Lasting Hematopoietic Memory Induced By Gestational Diabetes Contributes to Atherosclerosis Development in Adult Offspring

One in six pregnancies is affected by some form of gestational diabetes (GD), a prevalence that is steadily rising in the context of the worldwide epidemics of obesity and early diabetic onset. Early diagnosis and advances in maternal glucose control have greatly mitigated the perinatal consequences of gestational diabetes on the mothers and the offspring. However, these advances have only marginally impacted its long-term consequences. As such, exposure to hyperglycemia in utero remains associated with long-term morbidities in adult offspring, including an increased risk of atherosclerosis and cardiovascular diseases. The mechanisms driving this pathological transmission across generations have not been established. We hypothesized that the hematopoietic system and particularly the hematopoietic stem cells (HSCs) are stably altered by GD and are essential "effectors" of its long-term pathological effects in offspring.

We established two reliable genetic and pharmacological GD mouse models that reproduce the perinatal adverse features of the human pathology. In these models, we showed that gestational diabetes alters the hematopoiesis of the adult offspring even as they do not present any signs of diabetes. At steady state, we observed the skewing of the hematopoietic production toward the myeloid lineage and the expansion of key myeloid progenitors including multipotential progenitor (MPP) and granulocyte-macrophage progenitor (GMP) compartments. We also found marks of activation of the hematopoietic stem cell (HSC) compartment, which were associated with a loss of fitness in regenerative condition. Upon inflammatory challenge, GD offspring displayed an altered hematopoietic stress response that mimics the phenotype observed in mice with overt diabetes. Similar results were obtained in vitro as bone marrow-derived macrophages (BMDM) generated from GD offspring displayed functional characteristics similar to those generated from diabetic mice. Altogether these results suggest that non-diabetic GD offspring maintain, during adulthood, some hematopoietic features associated with diabetes. Importantly, this GD hematopoietic memory persisted in a hematopoietic transplantation setting, therefore indicating that this property is supported by intrinsic HSC alterations. We used atherosclerosis-prone Apoe knock-out mice to assess the impact of GD hematopoietic memory on the health of adult offspring. Consistent with the observations made in human populations, we found that GD favors atherosclerosis development in offspring with increased inflammation and cartilaginous metaplasia in the aortic valves. Importantly, similar results were obtained in a transplantation setting, therefore indicating that the alterations of the hematopoietic system in GD offspring directly contribute to the atherosclerotic phenotype.

Altogether these results establish that the hematopoietic system of GD offspring maintains a long-term functional glycemic memory, even as these animals are not displaying any gross metabolic disruptions. Importantly, these results show that this hematopoietic diabetic memory contributes to the development of pathology in the adult offspring born to diabetic pregnancy.

No relevant conflicts of interest to declare.