Research Article Title TET2 mutations affect non-CpG ... · Neumann2, Rong He2, Rodolphe Taby2, Aparna Vasanthakumar4, Trisha Macrae4, Kelly R. Ostler4, Hagop M. Kantarjian2, Shoudan

Yamazaki et al., p 1

Research Article

Title

TET2 mutations affect non-CpG island DNA methylation at enhancers and transcription factor

binding sites in chronic myelomonocytic leukemia

Authors and affiliations

Jumpei Yamazaki1,2, Jaroslav Jelinek1,2, Yue Lu3, Matteo Cesaroni1, Jozef Madzo1,4, Frank

Neumann2, Rong He2, Rodolphe Taby2, Aparna Vasanthakumar4, Trisha Macrae4, Kelly R.

Ostler4, Hagop M. Kantarjian2, Shoudan Liang5, Marcos R. Estecio2,3, Lucy A. Godley4 and Jean-

Pierre J. Issa1,2

1Fels Institute for Cancer Research and Molecular Biology, Temple University, Philadelphia, PA

19140, USA

2Department of Leukemia, The University of Texas MD Anderson Cancer Center, Houston, TX

77030, USA.

3Department of Molecular Carcinogenesis, The University of Texas MD Anderson Cancer Center,

Houston, TX 77030, USA.

4Section of Hematology/Oncology, Department of Medicine, The University of Chicago, Chicago,

IL 60637, USA.

5Department of Bioinformatics and Computational Biology, The University of Texas MD

on March 22, 2020. © 2015 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on May 13, 2015; DOI: 10.1158/0008-5472.CAN-14-0739

http://cancerres.aacrjournals.org/


Anderson Cancer Center, Houston, TX

Running Title: Effects of TET2 mutations on DNA methylation in CMML

Keywords: TET2, DNA methylation, mutation, CMML, genome-wide profiling

Precis: Results show how mutations in a leukemia-associated gene act epigenetically by

influencing the DNA methylation of hematopoietic-specific enhancers that impact leukemia

development.

Financial support: This work was supported by the National Institutes of Health grants

CA100632, CA121104, and CA049639 (to JPI) and CA129831 and CA129831-03S1 (to LAG), and

supported by a Stand Up to Cancer grant from the American Association for Cancer Research.

JPI is an American Cancer Society Clinical Research professor supported by a generous gift from

the F. M. Kirby Foundation.

Authors’ contributions

Contribution: J.Y. and J.-P.J.I. designed the study; J.Y., R.T., A.V., T.M., and K.O. performed

sequence analysis, analyzed sequence traces, and validated mutation; J.Y., F. N., and R. H.

performed DREAM; J.Y., J.J., Y.L., and S.L. analyzed DNA methylation data; J.Y., T.M., and J.M.

performed and analyzed 5hmc pull-down assay; J.Y. and J.-P.J.I. wrote the manuscript with

assistance from J.J., H.M.K., M.R.E., and L.A.G.

Correspondence:

Jean-Pierre J. Issa,





Fels Institute for Cancer Research and Molecular Biology, Temple University

3307 N. Broad Street, Room 154 Pharmacy Allied Health Building, Philadelphia, PA 19140

Email: [email protected]

Phone: 215-707-1454.

Conflicts of Interest: The authors declare no competing financial interests.

Word count: 4408

Total number of figures and tables: 6 figures, 7 supplementary figures, and 2 supplementary

tables





Abstract

TET2 enzymatically converts 5-methylcytosine to 5-hydroxymethylcytosine as well as other

covalently-modified cytosines and its mutations are common in myeloid leukemia. However,

the exact mechanism and the extent to which TET2 mutations affect DNA methylation remain

in question. Here we report on DNA methylomes in TET2 wild type (TET2-WT) and mutant

(TET2-MT) cases of chronic myelomonocytic leukemia (CMML). We analyzed 85,134 CpG sites

(28,114 sites in CpG islands (CGIs) and 57,020 in non-CpG islands (NCGIs)). TET2 mutations do

not explain genome-wide differences in DNA methylation in CMML, and we found few and

inconsistent differences at CGIs between TET2-WT and TET2-MT cases. By contrast, we

identified 409 (0.71%) TET2-specific differentially methylated CpGs (tet2-DMCs) in NCGIs, 86%

of which were hypermethylated in TET2-MT cases, suggesting a strikingly different biology of

the effects of TET2 mutations at CGIs and NCGIs. DNA methylation of tet2-DMCs at promoters

and non-promoters repressed gene expression. Tet2-DMCs showed significant enrichment at

hematopoietic-specific enhancers marked by H3K4me1, and at binding sites for the

transcription factor p300. Tet2-DMCs showed significantly lower 5-hydroxymethylcytosine in

TET2-MT cases. We conclude that leukemia-associated TET2 mutations affect DNA methylation

at NCGI regions containing hematopoietic-specific enhancers and transcription factor binding

sites.





Introduction

TET2 (ten-eleven translocation (TET) oncogene family member 2) is a tumor suppressor

gene on chromosome 4q24 (1). TET2 mutations were first described in myeloproliferative

neoplasms (MPN) (1), and were later also described in systemic mastocytosis (2), chronic

myelomonocytic leukemia (CMML) (3), myelodysplastic syndrome (MDS) (4), MDS/MPN (5),

and acute myeloid leukemia (AML) (6). The incidence of TET2 gene alterations ranges from 10-

50% in myeloid malignancies, with the highest frequency of mutations found in CMML, where

TET2 mutations were noted in 35-50% of cases (5-7). As first reported for TET1 (8), TET2

converts 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) (9) as well as other

covalently-modified cytosines (10,11) in embryonic stem cells, and thus mutations of TET2 were

theorized to contribute to leukemogenesis by altering the epigenetic regulation of transcription

through DNA methylation. In fact, among the three members of the TET gene family (TET1,

TET2, and TET3), TET2 is the sole gene found to be frequently mutated in myeloid malignancies

(6) and to disrupt hematopoietic differentiation (12,13). Furthermore, in murine models, Tet2

deficiency impairs hematopoietic differentiation with the expansion of myeloid precursors

(14,15). However, the exact mechanism and the extent to which TET2 mutations affect DNA

methylation remain in question. There are conflicting reports (12,13,16,17) on the effect of

TET2 mutations on DNA methylation. It has been reported that overall loss of 5mC content

(hypomethylation) (12) was a remarkable characteristic of CMML patients with TET2 mutations.

Two groups studied TET2 mutant AMLs and CMML and identified a promoter hypermethylation

phenotype (13,17). Another group reported predominantly hypomethylation in TET2 mutant

CMML, and we previously reported no effect of TET2 mutations on DNA methylation in CpG





islands (CGIs) (16). The genome-wide studies reported conflicting results used microarray

analysis with limited validation. Here, we employed the quantitative, sequence-based digital

restriction enzyme analysis of methylation (DREAM) (18) method to study this issue and found

that hypermethylated sites in TET2-MT are mostly in non-CpG islands (NCGIs) and are enriched

at hematopoietic-specific enhancers marked by H3K4me1, and at binding sites for the

transcription factor p300.

Materials and Methods

Patients

We analyzed whole bone marrow or peripheral blood samples (bone marrow was not

available in one TET2-WT case) prior to treatment from 40 patients with CMML referred to The

University of Texas MD Anderson Cancer Center or The University of Chicago, or enrolled in a

multi-institution Phase III trial comparing decitabine with supportive care (19). The Institutional

Review Boards at The University of Texas MD Anderson Cancer Center and The University of

Chicago approved each institution’s respective protocols, and all patients gave informed

consent for the collection of residual tissues as per institutional guidelines and in accordance

with the Declaration of Helsinki.

Mutation analysis

For TET2 gene analysis, polymerase chain reaction (PCR) and direct sequencing of exons

3-11 were performed starting from 20 ng of genomic DNA, as previously described (1). PCR





amplicons were sequenced by Beckman Coulter Genomics (Beckman Coulter Genomics, MA).

All TET2 mutations were scored on both strands. Sequence traces were analyzed with SeqMan

Pro (DNASTAR, Inc., WI) and reviewed visually. TET2 anomalies were numbered according to

the European Molecular Biology Laboratory nucleotide sequence reference FM992369.

Previously annotated single nucleotide polymorphisms in the HapMap database (20) were

discarded. SIFT software (21) was used to determine the probability that a particular amino acid

substitution is tolerated. We used pyrosequencing to analyze mutations of the R132 residue in

IDH1, and residues R140 and R173 in IDH2 which have been reported in MDS (22) and

glioblastoma (23). Mutations encoding amino acid R882 residue in the DNMT3A gene (24,25)

were analyzed by pyrosequencing. Primer sequences are listed in Supplementary Table 1.

Digital restriction enzyme analysis of methylation (DREAM)

Genome-wide DNA methylation analysis using next-generation sequencing (18,26) was

performed for 20 samples for which a sufficient amount of DNA was available. Briefly, genomic

DNA (5 μg) was digested with 5 μl of FastDigest SmaI endonuclease (Fermentas, MD) for 3 h at

37 °C. Subsequently, 50 U (5 μl) of XmaI endonuclease (NEB) was added, and digestion

continued for an additional 16 h. The digested DNA was purified using a QIAquick PCR

purification kit (Qiagen, CA). The 3′ recessed ends of the DNA created by XmaI digestion were

filled in with 3′-dA tails added by Klenow DNA polymerase lacking 3′-to-5′ exonuclease activity

(New England Biolabs, MA) and a dCTP, dGTP, and dATP mix (0.4 mM of each). Illumina paired-

end sequencing adaptors were ligated using Rapid T4 DNA ligase (Enzymatics, MA). The ligation

mix was size-selected by electrophoresis in 2% agarose. A slice corresponding to a 250- to 500-





bp window, according to the DNA ladder, was cut out, and DNA was extracted from the

agarose. Eluted DNA was amplified with Illumina paired-end PCR primers using iProof High-

Fidelity DNA Polymerase (Bio-Rad, CA) and 18 cycles of amplification. The resulting sequencing

library was cleaned with AMPure magnetic beads (Beckman Coulter Genomics) and sequenced

on an Illumina Genome Analyzer II or HiSeq 2000 (Illumina, CA). Sequencing reads were

mapped to SmaI sites in the human genome (hg18), and signatures corresponding to

methylated and unmethylated CpGs were enumerated for each SmaI site. Methylation

frequencies for individual SmaI sites were then calculated. The methylation ratio is the ratio of

the number of tags starting with CCGGG divided by the total number of tags mapped to a given

SmaI site. We used at least 10 sequencing reads to analyze methylation levels at individual SmaI

sites. Based on technical replicate experiments (data not shown), we could distinguish

differences in methylation of >10% with an FDR of 2.4%. We used the UCSC definition of CpG

islands: GC content of 50% or greater, length > 200 bp, ratio greater than 0.6 of observed

number of CG dinucleotides to the expected number on the basis of the number of Gs and Cs in

the segment (27). Sites at promoter regions are defined as being located within -1 kb to +500 b

from transcription start sites (TSS) of RefSeq genes.

Identification of Tet2-DMCs

P-values for the methylation difference between TET2-MT and WT for each CpG site were

calculated using a combinatorial approach. We created 1000 pseudo-datasets by sampling

patients between the two groups of TET2 mutant and TET2 wild type patients without

replacement. For each CpG site we created the sampling distribution of the average difference





between methylation levels in group one and group two. The proportion of shuffled datasets

where the difference was greater than or equal to the real difference was defined as the p-

value. We identified 506 CpG sites with a p-value less than or equal to 0.01. After we set the

cutoff for the minimum difference of 5% to exclude low level changes of uncertain significance,

the number of CpG sites differentially methylated in TET2 mutant and TET2 wild type CMML

patients dropped from 506 to 472. We applied the same approach to a control dataset

generated from 20 samples of total white blood cells (WBC) from age-matched healthy subjects

to evaluate the probability of obtaining 472 differentially methylated CpG sites by creating

random combinations of 2 groups containing 8 and 12 samples. After reshuffling all possible

125970 combinations and applying the filter for a minimum 5% difference between the two

groups, the probability of obtaining 472 differentially methylated CpG sites by random

reshuffling data from healthy WBCs was 0.0074 (935 combinations of 125970 total).

Quantitative DNA methylation analyses by bisulfite pyrosequencing

We used bisulfite pyrosequencing to quantitatively assess DNA methylation (28) for

differentially methylated genes from the DREAM analysis as well as for AIM2 and SP140 from a

previous report.(16) We analyzed the same sites of these genes which were analyzed by

DREAM. The number of patients with successful results (mostly >90% success rate) varied

slightly for each gene. Primer sequences are listed in Supplementary Table 1.

Quantitative real-time PCR

RNA was isolated using Trizol (Invitrogen, NY). cDNA was synthesized using High Capacity cDNA





Reverse Transcription Kits (Applied Biosystems, NY). Quantitative PCR was performed on a

StepOne™ Real-time PCR System (Applied Biosystems) using SYBR Green gene expression

assays (Bio-Rad). Gene expression data was normalized to GAPDH. Primer sequences are listed

in Supplementary Table 1.

Gene Set Enrichment Analysis (GSEA)

For Gene Set Enrichment Analysis (GSEA), gene sets were downloaded from the Broad

Institute’s MSigDB website (29). Gene set permutations were used to determine the statistical

enrichment of the gene sets using the difference in gene expression between TET2-MT and

TET2-WT cases in a gene expression microarray data with TET2 mutational status in the

previous study was used (30).

Analysis for enrichment of transcription factor binding sites and enhancer sites

The locations of transcription factor binding sites and enhancer sites were downloaded

from ENCODE project data (31). We annotated individual SmaI sites in the DREAM analysis

depending upon whether each site was in peaks for transcription factors and enhancers

(marked by H3K4me1). We calculated fold enrichment scores of tet2-DMCs over all sites

analyzed for each transcription factor binding site and enhancer site. We also mapped

individual SmaI sites in the DREAM analysis to regulatory regions by through publicly available

Ensembl Regulatory Build data (32).

5hmc pull-down assay followed by qPCR





The 5-hmC affinity purification was performed as previously described (33). Briefly,

sonicated genomic DNA (15-30 µg) was labeled with chemically modified uridine

diphosphoglucose glucose (UDP-6-N3-Glu). The click chemistry reaction was performed by

addition of 150 μM Biotin-S-S-DBCO (dibenzylcylooctyne). After pull-down with streptavidin

magnetic beads (Dynabeads, MyOne Streptavidin C1, Invitrogen), DNA was released with 50

mM DTT and purified by MinElute Reaction Cleanup Kit (Qiagen). DNA concentration after

affinity enrichment was measured by the Quant-iT PicoGreen dsDNA quantitation assay

(Invitrogen). The enrichment for target loci was assessed by qPCR using the Power SYBR Green

assay (Applied Biosystems) and the 7500 Applied Biosystems PCR machine. Primer sequences

are listed in Supplementary Table 1.

Statistical analysis

Statistical analyses were performed using PRISM (GraphPad Software, Inc., CA). We

used the Mann-Whitney test to compare continuous variables of DNA methylation levels

between TET2-MT and TET2-WT cases. All p values were two-tailed. Unsupervised hierarchical

analyses were performed by ArrayTrack (http://edkb.fda.gov/webstart/arraytrack/) with

standard criteria. Principal component analysis was performed in R using the princomp function

in the stats package. Fisher's exact test was used to calculate the enrichment of tet2-DMCs over

all sites analyzed for enhancers and transcription factors within known regulatory regions of

CD14+ monocytes and the GM12878 lymphoblastoid cell line.

Results





TET2 mutation status in CMML

We first analyzed the mutational status of the TET2 coding sequence (exons 3-11) in

samples from 20 patients with CMML, according to WHO criteria. TET2 missense or nonsense

mutations were detected in 8 out of 20 patients (40%) studied. Five patients had a single

heterozygous mutation, two had a biallelic or homozygous mutation, and one had two

mutations. Altogether, nine mutations were identified, including two missense, four nonsense,

and three frameshift mutations. Detailed mutation information is shown in Table 1. Using SIFT

software (21), both of the identified missense mutations were predicted to affect protein

function. Furthermore, we identified an IDH2 R140Q mutation in 1 out of 20 CMML patients

(5%), and a mutation at the R882 residue in DNMT3A was found in the same patient (TET2-WT).

Genome-wide DNA methylation analysis

We employed digital restriction enzyme analysis of methylation (DREAM) (18) using

next-generation sequencing, which allowed us to identify differentially methylated sites in the

human genome for TET2-MT and TET2-WT cases at high resolution, independently of bisulfite

treatment. From all the samples used for DREAM, 5-94 million unique usable reads (quality

filtered and aligned to the human genome) were successfully generated for DNA methylation

analyses (Supplementary Table 2). Supplementary Fig. 1 shows representative DREAM data for

DNA methylation from two TET2-MT and two TET2-WT cases compared to normal peripheral

blood. Compared to normal blood, hypermethylation in CGIs and hypomethylation in NCGIs

were found in a considerable number of sites in all patients regardless of TET2 mutation status





(Supplementary Fig. 1). This was also the case when averages of DNA methylation levels in each

population were analyzed. Direct comparison of averages of DNA methylation in TET2-MT

versus TET2-WT cases revealed a slight increase of CGI hypermethylation in TET2-WT cases and

NCGI hypermethylation in TET2-MT cases (Fig. 1). There were no differences in the methylation

status of 7 classes of repeat sequences examined (SINE, LINE, LTR, etc.) (Supplementary Fig. 2).

Next, we analyzed 38,282 CpG sites (all those with >9 reads in all 20 patients) alongside

those in 5 normal blood samples using unsupervised hierarchical clustering analysis. Using

either all sites (Supplementary Fig. 3a) or only the most variable sites (Fig. 2a), we found that

while normal blood and CMML clustered separately, TET2-MT and TET2-WT cases were not

clearly separated. This was also the case for a setting which analyzed CGI or NCGI sites

separately (Supplementary Fig. 3b and Supplementary Fig. 3c and Figs. 2b and 2c), suggesting

that TET2 mutations do not explain genome-wide differences in DNA methylation in CMML.

PCA analysis also showed interspersed patterns between TET2-MT and TET2-WT while normal

peripheral blood controls are tightly clustered and separated from CMML patients (Fig.2d, 2e,

and 2f).

Differentially methylated sites in TET2 mutants are mostly NCGIs

To dissect the difference in DNA methylation between TET2-MT and TET2-WT cases

more systematically, we calculated average DNA methylation in each population for each site

analyzed which had >9 reads in more than 10 out of 20 patients. In this setting, 85,134 CpG

sites were analyzed (28,114 sites in CGIs and 57,020 sites in NCGIs). Volcano plots of these sites





revealed that TET2-MT cases have more NCGI methylated sites, while minor differences were

seen in CGIs (Figs. 3a, 3b, and 3c). Using permutation analysis to control for overfitting, we

found 472 CpG sites (0.55%) which were differentially methylated in TET2-MT and TET2-WT

cases (TET2-specific differentially methylated CpGs; tet2-DMCs) (see Methods). We found more

methylated sites in TET2-MT (375 sites) than in TET2-WT cases (97 sites), supporting previous

findings of an increased amount of 5mC in TET2-MT compared to TET2-WT cases.(16)

Interestingly, we found a strikingly different biology at CGI and NCGI sites. Of these tet2-DMCs,

13% are in CGIs (63 sites) and 87% are in NCGIs (409 sites). Thus, 0.22% of CGI sites and 0.71%

of NCGI sites were affected by TET2 mutations (p<0.0001 for the difference between CGI and

NCGI). Furthermore, among the 63 CGI sites, 62% were more methylated in TET2-WT (39 sites).

By contrast, among the 409 NCGI sites, 86% (351 sites) were more methylated in TET2-MT

(p<0.0001). The differences were confirmed in additional cases by bisulfite-pyrosequencing for

several genes (Figs. 3d and 3e). Next, we performed supervised hierarchical clustering analysis

for only tet2-DMCs and found clearer clusters than in the analysis of all the sites

(Supplementary Fig. 4 and Fig. 2), suggesting that TET2 mutations mostly affect tet2-DMCs.

To further gain insight into the characteristics of tet2-DMCs, we compared their DNA

methylation levels to that seen in normal blood. This analysis gave us an idea of the role of the

temporal and spatial difference of tet2-DMCs in leukemogenesis in CMML. We again found a

striking difference between CGIs and NCGIs. Tet2-DMCs in CGI sites were often unmethylated in

normal blood and hypermethylated in TET2-WT cases but not in TET2-MT cases (Fig. 4a). On the

other hand, there is a considerable number of tet2-DMCs in NCGIs whose DNA methylation

levels in normal blood range from intermediate to high (Fig. 4b). For these sites, TET2-WT cases





showed hypomethylation, while TET2-MT cases showed identical to higher methylation levels

when compared to normal blood. To clarify these findings, we plotted the difference in DNA

methylation between TET2-MT and TET2-WT cases for tet2-DMCs against DNA methylation

levels in normal peripheral blood. Overall, tet2-DMCs in CGIs are mainly the sites whose

methylation levels are higher in TET2-WT cases than in TET2-MT cases, and which are

unmethylated in normal blood (Fig. 4c). By contrast, tet2-DMCs in NCGIs are the sites whose

methylation levels are higher in TET2-MT than in TET2-WT cases, and are intermediately to fully

methylated in normal blood (Fig. 4d). In other words, TET2-WT CMML is characterized by loss of

methylation at these normally methylated sits while TET2-MT CMML shows preserved or

enhanced methylation at these sites.

Functional relevance of TET2 mutations to leukemogenesis in CMML

To address the possible mechanism of TET2 mutations in leukemogenesis in CMML, we

sought a functional relevance for methylated sites in TET2-MT cases by correlating them with

gene expression levels in TET2-MT, TET2-WT, and normal peripheral blood. We looked at

expression profiles for genes with tet2-DMCs at their promoters (-1,000 bp to +500 bp from

transcription start site), since DNA methylation at promoters is well known to be correlated

with gene expression. We measured gene expression levels for GGA2, AIM2, and SP140 which

have NCGI promoters hypermethylated in TET2-MT cases. Within the complement of NCGI

sites, TET2 mutations affected promoters and non-promoter sites equally. Indeed, the CpG sites

previously validated as potential TET2 targets were AIM2 and SP140, both in NCGI

promoters.(16) As expected, we found a strong correlation between DNA methylation and gene





repression (Fig. 5a). We also selected three genes with tet2-DMCs in NCGIs more methylated in

TET2-MT at non-promoter regions and found that gene expression changes also correlated with

DNA methylation at these sites (Fig. 5b). Interestingly, TET2-WT cases showed expression levels

similar to TET2-MT cases in at least two out of three genes analyzed, but had lower DNA

methylation levels compared to TET2-MT cases.

Finally, to understand the mechanisms by which these genes are regulated by tet2-

DMCs in NCGI, we focused on enhancers and transcription factors which are known to control

gene expression in cis and trans from distal regions such as gene-body and outside the

genes.(34) We found that 25% (104 sites out of 409) of tet2-DMCs in NCGI were shown to be at

enhancer sites marked by H3K4me1 in a lymphoblastoid cell line, which was significantly

enriched compared to all NCGI sites analyzed (16%, p<0.001) (Fig. 5c and Supplementary Fig. 5).

We also noticed that these sites are localized at several transcription factor binding sites as

well. Among all analyzed, binding sites for p300 in the lymphobastoid cell line were co-localized

with 2.0% of these sites, which was a fourfold enrichment compared to all NCGI sites analyzed

(0.5%, p=0.001) (Fig. 5d and Supplementary Fig. 5). We also observed that tet2-DMCs in NCGIs

were significantly enriched in enhancer regions and in regions flanking promoters but depleted

within active promoters in CD14+ monocytes and the GM12878 lymphoblastoid cell line

(supplementary Fig. 6). No enrichment was observed in CTCF binding sites. Altogether, our data

suggest that methylation at tet2-DMCs in NCGIs is linked with dysregulation of gene expression

through altering hematopoietic specific transcription factor binding sites and enhancers.

Decreased 5hmc at tet2-DMCs in TET2-MT cases





Since neither DREAM nor bisulfite-pyrosequencing can discriminate 5hmc from 5mc, we

developed an assay for assessing enrichment of 5hmc at tet2-DMCs by 5hmc labeling followed

by an affinity enrichment method. We quantified 5hmc amounts at 6 tet2-DMCs in 13 CMML

patients (4 TET2-MT and 9 TET2-WT cases) and found that 3 out 6 tet2-DMCs analyzed showed

significantly lower 5hmc enrichment in TET2-MT cases (Figure 6a) while the remaining 3 tet2-

DMCs also showed a trend toward lower 5hmc enrichment in TET2-MT cases. We calculated a

z-score of 5hmc enrichment for each tet2-DMC and found that average of z-scores for 6 tet2-

DMCs in TET2-MT cases were significantly lower than TET2-WT cases (median z-scores -0.73 vs

0.40, P=0.03) (Figure 6b). Hierarchical clustering analysis of 5hmc enrichment clearly separated

subsets of TET2-MT cases and TET2-WT cases (Figure 6c).





Discussion

Genome-wide screening for DNA methylation by DREAM for eight TET2-MT and twelve

TET2-WT cases revealed that the general tumor phenotype in DNA methylation,

hypermethylation in CGIs and hypomethylation in NCGIs, is found in patients with both TET2-

MT and TET2-WT. Unsupervised hierarchical clustering analysis revealed that TET2-MT and

TET2-WT cases are not clearly separated, suggesting that the effects of TET2 mutations are

relatively minor in the human CMML methylome. We moved on to further analysis to clarify the

characteristics of the difference in DNA methylation between TET2-MT and TET2-WT cases. We

found that 0.55% of all sites analyzed were differentially methylated with a high proportion of

hypermethylation in TET2-MT, supporting previous findings of an increased amount of 5mC in

mutants compared to wild-types (16).

Strikingly, tet2-DMCs are primarily at NCGI sites that had intermediate to high DNA

methylation levels in normal blood. For these sites, TET2-WT cases showed hypomethylation,

while TET2-MT cases showed identical to higher methylation levels compared to normal blood.

This suggests that TET2 mutations block hypomethylation in NCGIs during tumorigenesis.

Interestingly, while TET1 has a CXXC domain responsible for binding to unmethylated sites, and

is enriched at CGIs (35,36), TET2 lacks the CXXC domain (37). This could explain why TET2

mutations primarily affect NCGI sites that are methylated in normal blood.

To gain insight into the functional relevance of TET2 mutations in leukemogenesis, we

investigated the effects of tet2-DMC methylation on gene expression. Although we found

relatively small numbers of genes with tet2-DMCs at promoter regions, we found good





correlations between DNA methylation and gene expression. Importantly, these genes have

promoters in NCGIs. We also found a substantial number of NCGI sites hypermethylated at non-

promoter regions in TET2-MT cases and found good correlations between DNA methylation and

gene expression which are validated for three genes with tet2-DMCs in NCGIs at non-promoter

regions. Interestingly, TET2-WT cases showed expression levels similar to TET2-MT cases in at

least two out of three genes analyzed, but had lower DNA methylation levels compared to

TET2-MT cases, implying that there might be mechanisms different from DNA methylation

which result in gene expression deregulation in CMML.

This finding prompted us to analyze the link to transcription factor binding sites (and

enhancers), which are known to control gene expression in cis and trans from distal regions in

order to achieve precise differentiation (34). We found that tet2-DMCs in NCGI case are

significantly enriched in enhancer sites—both in a lymphoblastoid cell line and in normal

monocytes— as well as in binding sites for transcription factors including p300. These data

suggest that methylation at tet2-DMCs in NCGIs might lead to reduced binding of p300 to

target sites, which in turn deregulate gene expression for nearby genes. In support of this, we

found that differentially expressed genes in TET2-MT and TET2-WT cases are enriched in

interferon-related genesets, which are known to be downstream targets of p300

(Supplementary Fig. 7) (38). Furthermore, it has recently been shown that approximately half of

5-hydroxymethylcytosines are located in distal regulatory elements immediately adjacent to

the binding sites of transcription factors such as p300 and CTCF (35,39-41). It is also worth

noting that two recent papers also reported preferential regulation of enhancer DNA

methylation by TET2 (42,43), which is very consistent with our findings.





Since the methods we used for identifying tet2-DMCs are not capable of discriminating

5hmc from 5mc and it is important to prove that not only 5mc but also 5hmc is variable at tet2-

DMCs, we developed a 5hmc pull-down assay to address this question. We found that 5hmc

amounts at tet2-DMCs are significantly lower in TET2-MT than TET2-WT cases, suggesting that

TET2 mutations lead to enzymatically deficient TET2 function at tet2-DMCs. Other covalently-

modified cytosines such as 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) are also

possibly variable between TET2-MT and TET2-WT, and this has to be verified with more

sensitive and specific assays to detect these further oxidized derivatives which are far lower in

quantity.

The effect of TET2 mutations on DNA methylation has been controversial. There are

three reports in CMML including our previous report, two out of which supported

hypermethylation phenotypes in TET2-MT CMML (16,17) whereas one report showed a

remarkable hypomethylation phenotype in TET2-MT CMML (12). The dominant

hypermethylation phenotypes were also supported by three other reports in AML (13), diffuse

large B-cell lymphoma (44), and normal elderly individuals (45). Among these, only one report

(44) utilized genome-wide profiling of DNA methylation like our method and showed that TET2

mutations were primarily associated with hypermethylation within CGI and CpG-rich

promoters. Although our data where we found tet2-DMCs was associated with NCGI sites is

inconsistent with this result, it is possible that the effects of TET2 mutations could vary

depending on differences in disease-origins such as myeloid and lymphoid cells. Similar studies

with genome-wide profiling need to be performed to clarify this question.

Although the fact that we have used normal peripheral blood rather than sorted cells as a





control is a drawback of our study, we were mostly interested in a case-case comparison of

TET2-WT to TET2-MT CMML, which is unaffected by normal peripheral blood data. Moreover,

we have previously shown that there are very few differences in methylation between bone

marrow and blood in MDS and CMML and relatively little variation in methylation in different

subsets in blood from patients with AML (sorted CD34+ and CD3–/19– cells) (46).

In conclusion, our data suggest that TET2 mutations have a minor effect (<1%) on DNA

methylation throughout the human genome, and preferentially result in hypermethylation at

selected NCGI sites that are enriched at transcription factor binding sites and enhancers. Our

data are consistent with a model whereby transcription factors such as p300 recruit TET2 as

part of their mechanism of gene regulation, and provide an explanation for the differentiation

block seen in TET2 mutant hematopoietic cells.





Acknowledgements

The authors thank the patients who have contributed to our understanding of these disorders

and thank Jenna Al-Malawi for excellent editing of the manuscript. This work was supported by

the National Institutes of Health grants CA100632, CA121104, and CA049639 (to JPI) and

CA129831 and CA129831-03S1 (to LAG), and supported by a Stand Up to Cancer grant from the

American Association for Cancer Research. JPI is an American Cancer Society Clinical Research

professor supported by a generous gift from the F. M. Kirby Foundation.





References

1. Delhommeau F, Dupont S, Della Valle V, James C, Trannoy S, Masse A, et al. Mutation in TET2 in

myeloid cancers. N Engl J Med 2009;360(22):2289-301. 2. Tefferi A, Levine RL, Lim KH, Abdel-Wahab O, Lasho TL, Patel J, et al. Frequent TET2 mutations in

systemic mastocytosis: clinical, KITD816V and FIP1L1-PDGFRA correlates. Leukemia 2009;23(5):900-4.

3. Tefferi A, Lim KH, Abdel-Wahab O, Lasho TL, Patel J, Patnaik MM, et al. Detection of mutant TET2 in myeloid malignancies other than myeloproliferative neoplasms: CMML, MDS, MDS/MPN and AML. Leukemia 2009;23(7):1343-5.

4. Langemeijer SM, Kuiper RP, Berends M, Knops R, Aslanyan MG, Massop M, et al. Acquired mutations in TET2 are common in myelodysplastic syndromes. Nature genetics 2009;41(7):838-42.

5. Jankowska AM, Szpurka H, Tiu RV, Makishima H, Afable M, Huh J, et al. Loss of heterozygosity 4q24 and TET2 mutations associated with myelodysplastic/myeloproliferative neoplasms. Blood 2009;113(25):6403-10.

6. Abdel-Wahab O, Mullally A, Hedvat C, Garcia-Manero G, Patel J, Wadleigh M, et al. Genetic characterization of TET1, TET2, and TET3 alterations in myeloid malignancies. Blood 2009;114(1):144-7.

7. Kosmider O, Gelsi-Boyer V, Ciudad M, Racoeur C, Jooste V, Vey N, et al. TET2 gene mutation is a frequent and adverse event in chronic myelomonocytic leukemia. Haematologica 2009;94(12):1676-81.

8. Tahiliani M, Koh KP, Shen Y, Pastor WA, Bandukwala H, Brudno Y, et al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science 2009;324(5929):930-5.

9. Ito S, D'Alessio AC, Taranova OV, Hong K, Sowers LC, Zhang Y. Role of Tet proteins in 5mC to 5hmC conversion, ES-cell self-renewal and inner cell mass specification. Nature 2010;466(7310):1129-33.

10. Ito S, Shen L, Dai Q, Wu SC, Collins LB, Swenberg JA, et al. Tet proteins can convert 5-methylcytosine to 5-formylcytosine and 5-carboxylcytosine. Science 2011;333(6047):1300-3.

11. He YF, Li BZ, Li Z, Liu P, Wang Y, Tang Q, et al. Tet-mediated formation of 5-carboxylcytosine and its excision by TDG in mammalian DNA. Science 2011;333(6047):1303-7.

12. Ko M, Huang Y, Jankowska AM, Pape UJ, Tahiliani M, Bandukwala HS, et al. Impaired hydroxylation of 5-methylcytosine in myeloid cancers with mutant TET2. Nature 2010;468(7325):839-43.

13. Figueroa ME, Abdel-Wahab O, Lu C, Ward PS, Patel J, Shih A, et al. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation. Cancer cell 2010;18(6):553-67.

14. Quivoron C, Couronne L, Della Valle V, Lopez CK, Plo I, Wagner-Ballon O, et al. TET2 Inactivation Results in Pleiotropic Hematopoietic Abnormalities in Mouse and Is a Recurrent Event during Human Lymphomagenesis. Cancer cell 2011;20(1):25-38.

15. Moran-Crusio K, Reavie L, Shih A, Abdel-Wahab O, Ndiaye-Lobry D, Lobry C, et al. Tet2 loss leads to increased hematopoietic stem cell self-renewal and myeloid transformation. Cancer cell 2011;20(1):11-24.

16. Yamazaki J, Taby R, Vasanthakumar A, Macrae T, Ostler KR, Shen L, et al. Effects of TET2





mutations on DNA methylation in chronic myelomonocytic leukemia. Epigenetics 2012;7(2):201-7.

17. Perez C, Martinez-Calle N, Martin-Subero JI, Segura V, Delabesse E, Fernandez-Mercado M, et al. TET2 mutations are associated with specific 5-methylcytosine and 5-hydroxymethylcytosine profiles in patients with chronic myelomonocytic leukemia. PLoS One 2012;7(2):e31605.

18. Challen GA, Sun D, Jeong M, Luo M, Jelinek J, Berg JS, et al. Dnmt3a is essential for hematopoietic stem cell differentiation. Nature genetics 2012;44(1):23-31.

19. Kantarjian H, Issa JP, Rosenfeld CS, Bennett JM, Albitar M, DiPersio J, et al. Decitabine improves patient outcomes in myelodysplastic syndromes: results of a phase III randomized study. Cancer 2006;106(8):1794-803.

20. International HapMap C. The International HapMap Project. Nature 2003;426(6968):789-96. 21. Ng PC, Henikoff S. Predicting the effects of amino acid substitutions on protein function. Annu

Rev Genomics Hum Genet 2006;7:61-80. 22. Kosmider O, Gelsi-Boyer V, Slama L, Dreyfus F, Beyne-Rauzy O, Quesnel B, et al. Mutations of

IDH1 and IDH2 genes in early and accelerated phases of myelodysplastic syndromes and MDS/myeloproliferative neoplasms. Leukemia 2010;24(5):1094-6.

23. Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med 2009;360(8):765-73.

24. Ley TJ, Ding L, Walter MJ, McLellan MD, Lamprecht T, Larson DE, et al. DNMT3A mutations in acute myeloid leukemia. N Engl J Med 2010;363(25):2424-33.

25. Walter MJ, Ding L, Shen D, Shao J, Grillot M, McLellan M, et al. Recurrent DNMT3A mutations in patients with myelodysplastic syndromes. Leukemia 2011;25(7):1153-8.

26. Jelinek J, Liang S, Lu Y, He R, Ramagli LS, Shpall EJ, et al. Conserved DNA methylation patterns in healthy blood cells and extensive changes in leukemia measured by a new quantitative technique. Epigenetics 2012;7(12).

27. Gardiner-Garden M, Frommer M. CpG islands in vertebrate genomes. J Mol Biol 1987;196(2):261-82.

28. Colella S, Shen L, Baggerly KA, Issa JP, Krahe R. Sensitive and quantitative universal Pyrosequencing methylation analysis of CpG sites. Biotechniques 2003;35(1):146-50.

29. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102(43):15545-50.

30. Braun T, Itzykson R, Renneville A, de Renzis B, Dreyfus F, Laribi K, et al. Molecular predictors of response to decitabine in advanced chronic myelomonocytic leukemia: a phase 2 trial. Blood 2011;118(14):3824-31.

31. Consortium EP. An integrated encyclopedia of DNA elements in the human genome. Nature 2012;489(7414):57-74.

32. Flicek P, Amode MR, Barrell D, Beal K, Billis K, Brent S, et al. Ensembl 2014. Nucleic acids research 2014;42(Database issue):D749-55.

33. Song CX, Szulwach KE, Fu Y, Dai Q, Yi C, Li X, et al. Selective chemical labeling reveals the genome-wide distribution of 5-hydroxymethylcytosine. Nat Biotechnol 2011;29(1):68-72.

34. Stadler MB, Murr R, Burger L, Ivanek R, Lienert F, Scholer A, et al. DNA-binding factors shape the mouse methylome at distal regulatory regions. Nature 2011;480(7378):490-5.

35. Wu H, D'Alessio AC, Ito S, Xia K, Wang Z, Cui K, et al. Dual functions of Tet1 in transcriptional regulation in mouse embryonic stem cells. Nature 2011;473(7347):389-93.

36. Williams K, Christensen J, Pedersen MT, Johansen JV, Cloos PA, Rappsilber J, et al. TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity. Nature 2011;473(7347):343-8.





37. Williams K, Christensen J, Helin K. DNA methylation: TET proteins-guardians of CpG islands? EMBO Rep 2012;13(1):28-35.

38. Vo N, Goodman RH. CREB-binding protein and p300 in transcriptional regulation. J Biol Chem 2001;276(17):13505-8.

39. Yu M, Hon GC, Szulwach KE, Song CX, Zhang L, Kim A, et al. Base-resolution analysis of 5-hydroxymethylcytosine in the mammalian genome. Cell 2012;149(6):1368-80.

40. Ficz G, Branco MR, Seisenberger S, Santos F, Krueger F, Hore TA, et al. Dynamic regulation of 5-hydroxymethylcytosine in mouse ES cells and during differentiation. Nature 2011;473(7347):398-402.

41. Stroud H, Feng S, Morey Kinney S, Pradhan S, Jacobsen SE. 5-Hydroxymethylcytosine is associated with enhancers and gene bodies in human embryonic stem cells. Genome Biol 2011;12(6):R54.

42. Hon GC, Song CX, Du T, Jin F, Selvaraj S, Lee AY, et al. 5mC Oxidation by Tet2 Modulates Enhancer Activity and Timing of Transcriptome Reprogramming during Differentiation. Molecular cell 2014.

43. Lu F, Liu Y, Jiang L, Yamaguchi S, Zhang Y. Role of Tet proteins in enhancer activity and telomere elongation. Genes & development 2014;28(19):2103-19.

44. Asmar F, Punj V, Christensen J, Pedersen MT, Pedersen A, Nielsen AB, et al. Genome-wide profiling identifies a DNA methylation signature that associates with TET2 mutations in diffuse large B-cell lymphoma. Haematologica 2013;98(12):1912-20.

45. Busque L, Patel JP, Figueroa ME, Vasanthakumar A, Provost S, Hamilou Z, et al. Recurrent somatic TET2 mutations in normal elderly individuals with clonal hematopoiesis. Nature genetics 2012;44(11):1179-81.

46. Shen L, Kantarjian H, Guo Y, Lin E, Shan J, Huang X, et al. DNA methylation predicts survival and response to therapy in patients with myelodysplastic syndromes. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2010;28(4):605-13.





Table 1. TET2 mutation status of the samples studied for DREAM

MT; mutant * Biallelic/homozygous mutations † Previously reported

Patient Nucleotide change Amino acid change

MT1 c.5163C>T Q1435X*

MT2 c.4435G>T G1192V

MT3 Ins c.2519 (G) V553FS

MT4 c.5109G>T V1417F*†

MT5 Del 3509_3510 (TC) F883FS

MT6 4914 G>T E1352X

MT7 2508 C>T R550X†

MT8 4506 C>T R1216X





Figure Legends

Figure 1. Scatter plots for DNA methylation levels analyzed by DREAM. Average DNA

methylation levels of TET2-MT versus TET2-WT in CGIs (upper) and in NCGIs (lower).

Differentially methylated sites are shown in black dots. R2 values are denoted in the upper right

corner.

Figure 2. Unsupervised hierarchical analyses for DNA methylation levels with the top 1,000

variable sites. Samples include eight TET2-MT (shown in red), twelve TET2-WT (blue), and five

normal bloods (green) (and an average of the five normal bloods) analyzed by DREAM. The

sample from the patient with IDH2/DNMT3A mutations is shown in purple. Sites in (a)

CGIs+NCGIs, (b) CGIs, and (c) NCGIs were used. Also shown is Principal Component Analysis of

DNA methylation at the 1000 most variable CpG sites in all sites analyzed (d), CGI sites (e), and

NCGI sites (f). TET2-WT patients (blue) and TET2-MT patients (red) are interspersed while

normal peripheral blood controls (NPB, green) are tightly clustered and separated from CMML

patients. The axes show loadings of the first two principal components and their scale is

arbitrary.

Figure 3. Difference in DNA methylation levels of TET2-MT and TET2-WT cases. Volcano plots

with the difference in DNA methylation between averages of TET2-MT versus TET2-WT on the

x-axis, and the unadjusted p-value for each site on the y-axis for (a) sites in CGIs+NCGIs, (b) sites

in CGIs, and (c) sites in NCGIs. Also shown is validation of DNA methylation levels by bisulphite-





pyrosequencing for (d) two tet2-DMCs in CGIs and (e) three tet2-DMCs in NCGIs, for TET2-MT

and TET2-WT cases. DNA methylation levels of normal blood samples are shown in the light

gray rectangle.

Figure 4. DNA methylation levels in TET2-MT and TET2-WT cases and their change compared to

normal blood. Scatter plots for the DNA methylation levels of tet2-DMCs in (a) CGIs and (b)

NCGIs in TET2-MT and TET2-WT cases versus normal blood. Each dot represents each tet2-DMC

in TET2-MT (red) and TET2-WT (blue). Heatmaps for tet2-DMCs in (c) CGIs and (d) NCGIs with

the difference in DNA methylation between TET2-MT and TET2-WT plotted on the x-axis, DNA

methylation levels in normal blood plotted on the y-axis, and their density in number of tet2-

DMCs (density increases from blue to red).

Figure 5. Deregulation of gene expression for genes with tet2-DMCs and their enrichment at

enhancer and transcription factor binding sites. Shown are correlations between DNA

methylation and gene expression for genes with tet2-DMCs at promoters (a) and non-

promoters (b). Red: TET2-MT; blue: TET2-WT; green: normal blood. The sample from the

patient with IDH2/DNMT3A mutations is shown in purple. Gene expression levels are calculated

as 40- ∆CT to GAPDH. Linear regression curve is indicated with a black line. (c) Enrichment for

enhancer sites (marked by H3K4me1) with tet2-DMCs in NCGIs over all sites analyzed in NCGIs

in several cell lines. Red bars indicate significant enrichment. (d) Enrichment for transcription

factor binding sites in GM12878, a lymphoblastoid cell line, with tet2-DMCs in NCGIs over all

sites analyzed s in NCGIs. Red bar indicates significant enrichment.





Figure 6. Decreased 5hmc enrichment at tet2-DMCs in TET2-MT cases. 5hmc enrichment scores

of (a) each tet2-DMC or (b) average of 6 tet2-DMCs in TET2-MT and TET2-WT cases. (c)

Unsupervised hierarchical clustering analysis for 5hmc enrichment scores from TET2-MT and

TET2-WT cases.






















Published OnlineFirst May 13, 2015.Cancer Res Jumpei Yamazaki, Jaroslav Jelinek, Yue Lu, et al. myelomonocytic leukemiaenhancers and transcription factor binding sites in chronic TET2 mutations affect non-CpG island DNA methylation at

Updated version

10.1158/0008-5472.CAN-14-0739doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2016/06/18/0008-5472.CAN-14-0739.DC1

Access the most recent supplemental material at:

Manuscript

Authoredited. Author manuscripts have been peer reviewed and accepted for publication but have not yet been

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/early/2015/05/13/0008-5472.CAN-14-0739To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-0739

http://cancerres.aacrjournals.org/content/suppl/2016/06/18/0008-5472.CAN-14-0739.DC1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2015/05/13/0008-5472.CAN-14-0739


Documents

Research Article Title TET2 mutations affect non-CpG ... · Neumann2, Rong He2, Rodolphe Taby2, Aparna Vasanthakumar4, Trisha Macrae4, Kelly R. Ostler4, Hagop M. Kantarjian2, Shoudan