Download Implication of a Chromosome 15q15.2 Locus in

Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Cancer
Research
Prevention and Epidemiology
Implication of a Chromosome 15q15.2 Locus in
Regulating UBR1 and Predisposing Smokers to
MGMT Methylation in Lung
Shuguang Leng1, Guodong Wu1, Leonard B. Collins2, Cynthia L. Thomas1, Carmen S. Tellez1,
Andrew R. Jauregui3, Maria A. Picchi1, Xiequn Zhang1, Daniel E. Juri1, Dhimant Desai4,
Shantu G. Amin4, Richard E. Crowell5, Christine A. Stidley5,Yushi Liu1, James A. Swenberg2,
Yong Lin1, Marc G. Wathelet3, Frank D. Gilliland6, and Steven A. Belinsky1
Abstract
O6-Methylguanine-DNA methyltransferase (MGMT) is a
DNA repair enzyme that protects cells from carcinogenic effects
of alkylating agents; however, MGMT is silenced by promoter
hypermethylation during carcinogenesis. A single-nucleotide
polymorphism (SNP) in an enhancer in the MGMT promoter
was previously identified to be highly significantly associated
with risk for MGMT methylation in lung cancer and sputum
from smokers. To further genetic investigations, a genome-wide
association and replication study was conducted in two smoker
cohorts to identify novel loci for MGMT methylation in sputum
that were independent of the MGMT enhancer polymorphism.
Two novel trans-acting loci (15q15.2 and 17q24.3) that were
identified acted together with the enhancer SNP to empower
risk prediction for MGMT methylation. We found that the
predisposition to MGMT methylation arising from the
15q15.2 locus involved regulation of the ubiquitin protein
ligase E3 component UBR1. UBR1 attenuation reduced turnover of MGMT protein and increased repair of O6-methylguanine in nitrosomethylurea-treated human bronchial epithelial
cells, while also reducing MGMT promoter activity and abolishing MGMT induction. Overall, our results substantiate
reduced gene transcription as a major mechanism for predisposition to MGMT methylation in the lungs of smokers, and
support the importance of UBR1 in regulating MGMT homeostasis and DNA repair of alkylated DNA adducts in cells. Cancer
Introduction
some (1, 2). The recovery of MGMT activity after its inactivation
with alkylating agents is slow and results entirely from de novo
protein synthesis (3–5).
The predominant mechanism for MGMT gene inactivation in
carcinogenesis is hypermethylation of a CpG island within its
promoter–enhancer region that leads to transcriptional silencing
in lymphoma and several solid tumors, including brain, colon,
lung, and head and neck cancers (6). Silencing of this gene is
associated with an increased prevalence for G:C to A:T transition
mutations in K-ras and p53 genes in colorectal and brain tumors
respectively, consistent with the important role for this gene in
removing the highly promutagenic adduct O6-methylguanine
(O6MG; refs. 7 and 8). Although alkylating adducts such as
O6MG can be derived from endogenous methylation by S-adenosylmethionine, knockouts of MGMT failed to show enhanced
tumor formation unless treated with exogenous alkylating carcinogens (9). MGMT methylation is a common event in lung
adenocarcinomas and associated with tumor progression (10).
Concomitant methylation of a panel of cancer-relevant genes,
including MGMT, detected in lung epithelial cells present in
sputum predicts risk for subsequent lung cancer incidence in
moderate and heavy smokers (11, 12). In addition, MGMT
methylation is also a prognostic biomarker for response of glioblastoma patients to the alkylating agent temozolomide (13, 14).
Genome-wide profiling of allele-specific methylation (ASM)
using a methylation-sensitive single nucleotide polymorphism
(SNP) array identified >35,000 ASM sites across the genomes of
multiple tissue types that were likely mediated by parental-origin
6
O -Methylguanine-DNA methyltransferase (MGMT) is a DNA
repair enzyme that protects cells from carcinogenic effects of
alkylating agents by removing adducts from the O6 position of
guanine (1). This repair mechanism is conserved in all organisms
and involves transfer of methyl and other alkyl groups from the
O6 position of guanine in DNA to a cysteine residue within
enzymatic active site of MGMT protein. The irreversible binding
of the alkyl group to the MGMT protein functionally inactivates
the enzyme and leads to a structural alteration resulting in the
recognition by ubiquitin ligases and degradation by the protea1
Lung Cancer Program, Lovelace Respiratory Research Institute, Albuquerque, New Mexico. 2Department of Environmental Sciences and
Engineering, University of North Carolina at Chapel Hill, Chapel Hill,
North Carolina. 3Lung Fibrosis Program, Lovelace Respiratory
Research Institute, Albuquerque, New Mexico. 4Department of Pharmacology, Penn State College of Medicine, Hershey, Pennsylvania.
5
Department of Internal Medicine, University of New Mexico, Albuquerque, New Mexico. 6Keck School of Medicine, University of Southern California, Los Angeles, California.
Note: Supplementary data for this article are available at Cancer Research
Online (http://cancerres.aacrjournals.org/).
Corresponding Author: Steven A. Belinsky, Lung Cancer Program, Lovelace
Respiratory Research Institute, Albuquerque, NM 87108. Phone: 505-348-9465;
Fax: 505-348-4990; E-mail: [email protected]
doi: 10.1158/0008-5472.CAN-15-0243
2015 American Association for Cancer Research.
Res; 75(15); 3108–17. 2015 AACR.
3108 Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Chromosome 15q15.2 for MGMT Promoter Hypermethylation
(imprinting) or short distance cis-acting SNPs (15–18). This effort
was further expanded by running methylation quantitative trait
loci (meQTL) analyses that associated genome-wide SNP array
data with methylation array data to identify short and long
distance cis- and trans-acting germline meQTL in normal tissues
that included lung, adipose, peripheral blood lymphocyte, and
brain (19–25). In normal lung, 34,304 cis- and 585 trans-acting
meQTL were identified and enriched in CCTC-binding factor
(CTCF)-binding sites, DNase I hypersensitivity regions, and histone marks (19). Four of five established lung cancer risk loci
among persons with European ancestry are cis-acting meQTL (19).
The findings of SNP/haplotype dependent ASM and meQTL
contribute to understanding genetic-epigenetic interactions and
also provide functional relevance for disease risk loci discovered
from genome-wide association studies (GWAS).
In addition to germline meQTL, somatic meQTL were also
identified that affected the predisposition of cancer-relevant genes
for promoter hypermethylation acquired in carcinogenesis for
multiple tumor types (13, 26–32). This is illustrated by our
studies and others for MGMT methylation. An enhancer SNP
rs16906252 located at the first exon–intron boundary of MGMT
was highly significantly associated with risk for MGMT methylation in colorectal cancer, pleural mesothelioma, lung adenocarcinoma, glioblastoma, and diffuse large B-cell lymphoma
(13, 27, 29–32). Significant association was also seen for MGMT
methylation in sputum from smokers (27). Functional characterization confirmed the ASM pattern in heterozygotes for
rs16906252 with the "A" allele preferentially silenced by methylation in primary lung tumors and lung cancer cell lines (27).
Although rs16906252 is a major determinant for acquisition of
MGMT methylation in lung carcinogenesis, 20% of current and
former smokers carrying wild homozygote (G/G) of rs16906252
are methylated for MGMT in their sputum (27), suggesting that
other genetic or environmental factors may contribute to MGMT
methylation. To address this issue, a GWAS discovery and replication study was conducted in two smoker cohorts from New
Mexico to identify novel loci whose association with risk for
MGMT methylation detected in sputum was independent of the
enhancer SNP. Subsequent hypothesis-driven studies were
designed based on the genetic associations detected to explore
the underlying biological mechanisms.
Sputum collection and measurement of MGMT promoter
methylation
Induced and spontaneous morning sputum samples were
collected in LSC and VSC, respectively. Sputum samples were
washed multiple times using Saccomanno's fixative prior to longterm storage at 80 C. Sputum adequacy defined as the presence
of deep lung macrophages or Curschmann's spiral (12) was
assessed by cytotechnologists and subjects with cancer cells
detected in sputum slides were removed from this study. Sputum
DNA was isolated and bisulfite converted. Methylation status for
CpGs around the transcriptional start site of MGMT was assessed
using two stage methylation-specific PCR (33). Methylation status
was scored as 0 (unmethylated) or 1 (methylated). Our assay can
reproducibly detect one methylated allele in a background of
10,000 unmethylated alleles (12).
Materials and Methods
Imputation of chromosome 15
Imputation of chromosome 15 in the LSC was conducted
using BEAGLE (version 3.3.2) with phased haplotype data of
EUR populations (n for chromosomes ¼ 758, EUR.chr15.phase1_release_v3.20101123) from 1,000 genomes project pilot 1
study as the reference panel. Masked analyses on 20% of SNPs on
chromosome 15 identified a high correlation of the observed
versus expected allelic frequencies (Pearson correlation coefficient ¼ 0.98). The estimated allelic dosages for SNPs (n ¼
165,451) with dosage R2 0.3 and MAF 0.05 in the LSC and
EUR reference populations were included for assessing genetic
association with risk for gene methylation.
Study populations
Longitudinal cohorts that enrolled current and former smokers
were used for the GWAS discovery (the Lovelace smokers cohort,
LSC) and the replication (the veterans smokers cohort, VSC). The
LSC began recruitment of female smokers in 2001 and expanded
to include male smokers in 2004. Enrollment is restricted to
current and former smokers ages 40 to 74 years with a minimum
of 15 pack-years of smoking. The VSC began recruitment of
smokers in 2000 using enrollment criteria and procedures similar
to the LSC. Most participants in the VSC are males with greater
than 20 pack-years smoking history. In total, 1,675 lung cancerfree subjects (1,200 subjects from LSC and 430 subjects from VSC)
were included in this study. A detailed description of subject
enrollment and follow has been provided previously (27). All
samples were collected with informed consent. The study was
approved by the Western Institutional Review Board and New
Mexico VA Health Care System.
www.aacrjournals.org
GWAS genotyping and quality assurance
The HumanOmni2.5-4v1-H BeadChip (Illumina) was used to
genotype 2,450,000 SNPs in 1200 Caucasian smokers from the
LSC. We removed 37 subjects due to low call rate (<95%, n ¼ 7),
low heterozygosity (n ¼ 1), low Caucasian ancestry (<85%, n ¼ 2),
and high relatedness with other samples (n ¼ 27). Furthermore,
SNPs were excluded if they had a call rate of <90%, a minor allele
frequency (MAF) <0.008, or P < 108 for Hardy–Weinberg equilibrium test, or were on Y or pseudo-autosomal region of X
chromosomes. The MAF cutoff is a technical one to identify at
least 20 heterozygotes for accurate genotype clustering required
by GenomeStudio. After quality assessment, 1,163 subjects with
1,599,980 SNPs remained in the genetic association analysis
(Supplementary Table S1). The genomic coordinate is based on
hg18 unless mentioned otherwise.
SNP selection for replication
Four SNPs (rs72887860, rs62287262, rs16957091, and
rs997781) associated with risk for MGMT methylation in LSC
(P < 105 in models with adjustment for rs16906252) and with
MAF > 0.1 were selected for replication using the VSC (n ¼ 430;
Supplementary Table S2). Rs16957091 was genotyped as the tag
SNPs at chromosome 15q15.2 locus for replication. These SNPs
were genotyped using the TaqMan assay (Life Technologies) in the
VSC.
Enrichment of SNPs in ENCODE-annotated regulatory regions
The enrichment of SNPs in ENCODE-annotated regulatory
regions within the 472-kb haplotype block at chromosome
15q15.2 locus was assessed using FunciSNP (34). Peak files
included CTCF-binding sites, DNase I hypersensitivity locations,
Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
3109
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Leng et al.
DNase digital genomic foot printing, and regions enriched in
active (H3K4me3 and H3K36me3) and repressive (H3K27me3)
histone modification markers from small airway epithelial cells
(SAEC). Known FAIRE-Seq (Formaldehyde-Assisted Isolation of
Regulatory Elements), DNase I hypersensitivity location, and gene
promoter were included as well.
Stable siRNA-mediated knockdown of UBR1 in human BEC
lines
Stable UBR1 knockdown (KD) cell lines were established
using three immortalized human bronchial epithelial cell lines
(HBEC1, HBEC4, and HBEC26; ref. 35). The pSUPER.retro.
puro (Oligo Engine) was modified to encode resistance to
hygromycin (pSUPER.retro.hyg). A pair of oligonucleotides
were annealed and inserted into the pSUPER.retro.hyg. The
two oligonucleotide sequences are as follows: 50 -GATCCCCGGATCGGAATCTATTAAGATTCAAGAGATCTTAATAGATTCCGATCCTTTTTA-30 and 5 0 -AGCTTAAAAAGGATCGGAATCTATTAAGATCTCTTGAATCTTAATAGATTCCGATCCGGG-3 0 ).
The resulting construct uses the H1 RNA promoter to express
shRNAs that are converted by endogenous enzymes to
siRNAs targeting UBR1 mRNA. These constructs were verified by sequencing and transfected in GP2-293 cells along
with a plasmid encoding the VSV-G protein. The resultant
pseudotyped retroviral particles were used to transduce
HBECs and drug resistant cells or clones were obtained after
selection with hygromycin. Control cells or clones were
established following the same procedure except pSUPER.
retro.hyg plasmid backbone was used. On average, 70% to
80% KD was achieved at the mRNA level in all HBECs.
Western blots suggested that minimal residual UBR1 protein
remained in KD HBECs compared with control lines (Fig.
1A, not shown).
Promoter activity assays
A 1,528-bp fragment containing the entire promoter and CpG
island of MGMT with different haplotype alleles was previously
cloned into the pGL2-basic Luciferase Reporter Vector (Promega)
upstream of the luciferase coding sequence (27). Luciferase
reporter constructs that contain HAP1 (wild haplotype allele) or
HAP4 (haplotype allele carrying A allele of rs16906252) were
transiently transfected in control HBEC4 and HBEC4 with UBR1
KD using Lipofectamine 3000. Transfection in HBEC1 and
HBEC26 was conducted using the Neon electroporation system.
Cells were harvested 48 hours posttransfection and reporter
activity was measured using the Dual Luciferase Assay System
(Promega).
Real-time PCR for quantifying gene expression
RNA was isolated from primary human BEC cultures (n ¼
54), normal lungs (n ¼ 6), and cell lines. TaqMan real-time PCR
was conducted to quantify gene expression in cDNA using the
delta threshold cycle method with b-actin as the endogenous
control.
Western blot and protein expression quantification
Western blot was conducted to quantify protein levels of
UBR1 and MGMT using antibodies from Abcam (anti-UBR1
antibody, ab42420) and Invitrogen (mouse anti-MGMT,
35-7000). Protein extract (90 mg) was digested with 100 ng
endoproteinase GluC in a 25 mL reaction mixture before being
resolved in PAGE gels (36). Na€ve and suicide MGMT protein
levels were quantified by densitometry (Image Lab; Bio-Rad) of
the individual bands.
Quantification of O6MG
HBECs were treated with 1 mmol/L nitrosomethylurea (MNU)
in the dark for 30 minutes. Cells were harvested at 0, 24, 48, and
72 hours post-MNU treatment. O6MG was quantified in genomic
DNA using HPLC-MS (37). The average amount recovered for 50
fmol positive controls ([15N5-13CD3]-O6-methyl-20 -deoxyguanosine) was 47 8 fmol (n ¼ 4). The limit of detection was 1 fmol,
whereas limit of quantitation was 2.5 fmol per injection. Genomic
O6MG level was expressed as the ratio between the amount of
O6MG and dG within each DNA sample. The percentage of
Figure 1.
Effect of UBR1 KD on MGMT turnover
and DNA repair. A, Western blot for
UBR1 KD in HBEC1. Stable transfection
with pSUPER.retro.hyg expressing
siRNA targeting UBR1 led to almost
undetectable UBR1 protein in HBEC1. B,
the ratios of na€ve versus suicide MGMT
were higher in control HBECs
compared with UBR1 KD lines (0.95 0.29 vs. 0.71 0.30, paired t test P ¼
0.037), with largest difference seen in
HBEC26 (0.36). C, Western blot for
MGMT protein in control HBEC26 and
UBR1 KD line with and without GluC
digest. D, DNA repair for O6MG is
slower in control HBECs than UBR1 KD
lines (percentage of O6MG left
unrepaired at 24 hours post-MNU
treatment; 0.22 0.0049 vs. 0.11 0.02, paired t test, P ¼ 0.011). Results
are summarized with mean and
standard deviation based on
experiments from three HBEC lines.
3110 Cancer Res; 75(15) August 1, 2015
Cancer Research
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Chromosome 15q15.2 for MGMT Promoter Hypermethylation
O6MG left unrepaired at 24 hours post-MNU treatment was
further standardized by taking into account the cell growth rate.
The cytokinesis-block micronucleus assay
Phytohemagglutinin-stimulated exponentially growing lymphocyte cultures established from LSC members (n ¼ 107) were
treated with 1 mmol/L MNU for 4 hours and then incubated with
4.5 mg/mL cytochalasin B for an additional 24 hours before
harvest (38). The dose selected was within the linear dose–
response range and caused obvious genotoxicity, but minimal
cytotoxicity. The micronuclei and nucleoplasmic bridge were
scored to allow the calculation of the percentage of cells with
chromosomal damages.
Statistical analysis
In the GWAS discovery, the genetic association was assessed in
1,163 subjects using logistic regression with adjustment for age
and sex. The genotype of rs16906252 was further included for
covariate adjustment in models that would identify novel SNPs
whose association with risk for MGMT methylation was independent of rs16906252. SNPs were assessed under an additive
genetic model. Odds ratio (OR) and 95% confidence intervals
(95% CI) were calculated to quantify the magnitude of the
association per allele.
In the VSC replication, logistic regression was applied to assess
the association between each individual SNP and risk for MGMT
methylation in 430 subjects with adjustment for age, ethnicity
(Hispanic versus non-Hispanic White), and rs16906252. The
inclusion of Hispanics (n ¼ 141) in the analysis is mainly due
to the sample size consideration. However, the inclusion of both
ethnicities should not affect our results because the prevalence of
MGMT methylation and the distribution of the four SNPs are not
different between the two major ethnicities in the VSC. Results
from the two stages were combined by a meta-analysis using the
inverse variance weighting method (39).
Generalized linear models (GLM) were used to assess the
associations between SNPs under the additive inheritance model
and gene expression (delta Ct relative to b-actin) in primary
human BECs and normal lung tissues; between UBR1 KD status,
promoter haplotype, and MGMT promoter activity in HBEC4;
between UBR1 KD status and induction of MGMT expression
upon MNU treatment; and between rs16857091 and log10
transformed percentage of cells with chromosomal damage. The
GLM included design variables and variables associated with each
of the outcomes (Supplementary Table S3). The paired t test was
used to compare the difference of the ratio of na€ve versus suicide
MGMT and O6MG between UBR1 KD and control cell lines. All
statistical tests were two-sided. Statistical analyses were conducted
in SAS 9.2, R 2.14, and PLINK 1.06.
Table 1. Associations between four SNPs and risk for MGMT methylation in LSC
LSC
SNP
Location
Allelea
MAFb
OR (95% CI)
rs72887860
Chr3:61999431
G/A
0.14
1.87 (1.42–2.47)
rs62287262
Chr4:6601119
A/G
0.40
0.58 (0.47–0.72)
rs16957091
Chr15:40804718
T/G
0.23
1.72 (1.36–2.17)
rs997781
Chr17:64629283
G/A
0.29
1.64 (1.33–2.03)
Results
GWAS discovery in LSC
Comparison of the observed and expected distribution of
association estimates calculated for the GWAS discovery indicated
no evidence of population stratification (inflation factor ¼ 1.003;
Supplementary Fig. S1). GWAS discovery was conducted in 1,163
LSC members with methylation of MGMT detected in 296 subjects (25.5%). Thirty-six common SNPs (MAF > 0.1) were associated with risk for MGMT methylation with P < 105 (Supplementary Table S2). Enhancer SNP rs16906252 (not included on
the HumanOmni2.5-4v1-H BeadChip) previously shown to be
highly significantly associated with risk for MGMT methylation
was genotyped in the 1,163 LSC members and the association
clearly reached genome wide significance (P ¼ 4.3 1031). SNP
rs16906252 was included in the logistic regression model to allow
for identification of novel SNPs with associations independent of
rs16906252 and this identified 12 common SNPs with P < 105
(Supplementary Table S2). SNPs located around the MGMT
promoter became nonsignificant (Ps > 0.94; Supplementary
Table S2 and Fig. S2), suggesting the associations seen for these
SNPs reflected some degree of linkage disequilibrium (LD) with
rs16906252. However, multiple associations around chromosome 15q15.2 locus remained statistically significant and thus
likely independent of rs16906252.
Replication in VSC
Four SNPs (rs72887860, rs62287262, rs16957091, and
rs997781) associated with risk for MGMT methylation in LSC
with P < 105 and MAF > 0.1 were selected for replication in 430
members from the VSC with MGMT methylation detected in 107
subjects (24.9%). Rs16957091 was genotyped as the tag SNP on
chromosome 15q15.2. Associations were replicated on chromosome 15q15.2 (rs16957091, P ¼ 0.016) and 17q24.3 loci
(rs997781, P ¼ 0.016). Meta-analysis combining the results for
these two SNPs from LSC and VSC identified associations of near
genome-wide significance (Table 1). Data from the LSC and VSC
cohorts were pooled to assess change in the area under the curve
(AUC) of the receiver operating characteristic (ROC) by adding
rs16957091, rs997781, and rs16906252 to the basic model containing age, sex, race, and cohort identifier only (Supplementary
Table S4). Adding all three SNPs into model increased the ROC
AUC by 19.6% (from 54.2% to 73.8%, P ¼ 2.97 1057).
Association based on imputed genotypes at chromosome
17q24.3 and 15q15.2 loci
Rs997781 is located within a 131-kb haplotype block
(chr17:64593108-64724190, hg18) on chromosome 17q24.3 containing ABCA6 and ABCA10 (not shown). Rs16957091 is located
within a 472-kb haplotype block (chr15:40709948-41181608,
and VSC
Pc
7.5 106
1.4 106
4.5 106
4.2 106
VSC
OR (95% CI)
0.91 (0.57–1.45)
1.20 (0.84–1.73)
1.63 (1.10–2.41)
1.57 (1.09–2.25)
Pd
0.69
0.32
0.016
0.016
Meta-analysis
OR (95% CI)
1.70 (1.39–2.07)
1.63 (1.36–1.95)
P
2.5 107
2.1 107
a
Second allele is the minor allele.
MAF was assessed in LSC.
c
Logistic regression with adjustment for age, sex, and rs16906252.
d
Logistic regression with adjustment for age, ethnicity (Hispanic vs. non-Hispanic White), and rs16906252.
b
www.aacrjournals.org
Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
3111
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Leng et al.
Figure 2.
Imputation of chromosome 15q15.2.
Imputation was conducted using
phased genotype data of EUR
populations from 1,000 Genomes
project. Genetic associations are
plotted for 941 variants (MAF 0.05)
within the region of Chr15: 40304748–
41304718 (hg18) that include 758
2
imputed variants (dosage R is 0.3).
Circles, genotyped SNPs; squares,
imputed SNPs; purple circle,
rs16957091. Degree of LD with
rs16957091 is schemed as the gradient
of red to dark blue color, with red as
2
high LD (R ¼ 0.8–1) and blue as low
2
LD (R ¼ 0–0.2). No SNPs in LD
2
(R > 0.2) with rs16957091 can be
detected outside of this region on
chromosome 15.
hg18) on chromosome 15q15.2 that contains STARD9, CDAN1,
TTBK2, and UBR1 (Fig. 2). Genotype imputation successfully
(dosage R2 0.3) estimated the allelic dosage for 126,145
common SNPs (MAF 0.05) on chromosome 15 that were
not genotyped by the Illumina HumanOmni2.5M chip. Association analyses based on imputation results on chromosome 15
identified five SNPs associated with risk for MGMT methylation
with P-values slightly lower than that seen for rs16957091,
although all five SNPs are in almost perfect LD with rs16957091
(R2s 0.95; Fig. 2). Imputation at chromosome 17q24.3 did not
find any associations that were more significant than that seen
for rs997781.
Rs16957091 affects UBR1 expression
The enrichment of SNPs in ENCODE-annotated regulatory
regions within the 472-kb haplotype block (chr15:4070994841181608, hg18) at chromosome 15q15.2 locus was assessed
using FunciSNP with data from SAEC. Multiple chromatin features surrounding STARD9, CDAN1, TTBK2, and UBR1 were
identified to overlap with SNPs in LD with rs16957091 (n ¼
108, R2 0.2, P-values by Benjamini–Hochberg adjustment
<0.01; Supplementary Fig. S3). Thus, the effect of rs16957091
on the expression of these four genes was assessed in 54 human
primary BEC cultures and 59 normal lung tissues. Among the four
genes within this block, UBR1 has the highest gene expression in
human primary BECs (Fig. 3A). A significant correlation of gene
expression among UBR1, CDAN1, and TTBK2 was identified
3112 Cancer Res; 75(15) August 1, 2015
(pairwise Pearson correlation coefficients ¼ 0.46–0.67, Ps 0.0005). However, STARD9 expression was not correlated with
the other three genes within the same haplotype block (Ps > 0.07)
possibly due to genomic recombination detected within STARD9.
A significant association was observed in human BECs (Ptrend ¼
0.018) and replicated in normal lung tissues (Ptrend ¼ 0.018) with
the variant allele (G) of rs16957091 associated with reduction in
UBR1 gene expression (Fig. 3A). The effect of rs16957091 on
STARD9, CDAN1, and TTBK2 gene expression was not statistically
significant (Ps > 0.32).
Experiments were designed to exam whether rs16957091–
UBR1 expression correlation was caused by sequence-dependent
allele-specific gene expression. Nine SNPs were annotated to any
UBR1 exons or 50 or 30 untranslated regions with MAF ranged from
0 to 0.12. LD analysis between rs16957091 and each of these nine
SNPs using 1,000 Genomes CEU population identified low LD
between rs16957091 and rs3917223 (R2 ¼ 0.34). The LD estimates between rs16957091 and each of the other eight SNPs are
very low (R2 ¼ 0–0.13). Sequencing genomic DNA and cDNA
containing rs3917223 from seven BECs heterozygous for
rs16957091 and rs3917223 did not identify an allele-specific
gene expression pattern for rs3917223.
Reduced UBR1 expression affects MGMT turnover and DNA
repair
UBR1 contains binding sites recognizing internal degrons of
MGMT protein and the UBR1/RAD6-dependent N-end rule
Cancer Research
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Chromosome 15q15.2 for MGMT Promoter Hypermethylation
Figure 3.
Rs16957091 affects expression of UBR1 and DNA repair capacity towards MNU
treatment. A, G allele of rs16957091 is associated with reduced gene
expression of UBR1 in primary human BECs (n ¼ 54; P ¼ 0.018) and normal
lungs (n ¼ 59; P ¼ 0.018). Y axis was truncated between 0.0075 and 0.015. B,
G allele of rs16957091 is associated with lower percentage of cells with
chromosomal damage (n ¼ 107, P ¼ 0.036), indicating better repair towards
MNU treatment. Results are summarized with mean and standard error.
pathway is a major mechanism targeting suicide MGMT protein
by alkylating agents and also unmodified MGMT for subsequent degradation via the 26S proteasome in eukaryotes (40,
41). Recovery of MGMT activity after its inactivation results
entirely from de novo protein synthesis (4). Our previous
studies demonstrated reduced gene transcription associated
with the enhancer SNP rs16906252 as a major mechanism
predisposing MGMT for promoter hypermethylation in the
lungs of smokers (27). Thus, the lower expression of UBR1
associated with rs16957091 variant allele could result in slower
turnover of the MGMT protein under normal physiological
conditions and/or following repair in response to challenge by
the alkylating agents, that in turn may lead to reduced MGMT
transcription.
UBR1 stable KD cell lines were established from three immortalized HBECs (HBEC1, HBEC4, and HBEC26) to test this hypothesis. A protease V8 digestion method (36) was used to distinguish
na€ve and suicide MGMT in HBEC1 treated with different MNU
concentrations (0, 0.5, 1, and 2.5 mmol/L) for 30 minutes. The
www.aacrjournals.org
ratio of na€ve versus suicide MGMT decreased with increasing
concentrations of MNU treatment with almost no active MGMT
left at the highest MNU concentration (not shown). Interestingly, without MNU treatment, protease V8 cutting is still
detected supporting formation of endogenous alkylating DNA
adducts as a part of normal cellular metabolism in these cells
(1, 42), although the assay specificity for detecting suicide
MGMT may not be perfect due to its potential binding to other
proteins. The ratio of na€ve versus suicide MGMT was higher in
control HBECs compared with UBR1 KD lines (0.95 vs. 0.71,
paired t test P ¼ 0.037; Fig. 1B) with the largest difference seen
in HBEC26 (0.36; Fig. 1C), suggesting accumulation of suicide
MGMT in UBR1 KD lines. UBR1 knockout cells are resistant to
alkylating agent (MNNG) induced genotoxicity probably
because UBR1 targets both alkylated and unmodified MGMT
(41). Thus, HBECs were treated with 1 mmol/L MNU and
O6MG was quantified in genomic DNA recovered from cells
at 0, 24, 48, and 72 hours post-MNU treatment. On average,
78% of O6MG adducts were repaired within 24 hours in
control HBECs. In contrast, 89% O6MG were repaired within
24 hours in UBR1 KD lines (Fig. 1D). Rates of repair within
24 hours differed significantly between control and UBR1 KD
(paired t test, P ¼ 0.011; Fig. 1D). The modest but statistically
significant effect of UBR1 KD on MGMT degradation and repair
of O6-methylguanine may reflect the complexity of constitutive
and exogenous alkylating agent-accelerated degradation of
MGMT. In addition to the UBR1/RAD6-dependent N-end rule
pathway, the UFD4/UBC4-dependent ubiquitin fusion degradation pathway functions as an enhancer of the processivity of
MGMT polyubiquitylation by the N-end rule pathway through
physical interaction between the HECT-type UFD4 E3 and the
RING-type UBR1 E3 (40, 41).
Because rs16957091 reduced UBR1 expression and UBR1 KD
was associated with faster DNA repair of O6MG induced by
MNU, we further assessed whether rs16957091 is associated
with better repair in cultured lymphocytes collected from LSC
subjects (n ¼ 107). A significant inverse association (P ¼
0.036; Fig. 3B) between the number of rs16957091 G alleles
and percentage of cells with chromosomal damage was
observed, suggesting that the G allele associated with reduced
UBR1 expression is associated with greater DNA repair of MNUinduced DNA damage.
UBR1 expression levels affect MGMT transcription
Reduced gene transcription is a major mechanism predisposing
MGMT for promoter hypermethylation in multiple tissue types
(13, 27); therefore, a luciferase reporter assay was used to assess
whether UBR1 KD would result in reduced MGMT promoter
activity. Consistent with our previous studies and others, HAP4
(rs16906252 carrier) was associated with a 35% (P < 0.0001) and
22% (P ¼ 0.0019) reduction in promoter activity compared with
HAP1 (wild type) in control and UBR1 KD HBEC1 lines, respectively (Fig. 4A). Excitingly, UBR1 KD reduced HAP1 promoter
activity by 25% compared with control line (P ¼ 0.0009). Reduction of promoter activity (12%) in UBR1 KD line was also
observed for HAP4 (P ¼ 0.089), although to a lesser degree that
likely stemmed from the inefficient transcription regulatory
mechanism due to carrying rs16906252. The combination
of UBR1 KD and HAP4 reduced promoter activity by 50% (P <
0.0001) compared with wild-type UBR1 and HAP1. The reduction
of HAP1 promoter activity in the UBR1 KD HBEC4 line was
Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
3113
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Leng et al.
Figure 4.
Effect of UBR1 KD on MGMT promoter
activity and induction of MGMT
expression following MNU treatment. A,
both UBR1 KD and HAP4 (rs16906252
carrier) reduced MGMT promoter
activity in HBEC4. The combination of
UBR1 KD and HAP4 reduced promoter
activity by almost 50% (P < 0.0001)
compared with wild-type UBR1 and
HAP1. The reduction of HAP1 promoter
activity in UBR1 KD HBEC4 lines was
replicated in HBEC1 (reduced by 23%;
P ¼ 0.0046) and in HBEC26 (reduced
8
by 28%; P ¼ 2.5 10 ). B, MNU
treatment at a noncytotoxic dose (1
mmol/L) induced MGMT transcription
in control HBEC4 and HBEC26 at 24 and
48 hours post-MNU treatment. UBR1 KD
largely abolished this induction in
HBEC4 and HBEC26. Although no
induction was observed in HBEC1,
control line still had higher MGMT
expression than UBR1 KD line (P ¼
0.086). Results are summarized with
mean and standard deviation.
replicated in HBEC1 (reduced by 23%, P ¼ 0.0046) and HBEC26
(reduced by 28%, P ¼ 2.5 108; Fig. 4A). These findings suggest
that impaired MGMT turnover due to UBR1 KD may lead to
reduced MGMT transcription, an important factor in the predisposition for promoter hypermethylation of MGMT.
The recovery of MGMT activity after its inactivation by
alkylating agent results entirely from de novo protein synthesis
and this rate differs between pulmonary cell types in the rat
(4, 43). Furthermore, a large induction of MGMT measured in
steady-state RNA has been observed in rat liver cell lines >24
hours post-alkylating agent treatment and was due to activation
of the MGMT promoter (3, 5). Strong induction of MGMT
expression in steady state RNA was also observed in control
HBEC4 and HBEC26 at 24 and 48 hours post-MNU treatment
(1 mmol/L), but not in control HBEC1 (Fig. 4B). In contrast,
only minimal induction was observed in HBEC4 and HBEC26
with UBR1 KD at 48 hours post-MNU treatment (Fig. 4B). GLM
was conducted to assess the effect of UBR1 status (KD vs. WT)
and time point (24 hours vs. 48 hours) on induction of MGMT
expression in HBEC4 and HBEC26. MGMT induction at 24 and
48 hours post-MNU treatment was not statistically different in
either cell lines (Ps > 0.11). However, statistically significant
differences were identified for MGMT induction between control and UBR1 KD in HBEC4 (least square means and standard
error, 1.68 0.068 vs. 1.17 0.068, P ¼ 0.0057) and HBEC26
(least square means and standard error, 1.43 0.067 vs. 1.13 0.067, P ¼ 0.032) in the GLM with adjustment for assay batch
and time point (24 hours vs. 48 hours).
3114 Cancer Res; 75(15) August 1, 2015
Discussion
Our GWAS study identified cis- and trans-acting SNPs that affect
silencing of a DNA repair gene (MGMT) by promoter hypermethylation acquired during lung carcinogenesis. Two novel
trans-acting loci at chromosome 15q15.2 and 17q24.3 were
identified as being associated with risk for MGMT methylation
detected in sputum and their effects were independent of the
enhancer SNP rs16906252. These three SNPs together improved
the ROC AUC by 20% in a logistic regression that explored
determinants for risk for MGMT methylation in current and
former smokers. Thus, SNPs affecting the predisposition of
MGMT silencing by promoter hypermethylation, together with
SNPs affecting methylation of other important cancer-relevant
genes acquired during carcinogenesis may lead to the establishment of a polygenic marker that could improve current risk
prediction models for these cancers (18).
MGMT methylation status is a validated prognostic biomarker
for the response of glioblastoma patients to the alkylating agent
temozolomide (14). Our previous study and others also suggested that MGMT methylation status may affect the response of
lung tumors to temozolomide treatment (27, 44). The assessment
of MGMT promoter activity using HEBC4 identified a 50%
reduction in luciferase reporter activity for the combination of
UBR1 KD and HAP4 (rs16906252 carrier) compared with wildtype UBR1 and HAP1 (non-rs16906252 carrier). This functional
assessment strongly supports the epidemiological findings that
subjects carrying variant alleles for cis- (rs16906252) and trans-
Cancer Research
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Chromosome 15q15.2 for MGMT Promoter Hypermethylation
(rs16957091) acting SNPs that regulate MGMT transcription had
the highest risk for MGMT methylation (OR, 17.9; 95% CI, 10.7–
30.1; P ¼ 5.8 1028; Supplementary Table S5). Furthermore,
these two SNPs should also augment MGMT methylation status
for the better stratification of patients for alkylating therapy due to
their association with reduced MGMT transcription in MGMT
unmethylated cells.
The mechanism for trans-acting meQTL is largely unknown.
Among 585 trans-acting meQTL identified in normal lungs,
approximately 196 trans-acting meQTL were mediated by methylation levels of CpG sites nearby the trans-acting SNPs (19).
Ontology classification of genes (n ¼ 115) in the region of the
proximal CpG sites mediating the trans-acting associations identified GTPase activity related genes, genes regulating transcription,
and genetic imprinting as the top three categories involved in
DNA methylation regulation. A unique mechanism underlying
the predisposition of MGMT methylation in lungs of smokers
involving the regulation of UBR1 expression by rs16957091
within the chromosome 15q15.2 locus was elucidated (Fig. 5).
Because reduced gene transcription is a major mechanism predisposing MGMT for promoter hypermethylation in multiple
tissue types (13, 27), three HBEC lines with stable UBR1 KD
were established to assess whether reduced UBR1 expression
could affect the MGMT transcription. The functional validity of
UBR1 KD was confirmed by showing compromised degradation
of suicide MGMT or unmodified MGMT and increased DNA
repair towards MNU induced O6MG in UBR1 KD lines compared
with control HBECs. These effects by UBR1 are consistent with its
role of binding to the internal degrons of suicide MGMT protein
and even unmodified MGMT for subsequent degradation via the
26S proteasome in eukaryotes (1, 40, 41). In addition, greater
DNA repair towards O6MG in UBR1 KD lines may be due to the
accumulation of unmodified MGMT resulting from UBR1 KD
(41). More importantly, stable UBR1 KD reduced MGMT promoter activity and abolished the induction of MGMT expression
in HBEC models treated with noncytotoxic MNU. These findings
suggest that compromised MGMT degradation may not be able to
provide an efficient positive feedback to the MGMT transcriptional machinery under normal physiological conditions or in
response to alkylating agent exposure. Thus, our findings not only
substantiate the reduced gene transcription as a major determinant for MGMT methylation in the lungs of smokers, but also
support the importance of the UBR1-mediated N-end rule pathway in regulating MGMT homeostasis and DNA repair towards
alkylated DNA adducts.
The mechanism for how rs16957091, located 223 kb from
UBR1 regulates its expression was explored by assessing the
enrichment of rs16957091 and SNPs in moderate and high LD
with it (R2 0.2) in ENCODE-annotated regulatory regions
surrounding UBR1 using SAEC data. Multiple chromatin features
surrounding UBR1 were identified to overlap with SNPs (n ¼ 22)
in LD with rs16957091 (Supplementary Fig. S3), suggesting that a
combination of variant alleles of these SNPs may form a genomic
backbone associated with a particular active or repressive chromatin conformation that affects UBR1 expression. In addition,
SNP rs68062403 located at þ178 relative to the transcription start
site of UBR1 is in moderate LD (R2 ¼ 0.25) with rs16957091. A
search of the ChIP-seq database identified multiple transcription
factors (POLR2A, YY1, and TAF1) overlapping with the region
where rs68062403 resides in A549, a lung carcinoma cell line.
Furthermore, rs68062403 is located within an open chromatin
region likely to be enriched in gene regulatory elements in SAEC
(Supplementary Fig. S3) and A549 (not shown). Thus, the association seen in this study could be due to the LD with SNP
Figure 5.
Conceptual diagram for cis- and transacting SNPs for MGMT promoter
hypermethylation. The UBR1/RAD6mediated N-end rule pathway regulates
MGMT homeostasis under normal
MGMT turnover and degradation of
suicide MGMT upon alkylating agent
challenge. Reduced UBR1 expression
due to SNPs at 15q15.2 locus likely
induces the accumulation of
unmodified MGMT that is associated
with better DNA repair towards
alkylating agent induced DNA adducts.
In addition, the compromised
degradation of unmodified MGMT and
suicide MGMT is associated with
reduced MGMT promoter activity and
expression induction, although the
molecular mechanism is still unknown.
This study also verified MGMT enhancer
SNP (rs16906252) as the driver cisacting SNP for MGMT methylation
through reducing MGMT transcription.
Thus, both cis- and trans-acting SNPs
could affect the MGMT transcriptional
machinery that predisposes MGMT for
promoter hypermethylation in lungs of
smokers.
www.aacrjournals.org
Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
3115
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Leng et al.
rs68062403 that in turn may affect the binding of a transcription
factor.
Our studies provide a proof-of-concept for a unique mechanism underlying the predisposition of MGMT methylation in
lungs of smokers that involves the regulation of MGMT homeostasis by the UBR1/RAD6-mediated N-end rule pathway (Fig. 5).
The cis- and trans-acting SNPs affecting MGMT methylation
should have strong clinical translational significance in terms of
their utility in cancer risk assessment and patient stratification for
alkylating agent chemotherapy.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: S. Leng, C.S. Tellez, D. Desai, C.A. Stidley, Y. Liu, Y. Lin,
M.G. Wathelet, F.D. Gilliland, S.A. Belinsky
Development of methodology: S. Leng, L.B. Collins, C.S. Tellez, A.R. Jauregui,
S.G. Amin, Y. Liu, J.A. Swenberg, M.G. Wathelet, S.A. Belinsky
Acquisition of data (provided animals, acquired and managed patients,
provided facilities, etc.): S. Leng, L.B. Collins, C.S. Tellez, R.E. Crowell,
S.A. Belinsky
Analysis and interpretation of data (e.g., statistical analysis, biostatistics,
computational analysis): S. Leng, C.S. Tellez, X. Zhang, C.A. Stidley, Y. Liu,
S.A. Belinsky
Writing, review, and/or revision of the manuscript: S. Leng, C.S. Tellez,
A.R. Jauregui, M.A. Picchi, D. Desai, S.G. Amin, C.A. Stidley, M.G. Wathelet,
F.D. Gilliland, S.A. Belinsky
Administrative, technical, or material support (i.e., reporting or organizing
data, constructing databases): C.L. Thomas, M.A. Picchi, D.E. Juri, D. Desai, S.
A. Belinsky
Study supervision: S. Leng, S.A. Belinsky
Other (statistical analysis): G. Wu
Other (primarily worked with Dr. Belinsky and Shuguang Leng on the DNA
adduct data, in collaboration with Leonard Collins in the mass spectrometry
facility): J.A. Swenberg
Acknowledgments
The author thank Mr. Randall P. Willink, Ms. Kieu C. Do, and Ms. Mabel T.
Padilla from LRRI for assistance in conducting methylation-specific PCR and
Western blot assay, and Mr. Thomas J. Gagliano from LRRI for the scientific
editing of all figures and conceptualization of the diagram.
Grant Support
This work was primarily supported by the National Cancer Institute grant
R01 CA097356 (S.A. Belinsky). The State of New Mexico as a direct appropriation from the Tobacco Settlement Fund to S.A. Belinsky through collaboration
with the University of New Mexico provided initial support to establish the LSC.
Additional support was provided by NIH/NCI P30 CA118100 and NIEHS P30ES10126.
The costs of publication of this article were defrayed in part by the payment of
page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received January 26, 2015; revised May 22, 2015; accepted May 22, 2015;
published OnlineFirst July 16, 2015.
References
1. Pegg AE. Multifaceted roles of alkyltransferase and related proteins in DNA
repair, DNA damage, resistance to chemotherapy, and research tools.
Chem Res Toxicol 2011;24:618–39.
2. Srivenugopal KS, Yuan XH, Friedman HS, Ali-Osman F. Ubiquitinationdependent proteolysis of O6-methylguanine-DNA methyltransferase in
human and murine tumor cells following inactivation with O6-benzylguanine or 1,3-bis(2-chloroethyl)-1-nitrosourea. Biochemistry 1996;
35:1328–34.
3. Fritz G, Tano K, Mitra S, Kaina B. Inducibility of the DNA repair
gene encoding O6-methylguanine-DNA methyltransferase in mammalian cells by DNA-damaging treatments. Mol Cell Biol 1991;
11:4660–8.
4. Pegg AE, Wiest L, Mummert C, Stine L, Moschel RC, Dolan ME. Use of
antibodies to human O6-alkylguanine-DNA alkyltransferase to study the
content of this protein in cells treated with O6-benzylguanine or N-methylN0 -nitro-N-nitrosoguanidine. Carcinogenesis 1991;12:1679–83.
5. Grombacher T, Mitra S, Kaina B. Induction of the alkyltransferase (MGMT)
gene by DNA damaging agents and the glucocorticoid dexamethasone and
comparison with the response of base excision repair genes. Carcinogenesis
1996;17:2329–36.
6. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of
the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia.
Cancer Res 1999;59:793–7.
7. Esteller M, Toyota M, Sanchez-Cespedes M, Capella G, Peinado MA,
Watkins DN, et al. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is associated
with G to A mutations in K-ras in colorectal tumorigenesis. Cancer Res
2000;60:2368–71.
8. Yin D, Xie D, Hofmann WK, Zhang W, Asotra K, Wong R, et al. DNA repair
gene O6-methylguanine-DNA methyltransferase: promoter hypermethylation associated with decreased expression and G:C to A:T mutations of
p53 in brain tumors. Mol Carcinog 2003;36:23–31.
9. Sakumi K, Shiraishi A, Shimizu S, Tsuzuki T, Ishikawa T, Sekiguchi M.
Methylnitrosourea-induced tumorigenesis in MGMT gene knockout mice.
Cancer Res 1997;57:2415–8.
3116 Cancer Res; 75(15) August 1, 2015
10. Pulling LC, Divine KK, Klinge DM, Gilliland FD, Kang T, Schwartz AG, et al.
Promoter hypermethylation of the O6-methylguanine-DNA methyltransferase gene: more common in lung adenocarcinomas from never-smokers
than smokers and associated with tumor progression. Cancer Res
2003;63:4842–8.
11. Belinsky SA. Gene-promoter hypermethylation as a biomarker in lung
cancer. Nat Rev Cancer 2004;4:707–17.
12. Belinsky SA, Liechty KC, Gentry FD, Wolf HJ, Rogers J, Vu K, et al. Promoter
hypermethylation of multiple genes in sputum precedes lung cancer
incidence in a high-risk cohort. Cancer Res 2006;66:3338–44.
13. McDonald KL, Rapkins RW, Olivier J, Zhao L, Nozue K, Lu D, et al. The T
genotype of the MGMT C>T (rs16906252) enhancer single-nucleotide
polymorphism (SNP) is associated with promoter methylation and longer
survival in glioblastoma patients. Eur J Cancer 2013;49:360–8.
14. Weller M, Stupp R, Reifenberger G, Brandes AA, van den Bent MJ, Wick W,
et al. MGMT promoter methylation in malignant gliomas: ready for
personalized medicine? Nat Rev Neurol 2010;6:39–51.
15. Schalkwyk LC, Meaburn EL, Smith R, Dempster EL, Jeffries AR, Davies MN,
et al. Allelic skewing of DNA methylation is widespread across the genome.
Am J Hum Genet 2010;86:196–212.
16. Kerkel K, Spadola A, Yuan E, Kosek J, Jiang L, Hod E, et al. Genomic surveys
by methylation-sensitive SNP analysis identify sequence-dependent allelespecific DNA methylation. Nat Genet 2008;40:904–8.
17. Paliwal A, Temkin AM, Kerkel K, Yale A, Yotova I, Drost N, et al.
Comparative anatomy of chromosomal domains with imprinted and
non-imprinted allele-specific DNA methylation. PLoS Genet 2013;9:
e1003622.
18. Tycko B. Mapping allele-specific DNA methylation: a new tool for maximizing information from GWAS. Am J Hum Genet 2010;86:109–12.
19. Shi J, Marconett CN, Duan J, Hyland PL, Li P, Wang Z, et al. Characterizing
the genetic basis of methylome diversity in histologically normal human
lung tissue. Nat Commun 2014;5:3365.
20. Drong AW, Nicholson G, Hedman AK, Meduri E, Grundberg E, Small KS,
et al. The presence of methylation quantitative trait loci indicates a direct
genetic influence on the level of DNA methylation in adipose tissue. PLoS
One 2013;8:e55923.
Cancer Research
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Chromosome 15q15.2 for MGMT Promoter Hypermethylation
21. Fraser HB, Lam LL, Neumann SM, Kobor MS. Population-specificity of
human DNA methylation. Genome Biol 2012;13:R8.
22. Gibbs JR, van der Brug MP, Hernandez DG, Traynor BJ, Nalls MA,
Lai SL, et al. Abundant quantitative trait loci exist for DNA
methylation and gene expression in human brain. PLoS Genet
2010;6:e1000952.
23. Heyn H, Moran S, Hernando-Herraez I, Sayols S, Gomez A, Sandoval J, et al.
DNA methylation contributes to natural human variation. Genome Res
2013;23:1363–72.
24. Heyn H, Sayols S, Moutinho C, Vidal E, Sanchez-Mut JV, Stefansson OA,
et al. Linkage of DNA methylation quantitative trait loci to human cancer
risk. Cell Rep 2014;7:331–8.
25. Zhang D, Cheng L, Badner JA, Chen C, Chen Q, Luo W, et al. Genetic control
of individual differences in gene-specific methylation in human brain. Am J
Hum Genet 2010;86:411–9.
26. Boumber YA, Kondo Y, Chen X, Shen L, Guo Y, Tellez C, et al. An Sp1/Sp3
binding polymorphism confers methylation protection. PLoS Genet
2008;4:e1000162.
27. Leng S, Bernauer AM, Hong C, Do KC, Yingling CM, Flores KG, et al. The
A/G allele of rs16906252 predicts for MGMT methylation and is selectively
silenced in premalignant lesions from smokers and in lung adenocarcinomas. Clin Cancer Res 2011;17:2014–23.
28. Ronneberg JA, Tost J, Solvang HK, Alnaes GI, Johansen FE, Brendeford
EM, et al. GSTP1 promoter haplotypes affect DNA methylation levels
and promoter activity in breast carcinomas. Cancer Res 2008;68:
5562–71.
29. Hawkins NJ, Lee JH, Wong JJ, Kwok CT, Ward RL, Hitchins MP. MGMT
methylation is associated primarily with the germline C>T SNP
(rs16906252) in colorectal cancer and normal colonic mucosa. Mod
Pathol 2009;22:1588–99.
30. Kristensen LS, Nielsen HM, Hager H, Hansen LL. Methylation of MGMT in
malignant pleural mesothelioma occurs in a subset of patients and is
associated with the T allele of the rs16906252 MGMT promoter SNP. Lung
Cancer 2011;71:130–6.
31. Kristensen LS, Treppendahl MB, Asmar F, Girkov MS, Nielsen HM, Kjeldsen
TE, et al. Investigation of MGMT and DAPK1 methylation patterns in
diffuse large B-cell lymphoma using allelic MSP-pyrosequencing. Sci Rep
2013;3:2789.
32. Ogino S, Hazra A, Tranah GJ, Kirkner GJ, Kawasaki T, Nosho K, et al. MGMT
germline polymorphism is associated with somatic MGMT promoter
methylation and gene silencing in colorectal cancer. Carcinogenesis
2007;28:1985–90.
www.aacrjournals.org
33. Leng S, Liu Y, Thomas CL, Gauderman WJ, Picchi MA, Bruse SE, et al. Native
American ancestry affects the risk for gene methylation in the lungs of
Hispanic smokers from New Mexico. Am J Respir Crit Care Med
2013;188:1110–6.
34. Coetzee SG, Rhie SK, Berman BP, Coetzee GA, Noushmehr H. FunciSNP:
an R/bioconductor tool integrating functional non-coding data sets
with genetic association studies to identify candidate regulatory SNPs.
Nucleic Acids Res 2012;40:e139.
35. Ramirez RD, Sheridan S, Girard L, Sato M, Kim Y, Pollack J, et al. Immortalization of human bronchial epithelial cells in the absence of viral
oncoproteins. Cancer Res 2004;64:9027–34.
36. Ayi TC, Oh HK, Lee TK, Li BF. A method for simultaneous identification of
human active and active-site alkylated O6-methylguanine-DNA methyltransferase and its possible application for monitoring human exposure to
alkylating carcinogens. Cancer Res 1994;54:3726–31.
37. Sharma V, Collins LB, Clement JM, Zhang Z, Nakamura J, Swenberg JA.
Molecular dosimetry of endogenous and exogenous O(6)-methyl-dG and
N7-methyl-G adducts following low dose [D3]-methylnitrosourea exposures in cultured human cells. Chem Res Toxicol 2014;27:480–2.
38. Duan H, Leng S, Pan Z, Dai Y, Niu Y, Huang C, et al. Biomarkers measured
by cytokinesis-block micronucleus cytome assay for evaluating genetic
damages induced by polycyclic aromatic hydrocarbons. Mutat Res
2009;677:93–9.
39. Egger M, Smith GD, Altman DG. Systematic reviews in health care: metaanalysis in context. London: BMJ; 2001. xviii, 487 p. p.
40. Hwang CS, Shemorry A, Auerbach D, Varshavsky A. The N-end rule
pathway is mediated by a complex of the RING-type Ubr1 and HECTtype Ufd4 ubiquitin ligases. Nat Cell Biol 2010;12:1177–85.
41. Hwang CS, Shemorry A, Varshavsky A. Two proteolytic pathways regulate
DNA repair by cotargeting the Mgt1 alkylguanine transferase. Proc Natl
Acad Sci U S A 2009;106:2142–7.
42. Nakamura J, Mutlu E, Sharma V, Collins L, Bodnar W, Yu R, et al. The
endogenous exposome. DNA Repair 2014;19:3–13.
43. Belinsky SA, Dolan ME, White CM, Maronpot RR, Pegg AE, Anderson MW.
Cell specific differences in O6-methylguanine-DNA methyltransferase
activity and removal of O6-methylguanine in rat pulmonary cells.
Carcinogenesis 1988;9:2053–8.
44. Pietanza MC, Kadota K, Huberman K, Sima CS, Fiore JJ, Sumner DK, et al.
Phase II trial of temozolomide in patients with relapsed sensitive or
refractory small cell lung cancer, with assessment of methylguanine-DNA
methyltransferase as a potential biomarker. Clin Cancer Res 2012;
18:1138–45.
Cancer Res; 75(15) August 1, 2015
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.
3117
Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-15-0243
Implication of a Chromosome 15q15.2 Locus in Regulating UBR1
and Predisposing Smokers to MGMT Methylation in Lung
Shuguang Leng, Guodong Wu, Leonard B. Collins, et al.
Cancer Res 2015;75:3108-3117. Published OnlineFirst July 16, 2015.
Updated version
Cited articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
doi:10.1158/0008-5472.CAN-15-0243
This article cites 43 articles, 18 of which you can access for free at:
http://cancerres.aacrjournals.org/content/75/15/3108.full.html#ref-list-1
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected].
To request permission to re-use all or part of this article, contact the AACR Publications Department at
[email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2015 American Association for Cancer Research.