Macrophage JAK2 deficiency protects against high-fat diet-induced inflammation

Abstract

During obesity, macrophages can infiltrate metabolic tissues, and contribute to chronic low-grade inflammation, and mediate insulin resistance and diabetes. Recent studies have elucidated the metabolic role of JAK2, a key mediator downstream of various cytokines and growth factors. Our study addresses the essential role of macrophage JAK2 in the pathogenesis to obesity-associated inflammation and insulin resistance. During high-fat diet (HFD) feeding, macrophage-specific JAK2 knockout (M-JAK2^-/-) mice gained less body weight compared to wildtype littermate control (M-JAK2^+/+) mice and were protected from HFD-induced systemic insulin resistance. Histological analysis revealed smaller adipocytes and qPCR analysis showed upregulated expression of some adipogenesis markers in visceral adipose tissue (VAT) of HFD-fed M-JAK2^-/- mice. There were decreased crown-like structures in VAT along with reduced mRNA expression of some macrophage markers and chemokines in liver and VAT of HFD-fed M-JAK2^-/- mice. Peritoneal macrophages from M-JAK2^-/- mice and Jak2 knockdown in macrophage cell line RAW 264.7 also showed lower levels of chemokine expression and reduced phosphorylated STAT3. However, leptin-dependent effects on augmenting chemokine expression in RAW 264.7 cells did not require JAK2. Collectively, our findings show that macrophage JAK2 deficiency improves systemic insulin sensitivity and reduces inflammation in VAT and liver in response to metabolic stress.

Figure 1

Generation and characterization of M-JAK2^+/+ and M-JAK2^−/− mice. (a) mRNA expression and (b) protein levels for JAK2 in thioglycolate-elicited peritoneal macrophages (PMs) from 8- to 10-week old M-JAK2^+/+ and M-JAK2^−/− mice (n = 6–7). (c) Representative Western blots of protein lysates from PMs, liver, visceral adipose tissue (VAT), skeletal muscle (SM), and hypothalamus from M-JAK2^+/+ and M-JAK2^−/− mice for JAK2 and tubulin, as a loading control. Full blots in Supplementary Fig. S6. (d) Body weight measurements of male and female M-JAK2^+/+ and M-JAK2^−/− mice fed NCD (n = 12-13 males, n = 12–13 females) or HFD (n = 13–14 males, n = 11–13 females); *HFD M-JAK2^+/+ compared to HFD M-JAK2^−/− mice, † NCD M-JAK2^+/+ compared to HFD M-JAK2^+/+ mice, § NCD M-JAK2^−/− compared to HFD M-JAK2^−/−. (e) Tissue weights relative to total body weight of pancreas (Panc.), spleen, liver, kidney, heart, and lungs; and (f) absolute weights and (g) weights relative to total body weight of adipose depots: epidydimal (epi., male), parametrial (para., female), retroperitoneal (retro.), mesenteric (mes.), inguinal (ing.), and interscapular brown adipose tissue (BAT) in M-JAK2^+/+ and M-JAK2^−/− male and female mice fed NCD (n = 5–8 males, n = 9 females) or HFD (n = 7–9 males, n = 10–11 females). (h) Representative image of whole animal MRI scan; (i) total VAT volume and (j) percentage of abdominal adipose volume relative to total body volume determined using MRI in M-JAK2^+/+ and M-JAK2^−/− male mice fed HFD (n = 4). All results are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Decreased adipocyte hypertrophy in HFD-fed M-JAK2^−/− mice. (a) Representative micrographs of H&E stained VAT sections of HFD-fed M-JAK2^+/+ and M-JAK2^−/− mice. Quantitative analysis of adipocytes in VAT sections from HFD fed M-JAK2^+/+ and M-JAK2^−/− (n = 7–9): (b) average adipocyte area, (c) adipocyte area distribution, (d) average adipocyte diameter, (e) adipocyte diameter distribution, and (f) adipocyte number per field. (g) mRNA expression of genes involved in adipogenesis in VAT of HFD fed M-JAK2^+/+ and M-JAK2^−/− mice (n = 8–10). All results are mean ± SEM; *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure 3

Increased systemic insulin sensitivity in HFD fed M-JAK2^−/− mice. Male and female M-JAK2^+/+ and M-JAK2^−/− mice subjected to glucose tolerance test during (a) NCD feeding (n = 5–8 males, n = 9 females) or (b) HFD feeding (n = 9–10 males, n = 11–13 females) feeding. Male and female M-JAK2^+/+ and M-JAK2^−/− mice subjected to insulin tolerance test during (c) NCD feeding (n = 5–8 males, n = 8 females) or (d) HFD feeding (n = 7–9 males, n = 9–12 females) feeding. Area under the curve (AUC) was measured using baseline glucose values. (e) Random and fasting blood glucose levels measured in male and female M-JAK2^+/+ and M-JAK2^−/− mice during NCD feeding (n = 5–8 males, n = 9 females) or HFD feeding (n = 9–10 males, n = 9–12 females). (f) Fasting serum insulin levels and (g) homeostasis model assessment of insulin resistance (HOMA-IR) calculated in male and female M-JAK2^+/+ and M-JAK2^−/− mice fed NCD (n = 6–7 males, n = 7–8 females) or HFD (n = 9–10 males, n = 8–10 females). Western blots of protein lysates from (h) visceral adipose tissue (VAT), (j) liver and (l) skeletal muscle (SM) for phospho (Ser473)-AKT (pAKT), total AKT (tAKT) and GAPDH as a loading control, and quantification of protein levels expressed as fold change over M-JAK2^+/+ injected without insulin in (i) VAT, (k) liver and (m) SM from HFD fed M-JAK2^+/+ and M-JAK2^−/− mice injected without (−) or with ( + ) insulin (n = 3–4). Full blots in Supplementary Fig. S7. All results are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Decreased CLS and inflammation in VAT of HFD fed M-JAK2^−/− mice. (a) Representative micrographs of macrophage marker F4/80 stained VAT sections of HFD-fed M-JAK2^+/+ and M-JAK2^−/− mice; crown-like structure (CLS) are indicated by arrowheads. (b) Quantification of CLS from VAT sections in at least 15 fields at 100X (n = 7–8). mRNA expression in VAT from HFD-fed M-JAK2^+/+ and M-JAK2^−/− mice for (c) macrophage (Mϕ) and M1 markers, (d) M2 markers, and (e) chemokine receptors and its associated ligands (n = 8–11). All results are mean ± SEM; *p < 0.05.

Figure 5

Reduced hepatic inflammation in HFD fed M-JAK2^−/− mice. (a) Representative micrographs of H&E (top) and Oil Red O (bottom) stained liver sections from NCD or HFD fed M-JAK2^+/+ and M-JAK2^−/− mice. Quantification of (b) total triacylglycerol (TG) levels and (c) amount of saturated (SFAs), mono-unsaturated (MUFAs), Omega-6, Omega-3, polyunsaturated (PUFAs) and highly unsaturated (HUFAs) fatty acids (FA) of TGs in liver from NCD- (n = 4–6) and HFD- (n = 9–10) fed M-JAK2^+/+ and M-JAK2^−/− male mice. mRNA expression in liver from HFD-fed M-JAK2^+/+ and M-JAK2^−/− mice for (d) macrophage (Mϕ) and M1 markers, (e) M2 markers, and (f) chemokine receptors and its associated ligands (n = 8–9). All results are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

Characterization of circulating leukocytes, and macrophages in liver and VAT. (a) Total white blood cells, and (b–g) frequency and absolute number of (b) Ly6C high (Ly6C^hi) monocytes, (c) Ly6C low (Ly6C^lo) monocytes, (d) neutrophils, (e) eosinophils, (f) B lymphocytes, and (g) T lymphocytes in M-JAK2^+/+ and M-JAK2^−/− mice fed a NCD (n = 5–7) or a HFD (n = 6–7). (h) Absolute numbers of CD45⁺CD11b⁺F4/80⁺ macrophages in liver, visceral adipose tissue (VAT) and spleen, (i) frequency and (j) absolute numbers of CD11c⁺CD206⁻ and CD11c⁻CD206⁺ macrophages in VAT and liver, and (k) expression of the macrophage activation markers CD80 and CD86 in the VAT and liver of M-JAK2^+/+ and M-JAK2^−/− mice fed HFD (n = 8–9 mice per group for VAT and n = 5–6 per group for spleen and liver, two independent experiments). Mean fluorescent intensity (MFI). All results are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 7

Cytokine measurements in peritoneal macrophages. mRNA expression as fold change over NCD-fed M-JAK2^+/+ mice for (a) M1 markers, (b) M2 markers, and (c) chemokine receptors and its associated ligands in thioglycolate-elicited peritoneal macrophages from NCD (n = 3–4) or HFD (n = 5) fed M-JAK2^+/+ and M-JAK2^−/− mice. All results are mean ± SEM; *p < 0.05.

Figure 8

Chemokine expression in response to leptin in RAW 264.7 cells. (a) mRNA expression of chemokines as fold change over vehicle in macrophage cell line RAW 264.7 after stimulation with vehicle or varying concentrations of leptin. Three independent experiments performed in triplicates. (b) mRNA expression of Jak2 in RAW 264.7 cells transfected with scramble siRNA control (siScr) or Jak2 siRNA (siJak2) as fold change over siScr. (c) Quantification of total JAK2 levels relative to GAPDH, expressed as fold change over siScr, in RAW 264.7 cells transfected with siScr or siJak2 (n = 7). (d) Western blots of RAW 264.7 cells transfected with siScr or siJak2 for total JAK2 and GAPDH as a loading control. Full-length blots presented in Supplementary Fig. S9. (e) mRNA expression as fold change over vehicle-treated siScr of chemokines in response to vehicle or leptin (50 nM or 100 nM) in RAW 264.7 cells transfected with siScr or siJak2. Four independent experiments performed in triplicates; *siScr compared to siJak2, ^†vehicle treated siScr compared to leptin-treated siScr, § vehicle-treated siJAK2 compared to leptin-treated siJak2. (f) Western blots for phospho-STAT3 (pSTAT3), total STAT3 (tSTAT3), phospho-STAT5 (pSTAT5), total STAT5 (tSTAT5), and GAPDH as a loading control in RAW 264.7 cells transfected with siScr or siJak2. Full blots in Supplementary Fig. S9. Quantification of protein levels as a fold change over siScr for (g) STAT3 and (h) STAT5 (n = 7). All results are mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.