Local proliferation of macrophages contributes to obesity-associated adipose tissue inflammation

Abstract

Adipose tissue (AT) of obese mice and humans accumulates immune cells, which secrete cytokines that can promote insulin resistance. AT macrophages (ATMs) are thought to originate from bone-marrow-derived monocytes, which infiltrate the tissue from the circulation. Here, we show that a major fraction of macrophages unexpectedly undergo cell division locally within AT, as detected by Ki67 expression and 5-ethynyl-2'-deoxyuridine incorporation. Macrophages within the visceral AT (VAT), but not those in other tissues (including liver and spleen), displayed increased proliferation in obesity. Importantly, depletion of blood monocytes had no impact on ATM content, whereas their proliferation in situ continued. Treatment with monocyte chemotactic protein 1 (MCP-1) induced macrophage cell division in AT explants, whereas mcp-1 deficiency in vivo decreased ATM proliferation. These results reveal that, in addition to blood monocyte recruitment, in situ proliferation driven by MCP-1 is an important process by which macrophages accumulate in the VAT in obesity.

Figure 1. Adipose tissue macrophages express the cell division marker Ki67

See also Figure S1. SVF from VAT of WT and ob/ob mice was isolated and analyzed by flow cytometry. (A) Representative flow cytometry dot plots. (B) Percentage of macrophages in SVF. (C) Representative flow cytometry dot plots of ATMs stained with Ki67. (D) Percentage of macrophages expressing Ki67. n=30-31 from 6 independent experiments. (E) Microscopy of plated SVF stained with antibodies against F4/80 (red) and Ki67 (green). Nuclei were stained with DAPI (Blue). 63x magnification images. Scale bar = 5 μm. (F) VAT of ob/ob mice containing CLS stained with antibodies against F4/80 (red) and Ki67 (green). Nuclei were stained with DAPI (Blue). 20x magnification images. Scale bar = 40 μm. (G) SVF from VAT of mice fed a ND or HFD for 7 weeks was isolated and analyzed by flow cytometry. Representative flow cytometry dot plots. (H) Mean percentage of macrophages in SVF. (I) Representative flow cytometry dot plots of ATMs stained with Ki67. (J) Percentage of macrophages expressing Ki67. n=30-31 from 6 independent experiments. (K) Representative dot plot of SVF from human SAT stained with Ki67. (L) Percentage of macrophages in SVF from VAT of mice fed a HFD for 7 weeks and fasted for 24 hours. n=8-18. (M) Number of macrophages in SVF from VAT of mice fed a HFD and fasted for 24 hours. n=5. (N) Percentage of macrophages expressing Ki67 in fasted mice. n=8-18. (O) Percentage of non-macrophages (CD11b−/F4/80−) expressing Ki67. n=8-18. All graphs are expressed as mean ± s.e.m. Statistical significance was determined by Student’s t-test. ***p<0.001; **p<0.01; *p<0.05 .

Figure 2. Obesity increases macrophage proliferation specifically in adipose tissue

See also Figure S2. WT and ob/ob mice were i.p. injected with EdU and (A) AT, (B) spleen, (C) liver and (D) blood were collected and digested 3 hours after treatment. All cells were stained and analyzed by flow cytometry. Representative dot plots depict the EdU incorporation into all cells of the respective tissues or blood monocytes. (E) Mean percentage of EdU incorporation rate of the macrophages of each tissue ± s.e.m. (F) Percentage of macrophages in each tissue. n= 14-15 from 3 independent experiments for AT and blood, and n=9-10 from 2 independent experiments for spleen and liver. Statistical significance was determined by Student’s t-test. ***p<0.001.

Figure 3. Adipose tissue macrophages proliferate independently of monocyte recruitment

See also Figure S3. ob/ob mice were i.v. injected with either PBS-Lipo or Clod-Lipo every 16 hours. (A) Example flow cytometry dot-plots of CD11b⁺ blood cells show depletion of monocytes, and (B) the quantitation of blood monocytes expressed as a percentage of total blood cells. n=10 for 16h-48h of liposome treatment and n=4-5 for the 96h time point. (C) Macrophage content in the AT of PBS-liposome and clodronate-liposome-treated ob/ob mice; n=5 mice per group. 18 hours after initial injection, the mice were given drinking water containing EdU. (D) Diagram representing experimental design of treatment. (E) Representative flow cytograms and (F) quantification of EdU incorporation into ATMs during 32h and 80h of exposure to EdU drinking water in PBSLipo-treated and monocyte-depleted Clod-Lipo-treated ob/ob mice. (G) Quantification of EdU incorporation into ATMs during 80h of exposure to EdU drinking water in VAT and SAT in lean WT and ob/ob obese mice. n= 5 mice per group. All graphs are expressed as mean ± s.e.m. Statistical significance was determined by Student’s t-test or two-way ANOVA followed by Tukey post-test. ***p<0.001; **p<0.01; *p<0.05. (H) VAT of ob/ob mice containing CLS stained with antibodies against F4/80 (red) and EdU (green). Nuclei were stained with DAPI (Blue). 20x magnification images. Scale bar = 40 μm.

Figure 4. MCP-1 is required for optimal adipose tissue macrophage proliferation

See also Figure S4. VAT was isolated from mice fed a ND or HFD for 7 weeks. (A) IL-4, (B) M-CSF, (C) OPN and (D) MCP-1 expression was measured by RT-PCR. n=5. (E) Expression of OPN and (F) MCP-1 in SAT, VAT and liver from mice fed a ND or HFD for 7 weeks. n=5. (G) Expression of MCP-1 in VAT of mice fed a HFD for 7 weeks and fasted for 24 hours. (H) Percentage and (I) number of macrophages in SVF from VAT of MCP-1 KO and WT mice fed a HFD for 6 weeks and fasted for 18 hours. n=5. (J) Percentage and (K) number of EdU⁺ macrophages in AT of MCP-1 KO and WT mice. (L) Number of blood monocytes in MCP-1 KO and WT mice. Explants from VAT of 5 ob/ob mice were treated with 1 and 10 ng/ml of MCP-1 in presence of 10 μM of EdU for 48 hours. Graph represents the number of (M) macrophages and (N) EdU⁺ macrophages in explants. (O) Body weight of MCP-1 KO and WT mice fed a HFD for 6 weeks. (P) GTT and (Q) fasting glycaemia. All graphs are expressed as mean ± s.e.m. Statistical significance was determined by Student’s t-test or two-way ANOVA followed by Tukey post-test. ***p<0.001; **p<0.01; *p<0.05.