Metabolic Programming of the Fetal Liver Hematopoietic Stem and Progenitor Cell Pool

Evidence in several organ systems demonstrates that pregnancy presents a window of vulnerability for establishing a foundation for health or chronic disease. Overnutrition and the complex metabolic changes that can accompany it can result in permanent phenotypic changes and a predisposition to metabolic syndrome, inflammatory or immune-mediated diseases. We previously reported that prenatal overnutrition stunted fetal liver size. Herein, we hypothesize that this might perturb hematopoietic stem and progenitor cell (HSPC) expansion.

To test the effects of a high-fat diet (HFD) and maternal obesity on offspring hematopoiesis, we used a mouse model of diet-induced obesity, feeding female mice a HFD or control diet starting at 5-7 weeks of age and keeping them on the respective diet during subsequent breeding and pregnancy. We then studied offspring at gestational day 14.5 by immunophenotyping, gene expression analysis, qRT-PCR, and transplantation. Fetal livers from HFD offspring had 51% fewer c-Kit⁺ Sca-1⁺ Lin^lo/- and 27% fewer AA4.1⁺ Sca-1⁺ Lin^lo/- (ASL) hematopoietic stem and progenitor cells (HSPC) than controls. This restriction in HSPC numbers was not due to apoptosis or increased reactive oxygen species, as tested by flow cytometry.

To determine whether there might be an increase in hematopoietic differentiation to account for relative HSPC deficiencies in HFD livers, we examined hematopoietic lineage subsets. HFD fetal livers had a relative increase in myeloid (Gr-1⁺/Ter119⁺) and B220⁺ lymphoid cells, with comparable proportion of CD3⁺ cells to controls. Taken together, these results suggest that chronic HFD fetal programming skews fetal liver HSPCs toward differentiation.

When we examined global gene expression of male HFD fetal livers versus controls by RNA-seq, we found differential expression of 125 genes. Among the upregulated transcripts, several were involved in hematopoietic regulation, stress response, and HSPC migration. We then used qRT-PCR to test for expression of several of these genes, along with genes critically involved in fetal HSPC self-renewal, within an HSPC-enriched (Sca-1⁺) population of chronic HFD fetal liver cells. As in RNA-seq, Matrixmetalloproteinase-8 and 9 (Mmp8, Mmp9), which are involved in cell mobilization, were upregulated in HFD-programmed cells. Early growth response-1 (Egr-1) was downregulated as well, further suggesting premature migration of HSPCs from HFD fetal liver. Hmga2, which is implicated in fetal stem cell self-renewal, and its direct target, Igf2bp2, were significantly downregulated in chronic HFD Sca-1⁺ cells. Along with the immunophenotyping data, these findings suggest that maternal obesity and HFD bias HSPCs toward differentiation, at the expense of self-renewal. To dissect the direct metabolic impact, we studied fetal livers from timed pregnancy cohorts after acute HFD exposure or diet reversal in obese dams, which partially ameliorated several molecular and immunophenotypic endpoints.

Finally, we performed a functional test of chronic HFD fetal liver cells by transplantation. A non-competitive transplant into irradiated male recipients yielded no difference in chimerism between HFD or control fetal liver-engrafted animals. Next, we preconditioned a cohort of female and male animals on HFD (or control diet) for 11 weeks, irradiated them, and then competitively transplanted them with a 1:1 ratio of HFD and control fetal liver cells. HFD-programmed fetal liver HSPCs engrafted HFD-conditioned male recipients at significantly lower rates than in HFD-conditioned female or control recipients of either sex. HFD-programmed donor cells retained the significant bias toward the myeloid (Gr-1⁺/Mac-1⁺) lineage, noted in the primary graft cells, and away from the B220⁺ B cell lineage in HFD-conditioned males.

In aggregate, prenatal HFD and maternal obesity suppress self-renewal in favor of HSPC differentiation during a time of critical developmental expansion. This suggests an HSPC defect that appears at least partly specified by the stem cell microenvironment. Our work is the first to demonstrate metabolic vulnerability of the hematopoietic stem and progenitor cell compartment and establishes the hematopoietic system as a target for in utero developmental programming.