Maternal diet-induced microRNAs and mTOR underlie β cell dysfunction in offspring

Abstract

A maternal diet that is low in protein increases the susceptibility of offspring to type 2 diabetes by inducing long-term alterations in β cell mass and function. Nutrients and growth factor signaling converge through mTOR, suggesting that this pathway participates in β cell programming during fetal development. Here, we revealed that newborns of dams exposed to low-protein diet (LP0.5) throughout pregnancy exhibited decreased insulin levels, a lower β cell fraction, and reduced mTOR signaling. Adult offspring of LP0.5-exposed mothers exhibited glucose intolerance as a result of an insulin secretory defect and not β cell mass reduction. The β cell insulin secretory defect was distal to glucose-dependent Ca2+ influx and resulted from reduced proinsulin biosynthesis and insulin content. Islets from offspring of LP0.5-fed dams exhibited reduced mTOR and increased expression of a subset of microRNAs, and blockade of microRNA-199a-3p and -342 in these islets restored mTOR and insulin secretion to normal. Finally, transient β cell activation of mTORC1 signaling in offspring during the last week of pregnancy of mothers fed a LP0.5 rescued the defect in the neonatal β cell fraction and metabolic abnormalities in the adult. Together, these findings indicate that a maternal low-protein diet alters microRNA and mTOR expression in the offspring, influencing insulin secretion and glucose homeostasis.

Figure 9. Transient activation of mTORC1 in β cells during development rescues neonatal β cell fraction, glucose intolerance, and insulin secretion dysfunction in LP0.5 mice.

(A and B) Newborn β cell/pancreas area in Tsc2^fl/fl, RIPCre Tsc2^fl/fl, RIPCre, or RIPCre Rheb Ctrl or LP0.5 mice (n = 3–5). (C and D) β Cell proliferation in Tsc2^fl/fl, RIPCre Tsc2^fl/fl, RIPCre, or RIPCre Rheb Ctrl or LP0.5 mice (n = 3–5). (E) Intraperitoneal glucose test in 12-week old male RIPCre Rheb and wild-type RIPCre mice without Dox (n = 6). (F and G) Intraperitoneal glucose test and glucose-stimulated insulin secretion in RIPCre Rheb (LP0.5) and wild-type Rheb (LP0.5) mice exposed to Dox throughout life (n = 3). (H–M) mTOR protein levels and phosphorylated S6 (Ser240) in islets from Rheb and RIPCre Rheb Ctrl or LP0.5 mice after exposure to Dox throughout life (n = 3). (N–Q) Pdx1 protein levels in islets from Rheb and RIPCre Rheb Ctrl or LP0.5 mice with Dox after birth to death (n = 3–6). *P < 0.05 vs. Ctrl, ^#P < 0.05 vs. LP0.5.

Figure 8. miR-199a-3p and -342 regulate mTOR protein levels and insulin secretion in β cells.

(A and B) mTOR (A) and Pdx1 (B) protein levels in dispersed islet cells treated with or without anti–miR-199a-3p (199). (C) mTOR protein levels in dispersed islet cells treated with or without anti–miR-342. (D and E) Insulin secretion (D) and insulin content (E) in dispersed β cells from 12-week-old male LP0.5 and control islets treated with or without anti–miR-199, -342, or -375. (F and G) Insulin secretion (F) and insulin content normalized to DNA (ng/ml) (G) in β cells from 12-week-old male control islets treated with or without miR-152 and -342 mimics. n = 3–6. *P < 0.05 vs. Ctrl, ^#P < 0.05 vs. LP0.5, **P < 0.05 vs. Ctrl in high glucose (16 mM glucose).

Figure 7. LP0.5 exposure during pregnancy regulates mTOR signaling in adult mice by altering specific miRs.

(A) mTOR transcription message from islets of 12-week-old male LP0.5 and Ctrl mice. (B–D) miRs differentially expressed in islets of 12-week-old male LP0.5 and Ctrl mice. Expression of the indicated miRs was measured by qRT-PCR. Small enriched RNA was purified by using MiRVana. Results are expressed as fold changes and correspond to the mean ± SEM of LP0.5 and Ctrl mice (n = 4). *P < 0.05 vs. Ctrl.

Figure 6. LP0.5 exposure during pregnancy decreases mTOR signaling in neonatal and adult β cells.

(A) Immunofluorescence staining of phosphorylated ribosomal protein S6 (pS6, Ser240, red) and insulin (green) in neonatal islets (original magnification, ×40) from LP0.5 and Ctrl mice. (B–E) mTOR protein levels in 12-week-old islets from LP0.5 and Ctrl mice. A representative Western blot of mTOR in male (B) and female (D) mice. Quantification of mTOR levels normalized to β-actin in male LP0.5 islets (C) and female LP0.5 islets (E). (F) Phosphorylated ribosomal protein S6, ERK1/2, and AKT1/2 levels in 12-week-old islets from LP0.5 and Ctrl mice. (G–I) Quantification of blots in F. D (mTOR and actin) and F (pS6 Ser240 and actin) are from the same experiment. *P < 0.05 vs. Ctrl, n ≥ 4.

Figure 5. LP0.5 exposure during pregnancy alters Ins2 and Pdx1 mRNA and β cell insulin content in adult mice.

(A and B) Insulin content was determined and normalized to total islet DNA in 12-week-old female (A) and male (B) LP0.5 and Ctrl mice. (C and D) Quantitative PCR was performed on isolated mRNA from 12-week-old male LP0.5 and Ctrl islets for Ins1 and Ins2. (E and F) Pdx1 mRNA expression (E) and total protein levels from 12-week-old LP0.5 and Ctrl islets (F). (G) Quantification of F. (H–M) mRNA expression levels of Neurod1, Hnf1a, Hnf4a, Tcf2, Igf2, and Gkc were measured in male 12-week-old LP0.5 and Ctrl islets by qPCR. Transcript levels were normalized to β-actin. n = 4–6 for C–M. *P < 0.05 vs. Ctrl, n ≥ 3.

Figure 4. LP0.5 adult offspring have insulin secretion defect that is distal to calcium signaling.

(A and B) Intraperitoneal glucose tests (A) and insulin tolerance tests (B) in 12-week-old male LP0.5 and Ctrl mice. (C) In vivo glucose-stimulated insulin secretion in 12-week-old male LP0.5 and Ctrl mice. (D) Glucose-stimulated insulin secretion in isolated islets from adult LP0.5 and Ctrl mice in response to 2 mM and 16 mM glucose. Secreted insulin was normalized to total content. (E) Total islet insulin content in D was measured and normalized to total islet DNA. (F) Insulin secretion in response to 30 mM KCl. (G and H) Representative recordings of intracellular calcium are shown in islets isolated from Ctrl or LP0.5 mice stimulated with 8 mM (8G) and 11 mM glucose (11G). (I) Average oscillatory period was determined by fast Fourier transform (FFT) from >20 minutes of recording. Numbers on bars indicate number of islets measured in ≥4 independent experiments from 2 cohorts. (J) Average oscillatory baseline, peak, and amplitude (Δ340/380) are displayed as in I. No significant differences were found. *P < 0.05 vs. Ctrl, **P < 0.05 vs. Ctrl 2 mM glucose, ^#P < 0.05 vs. Ctrl 22 mM glucose.

Figure 3. LP0.5 offspring exhibit abnormal neonatal β cell fraction.

(A) β Cell/pancreas area in newborn LP0.5 and Ctrl mice. (B) β Cell proliferation, as measured by Ki67 staining, in newborn LP0.5 and Ctrl mice. (C) Average number of islets per pancreas in LP0.5 and Ctrl mice at birth. (D) Average β cell number per islet in LP0.5 and Ctrl mice at birth. (E and F) Normalized β cell/pancreas area in 3-month-old male (E) or female (F) LP0.5 and Ctrl mice. (G and H) β Cell mass in 3-month-old male (G) or female (H) LP0.5 and Ctrl mice. *P < 0.05 vs. Ctrl, n ≥ 5.

Figure 2. LP0.5 adult offspring exhibited impaired glucose tolerance and enhanced insulin sensitivity.

(A, B, E, and F) Fed or fasting glucose levels in 4-week-old male (A and B) and female (E and F) LP.05 and Ctrl mice. (C, D, G, and H) Fed or fasting insulin levels in 4-week-old male (C and D) and female (G and H) LP.05 and Ctrl mice. (I–L) Intraperitoneal glucose and insulin tolerance tests were performed in 5- and 6-week-old male (I and K) and female (J and L) LP0.5 and Ctrl mice. *P < 0.05 vs. Ctrl. n = 5 for both male and female unless otherwise noted.

Figure 1. Maternal low-protein diet model and phenotypes of newborn offspring.

(A) Pregnant C57BL/6 mice were exposed to diets, Ctrl (23% protein) or LP0.5 (isocaloric and 9% protein), throughout pregnancy. After delivery, LP0.5 and Ctrl dams and offspring were introduced to Ctrl. Pancreata were collected at birth and on P84. Metabolic studies of male and female LP0.5 and Ctrl offspring were conducted between day 42 and 84. (B) Maternal body weight during the adaptation period (D0–D30), during pregnancy (P.05–P16), and after pregnancy (PP1–PP15). (C) Maternal fed glucose levels of LP0.5 and Ctrl dams on pregnancy day 15. (D and E) Litter size on P1 (D) and P21 (E). (F and G) Newborn body weight (F) and body length (G). (H and I) Newborn blood glucose (H) and insulin (I) levels in LP0.5 and Ctrl offspring. n = 4 dams were used for maternal body weight assessment (B and C). n = 16–27 mice were used for D–F. n = 8 mice were used for G–I. *P < 0.05 vs. Ctrl.