Page 1 of 46 Diabetes

The p300 and CBP transcriptional coactivators are required for beta cell and alpha cell

proliferation

Chi Kin Wong1,2, Adam K WadeVallance2, Dan S Luciani2,3, Paul K Brindle4, Francis C

Lynn2,3,5, William T Gibson1,2

1. Department of Medical Genetics, University of British Columbia, Vancouver, BC,

Canada

2. BC Children’s Hospital Research Institute, Vancouver, BC, Canada

3. Department of Surgery, University of British Columbia, Vancouver, BC, Canada

4. St. Jude Children’s Research Hospital, Memphis, TN, USA

5. Departments of Cellular & Physiological Sciences, University of British Columbia,

Vancouver, BC, Canada

*Corresponding author: Chi Kin Wong, [email protected], (604)8752000 ext 6783

Running title: roles of p300/CBP in pancreatic islets

Word counts: 4000

Number of table: 0

Number of figure: 6

Diabetes Publish Ahead of Print, published online December 19, 2017 Diabetes Page 2 of 46

Abstract p300 (EP300) and CBP (CREBBP) are transcriptional coactivators with histone acetyltransferase activity. Various beta cell transcription factors can recruit p300/CBP, and thus the coactivators could be important for beta cell function and health in vivo. We hypothesized that p300/CBP contribute to the development and proper function of pancreatic islets. To test this, we bred and studied mice lacking p300/CBP in their islets. Mice lacking either p300 or CBP in islets developed glucose intolerance attributable to impaired insulin secretion, together with reduced alpha and beta cell area and islet insulin content. These phenotypes were exacerbated in mice with only a single copy of p300 or CBP expressed in islets. Removing p300 in pancreatic endocrine progenitors impaired proliferation of neonatal alpha and beta cells. Mice lacking all four copies of p300/CBP in pancreatic endocrine progenitors failed to establish alpha and beta cell mass postnatally. Transcriptomic analyses revealed significant overlaps between p300/CBP downregulated genes and genes downregulated in Hnf1αnull islets and Nkx2.2null islets, among others. Furthermore, p300/CBP are important for the acetylation of H3K27 at loci downregulated in Hnf1αnull islets. We conclude that p300 and CBP are limiting cofactors for islet development, and hence for postnatal glucose homeostasis, with some functional redundancy.

Introduction

The expression of specific transcription factors both determines and maintains the identities of pancreatic endocrine cells through activation of endocrine genes. For example, Pdx1, MafA and

NeuroD1 form transcriptional complex at the insulin promoters and enhancers to activate insulin expression synergistically in beta cells (1). These transcription factors also recruit coregulators Page 3 of 46 Diabetes

to finetune gene expression. p300 and CBP (p300/CBP) are transcription coactivators that share

more than 60% protein sequence identity and exhibit highly similar function. These coactivators

acetylate lysine residues on histones to modulate chromatin structure or function, and lysine

residues on nonhistone proteins to modulate their activities (2). While p300/CBP can acetylate

most histone proteins, they are absolutely essential for acetylating histone H3 lysine 27 (4). The

H3K27Ac mark tags tissuespecific promoters and enhancers and signals transcription of the

tissuespecific target genes (3,4).

p300/CBP appear to regulate important beta cell functions in vitro. For instance, p300/CBP

coactivate insulin gene expression in vitro by binding synergistically to Pdx1 and NeuroD1/E47

(5). siRNA knock down of p300/CBP in INS1 cells reduced glucosestimulated insulin gene

expression (6). In contrast, CRISPRCas9mediated deletion of p300 in INS1 832/13 cells

induced a subtle increase in glucosestimulated insulin secretion and reduced high glucose

mediated apoptosis (7). Mice with the S436A variant in both copies of CBP, a mutation which

render CBP unresponsive to insulintriggered phosphorylation, had increased islet mass but

relatively normal beta cell function (8). These data left unresolved whether p300/CBP expression

in pancreatic islets is necessary for establishing glucose homeostasis in vivo. We hypothesized

that removal of p300/CBP from pancreatic endocrine progenitors would lead to postnatal glucose

intolerance due to defects in islet mass and function. In this study, we generated and phenotyped

Neurog3Cre driven pancreatic isletspecific p300 and CBP knockout mice to study the roles of

these coactivators in pancreatic islets in vivo.

Research Design and Methods Diabetes Page 4 of 46

Animals

All procedures were approved by the University of British Columbia Animal Care Committee.

Mice housed at BC Children’s Hospital Research Institute were under 12hour daynight cycle

with ad libitum access to standard chow (Teklad 2918; Envigo, UK) and water. Neurog3Cre mice were obtained from Dr. Francis Lynn (9). Ep300fl/fl mice were obtained from Dr. Paul

Brindle (10). Crebbpfl/fl mice were purchased from The Jackson Laboratory (ME, USA) (13). All

mice were kept on C57/BL6J background. Crenegative littermates from each breeding setup

were used as wildtype (WT) control. Timed matings were used to study embryonic day 15.5

(E15.5) and day 18.5 (E18.5), postnatal day 0 (P0) and day 7 (P7) mice; the morning when a

vaginal plug was found on the dams was designated as E0.5.

Metabolic phenotyping

For glucose tolerance test, mice were fasted for five hours and then injected intraperitoneally

with 2 g/kg glucose. For insulin tolerance test, mice were injected with 0.7 U/kg Humulin R (Eli

Lily, IN, USA). Blood was sampled from mouse tails and blood glucose levels were assessed

using a Onetouch UltraMini glucometer (Johnson & Johnson, NJ, USA). For plasma insulin

measurement, mice were fasted for five hours and blood was sampled from the saphenous veins

using heparinized capillary tubes before and after glucose injection. Heparinized blood was

centrifuged at 2,000 g for 15 minutes at 4oC to separate plasma. Body composition analysis and

metabolic cage experiments were performed as previously described (11).

Analyte measurement Page 5 of 46 Diabetes

Insulin was quantified using STELLUX Chemiluminescent Immunoassays (ALPCO, NH, USA).

Glucagon was quantified using Glucagon ELISA (Mercodia, Sweden). Somatostatin was

quantified using Somatostatin EIA Kit (Phoenix Pharmaceuticals, CA, USA). Total GLP1 was

quantified using Multi Species GLP1 Total ELISA (Merck Millipore, MA, USA).

Ex vivo islet assays

Mouse pancreatic islets were isolated as described previously (12). For glucosestimulated

insulin secretion assay, overnight recovered islets were incubated in KrebsRinger buffer (KRB)

containing 2.8 mM glucose for 1 hour at 37°C. After the preincubation, 30 islets were incubated

in KRB containing either 2.8 mM glucose, 16.7 mM glucose or 2.8 mM glucose with 30 mM

KCl for 1 hour at 37°C. Supernatants were collected for insulin measurement. Fura2 calcium

imaging was performed as described before (13). Perifusion assays were performed on a Biorep

Perifusion system per manufacturer’s instruction using 100 islets per chamber (Biorep, FL,

USA).

Immunofluorescence staining

Adult pancreata were fixed in 10% formalin for 24 hours at 4oC, dehydrated, embedded in

paraffin and 5 µm serial sections were made. A total of four to five sections each separated by

150 µm were obtained per adult pancreas. For E15.5 and E18.5, sections were obtained from the

entire pancreas and sections separated by 30 µm were stained. For P0 and P7, sections separated

by 60 µm were obtained. These sections were stained for insulin, glucagon, somatostatin,

ghrelin, chromogranin A and/or Ki67. Other proteins were stained using randomly chosen

sections. After blocking, the sections were incubated with primary antibodies overnight at 4oC Diabetes Page 6 of 46

followed by incubation with secondary antibodies. Primary antibodies used include rabbit anti p300 (N15 + C20 1:1, 1/50; Santa Cruz, TX, USA), rabbit antiCBP (1/200; CST, MA, USA),

guinea pig antiinsulin (1/200; Abcam, UK), mouse antiglucagon (1/1000; Abcam), rabbit anti

somatostatin (1/400; Abcam), goat antisomatostatin (1/200; Santa Cruz), goat antighrelin

(1/100; Santa Cruz), rabbit antiKi67 (1/200; CST), rabbit antichromogranin A (1/200; Abcam),

goat antichromogranin A (1/200; Santa Cruz), mouse antiNgn3 (1/50; DSHB, IA, USA), rabbit

antiH3K27Ac (1/200; CST) and rabbit antiH3K27me3 (1/200; CST). The TUNEL assays were performed with the In Situ Cell Death Detection Kit (SigmaAldrich, MO, USA). Representative

images of individual islets were taken on a SP5 confocal microscope (Leica, Germany). Images

of whole pancreas sections were tiled on a BX61 fluorescence microscope (Olympus, Japan) and

quantified using Fiji (14). Islet endocrine area were calculated by dividing the corresponding

stained area by the total pancreas area. For E15.5 studies, total number of cells per population

were counted and normalized to total DAPI count per section.

Transcriptomic analyses

Total RNA was extracted from islets using TRIzol (ThemoFisher, MA, USA) and RNeasy Micro

Kit (QIAgen, Germany). For RNAseq, six WT, three p300IsletKO, three CBPIsletKO and three

CBPHet; p300KO samples were sequenced in two batches at UBC Biomedical Research Centre

Sequencing Core. RNA samples with RIN higher than 8.5 as measured on Bioanalyzer 2100

(Agilent, CA, USA) were prepared into sequencing libraries on NeoPrep using TruSeq Stranded mRNA Library Prep kit (Illumina, CA, USA). Each library was sequenced to a depth of at least

20 million paired end reads on a NextSeq 500 (Illumina, CA, USA). Quality filtered reads were aligned to reference mouse genome mm10 with TopHat2 (15). Count tables for the aligned reads Page 7 of 46 Diabetes

were generated and batch effects were corrected using SVA (16), followed by calling of

differentially expressed genes using DESeq2 with default settings in R (17). Genes were defined

as differentially expressed based on adjusted p value < 0.05. The Venn diagram for overlapping

downregulated genes was generated using BioVenn (18). Downregulated gene lists were

uploaded to Webgestalt for gene ontology (GO) terms analysis and transcription factor target

prediction (19), using a reference list of 15,999 islet genes from the WT islet transcriptome for

accurate overrepresentation analyses. The lowest false discovery rate (FDR) from Webgestalt

was capped at 1015. Manual gene set enrichment analyses (GSEA) were performed by applying

hypergeometric test with Bonferroni correction to the overlapped genes between different gene

sets. All RNAseq data were deposited in the Gene Expression Omnibus (GEO) database

(GSE101537).

For qPCR, RNAs were reverse transcribed using the iScript cDNA synthesis kit (BioRad, CA,

USA) and quantified by SYBR greenbased reaction using GoTaq qPCR Master Mix (Promega,

WI, USA) and primers (see Supplementary Table 1) on ABI 7500 RealTime PCR system

(ThermoFisher, MA, USA). Data were normalized to 18s rRNA and fold changes were

calculated with the ∆∆CT method (20).

Western blotting

Islet nuclear extracts were prepared using the NEPER Nuclear and Cytoplasmic Extraction Kit

(ThermoFisher). One µg of nuclear lysate was subject to Western blotting using rabbit antip300

(N15 + C20 1:1, 1/1000; Santa Cruz), rabbit antiCBP (1/1000; Santa Cruz) and mouse anti

TBP (1/5000; Abcam) as described previously (21). Diabetes Page 8 of 46

Low input native chromatin immunoprecipitation (NChIP)

The low input NChIP protocol was carried out as previously described (22). 100 islets were

dispersed in 0.05% trypsin and flash frozen prior to lysis. Buffers were supplemented with 10

mM sodium butyrate to retain acetylation signals. Cells were lysed and digested with MNase for

5 minutes as validated using wildtype islet cells. Chromatin equivalent to 10 islets per sample

was subjected to ChIP using rabbit antiH3K27Ac or rabbit antiH3K27me3 (both 2 µL/ChIP;

CST). ChIP’d DNA was quantified by qPCR and expressed as percentage input.

Statistics

Data were shown as mean ± standard deviation. Statistical significance was tested using

Student’s t test, oneway ANOVA or twoway ANOVA, as appropriate, with p < 0.05 considered statistically significant.

Results

Mice lacking p300 in pancreatic islets develop glucose intolerance due to hypoinsulinemia

We first characterized Neurog3Cre; Ep300fl/WT mice which did not develop glucose intolerance

up to 24 weeks old. They also appeared phenotypically identical to Crenegative Ep300fl/fl mice

and to mice bearing Neurog3Cre transgene alone (not shown). We then made Neurog3Cre;

Ep300fl/fl mice (p300IsletKO mice). At E15.5, p300 was effectively removed using Neurog3Cre in up to 95% of chromogranin Apositive cells, while Ngn3positive progenitors remained p300 positive (Supplementary Figure 1). This is likely due to a time lag between the onset of Cre expression and the onset of recombination. Among the chromogranin Apositive cells, recombination occurred in newly formed beta cells, alpha cells, delta cells and epsilon cells Page 9 of 46 Diabetes

(delta cells in Supplementary Figure 1; other cell types not shown). In p300IsletKO mouse

pancreata, p300 was absent in islet nuclei but expressed normally in the exocrine tissues (Figure

1A & 1B). The protein levels of CBP were similar between WT and p300null islets, indicating

the absence of compensatory overexpression of CBP paralog.

p300IsletKO mice were glucose intolerant at eight weeks of age but display normal insulin

tolerance (Figure 1C & 1D). Plasma insulin levels of p300IsletKO mice were 60% lower than that

of WT both before and after glucose injection (Figure 1E). Theoretically, defective glucose

metabolism in p300IsletKO mice could be in part due to the Neurog3Cre mediated recombination

outside of islets such as in ventromedial hypothalamus and enteroendocrine cells (9,23). We

found that WT and p300IsletKO mice had similar body composition, energy expenditure,

locomotor activity and food intake (Supplementary Figure 2A to 2D). Also, plasma total GLP1

levels were normal in p300IsletKO mice (Supplementary Figure 2E). In the absence of observable

extrapancreatic phenotypes, the glucose intolerance and hypoinsulinemia can be attributed to the

recombination within pancreatic islets rather than other Neurog3expressing tissues.

p300IsletKO mice have reduced islet area and islet insulin content

To examine how p300IsletKO mice developed glucose intolerance and hypoinsulinemia, we first

quantified islet cell area in adult mice. While the weights of p300IsletKO mouse pancreata were

comparable to WT mice (Supplementary Figure 3A), they showed a 25% reduction in alpha cell

and beta cell area (Figure 2A). Delta cell area was unaffected. The reduced alpha cell and beta

cell area was attributable to reduced number of islets rather than islet size (Figure 2B &

Supplementary Figure 3B). Betatoalpha cell ratios were similar between WT and p300IsletKO Diabetes Page 10 of 46

mouse pancreata (Supplementary Figure 3C). Insulin content of p300null islets was reduced by

19%, while glucagon content was unchanged (Figure 2C). The somatostatin content was elevated by nearly twofold. Fasting plasma glucagon levels of p300IsletKO mice were unaffected

(Supplementary Figure 3D). To assess the function of p300null islet explants, we stimulated the

islets with low glucose, high glucose or KCl and measured their insulin release. While both WT

and p300null islets responded similarly to low or high glucose, p300null islets had increased

insulin secretion and calcium response upon KCl stimulation (Figure 2D & 2E). Glucagon

secretion from p300null islets under low glucose condition was not significantly different from

the WT islets (Supplementary Figure 3E). Thus, the combined defects in islet mass and islet

insulin content result in deficient glucosestimulated insulin secretion in p300IsletKO mice.

CBPIsletKO mice share similar beta cell phenotypes to p300IsletKO mice

To understand whether p300 and CBP function similarly in pancreatic islets, we generated

Neurog3Cre; Crebbpfl/fl (CBPIsletKO) mice. The deletion of CBP in islets was not compensated for by the overexpression of p300 (Figure 3A). CBPIsletKO mice developed glucose phenotypes similar to p300IsletKO mice, including glucose intolerance at eight weeks of age without insulin resistance, and defective insulin release upon glucose injection (Figure 3B, 3C & 3D). Unlike p300IsletKO mice, fasting plasma insulin levels of CBPIsletKO mice did not differ from controls

(Figure 3D). The pancreata of CBPIsletKO mice had 40% less alpha cell area and 30% less beta cell area, with delta cell area unaffected (Figure 3E). In contrast to p300null islets, CBPnull islets had reduced glucagon content but normal somatostatin content as compared to control

(Figure 3F). Similar to p300null islets, CBPnull islets had reduced insulin content but their beta cell secretory function appeared normal (Figure 3F & 3G). Overall, both p300IsletKO and Page 11 of 46 Diabetes

CBPIsletKO mice exhibited reduced beta cell area and islet insulin contents, features that explain

their glucose intolerance.

Mice with only a single copy of p300 or CBP in islets develop more severe glucose and islet

phenotypes than mice lacking p300 or CBP alone in islets

Since mice lacking p300 or CBP had similar phenotypes, we hypothesized that p300IsletKO or

CBPIsletKO mice lacking an additional copy of p300 or CBP would develop more severe glucose

phenotypes. To test this, we generated Neurog3Cre; Crebbpfl/WT; Ep300fl/fl mice and Neurog3

Cre; Crebbpfl/fl; Ep300fl/WT mice (CBPHet; p300KO and CBPKO; p300Het mice, respectively). These

triallelic mice developed severe glucose intolerance by eight weeks of age with no defects in

insulin tolerance (Figure 4A & 4B; CBPKO; p300Het mice data in Supplementary Figure 4).

Unlike the biallelic mice, triallelic p300/CBP mice of either genotype failed to mount an insulin

response to glucose challenge (Figure 4C). In addition, CBPHet; p300KO mice had 58% less alpha

cell area and 45% less beta cell area (Figure 4D). Their islets contained 72% less insulin than

WT islets (Figure 4E). These phenotypes recapitulated those seen in p300IsletKO or CBPIsletKO

mice despite being more severe.

Expression of p300/CBP is necessary for neonatal beta cell and alpha cell proliferation

Since reduced alpha cell and beta cell area could be caused by defects in differentiation,

proliferation and/or apoptosis, we examined these processes throughout pancreas development in

WT and p300IsletKO mice. We first excluded apoptosis as a possible cause of islet cell loss by

performing TUNEL assays on pancreata from E18.5 and adult p300IsletKO mice; apoptotic events

in p300null islets were as rare as in WT islets (not shown). Next, the number of newly Diabetes Page 12 of 46

differentiated endocrine cells and Ngn3positive endocrine progenitors were unaffected in E15.5 p300IsletKO mouse pancreata (Supplementary Figure 5A). At E18.5, alpha cell, beta cell and pan

endocrine cell area were normal in p300IsletKO mouse pancreata (Supplementary Figure 5B). At

P7, the alpha cell and beta cell area were reduced in p300IsletKO mouse pancreata (Figure 5A).

The percentages of Ki67+ alpha and beta cells in these pancreata were lower than that of WT;

however, the overall percentage of Ki67+ cells in the pancreata was unchanged (Figure 5B &

5C). This indicated that the proliferation of neonatal alpha cells and beta cells was reduced in p300null islets. Hence, the reduced alpha cell and beta cell mass in the adult p300IsletKO mouse pancreata originated after E18.5 and is attributable to impaired proliferation.

We attempted to breed for Neurog3Cre; Crebbpfl/fl; Ep300fl/fl mice (p300/CBP double knockout mice) but we did not observe any of the double knockout mice in a cohort of 59 pups at the weaning age, in contrast to the triallelic p300/CBP mice which were observed at the expected

Mendelian ratio (Supplementary Table 2). We speculated that these p300/CBP double knockout mice might die shortly after birth due to failure to establish sufficient beta cell mass. At P0, some double KO pups survived but their pancreata lacked alpha cells and beta cells completely (Figure

5E & 5F). Surprisingly, their delta cell and epsilon cell populations were unaffected (Figure 5E).

Immunostaining showed that a few epsilon cells, which are normally absent in adult mouse pancreata, persist in the biallelic and triallelic mouse pancreata (Supplementary Figure 6A).

Thus, at least one allele of p300 or CBP is necessary for normal development of alpha cells and beta cells, but not for delta cells nor epsilon cells.

Page 13 of 46 Diabetes

Loss of p300/CBP impairs genes associated with multiple islet/beta cell transcription

factors and impairs the coactivation of Hnf1αassociated genes in vivo

As p300/CBP are transcriptional coactivators, loss of p300/CBP may reduce the expression of

genes important for islet function or development. We examined gene expression by performing

RNAseq on islet mRNAs from WT, p300IsletKO, CBPIsletKO and CBPHet; p300KO mice. We

identified 761 (477 down, 284 up), 923 (513 down, 410 up) and 5,589 (2,411 down, 3,178 up)

differentially expressed genes in p300null, CBPnull and triallelic islets relative to WT islets,

respectively (Supplementary Table 3).

We focused our analyses on the downregulated genes. Aggregation of the downregulated gene

sets revealed 230 downregulated genes overlapped between p300null islets (48.2%) and CBP

null islets (44.8%) (Figure 6A). The genes downregulated in CBPHet; p300KO islets overlapped

with 436 (91.4%) and 437 (85.2%) of the downregulated genes from p300null and CBPnull

islets, respectively. Enrichment analyses of the Biological Process GO terms on all three sets

suggested three common themes of genes downregulated by loss of p300/CBP: lipid metabolic

process, regulation of hormone levels, and ion transport (Figure 6B and Supplementary Table 4).

Transcription factor target prediction from Webgestalt showed that all three gene sets were

significantly enriched for the predicted transcription factor Hnf1α (Figure 6C and Supplementary

Table 5). We also performed GSEA by comparing our gene sets to published downregulated

genes in mouse islets lacking factors important for beta cell development and function, including

Pdx1, NeuroD1, Hnf1α, Pax6, MafA, Nkx6.1 and Nkx2.2 (2430). Our gene sets overlapped

more significantly with the gene sets of Hnf1α and Nkx2.2, followed by MafA, Nkx6.1, Pdx1

and NeuroD1 (Figure 6D; Supplementary Table 6 and 7). Diabetes Page 14 of 46

Since Hnf1α could recruit p300/CBP for coactivation (31), we further examined the genes that

overlap between the Hnf1α gene set and the gene sets we had defined as downregulated genes in p300null islets, CBPnull islets and CBPHet; p300KO islets. Tmem27, a known Hnf1αmediated

regulator of beta cell proliferation (32), was reduced in all three models. Other loci

downregulated in Hnf1αnull islets including Pklr, Slc2a2 and G6pc2, were also downregulated

in triallelic islets as validated by qPCR (Figure 6E) (24). Beta cell transcription factors were not

specifically downregulated in either p300null or CBPnull islets, whereas Hnf4a, Hnf1b and

NeuroD1 were downregulated in triallelic islets (Supplementary Table 8). Insulin processing

genes were not altered in the biallelic mouse islets. Ins1 and Ins2 mRNAs were normally

expressed in in the biallelic mouse islets, although both were downregulated by more than 50%

in the triallelic islets.

Since p300/CBP coactivate transcription factors in part by acetylating H3K27 at target promoters

and enhancers, we hypothesized that the loss of p300/CBP would reduce H3K27 acetylation at

the loci downregulated in Hnf1αnull islets. We assessed the acetylation and methylation statuses

of H3K27 at various loci using lowinput native ChIP, and found that there were significantly

less H3K27Ac at the promoters of G6pc2, Hnf4a, Pklr and Tmem27 in the triallelic islets (Figure

6F; negative loci in Supplementary Figure 6B). Pdx1associated genes also showed reduced

H3K27Ac at their promoters and enhancers in the triallelic islets (Figure 6G). The H3K27Ac

levels at these loci were reduced in CBPnull islets, although the reduction did not reach

statistical significance. These locispecific H3K27Ac levels clearly correlated with the total

dosages of p300/CBP in the cells. We confirmed an approximately 60% reduction of H3K27Ac

globally in the triallelic islet nuclei (Figure 6H). Total and locispecific H3K27me3 levels were Page 15 of 46 Diabetes

unaffected in triallelic islets (Figure 6H & Supplementary Figure 6B). Overall, the reduced

dosages of p300/CBP impaired coactivation of downregulated genes in Hnf1αnull islets, which

we attribute to reductions in global and locispecific H3K27Ac levels.

Discussion

In this study, the loss of either p300 or CBP alone in the pancreatic islets was sufficient to

perturb whole body glucose homeostasis. Mice lacking p300 or CBP in islets developed similar

beta cell phenotypes including reduced beta cell area and insulin content. Mechanistically,

p300/CBP are known to coactivate Pdx1, NeuroD1, Hnf4α and Hnf1α/β in vitro (3335). Our

RNAseq data suggested that genes downregulated in Hnf1αnull islets became downregulated

once the dosages of either p300 or CBP were reduced in the islets. Hnf1α/β are homeobox

transcription factors that are critical for pancreas and beta cell development (36,37). In particular,

impaired Hnf1α coactivation in our mouse models could attenuate beta cell proliferation through

genes such as Tmem27 (32). The role of Hnf1α in alpha cells remains unclear, although high

levels of HNF1α were found in FACSsorted human alpha cells, thereby implying p300/CBP

might also regulate aspects of alpha cell biology, such as proliferation, through HNF1α (38).

p300/CBP bind to Hnf1α/β through their transactivation domains, and coactivate their

downstream targets by acetylating the histones bound to regulatory elements affiliated with these

targets (39). The observed loss of H3K27Ac in triallelic islets at loci downregulated in Hnf1α

null islets appears to be in line with such a mechanism.

While Hnf1α might be one of the targets in p300null/CBPnull islets, coactivation of other

transcription factors could also account for the phenotypes of p300/CBPnull islets. Our data Diabetes Page 16 of 46

suggest that p300/CBP do not appear to have major importance in the development of delta cells

or epsilon cells. The lack of effect on delta cells and epsilon cells in the double KO mice shows

striking similarity to the phenotypes of Nkx2.2null mice (40,41). Although Nkx2.2 is not known

to interact with p300/CBP, the significant overlapping between the gene set of Nkx2.2 and our p300/CBP gene sets suggested that Nkx2.2 might mediate some of the phenotypes seen in the p300/CBP mutant mice. Intriguingly, Nkx2.2 is mainly known for its repressor function, so p300/CBP might be recruited by Nkx2.2 to initiate its activator function instead (30,42). In the future, it will be interesting to explore whether p300/CBP interact physically with Nkx2.2 and acetylate the H3K27 residues at Nkx2.2associated loci, and whether the genomic occupancy of

Nkx2.2 or Hnf1α in islets is affected by p300 deletion.

Both Ins1 and Ins2 mRNAs were reduced in the triallelic p300/CBP islets but not in p300null islets or CBPnull islets. Reduced transcriptional activities of MafA and Nkx6.1, which are not known to recruit p300/CBP previously, might contribute indirectly to the reduced insulin gene expression seen in triallelic mice. Alternatively, p300/CBP might regulate insulin gene expression by binding to Pdx1 and NeuroD1 (6,35). The acetylation of H3K27 at the Ins1 promoter correlated with the dosages of p300/CBP present in the islets. The reduced insulin gene expression in triallelic islets could be a consequence of less p300/CBP available to beta cell transcription factors, which in turns impairs the acetylation of H3K27 at insulin promoters.

Taken together, p300/CBP may coordinate transcriptional networks in beta cells by coactivating various beta cell transcription factors, perhaps through Hnf1α/β and/or Nkx2.2. Mutations in many of these transcription factors are known to cause monogenic diabetes, including HNF1A, Page 17 of 46 Diabetes

HNF1B, PDX1 and NEUROD1 (43), suggesting that p300/CBP could have relevancy to the

underlying pathophysiology.

Overall, mice lacking p300 or CBP alone in islets developed glucose intolerance and

hypoinsulinemia associated with reduced islet area and insulin content. Mice lacking three copies

of p300/CBP in islets developed similar yet exacerbated phenotypes. Mice lacking all copies of

p300/CBP died postnatally due to their failure to establish beta cell mass. Islet genes mediated by

p300/CBP overlapped significantly with genes downregulated in islets lacking transcription

factors such as Hnf1α and Nkx2.2. p300/CBP expression was required to acetylate H3K27 at the

loci downregulated in Hnf1αnull islets including Slc2a2, Pklr, Hnf4a, and particularly Tmem27,

which could regulate beta cell proliferation. Thus, the expression of p300/CBP family of

coactivators in islets is critical to drive beta cell genesis and to maintain beta cell proliferation

and insulin content. In the pancreatic endocrine lineage, p300 and CBP serve as functionally

similar yet limiting cofactors to coordinate various islet transcription factors and maintain whole

body glucose homeostasis.

Acknowledgements

CKW and WTG conceived the study and designed the experiments. CKW wrote the manuscript.

CKW and AKW generated and analyzed the data. DSL helped with calcium imaging experiment.

All authors contributed to the study design and revised the article’s intellectual content. All

authors revised the manuscript and approved the final version. CKW and WTG are the

guarantors of this work and, as such, have full access to all the data in the study and take

responsibility for the integrity of the data and the accuracy of the data analysis. Diabetes Page 18 of 46

We are grateful for the institutional support from the The Canucks for Kids Childhood Diabetes

Laboratories at BCCHR and Dr. Jingsong Wang (BCCHR) for technical assistance at Imaging

Core. We would also like to thank Ryan Vander Werff and the UBC BRC Sequencing Core for

their support on RNAseq experiments, Dr. Julie Brind’Amour and Dr. Matthew Lorincz (UBC)

for their advice on low input ChIP, and Dr. Lawryn Kasper (St. Jude Children’s Research

Hospital) for advice on p300 western blotting.

CKW’s salary is supported by a BCCH Research Institute graduate studentship, and WTG’s

investigator salary is supported by BCCHRI Intramural IGAP award. This study was supported by grants to WTG from NSERC (RGPIN 40257611) and CIHR (MOP119595 and PJT

148695).

The authors declare that there is no duality of interest associated with this manuscript.

References

1. Melloul D, Marshak S, Cerasi E. Regulation of insulin gene transcription. Diabetologia. SpringerVerlag; 2002 Mar;45(3):309–26.

2. Bedford DC, Brindle PK. Is histone acetylation the most important physiological function for CBP and p300? Aging (Albany NY). 2012 Apr;4(4):247–55.

3. Heinz S, Romanoski CE, Benner C, Glass CK. The selection and function of cell type specific enhancers. Nat Rev Mol Cell Biol. 2015 Feb 4;16(3):144–54.

4. Bose DA, Donahue G, Reinberg D, Shiekhattar R, Bonasio R, Berger SL. RNA Binding to CBP Stimulates Histone Acetylation and Transcription. Cell. 2017 Jan 12;168(12):135– 149.e22.

5. Qiu Y, Guo M, Huang S, Stein R. Acetylation of the BETA2 transcription factor by p300 associated factor is important in insulin gene expression. J Biol Chem. 2004 Mar Page 19 of 46 Diabetes

12;279(11):9796–802.

6. Sampley ML, Ozcan S. Regulation of insulin gene transcription by multiple histone acetyltransferases. DNA Cell Biol. 2012 Jan;31(1):8–14.

7. Bompada P, Atac D, Luan C, Andersson R, Omella JD, Laakso EO, et al. Histone acetylation of glucoseinduced thioredoxininteracting protein gene expression in pancreatic islets. Int J Biochem Cell Biol. 2016 Oct 29;81:82–91.

8. Hussain MA, Porras DL, Rowe MH, West JR, Song WJ, Schreiber WE, et al. Increased pancreatic betacell proliferation mediated by CREB binding protein gene activation. Mol Cell Biol. American Society for Microbiology; 2006 Oct;26(20):7747–59.

9. Schonhoff SE, GielMoloney M, Leiter AB. Neurogenin 3expressing progenitor cells in the gastrointestinal tract differentiate into both endocrine and nonendocrine cell types. Dev Biol. 2004 Jun 15;270(2):443–54.

10. Kasper LH, Fukuyama T, Biesen MA, Boussouar F, Tong C, de Pauw A, et al. Conditional knockout mice reveal distinct functions for the global transcriptional coactivators CBP and p300 in Tcell development. Mol Cell Biol. 2006 Feb;26(3):789– 809.

11. Wong CK, Botta A, Pither J, Dai C, Gibson WT, Ghosh S. A highfat diet rich in corn oil reduces spontaneous locomotor activity and induces insulin resistance in mice. J Nutr Biochem. 2015 Apr;26(4):319–26.

12. Li DS, Yuan YH, Tu HJ, Liang QL, Dai LJ. A protocol for islet isolation from mouse pancreas. Nat Protoc. 2009;4(11):1649–52.

13. Luciani DS, White SA, Widenmaier SB, Saran VV, Taghizadeh F, Hu X, et al. Bcl2 and BclxL suppress glucose signaling in pancreatic βcells. Diabetes. American Diabetes Association; 2013 Jan;62(1):170–82.

14. Schindelin J, ArgandaCarreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an opensource platform for biologicalimage analysis. Nat Methods. 2012 Jun 28;9(7):676– 82.

15. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. BioMed Central; 2013 Apr 25;14(4):R36.

16. Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in highthroughput experiments. Bioinformatics. 2012 Mar 15;28(6):882–3.

17. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNAseq data with DESeq2. Genome Biol. BioMed Central; 2014;15(12):550. Diabetes Page 20 of 46

18. Hulsen T, de Vlieg J, Alkema W. BioVenn a web application for the comparison and visualization of biological lists using areaproportional Venn diagrams. BMC Genomics. BioMed Central; 2008 Oct 16;9(1):488.

19. Wang J, Duncan D, Shi Z, Zhang B. WEBbased GEne SeT AnaLysis Toolkit (WebGestalt): update 2013. Nucleic Acids Research. 2013 Jul;41(Web Server issue):W77–83.

20. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using realtime quantitative PCR and the 2(Delta Delta C(T)) Method. Methods. 2001 Dec;25(4):402–8.

21. Kasper LH, Boussouar F, Ney PA, Jackson CW, Rehg J, van Deursen JM, et al. A transcriptionfactorbinding surface of coactivator p300 is required for haematopoiesis. Nature. 2002 Oct 17;419(6908):738–43.

22. Brind'Amour J, Liu S, Hudson M, Chen C, Karimi MM, Lorincz MC. An ultralowinput native ChIPseq protocol for genomewide profiling of rare cell populations. Nat Commun. 2015 Jan 21;6:6033.

23. Song J, Xu Y, Hu X, Choi B, Tong Q. Brain expression of Cre recombinase driven by pancreasspecific promoters. Genesis. 2010 Nov;48(11):628–34.

24. Servitja JM, Pignatelli M, Maestro MA, Cardalda C, Boj SF, Lozano J, et al. Hnf1alpha (MODY3) controls tissuespecific transcriptional programs and exerts opposed effects on cell growth in pancreatic islets and liver. Mol Cell Biol. American Society for Microbiology; 2009 Jun;29(11):2945–59.

25. Gu C, Stein GH, Pan N, Goebbels S, Hornberg H, Nave KA, et al. Pancreatic beta cells require NeuroD to achieve and maintain functional maturity. Cell Metabolism. 2010 Apr 7;11(4):298–310.

26. Taylor BL, Liu FF, Sander M. Nkx6.1 is essential for maintaining the functional state of pancreatic beta cells. Cell Rep. 2013 Sep 26;4(6):1262–75.

27. Gao T, McKenna B, Li C, Reichert M, Nguyen J, Singh T, et al. Pdx1 maintains β cell identity and function by repressing an α cell program. Cell Metabolism. 2014 Feb 4;19(2):259–71.

28. Hang Y, Yamamoto T, Benninger RKP, Brissova M, Guo M, Bush W, et al. The MafA transcription factor becomes essential to islet βcells soon after birth. Diabetes. American Diabetes Association; 2014 Jun;63(6):1994–2005.

29. Swisa A, Avrahami D, Eden N, Zhang J, Feleke E, Dahan T, et al. PAX6 maintains β cell identity by repressing genes of alternative islet cell types. J Clin Invest. American Society for Clinical Investigation; 2017 Jan 3;127(1):230–43.

30. Gutiérrez GD, Bender AS, Cirulli V, Mastracci TL, Kelly SM, Tsirigos A, et al. Pancreatic β cell identity requires continual repression of nonβ cell programs. J Clin Page 21 of 46 Diabetes

Invest. American Society for Clinical Investigation; 2017 Jan 3;127(1):244–59.

31. Soutoglou E, Papafotiou G, Katrakili N, Talianidis I. Transcriptional activation by hepatocyte nuclear factor1 requires synergism between multiple coactivator proteins. J Biol Chem. 2000 Apr 28;275(17):12515–20.

32. Akpinar P, Kuwajima S, Krützfeldt J, Stoffel M. Tmem27: a cleaved and shed plasma membrane protein that stimulates pancreatic beta cell proliferation. Cell Metabolism. 2005 Dec;2(6):385–97.

33. Dell H, HadzopoulouCladaras M. CREBbinding protein is a transcriptional coactivator for hepatocyte nuclear factor4 and enhances apolipoprotein gene expression. J Biol Chem. 1999 Mar 26;274(13):9013–21.

34. Ban N, Yamada Y, Someya Y, Miyawaki K, Ihara Y, Hosokawa M, et al. Hepatocyte nuclear factor1alpha recruits the transcriptional coactivator p300 on the GLUT2 gene promoter. Diabetes. 2002 May;51(5):1409–18.

35. Qiu Y, Guo M, Huang S, Stein R. Insulin gene transcription is mediated by interactions between the p300 coactivator and PDX1, BETA2, and E47. Mol Cell Biol. 2002 Jan;22(2):412–20.

36. Pontoglio M, Sreenan S, Roe M, Pugh W, Ostrega D, Doyen A, et al. Defective insulin secretion in hepatocyte nuclear factor 1alphadeficient mice. J Clin Invest. American Society for Clinical Investigation; 1998 May 15;101(10):2215–22.

37. Haumaitre C, Barbacci E, Jenny M, Ott MO, Gradwohl G, Cereghini S. Lack of TCF2/vHNF1 in mice leads to pancreas agenesis. Proc Natl Acad Sci USA. 2005 Feb 1;102(5):1490–5.

38. Bramswig NC, Everett LJ, Schug J, Dorrell C, Liu C, Luo Y, et al. Epigenomic plasticity enables human pancreatic α to β cell reprogramming. J Clin Invest. 2013 Mar;123(3):1275–84.

39. Barbacci E, Chalkiadaki A, Masdeu C, Haumaitre C, Lokmane L, Loirat C, et al. HNF1beta/TCF2 mutations impair transactivation potential through altered coregulator recruitment. Hum Mol Genet. 2004 Dec 15;13(24):3139–49.

40. Prado CL, PughBernard AE, Elghazi L, SosaPineda B, Sussel L. Ghrelin cells replace insulinproducing beta cells in two mouse models of pancreas development. Proc Natl Acad Sci USA. 2004 Mar 2;101(9):2924–9.

41. Mastracci TL, Wilcox CL, Arnes L, Panea C, Golden JA, May CL, et al. Nkx2.2 and Arx genetically interact to regulate pancreatic endocrine cell development and endocrine hormone expression. Dev Biol. 2011 Nov 1;359(1):1–11.

42. Doyle MJ, Sussel L. Nkx2.2 regulates betacell function in the mature islet. Diabetes. American Diabetes Association; 2007 Aug;56(8):1999–2007. Diabetes Page 22 of 46

43. Fajans SS, Bell GI, Polonsky KS. Molecular mechanisms and clinical pathophysiology of maturityonset diabetes of the young. N Engl J Med. 2001 Sep 27;345(13):971–80.

Figure Legends

Figure 1. Mice lacking p300 in pancreatic islets develop glucose intolerance due to

hypoinsulinemia. (A) Western blotting for p300 and CBP in isolated WT or p300null islet

nuclear extracts. TBP was used as loading control. The experiment was replicated once. (B)

Representative immunofluorescence images of p300 and CBP in the pancreatic islets of WT and p300IsletKO mice. n = 3. Scale bar = 50 µm. (C) Glucose tolerance test of eightweekold WT or p300IsletKO mice. n = 8 – 9. (D) Insulin tolerance test of nineweekold WT and p300IsletKO mice. n

= 5 – 7. (E) Plasma insulin measurement before and 15 minutes after glucose injection. n = 4 – 5.

Twoway ANOVA for Figure C, D and E. * p < 0.05, ** p < 0.01, *** p < 0.001

Figure 2. p300IsletKO mice display defects in islet mass and islet insulin content.

(A) Quantification of beta cell, alpha cell and delta cell area of adult WT and p300IsletKO mice pancreata. n = 8 for beta cells; n = 4 – 5 for alpha cells and delta cells. (B) Islet density of adult

WT and p300IsletKO mice pancreata. n = 5. (C) Insulin, glucagon and somatostatin content of WT and p300null islets. n = 5 – 6. (D) Glucosestimulated insulin secretion assays on isolated WT and p300null islets. n = 5 – 7. (E) Fura2 calcium imaging on isolated WT and p300null islets.

n = 3. Student’s t test for Figure A, B, C and D. Twoway ANOVA for Figure E. * p < 0.05

Figure 3. CBPIsletKO mice share similar phenotypes with p300IsletKO mice. (A) Representative

immunofluorescence images of p300 and CBP in the pancreatic islets of WT and CBPIsletKO

mice. Scale bar = 50 µm. (B) Glucose tolerance test of eightweekold WT and CBPIsletKO mice. Page 23 of 46 Diabetes

n = 4 – 11. (C) Insulin tolerance test of nineweekold WT and CBPIsletKO mice. n = 5 – 8. (D)

Plasma insulin measurement of WT and CBPIsletKO mice before and 15 minutes after glucose

injection. n = 6 – 7. (E) Quantification of beta cell, alpha cell and delta cell area of adult WT and

CBPIsletKO mice pancreata as % of total pancreas area. n = 4 – 5 for alpha cells and beta cells; n =

5 – 6 for delta cells. (F) Islet insulin, glucagon and somatostatin content of WT and CBPnull

islets as quantified by ELISA. n = 4. (G) Perifusion assay for insulin secretion of WT and CBP

null islets. n = 3. Twoway ANOVA for Figure B, C and D. Student’s t test for Figure E and F. *

p < 0.05, ** p < 0.01, *** p < 0.001

Figure 4. Triallelic deletion of p300/CBP in islets leads to severe glucose intolerance. (A)

Intraperitoneal glucose tolerance test of eightweekold WT and CBPHet; p300KO mice. n = 6. (B)

Insulin tolerance test of adult WT and CBPHet; p300KO mice. n = 6 – 7. (C) Plasma insulin

measurement of WT and CBPHet; p300KO mice before and 15 minutes after glucose injection. n =

7. (D) Quantification of beta cell, alpha cell and delta cell area of adult WT and CBPHet; p300KO

mouse pancreata as % of total pancreas area. n = 4 – 6 for alpha cells and beta cells; n = 4 for

delta cells. (E) Islet insulin content of WT and CBPHet; p300KO islets as quantified by ELISA. n

= 4. Twoway ANOVA for Figure A, B and C. Student’s t test for Figure D and E. * p < 0.05,

** p < 0.01, *** p < 0.001

Figure 5. Expression of p300/CBP is necessary for beta cell and alpha cell development. (A)

Quantification of beta cell, alpha cell and delta cell area of P7 WT and p300IsletKO mouse

pancreata as % of total pancreas area. n = 4 – 6. (B) Representative immunofluorescence images

of insulin, glucagon and Ki67 in P7 WT and p300IsletKO mouse pancreata. Scale bar = 50 µm. (C) Diabetes Page 24 of 46

Quantification of Ki67+ beta cells, alpha cells and all pancreatic cells in P7 WT and p300IsletKO

mouse pancreata as % of total beta cells, alpha cells and total pancreatic cells. n = 8. (D)

Quantification of beta cell, alpha cell, delta cell, epsilon cell and chromogranin Apositive pan

endocrine cell area of P0 WT and p300/CBP double knockout (dKO) mouse pancreata as % of

total pancreas area. n = 3. (E) Representative immunofluorescence images of insulin, glucagon, somatostatin, ghrelin, chromogranin A and DAPI in P0 WT and p300/CBP double knockout mouse pancreata. Scale bar = 50 µm. Student’s t test for Figure A, C and D. * p < 0.05, ** p <

0.01

Figure 6. Loss of p300/CBP impairs coactivation of Hnf1α through reduced H3K27 acetylation. (A) Venn diagram of overlapping downregulated genes of p300IsletKO, CBPIsletKO and

CBPHet; p300KO mouse islets as compared to WT islets. (B) The three Biological Processes GO

terms commonly overrepresented in the downregulated genes of p300IsletKO, CBPIsletKO and

CBPHet; p300KO mouse islets. All significantly overrepresented GO terms and their associated genes can be found in Supplementary Table 4. (C) Transcription factor target analysis by

Webgestalt on the downregulated genes of p300IsletKO, CBPIsletKO and CBPHet; p300KO mouse

islets. Hnf1α was commonly overrepresented in all three gene sets. (D) Gene set enrichment

analysis on downregulated gene sets derived from microarray or RNAseq data of mice lacking beta cell transcription factors in islets or beta cells. Random 1 and 2 were control gene lists

generated randomly from the 15,999 genes in the WT reference list. (E) qPCR of islet Hnf1α

associated genes in WT and CBPHet; p300KO mouse islets. n = 5 – 6. (F) Low input native ChIP

for H3K27Ac at Hnf1αassociated genes in WT, CBPIsletKO and CBPHet; p300KO mouse islets. n =

3 – 5. (G) Low input native ChIP for H3K27Ac at Pdx1associated loci in in WT, CBPIsletKO and Page 25 of 46 Diabetes

CBPHet; p300KO mouse islets. n = 3 – 5. (H) Representative immunofluorescence images of

insulin, H3K27Ac and H3K27me3 in WT and CBPHet; p300KO mouse islets. Scale bar = 50 µm.

Student’s t test for Figure E. Oneway ANOVA for Figure F and G. * p < 0.05, ** p < 0.01, ***

p < 0.001 Diabetes Page 26 of 46

722x541mm (72 x 72 DPI)

Page 27 of 46 Diabetes

722x541mm (72 x 72 DPI)

Diabetes Page 28 of 46

722x541mm (72 x 72 DPI)

Page 29 of 46 Diabetes

722x541mm (72 x 72 DPI)

Diabetes Page 30 of 46

722x541mm (72 x 72 DPI)

Page 31 of 46 Diabetes