou L, Zhang X, Gawron AJ, Liu J. Surrogate tissue
telomerelength and cancer risk: shorter or longer? Cancer Lett
2012;319:130–5.
3. Giaimo S, d'Adda di Fagagna F. Is cellular senescence an
exampleof antagonistic pleiotropy? Aging Cell 2012;11:378–83.
4. Hou L, Joyce BT, Gao T, Liu L, Zheng Y, Penedo FJ, et al.
Bloodtelomere length attrition and cancer development in the
norma-tive aging study cohort. EBioMedicine 2015;2:591–6.
5. Mirabello L, Yu K, Kraft P, De Vivo I, Hunter DJ, Prescott J,
et al.The association of telomere length and genetic variation
intelomere biology genes. Hum Mutat 2010;31:1050–8.
6. Mirabello L, YeagerM, Chowdhury S, Qi L, Deng X,Wang Z, et
al.Worldwide genetic structure in 37 genes important in
telomerebiology. Heredity 2012;108:124–33.
7. Codd V, Nelson CP, Albrecht E, ManginoM, Deelen J, Buxton
JL,et al. Identification of seven loci affecting mean telomere
lengthand their association with disease. Nat Genet
2013;45:422–7.
8. Buxton JL, Suderman M, Pappas JJ, Borghol N, McArdle
W,Blakemore AI, et al. Human leukocyte telomere length is
associ-ated with DNA methylation levels in multiple subtelomeric
andimprinted loci. Sci Rep 2014;4:4954.
9. Wong JY, De Vivo I, Lin X, Grashow R, Cavallari J, Christiani
DC.The association between global DNA methylation and
telomerelength in a longitudinal study of boilermakers. Genet
Epidemiol2014;38:254–64.
10. Dilley RL, Verma P, Cho NW, Winters HD, Wondisford
AR,Greenberg RA. Break-induced telomere synthesis underlies
alter-native telomere maintenance. Nature 2016;539:54–8.
11. Dimitrova N, Chen YC, Spector DL, de Lange T. 53BP1
promotesnon-homologous end joining of telomeres by increasing
chro-matin mobility. Nature 2008;456:524–8.
12. Feng X, Luo Z, Jiang S, Li F, Han X, Hu Y, et al. The
telomere-associated homeobox-containing protein TAH1/HMBOX1
parti-cipates in telomere maintenance in ALT cells. J Cell Sci
2013;126(Pt 17):3982–9.
13. Frias C, Garcia-Aranda C, De Juan C, Moran A, Ortega P,
GomezA, et al. Telomere shortening is associated with poor
prognosisand telomerase activity correlateswithDNA repair
impairment innon-small cell lung cancer. Lung Cancer
2008;60:416–25.
14. Jia G, Su L, Singhal S, Liu X. Emerging roles of SIRT6 on
telomeremaintenance, DNA repair, metabolism and mammalian aging.Mol
Cell Biochem 2012;364:345–50.
15. Karami S, Han Y, Pande M, Cheng I, Rudd J, Pierce BL, et
al.Telomere structure andmaintenance gene variants and risk
offivecancer types. Int J Cancer 2016;139:2655–70.
16. Kim S, Bi X, Czarny-Ratajczak M, Dai J, Welsh DA, Myers L,
et al.Telomere maintenance genes SIRT1 and XRCC6 impact age-related
decline in telomere length but only SIRT1 is associatedwith human
longevity. Biogerontology 2012;13:119–31.
17. Kim S, Parks CG, Xu Z, Carswell G, DeRoo LA, Sandler DP, et
al.Association between genetic variants in DNA and histone
meth-ylation and telomere length. PLoS One 2012;7:e40504.
18. Lee JH, Cheng R, Honig LS, Feitosa M, Kammerer CM, Kang
MS,et al. Genome wide association and linkage analyses
identifiedthree loci-4q25, 17q23.2, and 10q11.21-associated with
varia-tion in leukocyte telomere length: the Long Life Family
Study.Front Genet 2013;4:310.
19. Levy D, Neuhausen SL, Hunt SC, Kimura M, Hwang SJ, Chen W,et
al. Genome-wide association identifies OBFC1 as a locus
involved in human leukocyte telomere biology. Proc Natl AcadSci
U S A 2010;107:9293–8.
20. Mangino M, Christiansen L, Stone R, Hunt SC, Horvath
K,Eisenberg DT, et al. DCAF4, a novel gene associated with
leuco-cyte telomere length. J Med Genet 2015;52:157–62.
21. MarchesiniM,Matocci R, Tasselli L, Cambiaghi V, Orleth A,
FuriaL, et al. PML is required for telomere stability in
non-neoplastichuman cells. Oncogene 2016;35:1811–21.
22. Mendez-Bermudez A, Royle NJ. Deficiency in DNA
mismatchrepair increases the rate of telomere shortening in normal
humancells. Hum Mutat 2011;32:939–46.
23. Mirabello L, Richards EG, Duong LM, Yu K, Wang Z, Cawthon
R,et al. Telomere length and variation in telomere biology genes
inindividuals with osteosarcoma. Int J Mol Epidemiol
Genet2011;2:19–29.
24. Pooley KA, Bojesen SE, Weischer M, Nielsen SF, Thompson
D,AminAlOlamaA, et al. A genome-wide association scan (GWAS)formean
telomere lengthwithin theCOGSproject: identified locishow little
association with hormone-related cancer risk. HumMol Genet
2013;22:5056–64.
25. Potts PR, Yu H. The SMC5/6 complex maintains telomere
lengthin ALT cancer cells through SUMOylation of
telomere-bindingproteins. Nat Struct Mol Biol 2007;14:581–90.
26. Saharia A, Teasley DC, Duxin JP, Dao B, Chiappinelli KB,
StewartSA. FEN1ensures telomere stability by facilitating
replication forkre-initiation. J Biol Chem 2010;285:27057–66.
27. Stern JL, Zyner KG, Pickett HA, Cohen SB, Bryan TM.
Telomeraserecruitment requires both TCAB1 and Cajal bodies
independent-ly. Mol Cell Biol 2012;32:2384–95.
28. Wilson JS, Tejera AM, Castor D, Toth R, Blasco MA, Rouse
J.Localization-dependent and -independent roles of SLX4 in
reg-ulating telomeres. Cell Rep 2013;4:853–60.
29. Zee RY, Ridker PM, Chasman DI. Genetic variants in
eleventelomere-associated genes and the risk of incident
cardio/cere-brovascular disease: The Women's Genome Health Study.
ClinChim Acta 2011;412:199–202.
30. Triche TJ Jr, Weisenberger DJ, Van Den Berg D, Laird PW,
Sieg-mund KD. Low-level processing of Illumina Infinium
DNAMethylation BeadArrays. Nucleic Acids Res 2013;41:e90.
31. Davis S, Du P, Bilke S, Triche T Jr, Bootwalla M.
methylumi:Handle Illuminamethylation data. 2015. Available from:
https://rdrr.io/bioc/methylumi/.
32. Teschendorff AE, Marabita F, Lechner M, Bartlett T, Tegner
J,Gomez-Cabrero D, et al. A beta-mixture quantile normaliza-tion
method for correcting probe design bias in IlluminaInfinium 450 k
DNA methylation data. Bioinformatics 2013;29:189–96.
33. Pidsley R, CC YW, Volta M, Lunnon K, Mill J, Schalkwyk LC.
Adata-driven approach to preprocessing Illumina 450K methyla-tion
array data. BMC Genomics 2013;14:293.
34. Liu L,HuangX. Joint analysis of correlated repeatedmeasures
andrecurrent events processes in the presence of death, with
appli-cation to a study on acquired immune deficiency syndrome. J
RStat Soc Ser C Appl Stat 2009;58:65–81.
35. Liu Y, Liu L. Joint models for longitudinal data and
time-to-eventoccurrence. Routledge International Handbook of
AdvancedQuantitative Methods in Nursing Research. Taylor and
FrancisInc., New York, NY; 2015.
36. Houseman EA, Accomando WP, Koestler DC, Christensen
BC,Marsit CJ, NelsonHH, et al. DNAmethylation arrays as
surrogatemeasures of cell mixture distribution. BMC Bioinformatics
2012;13:86.
Joyce et al.
Cancer Prev Res; 11(8) August 2018 Cancer Prevention
Research520
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
https://rdrr.io/bioc/methylumi/https://rdrr.io/bioc/methylumi/http://cancerpreventionresearch.aacrjournals.org/
-
37. Zheng Y, Joyce BT, Colicino E, Liu L, ZhangW, Dai Q, et al.
Bloodepigenetic age may predict cancer incidence and mortality.
EBio-Medicine 2016;5:68–73.
38. Bellodi C, Krasnykh O, Haynes N, Theodoropoulou M, Peng
G,Montanaro L, et al. Loss of function of the tumor suppressorDKC1
perturbs p27 translation control and contributes to pitu-itary
tumorigenesis. Cancer Res 2010;70:6026–35.
39. Smith IM, Mithani SK, Mydlarz WK, Chang SS, Califano
JA.Inactivation of the tumor suppressor genes causing the
hereditarysyndromes predisposing to head and neck cancer via
promoterhypermethylation in sporadic head and neck cancers. ORL
JOtorhinolaryngol Relat Spec 2010;72:44–50.
40. Montanaro L. Dyskerin and cancer: more than telomerase.
Thedefect in mRNA translation helps in explaining how a
prolifer-ative defect leads to cancer. J Pathol 2010;222:345–9.
41. Gao J, RoyS, TongL,ArgosM, Jasmine F,RahamanR, et al.
Arsenicexposure, telomere length, and expression of
telomere-relatedgenes among Bangladeshi individuals. Environ Res
2015;136:462–9.
42. ZhangH,WangDW,Adell G, Sun XF.WRAP53 is an
independentprognostic factor in rectal cancer- a study of Swedish
clinical trialof preoperative radiotherapy in rectal cancer
patients. BMCCancer 2012;12:294.
43. Aoude LG, Pritchard AL, Robles-Espinoza CD, Wadt K,
HarlandM, Choi J, et al. Nonsense mutations in the shelterin
complexgenes ACD and TERF2IP in familial melanoma. J Natl Cancer
Inst2015;107:pii:dju408.
44. Hartmann K, Illing A, Leithauser F, Baisantry A,
Quintanilla-Martinez L, Rudolph KL. Gene dosage reductions of Trf1
and/or Tin2 induce telomere DNA damage and lymphoma formationin
aging mice. Leukemia 2016;30:749–53.
45. Hoxha M, Fabris S, Agnelli L, Bollati V, Cutrona G, Matis S,
et al.Relevance of telomere/telomerase system impairment in
earlystage chronic lymphocytic leukemia. Genes Chromosomes Can-cer
2014;53:612–21.
46. Fachin AL, Mello SS, Sandrin-Garcia P, Junta CM,
Ghilardi-NettoT, Donadi EA, et al. Gene expression profiles in
radiation workersoccupationally exposed to ionizing radiation. J
Radiat Res2009;50:61–71.
47. Zhang X, Zhao Q, Zhu W, Liu T, Xie SH, Zhong LX, et al.
Theassociation of telomere length in peripheral blood cells
withcancer risk: a systematic review and meta-analysis of
prospec-tive studies. Cancer Epidemiol Biomarkers Prev
2017;26:1381–90.
48. Ellinger J, Bachmann A, Goke F, Behbahani TE, Baumann
C,Heukamp LC, et al. Alterations of global histone H3K9 andH3K27
methylation levels in bladder cancer. Urol Int 2014;93:113–8.
49. Song Y, Li ZX, Liu X, Wang R, Li LW, Zhang Q. The
Wnt/beta-catenin and PI3K/Akt signaling pathways promote EMT in
gastriccancer by epigenetic regulation via H3 lysine 27
acetylation.Tumour Biol 2017;39:1010428317712617.
50. Okitsu CY, Hsieh CL. DNA methylation dictates histone
H3K4methylation. Mol Cell Biol 2007;27:2746–57.
www.aacrjournals.org Cancer Prev Res; 11(8) August 2018 521
Telomere Gene Methylation and Cancer
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/
-
2018;11:511-522. Published OnlineFirst June 12, 2018.Cancer Prev
Res Brian T. Joyce, Yinan Zheng, Drew Nannini, et al. DNA
Methylation of Telomere-Related Genes and Cancer Risk
Updated version
10.1158/1940-6207.CAPR-17-0413doi:
Access the most recent version of this article at:
Material
Supplementary
1
http://cancerpreventionresearch.aacrjournals.org/content/suppl/2018/06/12/1940-6207.CAPR-17-0413.DCAccess
the most recent supplemental material at:
Cited articles
http://cancerpreventionresearch.aacrjournals.org/content/11/8/511.full#ref-list-1
This article cites 48 articles, 8 of which you can access for
free at:
E-mail alerts related to this article or journal.Sign up to
receive free email-alerts
Subscriptions
Reprints and
[email protected]
To order reprints of this article or to subscribe to the
journal, contact the AACR Publications Department at
Permissions
Rightslink site. Click on "Request Permissions" which will take
you to the Copyright Clearance Center's (CCC)
.http://cancerpreventionresearch.aacrjournals.org/content/11/8/511To
request permission to re-use all or part of this article, use this
link
Research. on April 10, 2020. © 2018 American Association for
Cancercancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst June 12, 2018; DOI:
10.1158/1940-6207.CAPR-17-0413
http://cancerpreventionresearch.aacrjournals.org/lookup/doi/10.1158/1940-6207.CAPR-17-0413http://cancerpreventionresearch.aacrjournals.org/content/suppl/2018/06/12/1940-6207.CAPR-17-0413.DC1http://cancerpreventionresearch.aacrjournals.org/content/suppl/2018/06/12/1940-6207.CAPR-17-0413.DC1http://cancerpreventionresearch.aacrjournals.org/content/11/8/511.full#ref-list-1http://cancerpreventionresearch.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerpreventionresearch.aacrjournals.org/content/11/8/511http://cancerpreventionresearch.aacrjournals.org/
/ColorImageDict > /JPEG2000ColorACSImageDict >
/JPEG2000ColorImageDict > /AntiAliasGrayImages false
/CropGrayImages false /GrayImageMinResolution 200
/GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true
/GrayImageDownsampleType /Bicubic /GrayImageResolution 300
/GrayImageDepth -1 /GrayImageMinDownsampleDepth 2
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict >
/GrayImageDict > /JPEG2000GrayACSImageDict >
/JPEG2000GrayImageDict > /AntiAliasMonoImages false
/CropMonoImages false /MonoImageMinResolution 600
/MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true
/MonoImageDownsampleType /Bicubic /MonoImageResolution 900
/MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000
/EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode
/MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None
] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false
/PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000
0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true
/PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier ()
/PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped
/False
/CreateJDFFile false /Description > /Namespace [ (Adobe)
(Common) (1.0) ] /OtherNamespaces [ > /FormElements false
/GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks
false /IncludeInteractive false /IncludeLayers false
/IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000
/MultimediaHandling /UseObjectSettings /Namespace [ (Adobe)
(CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA
/PageMarksFile /RomanDefault /PreserveEditing true
/UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling
/LeaveUntagged /UseDocumentBleed false >> > ]>>
setdistillerparams> setpagedevice