Hematologic DNMT3A reduction and high-fat diet synergize to promote weight gain and tissue inflammation - PubMed (original) (raw)

. 2024 Feb 5;27(3):109122.

doi: 10.1016/j.isci.2024.109122. eCollection 2024 Mar 15.

Jaime M Reyes 1 2 3, Linda Zhang 1 2 4 5, Angelina S Bortoletto 1 2 4 5, Carina Rosas 1 2, Chun-Wei Chen 1 2 3, Sarah M Waldvogel 1 2 6 5, Anna G Guzman 1 2, Rogelio Aguilar 1 7, Sinjini Gupta 1 2, Ling Liu 8 9, Matthew T Buckley 10, Kalyani R Patel 11, Andrea N Marcogliese 11, Yumei Li 3, Choladda V Curry 11, Thomas A Rando 8 9, Anne Brunet 10 9, Ronald J Parchem 1 2, Rachel E Rau 1 7, Margaret A Goodell 1 2 3

Affiliations

Hematologic DNMT3A reduction and high-fat diet synergize to promote weight gain and tissue inflammation

Jaime M Reyes et al. iScience. 2024.

Erratum in

Abstract

During aging, blood cell production becomes dominated by a limited number of variant hematopoietic stem cell (HSC) clones. Differentiated progeny of variant HSCs are thought to mediate the detrimental effects of such clonal hematopoiesis on organismal health, but the mechanisms are poorly understood. While somatic mutations in DNA methyltransferase 3A (DNMT3A) frequently drive clonal dominance, the aging milieu also likely contributes. Here, we examined in mice the interaction between high-fat diet (HFD) and reduced DNMT3A in hematopoietic cells; strikingly, this combination led to weight gain. HFD amplified pro-inflammatory pathways and upregulated inflammation-associated genes in mutant cells along a pro-myeloid trajectory. Aberrant DNA methylation during myeloid differentiation and in response to HFD led to pro-inflammatory activation and maintenance of stemness genes. These findings suggest that reduced DNMT3A in hematopoietic cells contributes to weight gain, inflammation, and metabolic dysfunction, highlighting a role for DNMT3A loss in the development of metabolic disorders.

Keywords: Epigenetics; Immunology; Physiology; Stem cells research; Transcriptomics.

© 2024 The Author(s).

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

None

Graphical abstract

Figure 1

Figure 1

Circulating hematopoietic cells with reduced DNMT3A accelerate weight gain in mice on high-fat diet (A) Experimental design. Bone marrow from wild-type (WT) or Dnmt3a+/− (HET) mice was extracted and infused into irradiated WT recipients. Four weeks after transplantation, recipients were divided into groups and placed on high-fat diet (HFD) or regular diet (REG). Weight was monitored monthly. Donor cells bear the CD45.2 marker on all nucleated blood cells, while recipients bear the CD45.1 marker, allowing donor and recipients to be tracked in all blood and bone marrow lineages. (B) Weights over time of mice engrafted with WT or _Dnmt3a_-HET bone marrow fed either HFD or regular diet (REG). Two-way ANOVA test was used to determine significance. N = 8–10 male mice per group. (∗ denotes p value <0.05). Data are represented as mean ± SEM. Experiment repeated twice. (C) Quantification of flow-cytometric data depicting the percent of bone marrow common lymphoid progenitors (CLPs), common myeloid progenitors (CMPs), granulocyte monocyte progenitors (GMPs), and megakaryocytic erythrocytic progenitors (MEPs) from indicated 6 months post-transplant recipient mice (regular chow- REG, high-fat diet -HFD). Gating strategy is in Figure S1, and two-way ANOVA test was used to determine significance. Data are represented as mean ± SEM. (D) Relative proportion of B cells (B), T cells (T), and Myeloid cells (M) as a percentage of total donor cells (CD45.2) measured by flow cytometry 10 months after transplantation. Lineages defined as B220+ (B), CD4+ and CD8+ (T) and CD11b+ (M). Data are represented as mean ± SEM. (E) Quantification of flow-cytometric data depicting the percent of donor-derived Ly6g+ neutrophils as a proportion of total CD11b+ myeloid cells in peripheral blood 10 months post-transplant. Statistical significance was determined using a one-way ANOVA. Data are represented as mean ± SEM. (F) Quantification of flow-cytometric data depicting the percent of donor-derived Ly6C+ monocytes as a proportion of total CD11b+ myeloid cells in peripheral blood 10 months post-transplant. Statistical significance was determined using a one-way ANOVA. Data are represented as mean ± SEM.

Figure 2

Figure 2

High-fat diet coupled with circulating Dnmt3a+/− cells impacts glucose metabolism (A) Quantification of macrophages (F4/80+, CD11b+) as a percentage of donor-derived (CD45.2) myeloid cells (CD11b+) extracted from subcutaneous adipose of receipt mice in REG or HFD at 6 months compared to 10 months post-transplant. Tissue was processed to isolate stromal vascular cells which were then subjected to flow-cytometric analyses with the denoted markers. Statistical significance was determined using two-way ANOVA. Data are represented as mean ± SEM. (B) Microscopic imaging of subcutaneous adipose depots collected 40 months after transplantation (as in Figure 1A). Tissues were collected in 4% formalin, embedded in paraffin, and stained with hematoxylin and eosin (H&E), DAPI, and antibody against F4/80 to detect adipocytes, all nuclei (blue), and F4/80 macrophages (brown), respectively. Arrows denote inflammatory crown structures surrounding adipocytes. Image is taken at 20× magnification. Scale bars (bottom left) indicate 100 μm. (C) Quantification of F4/80 macrophages as a fraction of total cells per islet. Two islets were quantified for each of three mice for each group using the Analyze Particles function of ImageJ (See STAR Methods for details). Significance testing was performed using one-way ANOVA. Data are represented as mean ± SEM. (D) Glucose tolerance test 10 months after transplantation in mice fed HFD. Mice were fasted for 12 h prior to testing, infused with glucose intravenously at time zero, and measured for glucose in the peripheral blood in 30-min intervals. N = 10–12 per group. Significance testing was performed using one-way ANOVA. Data points are presented as mean ± SE. ∗ pval <0.05, ∗∗ pval <0.01. (E) Representative pancreatic beta islets from recipient mice (as in Figure 1A) fed an HFD. Pancreases were removed 10 months after transplantation, frozen sectioned, and immunostained for F4/80 macrophages (purple), Insulin (green), or all nuclei (DAPI, blue). Scale bars (bottom left) indicate 100 μm. (F) Quantification of insulin intensity per islet area. Two islets were quantified for each of three mice for each condition. Insulin expression was determined by analyzing the total insulin signal area per islet area using the Measure and Histogram functions of ImageJ. Significance testing was performed using ANOVA. Data are represented as mean ± SEM.

Figure 3

Figure 3

Pro-myeloid stem and progenitor cells from _Dnmt3a_-HET under HFD are transcriptionally enriched for pro-inflammatory pathways (A) Uniform manifold approximation and projection (UMAP) of hematopoietic progenitors (Lineageneg Sca1+ cKit+, “LSK” cells) (n = 28,984) from mice 10-month after transplantation mice (n = 2 each from separate transplant experiment). Colors indicate annotated cell type. Cell types were annotated using a mapped reference (See STAR Methods) dataset and SingleR. HSC, hematopoietic stem cell; LMPP, lympho-myeloid multipotent progenitor; MyMPP, myeloid multipotent progenitor; CLP, common lymphoid progenitor; MEP, megakaryocyte-erythroid progenitor; Mast, mast cell. (B) Cell type fractions (as defined in (a)) per condition. Values indicate percent of total cells per condition. (C) Force-directed graph layout of cell clusters within LSK overlaid with lineage trajectories (solid lines). Red indicates the pro-myeloid trajectory. (D) Tnf-α gene expression plot for WT or HET cells under REG along the pro-myeloid trajectory. The smoothed line indicates average gene expression along the pseudotime ordered trajectory. Shaded area indicates standard error (SE) for either WT REG (gray) or HET REG (red). (E) Tnf-α gene expression plot for WT or HET cells under HFD along the pro-myeloid trajectory. Average expression indicated for WT HFD (black) or HET HFD (blue). (F) Gene set enrichment analysis for genes differentially expressed between HET and WT cells within the pro-myeloid trajectory under REG (left) or HFD (right). Bars indicate changes in pathway ranks between diets with those in red and blue indicating significantly enriched pathways in HET or WT, respectively. Significance determined using Benjamin-Hochberg adjusted p value. p value <0.05.

Figure 4

Figure 4

IL-6 is differentially upregulated in recipient mice with HET bone marrow (A) Uniform manifold approximation and projection (UMAP) of myeloid-derived hematopoietic cells (CD45+, CD11b+ cells) (n = 724) isolated from adipose tissue of 12-months-old WT and germline Dnmt3a-HET mice. Colors indicate annotated cell type. Cell types were annotated using SingleR according ImmGen reference (Top). Clusters were then identified using Seruat function find neighbors (bottom). (B) Quantification of IL-6 expression in WT and HET in UMAP representation. (C) Quantification of IL-6 expression across myeloid cell clusters as defined in (A). (D) Enzyme-linked immunosorbent assay (ELISA) for IL-6 in plasma derived from mice 10 months post-transplant (as in Figure 1A) of WT or HET bone marrow and fed REG or HFD. Significance was determined using one-way ANOVA. Data are represented as mean ± SEM. (E) Flow-cytometric quantification of IL-6 using fluorescent antibody crosslinking from cKIT+ hematopoietic stem and progenitor cells. Cells were enriched using magnetic column and then assessed for IL-6 fluorescence. Significance was determined using one-way ANOVA. Data are represented as mean ± SEM. (F) Flow-cytometric quantification of IL-6 using fluorescent antibody crosslinking from adipose tissues derived from transplanted mice fed HFD. Tissue was processed to isolate stromal vascular cells and subject to flow-cytometric analysis with the denoted markers. Cells were gated by Cd45 positivity and assessed for IL-6 fluorescence. Significance was determined using one-way ANOVA. Data are represented as mean ± SEM.

Figure 5

Figure 5

IL-6 exposure reinforces pro-inflammatory pathways in HET bone marrow-derived macrophages (A) Experimental design. Bone marrow-derived cells from 2-month-old WT or HET (n = 3 per condition) were plated and treated with 50 ng/mL MCSF for 7 days. Cells were then treated with IL-6 at 10 ng/mL for an additional 2 days to mimic growth under inflammatory stimulus. (B) Flow-cytometric quantification of IL-6 using fluorescent antibody crosslinking of in vitro differentiated CD11b cells as denoted from day 0 of bone marrow isolation (NT) to day 7 of differentiation. Significance was determined using one-way ANOVA. Data are represented as mean ± SEM. (C) Phagocytic activity of in vitro differentiated WT and HET cells at day 7 of differentiation with or without IL-6 treatment (added on day 6). On day 7 cells were incubated with PE + magnetic beads for the denoted time points and following phagocytic activity was measured. Macrophages were gated based on CD11b+, F4/80+ and incorporation of the PE + beads was then determined. Two-way ANOVA was used to determine statistical significance. Data are represented as mean ± SEM. (D) Heatmap of the top 200 most variable genes across all samples. Genes were subset by expression patterns across conditions (red indicates higher expression, blue indicates lower expression). Group I (blue) includes genes that were upregulated in NT samples; Group II (gold) indicates genes upregulated only in HET_NT samples; Group III (pink) indicates genes upregulated upon treatment by IL-6; Group IV (green) indicates genes upregulated in WT_NT only; Group V (red) indicates genes that were upregulated in HET_NT and were further upregulated in HET_IL-6 samples. Only the top 200 variable genes are shown, sub-setting and testing were performed using the full dataset of expressed genes. (E) Mean fluorescence intensity (MFI) of the cells from (i). two-way ANOVA was used to determine statistical significance. Data are represented as mean ± SEM. (F) Differential transcript expression from mature F4/80 macrophages as described in (G) and treated with IL-6. Bulk RNA-seq was performed. Red dots indicate transcripts with p value <0.005 and fold change > 2-fold. Significance testing was performed using Bejamini-Hochberg method to correct p values for multiple hypothesis testing. (G) GSEA analysis for the hallmark pathways gene sets from MSigDB on IL-6-treated WT versus HET macrophages. Red and blue bars (HET and WT respectively) indicate significantly enriched pathways determined by GSEA normalized ES. Adjusted p value < 0.05.

Figure 6

Figure 6

Myeloid progenitors with reduced DNMT3A lack are unable to gain DNA methylation at critical gene regulatory loci (A) Schematic strategy to define differentially methylated regions (DMRs) as defined for WT- and HET-derived common myeloid progenitors (CMPs) and adipose tissue-isolated macrophages (MACs) 6 moths post-transplant from recipients fed regular diet (REG) or high-fat diet (HFD). Diet-DMRs are defined as regions that gain/loss (∼25%) methylation in HFD compared to REG for CMP or MAC. DNMT3A-dependent (D3a dep) Diet-DMRs are regions that had little change (<10%) in HET HFD compared to REG. Differentiation DMRs (Diff-DMRs) are defined as regions that gain/loss (∼25%) methylation across differentiation from CMP to MAC. DNMT3A-dependent (D3a dep) Diff-DMRs are regions that had little change (<10%) across differentiation in HET samples. (B) Violin plot of mean methylation for hypermethylated Diet-DMRs and their mean methylation in WT and HET CMP and macrophages. Boxplots indicate +/− SE mean. (C) Violin plot of mean methylation for hypermethylated Diff-DMRs and their mean methylation in WT and HET CMP and macrophages. Boxplots indicate +/− SE mean. (D) Histogram displaying change in mean methylation in HFD compared to regular diet of D3a-dependent Diff-DMRs. These regions were defined when comparing MAC REG to CMP REG. We then calculated mean difference in DNA methylation in HFD compared to REG in either CMPs (blue) or MAC (red). A rightward shift in the line indicates a gain of methylation in HFD for the same loci that gain methylation in differentiation. (E) Mean methylation +/1 1 kb from D3a-dependent Diff-DMRs. Methylation difference in edges is marked with yellow box (left). For the same regions defined in (e) we plotted mean methylation +/1 1 kb in HFD samples (right). (F) Whole-genome DNA methylation track from WT and HET cells. Vertical yellow boxes indicate regions that gain DNA methylation during differentiation in WT-CMPs compared to MAC but fail to gain similar degree of methylation in HET cells.

References

    1. Libby P., Sidlow R., Lin A.E., Gupta D., Jones L.W., Moslehi J., Zeiher A., Jaiswal S., Schulz C., Blankstein R., et al. Clonal Hematopoiesis: Crossroads of Aging, Cardiovascular Disease, and Cancer: JACC Review Topic of the Week. J. Am. Coll. Cardiol. 2019;74:567–577. doi: 10.1016/j.jacc.2019.06.007. -DOI -PMC -PubMed
    1. Luis T.C., Wilkinson A.C., Beerman I., Jaiswal S., Shlush L.I. Biological implications of clonal hematopoiesis. Exp. Hematol. 2019;77:1–5. doi: 10.1016/j.exphem.2019.08.004. -DOI -PMC -PubMed
    1. Bick A.G., Weinstock J.S., Nandakumar S.K., Fulco C.P., Bao E.L., Zekavat S.M., Szeto M.D., Liao X., Leventhal M.J., Nasser J., et al. Inherited causes of clonal haematopoiesis in 97,691 whole genomes. Nature. 2020;586:763–768. doi: 10.1038/s41586-020-2819-2. -DOI -PMC -PubMed
    1. Jaiswal S., Fontanillas P., Flannick J., Manning A., Grauman P.V., Mar B.G., Lindsley R.C., Mermel C.H., Burtt N., Chavez A., et al. Age-related clonal hematopoiesis associated with adverse outcomes. N. Engl. J. Med. 2014;371:2488–2498. doi: 10.1056/NEJMoa1408617. -DOI -PMC -PubMed
    1. Genovese G., Kähler A.K., Handsaker R.E., Lindberg J., Rose S.A., Bakhoum S.F., Chambert K., Mick E., Neale B.M., Fromer M., et al. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. N. Engl. J. Med. 2014;371:2477–2487. doi: 10.1056/NEJMoa1409405. -DOI -PMC -PubMed

Grants and funding

LinkOut - more resources