Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Dynamic Epigenetic Regulation by Menin During Pancreatic

Islet Tumor Formation

Wenchu Lin1,2,3*, Hideo Watanabe1, 2,3,*, Shouyong Peng1, 2,3, Joshua M. Francis1,2,3,

Nathan Kaplan1,2,3, Chandra Sekhar Pedamallu 1,2,3,, Aruna Ramachandran1,2,3, Agoston

Agoston2, Adam J. Bass1,2,3, Matthew Meyerson1,2,3

1Department of Medical Oncology, Center for Cancer Genome Discovery, Dana-Farber

Cancer Institute, Boston, MA 02215, USA

2Department of Pathology, Brigham and Women's Hospital and Harvard Medical School

Boston, MA 02115, USA

3Cancer program, Broad Institute of Harvard and MIT, Cambridge, MA 02142, USA

*equal contribution

Correspondence: [email protected]

Running Title: Epigenetic regulation by the Men1 tumor suppressor

Key words: Men1, ChIP-seq, H3K4me3, H3K27me3, Igf2bp2

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

1

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Dynamic Epigenetic Regulation by Menin During Pancreatic

Islet Tumor Formation

Abstract

The tumor suppressor gene MEN1 is frequently mutated in sporadic pancreatic

neuroendocrine tumors (PanNETs) and is responsible for the familial Multiple Endocrine

Neoplasia type 1 (MEN-1) cancer syndrome. Menin, the product of MEN1,

associates with the histone methyltransferases (HMT) MLL1 (KMT2A) and MLL4

(KMT2B) to form menin-HMT complexes in both human and mouse model systems. To

elucidate the role of methylation of histone H3 at lysine 4 (H3K4) mediated by menin-

HMT complexes during pancreatic neuroendocrine tumor formation, genome-wide

histone H3 lysine 4 trimethylation (H3K4me3) signals were mapped in pancreatic islets

using unbiased chromatin immunoprecipitation coupled with next-generation sequencing

(ChIP-seq). Integrative analysis of gene expression profiles and histone H3K4me3 levels

identified a number of transcripts and target dependent on menin. In the absence of

Men1, histone H3K27me3 levels are enriched, with a concomitant decrease in H3K4me3

within the promoters of these target genes. In particular, expression of the insulin-like

growth factor 2 mRNA binding protein 2 (IGF2BP2) gene is subject to dynamic

epigenetic regulation by Men1-dependent histone modification in a time-dependent

manner. Decreased expression of IGF2BP2 in Men1-deficient hyperplastic pancreatic

islets is partially reversed by ablation of RBP2 (KDM5A), a histone H3K4-specific

demethylase of the jumonji, AT-rich interactive domain 1 (JARID1) family. Taken

2

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

together, these data demonstrate that loss of Men1 in pancreatic islet cells alters the

epigenetic landscape of its target genes.

Implications: Epigenetic profiling and gene expression analysis in Men1-deficient

pancreatic islet cells reveals vital insight into the molecular events that occur during the

progression of pancreatic islet tumorigenesis.

Introduction

Multiple endocrine neoplasia type 1 (MEN-1) is an autosomal dominant

syndrome, characterized by multiple tumors in endocrine tissues such as the pituitary

gland, parathyroid gland and pancreatic islets (1). Linkage studies and positional cloning

identified the causative gene, MEN1, for this disorder. Over 1300 mutations, typically

truncating, have been identified in MEN1 (2, 3). The importance of MEN1 inactivation in

tumorigenesis is highlighted by the frequency of MEN1 mutations in sporadic endocrine

tumors—44% in pancreatic neuroendocrine tumors and 35% in parathyroid adenomas (4,

5). Heterozygous and conditional Men1 knockout mice develop tumors in multiple

neuroendocrine tissues, recapitulating the spectrum of tumors in MEN-1 syndrome, with

Men1 conditional knockout animals demonstrating a shorter latency (6-9). Although

Men1 mutations are primarily associated with neuroendocrine cancers, several lines of

evidence demonstrate that Men1 can also be mutated in non-neuroendocrine tumors such

as lung cancer, melanoma and liver cancer (10-12).

Several studies have implicated that menin, the protein product of MEN1, is

involved in transcriptional regulation, cell cycle control, protein degradation and genome

instability through interaction with a number of transcription factors such as JunD, NF-

3

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

B and members of the Smad family (3, 13-15). In addition, we and others have shown

that menin is physically associated with Trithorax-like complexes containing the histone

methyltransferases MLL1 (KMT2A) and MLL4 (KMT2B, previously MLL2), to

promote trimethylation of histone H3 at lysine 4 (H3K4me3) (16, 17). Surprisingly,

menin also binds to the MLL fusion protein in leukemia cells to upregulate HoxA9 gene

expression, thus promoting oncogenic activity in MLL-associated leukemiogenesis (18).

Genome-wide analysis by chromatin immunoprecipitation coupled with DNA microarray

analysis (ChIP-chip) (19) has revealed that menin colocalizes with MLL at gene

promoters in various cell types, suggesting that menin regulates transcription in

cooperation with MLL in multiple tissues (20).

Currently there are no targeted therapies directed toward patients harboring

pancreatic neuroendocrine cancer tumors with MEN1 mutations. Thus there is a critical

need to deepen our understanding of the biology of these cancers to develop more

effective therapeutic approaches. Given the inactivation of menin in multiple endocrine

cancers and the reversibility of histone H3K4me3, we were interested in the role of

enzymes that potentially antagonize the histone methylation activity of menin. Rbp2

(Kdm5a, Jarid1a), initially identified as Retinoblastoma-binding protein 2, is a member of

the Jumonji (JMJ) domain-containing family of histone demethylases, with roles in

chromatin modification and transcriptional regulation (21). Loss of Rbp2 recruitment to

the CDKN1B gene is highly correlated with increased histone H3K4me3 levels and

elevated gene expression (21-23). We have previously demonstrated in murine models

that inactivation of the histone demethylase Rbp2 significantly inhibits tumor growth in

Men1-deficent mice (24). We also demonstrated that alterations in gene expression

4

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

patterns upon Men1 loss in pancreatic islets are partially reversed by Rbp2 loss in these

cells (24). Collectively, these observations support the notion that (a) histone methylation

plays a key role in Men1 deletion-mediated tumorigenesis in neuroendocrine cells, and

(b) the demethylase enzyme activity of Rbp2 antagonizes the histone methyltransferase

activity associated with menin at gene loci such as CDKN1B. Although loss of function

of menin is known to play an important role in tumor initiation and progression in

endocrine tissues (7), there is limited information on the mechanisms linking menin-

HMT complexes to neuroendocrine-specific hyperplasia and tumorigenesis.

Men1-deficient mice can take up to a year to accumulate the numerous genetic

and epigenetic alterations that result in tumor formation, a slow process during which

pancreatic islet cells transform from normal to a hyperplastic and finally a malignant state

(7). Thus, this period represents a window of opportunity to investigate early events

leading to tumorigenesis and to address the role of menin-HMT complexes in modulating

cell proliferation and behavior at this pre-cancerous stage.

To investigate tumor formation mediated by alterations in H3K4me3 levels and to

identify gene targets of menin-HMT complexes, we conducted epigenetic profiling of

Men1-deficent pancreatic islets in 2 month-old Men1 conditional knockout mice and

control wild-type littermates, Using chromatin immunoprecipitation techniques coupled

with next-generation sequencing (ChIP-seq), we found that Men1 loss lowered H3K4me3

levels at select target gene promoters, resulting in downregulation of gene expression.

Additionally, loss of H3K4me3 correlated with increased H3K27me3 levels, consistent

with the known association of H3K27me3 with gene repression (25). Our study is the

first to identify gene targets of menin-HMT complexes in mouse pancreatic islets in vivo

5

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

along with the time course of the epigenetic changes accompanying mouse pancreatic

neuroendocrine tumor formation.

Materials and methods

Mouse Experiments

Creation and genotyping of RIP-Cre mice, Men1 KO mice and Men1 Rbp2 double KO

mice has been described previously (24). Mice were maintained on a mixed 129s6,

FVB/N, and C57BL/6 background. All procedures were carried out in in accordance with

the National Research Council Guide for the Care and Use of laboratory Animals and

were approved by the Institutional Animal Care and Use Committee of the Dana-Farber

Cancer Institute.

Isolation of mouse pancreatic islets

Pancreatic islets were isolated as previously described (24).

Histological and Immunohistochemical analysis for pancreatic tissues

Pancreata were collected from mice at indicated time points and fixed in 4%

paraformaldehyde for 2 hours followed by dehydration and paraffin embedding.

Histopathological analysis was carried out on five-micrometer sections stained with

hematoxylin and eosin. Islet morphology and tumors were examined in at least three cut

sections for each pancreas after staining with hematoxylin and eosin. Appropriate

positive and negative controls were run on matched sections for all applied antibodies.

Immunohistochemical staining was performed on serial sections using antibodies against

H3K4me3 (Active Motif, catalog no. 36159, 1:500), H3K27me3 (Cell Signaling

Technology, catalog no. 9733, 1:100) and Igf2bp2 (Abcam, Ab124930, 1:1000). Sections

6

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

were counterstained in Meyer’s hemotoxylin, mounted and photographed using an

Olympus microscope.

ChIP-seq

For each ChIP experiment, islets from at least 4 adult mice were purified by collagenase

digestion and gradient centrifugation, with subsequent hand picking. ChIP was performed

as described (26) using 4 g of anti-H3K4me3 (Active Motif, catalog no. 36159) or anti-

H3K27me3 antibodies (Cell Signaling Technology, catalog no. 9733). 5 to 50 ng of DNA

was used for library construction. DNA was prepared for sequencing by Illumina cluster

generation using a SPRI-work system with 100-300 bp size selection followed by

enrichment with barcoded PCR primers for multiplexing. Sequencing was performed on a

Hiseq2000 machine for 40 nucleotides from a single end, at the MIT BioMicro Center.

Barcode-separated FASTQ files were generated from QSEQ files.

ChIP-Seq data analysis

Forty nucleotides of sequenced reads were aligned to the mouse reference genome (mm9

assembly), using Bowtie aligner (27). Only those reads/tags that mapped to unique

genomic locations with at most two mismatches were retained for further analysis.

Histone mark peaks were detected using MACS (version 1.4.2) as previously described

(28), with a p-value cutoff of 10-6 and with default values for other parameters.

Quantitative changes in H3K4me3 upon Men1 deletion was assessed using MAnorm

-5 algorithm (29) with a p-value cutoff of 10 and >1 log2 fold change. H3K4me3 peaks

that significantly decreased in Men1f/f compared to RIP-Cre control islets were assigned

to the most adjacent gene within 30kb; these genes were assessed for overlap with genes

downregulated in Men1f/f mice. H3K4me3 ChIP signals were plotted around transcription

7

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

start sites (TSS) of both upregulated or downregulated genes in Men1f/f islets. H3K27me3

ChIP signals were plotted around the peaks of either unchanged or decreased levels of

H3K4me3 in Men1f/f compared to RIP-Cre control islets.

Data availability

All ChIP-Seq data generated in this study have been deposited in the NCBI GEO

repository (accession number GSE63020).

RNA isolation and RT-qPCR

Total RNA was isolated using the RNeasy kit (Qiagen) from 100 to 300 mouse pancreatic

islets purified from 2 mice with different genotypes. RNA quality was assessed on the

Agilent Bioanalyzer. For RT-PCR, DNase I (Qiagen)-treated RNA samples were reverse

transcribed using oligo-dT and SuperScript III (Invitrogen), with first strand cDNA used

for PCR using SYBR green PCR mix (Qiagen, Valencia, California, United States) in an

Applied Biosystems 7300 Real Time-PCR system (Foster City, California, United

States). Standard ChIP with H3K4me3 and H3K27me3 antibodies was performed on

mouse pancreatic islets in duplicate. PCR primer pairs were designed to amplify 150- to

200-bp fragments from select genomic regions. Primer sequences are listed in Table S5.

Results

Inactivation of Men1 does not alter global H3K4me3 levels in pancreatic beta cells.

To address whether Men1 loss causes global changes in histone H3K4

methylation, we assessed H3K4me3 levels in Men1-deficient mouse pancreatic islets and

control RIP-Cre islets by immunohistochemistry (IHC). We detected no significant

change in the overall levels of H3K4me3 in Men1-deficient islets compared to wild-type

islets (Fig. 1A). Our finding is consistent with earlier studies demonstrating that in

8

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

contrast to Set1a and Set1b, the major H3K4 trimethylases in mammalian cells (30),

MLL1 and MLL4, the HMTs known to associate with menin, are responsible for H3K4

trimethylation of only a subset of loci (31). Our observations are also in line with studies

showing that MLL1 loss decreased H3K4me3 levels in less than 5 percent of genes in

mouse embryonic fibroblasts (MEFs) and that MLL4 knockdown had no impact on

overall H3K4me3 levels in mouse embryonic stem cells (31, 32). We previously reported

that inactivation of Rbp2 could partially rescue the tumor phenotype in Men1-deficient

mice (24). Immunohistochemistry revealed no appreciable change in overall H3K4me3

levels in Men1/Rbp2 double knockout islets compared with wild-type or Men1-deficient

islets (Fig. 1A, right panel), suggesting that menin-HMT complexes potentially regulate

histone modifications for only a subset of genes in mouse pancreatic islets.

Menin-dependent H3K4me3 is altered during early stages of pancreatic

neuroendocrine tumor formation.

To investigate -specific H3K4 trimethylation potentially regulated by menin-

HMTs in pancreatic islets, we performed ChIP-seq to identify genome-wide H3K4me3

occupancy in islets from 2 month-old RIP-Cre (control) mice or Men1-deficient mice.

We identified >22,000 H3K4me3 peaks, using the MACS algorithm (28). As expected,

H3K4me3 marks were most frequently observed at proximal promoter regions, near

transcriptional start sites (TSSs)(16) (Table S1 & S2). Comparable profiles were

observed in two independently purified batches of islets from RIP-Cre mice (data not

shown). We next compared H3K4me3 signals in Men1-deficient pancreatic and control

islets. Consistent with our predictions from the IHC results described above, we observed

differential H3K4me3 signals only in a subset of regions (Fig. 1B). Among the total 1565

9

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

differential peaks identified in control cells (p-value < 10-5, fold change >2), only 815

peaks (~3.5% of total H3K4me3 peaks) showed a decrease in H3K4me3 signals in Men1-

deficient islets (Fig. 1B, and Table S3), representing potential menin-HMT targets.

Integrative analysis of H3K4 trimethylation and gene expression identifies menin-

HMT gene targets.

To determine whether the regulation of differentially expressed genes in Men1-

deficient islets is dependent on menin-mediated H3K4me3, we integrated gene

expression data (24) with H3K4me3 profile data. Genes whose expression was

downregulated at least two-fold in Men1-deficient islets showed a significant reduction in

H3K4me3 levels (Fig. 2A). In contrast, H3K4me3 levels were unchanged at genes

upregulated at least two-fold upon menin loss (Fig. 2B). Notably the H3K4me3 signal in

downregulated loci was also generally lower than for genes showing increased expression

upon Men1 loss (compare profiles of Figs. 2A, B). These data suggest a direct role for

menin-dependent H3K4me3 in activating gene expression from its target promoters in

mouse pancreatic islets. The role of H3K4me3 in menin-dependent repression of gene

expression however remains to be elucidated.

To identify genes that were downregulated and associated with reduced

H3K4me3 levels in Men1-deficient islets we intersected these two datasets, revealing that

50 out of 365 genes showing reduced gene expression (14% overlap rate vs 1.7%

expected random overlap rate; p-value<0.001) also exhibited lower H3K4me3 levels

(Fig. 2C). Many of these targets are linked to type 2 diabetes and beta cell functions (also

see Table S4), with Igf2bp2 (insulin-like growth factor 2 mRNA binding protein 2)

emerging as one of the most significant genes in this category. In contrast, of the 1227

10

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

genes that showed increased expression upon Men-1 loss, only 40 (3.1% overlap rate vs

5.7% expected random overlap rate; not enriched) showed a corresponding increase in

H3K4me3 (Supplementary Fig. S1). These data indicate that gene repression mediated by

menin under wild-type conditions likely occurs by mechanisms other than via reduction

of H3K4 tri-methylation.

Men1-loss does not change histone modifications at known menin gene targets in

mouse pancreatic islets, at age of 2 months.

A number of genes including Cdkn1b, Cdkn2c, Mnx1, Hoxa9, and Hoxc6, have

been previously shown to be regulated by menin or menin-HMT complexes (16, 18, 20,

33, 34). We, however, did not observe a significant difference in enrichment of

H3K4me3 or reduction of H3K27me3 at the Cdkn1b, Cdkn2c, and Mnx1 promoters in

Men1-deficient islets (Supplementary Fig. S2), despite Cdkn1b and Cdkn2c being highly

expressed in mouse pancreatic islets (WL, unpublished observations). These findings,

although consistent with our prior observations (24) where we reported no significant

difference in expression of these genes upon Men1 loss, are at odds with other studies

using mouse embryonic fibroblasts (MEFs) and pancreatic islets that have demonstrated

the role of trimethylation of H3K4 in transcriptional activation of Cdkn1b and Cdkn2c

(34). It is possible that expression of Cdkn1b and Cdkn2c does not require menin-HMT

complexes at this stage.

Menin is also known to regulate the majority of Hox genes in MEFs (31). Hox

genes are typically bivalently modified in embryonic stem cells (32); however we did not

observe strong H3K4me3 signals at these loci in pancreatic islets (Supplementary Fig.

S3) although we detected strong H3K27me3 association with four Hox gene clusters

11

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

(Supplementary Fig. S3). Since H3K27me3 is associated with repression of gene

expression, our results are consistent with microarray data showing that most Hox genes

were silenced or expressed at very low levels (data not shown). Thus, at the time point

being evaluated, menin-driven gene expression appears to occur via both H3K4me3-

dependent and –independent mechanisms.

Loci with increased H3K27me3 signals in Men1-deficient islets are associated with

decreased H3K4me3.

Trimethylation of histone H3K4 is mediated by trithorax group including

MLL1 and MLL4 and can antagonize H3K27 trimethylation mediated by polycomb

group complexes such as PRC2 (35, 36). To investigate this potential inverse correlation,

we evaluated H3K27me3 levels by IHC in control, Men1 single knockout and

Men1/Rbp2 double knockout islets. As observed for H3K4me3, we did not detect

significant differences in overall expression of H3K27me3 (Fig. 3A). We next asked

whether loss of H3K4me3 in genes regulated by menin-HMT complexes altered

H3K27me3 levels. We compared H3K27me3 signals on H3K4me3-marked regions

classified as unchanged or decreased by Men1 loss (Figs. 1B gray and green points and

3B) (Fig. 3B). We detected enrichment of H3K27me3 at promoters that showed

decreased levels of H3K4me3 in Men1-deficient islets, but not at promoters with

unaltered H3K4me3 occupancy (compare red tracings in Figs 3C & 3D). In total, 37 out

of 50 genes with both decreased expression and decreased H3K4me3 in Men1-deficient

islets showed enhanced H3K27me3 (Supplementary Fig. S4).

12

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Dynamic time-dependent changes in H3K4me3 and H3K27me3 at the Igf2bp2

promoter.

Next, we sought to establish how H3K4me3 and H3K27me3 signals change over

time during pancreatic islet tumorigenesis. We focused on the Igf2bp2 locus since

Igf2bp2 was one of the two most down-regulated genes in Men1-deficient islets

compared to RIP-Cre islets based on analysis of microarray data, and because the Igf2bp2

promoter showed the most dramatic change in H3K4me3 signal, of the 50 genes

identified in our integrative analysis. We isolated pancreatic islets from control and Men1

knockout mice at 2, 6 and 12 months of age and tumors from Men1-deficient mice at 12

months. ChIP-seq to assess changes in H3K4me3 and H3K27me3 signals over time (Fig.

4A) showed, at 2 months, that H3K4me3 signals at the Igf2bp2 promoter in RIP-Cre

(control) islets (shown in blue) localized to two peaks; upstream of exon 1 (Peak 1) and

between exons 1 and 2 (Peak 2). In Men1-deficient islets, both peaks were decreased,

with Peak 2 showing a greater reduction. At 6 months, H3K4me3 signals were reduced in

both control and Men1-null islets compared to 2 months. Finally at 12 months we

observed a further reduction in H3K4me3 levels in wild-type mice and in tumors from

Men1-deficient animals, with a barely detectable Peak 2 in tumors (Fig. 4A).

In contrast, H3K27me3 occupancy at the Igf2bp2 promoter exhibited an inverse

trend over time. We detected no appreciable H3K27me3 signal in RIP-Cre islets from 2-

month-old mice whereas Men1-deficient islets showed enhanced H3K27me3, in regions

distinct from H3K4me3–occupied areas. We observed a slight increase in H3K27me3

signals in the RIP-Cre islets and strong H3K27me3 signals in the Men1-deficient islets

from 6-month-old mice, including an overlap with H3K4me3 Peak 2. This trend

13

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

continued at 12 months, where we observed further enhancement of the H3K27me3

signal in control islets and a robust H3K27me3 signal in Men1-deficient tumors from 12-

month-old mice, spreading over the Igf2bp2 promoter, upstream regulatory and coding

regions (Fig. 4B). Many other genes also show a similar inverse relationship between

H3K4me3 and H3K27me3 signals that was consistent during the progression of

pancreatic islet tumorigenesis; examples include Gata6 and Oxtr (Supplementary Fig.

S5).

Igf2bp2 expression is epigenetically regulated during pancreatic islet tumor

formation.

Analysis of our previously published microarray data demonstrated that Igf2bp2

expression is downregulated in Men1-deficent pancreatic islets and is partially restored in

Men1;Rbp2 double knockout islets (Fig. 5A). We verified these observations by qPCR

analysis, confirming that Igf2bp2 expression was decreased by 60% upon Men1 ablation

and rescued to near-baseline levels in the Men1 Rbp2 double knockout condition (Fig.

5B). ChIP-PCR revealed that H3K4me3 levels at the Igfbp2 locus in Men1-deficient

islets decreased to <60% of control (consistent with the ChIP-seq data) and was restored

to baseline in Men1 Rbp2 double knockout islets, whereas H3K4me3 signals in control

regions (p53 and Gapdh) were unaltered (Fig. 5C). In order to correlate the changes in

Igf2bp2 mRNA expression with protein levels, we evaluated Igf2bp2 expression by IHC

in pancreatic islets. We detected a marked reduction in Igf2bp2 protein in Men1-deficient

islets compared to control RIP-Cre (Fig. 5D). Consistent with transcript levels, Igf2bp2

protein levels were restored in Men1/Rbp2 double knockout islets. Collectively, these

data reveal that Igf2bp2 is epigenetically regulated by menin-HMT complexes in mouse

14

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

pancreatic islets and that epigenetic changes occurring as a consequence of Men1 loss are

partially restored by ablation of the Rbp2 histone demethylase.

Discussion

Menin-HMT targets associated with tumor progression.

Few genes to date have been established as targets of menin; these include Hoxc6,

Hoxc8, Hoxa9, Cdkn2c, Cdkn1b and Mnx1 (16, 20, 33, 34). Menin is known to associate

with a number of histone methyltransferases including MLL1 and MLL4 (16, 17). Here

we have investigated menin-induced epigenetic modifications during pancreatic

neuroendocrine tumorigenesis. We sought to identify direct targets of menin-HMT

complexes in mouse pancreatic islets via integrative analysis of ChIP-seq and gene

expression data and have identified a number of genes regulated by menin-mediated

H3K4 trimethylation in pancreatic islets. A similar approach has been employed by

others (37), using mouse embryonic stem cells and mouse pancreatic islet-like endocrine

cells (PILECs) as models to study menin-mediated H3K4me3 during ES cell

differentiation. Consistent with our observations, the authors did not observe a correlation

between downregulation of gene expression and decreased H3K4me3 in ES cells upon

Men1 loss. However, upon differentiation of ES cells into PILEC, decreased H3K4me3

levels were associated with downregulated genes, indicating a role for menin-HMT

complexes during ES cell differentiation and lineage-specification upon Men1 loss (37).

In contrast to previous studies in different experimental systems (33, 34), we did

not find trimethylation of H3K4 at the Cdkn2c and Cdkn1b genes to be dependent on

menin-HMTs. Thus, regulation of Cdk inhibitor genes at this stage may be driven by

15

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

other mechanisms or may be independent of menin. Menin has also been reported to play

a critical role during embryogenesis and MLL1-mediated leukemogenesis via regulation

of Hox gene expression (31).

Although several previous studies have focused on chromatin and gene expression

targets of menin (20, 37), we have identified different targets in our study. Previous

studies utilized different experimental systems, either non-hyperplastic islets (20) as

opposed to the hyperplastic islets in our study, or Men1-deficient mESCs and mouse

pancreatic islet-like endocrine cells (PILECs) (37). Furthermore, in studies of mouse

embryos, we previously identified HoxC6 and HoxC8 as menin targets (16), but did not

find these genes as targets in the current study, suggesting the possibility that Hox genes

might be the targets of menin during embryonic stages rather than in adult stages.

Indeed, in adult mouse pancreatic islets, we did not observe significant expression of Hox

genes or significant trimethylation of H3K4me3 at Hox gene promoters but did observe

high levels of H3K27 methylation at Hox gene promoters (Supplementary Fig. S3). Thus

we believe that the differences between the current and previous studies are likely to

represent the effect of Men1 under different biological circumstances, although we can

not exclude the possibility that some of these differences represent the effects of variation

in experimental conditions. Further independent studies in consistent cell types and

tissues will be required to clarify this issue in the long term.

Anti-correlation between H3K4me3 and H3K27me3 signals.

We observed that H3K27me3 signals were enriched in regions showing decreased

H3K4me3. This may arise from either a mixed population of individual cells with each

feature or represent truly bivalent domains within a single cell. In the former case, this

16

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

likely reflects an indirect effect of Men1 loss since menin has not been known to

associate with any histone K27 demethylase. However, since decreased gene expression

is accompanied by decreased H3K4me3 at menin target gene promoters (Fig. 2A), it is

conceivable that these epigenetic marks may coexist within a given cell.

Genes regulating cell fate decisions during embryonic development are often

characterized by dual H3K4me3 and H3K27me3 marks (38), a bivalent mark indicating a

“poised” state which can either be activated or repressed during lineage specification.

The MLL4 methyltransferase that interacts with menin (16, 17) has been reported to

regulate bivalent promoters in mouse embryonic stem cells (32). During pancreatic islet

cell differentiation some bivalent marks may keep genes silent and poised, capable of

switching to either an activated or repressed state in response to the appropriate signal;

genes associated with H3K4me3 alone, however, are inactivated by additional

trimethylation of H3K27 (38) (Fig. 6). To our knowledge this is the first study providing

evidence that a significant number of menin-dependent mouse genes are subject to

bivalent histone modification and regulation under physiological conditions.

Igf2bp2 functions during cell differentiation and tumorigenesis.

We have identified Igf2bp2 as a menin gene target that shows an increase in

H3K4me3 levels and reduction in H3K27me3, upon Men1 loss. We also observe that this

inverse correlation is enhanced with increasing age and is most dramatic in Men1-

deficient tumors. We found that Igf2bp2 is the first gene that shows an inverse correlation

between H3K4me3 and H3K27me3 levels in neuroendocrine tumor formation in vivo

and thus may represent a new class of menin-regulated genes associated with this pattern

of dual histone modifications. Accordingly, Igf2bp2 expression can be adjusted not only

17

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

by a decrease in H3K4me3 but also by enhancement of H3K27me3. Igf2bp2 is a

developmental gene highly expressed during embryogenesis, and gradually silenced in

the adult (39, 40). It is the major Igf2 binding protein family member expressed in adult

pancreatic islets (WL, unpublished observations). Although the precise function of

Igf2bp2 is unclear, it has been reported that Igf2bp2 interferes with Igf2 translation by

associating with the 5’ end of the Igf2 transcript during embryonic development (39). It is

unknown whether Igf2bp2 also functions as an inhibitor of Igf2 or other genes in the

adult. Thus, misregulation of Igf2bp2, as observed for other developmental genes, might

play a role in tumorigenesis.

Dynamic epigenetic regulation of Igf2bp2 by Men1 and Rbp2 under physiological

conditions.

Interestingly, menin has previously been shown to bind the promoter of the Igfbp2

(insulin-like growth factor binding protein 2) and to repress Igfbp2 expression in MEFs

(mouse embryonic fibroblasts) (41, 42), in addition to the effect on Igf2bp2 that we report

here. We do not see significant changes for either H3K4me3 or H3K27me3 levels at the

Igfbp2 locus in pancreatic islets harvested from mice at age of 2 months, suggesting that

the mechanism for Igfbp2 expression by menin is different from that for Igf2bp2

expression.

In contrast, Igf2bp2 expression is dynamically regulated by epigenetic changes

driven by menin-HMT complexes, and also modulated by the Rbp2 histone demethylase.

We observed that Igf2bp2 expression is decreased during the hyperplasia stage in Men1-

deficient pancreatic islets and is accompanied by changes in H3K4 and H3K27 histone

methylation at the Igf2bp2 promoter. These effects are partially reversed by deletion of

18

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

the Rbp2 histone demethylase, implying that inactivation of Rbp2 counteracts epigenetic

changes induced by menin-HMT complexes. One explanation for this phenomenon may

be that Rbp2, in association with the PRC2 complex (43), binds the Igf2bp2 promoter to

reduce its expression; loss of Rbp2 may relieve this repression, resulting in gene

activation.

In conclusion, we have identified several genes, notably Igf2bp2, as being

regulated by menin-mediated H3K4me3 and observe epigenetic changes in these targets

over time, strongly suggesting a role for these genes in pancreatic islet tumorigenesis

induced upon ablation of Men1.

Acknowledgements

We thank members of the Meyerson laboratory for critical reading of the manuscript and

helpful discussions; Harvard Specialized Histopathology Services-Longwood for

histology; the BioMicro Center for library preparation and Illumina sequencing and the

Joslin Diabetes Center Core facility for islet isolation.

Grant support

This work was supported by an anonymous gift (to M.M).

19

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

References:

1. Agarwal SK, Lee Burns A, Sukhodolets KE, Kennedy PA, Obungu VH, Hickman AB, et al. Molecular pathology of the MEN1 gene. Annals of the New York Academy of Sciences. 2004;1014:189-98. 2. Chandrasekharappa SC, Guru SC, Manickam P, Olufemi SE, Collins FS, Emmert-Buck MR, et al. Positional cloning of the gene for multiple endocrine neoplasia- type 1. Science. 1997;276(5311):404-7. 3. Lemos MC, Harding B, Shalet SM, Thakker RV. A novel MEN1 intronic mutation associated with multiple endocrine neoplasia type 1. Clin Endocrinol (Oxf). 2007;66(5):709-13. 4. Jiao Y, Shi C, Edil BH, de Wilde RF, Klimstra DS, Maitra A, et al. DAXX/ATRX, MEN1, and mTOR pathway genes are frequently altered in pancreatic neuroendocrine tumors. Science.331(6021):1199-203. 5. Cromer MK, Starker LF, Choi M, Udelsman R, Nelson-Williams C, Lifton RP, et al. Identification of somatic mutations in parathyroid tumors using whole-exome sequencing. J Clin Endocrinol Metab.97(9):E1774-81. 6. Crabtree JS, Scacheri PC, Ward JM, Garrett-Beal L, Emmert-Buck MR, Edgemon KA, et al. A mouse model of multiple endocrine neoplasia, type 1, develops multiple endocrine tumors. Proc Natl Acad Sci U S A. 2001;98(3):1118-23. 7. Crabtree JS, Scacheri PC, Ward JM, McNally SR, Swain GP, Montagna C, et al. Of mice and MEN1: Insulinomas in a conditional mouse knockout. Mol Cell Biol. 2003;23(17):6075-85. 8. Bertolino P, Tong WM, Galendo D, Wang ZQ, Zhang CX. Heterozygous Men1 mutant mice develop a range of endocrine tumors mimicking multiple endocrine neoplasia type 1. Mol Endocrinol. 2003;17(9):1880-92. 9. Biondi CA, Gartside MG, Waring P, Loffler KA, Stark MS, Magnuson MA, et al. Conditional inactivation of the MEN1 gene leads to pancreatic and pituitary tumorigenesis but does not affect normal development of these tissues. Mol Cell Biol. 2004;24(8):3125-31. 10. Gao SB, Feng ZJ, Xu B, Wu Y, Yin P, Yang Y, et al. Suppression of lung adenocarcinoma through menin and polycomb gene-mediated repression of growth factor pleiotrophin. Oncogene. 2009;28(46):4095-104. 11. Fang M, Xia F, Mahalingam M, Virbasius CM, Wajapeyee N, Green MR. MEN1 is a melanoma tumor suppressor that preserves genomic integrity by stimulating transcription of genes that promote homologous recombination-directed DNA repair. Mol Cell Biol.33(13):2635-47. 12. Xu B, Li SH, Zheng R, Gao SB, Ding LH, Yin ZY, et al. Menin promotes hepatocellular carcinogenesis and epigenetically up-regulates Yap1 transcription. Proc Natl Acad Sci U S A.110(43):17480-5. 13. Agarwal SK, Guru SC, Heppner C, Erdos MR, Collins RM, Park SY, et al. Menin interacts with the AP1 transcription factor JunD and represses JunD-activated transcription. Cell. 1999;96(1):143-52. 14. Heppner C, Bilimoria KY, Agarwal SK, Kester M, Whitty LJ, Guru SC, et al. The tumor suppressor protein menin interacts with NF-kappaB proteins and inhibits NF- kappaB-mediated transactivation. Oncogene. 2001;20(36):4917-25.

20

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

15. Kaji H, Canaff L, Lebrun JJ, Goltzman D, Hendy GN. Inactivation of menin, a Smad3-interacting protein, blocks transforming growth factor type beta signaling. Proc Natl Acad Sci U S A. 2001;98(7):3837-42. 16. Hughes CM, Rozenblatt-Rosen O, Milne TA, Copeland TD, Levine SS, Lee JC, et al. Menin associates with a trithorax family histone methyltransferase complex and with the hoxc8 locus. Mol Cell. 2004;13(4):587-97. 17. Yokoyama A, Wang Z, Wysocka J, Sanyal M, Aufiero DJ, Kitabayashi I, et al. Leukemia proto-oncoprotein MLL forms a SET1-like histone methyltransferase complex with menin to regulate Hox gene expression. Mol Cell Biol. 2004;24(13):5639-49. 18. Yokoyama A, Somervaille TC, Smith KS, Rozenblatt-Rosen O, Meyerson M, Cleary ML. The menin tumor suppressor protein is an essential oncogenic cofactor for MLL-associated leukemogenesis. Cell. 2005;123(2):207-18. 19. Odom DT, Zizlsperger N, Gordon DB, Bell GW, Rinaldi NJ, Murray HL, et al. Control of pancreas and liver gene expression by HNF transcription factors. Science. 2004;303(5662):1378-81. 20. Scacheri PC, Davis S, Odom DT, Crawford GE, Perkins S, Halawi MJ, et al. Genome-wide analysis of menin binding provides insights into MEN1 tumorigenesis. PLoS Genet. 2006;2(4):e51. 21. Klose RJ, Yan Q, Tothova Z, Yamane K, Erdjument-Bromage H, Tempst P, et al. The retinoblastoma binding protein RBP2 is an H3K4 demethylase. Cell. 2007;128(5):889-900. 22. Christensen J, Agger K, Cloos PA, Pasini D, Rose S, Sennels L, et al. RBP2 belongs to a family of demethylases, specific for tri-and dimethylated lysine 4 on histone 3. Cell. 2007;128(6):1063-76. 23. Zeng J, Ge Z, Wang L, Li Q, Wang N, Bjorkholm M, et al. The histone demethylase RBP2 Is overexpressed in gastric cancer and its inhibition triggers senescence of cancer cells. Gastroenterology.138(3):981-92. 24. Lin W, Cao J, Liu J, Beshiri ML, Fujiwara Y, Francis J, et al. Loss of the retinoblastoma binding protein 2 (RBP2) histone demethylase suppresses tumorigenesis in mice lacking Rb1 or Men1. Proc Natl Acad Sci U S A.108(33):13379-86. 25. Young MD, Willson TA, Wakefield MJ, Trounson E, Hilton DJ, Blewitt ME, et al. ChIP-seq analysis reveals distinct H3K27me3 profiles that correlate with transcriptional activity. Nucleic Acids Res. 2011;39(17):7415-27. 26. Wederell ED, Bilenky M, Cullum R, Thiessen N, Dagpinar M, Delaney A, et al. Global analysis of in vivo Foxa2-binding sites in mouse adult liver using massively parallel sequencing. Nucleic Acids Res. 2008;36(14):4549-64. 27. Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the . Genome Biol. 2009;10(3):R25. 28. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008;9(9):R137. 29. Shao Z, Zhang Y, Yuan GC, Orkin SH, Waxman DJ. MAnorm: a robust model for quantitative comparison of ChIP-Seq data sets. Genome Biol.13(3):R16. 30. Wu M, Wang PF, Lee JS, Martin-Brown S, Florens L, Washburn M, et al. Molecular regulation of H3K4 trimethylation by Wdr82, a component of human Set1/COMPASS. Mol Cell Biol. 2008;28(24):7337-44.

21

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

31. Wang P, Lin C, Smith ER, Guo H, Sanderson BW, Wu M, et al. Global analysis of H3K4 methylation defines MLL family member targets and points to a role for MLL1- mediated H3K4 methylation in the regulation of transcriptional initiation by RNA polymerase II. Mol Cell Biol. 2009;29(22):6074-85. 32. Hu D, Gao X, Morgan MA, Herz HM, Smith ER, Shilatifard A. The MLL3/MLL4 branches of the COMPASS family function as major histone H3K4 monomethylases at enhancers. Mol Cell Biol.33(23):4745-54. 33. Milne TA, Hughes CM, Lloyd R, Yang Z, Rozenblatt-Rosen O, Dou Y, et al. Menin and MLL cooperatively regulate expression of cyclin-dependent kinase inhibitors. Proc Natl Acad Sci U S A. 2005;102(3):749-54. 34. Karnik SK, Hughes CM, Gu X, Rozenblatt-Rosen O, McLean GW, Xiong Y, et al. Menin regulates pancreatic islet growth by promoting histone methylation and expression of genes encoding p27Kip1 and p18INK4c. Proc Natl Acad Sci U S A. 2005;102(41):14659-64. 35. Papp B, Muller J. Histone trimethylation and the maintenance of transcriptional ON and OFF states by trxG and PcG proteins. Genes Dev. 2006;20(15):2041-54. 36. Schmitges FW, Prusty AB, Faty M, Stutzer A, Lingaraju GM, Aiwazian J, et al. Histone methylation by PRC2 is inhibited by active chromatin marks. Mol Cell.42(3):330-41. 37. Agarwal SK, Jothi R. Genome-wide characterization of menin-dependent H3K4me3 reveals a specific role for menin in the regulation of genes implicated in MEN1-like tumors. PLoS One.7(5):e37952. 38. Voigt P, Tee WW, Reinberg D. A double take on bivalent promoters. Genes Dev.27(12):1318-38. 39. Nielsen J, Christiansen J, Lykke-Andersen J, Johnsen AH, Wewer UM, Nielsen FC. A family of insulin-like growth factor II mRNA-binding proteins represses translation in late development. Mol Cell Biol. 1999;19(2):1262-70. 40. Christiansen J, Kolte AM, Hansen T, Nielsen FC. IGF2 mRNA-binding protein 2: biological function and putative role in type 2 diabetes. J Mol Endocrinol. 2009;43(5):187-95. 41. La P, Schnepp RW, C DP, A CS, Hua X. Tumor suppressor menin regulates expression of insulin-like growth factor binding protein 2. Endocrinology. 2004;145(7):3443-50. 42. La P, Desmond A, Hou Z, Silva AC, Schnepp RW, Hua X. Tumor suppressor menin: the essential role of nuclear localization signal domains in coordinating gene expression. Oncogene. 2006;25(25):3537-46. 43. Pasini D, Hansen KH, Christensen J, Agger K, Cloos PA, Helin K. Coordinated regulation of transcriptional repression by the RBP2 H3K4 demethylase and Polycomb- Repressive Complex 2. Genes Dev. 2008;22(10):1345-55. 44. Ho JW, Bishop E, Karchenko PV, Negre N, White KP, Park PJ. ChIP-chip versus ChIP-seq: lessons for experimental design and data analysis. BMC Genomics.12:134.

22

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure Legends

Fig 1. H3K4me3 profiles in mouse pancreatic islets. A: Total H3K4me3 levels were

evaluated in pancreatic islet cells by immunohistochemistry (IHC). Islets were purified

from 2 month old mice in which Men1 was deleted using an islet-specific Cre driver,

either alone (Men1f/f; RIP-Cre) or in combination with Rbp2 (Men1f/f; Rbp2f/f; RIP-Cre).

Islets from mice bearing the Cre driver alone (RIP-Cre) were used as controls. B:

Volcano plot illustrating differential changes in H3K4me3 levels in Men1-/- islets vs wild-

type controls (left panel: green dots represent decreases in the Men1 KO while red dots

represent increases). Heat map of those loci from the volcano plot showing significant

alterations of H3K4me3 levels in Men1-null islets compared to wild-type islets (right

panel).

Fig 2. H3K4me3 levels are significantly decreased in genes down-regulated in Men1-

deficient islets. H3K4me3 levels were assessed within a 3 kb window of the

transcriptional start site of genes up- or down-regulated in Men1-depleted islet cells

compared to wild-type. A: A significant reduction in H3K4me3 signal was observed in

genes that were down-regulated in Men1 KO islets. B: no difference was apparent in the

promoters of genes up-regulated in Men1 KO islets . C: Overlap of genes showing

decreased expression in Men1 KO islets with those showing reduction in H3K4me3

levels identified 50 common targets.

Fig 3. Increased H3K27me3 levels are observed in regions with decreased

H3K4me3. A: H3K27me3 levels were evaluated in pancreatic islet cells by

immunohistochemical (IHC) analysis. Islets were purified from mice in which Men1 was

deleted using an islet-specific Cre driver either alone (Men1f/f; RIP-Cre) or in

23

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

combination with Rbp2 (Men1f/f; Rbp2f/f; RIP-Cre). Islets from mice bearing the Cre

driver alone (RIP-Cre) served as controls. B: Regions of H3K4me3 occupancy were

classified according to changes in baseline levels occurring upon Men1 loss. C,D:

H3K27me3 levels were assessed within a 20 kb window spanning the center of the

H3K4me3 signal in loci exhibiting decreased H3K4 trimethylation (red trace) or no

change (purple) in RIP-Cre (C) or Men1-depleted (D) islet cells. A significant enrichment

in H3K27me3 signal was observed in genes associated with decreased H3K4me3 level in

Men1 KO cells (red peaks in area defined by green dotted lines in 3D), but not in wild-

type cells (3C).

Fig 4. Time-dependent epigenetic changes at the Igf2bp2 promoter.

A: H3K4me3 levels progressively decrease with age in Men1-/- islets (red tracks) relative

to RIP-Cre islets (blue tracks). c.p.m: counts per million mapped reads. B: H3K27me3

levels are higher in Men1-/- islets (red) than in RIP-Cre islets (blue) in 2 month old mice

and continue to increase at 6 and 12 months.

Fig 5. Igf2bp2, a menin-HMT target, is epigenetically regulated in mouse pancreatic

islets. A: Relative expression of Igf2bp2 in islets from wild-type, Men1-/- and Men1-/-;

Rbp2-/- mice as determined by microarray analysis, and subsequent validation by qPCR

(B). C: ChIP-qPCR showing H3K4me3 levels at the Igf2bp2, Trp53 and Gapdh

transcriptional start sites in control, Men1-/- or Men1-/-;Rbp2-/- islet cells. Data are the

mean±SEM of three independent experiments. **, P<0.01. D: Detection of Igfbp2 by

immunohistochemistry in wild type, Men1-/- and Men1-/-; Rbp2-/- pancreatic islets in 2

month old mice.

24

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig 6. Proposed model for epigenetic regulation by menin and effect on gene

expression.

Histone marks and expression status of genes regulated by menin-HMT complexes in

wild-type (top panel) and Men1-null (bottom panel) mouse pancreatic islets. Increased

H3K4me3 marks mediated by menin-HMT complexes opens chromatin, allowing for

binding of promoter-specific transcription factors (44), general transcription factors

(GTFs) and RNA polymerase II (Pol II), to facilitate gene expression. Loss of the

H3K4me3 mark due to deletion of Men1, along with addition of the repressive

H3K27me3 mark—possibly by PRC2 (Polycomb Repressive Complex 2) complexes—

results in closed chromatin and repression of gene expression. Green triangles: H3K4me3

marks on chromatin; red circles: H3K27me3 marks.

25

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

f/f f/f A RIP-Cre Men1f/f; RIP-Cre Men1 ; Rbp2 ; RIP-Cre

H3K4me3

100μm 100μm 100μm

B 140 H3K4me3 Decreased 30 Increased 120

25 100

20 80

-log10 P-value -log10 60 15

10 40

5 20

0 0 -400 0 400-400 0 400 -2-4-6 0 2 4 6 RIP-Cre Men1f/f; log2 fold change (Men1 KO vs Cre) RIP-Cre

Figure 1. Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

A B H3K4me3 profile around TSS H3K4me3 profile around TSS of down-regulated genes of up-regulated genes 100 100 Men1KO RIP-CreMen1KO RIP-Cre Cre f/f Men1Cre f/f;RIP-Cre Men1 ;RIP-Cre 80 80

60 60

40 40 H3K4me3Signal H3K4me3Signal 20 20

0 0 -3000-2000-1000 0 1000 2000 3000 -3000-2000-1000 0 1000 2000 3000 Relative Distance (bp) Relative Distance (bp)

C

Down- Genes with regulated 50 decreased genes H3K4me3 315 573

Igf2bp2 H2-T22 Eps8l2 Prom1 Gucy2c Pik3c2g Igf2bp2 Trpm3 Gem Pdk4 Cryba2 Cldn4 Tat Cnksr1 Igsf5 Gpr119 Rasgrf2 Anks4b Oxtr Kcnk16 Cyp4f39 Filip1 Tmprss2 Fev Il17re Vil1 H2-T23 Igsf21 Ttc22 Bcmo1 Rps4y2 Klc3 Tmem71 Ring1 Scn3a Efemp2 Mapk15 Il17rc Gata6 Abca14 Zfp236 Mica13 Rtkn H2-Q8 Myh14 Mdk Efnb3 Arnt12 Nr1h4 Zbtb17 4930444G20Rik

RIP-Cre

Men1f/f;RIP-Cre

Figure 2. Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. A RIP-Cre Men1f/f; RIP-Cre Men1f/f; Rbp2f/f; RIP-Cre Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

H3K27me3

100μm 100μm 100μm

B Decreased upon Men1 loss

H3K4me3 occupancy Unchanged upon Men1 loss in Men1 KO vs RIP-Cre Increased upon Men1 loss

C H3K27me3 profile centered around H3K4me3 peaks in RIP-Cre islets Regions with decreased H3K4me3 Regions with unchanged H3K4me3 0.60

0.55

0.50

Averaged profile Averaged 0.45

0.40 1500 1500 -20000 -10000 0 10000 20000 Relative distance from H3K4me3 peak (bp) D H3K27me3 profile centered around H3K4me3 peaks in Men1 KO islets Regions with decreased H3K4me3 0.65 Regions with unchanged H3K4me3 0.60

0.55

0.50

Averaged profile Averaged 0.45 0.40

-20000 -10000 1500 0 1500 10000 20000 Relative distance from H3K4me3 peak (bp)

Figure 3.

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

A 1kb Peak 2 Peak 1 12

c.p.m RIP-Cre 12 2 mths Men1f/f;RIP-Cre 12 2 mths RIP-Cre 12 6 mths

H3K4me3 Men1f/f;RIP-Cre 12 6 mths

RIP-Cre 12 12 mths Men1 KO tumor Igf2bp2 Igf2bp2 12 mths B 1

c.p.m RIP-Cre 1[0 - 18] Men1KO_2M_K27me3_cnv_model_swig_ 2 mths Men1f/f;RIP-Cre [0 - 12] 1 2 mths RIP-Cre [0 - 20] 1 6 mths H3K27me3 Men1f/f;RIP-Cre [0 - 40] 1 6 mths RIP-Cre [0 - 15] 1 12 mths Men1 KO tumor Igf2bp2 Igf2bp2 12 mths

Figure 4. Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

A Microarray data B Q-RT-PCR data

1.2 p<0.001 p<0.001 G enotype Relative * * * * Igf2bp2 1.0 expression 0.8 RIP -Cre 1 f/f 0.6 M en1 ;RIP - 0.31 Cre 0.4 f/f f/f M en1 ;Rbp2 ; 0.61 RIP -Cre Relativeexpression 0.2

RIP-Cre Men1f/f; Men1f/f; RIP-Cre Rbp2f/f; RIP-Cre C ChIP-PCR data 1.6 1.6 * * 1.4 ns 1.4 ns 1.21.2 * * RIP-CreR 1.01 MMen1f/f;RIP-Cre 0.80.8 MMen1f/f;Rbp2f/f;RIP-Cre 0.60.6 IgGIg signal/anti-H3K4me3 0.40.4 signal (RIP-Cre) Relative level of H3K4me3Relative of level 0.20.2 0 Igf2bp2 TSS P53 TSS Gapdh TSS

D IHC f/f f/f RIP-cre Men1 ;RIP-cre Men1f/f ;Rbp2 ;RIP-cre

Igf2bp2

Figure 5. Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Ash2l Mll1/Mll2 Wdr5 menin Rbbp5 Men1 Histone Transcription marks

Pol II WT High ON TF GTFs H3K4me3

Ash2l Mll1/Mll2 Wdr5 Rbbp5

Low Null H3K4me3 OFF High H3K27me3 PRC2 ?

DownloadedFigure from 6. mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research. Author Manuscript Published OnlineFirst on December 23, 2014; DOI: 10.1158/1541-7786.MCR-14-0457 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Dynamic Epigenetic Regulation by Menin During Pancreatic Islet Tumor Formation

Wenchu Lin, Hideo Watanabe, Shouyong Peng, et al.

Mol Cancer Res Published OnlineFirst December 23, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-14-0457

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2014/12/24/1541-7786.MCR-14-0457.DC1

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

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2014/12/23/1541-7786.MCR-14-0457. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2014 American Association for Cancer Research.