Fetal growth is driven by a bimodal fetal beta cell response and reshaped by a ketogenic diet in a mouse model of maternal diabetes

AIMS/HYPOTHESIS: Maternal diabetes confers two opposing risks to fetal growth, resulting in macrosomia in mild cases and intrauterine growth restriction (IUGR) in severe cases. The mechanisms governing these divergent responses are poorly understood, given the intimate regulation of insulin by glucose and insulin's fetal growth-promoting effects. We hypothesised that the degree of maternal hyperglycaemia dictates a bimodal pattern of fetal insulin secretion that determines fetal growth, and that use of a ketogenic diet (KD) as a nutritional intervention could modify this outcome. METHODS: We used the Insulin-rtTA;TET-DTA mouse model to induce preconception diabetes. Dams were stratified based on maternal blood glucose, namely non-diabetes (glucose <9.6 mmol/l), mild diabetes (glucose range 9.6-16.7 mmol/l) or severe diabetes (glucose >16.7 mmol/l), and maintained on either a normal diet or a KD. We assessed fetal growth and plasma C-peptide, performed islet functional assays ex vivo, and characterised changes in plasma metabolites. Fetal pancreases were analysed by immunohistochemistry for beta cell area, proliferation, maturation and mechanistic target of rapamycin complex 1 (mTORC1) activity. RESULTS: Mild maternal diabetes induced fetal macrosomia, driven by beta cell hyperplasia, hyperinsulinaemia and premature beta cell functional maturation, as reflected by glucose-stimulated insulin secretion and upregulated MafA expression. This was associated with strong activation of the mTORC1 pathway. In contrast, severe diabetes caused IUGR associated with reduced beta cell mass and profound functional impairment. The KD had divergent effects: it normalised fetal growth in the mild diabetes group by preventing beta cell proliferation and premature maturation, thereby reducing insulin secretion, but failed to rescue IUGR in the severe diabetes group, despite partially restoring beta cell function. Notably, the KD uncoupled the positive correlation between fetal insulin and body weight, revealing a primary, insulin-independent, growth-restrictive effect. CONCLUSIONS/INTERPRETATION: Fetal growth in a mouse model of diabetes in pregnancy is governed by a bimodal beta cell response to the maternal glycaemic environment, orchestrated at the molecular level by the mTORC1 pathway. A KD can prevent diabetes-derived macrosomia by reducing beta cell stimulation and through insulin-independent mechanisms, but cannot reverse IUGR, warranting further studies of its role in diabetes during pregnancy.