Diabetes Volume 66, July 2017 1879

Prolonged Elimination of Negative Feedback Control Mechanisms Along the Signaling Pathway Impairs b-Cell Function In Vivo

Roi Isaac, Yaron Vinik, Sigalit Boura-Halfon, Lydia Farack, Sarina Streim, Eytan Elhanany, Zvi Kam, and Yehiel Zick

Diabetes 2017;66:1879–1889 | https://doi.org/10.2337/db16-0827

Cellular stress and proinflammatory cytokines induce critical roles in pancreatic b-cells (1). Decreased IRS2 ex- phosphorylation of insulin receptor substrate (IRS) pro- pression causes b-cell apoptosis (1,2) and mice lacking IRS2 teins at Ser sites that inhibit insulin and IGF-I signaling. We develop diabetes 8–10 weeks after birth due to reduced therefore examined the effects of mutation of five “inhib- b-cell mass and impaired b-cell function (1). Conversely, itory” Ser phosphorylation sites on IRS2 function in trans- increased IRS2 expression promotes b-cell survival (3) 2 2 genic mice that overexpress, selectively in pancreatic and prevents diabetes in Irs2 / mice (4). b-cells, either wild-type (WT) or a mutated IRS2 STUDIES ISLET 5A IRS have a conserved amino terminus, which (IRS2 ). Islets size, number, and mRNA levels of catalase contains a pleckstrin homology domain flanked by a P-Tyr and superoxide dismutase were increased, whereas binding domain that interacts with the juxtamembrane those of nitric oxide synthase were decreased, in 7- to domains of the insulin and IGF-I receptors (5,6). IRS pro- 10-week-old IRS25A-b mice compared with IRS2WT-b teins undergo phosphorylation on multiple Tyr residues at mice. However, glucose homeostasis and insulin secre- their COOH-terminal region, which serves as a docking site tion in IRS25A-b mice were impaired when compared with IRS2WT-b mice or to nontransgenic mice. This was asso- for SH2-containing proteins that further propagate insulin ciated with reduced mRNA levels of Glut2 and islet b-cell and IGF-I signals (7). . transcription factors such as Nkx6.1 and MafA. Similarly, IRS proteins contain 70 potential Ser/Thr phosphory- components mediating the unfolded protein response lationsitesforkinasessuchascAMP-dependentproteinki- were decreased in islets of IRS25A-b mice in accordance nase, protein kinase C, and Akt (reviewed by Shanik et al. [8] with their decreased insulin secretion. The beneficial ef- and Boura-Halfon and Zick [9]). Insulin-induced Ser/Thr fects of IRS25A on b-cell proliferation and b-cell transcrip- phosphorylation of IRS proteins dissociates them from tion factors were evident only in 5- to 8-day-old mice. the insulin receptor (IR), prevents their Tyr phosphoryla- These findings suggest that elimination of inhibitory Ser tion, and inhibits their interactions with downstream effec- phosphorylation sites of IRS2 exerts short-term beneficial tors (10). This serves as a physiological negative-feedback effects in vivo; however, their sustained elimination leads control mechanism, which is used by insulin and IGF-I to to impaired b-cell function. turn off their own signaling cascades. However, proinflam- matory cytokines and other inducers of insulin resistance take advantage of this mechanism. By activating a number Insulin and IGF-I actions are mediated by their receptors of IRS kinases, they uncouple the IRS proteins from the IR that function as Tyr-kinases. Key substrates for these or IGF-I receptor and inhibit their biological activities (11). receptors are the insulin receptor substrate (IRS) proteins Accordingly, the mutation of selected inhibitory Ser sites IRS1 and IRS2, which integrate many of the pleiotropic of IRS1, located in close proximity to its P-Tyr binding effects of insulin and IGF-I. IRS proteins, mainly IRS2, play domain, renders it less prone to the action of IRS kinases.

Department of Molecular Cell Biology, Weizmann Institute of Science, Rehovot, © 2017 by the American Diabetes Association. Readers may use this article as Israel long as the work is properly cited, the use is educational and not for profit, and the Corresponding author: Yehiel Zick, [email protected]. work is not altered. More information is available at http://www.diabetesjournals .org/content/license. Received 7 July 2016 and accepted 6 April 2017. This article contains Supplementary Data online at http://diabetes .diabetesjournals.org/lookup/suppl/doi:10.2337/db16-0827/-/DC1. 1880 Effects of IRS2 on b-Cell Function In Vivo Diabetes Volume 66, July 2017

As a result, the mutated IRS1 better propagates insulin and CB6F1andICRmice,themicewerebackcrossedwith IGF-I actions (12,13). Similarly, the introduction of IRS2 C57BL/6 mice to achieve a homogeneous genetic back- proteins mutated at five potential inhibitory Ser sites ground. After six backcrosses, the new litters expressed at (IRS25A)intopancreaticb-cells for a short period of time least ;98% of the genetic background of C57BL/6 mice. better promoted insulin signaling and improved b-cell sur- vival and function in culture. Furthermore, when islets Densitometry and Statistical Analysis expressing IRS25A were transplanted into diabetic mice, The intensity of bands in autoradiograms was determined they supported glucose homeostasis better than those ex- by densitometry carried out on exposures within the linear pressing IRS2WT (14). These findings suggest that short- range. Graphic analysis was performed with National Insti- term elimination of negative-feedback control mechanisms tutes of Health image software. Results were analyzed using along the insulin/IGF-I signaling pathway improves b-cell nonparametric comparisons based upon Mann-Whitney U , fi function under stress. tests. A P value of 0.05 was considered signi cant. In To further assess the physiological significance of per- those studies for which the n values were low, the experi- manently eliminating “inhibitory” Ser phosphorylation sites, ments may be underpowered. To quantify the difference we have generated transgenic (Tg) mice that constitutively between glucose tolerance tests (GTTs), the area under overexpress either IRS2WT or IRS25A selectively in pancre- the curve (AUC) was calculated by the trapezoid rule. The 6 atic b-cells. Our findings indicate that although the overex- mean AUC SEM was determined from at least three pressed IRS2 proteins, mainly IRS25A,exertbeneficial effects individual curves for each experimental condition reported. – on the animals, mainly at a very young age (5 8days),a Additional Materials and Methods 5A b long-term expression of IRS2 impairs -cell function. The sources of materials, antibodies, and detailed methods fi fi 5A These ndings indicate that the bene cial effects of IRS2 concerning mouse genotyping; metabolic studies; mouse (14) are limited in duration and, in the long run, are over- diet; preparation of cell extracts; isolation and treatment of b ridden by its negative effects, leading eventually to -cell murine islets; islet morphology; RNA analysis; glucose- exhaustion and demise. stimulated insulin secretion (GSIS); immunofluorescence microscopy, and immunostaining analysis are provided in RESEARCH DESIGN AND METHODS the Supplementary Data. Mice Male C57BL/6J mice were housed under standard light/ RESULTS dark conditions and were given access to food and water ad Characterization of Irs25A-b Mice libitum; experiments were approved by the Animal Care and TransientexpressioninpancreaticisletsofIRS2proteins Use Committee of the Weizmann Institute of Science 5A mutated at five potential inhibitory Ser sites (IRS2 ; (Rehovot, Israel). S303A, S343A, S362A, S381A, S480A) improved insulin Generation of IRS2WT and IRS25A Tg Mice signaling as well as b-cell survival and function in culture Double-Tg heterozygous mice that express Myc-Irs2 pro- (14). Furthermore, when islets expressing Irs25A were trans- teins (wild type [WT] or 5A) selectively in pancreatic planted into diabetic mice, they supported glucose homeo- b-cells were generated by crossing mice that express stasis better than those expressing Irs2WT (14). Our goal Myc-Irs2 proteins (WT or 5A) driven by the tetracycline was to follow on these findings and determine the sustained operator–based promoter with mice that express the reverse effects of Irs25A expression under in vivo settings. To this tetracycline-regulated trans activator (rtTA)drivenbythe end, double-heterozygous Tg mice that overexpress both rat insulin 2 promoter fragment (15). In brief, pTet-Splice1 Myc-IRS2 proteins (WT or 5A), driven by the tetracycline plasmids containing the tetracycline response element and operator–basedpromoter,andthertTA, driven by the rat Myc-tagged Irs2WT or Irs25A were generated as described insulin-2 promoter fragment, were generated. These mice previously (14). The plasmids were linearized by digestion were expected to express Myc-IRS2 (WT or 5A) selectively with Xho1andNot1, gel purified, and microinjected into the in pancreatic b-cells in an inducible manner. pronuclei of CB6F1 zygotes. The mice that were born (F To examine expression of the IRS2 proteins in pancreatic generation) were scanned for the presence of Myc-Irs25A or b-cells, doxycycline was administered into the drinking wa- Myc-Irs2WT by PCR (vide infra), and positive mice were out- ter of Irs25A-b and Irs2WT-b mice. After 6 days, islets were bred with WT CB6F1 mice to generate heterozygote mice isolated and analyzed for Myc expression. As shown in Fig. 1, (F1 generation). Heterozygote mice were crossed with ICR there was no expression of Myc in the control Non-Tg mice expressing the rtTA under the control of the rat in- animals both at the mRNA and protein levels; however, sulin II promoter, consisting of the 9.5-kb 59 flanking re- there was a significant increase in Myc-Irs2 mRNA in islets gionofthegene(Rip7-rtTA) (16). Crossbreeding of the two isolated both from Irs25A-b and Irs2WT-b mice (Fig. 1A). Tg lines yielded a double-Tg heterozygote mice (Irs25A-b; Furthermore, the transgene was expressed at comparable Irs2WT-b) that activates their transgene after the adminis- levels in the two mouse lines (Fig. 1A). A marked compara- tration of tetracyclin to their drinking water. Since the bleincreaseinmRNAlevelsoftotalIrs2 (exogenous and original Tg mice were offspring of a crossbreeding of endogenous;eitherWTor5A)wasobservedintheTg diabetes.diabetesjournals.org Isaac and Associates 1881

Figure 1—Expression of Myc-Irs2 in vivo. Doxycycline (400 ng/mL doxycycline, 3% w/v sucrose) was administered in the drinking water of 8-week-old male mice (Non-Tg, Irs2WT-b,orIrs25A-b) for a period of 2 weeks (A and B)or6days(C and D). Mice were sacrificed, and islets were isolated. RNA was extracted, and mRNA levels of Myc-Irs2 (A)andTotal Irs2 (B) were determined by real-time PCR. Data were normalized to the content of actin mRNA. Protein extractions from the isolated islets were resolved by SDS-PAGE, immunoblotted with the indicated antibodies (C), and quantified (D). E: Eight-week-old male mice (Non-Tg, Irs2WT-b,orIrs25A-b)weresacrificed, and protein extractions (from isolated islets, 7.5 mg) were resolved by SDS-PAGE and immunoblotted with the indicated antibodies. F and G: The 3.5-month-old male and female mice were provided with regular drinking water (open bars) or with drinking water containing doxycycline (400 ng/mL doxycycline, 2 weeks) (black bars). Islets were isolated, RNA was extracted, and mRNA levels of Myc-Irs2 (F) and Total Irs2 (G) were determined by real-time PCR. Vertical lines in C and E indicate places where irrelevant lanes were spliced out. Data were normalized for the content of hypoxanthine phosphoribosyl transferase mRNA levels. Data shown in bar graphs are the mean 6 SEM of n $ 3 mice/group, except in E (n =2).***P < 0.001. A.U., arbitrary units; Dox, doxycycline; Myc, Myc-Irs2.

animals (Fig. 1B) that was accompanied by increased IRS2 suggesting that the TetOn system in all founders was leaky, expression in islets of the Tg animals, compared with Non- possibly due to a strong positional effects on the TetOn Tg ones, as revealed by immunoblotting using Myc or IRS2 minimal promoter (17), resulting in the expression of the antibodies (Fig. 1C–E). It revealed a 6- to 10-fold increase in transgenes from the time the insulin promoter was acti- total IRS2 protein levels in the Tg animals (Fig. 1E)thatis vated in b-cells, at approximately the 20-somite stage (18). similar to those previously reported (4) for Irs2 transgenes. To further assess the specificity of Myc-Irs2 expression, Immunohistochemistry of total IRS2 yielded similar, albeit Myc staining was performed on pancreas sections of Irs25A- lower, fold increases (vide infra). The mRNA levels of b and Irs2WT-b mice (without doxycycline administration). Myc-Irs2 and total Irs2 were also tested in other organs AsshowninFig.2A,IRS2WT and IRS25A were detected in such as the liver, lung, spleen, thymus, testicles, and kidney, most insulin-expressing cells; however, Myc staining was all of which were found negative with regard to Myc-Irs2 not detected in non–b-cells. Similarly, more intense (two- mRNA expression. Weak mRNA expression (2–4% that fold to threefold) staining of IRS2 itself was evident selec- of islets) was detected in the brain (Supplementary tively in b-cells of both young and adult mice (Fig. 2B and Fig. 1A and B). C). Of note, although the expression of Myc-IRS2WT was To determine whether the expression of Myc-IRS2 observed mainly in the cytoplasm, the expression of Myc- proteins is doxycycline dependent, mice were administered IRS25A was observed mostly in the nucleus (Fig. 2A). The doxycycline for 2 weeks. However, the transcription levels reason for this difference and its biological significance are of the exogenous Myc-Irs2 mRNA as well as total Irs2 mRNA currently unknown. were increased to comparable levels regardless of the ad- The differences in phenotypes between Irs25A-b and ministration of doxycycline (Fig. 1F and G). This effect was Irs2WT-b male mice (7–8 weeks old), fed a regular diet, also observed in two other mouse lines (data not shown), were evaluated next. As shown in Supplementary Fig. 2A, 1882 Effects of IRS2 on b-Cell Function In Vivo Diabetes Volume 66, July 2017

Figure 2—b-Cell–specific overexpression of Myc-IRS2 proteins in pancreata sections. A: Pancreas sections of 10-week-old male mice (Non-Tg, Irs2WT-b,orIrs25A-b) were stained with DAPI (blue) or were immunostained with anti-Myc or anti-insulin antibodies, followed by cyanine-3 (IRS2) or fluorescein (Insulin)-conjugated secondary antibodies, respectively. Original magnification 340. B: Pancreas sections of 6- to 8-day-old (top panels) and 10-week-old male mice (bottom panels) were stained with DAPI (blue) and immunostained with anti-IRS2 antibodies, followed by fluorescein-conjugated secondary antibodies. Original magnification 360. C: Bar graphs represent the mean of IRS2 intensity per islets. n =4–6 mice/group. *P < 0.05; ***P < 0.001.

there was no significant change in body weight between the of large islets (.30 mm2)inIrs25A-b mice (Fig. 3B), al- mouse lines, and fasting blood glucose levels of Irs2WT-b though the average number of b-cells/islet-area remained and Irs25A-b mice were indistinguishable compared with unaltered (Supplementary Fig. 3A). The overexpression of Non-Tg mice (Supplementary Fig. 2B). Irs2WT in b-cells also resulted in an ;1.5-fold increase in the number of islets per pancreas section compared with Effects of IRS2 Overexpression on Islet Size and Number Non-Tg mice. A further ;1.5-fold increase was observed in It has been previously shown that the overexpression of Irs25A-b mice compared with Irs2WT-b mice (Fig. 3C). Thus, IRS2 in b-cells of Tg mice does not change b-cell content our data support previous findings regarding the effects of per islet but leads to a twofold increase in islet area, mainly IRS2WT on islets morphology and density, whereas the over- due to an increased number of normal-sized islets (4). Con- expression of IRS25A further enhances these effects. Yet, sistent with these observations, the mean islet area in despite the hyperplastic nature of the islets derived from Irs2WT-b mice was increased by ;25% compared with their 10-week-old Irs25A-b mice, we failed to detect an increase in Non-Tg littermates, whereas the overexpression of IRS25A the cell proliferation marker Ki67 in these islets (Supple- increased islets area by ;60% compared with Non-Tg mice mentary Fig. 3B and C), although a trend of increased (Fig. 3A). This effect was mainly due to a greater proportion cell proliferation, determined by a twofold increase in diabetes.diabetesjournals.org Isaac and Associates 1883

Figure 3—Effects of Myc-IRS2 expression on islets number and size. A: Pancreas sections from 10-week-old male mice (Non-Tg; Irs2WT-b,or Irs25A-b)werefixed and embedded in paraffin for hematoxylin-eosin staining. Representative sections are shown. Original magnification 34. Bar graphs are the mean islet area for each animal type measured with panoramic viewer and ImageJ software. B: Bar graphs represent the distribution (%) of islets areas. C: Bar graphs represent the mean of number of islets per section (see RESEARCH DESIGN AND METHODS;Non-Tgn =5; Irs2WT-b n =3;Irs25A-b n = 4 mice). D:Paraffin-embedded sections were stained with DAPI and were immunostained with antiinsulin and anti- PCNA antibodies, followed by cyanine-3- (PCNA) or cyanine-2 (Insulin)-conjugated secondary antibodies. Bar graphs are the percentage of PCNA+/insulin+ cells. More than 900 insulin+ b-cells were counted per mouse (Non-Tg n =4;Irs2WT-b n =4;Irs25A-b n =3mice).*P < 0.05; **P < 0.01; ***P < 0.001.

proliferating cell nuclear antigen (PCNA)-positive cells, was supportive of the concept that IRS proteins can reduce observed in islets derived from Irs2WT-b mice, and an even the oxygen burden of pancreatic islets through the activa- greater increase was observed in islets derived from Irs25A-b tion of Akt and its downstream target FOXO1 (21) and mice (Fig. 3D and Supplementary Fig. 3D). The increase in exert beneficial effects on b-cell function. PCNA expression could, however, be accounted for, at least in part, by the fact that PCNA marks not only actively di- Effects of IRS2 Overexpression on GTTs and viding cells but also those in the process of DNA repair (19). Insulin Secretion The effects of IRS2 overexpression on GTT results and Overexpression of IRS Proteins Increases the mRNA insulin secretion were evaluated next. Unlike expectations, Levels of Antioxidative Proteins Irs2WT-b mice responded like Non-Tg animals, whereas b-Cells are highly sensitive to reactive oxygen species that Irs25A-b male mice were glucose intolerant (Fig. 5A). Ac- induce b-cell dysfunction (20). We therefore studied the cordingly, the AUC of Irs25A-b mice increased by ;27% mRNA levels of catalase (21) and superoxide dismutase compared with Non-Tg mice (Fig. 5B). The same trend (SOD) (22) in islets derived from wt and Irs2 Tg mice. As was also observed in IRS25A-b female mice (data not shown in Fig. 4, the mRNA levels of catalase, sod1,andsod2 shown). Peripheral insulin resistance and/or b-cell dysfunc- were significantly increased in islets derived from Irs2WT-b tion are the main reasons for impaired GTT results, there- mice, and an even greater elevation was observed in islets fore GSIS was analyzed in islets isolated from these mice. derived from Irs25A-b mice. These effects were evident in The GSIS of Irs2WT-b mice was comparable to that of the islets derived from mice fed regular chow (Fig. 4A)orhigh- Non-Tg mice (Fig. 5C); however, the GSIS of Irs25A-b mice fat diet (HFD) (Fig. 4B). Conversely, the mRNA levels of was reduced by ;67%. Although this effect could not be nitric oxide synthase-2 (nos2) were reduced 25% in islets explained by significant changes in total insulin content/ derived from Irs25A-b mice compared with Non-Tg ani- islet (Fig. 5D), a more sensitive immunostaining of individ- mals (Fig. 4C). These findings might be considered as ual b-cells revealed a significant increase in the mean insulin 1884 Effects of IRS2 on b-Cell Function In Vivo Diabetes Volume 66, July 2017

Figure 4—Expression of sod, catalase, and nos2 in islets of Irs2WT-b and Irs25A-b mice. The 7- to 9-week-old male and female mice (Non-Tg, Irs2WT-b,orIrs25A-b)werefedregularchow(A)orHFD(B and C) for 2 weeks. Mice were sacrificed, islets were isolated, and RNA was extracted. mRNA levels of sod1, sod2, catalase, and nos2 were determined by real-time PCR. Data were normalized for the content of actin mRNA levels. Regular chow n =4mice/group,HFDn $ 5 mice/group. *P < 0.05; **P < 0.01. A.U., arbitrary units. staining only in islets derived from Irs2WT-b mice (Fig. 5E). Irs25A-b mice were the most glucose intolerant (Fig. 7A), These results suggest that sustained overexpression of having the largest AUC (Fig. 7B). This was accompanied by a IRS25A in b-cells significantly inhibits GSIS, which may ac- marked reduction (;40%) in insulin content in islets iso- count for the impaired GTT results despite islet hyperplasia. lated from Irs25A-b mice, whereas islets isolated from Irs2WT-b mice showed comparable or even slightly higher Effects of Myc-IRS2 Expression on b-Cell Transcription Factors insulin content (Fig. 7C and D). These results suggest that when placed on an HFD, the overexpression of IRS2 in To better understand the molecular basis for the impaired b GTT results and GSIS in Irs25A-b mice, we examined the -cells, either WT or even more so 5A, impairs the clearance 5A b mRNA levels of a number of b-cell transcription factors. A of blood glucose and reduces insulin content in Irs2 - significant reduction of ;30% in MafA mRNA was observed mice. b in islets isolated from Irs2WT-b, and a much greater reduc- Notable changes in mRNA levels of a number of -cell tion (;60%) was observed in Irs25A-b mice, when compared transcription factors were observed as well. A slight reduc- fi with Non-Tg mice (Fig. 6A). Reduced mRNA and protein tion in MafA mRNA as well as signi cant reductions in Pdx1 levels of Nkx6.1 (;40–50%) were observed only in islets and Nkx6.1 mRNA levels was observed only in islets isolated 5A b isolated from Irs25A-b mice (Fig. 6B–D and Supplementary from IRS2 - mice (Fig. 7E). A reduction at the protein Fig. 4A). A similar ;35% reduction was observed in Glut2 levels of Nkx6.1 was detected as well (Fig. 7F and Supple- mRNA (Fig. 6E), a downstream target of Nkx6.1 (23). No mentary Fig. 4B and C). Similarly, reductions (yet not sig- fi change in Pdx1 mRNA was observed in the Tg animals ni cant) in NeuroD and Ngn3 mRNA levels were observed (Fig. 6A). withnochangesinmRNAlevelsofNkx2.2 and Kir6.2 (Fig. 7G). A marked reduction in mRNA levels of Top2a (24), a Characteristics of Tg Mice Placed on HFD marker of cell proliferation, was observed as well (Fig. 7H). The effects of IRS2 overexpression on animals placed on an HFD (60% kcal fat) were next evaluated. GTT experiments IRS25A Prevents the Induction of the Unfolded Protein revealed that 2 weeks of eating an HFD significantly im- Response paired, as expected, the glucose tolerance of Non-Tg mice. The efficacy of the unfolded protein response (UPR) influ- Further impairment was observed in Irs2WT-b mice, whereas ences the ability of islets to meet increased insulin demand diabetes.diabetesjournals.org Isaac and Associates 1885

Figure 5—GTT, GSIS, and insulin content of islets from Irs2WT-b or Irs25A-b mice maintained on regular chow. The 7- to 8-week-old male mice (Non-Tg, Irs2WT-b,orIrs25A-b) were maintained on regular chow. A GTT was performed on overnight-fasted mice. A: Blood samples were taken at the indicated time points (0–120 min), and glucose levels were determined. B: The AUCs of the GTT graphs were then calculated. Non-Tg n =20;Irs2WT-b n =7;Irs25A-b n =5(*P < 0.05; **P < 0.01 compared with Non-Tg mice). C: GSIS was carried out at low (2.5 mmol/L) and high (22.5 mmol/L) glucose concentrations on pancreatic islets (see RESEARCH DESIGN AND METHODS). The amounts of secreted insulin were normalized relative to the total insulin content. D: Total insulin content per 10 islets was normalized for total protein content. E: Pancreata sections were immunostained for insulin as described in Fig. 2, and the mean insulin intensity per islet was quantified. Data are shown as the mean 6 SEM for Non-Tg mice (n =7),Irs2WT-b mice (n =4),andIrs25A-b mice (n =5).*P < 0.05; **P < 0.01. A.U., arbitrary units.

under metabolic stressors such as chronic hyperglycemia or b-cell transcription factors using islets isolated from young insulin resistance (25). To examine the effects of sustained 10-day-old mice. Indeed, the mRNA levels of Pdx1, MafA, overexpression of IRS2 on the UPR, we measured the and Nkx6.1 were significantly increased twofold to threefold expression of three key players in this process (XBP1, ATF4, in islets isolated from young Irs2WT-b mice, but not Irs25A-b and CHOP). Although no changes were observed in spliced mice (Fig. 8A), suggesting that the transient beneficial ef- Xbp1 mRNA, we could show a significantreductioninmRNA fects of IRS25A, might have taken place at even earlier levels of Atf4 and Chop in islets from Irs25A-b mice that had times. To further corroborate these findings,evenyounger been placed on an HFD, whereas no such effect was ob- mice (5–8 days old) were analyzed. Indeed, when compared served in islets from Irs2WT-b mice (Fig. 7I). These findings with Non-Tg mice, higher b-cell proliferation (Ki67+ cells) suggest that the unaltered and even slightly reduced UPR was evident in Irs25A-b mice (Fig. 8B and Supplementary observed in islets derived from Irs25A-b mice could be a Fig. 5). Significantly a higher number of cyclinD1+ cells was natural downstream response to the reduced insulin pro- also evident both in Irs2WT-b and Irs25A-b mice (Fig. 8C) duction, insulin content, and GSIS in these cells (26,27). combined with the increased expression of PDX1 and NKX6.1 proteins (Fig. 8D and E and Supplementary Fig. 6), b Overexpression of IRS2 in Young Mice Increases -Cell indicating that the beneficial effects of IRS25A are indeed Proliferation and mRNA Levels of b-Cell Transcription evident only in very young animals. Factors Based on the above results, and previous findings (14), our working hypothesis predicted that the overexpression of DISCUSSION IRS2 might initially be beneficial, yet it leads to overuse This study was aimed to challenge in vivo the hypothesis of the b-cells, resulting in b-cell demise, similar to the ef- that selective expression in b-cells of IRS25A interferes with fects observed upon transition from an insulin-resistant negative-feedback control mechanisms along the insulin- state to overt diabetes (28). To further challenge this hy- signaling pathway and improve b-cell function under stress. pothesis, we assayed for the expression of a number of For this purpose, Tg mice that overexpress IRS2, either WT 1886 Effects of IRS2 on b-Cell Function In Vivo Diabetes Volume 66, July 2017

Figure 6—Expression of b-cell transcription factors in islets of Irs2WT-b and Irs25A-b mice. The 7- to 9-week-old male and female mice (Non-Tg, Irs2WT-b,orIrs25A-b) placed on regular chow were sacrificed, and islets were isolated. RNA was extracted, and mRNA levels of MafA, Nkx6.1, Pdx1 (A), and Glut2 (E) were determined by real-time PCR. Data were normalized for the content of Hprt mRNA levels. n $ 5 mice/group. B: Paraffin sections were stained with DAPI (blue) and were immunostained with anti-insulin and anti-Nkx6.1 antibodies, followed by Alexa Fluor 594 (Nkx6.1) or Cy2 (insulin)-conjugated secondary antibodies. C: Histograms represent the percentage of nuclei having a given Nkx6.1 intensity. D: Bar graphs represent the percentage of nuclei stained for Nkx6.1 at an intensity of >4 arbitrary units (A.U.) (see RESEARCH DESIGN + WT 5A AND METHODS). More than 800 insulin b-cells were counted per mouse. Non-Tg mice (n =6),Irs2 -b; Irs2 -b mice (n = 4 mice/group). *P < 0.05; **P < 0.01; ***P < 0.001.

or 5A selectively in b-cells, were generated, using the Tet Irs25A-b mice. These results were in accordance with pre- on/off system (29). However, unexpectedly, the IRS pro- vious findings (4) that reported an increase in the number teins were expressed in a constitutive rather than an in- of islets in Tg mice overexpressing Irs2WT in b-cells. ducible manner. Thus, whereas in a previous study (14) Yet, despite certain beneficial effects, most striking is the IRS2WT and IRS25A expression was induced in adult isolated observation that sustained overexpression of Irs25A in islets by infection with adenovirus, in our in vivo system b-cells in vivo is rather detrimental to islet functionality, the IRS2 proteins, under the control of an insulin promoter, as is evident by the reduced insulin immunostaining per were expressed from an early embryonic stage (18). islet and the impaired GTT results and GSIS of Irs25A-b We have previously shown that IRS25A was much more mice. These deleterious effects were even more pronounced effective than IRS2WT in protecting b-cells from apoptosis in mice placed on an HFD. At the molecular levels, we induced by proinflammatory cytokines (14). Furthermore, observed a significant reduction in mRNA levels of a num- when expressed in isolated adult mouse islets transplanted ber of b-cell transcription factors that are key players in into diabetic mice, IRS25A was more effective than IRS2WT maintaining b-cell function. These included MafA, Nkx6.1, in restoring normoglycemia (14). In the current study, we and Pdx1, which were markedly reduced in Irs25A-b mice, could show that IRS25A functions in vivo in some aspects especially in those placed on an HFD. The loss of MafA is an better than IRS2WT. The expression of IRS25A increased the early indicator of b-cell inactivity, and the subsequent def- mRNA levels of catalase and sod while lowering those of icit of the more impactful Nkx6.1 results in overt dysfunc- nos2. Assuming that the protein levels of these enzymes are tion associated with T2DM (31). Indeed, the reduced altered in the same way, this could reduce the oxidative expression of NKx6.1 could account for the reduced transcrip- burden inflicted upon b-cells and could exert beneficial ef- tion of its downstream target Glut2, observed in Irs25A-b fects on islet functionality and insulin secretion (30). The mice, and could explain the impaired GSIS observed in expression of IRS2 also induced islet hyperplasia, with an these animals because Glut2 is the main glucose transporter increased number of big islets (area .30 3 103 mm2)in in pancreatic islets (32). Reduced expression of some ele- Irs2WT-b miceandanevenbiggerincreaseinisletsfrom ments mediating the UPR, such as ATF4 and CHOP, that diabetes.diabetesjournals.org Isaac and Associates 1887

Figure 7—Effects of HFD on Irs2WT-b and Irs25A-b mice. The 7-week-old male mice (Non-Tg, Irs2WT-b,orIrs25A-b) were maintained on an HFD for 2 weeks. A GTT was performed on overnight-fasted mice. A: Blood samples were taken at the indicated time points (0–120 min), and glucose levels were determined. B: The AUCs of the GTT graphs were then calculated. Data are shown as the mean 6 SEM of Non-Tg mice (n =6)and Irs2WT-b; Irs25A-b mice (n = 3). Pancreata sections were immunostained for insulin (C) as described in Fig. 2, and the mean insulin intensity per islet was quantified (D) from two sections per animal (Non-Tg mice n =5;Irs2WT-b; Irs25A-b mice n =3).E and G–I: The 7- to 9-week-old male and female mice (Non-Tg, Irs2WT-b,andIrs25A-b) placed on an HFD for 2 weeks were sacrificed, and islets were isolated. RNA was extracted, and mRNA levels of the indicated were determined by real-time PCR (Xbp1-S; Spliced Xbp1). Data were normalized for the content of actin mRNA levels. Non-Tg mice n =8;Irs2WT-b; Irs25A-b mice n =5.F:Paraffin-embedded sections were stained with DAPI (blue) and were immunostained with anti-insulin and anti-Nkx6.1 antibodies followed by Alexa Fluor 594 (Nkx6.1) or Cy2 (insulin)-conjugated secondary anti- bodies. Bar graphs represent the percentage of nuclei stained for Nkx6.1 at an intensity of >3 arbitrary units (A.U.) (see RESEARCH DESIGN AND WT 5A METHODS). Non-Tg mice (n = 6 mice/group), Irs2 -b mice (n = 5 mice/group), and Irs2 -b mice (n =4mice/group).*P < 0.05; **P < 0.01. RC, regular chow.

were observed selectively in islets derived from Irs25A-b could stem from subtle differences in the strains of mice mice placed on an HFD could also reflect a downstream being used, because different origins of mice (in this case, response to the reduced insulin production and GSIS in C57BL6 substrains maintained by different vendors) can these cells. Of note, the mRNA levels of other transcription cause major differences in how mice respond (33). Further- factors (e.g., Ngn3, Nkx2.2, and Kir6.2) were unchanged, more, given that after six backcrosses the Tg mice expressed highlighting the complexity and selectivity of the inactiva- ;98% of the genetic background of C57BL/6 mice, the tion of islet-enriched transcription factors (31). remaining 2% difference could be important. The site of Our results are somewhat at variance with earlier findings insertion and copy number of the transgene or epigenetic (4) that described the beneficial effects of selective over- differences of unknown origin could also be contributing expression of Irs2WT in pancreatic b-cells. In these studies, factors. . the overexpression of Irs2 in WT mice (denoted rip13- IRS2) The essential role of IRS2 in maintaining proper b-cell improved pancreatic insulin content by 2.7-fold; however, growth and functionality is very well established (1–4). This glucose homeostasis remained unaltered (4). A possible includes the beneficial effects of overexpressing IRS25A on explanation for the different results in the two studies insulin signaling, GTT results, and GSIS in cultured adult 1888 Effects of IRS2 on b-Cell Function In Vivo Diabetes Volume 66, July 2017

Figure 8—Effects of overexpression of IRS2WT or IRS25A on young mice. A: Islets were isolated from 10-day-old mice. RNA was extracted, and mRNA levels of MafA, Nkx6.1,andPdx1 were determined by real-time PCR. Data were normalized for the content of hprt.Non-Tgmicen = 10; Irs2WT-b mice n =6;Irs25A-b mice n =3.B–E: Paraffin-embedded sections, prepared from islets isolated from 5- to 8-day-old mice were stained with DAPI and were immunostained with anti-insulin, and anti-Ki67, anti-CyclinD1, anti-Nkx6.1, or anti Pdx1 antibodies followed by Alexa Fluor 594 (Ki67/CyclinD1/Nkx6.1/Pdx1) or Cy2 (insulin)-conjugated secondary antibodies. Bar graphs represent the percentage of Ki67+/insulin+ (B) and CyclinD1+/insulin+ (C) cells. Bar graphs represent the percentage of insulin+ nuclei stained for Nkx6.1 (D)orPdx1(E)atanintensityof>3 + WT A.U. (D)or>2A.U.(E)(seeRESEARCH DESIGN AND METHODS). More than 600 insulin b-cells were counted/group. Non-Tg mice n =6;Irs2 -b mice n =4;Irs25A-b mice n =6.*P < 0.05; **P < 0.01; ***P < 0.001. A.U., arbitrary units.

mouse and human pancreatic islets (14). Therefore, the proliferation (Ki67 and cyclinD1) and the protein levels of phenotype of the Irs25A-b mice, which were glucose intol- Nkx6.1andPdx1weresignificantly increased in these young erant and had impaired GSIS, was counter to expectation. mice, which is in accordance with IRS25A being more effica- However, we could rationalize this phenomenon assuming cious than IRS2WT in promoting insulin signaling in b-cells that, indeed, the overexpression of Irs2WT and Irs25A exerts (9). initial beneficial effects on b-cell function and growth, The ability of potential promoters of b-cell function to which are exemplified by the islet hyperplasia observed in curtail b-cell activity was demonstrated in other model sys- the Tg animals as well as by the increased mRNA levels of tems as well. For example, the overexpression in b-cells of a catalase and sod that are expected to improve cellular re- constitutively active form of glucokinase, which was ex- ducing power. However, the sustained expression of Irs2WT pected to promote GSIS, was in fact deleterious to the cells, and even more so of Irs25A might have generated stress resulting in b-cell demise and apoptosis (34). In both model signals in the b-cells due to the overuse of the islet growth systems, b-cells initially responded by enhanced replication. and secretion machineries driven by the overexpressed IRS2 However, at a later stage, b-cell failure caused hyperglyce- proteins. This could lead to b-cell dysfunction by targeting mia and impaired glucose tolerance (Tornovsky-Babeay MafA, Nkx6.1, Pdx1,andGlut2. Support for this hypothesis et al. [34] and the present study). Also, in both models is provided by the fact that, indeed, in young Irs2WT-b mice the key b-cell transcription factors MafA and Nkx6.1, but at 10 days of age the mRNA levels of these transcription not others, were dramatically reduced. These observations factors are still elevated, whereas in Irs25A-b mice their levels strengthen the hypotheses that overexpression for a limited have already declined. The beneficial effects of IRS25A were period of time of proteins expected to be beneficial to b-cell evident only in even younger mice (5–8daysold).b-Cell function, such as an active mutated glucokinase or IRS2, diabetes.diabetesjournals.org Isaac and Associates 1889 indeed promotes b-cell function (14,34). However, allowing 12. Liu YF, Herschkovitz A, Boura-Halfon S, et al. Serine phosphorylation proximal the expression of such proteins for too long induces b-cell to its phosphotyrosine binding domain inhibits insulin receptor substrate 1 function stress, leading to their demise.Thetranslationalimpli- and promotes insulin resistance. Mol Cell Biol 2004;24:9668–9681 cation of this study is the need to identify the proper 13. Herschkovitz A, Liu YF, Ilan E, Ronen D, Boura-Halfon S, Zick Y. Common “ ” fi inhibitory serine sites phosphorylated by IRS-1 kinases, triggered by insulin and time window when the expression of bene cial proteins – b inducers of insulin resistance. J Biol Chem 2007;282:18018 18027 is supportive for human -cells without running the risk 14. Gurevitch D, Boura-Halfon S, Isaac R, et al. Elimination of negative feedback of expression that is too long and might end up being control mechanisms along the insulin signaling pathway improves beta-cell function detrimental. under stress. Diabetes 2010;59:2188–2197 15. Saunders TL. Inducible transgenic mouse models. Methods Mol Biol 2011;693: 103–115 Acknowledgments. The authors thank Dr. Sanford Sampson (Bar Ilan 16. Milo-Landesman D, Surana M, Berkovich I, et al. Correction of hypergly- University, Ramat Gan, Israel) for insightful discussions. cemia in diabetic mice transplanted with reversibly immortalized pancreatic Funding. This work was supported by JDRF (grant 40-2009-720 to Y.Z.) and the beta cells controlled by the tet-on regulatory system. Cell Transplant 2001;10: Israel Science Foundation (grant 759/09 to Y.Z.). 645–650 Duality of Interest. No potential conflicts of interest relevant to this article 17. Jaisser F. Inducible and gene modification in transgenic mice. were reported. J Am Soc Nephrol 2000;11(Suppl. 16):S95–S100 Author Contributions. R.I., Y.V., S.B.-H., L.F., and Z.K. conceived and 18. Gittes GK, Rutter WJ. Onset of cell-specific gene expression in the developing designed the experiments, performed the experiments and statistical analysis, and mouse pancreas. Proc Natl Acad Sci USA 1992;89:1128–1132 contributed to the discussion. S.S. performed the experiments and statistical 19. Juríková M, Danihel L’,PolákS, Varga I. Ki67, PCNA, and MCM proteins: analysis. E.E. conceived and designed the experiments, contributed to the discussion, markers of proliferation in the diagnosis of breast cancer. Acta Histochem 2016;118: and reviewed and edited the manuscript. Y.Z. conceived and designed the 544–552 experiments, contributed to the discussion, and wrote the manuscript. Y.Z. is the 20. Hou N, Torii S, Saito N, Hosaka M, Takeuchi T. Reactive oxygen species- guarantor of this work and, as such, had full access to all the data in the study and mediated pancreatic beta-cell death is regulated by interactions between stress- takes responsibility for the integrity of the data and the accuracy of the data analysis. activated protein kinases, p38 and c-Jun N-terminal kinase, and mitogen-activated Prior Presentation. Parts of this study were presented in abstract form at protein kinase phosphatases. Endocrinology 2008;149:1654–1665 the Keystone Symposium: Diabetes: New Insights into Molecular Mechanisms and 21. Glorieux C, Zamocky M, Sandoval JM, Verrax J, Calderon PB. Regulation of TherapeuticStrategies,Kyoto,Japan,25–29 October 2015. catalase expression in healthy and cancerous cells. Free Radic Biol Med 2015;87: 84–97 References 22. Macmillan-Crow LA, Cruthirds DL. Invited review: manganese superoxide dis- 1. Withers DJ, Gutierrez JS, Towery H, et al. Disruption of IRS-2 causes type mutase in disease. Free Radic Res 2001;34:325–336 2 diabetes in mice. Nature 1998;391:900–904 23. Taylor BL, Liu FF, Sander M. Nkx6.1 is essential for maintaining the functional 2. Lingohr MK, Dickson LM, Wrede CE, et al. Decreasing IRS-2 expression in state of pancreatic beta cells. Cell Rep 2013;4:1262–1275 pancreatic beta-cells (INS-1) promotes apoptosis, which can be compensated for by 24. Lu LY, Kuang H, Korakavi G, Yu X. Topoisomerase II regulates the maintenance introduction of IRS-4 expression. Mol Cell Endocrinol 2003;209:17–31 of DNA methylation. J Biol Chem 2015;290:851–860 3. Lin X, Taguchi A, Park S, et al. Dysregulation of insulin receptor substrate 2 in 25. Cnop M, Foufelle F, Velloso LA. Endoplasmic reticulum stress, obesity and beta cells and brain causes obesity and diabetes. J Clin Invest 2004;114:908–916 diabetes. Trends Mol Med 2012;18:59–68 4. Hennige AM, Burks DJ, Ozcan U, et al. Upregulation of insulin receptor sub- 26. Szabat M, Page MM, Panzhinskiy E, et al. Reduced Insulin Production Relieves strate-2 in pancreatic beta cells prevents diabetes. J Clin Invest 2003;112:1521–1532 Endoplasmic Reticulum Stress and Induces b Cell Proliferation. Cell Metab 2016;23: 5. Eck MJ, Dhe-Paganon S, Trüb T, Nolte RT, Shoelson SE. Structure of the IRS-1 179–193 PTB domain bound to the juxtamembrane region of the insulin receptor. Cell 1996; 27. Back SH, Kaufman RJ. Endoplasmic reticulum stress and type 2 diabetes. Annu 85:695–705 Rev Biochem 2012;81:767–793 6. Craparo A, O’Neill TJ, Gustafson TA. Non-SH2 domains within insulin receptor 28. Saltiel AR. Insulin resistance in the defense against obesity. Cell Metab 2012; substrate-1 and SHC mediate their phosphotyrosine-dependent interaction with the 15:798–804 NPEY motif of the insulin-like growth factor I receptor. J Biol Chem 1995;270: 29. Shockett P, Difilippantonio M, Hellman N, Schatz DG. A modified tetracycline- 15639–15643 regulated system provides autoregulatory, inducible gene expression in cultured cells 7. Khan AH, Pessin JE. Insulin regulation of glucose uptake: a complex interplay of and transgenic mice. Proc Natl Acad Sci USA 1995;92:6522–6526 intracellular signalling pathways. Diabetologia 2002;45:1475–1483 30. Wang J, Yang X, Zhang J. Bridges between mitochondrial oxidative stress, ER 8. ShanikMH,XuY,SkrhaJ,DanknerR,ZickY,RothJ.Insulinresistanceand stressandmTORsignalinginpancreaticb cells. Cell Signal 2016;28:1099–1104 hyperinsulinemia: is hyperinsulinemia the cart or the horse? Diabetes Care 2008;31 31. Guo S, Dai C, Guo M, et al. Inactivation of specific b cell transcription factors in (Suppl. 2):S262–S268 type 2 diabetes. J Clin Invest 2013;123:3305–3316 9. Boura-Halfon S, Zick Y. Phosphorylation of IRS proteins, insulin action and 32. Thorens B. GLUT2, glucose sensing and glucose homeostasis. Diabetologia insulin resistance. Am J Physiol Endocrinol Metab 2009;296:E581–E591 2015;58:221–232 10. Paz K, Hemi R, LeRoith D, et al. A molecular basis for insulin resistance. El- 33. Suzuki R, Tobe K, Terauchi Y, et al. Pdx1 expression in Irs2-deficient mouse evated serine/threonine phosphorylation of IRS-1 and IRS-2 inhibits their binding to beta-cells is regulated in a strain-dependent manner. J Biol Chem 2003;278:43691– the juxtamembrane region of the insulin receptor and impairs their ability to undergo 43698 insulin-induced tyrosine phosphorylation. J Biol Chem 1997;272:29911–29918 34. Tornovsky-Babeay S, Dadon D, Ziv O, et al. Type 2 diabetes and congenital 11. Zick Y. Insulin resistance: a phosphorylation-based uncoupling of insulin sig- hyperinsulinism cause DNA double-strand breaks and p53 activity in b cells. Cell naling. Trends Cell Biol 2001;11:437–441 Metab 2014;19:109–121