Myeloid HMG-CoA Reductase Determines Adipose Tissue Inflammation, Insulin Resistance, and Hepatic Steatosis in Diet-Induced Obese Mice

One of the characteristics of obesity is chronic low-grade inflammation, which is associated with accumulation of adipose tissue macrophages (ATMs). The ATMs, in turn, can contribute to insulin resistance and type 2 diabetes (1,2). The activation state of ATMs shifts from an alternatively activated state (M2) in lean mice to a classically activated state (M1) in obese mice (3). Adipose tissue inflammation and insulin resistance are ameliorated by inhibiting the accumulation of ATMs through genetic ablation of CD11c⁺ cells (4) or knockdown of C-C motif chemokine receptor 2 (CCR2) in macrophages (5). These findings indicate that targeting ATMs is a promising strategy to improve insulin resistance.

HMG-CoA reductase (HMGCR) is the rate-limiting enzyme in the mevalonate pathway (6). Inhibitors of HMGCR (statins) are widely used to prevent the occurrence of coronary heart disease and other atherosclerotic diseases, primarily by reducing blood cholesterol levels. The atheroprotective effects of statins have been ascribed not only to cholesterol lowering but also to various effects on vascular endothelial cells, vascular smooth muscle cells, and immune cells, including monocytes and macrophages (7). We recently reported that myeloid cell–specific reduction of Hmgcr protects against atherosclerosis by reducing migration of macrophages to the aorta in hypercholesterolemic mice (8). On the basis of these findings, we hypothesized that myeloid cell–specific reduction of Hmgcr also improves insulin sensitivity by reducing the recruitment of macrophages to adipose tissues in obesity.

Animal studies were performed according to the regulations of the Animal Care Committees of Jichi Medical University. Myeloid cell–specific Hmgcr reduction (Hmgcr^m−/m−) mice were generated by crossing Hmgcr^fl/fl mice (control) (9) with LysMcre mice (The Jackson Laboratory) (8). All mice used in this study had a C57BL/6J genetic background. Eight-week-old male mice were fed either normal chow diet (NCD) (12% kcal fat, CE-2; CLEA Japan, Inc.) or high-fat diet (HFD) (60% kcal fat from lard, D12492; Research Diets) for 24 weeks. Plasma levels of glucose, insulin, tumor necrosis factor-α (TNF-α), and adiponectin were measured by Glucose CII Test Wako Kit (Wako), Mouse Insulin ELISA Kit (Morinaga), Mouse TNF-α Quantikine HS ELISA Kit (R&D Systems), and Mouse/Rat Adiponectin ELISA Kit (Otsuka Pharmaceuticals), respectively. Plasma lipids and liver triglyceride (TG) content were measured as described previously (8,9).

Protein extracts from the tissues were subjected to SDS-PAGE and visualized with primary antibodies against p-Akt^Ser473 or Akt and chemiluminescence of horseradish peroxidase–conjugated secondary antibody (Cell Signaling Technology) using ECL Prime Western Blotting Detection Reagent (GE Healthcare).

The tissue sections were stained for F4/80 using a rat anti-F4/80 antibody (1:100 dilution; Bio-Rad), Histofine Simple Stain MAX PO (Nichirei), and diaminobenzidine (Sigma-Aldrich). Whole-mount immunostaining was performed as described previously (10) using the rat anti-F4/80 antibody followed by Alexa Fluor 488 goat anti-rat IgG (Thermo Fisher Scientific) or a rabbit anti-Ki67 antibody (Abcam) followed by Alexa Fluor 568 goat anti-rabbit IgG (Thermo Fisher Scientific). Apoptosis was determined by TUNEL staining using an In situ Apoptosis Detection Kit (Takara Bio). F4/80⁺ cells were sorted using MS Columns (Miltenyi Biotec) after incubation with biotin-conjugated anti-F4/80 antibody (eBioscience) followed by magnetic labeling with Anti-Biotin MicroBeads (Miltenyi Biotec) from the stromal vascular fraction of epididymal white adipose tissue (eWAT).

Chemotaxis of thioglycolate-elicited macrophages (TGEMs) was quantified as described previously (8) with slight modifications. Total quantitative real-time PCR was performed as described previously (8). The primer-probe sets used in the experiments are listed in Supplementary Table 1.

All data are presented as mean ± SD. GraphPad Prism 6 software was used for data analyses. Unpaired Student t test or repeated-measures ANOVA with Bonferroni multiple comparison test was used for comparisons as appropriate. Differences were considered significant at P < 0.05.

The data and critical resources supporting their reported findings, methods, and conclusions are available from the corresponding author upon reasonable request.

There were no significant differences in the body weight, weights of eWAT, weights of liver, and food intake between the two groups under either the NCD or the HFD condition (Supplementary Fig. 1A–D). Body compositions estimated by CT scan and plasma lipids were not significantly different between the two groups under the HFD condition (Supplementary Fig. 1E–I).

Fasting plasma glucose concentration, insulin concentration, and HOMA of insulin resistance (HOMA-IR) were lower Hmgcr^m−/m− mice than in Hmgcr^fl/fl mice under the HFD condition (Fig. 1A–C). Hmgcr^m−/m− mice showed significant improvement in glucose tolerance by glucose tolerance test compared with Hmgcr^fl/fl mice under the HFD condition (Fig. 1D). Results of insulin tolerance tests showed that Hmgcr^m−/m− mice had significant improvement in insulin sensitivity compared with Hmgcr^fl/fl mice under the HFD condition (Fig. 1E). These parameters were not significantly different between the two groups under the NCD condition (Supplementary Fig. 2A–E).

To confirm the improvement of systemic insulin sensitivity, we performed immunoblot analysis for p-Akt^Ser473 in the liver, eWAT, and gastrocnemius muscle (Fig. 1F–K). Of note, the magnitudes of insulin-stimulated increases in p-Akt^Ser473 levels were significantly higher in Hmgcr^m−/m− mice than in Hmgcr^fl/fl mice fed HFD. These results indicate that reduction of Hmgcr in myeloid cells attenuates HFD-induced insulin resistance.

ATMs typically surround dead adipocytes, known as crown-like structures (CLSs) (11). The number of F4/80⁺ CLSs was reduced by 77% in Hmgcr^m−/m− mice compared with Hmgcr^fl/fl mice under the HFD condition (Fig. 2A and B). The mRNA expression levels of marker genes for general macrophages, such as F4/80 and Cd68, and M1, such as Cd11c, Tnf-α, Il-1β, and Mcp-1, were significantly lower in Hmgcr^m−/m− mice than in Hmgcr^fl/fl mice under the HFD condition (Fig. 2C–I). On the other hand, the mRNA expression levels of M2 markers, such as Cd206 and Cd163, were not different between the two groups (Fig. 2J and K).

To determine how the reduction of myeloid Hmgcr alleviated insulin resistance, we measured mRNA expression levels and plasma concentrations of adiponectin, which is an adipocyte-derived insulin-sensitizing hormone (12), and measured plasma concentrations of TNF-α which antagonizes insulin action (13). Hmgcr^m−/m− mice had significantly higher mRNA expression levels and plasma concentrations of adiponectin than Hmgcr^fl/fl mice under the HFD condition (Fig. 2L and M). Furthermore, plasma TNF-α concentrations were significantly lower in Hmgcr^m−/m− mice under the HFD condition (Fig. 2N). To investigate the inflammatory changes of ATMs, we collected ATMs from eWAT from mice fed an HFD. The mRNA expression levels of genes related to cholesterol metabolism and proinflammatory cytokines in F4/80⁺ cells were not significantly different between the two groups under the HFD condition except for Hmgcr (Fig. 2O and P). These results indicate that the reduction of Hmgcr in myeloid cells attenuates ATM accumulation and HFD-induced inflammation in WAT.

Liver TG content reduced significantly by 52% in Hmgcr^m−/m− mice compared with Hmgcr^fl/fl mice under the HFD condition (Fig. 3A and B). The numbers of F4/80⁺ macrophages in the liver were not different between the two groups (Fig. 3C and D). In agreement with this, the mRNA expression levels of marker genes for general macrophages, those for M1, and those for M2 were not different between the two groups (Fig. 3E–L). These results indicate that the reduction of Hmgcr in myeloid cells inhibits HFD-induced hepatic steatosis without affecting the number of liver macrophages or expression of proinflammatory cytokines in the liver.

We hypothesized that the attenuation of ATM accumulation in Hmgcr^m−/m− mice fed HFD was due to impaired chemotaxis to MCP-1, which is critical for ATM accumulation during obesity (14). The mRNA expression levels of Ccr2, a receptor for MCP-1, was not different between the two groups of F4/80⁺ cells isolated from eWAT (Supplementary Fig. 3A). On the other hand, chemotaxis of macrophages toward MCP-1 was reduced by 54% in TGEMs from Hmgcr^m−/m− mice compared with those from Hmgcr^fl/fl mice. The addition of mevalonate or geranylgeranyl pyrophosphate (GGPP) reversed the reduction of chemotaxis toward MCP-1 but not the addition of squalene or farnesyl pyrophosphate (FPP). These results suggest that GGPP plays an important role in macrophage chemotaxis (Fig. 4A and Supplementary Fig. 3B).

To evaluate the numbers of proliferating and/or apoptotic ATMs, which are important determinants of the number of ATMs (10,15), we performed whole-mount immunostaining for Ki67 and TUNEL (Fig. 4B–E). Neither Ki67⁺ ATMs nor TUNEL⁺ ATMs were significantly different between the two groups.

In obese mice, adipose tissues are reported to recruit bone marrow–derived macrophages (2). Inhibiting chemotaxis of macrophages, such as knockdown of CCR2, has been reported to alleviate inflammation in adipose tissues and insulin resistance, primarily as a result of the reduction of ATMs (2,5). Consistent with these reports, the current study clearly shows that reduced chemotaxis of macrophages with reduced Hmgcr is associated with a decreased number of ATMs and amelioration of insulin resistance and fatty liver.

How did the reduced supply of GGPP impair chemotaxis in Hmgcr^m−/m− TGEMs? It is well known that Rho GTPases play a fundamental role in the control of cell shape and motility (16), and GGPP plays an essential role for the function of these small GTPases, including Rho GTPases, through protein prenylation (7). However, studies using macrophages isolated from knockout mice have shown that Rac1, Rac2, RhoA, RhoB, and RhoC are not essential for macrophage migration, although they do affect cell shape and adhesion (17,18). To further complicate the issue, deficiency of geranylgeranyltransferase type 1 (GGTase 1), which supposedly activate Rho GTPases by geranylgeranylation, in macrophages increases interleukin-1β production by increasing active GTP-bound Rac1, Cdc42, and RhoA in mice (19). Consistently, we found increased expression of proinflammatory cytokines when stimulated with lipopolysaccharide and increased amounts of membrane-bound Rac1, Cdc42, and RhoA in Hmgcr^m−/m− TGEMs (8).

Akula et al. (20) have reported the results that would solve the mystery. According to them, protein geranylgeranylation enables Toll-like receptor–induced activation of phosphatidylinositol-3-OH kinase [PI(3)K] by promoting the interaction between the small GTPase Kras and PI(3)K catalytic subunit p110δ. In the absence of geranylgeranylation, compromised PI(3)K activity allows an unchecked toll-like receptor–induced inflammatory response and constitutive activation of pyrin inflammasome. Taken together, it is unlikely that the reduced chemotaxis of Hmgcr^m−/m− TGEMs is caused by inactivation of Rho GTPases.

More specifically, G-protein subtypes γ2 and 3, which are essential for intracellular signaling of CCR2 in macrophages, are geranylgeranylated (21). Fluvastatin inhibits the migration of RAW264.7 cells by decreasing PI(3)K-mediated production of phosphatidylinositol (3,4,5)-trisphosphate (22), suggesting that similar defective downstream signaling underlies the impaired chemotaxis of Hmgcr^m−/m− TGEMs. It is also possible that decreased supply of GGPP impairs geranylgeranylation of Rab GTPases, which are catalyzed by GGTase 2, also known as Rab GGTase, thereby impairing cell migration (23). Moreover, a decrease in cholesterol in the lipid raft of plasma membrane may inhibit the migratory activity of macrophages, as reported in mice whose fatty acid synthase (Fas) is ablated in myeloid cells (24). Further studies are warranted to clarify the mechanism behind the impaired migratory activity of Hmgcr^m−/m− TGEMs.

Why did HFD elicit inflammation in WAT but not in the liver? van der Heijden et al. (25) showed that AT inflammation was present after 24 weeks of HFD, whereas hepatic inflammation was not detected until 40 weeks of HFD, indicating that AT inflammation is established before the development of hepatic inflammation.

In conclusion, reduction of myeloid Hmgcr improves glucose tolerance and insulin sensitivity by decreasing the number of ATMs and inflammation of adipose tissues by reducing the chemotaxis of macrophages to adipose tissues. Moreover, reduction of myeloid Hmgcr attenuates hepatic steatosis. Therefore, reducing myeloid HMGCR may be a promising strategy for improving insulin resistance and hepatic steatosis in obesity.

The authors thank Mika Hayashi, Nozomi Takatsuto, and Mihoko Sejimo (Division of Endocrinology and Metabolism, Department of Internal Medicine, Jichi Medical University) for excellent technical support. The authors also thank Biopathology Institute Co., Ltd., for helping with the performance of histological analyses.

Funding. This study was supported by Grants-in-Aid for Scientific Research-KAKENHI and Program for the Strategic Research Foundation at Private Universities (2011-2015) Cooperative Basic and Clinical Research on Circadian Medicine and Non-Communicable Diseases from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Duality of Interest. This study was supported by unrestricted grants from Astellas Pharma, Daiichi Sankyo Co, Shionogi Co, Boehriger Ingelheim Japan, Ono Pharma, Mitsubishi Tanabe Pharma, Takeda Pharma Co, Toyama Chemical Co, Teijin, Sumitomo Dainippon Pharma, Sanofi K.K., Novo Nordisk Pharma, MSD K.K., Pfizer Japan, Novartis Pharma, and Eli Lilly Co. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. A.T. and S.T. designed and performed the experiments, maintained the mice, analyzed and interpreted the data, and wrote the manuscript. S.N. designed the experiments, maintained the mice, and contributed to the discussion. D.Y., A.M., T.W., M.I., H.Y., C.E., M.T., and K.E. contributed to the discussion. S.I. designed the experiments and wrote the manuscript. S.I. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.