FIGURE 4.

RAPA/MET‐treatment normalized insulin sensitivity and ameliorated the effects of RAPA‐treatment on islet granulation. (a) UNT mice progressed toward hyperinsulinemia. MET‐treatment reduced this hyperinsulinemia (p = 0.01 vs. UNT after 12 weeks of treatment, Tukey–Kramer HSD). RAPA‐ and RAPA/MET‐treatment prevented hyperinsulinemia (p < 0.0001 for both vs. UNT after 12 weeks of treatment, Tukey–Kramer HSD). (b) UNT mice were insulin resistant, indicated by the insulin tolerance test (ITT, showing the percentage of circulating glucose of the pre‐injection basal level). RAPA‐treatment for 17 weeks improved insulin sensitivity (repeated measures MANOVA, p = 0.03 vs. UNT; one RAPA‐treated mouse with glucose values >600 mg/dl at 45 and 60 min after insulin injection was censored as an outlier). MET‐treatment had no effect. RAPA/MET‐treatment normalized insulin sensitivity to a level equivalent to that of untreated C57BL/6J males (repeated measures MANOVA, p = 0.005 vs. UNT). (c) UNT mice developed hyperglycemia as expected (Leiter et al., 2013; Leiter & Reifsnyder, 2004). RAPA‐treatment further elevated this hyperglycemia by 12 weeks of treatment (p = 0.03, Tukey–Kramer HSD). MET‐treatment had no effect. After treatment was initiated, RAPA/MET‐treatment reduced hyperglycemia compared to RAPA‐treatment alone (repeated measures MANOVA, 16–28 weeks of age, interaction of treatment with time, p = 0.03). Glucose levels in RAPA/MET‐treated mice were comparable to levels in UNT mice after 16 weeks of treatment. (d) RAPA/MET‐treatment prevented RAPA‐ driven elevation of HbA1c, measured 18 weeks after treatment was initiated. (e) All three treatments reduced hepatic Ppargc1a expression. Reduced hepatic Gck and increased G6pc expression in RAPA‐treated mice are associated with elevated hyperglycemia. RAPA/MET‐ treatment counteracts this hyperglycemic effect potentially by preventing the rapamycin‐associated elevation of G6pc. (f) All islets from 3 separate sections (per pancreas) were scored for granulation status (insulin storage): F = fully‐granulated; P = partially‐degranulated; or M = mostly‐to‐completely‐degranulated. Treatment affected the islet degranulation profile (p = 0.0006, MANOVA, Wilk's λ). RAPA‐treatment significantly exacerbated the phenotype (post hoc pairwise comparisons, p = 0.02, MANOVA, Wilk's λ). MET‐treatment had no effect versus UNT. RAPA/MET‐treatment prevented the effect of RAPA‐treatment on islet morphology (p = 0.03, MANOVA, Wilk's λ). (g) RAPA‐ and RAPA/MET‐treatments reduced pancreatic insulin content. (h) Representative islets illustrate effects of treatment on islet granulation (aldehyde fuchsin staining). The islet from an UNT mouse illustrates partial degranulation. The islet from a RAPA‐treated mouse illustrates the mostly degranulated condition of the majority of islets in RAPA‐treated mice. Islets from the MET‐ and RAPA/MET‐treated mice illustrate the partial degranulation typical of islets from these mice. All histologic pictures are at the same magnification. Mean ± SE. N = 9–10 mice per treatment except (b), N = 5–6 mice per treatment. Significance annotations and treatment abbreviations as in Figure 1