The homeobox transcription factor Nkx6.1 is sufficient to increase functional β-cell mass, where functional β-cell mass refers to the combination of β-cell proliferation, glucose-stimulated insulin secretion (GSIS) and β-cell survival. Here, we demonstrate that the histone deacetylase 1 (HDAC1), which is an early target of Nkx6.1, is sufficient to increase functional β-cell mass. We show that HDAC activity is necessary for Nkx6.1-mediated proliferation, and that HDAC1 is sufficient to increase β-cell proliferation in primary rat islets and the INS-1 832/13 β-cell line. The increase in HDAC1-mediated proliferation occurs while maintaining GSIS and increasing β-cell survival in response to apoptotic stimuli. We demonstrate that HDAC1 overexpression results in decreased expression of the cell cycle inhibitor Cdkn1b/p27 which is essential for inhibiting the G1 to S phase transition of the cell cycle. This corresponds with increased expression of key cell cycle activators, such as Cyclin A2, Cyclin B1 and E2F1, which are activated by activation of the Cdk4/Cdk6/Cyclin D holoenzymes due to down-regulation of Cdkn1b/p27. Finally, we demonstrate that overexpression of Cdkn1b/p27 inhibits HDAC1-mediated β-cell proliferation. Our data suggest that HDAC1 is critical for the Nkx6.1-mediated pathway that enhances functional β-cell mass.
Type 1 (T1D) and Type 2 diabetes (T2D) affect more than 9% of the American population, with the incidences continuing to increase at a startling rate. While autoimmunity is unique to T1D, and insulin resistance of the liver, muscle and adipose tissue is unique to T2D, both forms ultimately result in severe decreases in functional β-cell mass. Therefore, a cure for both types of diabetes must ultimately increase functional β-cell mass through increased β-cell proliferation, insulin secretion and cell survival.
β-cell proliferation rates are highest during the embryonic period and soon after birth. These rates decrease rapidly with age, decreasing to less than 1% of the β-cell population demonstrating replication markers at any given point [1–3]. However, studies in rodents and humans have clearly demonstrated that β-cell proliferation can increase during pregnancy and obesity [4–6]. These findings demonstrate that while the molecular mechanisms that control β-cell proliferation are tightly regulated, they are also intact and given the correct set of signals can be re-engaged resulting in increased functional β-cell mass through enhanced β-cell growth.
We have previously shown that the homeobox β-cell transcription factor Nkx6.1 is sufficient to induce proliferation of human and rodent β-cells [7,8]. Developmental studies have shown that the rapid period of embryonic β-cell replication is dependent on Nkx6.1 , and that Nkx6.1 inactivation causes rapid-onset diabetes and β-cell loss . Our published data show that Nkx6.1 overexpression results in significant induction of β-cell proliferation. Interestingly, while Nkx6.1 protein levels are observed as early as 24 h after transduction with AdCMV-Nkx6.1, there is a 72-h delay until β-cell proliferation is observed. This suggests that early Nkx6.1 target genes are essential to permit β-cells to re-enter the cell cycle and undergo cellular replication.
Histone deacetylases (HDACs) function by removing acetyl groups from histone (and non-histone protein) lysine residues. Histone acetylation has a direct effect on transcription, as acetylated histones increase the ability of RNA polymerase to access DNA thus enhancing gene transcription at certain sites . Therefore, HDACs down-regulate gene transcription by removing acetyl groups from histone tails, resulting in tighter histone DNA binding and impediment of RNA polymerase access. There are four classes of HDACs, HDAC 1, 2, 3 and 8, comprising HDAC class 1 .
HDAC1 plays an important role in cell cycle control. Deletion of HDAC1 results in embryonic lethality. Further studies have demonstrated that loss of HDAC1 in embryonic stem cells results in increased expression of the cell cycle inhibitors Cdkn1a/p21 and Cdkn1b/p27 and decreased proliferation. Ablation of Cdkn1a/p21 and Cdkn1b/p27 rescues the observed proliferation phenotype [13,14]. The observation that HDAC1 controls Cdkn1a/p21 and Cdkn1b/p27 gene expression, and ultimately cellular proliferation has been solidified by further studies that have also shown that HDAC1 deletion increases Cdkn1b/p27 expression [15–19], and others that show that HDAC1 overexpression decreases Cdkn1b/p27 expression [20,21].
Cellular proliferation is controlled by cell cycle activators and repressors. The cell cycle is driven by the concerted efforts of Cyclin-dependent-kinases (Cdks) and their respective Cyclins. Upon binding and activation, these Cdk/Cyclin complexes induce cellular replication by phosphorylating key genes that permit progression through the cell cycle [22,23]. Cdk/Cyclin complexes are inhibited by Cyclin-dependent kinase inhibitors (Cdkns). These fall into two families, the Cdkn1 family which comprises Cdkn1a/p21, Cdkn1b/p27 and Cdkn1c/p57 and the Cdkn2 family comprises Cdkn2a/p16, Cdkn2b/p15, Cdkn2c/p18 and Cdkn2d/p19. These inhibitors block cell cycle progression by binding to the Cdk/cyclin complexes, such as Cyclin D/Cdk4, and ultimately impede cellular replication .
In this study, we demonstrate that HDAC1 expression is increased in response to Nkx6.1 overexpression. We show that HDAC activity is necessary for Nkx6.1-mediated proliferation, and that HDAC1 is sufficient to induce functional β-cell proliferation. We also show that HDAC1 overexpression in β-cells maintains glucose-stimulated insulin secretion (GSIS) and increases β-cell survival when exposed to apoptotic stimuli. Furthermore, we establish that overexpression of HDAC1 results in down-regulation of the cell cycle inhibitor Cdkn1b/p27, which corresponds with increased expression of cell cycle activators. Finally, we demonstrate that overexpression of Cdkn1b/p27 is sufficient to impede HDAC1-mediated β-cell proliferation. Our data demonstrate that HDAC1 is necessary for Nkx6.1-mediated β-cell proliferation and suggest that HDAC1 could be leveraged as a mechanism for increasing functional β-cell mass as a treatment for diabetes.
INS-1 832/13 rat insulinoma β-cells were cultured, as previously described . Briefly, the INS-1-derived 832/13 rat β-cell line was maintained in complete RPMI 1640 medium with l-glutamine and 11.2 mM glucose supplemented with 50 U/ml penicillin, 50 µg/ml streptomycin, 10 mM HEPES, 10% fetal bovine serum and INS-1 supplement. For chemical inhibition of HDAC1, cells were cultured with 200 nM trichostatin A (TSA) in DMSO for 18 h .
Adenoviral cloning and preparation
Recombinant AdCMV-HDAC1 and AdCMV-Cdkn1b/p27 adenoviruses were generated and purified as previously described . HDAC1 overexpression was verified by RT-PCR (reverse transcription-polymerase chain reaction) and western blotting. Adenoviruses expressing AdCMV-Nkx6.1 and AdCMV-GFP have been described elsewhere [7,28,29].
Islet isolation and culture
Wistar rat breeding pairs were purchased from Harlan and maintained on standard chow diet (Teklad 7001; Harlan). Pups were weaned at 21 days. Male rats were allowed to feed ad libitum and were maintained on a 12-h light-dark cycle. Rats were age- and weight-matched for all islet experiments. Pancreatic islets were isolated from 5-week-old male rats as previously described . Primary rat islets were cultured in RPMI 1640 and supplemented with 10% FBS, 1% Fungizone antimycotic (Life Technologies) and 1% HEPES. Islet medium was changed every 24 h. All animal studies were approved and performed in accordance with Brigham Young University's IACUC guidelines (Protocol #16-0902). Islet isolation, culture and transduction were completed, as previously described [25,29,30].
Human pancreatic islets were provided by the NIDDK-funded Integrated Islet Distribution Program (IIDP) at City of Hope, NIH Grant # 2UC4DK098085. Human islets were cultured under conditions recommended by the IIDP, as previously described .
DNA synthesis rates in INS-1 832/13 β-cells, primary rat islets and primary human islets were measured by [methyl-3H]-thymidine incorporation, and were completed, as previously described [25,29,30]. Briefly, for INS-1 832/13 β-cells, cells were labeled with 3H-thymidine for 15 min, followed by processing for DNA 3H-thymidine measurements. For primary rat and human islets, islets were labeled for 24 h, beginning at 72 h of culture, after which islets were harvested at 96 h of culture and processed for DNA 3H-thymidine measurements.
HDAC activity assay
INS-1 832/13 β-cells were transduced with AdCMV-GFP, AdCMV-Nkx6.1 and AdCMV-HDAC1, or left untreated (No Virus-NV). After transduction, cells were cultured for 96 h, cells were harvested and HDAC activity was measured using the HDAC Activity Assay Kit, following the manufacturer's instructions (BioVision).
Cell viability assays
INS-1 832/13 β-cells were transduced with AdCMV-GFP or AdCMV-HDAC1 or left untreated (No Virus-NV). Twenty-four hours post adenoviral transduction, the cells were treated with apoptotic stimulants for 18 h (9.0 µM etoposide, 0.3 µM thapsigargin or 2.0 µM camptothecin). At the end of the treatment, time and percent cell viability were determined by completing cell counts of the treated cells when compared with untreated cells, as previously described . Alamar Blue and MTT assays were completed to measure viability after HDAC1 overexpression following the manufacturer's instructions, as previously described .
siRNA-mediated gene knockdown
siRNA-mediated gene knockdown of HDAC1, HDAC2, HDAC3, HDAC4, HDAC5, HDAC6, HDAC7, HDAC8, HDAC9 and HDAC10 was completed using OnTargetPlus siRNA, and following the manufacturer's instructions, as previously described (Dharmacon) .
Rat islet EdU (Invitrogen, Waltham, MA, U.S.A.) incorporation of dispersed rat islets, with DAPI and insulin (DAKO, Carpinteria, CA, U.S.A.) counterstaining was completed, as previously described . Briefly, islets were labeled with EdU for 24 h beginning at 72 h of culture, and islets were harvested at 96 h. Islets were dispersed using trypsin and plated on collagen-coated coverslips, after which cells were fixed in 4%PFA, permeabilized with Triton X-100 and processed for staining. Three sections containing ≥400 nuclei were evaluated for EdU signals using IMAGEJ software for each condition (National Institutes of Health, Bethesda, MD, U.S.A.).
Glucose-stimulated insulin secretion
GSIS assays from rat islets or INS-1 832/13 cells were completed, as previously described [30,31]. Briefly, INS-1 832/13 β-cells were grown to confluency, washed with PBS and preincubated in secretion assay buffer (SAB) for 1.5 h (114 mM NaCl, 4.7 mM KCl, 1.2 mM KH2PO4, 1.16 mM MgSO4, 20 mM HEPES, 2.5 mM CaCl2, 0.2% BSA, pH 7.2) containing 2.5 mM glucose. GSIS was performed by incubating quadruplicate replicate wells of cells cultured with test compounds at 0–100 µM in SAB containing 2.5 mM glucose for 1 h (basal), followed by 1 h in SAB with 16.7 mM glucose (glucose stimulation), followed by a collection of the respective buffers, as previously described. For total insulin content, β-cells stimulated with 16.7 mM glucose for 1 h were lysed in RIPA buffer with protease inhibitors (Life Technologies). Secreted insulin and total insulin was measured in SAB using a rat insulin RIA kit (MP Biomedicals), and normalized to total cellular protein concentration (determined by BCA assay), as previously described. Secreted insulin and total insulin was measured using a rat insulin RIA kit (MP Biomedicals, Santa Ana, CA, U.S.A.) .
Reverse transcription-polymerase chain reaction
RNA was harvested using TriReagent (Life Technologies) and cDNA was synthesized using the High-Capacity cDNA Reverse Transcription kit (Life Technologies), per the manufacturer's instructions and as previously described [22,25]. Real-time PCR was performed using the Life Technologies One Step Plus Sequence Detection System and Software (Life Technologies) using Taq-man assays on demand probes and primers for GFP, Nkx6.1, HDAC1, Cdet1, Cdc2a, Cdc6, Cdk2, Cyclin A2, Cyclin B1, Cyclin D2, Cyclin E1, E2F1, FoxM1, AURKA, HDAC7, Nr4a1, Nr4a3, Cdkn1a, Cdkn1b, Cdkn1c, Cdkn2a, Cdkn2b, Cdkn2c, Cdkn2d and PPIA (Life Technologies).
Cells or islets were washed in PBS and harvested in RIPA buffer followed by sonication. Protein concentration was quantified by BCA, and 30 µg was run per sample on 10% gels. Gel transfer, probing and visualization were completed, as previously described . Blots were probed with anti-Tubulin (ab6046, Abcam, Cambridge, MA, U.S.A., 1 : 1000), HDAC1 (ab19845-100, Abcam, Cambridge, MA, U.S.A., 1 : 1000), GFP (ab290, Abcam, Cambridge, MA, U.S.A., 1 : 1000) and Cdkn1b/p27 (3686S, Cell Signaling, Danvers, MA, U.S.A. 1 : 1000) antibodies. Bands were detected using a Licor Odyssey CLx and quantified using Image Studio.
All results are expressed as mean ± SEM. Data were analyzed using two-tailed Student t-test or two-way ANOVA where appropriate (Prism Software). Statistical significance was defined as P < 0.05.
Nkx6.1 induces HDAC1 expression in primary rat islets
We have shown that primary rat and human β-cells proliferate 72 h after adenoviral transduction with AdCMV-Nkx6.1 , suggesting Nkx6.1 target genes are necessary to induce β-cell proliferation. Analysis of previously published microarray data demonstrates that HDAC1 is induced within 48 h of Nkx6.1 overexpression . We measured HDAC1 mRNA levels in untreated primary rat islets immediately upon isolation (0 h) and of untreated islets or islets transduced with AdCMV-Nkx6.1 or AdCMV-GFP at 24, 48, 72 and 96 h after transduction. HDAC1 mRNA levels were significantly increased 48 h after Nkx6.1 overexpression and remained elevated throughout the 96-h time course (Figure 1A). Overexpression of Nkx6.1 in primary rat islets resulted in induction of HDAC1 when compared with untreated islets, or islets transduced with AdCMV-GFP (Figure 1B,C). These results demonstrate that HDAC1 is induced by Nkx6.1 overexpression in primary rat islets.
HDAC1 is induced by Nkx6.1 in β-cells.
HDAC1 inhibition impedes β-cell proliferation
As Nkx6.1 overexpression increases HDAC1 levels and given that Nkx6.1 overexpression induces primary β-cell proliferation, we hypothesized that HDAC activity may be necessary for β-cell proliferation. We tested this hypothesis by transducing primary rat islets with AdCMV-Nkx6.1, AdCMV-GFP or left untreated and cultured for 96-h, followed by 3H-thymidine incorporation to measure proliferation. After 72-h of culture, islets were treated with or without TSA, the pharmacological class I and II HDAC inhibitor. While Nkx6.1 induced a significant increase in proliferation, AdCMV-Nkx6.1 transduced islets that were treated with TSA failed to demonstrate proliferation (Figure 2A). This demonstrates that class I or II HDAC activity is necessary for Nkx6.1-mediated β-cell proliferation. To determine which of class I, IIa or IIb HDACs are necessary for β-cell proliferation, we knocked down expression of HDAC1, HDAC2, HDAC3, HDAC4, HDAC5, HDAC6, HDAC7, HDAC8, HDAC9 and HDAC10 in the INS-1 832/13 β-cell line. Only siRNA-mediated HDAC1 knockdown significantly impaired basal INS-1 832/13 β-cell proliferation (Figure 2B). Interestingly, HDAC8 knock down resulted in increased basal proliferation, which will be explored in the future. This suggests that HDAC1 may be needed for basal β-cell proliferation. These data demonstrate that HDAC activity, and more specifically HDAC1 activity, is critical for Nkx6.1-mediated proliferation.
HDAC1 inhibition impedes β-cell proliferation.
HDAC1 overexpression induces β-cell proliferation
To determine the effect of HDAC1 on β-cell proliferation and function, primary rat islets were transduced with AdCMV-GFP or AdCMV-HDAC1 or left untreated. HDAC1 protein levels were measured and demonstrated to increase 3-fold after overexpression (Figure 3A,B), with HDAC1 increasing 6-fold in INS-1 832/13 cells after adenoviral transduction (Supplemental Figure S1A,B). As anticipated, HDAC1 overexpression in INS-1 832/13 β-cells resulted in increased HDAC activity and improved cellular viability (Supplemental Figure S1C–E). The increased viability observed with HDAC1 overexpression could be the result of increased cell proliferation. HDAC1 overexpression in primary rat islets demonstrated increased proliferation as early as 48 h after adenoviral transduction, as measured by 3H-thymidine incorporation (Figure 3C). This proliferation is earlier than that observed with Nkx6.1, which occurred at 72 h after transduction, and corresponds with data demonstrating that HDAC1 is an early target of Nkx6.1. Similar results were observed with INS-1 832/13 β-cells transduced with serial dilutions of AdCMV-HDAC1, demonstrating a dose-dependent increase in HDAC1-mediated proliferation as measured 96 h post-transduction (Supplemental Figure S1F). To determine which islet cell type proliferates in response to HDAC1 overexpression, primary rat islets were transduced with AdCMV-GFP or AdCMV-HDAC1 or not transduced, and cultured with the thymidine analog EdU to mark replicating cells. Immunofluorescent analysis of labeled dispersed islets demonstrated that HDAC1 overexpression resulted in 2.5% insulin +EdU+ cells, while GFP overexpression resulted in less than 1% insulin + EdU+ cells. Furthermore, only 1.3% EdU+ cells were observed in the insulin-population, demonstrating greater β-cell proliferation (Figure 3D,E). Finally, a pilot experiment was completed to determine the effect of HDAC1 overexpression in human islets. Our results, using islets from five donors, demonstrate a significant increase in human islet proliferation in response to HDAC1 overexpression (Figure 3F). These data demonstrate that HDAC1 overexpression induces proliferation of β-cells, from the INS-1 832/13 β-cell line to primary rodent and human islets.
HDAC1 overexpression induces β-cell proliferation.
HDAC1 overexpression maintains glucose-stimulated insulin secretion
Our previous data demonstrated that Nkx6.1 overexpression not only induced β-cell proliferation but also enhanced glucose-stimulated insulin secretion (GSIS) . Therefore, we sought to determine the effect of HDAC1 overexpression on GSIS. INS-1 832/13 β-cells were transduced with AdCMV-GFP or AdCMV-HDAC or left untreated. Insulin secretion was measured 96 h after transduction. Significant changes were observed within each treatment group between non-stimulatory (2.5 mM glucose) and stimulatory (16.7 mM glucose) conditions; however, there were no significant differences between the group with the non-stimulatory or the stimulatory conditions (Figure 4A). Furthermore, HDAC1 overexpression resulted in no significant change in total insulin content (Figure 4B). These data demonstrate that HDAC1 overexpression maintains insulin secretion and insulin content, and ultimately does not inhibit β-cell function. Furthermore, as no changes are observed in GSIS, this precludes improvements to mitochondrial fitness or content. Finally, these results are consistent with previous findings that Nkx6.1-mediated improvement of GSIS was due to the expression of VGF, and that this was independent of the observed proliferation phenotype [29,31]. Our results place HDAC1 outside of the pathway necessary for enhancing GSIS.
Overexpression of HDAC1 maintains glucose-stimulated insulin secretion.
HDAC1 overexpression protects against apoptotic stimuli
To determine the effects of HDAC1 overexpression on cell survival, INS-1 832/13 β-cells were treated with AdCMV-GFP or AdCMV-HDAC1 or left untreated. Twenty-four hours after adenoviral transduction, cells were treated with camptothecin, thapsigargin or etoposide to induce apoptosis. Untransduced cells or AdCMV-GFP transduced cells demonstrated viability between 40 and 50% relative to vehicle-treated controls after treatment with the respective apoptotic stimuli. Interestingly, cells transduced with AdCMV-HDAC1 demonstrated a significant increase in cell viability when treated with etoposide or thapsigargin (Figure 5A,B) but not with camptothecin (Figure 5C). These results show that HDAC1 overexpression protects INS-1 832/13 β-cells from etoposide or thapsigargin-induced apoptosis.
HDAC1 overexpression protects β-cells from apoptosis.
HDAC1 overexpression induces changes in cell cycle gene expression
HDACs function by removing acetyl groups from lysine residues on histones. Decreased histone acetylation generally results in decreased gene expression, because chromatin assumes a more compact state . We hypothesized that HDAC1 overexpression would result in decreased expression of genes critical for inhibiting cell cycle progression, thus resulting in increased β-cell proliferation. Overexpression of either Nkx6.1 or HDAC1 in INS-1 832/13 β-cells resulted in greater than 93% nuclear localization, demonstrating the maintenance of the normal subcellular localization, and suggesting that HDAC1 function in the overexpression model is nuclear (Supplemental Figure S2A–D). The cell cycle is controlled by both activators and inhibitors. The inhibitors act as a braking mechanism to impede cell cycle progression. We measured expression of all cell cycle inhibitors (Cdkn1a, Cdkn1b, Cdkn1c, Cdkn2a, Cdkn2b, Cdkn2c and Cdkn2d) in untreated islets or islets transduced with AdCMV-GFP or AdCMV-HDAC1, and while no significant changes in expression were observed for most of the genes, we did observe a significant decrease in Cdkn1b/p27 expression (Figure 6A). Western blot analysis of primary rat islets transduced with AdCMV-HDAC1 demonstrated a significant decrease in Cdkn1b/p27 protein levels (Figure 6B,C). Similar results were observed for INS-1 832/13 β-cells transduced with AdCMV-HDAC1 (Supplemental Figure S1G,H).
HDAC1 overexpression induces a change in cell cycle gene expression.
Cdkn1b/p27 inhibits G1 to S phase cell cycle progression by binding and inhibiting Cyclin D/Cdk4 and Cyclin D/Cdk6 complex formation and activity. We hypothesized decreased expression of Cdkn1b/p27 would result in up-regulation of genes controlled through the Cyclin D complexes. To determine if HDAC1-mediated β-cell proliferation corresponds with changes to cell cycle gene expression, primary rat islets were transduced with AdCMV-GFP or AdCMV-HDAC1 or left untreated. HDAC1 overexpression resulted in no change in expression for Cyclin D2, Cyclin E1, AURKA, Nr4a1 or Nr4a3. HDAC1 overexpression did, however, enhance expression of Cyclin A2 Cyclin B1, E2F1, Cdc6, Cdk2, Cdt1 and Foxm1 (Figure 6D). These data demonstrate that HDAC1 overexpression in β-cells results in decreased Cdkn1b/p27 expression, which corresponds with increased expression of genes involved in the G1 to S phase transition.
Cdkn1b/p27 overexpression impedes HDAC1-mediated β-cell proliferation
To determine if Cdkn1b/p27 down-regulation is necessary for HDAC1-mediated β-cell proliferation, we overexpressed Cdkn1b/p27 in the presence and absence of HDAC1 overexpression. INS-1 832/13 β-cells were cultured for 48 h after adenoviral transduction, and while HDAC1 overexpression resulted in increased β-cell proliferation, cells that were also transduced with AdCMV-Cdkn1b/p27 failed to replicate in response to HDAC1 overexpression (Figure 7). These findings suggest that HDAC1-mediated β-cell proliferation is dependent on Cdkn1b/p27 down-regulation, and in the absence of this down-regulation HDAC1 is unable to induce β-cell proliferation.
Overexpression of Cdkn1b/p27 inhibits HDAC1-mediated β-cell proliferation.
The β-cell transcription factor Nkx6.1 is sufficient to induce functional β-cell proliferation [7,29]. Nkx6.1-mediated proliferation is observed 72 h after overexpression in primary β-cells, suggesting that early Nkx6.1 targets may be necessary to permit β-cells to re-enter the cell cycle and proceed through cellular replication . Here, we show that HDAC1 expression is increased as a result of Nkx6.1 overexpression in β-cells. We demonstrate that HDAC activity is necessary for Nkx6.1-mediated β-cell proliferation, and that HDAC1 knockdown significantly decreases β-cell proliferation. Our data demonstrate that HDAC1 overexpression in primary rat β-cells or the INS-1 832/13 β-cell line is sufficient to induce proliferation, while maintaining insulin secretion and insulin content. We demonstrate that HDAC1 overexpression is sufficient to protect INS-1 832/13 β-cells from etoposide and thapsigargin-induced cell death. Concomitant with the observed β-cell proliferation, we demonstrate that HDAC1 overexpression corresponds with both down-regulation of the cell cycle inhibitor Cdkn1b/p27 and increased expression of key G1/S phase cell cycle activators. Finally, we demonstrate that overexpression of Cdkn1b/p27 is sufficient to impede HDAC1-mediated β-cell proliferation.
Based on these findings, we propose a model whereby expression of Cdkn1b/p27 in unmanipulated β-cells inhibits proliferation, potentially by blocking the activity of its canonical targets Cyclin D/Cdk4 and Cyclin D/Cdk6 (Figure 8A). With HDAC1 overexpression, a downstream target of Nkx6.1, Cdkn1b/p27 expression is decreased. Decreased Cdkn1b/p27 expression allows Cyclin D/Cdk complex formation. Cyclin D/Cdk4 and Cyclin D/Cdk6 complex formation results in cell cycle progression by phosphorylating pRb which releases the cell cycle transcription factor E2F1. Unfettered E2F1 results in up-regulation of cell cycle genes critical for DNA replication and cell division (Figure 8B). Our results demonstrate that HDAC1 is necessary for Nkx6.1-mediated activity, and sufficient to induce β-cell proliferation while maintaining GSIS and enhancing cell survival.
HDAC1 overexpression induces β-cell proliferation.
HDAC1 has previously been shown to inhibit Cdkn1b/p27 expression. The transcription factor NANOG mediates HDAC1 up-regulation in T-cells resulting in down-regulation of Cdkn1b/p27 . Conversely, treatment of T-cell acute lymphoblastic leukemia cells with HDAC1 chemical inhibitors results in cell cycle inhibition that corresponds with up-regulation of Cdkn1b/p27 . Up-regulation of HDAC1 resulting in deacetylation and subsequent down-regulation of Cdkn1b/p27 has been demonstrated in various tissues [20,21,35], while inhibition or down-regulation of HDAC1 corresponds with increased Cdkn1b/p27 expression and cell cycle inhibition [36–38]. These data support our finding that HDAC1 overexpression in the β-cell results in increased proliferation concomitant with decreased Cdkn1b/p27 expression. Cdkn1b/p27 has been defined as a critical gene in impeding β-cell proliferation, with down-regulation of Cdkn1b/p27 resulting in increased β-cell replication [39–41]. Our data clearly demonstrate that enhanced HDAC1 activity can result in Cdkn1b/p27 down-regulation (Figure 6A), leading to increased expression of cell cycle activators and ultimately increase β-cell proliferation (Figure 6B). These data are further strengthened by the observation that Cdkn1b/p27 overexpression is sufficient to inhibit HDAC1-mediated β-cell proliferation (Figure 7).
While various studies have demonstrated that HDAC1 can regulate expression of both Cdkn1a/p21 and Cdkn1b/p27, our results only demonstrated HDAC1-dependent down-regulation of Cdkn1b/p27. While there are many potential reasons for why this observation may occur (lack of a necessary HDAC1 binding factor in the β-cell, inaccessibility of the Cdkn1a/p21 promoter to HDAC1 in the β-cell, etc.), it is important to note that loss of Cdkn1a/p21 has been shown to be not sufficient for β-cell expansion . Furthermore, it is important to note that while HDAC1 co-inhibition of Cdkn1a/p21 and Cdkn1b/p27 expression (and that conversely loss of HDAC1 results in increased expression of these inhibitors) has been reported, various reports have shown that they are not always co-regulated. Inhibition of HDAC1 in hepatocellular carcinoma results in increased Cdkn1a/p21 expression . Likewise, up-regulation of HDAC1 in human aortic endothelial cells results in decreased Cdkn1a/p21 expression . These findings are contrasted with those that show that the use of HDAC inhibitors with acute promyelocytic leukemia cells results in Cdkn1b/p27 up-regulation [45,46], and that HDAC inhibitors used with human ovarian cancer cells induced Cdkn1b/p27 up-regulation . While these reports do not explain the mechanism by which p21 and p27 can be regulated independently, they do substantiate our observation. This area of research is ongoing in our laboratory.
As previously mentioned, β-cell proliferation rarely naturally occurs post-adolescence. Two exceptions to this are the increased β-cell proliferation observed during pregnancy and with obesity. While HDAC1 activity has not yet been defined in either of these biological processes, there are data that support that activity of HDAC1. Pregnancy-associated β-cell proliferation is associated with menin down-regulation. Decreased menin expression corresponds with a down-regulation of Cdkn1b/p27 expression [48–50]. Furthermore, menin binds HDAC1 . This may suggest that menin may sequester HDAC1, preventing HDAC1-mediated deacetylation of the Cdkn1b/p27 promoter and its eventual down-regulation. Obesity-associated β-cell proliferation is also associated with Cdkn1b/p27 down-regulation. In this model, GSK3-β is inhibited, which results in decreased Cdkn1b/p27 expression [41,52]. GSK3-β has been shown to interact with HDAC1 , and that HDAC inhibition enhances Cdkn1b/p27 through GSK3- β . While clear data demonstrating a connection between HDAC1 activity and pregnancy or obesity-related β-cell proliferation are not yet available, this area of research is ongoing in our group.
While increasing β-cell proliferation is one way of enhancing functional β-cell mass, these β-cells must be able to sense elevated blood glucose levels and secret insulin in response to this elevation. Our data demonstrate that HDAC1 overexpression maintains β-cell insulin content and GSIS. While various studies have demonstrated that HDAC3 appears to negatively regulate GSIS [55–57], little is known about the effect of HDAC1 manipulation on GSIS. Our data are supported by a recent study showing that rats treated with the broad histone inhibitor valproic acid result in decreased insulin content and insulin secretion, suggesting that HDAC activity is necessary for GSIS . These findings are, however, somewhat controversial because treatment with other class I HDAC inhibitors has been shown to maintain or enhance GSIS [59,60]. Other studies, looking at the effect of cytokines on impeding GSIS in the context of β-cell death, have shown that HDAC1 knockdown improves GSIS . As our study did not address the effect of HDAC1 overexpression in the presence of cytokine-mediated β-cell death, this would be an interesting avenue to explore.
Our data demonstrate that HDAC1 overexpression protects β-cells from etoposide and thapsigargin-mediated cell death, while having no effect on camptothecin-mediated cell death. The mechanism of these compounds varies. Camptothecin affects DNA topoisomerase I, etoposide affects DNA topoisomerase II and thapsigargin affects intracellular Ca2+ levels. While camptothecin and etoposide both effect DNA stability, camptothecin has its greatest effect during S phase, while etoposide has effects into G2 phase . This may explain why HDAC1 expressing cells do not have any greater amount of stability in the presence of camptothecin, as DNA replication is fully engaged. Other studies have looked at the role of class I HDACs in β-cell survival. Studies have shown that cytokine-mediated cell death is dependent primarily on HDAC3, with HDAC1 activity being unnecessary for cell death [61,63]. Studies have also shown that HDAC3, and not HDAC1, is necessary for palmitate-induced β-cell death . While we have not yet tested the role of HDAC1 overexpression in our model with either cytokine or palmitate-induced cell death, our data are supported by the previous findings. Our findings are also supported by research in other tissue types that show that HDAC inhibition magnifies etoposide-mediated cell death . Further studies show that HDAC inhibition potentiates thapsigargin-mediated cell death , supporting our findings that HDAC1 overexpression leads to protection from cell death.
In summary, we have shown that HDAC1 is a downstream Nkx6.1 target that is necessary and sufficient for β-cell proliferation. Furthermore, our findings show that HDAC1-mediated β-cell proliferation is dependent on Cdkn1b/p27 down-regulation. The present study demonstrates two potential points that may be manipulated to enhance functional β-cell mass. First, our data demonstrating that Cdkn1b/p27 down-regulation is necessary for HDAC1-mediated proliferation emphasizes the need to remove cell cycle braking mechanisms in order to enhance β-cell proliferation. Second, our data show that manipulation of HDAC1 can enhance β-cell proliferation. Together, these results emphasize the need to remove cell cycle braking mechanisms as well as stimulate cell cycle activators in seeking pathways to enhance functional β-cell mass. While there are many class I HDAC inhibitors, there are few class I HDAC activators (N-acetylthioureas serve as putative isozyme selective activators for human HDAC8 on a fluorescent substrate [67,68]). Given our results, HDAC1-specific agonists may be useful in terms of enhancing functional β-cell mass and should be explored.
cyclin-dependent kinase inhibitor
green florescent protein
glucose-stimulated insulin secretion
histone deacetylase 1
institutional animal care and use committee
NK6 homeobox 1
reverse transcription-polymerase chain reaction
short interfering RNA
type 1 diabetes
type 2 diabetes
J.S.T. conceived and managed the studies. J.S.T., M.A., C.D., M.C.A., A.H.L., C.J.S., K.B.K., T.J.A., K.H.H., A.H.-C., P.T.F., A.C.H. and L.A.L. prepared and analyzed all experiments. J.S.T., M.A., C.D., M.C.A., A.H.L., C.J.S., K.B.K., T.J.A., K.H.H,. A.H.-C., P.T.F., A.C.H. and L.A.L. prepared, wrote and edited the manuscript.
This work was supported by grants to J.S.T. from the American Diabetes Association (1-17-IBS-101) and the Diabetes Action and Research Foundation [Grant # 461], grant to P.T.F. from the NIH (DK099311), an American Diabetes Association Minority Undergraduate Fellowship (C.J.S.), Mentoring Environment Grant from the BYU Office of Research and Creative Activity (J.S.T.) and Undergraduate ORCA awards (C.D. and K.B.K.) and an endowment from the Kevin and Mimi Sayer Diabetes Research laboratory. Human pancreatic islets were provided by the NIDDK-funded IIDP at City of Hope, NIH Grant # 2UC4DK098085.
We thank Drs Thomson and Andersen for their comments and critiques of the present study. We thank the members of the Tessem laboratory for their feedback and suggestions.
The Authors declare that there are no competing interests associated with the manuscript.
Present address: Northwestern Medicine, Central DuPage Hospital, Northwestern University, Winfield, IL 60190, U.S.A.
Present address: Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, U.S.A.