Exposure to maternal obesity programs sex differences in pancreatic islets of the offspring in mice

This research was regulated under the UK Home Office Animals (Scientific Procedures) Act 1986 following ethical review by the University of Cambridge Animal Welfare and Ethical Review Board. The model has been described in detail previously [14]. Briefly, female C57BL/6J mice were randomised to either ad libitum or a standard chow RM1 diet (∼7% simple sugars, 3% fat, 50% polysaccharide, 15% protein [wt/wt], 10.74 kJ/g) or a highly palatable energy-rich obesogenic diet (∼10% simple sugars, 20% animal lard, 28% polysaccharide, 23% protein [wt/wt], 28.43 kJ/g) and sweetened condensed milk (Nestle) (∼16% fat, 33% simple sugars, 15% protein, 13.7 kJ/g) fortified with mineral and vitamin mix (AIN93G), for 6 weeks before mating for first pregnancy. Diets were purchased from Special Dietary Services (Witham, UK). Mice were purchased from Charles River Laboratories (Wilmington, MA, USA). The first litter was culled post weaning. This first pregnancy ensured mice were proven breeders. Mice were then re-mated for a second pregnancy. Dams were maintained on their respective diets throughout both pregnancies and lactation periods. Dams fed an obesogenic diet have around 40% body fat on day one of pregnancy and are hyperinsulinaemic and have impaired glucose tolerance in late gestation [15]. Litters were standardised to six pups on postnatal day two. Offspring from ‘Control’ and ‘Obese’ groups were weaned onto RM1 on postnatal day 21 and remained on this diet until the end of the study. Mice were housed in ventilated cages, 2–3 mice per cage with same-sex littermates, and maintained in a humidity-controlled room with a 12 h light/dark cycle, with free access to food and water. An IPGTT was performed on 7-week-old offspring and the offspring were euthanised at 8 weeks of age. Blood glucose was measured between 07:00 and 08:00 hours on the day of euthanasia from tail blood sample using a blood glucose analyser (AlphaTRAK, Abbot Logistics, the Netherlands). Mice were euthanised by carbon dioxide inhalation. Blood was taken by cardiac puncture for serum insulin analysis. Insulin concentration was determined by ELISA (Mercodia, Sweden).

Following a 4 h fast, glucose (1 g/kg) was injected into the intraperitoneal cavity. Blood glucose measurements were made using a blood glucose analyser (AlphaTRAK) at 0, 15, 30, 45, 60, 90 and 120 min. Tail blood was also collected at 0, 15 and 30 min into glass micro-haematocrit capillary tubes with sodium heparin (Hirschmann-Laborgeräte, Germany) for plasma insulin analysis. The AUC for glucose was calculated using the trapezoidal rule.

Islets were isolated as described in [16]. Islets were hand-picked under a stereo microscope and incubated at 37°C overnight in RPMI-1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin sulphate.

Groups of 20 islets were incubated in KRB [16] and 2.8 mmol/l glucose for 1 h at 37°C. Next, these groups of islets were incubated at 37°C for 1 h in KRB containing 2.8 mmol/l glucose, 16.7 mmol/l glucose or 10 mmol/l leucine/glutamine.

Islets were washed in PBS mixed with ethanol/hydrochloric acid and sonicated at 4°C. After centrifugation, the supernatant was stored at −20°C until assayed.

Oxygen consumption was measured by the XF24 Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, CA, USA). Groups of 50 islets (in triplicate) were pre-incubated in KRB with 2 mg/ml BSA and 2.8 mmol/l glucose for 30 min at 37°C. Respiration was measured in the presence of 16.7 mmol/l glucose, oligomycin, carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and rotenone as previously described [17]. Calculations of basal, glucose-stimulated, ATP-linked and proton leak-linked respiration and coupling efficiency were carried out as previously described [18].

This was performed as previously described [16]. Islets were lysed and the content of ATP at 2.8 or 16.7 mmol/l glucose was assayed using a luciferase-based luminescent assay (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions.

RNA was extracted (Qiagen, Manchester, UK) and reverse transcribed into cDNA (Fermentas, Waltham, MA, USA). Quantitative RT-PCR (qRT-PCR) was performed using predesigned TaqMan Gene Expression assays or the SYBR Green system (Thermo Fisher Scientific, Carlsbad, CA, USA). mRNA expression was determined by the ΔΔC_t method and normalised to Rplp0 and Hprt (females) and Actb and Hprt (males), expression of which was not influenced by maternal diet. Primer sequences are available in electronic supplementary material (ESM) Table 1.

Islets were lysed in RIPA buffer with protease inhibitor cocktail (Sigma, Gillingham, UK), subjected to SDS-PAGE and blotted using antibodies (ESM Table 2). Male and female samples were run on separate gels. Membranes were sectioned and some sections stripped (Restore Western Blot Stripping Buffer, Thermo Fischer Scientific) and re-probed to maximise the amount of data obtained from each western blot. Image Lab software version 5.2.1 (Bio-Rad, Hercules, CA, USA) was used to quantify the density of specific bands. Images of western blots showing statistically significant differences are included in ESM Fig. 1. All blots were made available to Diabetologia at the point of submission.

DNA was extracted (Qiagen) and quantified (Quant-iT PicoGreen dsDNA Reagent, Thermo Fisher Scientific). Equal quantities were amplified by qRT-PCR using primers for nuclear (Rplp0) and mitochondrial (mt-Nd5) DNA. Mitochondrial (mtDNA) content was determined as described in [19].

Samples were prepared as previously described [16]. Electron micrographs of at least 40 different beta cells from 2–4 mice/sex were taken for each maternal diet group. Granules (large dense-core vesicles) were defined as docked when the centre of the granule was located within 190 nm (a half-length of the mean granule diameter) from the plasma membrane. The distance between the centre of the granule and the plasma membrane was calculated using an in-house software programmed in MatLab 7 (MathWorks, Natick, MA, USA) [20]. Granule density was estimated according to [21] and was normalised to granule diameter assuming spherical geometry [20]. Mitochondrial lesion in beta cells was determined by Photoshop Elements software (Adobe Systems, San Jose, CA, USA). The lesion area was outlined and the size was measured by ImageJ (National Institutes of Health, Bethesda, MD, USA) and expressed as the area density.

Paraffin-embedded pancreas was exhaustively sectioned (5 μm thickness, 200 μm apart). Sections were incubated with primary antibody against insulin [Dako (Agilent Technologies] and glucagon (Bioss Antibodies, Woburn, MA, USA). The appropriate fluorescent-dye-conjugated secondary antibodies were used for identifying beta and alpha cells (Jackson ImmunoResearch, Ely, UK). Digital images were obtained using Axioscan Z1 Slide Scanner (Zeiss, Germany) and HALO image analysis platform (Indica Labs, Albuquerque, NM, USA) was used to calculate immune-positive and total tissue area. Beta and alpha cell mass was calculated as described in [22].

Investigators were not blind to group assignment and outcome assessment except for analyses of electron micrographs. Calculations were performed in IBM SPSS Statistics 23 (Armonk, NY, USA). ‘n’ represents the number of mice from separate litters. Data are presented as mean ± SEM of the indicated number of litters. mRNA expression, protein abundance, mtDNA content and AUC_glucose were analysed using unpaired Student’s t test. Blood glucose and insulin levels during IPGTT were measured by two-way (repeated measures) ANOVA followed by Bonferroni’s multiple comparisons test (note: each sex was analysed separately). All other data were analysed by two-way ANOVA. A significant interaction indicated sex differences in the impact of maternal obesity. In this case, a simple main effect analysis was performed to isolate the effects of maternal diet on each sex separately. A probability level of 5% (p < 0.05) was taken to be significant.

Body weight, (non-fasted) blood glucose and serum insulin levels were comparable in 8-week-old offspring between maternal diet groups (ESM Table 3). Male offspring, however, were heavier (p < 0.05) and had higher serum insulin (p < 0.01) (ESM Table 3).

Glucose tolerance in both sexes, as reflected by AUC, was also similar irrespective of maternal diet (Fig. 1a, b). We observed, however, that blood glucose was lower (p < 0.001) and plasma insulin higher (p < 0.05) in female offspring of obese dams 15 min after glucose administration compared with controls (Fig. 1b, c).

The amount of insulin secreted by beta cells depends on the mass and function of these cells. Beta cell mass was comparable between maternal diet groups (Fig. 2a). However, in line with having higher serum insulin, male offspring also had increased beta cell mass (p < 0.05). This was due to higher absolute pancreas weight (p < 0.05) and consequently greater total pancreatic tissue area (p < 0.05) with no difference in the relative number of insulin⁺ cells (ESM Fig. 2a–c). In contrast, alpha cell mass was comparable between sexes (Fig. 2b) owing to a lower number of glucagon⁺ cells (p < 0.01) in male compared with female offspring (ESM Fig. 2d).

We found differences in beta cell function between male and female offspring of obese dams. Basal (p < 0.05), GSIS (p < 0.01) and amino acid-stimulated (p < 0.05) insulin secretion were increased only in female offspring of obese dams (Fig. 2c). Islet insulin content was not different between groups (Fig. 2d).

To identify cellular components that may be contributing to enhanced GSIS in female offspring, we quantified GLUT2 protein abundance, the primary glucose transporter in rodent islets [23]. This was not altered by exposure to maternal obesity (Fig. 3a). Subsequent phosphorylation of glucose by glucokinase is the key step controlling glycolytic flux [24]. Glucokinase abundance was reduced in both male (p < 0.05) and female (p < 0.01) offspring of obese dams (Fig. 3b). This finding was unexpected given that the rate of glycolysis is an important determinant of GSIS from beta cells.

Mitochondrial metabolism couples glucose metabolism to insulin secretion. Since leucine/glutamine are mitochondrial fuels, this suggested that pathways downstream of glucokinase may be activated in these islets. We found sex differences in mitochondrial function in offspring of obese dams. Whilst exposure to maternal obesity resulted in higher basal respiration irrespective of offspring sex (p < 0.01) (Fig. 3c–e), glucose-stimulated (relative to basal) respiration was reduced in males (p < 0.05) but increased in female (p < 0.01) offspring of obese dams (Fig. 3c, d, f).

Respiration rate is mostly composed of ATP-linked and proton leak-linked respiration [25]. Similar to glucose-stimulated respiration, ATP-linked respiration was reduced in male offspring of obese dams (p < 0.05) (Fig. 3g). Moreover, coupling efficiency, which estimates the fraction of respiration used to drive ATP synthesis was also reduced in these islets (p < 0.01) (Fig. 3h). In contrast, ATP-linked respiration was increased in female offspring of obese dams (p < 0.01) (Fig. 3g). Finally, we also found that proton leak-linked respiration was higher in islets from maternal obesity-exposed offspring (p < 0.05) (Fig. 3i). This appears to be driven by a relatively large difference between female offspring groups due to lower proton leak-linked respiration in the control group.

Mitochondrial respiration/function is dependent on mtDNA, genes of the respiratory chain complexes that it encodes and its nuclear-encoded constituents. mRNA expression of some mitochondrial-encoded components of the respiratory complexes was significantly increased in maternal obesity-exposed female offspring (Fig. 4a). This was not due to higher levels of mitochondrial transcription factor A (TFAM) (Fig. 4b) but may be partly due to increased number of mitochondria as evidenced by higher mitochondrial area density (p < 0.05) (Fig. 4c). There was, however, no statistical difference in mtDNA content (p < 0.06; Fig. 4d). In addition to mitochondrial genes, mRNA expression of Sdha (Complex II) (p < 0.05) (Fig. 4a) and protein abundance of succinate dehydrogenase B (SDHB; Complex II) (p < 0.01), ubiquinol-cytochrome C reductase core protein 2 (UQCRC2; Complex III) (p < 0.05) and ATP5A (Complex V) (p < 0.01), which are nuclear-encoded, were also increased (Fig. 4e).

Our results suggest that higher mitochondrial respiration in female offspring of obese dams could be due to having more mitochondria and higher levels of respiratory complex proteins. The converse is not the case for male offspring (ESM Fig. 3a–d and Fig. 4c).

Mitochondria are a major source of intracellular ROS [26]. Basal ROS levels were comparable between offspring groups (Fig. 5a). In response to glucose stimulation, ROS levels were higher in islets of maternal obesity-exposed offspring (p < 0.05) (Fig. 5b). Thus, in the case of male offspring of obese dams, ROS levels are elevated even with reduced mitochondrial respiration. Furthermore, female but not male offspring of obese dams displayed increased Sod2 (p < 0.01) (but not Cat) mRNA expression (Fig. 5c, d) suggesting that islets from these male offspring may be more vulnerable to oxidative stress.

Activation of L-type voltage-gated Ca²⁺ (Ca_V1) channels are essential for insulin secretion [27, 28]. Exposure to maternal obesity resulted in sex-specific differences in expression of the two subtypes of Ca_V1 channels: Cacna1c and Cacna1d (which encode Ca_V1.2 and Ca_V1.3, respectively). Cacna1d expression was increased in females (p < 0.05) but both Cacna1c and Cacna1d expression was decreased in male (p < 0.001) offspring of obese dams (Fig. 6a, b). Downregulation of these genes has been observed in both human and rodent type 2 diabetic islets [29‐31].

Docking of secretory vesicles is required for insulin exocytosis [20, 32]. Insulin granule density was reduced in beta cells of maternal obesity-exposed offspring of both sexes (p < 0.01) (Fig. 6c). We found, however, that the area of docked granules was decreased only in male offspring of obese dams (p < 0.001) (Fig. 6d). Furthermore, the number of granules in the first fraction beneath the plasma membrane (0.19 μm from the plasma membrane) was reduced (p < 0.01) whilst those in the second fraction (0.38 μm from the plasma membrane) were increased (p < 0.001) in these mice (Fig. 6e). These results suggest that granules may be stacking in the second fraction (which is considered to be the reserve pool) owing to dysfunctional docking in beta cells of these mice. Expression of syntaxin-1A (STX1A) and syntaxin binding protein 1 (STXBP1) proteins, which are required for docking [33] was unchanged in both male and female offspring (Fig. 6g, h).

17β-oestradiol has been shown to protect females from beta cell death and hyperglycaemia via oestrogen receptor α (ERα) [34‐36]. In our model, ERα protein abundance was higher in female offspring of obese dams (p < 0.05) (Fig. 7a). Furthermore, expression of cleaved (activated) caspase-3, the key mediator of the apoptotic cascade in mammalian cells [37] was decreased (p < 0.05) in these offspring (Fig. 7b). In contrast, cleaved caspase-3 abundance and Casp3 mRNA expression were increased in male offspring of obese dams (p < 0.05) (Fig. 7b, c).

The mitochondrial phenotype in this study prompted us to investigate whether the mitochondrial intrinsic apoptosis pathway could be contributing to changes in caspase-3 activity. mRNA expression of Casp9, the initiator caspase involved in this pathway, was higher in male offspring of obese dams (p < 0.05) (Fig. 7d). Furthermore, the ratio of pro-apoptotic BAX to anti-apoptotic Bcl-2, which can profoundly influence the ability of a cell to respond to an apoptotic signal [38], was lower in female offspring of obese dams (p < 0.05) (Fig. 7e). This suggests a downregulation of the mitochondrial apoptosis pathway in these offspring. There were no differences in protein levels of BAX and Bcl-2 between offspring of maternal diet groups (ESM Fig. 4a, b).