Doc2b Protects β-Cells Against Inflammatory Damage and Enhances Function

In the U.S., type 1 diabetes currently affects an estimated 1.25 million people, and its prevalence is predicted to rise to 5 million Americans by 2050 (1–3). Type 1 diabetes is characterized by hyperglycemia caused by autoimmune destruction of islet β-cells, leading to a decline in β-cell function and mass. Increasing evidence shows that β-cell dysfunction precedes the clinical onset of type 1 diabetes (4,5). Indeed, multiple clinical trials show that patients with type 1 diabetes who display even minimal β-cell function at the onset of the study exhibit decreased microvascular complications, including retinopathy and neuropathy (6,7). Currently, the only therapy for patients with type 1 diabetes is exogenous insulin replacement. The danger in this approach, however, is that normal, glucose-dependent release of endogenous insulin is lost and that life-threatening hypoglycemic episodes and progressive destruction of β-cell mass persist. Thus, there is a high demand for alternate/adjuvant therapies that can preserve β-cell mass and function.

Normal β-cell function/insulin secretion is dependent on intact insulin secretory machinery, termed the soluble N-ethylmaleimide–sensitive factor attachment protein receptor (SNARE) complex. The SNARE complex consists of two target membrane (t-SNARE) proteins, syntaxin (STX) and SNAP25, and one vesicle-associated (v-SNARE) protein, VAMP (8–11). Upon glucose stimulation, one v-SNARE binds with two cognate t-SNARE proteins in a 1:1:1 heterotrimeric ratio (9). SNARE proteins facilitate biphasic insulin secretion using different requisite SNARE protein isoforms (12,13). The first phase of insulin secretion (first 5–10 min after glucose stimulation) uses STX1A, STX4, SNAP25 or SNAP23, and the v-SNARE VAMP2, whereas second-phase insulin secretion (>10 min after glucose stimulation) uses STX4, SNAP25 or SNAP23, and VAMP2 (11,14–19). Another v-SNARE protein, VAMP8, also functions in insulin secretion and is selectively required for glucagon-like peptide 1 (GLP-1)–enhanced insulin release (20). Assembly of the β-cell SNARE complex is regulated by accessory binding proteins, such as double C2 domain–containing protein β (Doc2b). Upon glucose stimulation, STX binds VAMP2, permitting SNARE complex assembly and fusion of insulin granules with the plasma membrane (21,22). Doc2b is essential for this mechanism to occur (22–24).

We recently showed that Doc2b levels are reduced in the residual islet β-cells from new-onset human type 1 diabetes donor pancreata; Doc2b deficiency was recapitulated by treatment of nondiabetic human islets with proinflammatory cytokines, ex vivo (25). Proinflammatory cytokines are used to mimic the diabetogenic milieu by promoting loss of β-cell mass, given that loss of β-cell mass in the context of type 1 diabetes is linked to apoptosis (26,27). Whether Doc2b plays a role in β-cell survival or apoptosis remains unknown.

Doc2b is a ubiquitously expressed, 46–50-kDa protein that localizes to the plasma membrane in β-cells and muscle/fat cells (22,28,29). Doc2b contains an N-terminal Munc13-interacting domain (MID) linked to C-terminal tandem C2 domains (C2A and C2B), which are motifs known to bind Ca²⁺ and phospholipids. Several laboratories have shown that overexpression of Doc2b enhances insulin secretion by 30–40% in vitro (24,29). Moreover, transgenic mice overexpressing Doc2b (simultaneously in the pancreas, skeletal muscle, and adipose tissue) show enhanced whole-body glucose homeostasis and insulin sensitivity, with enhanced islet glucose-stimulated insulin secretion (GSIS) ex vivo, but have normal fasting insulin levels (28). These data suggest that Doc2b enrichment is beneficial, although the investigation of detailed mechanisms for Doc2b actions in the β-cells has been limited by the use of a global transgene expression system. Doc2b-null knockout mice harbor dysfunctional islets (30,31), consistent with the observed low Doc2b transcript levels in islets of diabetic rodents (25,32). However, it remains unknown whether Doc2b deficiency is causal or consequential to the loss of β-cell function and diabetes onset.

In this study, we tested the hypothesis that Doc2b deficiency underlies β-cell dysfunction and susceptibility to diabetes and that selective β-cell Doc2b enrichment is sufficient to confer protection against proinflammatory stimuli. Data gained from classic whole-body Doc2b heterozygous knockout mice, inducible β-cell–specific transgenic Doc2b overexpressing (βDoc2b-dTg) mice, and Doc2b overexpression in clonal β-cells provide strong support for this hypothesis. Furthermore, we show that enrichment of a peptide containing the Doc2b tandem C2 domains (C2AB) is sufficient to confer the GSIS-boosting effect of full-length Doc2b and protects against thapsigargin-induced β-cell apoptosis.

The rabbit anti-Doc2b antibody used for detection of endogenous Doc2b levels and detection of hDoc2b-DDK was purchased from Proteintech (1:1,000 dilution, catalog no. 20574–1-AP; Rosemont, IL). The rabbit anti-cMyc antibody (1:1,000 dilution, catalog no. sc-40) and mouse antiglucagon (1:200 dilution, catalog no. sc-13091) antibodies were purchased from Santa Cruz Biotechnology (Dallas, TX). The rabbit anti-STX4 antibody used in the pulldown binding assay was generated in house (33). Antibodies to cleaved caspase 3 (CC3; 1:1,000 dilution, catalog no. 9661), CHOP (1:1,000 dilution, catalog no. 2895), PARP (1:1,000 dilution, catalog no. 9532), phospho-eIF2a (p-eIF2a), and total eIF2a (1:1,000 dilution, catalog nos. 9721 and 9722, respectively) were purchased from Cell Signaling (Danvers, MA). The tubulin antibody was purchased from Abcam (1:5,000 dilution, catalog no. ab56676; Cambridge, MA). The guinea pig anti-insulin was from Dako/Agilent (1:100 dilution, catalog no. A0564; Santa Clara, CA). The goat anti-rabbit- and anti-mouse-horseradish peroxidase secondary antibodies were purchased from Bio-Rad (catalog nos. 1706515 and 1706516, respectively; Hercules, CA). Thapsigargin was purchased from Sigma-Aldrich (catalog no. T9033). Glutathione sepharose 4B agarose beads were obtained from GE Life Sciences (Pittsburgh, PA). Enhanced chemiluminescence was purchased from Amersham Biosciences (Pittsburg, PA). Humulin R was obtained from Eli Lilly and Co. (Indianapolis, IN). The rat (r)Doc2b-GFP and rC2AB-GFP plasmids were gifts from U. Ashery (Tel-Aviv University, Ramat-Aviv, Israel). The rat CaMut-GFP plasmid was a gift from E. Chapman (University of Wisconsin).

Animals were maintained under protocols approved by the Institutional Animal Care and Use Committees at Indiana University School of Medicine and the City of Hope, according to the Guidelines for the Care and Use of Laboratory Animals. Doc2b^+/− knockout mice were generated as described previously (30). The rat Doc2b cDNA tagged with a C-terminal myc-his tag was subcloned into the 5′ Pme1 and 3′ BamH1 restriction sites in the pTRE-pIRES-eGFP vector (gift from Drs. S. Afelik and J. Jenssen, Cleveland Clinic), downstream of the tetracycline response element (TRE), to provide tetracycline/doxycycline (Dox) inducibility. The linearized plasmid was microinjected into C57BL/6J oocytes (Transgenic Core, Indiana University School of Medicine). Of two founders, one transmitted the transgene to offspring and was shipped to the City of Hope Animal Research Center for colony expansion. TRE-Doc2b^−/+ offspring crossed to rat insulin promoter (RIP)–rtTA^−/+ mice (purchased from Jax Laboratories) produced Dox-inducible β-cell–selective expression in the double transgenic offspring (βDoc2b-dTg) mice. The mice were generated and maintained on the C57BL/6J genetic background.

Female βDoc2b-dTg mice and single transgenic controls were fasted for 6 h (0800–1400 h) before the intraperitoneal glucose tolerance test (IPGTT; 2 mg/mL of glucose) and intraperitoneal insulin tolerance test (IPITT, insulin dose = 0.75 units/kg body weight), as previously described (30). Male Doc2b^+/− and wild-type (Wt) littermate mice (7 weeks old) were used for streptozotocin (STZ) studies; a pre-STZ IPGTT was performed on day 0. On days 1–5, mice were injected intraperitoneally in multiple low doses with freshly prepared STZ (35 mg/kg body weight, as previously described [34]). A post-STZ IPGTT was performed on day 10. Male mice were used for STZ experiments, as they are reportedly more sensitive to STZ treatment (35). Male βDoc2b-dTg mice (aged 9–10 weeks) were provided Dox-treated (2 mg/mL) or standard drinking water for 3 weeks prior to multiple-low-dose (MLD) STZ (35–40 mg/kg doses); IPGTT was assessed on day 24. On day 25, serum samples and tissues were collected, and the pancreata were fixed and paraffin embedded for TUNEL staining. Non–STZ-treated male and female mice were used for tissue harvest and evaluation of protein/mRNA expression. Serum insulin levels were measured via radioimmunoassay (Millipore).

Paraffin-embedded pancreatic tissue sections were prepared and assessed as previously described (36). To assess β-cell specificity, pancreatic sections were immunostained with guinea pig anti-insulin (1:100, catalog no. A0564; DAKO, Agilent Technologies, Santa Clara, CA), rabbit anti-myc (1:100, catalog no. C3956; Sigma-Aldrich, St. Louis, MO), and mouse anti-glucagon (1:100, catalog no. G2654; Sigma-Aldrich). Alexa Fluor 568 goat anti-rabbit (1:200, catalog no. A11011; Thermo Fisher Scientific), 488 goat anti-guinea pig (1:100, catalog no. A11073; Thermo Fisher Scientific), and 647 goat anti-mouse IgG (Heavy + Light Chains [H+L], 1:100, catalog no. A21236; Thermo Fisher Scientific) were used to detect myc, insulin, and glucagon, respectively. Slides were counterstained to mark the nuclei using 4′,6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories, Burlingame, CA) and viewed using a Keyence BZ X-700 fluorescence microscope (Leica Microsystems, Deerfield, IL), and images were acquired using the 20× objective. To assess GLUT2 staining, pancreatic sections were immunostained with goat anti-GLUT2 (1:100, catalog no. sc-7580; Santa Cruz Biotechnology, Inc., Dallas, TX) and guinea pig anti-insulin (1:100, catalog no. A0564; DAKO, Agilent Technologies). Cy3 anti-goat (1:200, catalog no. 715-166-147, Jackson ImmunoResearch) and Alexa Fluor 488 goat anti-guinea pig (1:100, catalog no. A11073; ThermoFisher Scientific, Waltham, MA) were used to detect GLUT2 and insulin, respectively.

To assess apoptosis, the pancreatic sections were immunostained with guinea pig anti-insulin, and the TUNEL In Situ Cell Death Detection Kit, Fluorescein (Roche, Mannheim, Germany) was used to stain apoptotic cells. The Alexa Fluor 488 goat anti-guinea pig IgG (H+L) (1:500 dilution, catalog no. A11073; Thermo Fisher Scientific) secondary antibody was used for detection of insulin. Sections were scanned using the Zeiss Axio Observer and LSM 700 microscopes and analyzed using Image-Pro Software (Media Cybernetics, Rockville, MD). The results are expressed as the percentage of cells positive for TUNEL staining relative to the total number of insulin-positive cells.

Human Doc2b (hDoc2b) adenovirus was generated by insertion of the hDoc2b-mycDDK cDNA from the pCMVhDoc2bmycDDK plasmid (OriGene Technology, Inc., Rockville, MD) into the pAd5CMVmpA adenoviral vector. The adenoviruses were packaged with EGFP to enable visualization of infection efficiency, and then were amplified and purified for use by Viraquest, Inc. (North Liberty, IA). Rat INS-1 832/13 β-cells (gift from C.B. Newgard, Duke University, Durham, NC, and P. Fueger, City of Hope; passage 70–80) were cultured in RPMI 1640 medium as previously described (37) and transduced with Ad5-CMV-hDoc2b prior to treatment with a cytokine mixture (10 ng/mL TNF-α, 100 ng/mL IFN-γ, and 5 ng/mL IL-1β; ProSpec, East Brunswick, NJ) for 16 h, as previously described (38). For evaluation of thapsigargin-induced apoptosis, INS-1 832/13 β-cells were transfected with rDoc2b-GFP, rC2AB-GFP, and rCaMut-GFP. Cells were transfected using the K2 Transfection System (Biontex, Munchen, Germany) and treated with 1 μmol/L thapsigargin for 6 h prior to harvest; concurrently, cells were treated with 1 mmol/L EGTA as a negative control. The cells were harvested in 1% NP-40 lysis buffer, and cleared detergent lysates were used for immunoblotting for apoptosis/endoplasmic reticulum (ER) stress markers.

Glutathione S-transferase (GST)–VAMP2 interaction assays were conducted as previously described (33). In brief, GST-VAMP2 protein conjugated to sepharose beads was mixed with cleared detergent cell lysates prepared from MIN6 β-cells that had been preincubated in serum-free media for 2 h and were harvested in 1% NP-40 lysis buffer, as previously described (39). Recombinantly expressed and purified Doc2b protein or C2AB peptide were added to the reactions, and reactions were rotated at 4°C for 2 h, beads were pelleted, and proteins were eluted for resolution on SDS-PAGE and immunoblotting.

GSIS analysis was performed as previously described (38). In brief, rat INS-1 832/13 β-cells and MIN6 β-cells were prepared for GSIS analysis using the K2 Transfection System (Biontex). INS-1 832/13 cells were placed in low-glucose/low-serum medium overnight prior to starvation in Krebs-Ringer bicarbonate buffer for 1 h and treated with glucose (basal: 2.8 mmol/L; glucose stimulated: 20 mmol/L) for 30 min. The MIN6 cells were serum starved in modified Krebs-Ringer bicarbonate buffer for 1 h prior to glucose treatment (basal: 2.8 mmol/L; glucose stimulated: 20 mmol/L). The supernatants were collected for measurement of insulin content using the Rat Insulin ELISA (Mercodia, Uppsala, Sweden) or Mouse Insulin ELISA (Alpco, Salem, NH).

Total RNA was isolated from mouse islets, hypothalamus, and cerebellum using the Qiagen RNAeasy Plus Mini Kit (Qiagen, Valencia, CA) and assessed using the QuantiTect SYBR Green RT-PCR kit (Qiagen). Primers used for the detection of rDoc2b-myc are as follows: forward 5′-ACTGTCTGAAGAACAAGGACAAGAG-3′ and reverse 5′-GAGTTTTTGTTCTACGTAAGCTTGG-3′. Hypoxanthine-guanine phosphoribosyltransferase primers were used for normalization: forward 5′-AAGCCTAAGATGAGCGCAAG-3′ and reverse 5′-TTACTAGGCAGATGGCCACA-3′. Primers used for the detection of rat/mouse Doc2b (nontagged) are as follows: forward 5′-CCAGCAAGGCAAATAAGCTC-3′ and reverse 5′-GTTGGGTTTCAGCTTCTTCA-3′.

Islet morphometry was conducted using anti-insulin–stained pancreatic sections, as previously described (38). The percentage of β-cell area was calculated using Keyence BZX-Analyzer software (Keyence Corporation, Itasca, IL). The data shown are representative of three pancreatic sections per mouse. The β-cell mass was calculated by multiplying the percentage of the β-cell area by the pancreas weight.

All data are expressed as the mean ± SD for all figures, except those representing IPGTT/IPITT or area under the curve (AUC) (mean ± SEM). The data were evaluated for statistical significance using Student t test for comparison of two groups, or ANOVA and Tukey post hoc test (GraphPad Software, La Jolla, CA) for more than two groups.

To assess whether Doc2b abundance impacts β-cell susceptibility to diabetogenic insults, we studied Doc2b^+/− mice prior to and after MLD-STZ induction of diabetes. Prior to MLD-STZ treatment, young 7-week-old male Doc2b^+/− and Wt littermate mice showed no differences in blood glucose levels (Fig. 1A and B) and β-cell mass and harbored the expected 50% reduction in Doc2b protein (Supplementary Fig. 1A and B). Ten days after the first STZ injection, the Doc2b^+/− mice showed significantly higher blood glucose levels than Wt mice (Fig. 1C and D). One day later, all pancreata were collected and processed for TUNEL staining, and the Doc2b^+/− mice showed greater than threefold higher levels of β-cell apoptosis than the Wt mice (Fig. 1E). Although the number of islets was similar in the Doc2b^+/− and Wt mice, the percentage of the islet area occupied by β-cells was decreased in the Doc2b^+/− mice, which also resulted in a reduced β-cell mass (Fig. 1F and G). Consistent with this, serum insulin levels in the Doc2b^+/− were lower than Wt mice post-STZ treatment, whereas prior to STZ, levels were similar to Wt mice (Supplementary Fig. 2). The body weights of Doc2b^+/− and Wt mice were equivalent before and after the brief STZ treatment (Supplementary Table 1).

To determine whether β-cell–specific overexpression of Doc2b is sufficient to enhance whole-body glucose tolerance, we generated a Dox-induced, β-cell–specific double transgenic mouse model (βDoc2b-dTg) (Fig. 2A). The Doc2b mRNA abundance in islets isolated from female mice was approximately four- to fivefold higher in Dox-induced βDoc2b-dTg mice (same for male mice, data not shown) than in non–Dox-treated βDoc2b-dTg control mice (Fig. 2B and Supplementary Fig. 3), and the presence of the myc-tagged Doc2b protein was clearly detectable (Fig. 2C). No transgene overexpression was detected in the hypothalamus or cerebellum, indicating no aberrant expression due to the RIP-rtTA (Fig. 2D). Immunofluorescent analysis of βDoc2b-dTg islets shows β-cell–specific overexpression of Doc2b-myc in Dox-treated βDoc2b-dTg mice (Fig. 2E).

Changes to whole-body glucose homeostasis were assessed by IPGTT in the βDoc2b-dTg mice. Dox-induced dTg mice showed significantly lower blood glucose levels (i.e., improved glucose tolerance) after glucose injection; basal fasting blood glucose levels were similar to those of non–Dox-treated dTg control mice (Fig. 3A and B) or single transgenic mice (Supplementary Fig. 4). Consistent with this, the serum insulin levels in Dox-treated βDoc2b-dTg mice were higher than controls within the first 10 min after glucose injection during IPGTT, but they were not statistically different under basal (time 0 in the IPGTT) conditions (Fig. 3C). No differences were seen between Dox-treated and non–Dox-treated βDoc2b-dTg mice in IPITTs, suggesting no changes to peripheral insulin sensitivity (Fig. 3D and E). βDoc2b-dTg mice had similar body weight and tissue weights relative to the control Dox⁺ or noninduced (Dox⁻) single transgenic mice and noninduced dTg mice (Supplementary Table 2). These data suggest that enrichment of Doc2b selectively in adult pancreatic islet β-cells is sufficient to improve whole-body glucose homeostasis.

Male βDoc2b-dTg mice were subjected to a longer (24-day) MLD-STZ protocol to determine if β-cell Doc2b enrichment protects islets from diabetogenic stress. IPGTT assessments at 24 days after initiation of the STZ protocol reveal increased blood glucose levels, with fasting levels ∼300 mg/dL in non–Dox-treated dTg mice (Fig. 4A and B). In contrast, Dox-induced βDoc2b-dTg mice were largely protected, even though the longer time after STZ treatment would be expected to increase β-cell destruction. Body weights of the Dox⁺ and non–Dox-treated mice were similar both before and after MLD-STZ treatment (Supplementary Table 3).

To determine if the improved glucose tolerance, robust serum insulin response, and protection from STZ-induced glucose intolerance in the βDoc2b-dTg mice were related to changes in β-cell apoptosis, pancreata from the STZ-treated dTg mice were assessed by TUNEL staining. Islets from Dox-treated βDoc2b-dTg mice contained ∼58% fewer apoptotic cells than islets from non–Dox-treated dTg mice (Fig. 4C). Although the islet numbers were similar in both groups (data not shown), the Dox-treated βDoc2b-dTg mice had a larger β-cell area, corresponding to higher β-cell mass, compared with non–Dox-treated dTg mice (Fig. 4D and E). Consistent with this, Dox-induced dTg mice had serum insulin levels similar to Wt mice pre-STZ, whereas non–Dox-treated dTg mice showed loss of insulin (Supplementary Fig. 5). GLUT2 expression in Dox⁺ and Dox⁻ βDoc2b-dTg mice was unchanged (Supplementary Fig. 6). These data suggest that overexpression of Doc2b protects islet β-cells against STZ-induced β-cell death.

To date, all studies of Doc2b overexpression in islets and in vivo function have used rat Doc2b; hDoc2b has yet to be tested. Toward this, we generated adenoviral particles encoding hDoc2b tagged on the C terminus with DDK. Ad-hDoc2b-DDK was packaged with GFP and was detected in ≥70% of rat INS-1 832/13 cells after viral transduction (Supplementary Fig. 7); the Ad vector control transduced ∼100% of cells. Treatment of Ad-hDoc2b–expressing INS-1 832/13 cells with proinflammatory cytokines for 16 h resulted in a 45% reduction in CC3 and >50% reduction in cleaved PARP, as compared with control cells (Fig. 5). Additionally, the levels of ER stress markers CHOP and p-eIF2a were reduced by ∼50% and ∼60%, respectively, in Doc2b-overexpressing β-cells (Fig. 5). These data show that Doc2b protects cells from apoptotic stimuli and reduces ER stress.

Doc2b is composed of three primary domains, MID, C2A, and C2B (Fig. 6A); the C2 domains are known calcium-sensing domains, binding two calcium ions each and contributing to Doc2b’s function. To further investigate Doc2b’s antiapoptotic actions and the role of the C2 domains in alleviating ER stress, GFP-tagged plasmids were generated that encode rDoc2b (Doc2b-GFP), a fragment containing an N-terminal truncation of the MID domain (C2AB-GFP), and CaMut-GFP, a mutated form of Doc2b-GFP carrying D→N substitutions in the amino acids implicated in calcium handling in the C2 domains (40) (Fig. 6A). INS-1 832/13 cells were treated with the ER stress inducer and SERCA2b pump inhibitor thapsigargin. Treatment of Doc2b-GFP– and C2AB-GFP–overexpressing INS-1 832/13 cells with 1 μmol/L thapsigargin for 6 h decreased the levels of CHOP and CC3, compared with control (GFP)-treated cells (Fig. 6B). These decreased levels were similar to the levels of CHOP in cells treated with the calcium chelator EGTA (Fig. 6B). The benefit of expressing Doc2b-GFP or C2AB-GFP was absent when the cells were transfected with the calcium binding–deficient mutant form of Doc2b, CaMut-GFP (Fig. 6A and B). These data suggest that Doc2b’s calcium binding capacity within the C2AB tandem domains is linked to its protective effect against ER stress and apoptosis.

Because the C2AB peptide was sufficient to protect β-cells against thapsigargin-induced apoptosis, we investigated the capacity of C2AB to enhance β-cell function. Full-length rDoc2b-GFP expression in MIN6 cells recapitulated the enhancement of GSIS seen with rDoc2b in prior reports (24,28,29) (Fig. 7A). Remarkably, C2AB enhanced GSIS by approximately two- to threefold, similar to full-length Doc2b (Fig. 7A). Neither C2A or C2B expression alone enhanced GSIS, despite confirmed expression of each peptide; if anything, the C2AB-GFP protein expression levels were slightly lower than those of the other proteins (Fig. 7B). The insulin content was also similar among all treatment groups (Supplementary Fig. 8A). These results were recapitulated when the constructs were tested in INS-1 832/13 cells (Fig. 7C and D and Supplementary Fig. 8B). Doc2b is required for SNARE complex formation in β-cells, possibly via increasing STX4 activation (i.e., interaction with its cognate v-SNARE partner, VAMP2 [28]). To determine if C2AB can replicate the underlying mechanism, the C2AB and Doc2b proteins (untagged) were tested for the ability to increase STX4 binding to GST-VAMP2 (Fig. 7E). Notably, the C2AB protein fully recapitulated the Doc2b-induced activation of STX4. These data suggest that C2AB includes a minimal region of Doc2b that is sufficient to confer enhanced GSIS, STX4 activation, and protection against apoptosis.

The data presented here demonstrate an important role for Doc2b in maintaining β-cell mass and function. We propose that Doc2b deficiency leaves β-cells more susceptible to diabetogenic damage and that overexpression of Doc2b in the β-cell enhances whole-body glucose homeostasis and prevents β-cell apoptosis and ER stress. We also show that C2AB comprises a minimal Doc2b region that is required for enhanced insulin secretion and that its antiapoptotic effects are linked to known sites of calcium binding within the C2AB region. Hence, delivery of C2AB could represent a novel candidate therapeutic strategy to promote and protect functional β-cell mass.

To our knowledge, this study is the first report showing protective, antidiabetic properties of Doc2b. We found that reducing the Doc2b levels in mice (Doc2b^+/−) increases susceptibility to MLD-STZ–induced β-cell destruction and glucose intolerance. On the other hand, enhancing Doc2b levels in the β-cell (βDoc2b-dTg) decreases susceptibility to MLD-STZ–induced β-cell destruction and glucose intolerance. INS-1 832/13 β-cells transduced to overexpress hDoc2b also exhibit protection from proinflammatory cytokine and thapsigargin-induced apoptosis, indicating that Doc2b protection occurs at the level of the β-cells, and the protective effect is not dependent upon interislet non–β-cell communication, nor is it due to a circulating factor in vivo. The role of Doc2b or other exocytosis proteins in β-cell survival mechanisms still remains largely unexplored; however, recent reports have shown involvement of exocytosis proteins, such as VAMP8, in β-cell proliferation (20). Also, in neurons, Doc2b has been implicated in proliferation and differentiation during early embryogenesis (41,42). However, the β-cell area and insulin content in islets from mice with increased or decreased Doc2b levels are similar to Wt mice without diabetogenic stimuli (28,30,31); these data do not support the hypothesis that Doc2b overexpression acts via promoting proliferation or differentiation in the β-cell per se. The data reported here suggest that Doc2b overexpression promotes β-cell survival and protects β-cell mass from apoptosis and ER stress, with the protection occurring at the level of the β-cell itself. This hypothesis is supported by our data showing the antiapoptotic actions of hDoc2b overexpression in clonal β-cell lines.

Doc2b is a C2 domain–containing protein in a family of proteins that includes Munc13, synaptotagmins, ferlins, rabphilin 3A, and RIM1/2 (43–45). These proteins have two aspects in common: all are calcium regulated/calcium sensing, and all are implicated in vesicle trafficking/exocytosis. The role of Doc2b as a calcium sensor in β-cells has been reported (24), and the tandem C2 domain region of Doc2b, referred to here as the C2AB, is sufficient to support membrane fusion in cell-free assays and shows stronger liposome binding than either of the isolated C2A and C2B domains (42). Indeed, our data show that C2AB can fully recapitulate the antiapoptotic effects of full-length Doc2b, likely through calcium handling. The CaMut form of C2AB has been previously reported to abolish Doc2b’s calcium binding capacity (40). ER stress mechanisms are known to be calcium-dependent events, and thus we propose that Doc2b may sequester calcium, resulting in reduced CC3, PARP, CHOP, and p-eIF2a.

C2AB also recapitulated the GSIS enhancement seen with full-length Doc2b, whereas the C2 domains individually did not. Interestingly, the C2AB protein also recapitulated the activating effect of full-length Doc2b upon STX4, consistent with prior work showing Doc2b to promote SNARE protein assembly (21,23,24,46). C2AB may confer its SNARE complex–promoting effect, in part, by acting directly on the membrane, as recent work shows that C2AB is sufficient to assemble in a ring-like oligomer (47) and act directly on membranes to regulate hemifusion events in neurons (48). These findings are important starting points for further mechanistic investigation of how the exocytotic function of Doc2b might protect β-cell function.

In summary, the data presented here show a novel, antiapoptotic role for Doc2b in the β-cell. We propose that β-cell–specific overexpression of Doc2b is sufficient to enhance whole-body glucose homeostasis and to protect the β-cell mass against type 1 diabetes–related stress. Furthermore, we show that the tandem C2AB domain is sufficient to enhance insulin secretion and protect against apoptosis and that this may be due to calcium handling. Doc2b thus carries the potential to be a therapeutic target for prevention/management of type 1 diabetes.

Acknowledgments. The authors are grateful to Drs. Jeffrey Elmendorf, Carmella Evans-Molina, A.J. Baucum (all from Indiana University), and Patrick Fueger (now at City of Hope) for their sage advice and feedback on this project emanating from the doctoral dissertation of A.A. The authors also thank Drs. Edwin Chapman and Uri Ashery for the calcium ligand mutant (CaMut-GFP) and C2AB-GFP plasmids, respectively. The authors thank Jeff Rawson (City of Hope) and the City of Hope Islet Core for isolating the mouse islets used for this study. Dr. Nancy Linford (Linford Biomedical, Iceland) provided editing assistance.

Funding. This work was supported by grants from the American Heart Association (17POST33661194 to J.Z. and 15PRE21970002 to R.T.), the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK067912 and R01 DK102233 to D.C.T.; Indiana CTSI Predoctoral Award funded in part by grant UL1TR001108 [A. Shekhar, PI] to A.A.), and JDRF (INO-2014-165-A-V to D.C.T.). Additional financial support was provided to D.C.T. from the Ruth and Robert Lanman Endowment, the George Schaeffer Award, and an Excellence Award through the City of Hope. Research reported in this publication also includes work performed in the Integrative Genomics Core, Drug Discovery and Structural Biology Core, Light Microscopy/Digital Imaging Core, Pathology Research Services Core, and the Comprehensive Metabolic Phenotyping Core (all affiliated with City of Hope), all supported by the National Cancer Institute of the National Institutes of Health under award number P30CA33572.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.A. performed the majority of the studies, contributed to the discussion, and wrote and edited the manuscript. E.O. designed and generated the TRE-Doc2b mice and contributed to the discussion. E.M.O. assisted with in vivo studies, TUNEL staining, immunohistochemistry, and islet cell morphometrics. J.Z. performed quantitative real-time PCR analyses and contributed to discussion. M.A. generated the Doc2b adenovirus and contributed to discussion. A.S.M.M. assisted with immunofluorescent imaging of pancreata. R.T., V.A.S., and R.V. assisted with in vivo studies and contributed to the discussion. D.C.T. conceived of the study; contributed to the discussion; and wrote, reviewed, and edited the manuscript. All authors read and approved the final version of the manuscript. D.C.T. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Portions of this work were presented at the Levine-Riggs Symposium, Orlando, FL, 28–31 March 2017, and the 76th Scientific Sessions of the American Diabetes Association, New Orleans, LA, 10–14 June 2016.