Gestational diabetes in mice induces hematopoietic memory that affects the long-term health of the offspring

Abstract

Gestational diabetes is a common medical complication of pregnancy that is associated with adverse perinatal outcomes and an increased risk of metabolic diseases and atherosclerosis in adult offspring. The mechanisms responsible for this delayed pathological transmission remain unknown. In mouse models, we found that the development of atherosclerosis in adult offspring born to diabetic pregnancy can be in part linked to hematopoietic alterations. Although they do not show any gross metabolic disruptions, the adult offspring maintain hematopoietic features associated with diabetes, indicating the acquisition of a lasting diabetic hematopoietic memory. We show that the induction of this hematopoietic memory during gestation relies on the activity of the advanced glycation end product receptor (AGER) and the nucleotide binding and oligomerization domain-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome, which lead to increased placental inflammation. In adult offspring, we find that this memory is associated with DNA methyltransferase 1 (DNMT1) upregulation and epigenetic changes in hematopoietic progenitors. Together, our results demonstrate that the hematopoietic system can acquire a lasting memory of gestational diabetes and that this memory constitutes a pathway connecting gestational health to adult pathologies.

Keywords: Cardiovascular disease; Diabetes; Hematology; Hematopoietic stem cells; Inflammation.

Figure 1. Modeling diabetes during pregnancy.

(A and B) Breeding and treatment schematic used to generate genetic (A) and pharmacological (B) mouse models of GD. (C and D) Dam nonfasting glycemia before, during, and after pregnancy in genetic (C) and pharmacological (D) mouse models (C: n = 17–20/group, D: n = 6–20/group, n = 6–8/group at postweaning age). (E and F) Weaning-age (4 weeks old) and adult-age (8 weeks old) body weight (left panels) and nonfasting glycemia (right panels) of offspring born to diabetic pregnancy in the genetic (E) and pharmacological (F) mouse models (n = 5–8/group). Data are represented as means ± SD. Two-way ANOVA with Šidák’s post hoc test. ***P ≤ 0.0005; ****P ≤ 0.0001.

Figure 2. GD promotes atherosclerosis development in adult offspring.

(A) Experimental schematic used to assess atherosclerosis development in Apoe^–/– GD offspring. (B) STZ-treated Apoe^–/– dam nonfasting glycemia before and during pregnancy (n = 3). (C) Adult body weight (left panel) and nonfasting glycemia (right panel) of Apoe^–/– offspring born to diabetic pregnancy (n = 4–7). (D) Histological criteria and atherosclerosis severity score in adult Ctrl and GD Apoe^–/– offspring (n = 7–5/group). (E) Example of histological analysis of H&E-stained aortic valve from Apoe^–/– offspring born to diabetic pregnancy compared with Ctrl. Original magnification, ×100; insets, ×200. (F) Experimental schematic used to assess atherosclerosis in Apoe^–/– recipients transplanted with BM isolated from adult Ctrl and GD Apoe^–/– offspring. (G) Adult body weight (left panel) and nonfasting glycemia (right panel) of Apoe^–/– recipient mice transplanted with BM cells isolated from Ctrl or GD offspring (n = 5). (H) Atherosclerosis severity in Apoe^–/– recipient mice (n = 11–12/group). (I) Example of histological analysis of aortic valve from Apoe^–/– recipient mice. Original magnification, ×100. Data are represented as means ± SD (B, C, and G). Two-way ANOVA with Šidák’s post hoc test (B); χ² test for trend (D and H). *P ≤ 0.05; **P ≤ 0.01.

Figure 3. Offspring born to diabetic pregnancy display altered steady-state hematopoiesis.

(A) Schematic of the murine hematopoietic hierarchy. (B) BM cellularity and absolute numbers of BM myeloid/lymphoid cells in adult WT^Akita offspring (n = 12). (C and D) Absolute numbers of HSPC populations in the BM of adult Ctrl and WT^Akita offspring (n = 11–12). CMP, common myeloid progenitor (Lineage^–c-Kit⁺Sca-1^–CD34⁺FcγR^–); MEP, megakaryocyte/erythroid progenitor (Lineage^–c-Kit⁺Sca-1^–CD34^–FcγR^–). (E) Percentage of HSC distribution in cell-cycle phases in adult Ctrl and WT^Akita offspring. Right panel shows representative FACS plots for Ki67/ Hoechst 33342 staining (n = 10). (F) Percentage of HSCs isolated from Ctrl and WT^Akita offspring that present FOXO3 nuclear localization at steady state (n = 6 with 50 individual cells analyzed for each). Right panel shows representative images of FOXO3 immunofluorescence analysis. Scale bar: 10 μm. (G and H) Competitive hematopoietic reconstitution assay for HSCs isolated from Ctrl (n = 6) and WT^Akita (n = 7) offspring from 1 experiment: PB chimerism over time (G) and BM chimerism for HSC subsets, 20 weeks after transplantation (H). Data are represented as means ± SD. Two-way ANOVA with Šidák’s post hoc test (B, C, D, G, and H) or with Tukey’s post hoc test (E); unpaired 2-tailed Student’s t test (F). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0005; ****P ≤ 0.0001.

Figure 4. GD altered the inflammatory hematopoietic response in offspring.

(A) Schematic of the experimental design for the in vivo LPS inflammatory challenge. (B) BM cellularity and absolute numbers of BM myeloid cells 3 days after LPS treatment. (C) Absolute numbers of BM MPP3 and GMP cells 3 days after LPS treatment (n = 3–6/group). (D) Inflammatory cytokine concentration in the mouse serum 3 days after LPS treatment (n = 4/group). (E) Schematic of experimental design to generate and activate BMDMs. (F) Absolute numbers of BMDMs in culture 24 hours after treatment with PBS or LPS/IFN-γ (n = 4–9/group). (G) Schematic of experimental design describing BMDM generation from transplanted mice. (H) Following transplantation, absolute numbers of BMDMs in culture 24 hours after treatment with PBS or LPS/IFN-γ (n = 4/group). Data are represented as means ± SD. One-way ANOVA with Tukey’s post hoc test (B and C) or 2-way ANOVA with Šidák’s post hoc test (D, F and G). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0005; ****P ≤ 0.0001.

Figure 5. AGER and NLRP3 function are necessary for the induction of the GD hematopoietic memory in offspring.

(A) Schematic of the experimental design. (B–D) Hematopoietic readout for offspring born to normal or diabetic pregnancy: absolute numbers of BM GMP cells (left graphs) and absolute numbers of BMDMs in culture 24 hours after treatment with PBS or LPS/IFN-γ (right graphs) in WT (B) (data also presented in Supplemental Figure 3D and Figure 4F), Ager^–/– (C), and Nlrp3^–/– backgrounds (D). (C: n = 8–10 for GMP; n = 10–20 for BMDM; D: n = 6 for GMP, n = 6–12 for BMDM). (E) Immune cell composition in placenta: percentages in CD45⁺ cells (n = 4–7/group). (F) RT-PCR analysis showing the expression of the Il6 and Ccl2 inflammatory cytokine genes. Results are expressed as fold change relative to Ctrls, set at 1 (n = 3). (G) RNA-Seq analysis on placental CD45⁺ cells isolated at G17 from Ctrl, WT^STZ, and Nlrp3^STZ dams (n = 3/group). Principal component analysis (PCA) of RNA-Seq data (left panel). Differential gene signature: heatmap using differentially expressed genes (DEGs) with log₂FC> |0.58| and FDR < 0.05 in all comparisons (central panel). GSEA of genes differentially expressed in WT^STZ cells compared with Ctrl and Nlrp3^STZ conditions (right panel). Data are represented as means ± SD. Unpaired 2-tailed Student’s t tests (B–D, left graphs); 1-way ANOVA with Tukey’s post hoc test (F) or 2-way ANOVA with Šidák’s post hoc test (B–D, right graphs; E). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0005; ****P ≤ 0.0001.

Figure 6. DNA methylation contributes to the maintenance of GD hematopoietic memory in adult.

(A) Immunoblot analysis of DNMT1 and DNMT3A in LSK cells isolated from Ctrl, GD offspring, (WT^Akita and WT^STZ), and adult mice with full-blown diabetes triggered by the Ins2^Akita mutation or STZ treatment. (B) Schematic of experimental design for in vivo treatment with vehicle or 5-azadC. (C) BM cellularity in WT^STZ offspring directly after 4 weeks of 5-azadC treatment (n = 5/group) or a 2-week recovery period after treatment (n = 4/group). (D) Immunoblot analysis of DNMT1 and DNMT3A in c-Kit⁺ cells isolated from WT^STZ after 5-azadC treatment or recovery period. (E) Absolute numbers of BMDMs in culture 24 hours after treatment with PBS or LPS/IFN-γ for 5-azadC–treated WT^STZ offspring after treatment (n = 12) or recovery period (n = 8). Data are represented as means ± SD. Unpaired 2-tailed Student’s t tests (C) and 2-way ANOVA with Šidák’s post hoc test (E). *P ≤ 0.05; ***P ≤ 0.0005.