Genetic Disruption of Adenosine Kinase in Mouse Pancreatic β-Cells Protects Against High-Fat Diet–Induced Glucose Intolerance

Diabetes is a pathologic state of disrupted glucose homeostasis characterized by an absolute or relative insulin deficiency and a loss of insulin-producing β-cells. In type 2 diabetes (T2D), β-cell failure results from a multifactorial process initiated by insulin resistance, often in the setting of obesity (1–3). In T2D, a variety of insults contribute to progressive β-cell failure, including endoplasmic reticulum stress, inflammatory cytokines, excess reactive oxygen species, and glycolipid toxicity (2). β-Cell loss occurs through a combination of increased apoptosis and dedifferentiation, although the relative contribution of these outcomes remains unclear (3–6). Presently, a major research goal is to understand the molecular mechanisms of β-cell failure and devise strategies to reverse this process.

Although T2D is accompanied by reduced insulin secretion in late disease, increased insulin secretion is an early adaptation to insulin resistance (7,8). Of note, individuals without diabetes with a high genetic risk for diabetes have a reduced glucose-stimulated insulin response (9), but whether this is a consequence of defective β-cell function or deficient β-cell mass is unclear. T2D-associated risk alleles implicate genes that participate in both processes (e.g., CDKN2A, KCNQ1 [10–12]). Murine studies have demonstrated a central role for β-cell mass plasticity in the accommodation of obesity-associated insulin resistance (13). Although adaptive β-cell expansion is less evident in humans, obese humans without diabetes have a 1.5-fold increase in β-cell mass and increased β-cell number (14). Hence, human β-cell mass possibly exhibits modest plasticity that influences an individual’s susceptibility to T2D (15).

In mature animals, numerous potential sources of new β-cells have been identified (16); however, the usual source of new β-cells is previously existing β-cells (17,18). Consequently, understanding the signals that control self-duplication is critical to understanding how β-cell mass is controlled. To identify molecular mechanisms that regulate β-cell growth, we developed a primary islet cell–based small-molecule screening platform (19,20). With this platform, we uncovered the β-cell replication-promoting activity of adenosine kinase inhibitors (ADKIs). ADK is a broadly expressed metabolic enzyme that controls extracellular and intracellular adenosine pools through its enzymatic activity: conversion of adenosine to AMP (21). Several lines of evidence indicate that ADKIs promote β-cell replication, in part through ADK inhibition: Multiple structurally dissimilar ADKIs promote β-cell replication, ADK-directed RNA interference triggers cell autonomous β-cell replication, and an independent screen for β-cell regeneration-promoting compounds identifies distinct ADKIs (19,22). Additional activities of some ADKIs contribute to their β-cell replication-promoting activity. For example, 5-iodotubercidin (5-IT) has been shown to promote human β-cell replication through inhibition of the dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A) (23–25). To investigate the function of ADK in β-cells, we generated mice conditionally targeted at the ADK locus and tested the hypothesis that ADK acts as a negative regulator of β-cell replication and limits the adaptive response of β-cells to diabetogenic challenge.

All animal work was approved and carried out in accordance with our institutional animal care and use committee and the Guide for the Care and Use of Laboratory Animals. Adk-targeted mice were generated from International Knockout Mouse Consortium clone EUCE0154a03 (26). Genotyping of the following ADK-targeted mice was performed by PCR: wild-type allele (ADK_F [5′-AGCCTAGACTACACAACAAG-3′] and ADK_R [5′-GCTCAATCACCTAGATGGCC-3′]), Adk-targeted allele (ADK¹) (ADK_F and B32 [5′-CAAGGCGATTAAGTTGGGTAACG-3′]), and nonmutagenic orientation (ADK²) (B32 and ADK_6898 [5′-TCAAGCCCTTTGTACACCCTAAG-3′]) (Fig. 1A). The FLP deleter strain (FLPo-10, JAX 011065) was used to convert ADK¹ to ADK². Two Cre-expressing strains were used for constitutive [Tg(Ins2-Cre)^25Mgn, JAX 0035731] and conditional (Ins1-Cre/ERT^1Lphi, JAX 024709) recombination. For Cre/ERT activation, tamoxifen (100 mg/kg, #13258; Cayman Chemical) was dissolved in 100% ethanol solubilized in sterile warm corn oil (10 mg/mL, C8267; Sigma) and administered by intraperitoneal injection on 4 consecutive days. β-Galactosidase activity was used to confirm efficient tamoxifen-induced recombination. One week after tamoxifen injection, pancreata were excised and fixed (4% paraformaldehyde/PBS at 4°C for 30 min), washed with PBS, and placed in 30% sucrose/PBS for 48 h at 4°C before freezing in optimal cutting temperature compound (Tissue Tek). Animals were fed normal chow (NC) (2018; Harlan Teklad) or a high-fat diet (HFD) (D12492; Research Diets) as indicated. Experiments were conducted in female mice unless stated otherwise. Mice were of mixed background (129P2/OlaHsd and C57BL/6J).

Isolated islets and homogenized liver tissue lysates were used for SDS-PAGE (RIPA Lysis System, sc-24948; Santa Cruz Biotechnology). Islet-restricted ADK disruption was assessed by using liver and islet lysates from adult βADKO and control animals. Islets were isolated as previously described (20). Anti-ADK (1:200, sc-32908) and anti-enolase (1:500, sc-15343) were used for protein detection.

All glucose physiology experiments were performed on age- and sex-matched cohorts. Glucose measurements were made at the indicated times with a TRUEresult glucometer and TRUEtest strips. For fasting glucose measurement and intraperitoneal glucose tolerance testing (IPGTT), mice were weighed and fasted overnight (14 h, 6:00 p.m.–8:00 a.m.). Glucose levels were obtained at 0 min before injection of glucose 1.5–2.0 g/kg i.p. (10 μL/g) and at the subsequent indicated time points. Random glucose levels were obtained for fed mice at 10:00 a.m. Intraperitoneal insulin tolerance testing (IPITT) was performed by using 0.5–1.5 units/kg Humulin (R-100; Eli Lilly) as indicated after a 4-h fast starting at 8:00 a.m. In vivo glucose-stimulated insulin secretion (GSIS) was measured in blood collected from overnight fasted mice at time 0 and after injection of glucose 3 g/kg i.p. Insulin was measured with Alpco Mouse Ultrasensitive Insulin ELISA kits (80-INSMSU-E01). ADK^2/+ Rip-Cre animals displayed no detectable phenotype and were included with ADK^+/+ Rip-Cre as control animals.

Tissue β-galactosidase staining was performed on fresh frozen 25-μm sections that were fixed for 5 min with cold 4% paraformaldehyde/PBS and washed in cold PBS. Slides were incubated in X-gal staining solution containing PBS with potassium ferricyanide (5 mmol/L, P8131; Sigma), potassium ferricyanide trihydrate (5 mmol/L, P3289; Sigma), magnesium chloride (2 mmol/L; Promega), and X-gal (20 mg/100 mL; Promega) at 37°C. Slides were counterstained as indicated. Antigen retrieval was performed by heating slides to 90°C for 10 min in 10 mmol/L sodium citrate (pH 6.0) solution. Primary antibodies were incubated overnight at 4°C. To measure islet size distribution, pancreatic sections were insulin (1:300, A0564; Dako) or glucagon (1:1,000, A0565; Dako) stained along with DAPI. Affinity-purified secondary antibodies (donkey) were obtained from Jackson ImmunoResearch (1:300). Entire sections (eight per mouse) were scanned with an inverted Nikon spinning disk confocal microscope. Images were stitched together and analyzed with Volocity 6 image analysis software (PerkinElmer). Cellular clusters of >900 μm² were counted. Total pancreatic area was determined by the DAPI-positive area. The β-cell mass was calculated by multiplying pancreatic mass by the percentage of pancreatic area that was insulin positive. β-Cell replication indices were determined by BrdU incorporation (1:50, M074401-8; Dako) and/or ki67 expression (1:300, 556003; BD Bioscience); ki67 detection was performed by using biotin amplification (1:200; Jackson ImmunoResearch). β-Cells were identified by the expression of insulin or PDX-1 (1:200, AF2419; R&D Systems); use of insulin staining to confirm findings obtained with PDX-1 is necessary given that δ-cells also express PDX-1 (27). A minimum of 2,000 β-cells from nonconsecutive sections (>50 μmol/L apart) were used to determine the β-cell replication rate for each animal. Replication events were adjudicated in a blinded fashion. Mice were provided BrdU-containing water (0.8 mg/mL, 228590100; Acros) for 1 week in opaque bottles that were changed every 2 days.

Islets were isolated from 36-week vehicle-injected (n = 4) and tamoxifen-injected (n = 4) iβADKO mice. Injections were performed at age 18 weeks to avoid any developmental impact, and islets were harvested after an 18-week delay to avoid potential short-term tamoxifen and recombination effects. Islets were dispersed, plated (allowed 60 h to attach), treated (60 h duration), fixed, and assayed (PDX-1 and ki67 staining) as previously described (19,20).

Experimental data are presented as scatter plots with an adjacent representation of the statistical mean that includes an error bar representing the SD. Statistically significant differences were determined by using Student two-tailed t test where P ≤ 0.05 was taken to be significant. Experimental results were confirmed in independent experimentation in all cases except for the in vivo β-cell replication experiments.

To study the in vivo function of ADK in β-cells, we generated mice conditionally targeted at the Adk locus (Fig. 1A). The integrated mutagenic orientation (ADK¹) placed a highly efficient splice-acceptor sequence (SA)/βgeo expression cassette (β-galactosidase-neomycin resistance fusion gene) in frame with exons 1a and 1b, the ADK transcriptional start sites that encode ADK-short (cytoplasmic) and ADK-long (nuclear) isoforms, respectively (28). β-Galactosidase activity was used to assess ADK expression in the pancreas of Adk^1/+ mice (Fig. 1B, left panels). ADK was highly expressed in the exocrine pancreas and modestly expressed in the endocrine pancreas where the nuclear-located isoform (ADK-long) predominates. Newborn litters from heterozygous (Adk^1/+ × Adk^1/+) breeding pairs contained Adk^1/+ pups at Mendelian frequency. However, consistent with the published phenotype of ADK-null animals, no ADK^1/1 mice were identified at weaning (>150 mice screened) (29). Furthermore, ADK protein was nearly undetectable in hepatocyte lysates from Adk^1/1 newborn pups (Fig. 1C). These results indicate efficient disruption of ADK expression by the targeting strategy.

We disrupted ADK expression in β-cells (βADKO mice) by using the Rip-Cre driver strain (30). We confirmed β-cell–restricted disruption by measuring β-galactosidase activity in pancreatic tissue sections from 1) mice harboring the gene-trap vector in the nonmutagenic orientation (Adk^2/+ mice), 2) Rip-Cre mice, and 3) Rip-Cre Adk^2/+ mice (βADKO-Het) (Fig. 1B). As anticipated, pancreatic sections from Adk^2/+ and Rip-Cre animals lacked β-galactosidase activity, whereas βADKO-Het sections displayed β-cell–restricted β-galactosidase activity. Of note, Rip-Cre mice have recombination activity in the brain and exhibit glucose intolerance (31,32). Consequently, all studies included Rip-Cre control animals. Next, we assessed ADK disruption by Western blot of liver and islet lysates from Rip-Cre, βADKO-Het, and βADKO mice (Fig. 1D). Whereas hepatic ADK expression was similar among genotypes, islet ADK expression was reduced in βADKO animals. Residual ADK expression in islet samples might reflect exocrine and/or insulin-negative islet cell contamination. Taken together, these data indicate successful disruption of ADK expression in pancreatic β-cells.

Young βADKO animals were heavier than Rip-Cre control littermates (Fig. 2A). At 8 weeks of age, female βADKO mice weighed ∼2 g more than Rip-Cre control animals (18.3 ± 0.3 vs. 16.1 ± 0.6 g; P < 0.01). As mice aged, the body weights of βADKO and Rip-Cre mice converged; no statistical difference was present beyond 14 weeks (P > 0.05). By 52 weeks, the body weights of Rip-Cre (28.9 ± 1.4 g) and βADKO (29.2 ± 1.5 g) were indistinguishable. HFD-fed 13-week βADKO and Rip-Cre mice demonstrated increased weight gain compared with genotype-matched NC-fed mice beginning 2 weeks after HFD initiation (P < 0.05 for weeks 15–21). HFD-induced weight gain in Rip-Cre and βADKO mice was similar (P > 0.05 at every time point) (Fig. 2B). Thus, βADKO animals were transiently heavier than control animals but not predisposed to increased HFD-induced weight gain.

We next assessed the impact of β-cell–targeted ADK disruption on glucose tolerance by performing IPGTTs every 4–8 weeks on NC-fed Rip-Cre and βADKO mice. No difference was observed until 52 weeks when subtle improvement in the IPGTT of βADKO mice emerged (Fig. 2C and D). To assess a potential protective role of ADK disruption against diabetogenic challenge, IPGTTs were performed on Rip-Cre and βADKO mice placed on an HFD. After only 2 weeks of HFD, female βADKO mice displayed lower fasting glucose levels but similar glucose tolerance (Fig. 2E). After 6 weeks and 18 weeks of HFD, female βADKO mice demonstrated substantial improvement in IPGTT (Fig. 2F and G). A similar improvement in IPGTT was also observed in HFD-fed male βADKO mice (Fig. 2H). Fasting and random-fed glucose values were also significantly improved in HFD-fed βADKO mice (Fig. 2I and J). Thus, βADKO animals were resistant to HFD-dependent impairment of glucose homeostasis.

To assess the physiologic basis of improved IPGTT in HFD-fed βADKO mice, we performed IPITT (Fig. 3A and B). As anticipated, βADKO fasting glucose values were lower than that of controls (154 ± 19 vs. 214 ± 34 mg/dL; P < 0.01). Consequently, we compared glucose clearance by βADKO and Rip-Cre animals with (Fig. 3A) and without (Fig. 3B) normalizing to starting glucose values. βADKO mice displayed more-robust insulin responsiveness at 30 and 60 min after insulin injection. However, the absolute drop in blood glucose levels in response to insulin was greater for Rip-Cre mice than for βADKO mice (156 ± 29 vs. 124 ± 18 mg/dL, respectively; P < 0.05). These results indicate that HFD-fed βADKO mice had modestly enhanced insulin sensitivity, but given the subtlety of this difference, other factors were likely contributing to the improved glucose tolerance of the βADKO mice.

We next evaluated whether enhanced insulin secretion contributed to the improved glucose tolerance of βADKO mice by measuring GSIS in NC- and HFD-fed Rip-Cre and βADKO mice (Fig. 3C). As anticipated HFD-fed mice displayed increased insulin secretion relative to NC-fed mice. Both NC- and HFD-fed βADKO mice displayed increased insulin secretion compared with Rip-Cre controls (Fig. 3C and D). The increased GSIS of NC-fed mice was unexpected because no difference in IPGTT was detected. However, this difference was subtle (NC area under the curve [AUC] Rip-Cre 8.1 ± 2.1 vs. βADKO 13.6 ± 1.9; P < 0.05) compared with HFD-fed mice (HFD AUC Rip-Cre 32.31 ± 3.7 vs. βADKO 61.4 ± 9.9; P < 0.05).

Islet isolations from HFD-fed βADKO mice demonstrated increased yield (1.3 ± 0.2-fold; P < 0.01) compared with Rip-Cre mice (Fig. 4A). This observation led us to consider two potential explanations: 1) HFD-fed βADKO mice had more islets, or 2) HFD-fed βADKO mice had larger islets that were more efficiently isolated. To assess these possibilities, islet size and number were analyzed in pancreatic sections from HFD-fed Rip-Cre and βADKO animals. Visual inspection of insulin-stained sections gave the impression that βADKO islets were enlarged (Fig. 4B). To quantitate this impression, we measured the insulin-stained area. Consistent with visual observation, βADKO animals demonstrated a right shift in insulin cluster size (Fig. 4C). In addition, the average insulin-positive cluster size was increased in βADKO mice (average insulin area, βADKO 7,790 ± 694 vs. Rip-Cre 3,857 ± 952 μm²; P < 0.01) (Fig. 4D). Furthermore, the insulin-positive area and β-cell mass but not pancreatic area or pancreatic mass were increased in βADKO mice (Fig. 4E–I). No increase in the number of insulin-positive clusters was detected (Rip-Cre 6.34 ± 1.9 vs. βADKO 7.6 ± 0.4; P = 0.4) (Fig. 4J). No change in α-cell area was detected (Fig. 4K), and β-cell proliferation in HFD-fed mice, as measured by BrdU incorporation, demonstrated a trend toward an increase in the βADKO mice (Fig. 4L). Taken together, these results indicate that β-cell expansion rather than increased islet number was present in HFD-fed βADKO mice.

To mitigate deficiencies of the Rip-Cre transgene Tg(Ins2-Cre)^25Mgn, we reevaluated our findings with the inducible and β-cell–specific Cre-expressing Mip-Cre/ERT mouse line (Ins1-Cre/ERT^1Lphi) (31,32). Because Mip-Cre/ERT mice have tamoxifen-independent growth hormone expression, we used vehicle-treated iβADKO mice as control subjects (33).

We evaluated the impact of tamoxifen-induced β-cell–specific ADK disruption on glucose homeostasis and β-cell replication in mature animals (Fig. 5A). Tamoxifen-dependent disruption of ADK expression was confirmed by induction of β-galactosidase activity (Fig. 5B). Similar to NC-fed βADKO mice, NC-fed vehicle- and tamoxifen-treated iβADKO mice demonstrated comparable IPGTTs (Fig. 5C). Consequently, tamoxifen treatment had no impact on the IPGTT. Also consistent with the βADKO phenotype, HFD-fed tamoxifen-treated mice demonstrated significantly lower fed glucose values (Fig. 5D) and enhanced IPGTT after 3 (Fig. 5E) and 11 (Fig. 5F) weeks of HFD. By contrast, insulin sensitivity measured by IPITT was unchanged (Fig. 5G), perhaps indicating an impact of ectopic recombinase activity in βADKO mice. Finally, we assessed whether disruption of ADK enhanced in vivo GSIS. Indeed, tamoxifen-treated mice demonstrated increased insulin secretion after glucose challenge (15-min vehicle treatment 0.15 ± 0.02 vs. tamoxifen-treatment 0.24 ± 0.04 ng/mL; P = 0.01) (Fig. 5H). These data indicate that β-cell–specific disruption of ADK in adult mice was protective against HFD-induced glucose intolerance and enhanced GSIS in vivo.

We next assessed the impact of ADK disruption on mature β-cell proliferation. We determined the percentage of insulin-positive cells that coexpressed ki67 in vehicle- and tamoxifen-treated mice fed either NC or HFD. Loss of ADK expression had no effect on β-cell replication in NC-fed mice (replication of vehicle 0.36 ± 0.1% vs. tamoxifen 0.50 ± 0.10%; P = 0.38); however, ki67 expression was significantly increased in HFD-fed tamoxifen-treated mice (vehicle 1.59 ± 0.10% vs. tamoxifen 2.70 ± 0.32%; P < 0.05) (Fig. 5I). We confirmed this finding by using distinct β-cell (PDX-1) and replication (BrdU incorporation) markers (vehicle 1.00 ± 0.26% vs. tamoxifen 2.33 ± 0.26%; P < 0.05) (Fig. 5J). Representative images used to quantify β-cell replication are shown in Fig. 5K.

Finally, we compared the in vitro basal and compound-induced replication index of β-cells obtained from vehicle- and tamoxifen-injected iβADKO mice (Fig. 5L). Islet ADK deletion was confirmed by Western blot (Fig. 5M). ADK-deficient β-cells (tamoxifen-injected) displayed increased basal replication compared with ADK-expressing β-cells (Fig. 5L). In addition, ADK-expressing β-cells displayed an approximately twofold replication increase after treatment with 5-IT, an ADK and DYRK1A/B inhibitor (Fig. 5L) (19,20). Similarly, 5-IT treatment of ADK-deficient β-cells demonstrated an ∼1.5-fold increase in β-cell replication. Therefore, disruption of ADK expression in β-cells increases the basal β-cell replication index in vitro but does not eliminate 5-IT–dependent replication induction. These results indicate that ADK negatively regulates β-cell replication in vitro and that the β-cell replication-promoting activity of 5-IT was not mediated entirely through ADK inhibition.

Identification of novel strategies to enhance in vivo β-cell proliferation and mass while retaining optimum function is an attractive therapeutic strategy for diabetes. Previously, we discovered that short-term treatment with ADKIs stimulated rodent and porcine β-cell replication (19). In the current study, we used a novel genetic mouse model to study the function of ADK in β-cells and to contemplate the potential utility of ADK inhibition. This study addressed two critical questions: 1) is long-term disruption of ADK detrimental to β-cell function, and 2) are mice lacking ADK expression in their β-cells protected against diabetogenic insults, such as aging and HFD? By using two different mouse models, we found β-cell–selective disruption of ADK expression to be well tolerated and protective against HFD challenge.

Currently available ADKIs are not amenable to long-term in vivo treatment because of associated toxicity (19,34). In addition, global disruption of ADK expression results in perinatal lethality (29). These findings raise concern about potential detrimental effects of ADK disruption on β-cell viability and function. We found that ADK expression in the islet was low relative to the exocrine pancreas and that disruption of ADK had a limited impact on β-cell development, function, and growth. NC-fed mice lacking β-cell expression of ADK, either constitutively or acutely, displayed glucose homeostasis parameters similar to control mice, with no detectable difference in β-cell replication or mass. Hence, disruption of ADK in β-cells was well tolerated.

Although NC-fed ADK-deficient mice were essentially indistinguishable from control animals, a subtle but consistent improvement in IPGTT emerged as animals aged beyond 1 year. To better gauge the potential benefit of ADK disruption on glucose homeostasis, βADKO mice were challenged with an HFD. Indeed, HFD-fed mice harboring ADK-deficient β-cells displayed significantly enhanced IPGTT, enhanced in vivo GSIS, and increased β-cell mass. No change in α-cell mass was observed, indicating that the impact was cell autonomous and/or lineage specific. In addition, we found no evidence for islet neogenesis because islet density (insulin clusters per section) was unchanged in HFD-fed βADKO mice. Although βADKO mice maintained on HFD for several months displayed increased β-cell mass, active β-cell replication was not significantly increased in these animals. By contrast, β-cell replication was increased 1 week after conditional disruption of ADK expression. These observations are consistent with prior work demonstrating that β-cell replication rates initially increase in response to HFD but decline with prolonged exposure (35). Of note, ADK-deficient β-cells demonstrated an increased in vitro replication index compared with control β-cells. The ADKI 5-IT that promotes human β-cell replication, in part through DYRK1A/B inhibition (25), further enhanced the replication of ADK-deficient β-cells, suggesting that inhibition of both ADK and DYRK1A/B contribute to rodent β-cell replication control. In summary, disruption of ADK expression in β-cells was protective against HFD-induced glucose intolerance in mice in part as a result of an increased β-cell replication response, β-cell mass expansion, and enhanced insulin secretion.

Numerous mouse models of β-cell–specific gene disruption yield altered β-cell mass and/or glucose homeostasis (36). However, the majority of these models demonstrate an impact under basal conditions. Disruption of ADK expression in β-cells is distinct in that essentially no phenotype has been identified in unchallenged animals. These findings suggest that ADK plays a role in dampening the adaptive β-cell response to metabolic challenge. Defining factors that specifically modulate the adaptive β-cell response is an important avenue for uncovering therapeutic strategies for T2D. Similar to ADK, connective tissue growth factor (CTGF) contributes to the adaptive response of β-cells. Overexpression of CTGF in mature β-cells has no impact on β-cell proliferation but enhances β-cell expansion under diabetogenic conditions (37). Similarly, β-cell–specific deletion of the prolactin receptor has no impact on β-cell mass or function under basal conditions but predisposes female mice to gestational diabetes mellitus by dampening the β-cell replication response during pregnancy (38). Although the molecular links among CTGF, prolactin signaling, and ADK are not immediately obvious, these models highlight the potential to therapeutically manipulate the adaptive response of β-cells. Perhaps more relevant, β-cell–specific deletion of the adenosine receptor 2a (Adora2a) in mice was recently shown to impair glucose regulation and β-cell proliferation in pregnancy while having no impact on these parameters under basal conditions (39), indicating a role for adenosine signaling in promoting the adaptive β-cell response to an insulin-resistant state. A primary function of ADK is to reduce extracellular adenosine levels through adenosine phosphorylation; hence, disruption of ADK in β-cells might promote an adaptive β-cell response in vivo by increasing extracellular adenosine levels and augmenting Adora2a-dependent signaling.

The current study has notable limitations. First, caution must be taken with use of the Rip-Cre and MIP-Cre/ERT transgenic lines (32,33). Prior studies have shown detrimental effects of the Rip-Cre cassette on glucose tolerance and insulin secretion, findings opposite to what we observed in the βADKO mice. Consequently, impacts of the Rip-Cre line are anticipated to bias results away from the observed protective effect of ADK deletion on HFD-induced glucose intolerance. In addition, complementary findings obtained with the constitutive and inducible Cre driver strains excluded developmental impacts and ectopic gene deletion as the probable basis of the enhanced β-cell replication, mass, and function we observed. Second, the applicability of the findings to human β-cells is unknown because adult human β-cells have limited regenerative capacity (14,40). Finally, the potential applications of in vivo ADK inhibition may be limited by the function of ADK in other tissues, such as the liver. Indeed, identifying methods for lineage-restricted drug delivery is likely to be a critical hurdle for developing regenerative therapies for diabetes. Future efforts will focus on better understanding the mechanism by which ADK disruption enhances the adaptive response of β-cells to HFD challenge.

Acknowledgments. The authors thank Fredric Kraemer (Stanford University School of Medicine) for generous support of this work. The authors also thank Sara Sun (Stanford University) for technical contributions to this work.

Funding. This research was supported by the Friedenrich BII Diabetes Fund, the SPARK Translational Research Program, and the Child Health Research Institute at Stanford University (National Institutes of Health National Center for Advancing Translational Sciences Clinical and Translational Science Award UL1-TR-001085 and National Institute of Diabetes and Digestive and Kidney Diseases grant R01-DK-101530).

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

Author Contributions. G.N. conceived and designed experiments, performed experiments, analyzed data, and approved the final version of the manuscript. Y.A. conceived and performed experiments, analyzed data, and assisted in writing the manuscript. Z.Z., H.H., and S.L. performed experiments, analyzed data, and approved the final version of the manuscript. N.A.A. analyzed data and assisted in writing of the manuscript. J.P.A. conceived and designed experiments, analyzed data, and wrote the manuscript. J.P.A. 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. Parts of this study were presented at the 97th Endocrine Society's Annual Meeting and Expo, San Diego, CA, 5–7 March 2015.