Sex-specific effects of maternal metformin intervention during glucose-intolerant obese pregnancy on body composition and metabolic health in aged mouse offspring

Abstract

Aims/hypothesis: Metformin is increasingly used to treat gestational diabetes (GDM) and pregnancies complicated by pregestational type 2 diabetes or polycystic ovary syndrome but data regarding long-term offspring outcome are lacking in both human studies and animal models. Using a mouse model, this study investigated the effects of maternal metformin intervention during obese glucose-intolerant pregnancy on adiposity, hepatic steatosis and markers of metabolic health of male and female offspring up to the age of 12 months.

Methods: C57BL/6J female mice were weaned onto either a control diet (Con) or, to induce pre-conception obesity, an obesogenic diet (Ob). The respective diets were maintained throughout pregnancy and lactation. These obese dams were then randomised to the untreated group or to receive 300 mg/kg oral metformin hydrochloride treatment (Ob-Met) daily during pregnancy. In male and female offspring, body weights and body composition were measured from 1 month until 12 months of age, when serum and tissues were collected for investigation of adipocyte cellularity (histology), adipose tissue inflammation (histology and quantitative RT-PCR), and hepatic steatosis and fibrosis (histochemistry and modified Folch assay).

Results: At 12 months of age, male Ob and Ob-Met offspring showed increased adiposity, adipocyte hypertrophy, elevated expression of proinflammatory genes, hyperleptinaemia and hepatic lipid accumulation compared with Con offspring. Male Ob-Met offspring failed to show hyperplasia between 8 weeks and 12 months, indicative of restricted adipose tissue expansion, resulting in increased immune cell infiltration and ectopic lipid deposition. Female Ob offspring were relatively protected from these phenotypes but Ob-Met female offspring showed increased adiposity, adipose tissue inflammation, hepatic lipid accumulation, hyperleptinaemia and hyperinsulinaemia compared with Con female offspring.

Conclusions/interpretation: Maternal metformin treatment of obese dams increased offspring metabolic risk factors in a sex- and age-dependent manner. These observations highlight the importance of following up offspring of both sexes beyond early adulthood after interventions during pregnancy. Our findings illustrate the complexity of balancing short-term benefits to mother and child vs any potential long-term metabolic effects on the offspring when prescribing therapeutic agents that cross the placenta.

Keywords: Developmental programming; Fatty liver; Gestational diabetes; Inflammation; Maternal obesity; Metformin; White adipose tissue.

Fig. 1

Schematic timeline representing the development of glucose intolerance in our mouse model of diet-induced obese pregnancy. Boxes refer to periods when mice were fed the high-fat high-sugar diet. The experimental pregnancy used for this study is Pregnancy 2. Reference numbers indicate the publications corresponding to the data. E16/18, embryonic day 16/18

Fig. 2

Mouse offspring body composition. (a–d) Fat mass (a, b) and lean mass (c, d) of male (a, c) and female (b, d) offspring until 6 months of age by TD-NMR. Numbers are n=9–12 (male Con), n=9–12 (male Ob), n=9–11 (male Ob-Met), n=11 or 12 (female Con), n=10–12 (female Ob) and n=11 or 12 (female Ob-Met) independent litters per group. (e, f) Body weight of male (e) and female (f) offspring until 12 months of age (n=12 independent litters per group; data are littermate means). *p<0.05 and **p<0.01, Con vs Ob; ^†p<0.05, ^††p<0.01 and ^†††p<0.001, Con vs Ob-Met; ^‡p<0.05 and ^‡‡p<0.01 Ob vs Ob-Met (two-way ANOVA with Tukey’s multiple comparison test). (g–j) total weight of WAT depots collected at 12 months of age in male (g, h) and female (i, j) offspring in absolute terms and relative to offspring body weight (n=10–12 independent litters per group except for n=8 for male Ob group). An outlier was excluded from (g) male Ob-Met (ROUT method, outlier excluded value 4.71 g). *p<0.05, **p<0.01 and ***p<0.001 (one-way ANOVA with Tukey’s multiple comparison test). Black circles, Con (offspring of control-fed dams); pink squares, Ob (offspring of obese dams); blue triangles, Ob-Met (offspring of obese metformin-treated dams); closed symbols, male offspring; open symbols, female offspring. BW, body weight

Fig. 3

Immune cell infiltration and inflammation in 12-month-old mouse gonadal adipose tissue. (a–h) Immune cell infiltration into gWAT of 12-month-old male (a–c) and female (d–f) offspring as assessed by the presence of CLS as seen on the representative images for male (g) and female (h) offspring (n=9–12 independent litters per group). Scale bar, 100 μm. Area density of CLS in WAT (a, d), CLS size (b, e) and percentage of adipocytes surrounded by CLS (c, f) are shown. (i, j) mRNA expression of macrophage markers and proinflammatory genes in gonadal WAT of 12-month-old male (i) and female (j) offspring relative to the expression of Ppia (fold change using the comparative C_t method) and expressed relative to Con offspring (n=11 or 12 independent litters except for n=8 male Ob offspring). Outliers were excluded as follows: from (i) Adgre1, Cd1c and Ccl2 for male Con (Grubb’s method, outlier excluded values 2.178-, 7.088- and 2.391-fold, respectively); from (j) Ccl2 and Tnf for female Ob (Grubb’s method, outlier excluded values 4.790- and 2.440-fold, respectively); and from (j) Itgax for one female Con and one female Ob (Grubb’s method, outlier excluded values 4.059- and 3.671-fold, respectively). *p<0.05 and **p<0.01; ^¶0.05<p<0.10 (one-way ANOVA with Tukey’s multiple comparison test). Circles, Con (offspring of control-fed dams); squares, Ob (offspring of obese dams); triangles, Ob-Met (offspring of obese metformin-treated dams); closed symbols, male offspring; open symbols, female offspring

Fig. 4

Cellularity of 12-month-old mouse offspring gWAT. (a, d) Cumulative frequency distribution of gonadal adipocyte size in male (a) and female (d) offspring shown as a curve reflecting the percentage of cells that falls below a particular adipocyte size, with representative images of H&E-stained sections for male (c) and female (f) offspring (n=11 or 12 independent litters per group). Scale bar, 100 μm. *p<0.05 for Con vs Ob; ^†p<0.05 for Con vs Ob-Met; ^¶0.05<p<0.10 (two-way ANOVA with Tukey’s multiple comparison test). (b, e) Estimated adipocyte number in the collected gWAT depot for male (b) and female (e) offspring. An outlier was excluded in (b) from male Ob-Met (Grubb’s method, outlier excluded value 3.27×10⁶ cells). *p<0.05 (one-way ANOVA with Tukey’s multiple comparison test). Black circles, Con (offspring of control-fed dams); pink squares, Ob (offspring of obese dams); blue triangles, Ob-Met (offspring of obese metformin-treated dams); closed symbols, male offspring; open symbols, female offspring

Fig. 5

Hepatic lipid accumulation in 12-month-old mouse offspring. (a, b, e, f) Fasted liver weight (a, e) and corresponding total hepatic lipid content as assessed by Folch assay (b, f) in male (a, b) and female (e, f) offspring (n=11 or 12 independent litters per group except for n=7 or 8 male Ob offspring). (c, d, g, h) Area density of lipid droplets in H&E-stained liver tissue (c, g) with representative images (d, h) of male (c, d) and female (g, h) offspring (n=11 or 12 independent litters per group). Scale bar, 500 μm. Pie charts show percentages of pathological scoring for hepatic steatosis: white, G0 (grade 0, absent); blue, G1 (grade 1, mild); light red, G2 (grade 2, moderate); dark red, G3 (grade 3, severe). Outliers were excluded from (a) for male Con offspring (ROUT method, 2.04 g and 1.08 g) and male Ob offspring (ROUT method, 3.9 g). An outlier was excluded from (g) for female Ob offspring (ROUT method, outlier excluded value 5.03%). *p<0.05 or **p<0.01; ^¶0.05<p<0.10 (one-way ANOVA with Tukey’s multiple comparison test). Closed symbols, male offspring; open symbols, female offspring