- Email: [email protected]
Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant
Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant Chang Zhang a, b, 1, Xianjun Luo a, 1, Da Zhang c, Bangli Deng a, Jingkai Tong d, Mo Zhang a, Liming Chen a, Hongquan Duan b, Wenyan Niu a, * a
Department of Immunomlogy, Key Laboratory of Immune Microenvironment and Disease (Ministry of Education), NHC Key Laboratory of Hormones and Development (Tianjin Medical University), Tianjin Key Laboratory of Metabolic Diseases, Tianjin Medical University Chu Hsien-I Memorial Hospital & Tianjin Institute of Endocrinology, Tianjin Medical University, Tianjin, 300070, China b School of Pharmacy, Research Center of Basic Medical Science, Tianjin Medical University, Tianjin, 300070, China c Department of Ultrasound, Tianjin Hospital, Tianjin, 300211, China d Department of Endocrinology, Tianjin First Center Hospital, Tianjin, 300192, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 October 2019 Accepted 24 October 2019 Available online xxx
Adipose tissue hypoxia occurs early in obesity and is associated with increased tissue macrophages and systemic inflammation that impacts muscle insulin responsiveness. We investigated how hypoxia interacted with adipocyte-macrophage crosstalk and inflammatory cytokine release, using co-culture and conditioned media (CM). Murine primary adipocytes from lean or obese mice were cultured under normoxic (21% O2) or hypoxic (1% O2) conditions. RAW264.7 macrophages were incubated under normoxic or hypoxic conditions with or without adipocyte conditioned media. Macrophage and adipocyte-macrophage co-culture CM were also collected. We found hypoxia did not elicit direct cytokine release from macrophages. However, adipocyte CM or adipocyte co-culture, synergistically stimulated TNFa and MCP-1 release from macrophages that was not further impacted by hypoxia. Exposure of muscle cells to elevated cytokines led to reduced insulin and muscle stress/inflammatory signaling. We conclude hypoxia or obesity induces release of inflammatory TNFa and MCP-1 from mice primary adipocytes but the two environmental conditions do not synergize to worsen macrophage signal transduction or insulin responsiveness. © 2019 Elsevier Inc. All rights reserved.
Keywords: Obesity Hypoxia Insulin signaling Conditioned medium Inflammation
1. Introduction Obesity is a “modern” disease with sedentary lifestyle and energy-dense diets having major roles in its development [1,2]. Obesity-associated chronic low-grade systemic inflammation is a consequence that can impact many metabolic diseases such as insulin resistance, atherosclerosis and fatty liver disease [3], by contributing to the onset of insulin resistance in adipose tissue, liver and skeletal muscle. An excess of adipose tissue is attributed to hypertrophy of adipocytes, meanwhile hypertrophic adipocytes endure less than adequate oxygen supply. Adipose tissue of genetic mice, diet-induced obese mice and obese human were found
* Corresponding author. Department of Immunology, Key Laboratory of Immune Microenvironment and Disease (Ministry of Education), Tianjin Medical University, Tianjin, 300070, China. E-mail address: [email protected] (W. Niu). 1 These authors contribute equally.
hypoxia [4,5]. Adipose inflammation is characterized with macrophage infiltration into the tissue [4,5]. Obese adipose tissue secretes fatty acids and inflammatory cytokines. Lipotoxicity and inflammation in other metabolic tissues are two main factors of obesity-induced the whole-body insulin resistance. Specifically, inflammatory cytokines released from obese adipose tissue are thought to contribute whole body insulin resistance by affecting insulin signaling in muscle and liver. Adipose tissue macrophages are a major source of inflammatory cytokines, such as TNFa and MCP-1 in the adipose tissue [6,7]. We previously showed that hypoxic adipocytes secrete inflammatory cytokines to evoke muscle cell insulin resistance [8]. However, the effect of factors released by macrophages under different conditions on insulin signaling is not clear. Skeletal muscles play an important role in the regulation of whole-body glucose metabolism. The majority of glucose is taken by skeletal muscle upon insulin stimulation after a meal. Skeletal muscle insulin resistance contributes to type 2 diabetes (T2DM)
https://doi.org/10.1016/j.bbrc.2019.10.162 0006-291X/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
2
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
and accompanies the development of T2DM. Upon binding to its receptor, insulin elicits a cascade of intracellular signals through IRS1, PI3K, Akt and AS160, leading GLUT4 translocation to the cell surface to transport more glucose into the cell [9]. Impaired insulin signal transduction leads to insulin resistance and at the molecular level, insulin resistance in skeletal muscle may manifest blunted sensitivity and responsiveness of the insulin signaling pathway that is associated with elevated serine phosphorylation of IRS1 via activation of the stress kinase c-Jun NH2-terminal kinase (JNK) and/ or the negative feed-back of p70S6-kinase (S6K) [10e12]. Systemic low-level inflammation caused by obesity induces insulin resistance in skeletal muscle due to secreted factors by hypoxic adipose tissue [8]. Here, we hypothesize that obese or hypoxic adipocytes affect adipose tissue macrophages to release inflammatory factors which may affect insulin signaling in muscle cells. 2. Material and methods 2.1. Materials Dulbecco’s modified Eagle’s medium (DMEM), horse serum (HS), and trypsin-EDTA were obtained from Invitrogen (Carlsbad, CA). Fetal bovine serum (FBS) was from BioInd (Israel). Dexamethasone, 3-isobutyl-1-methylxanthine, porcine insulin, collagenase, protease inhibitor cocktail, anti-a-actinin-1 and all other chemicals unless otherwise noted were purchased from Sigma Chemical (St. Louis, MO). Human insulin (Humulin R) was from Eli Lilly Canada (Toronto, ON, Canada). ELISA kits for TNFa, MCP-1 were obtained from R&D Systems, Inc. (USA). The Pierce BCA protein detection kit was purchased from Thermo Fisher Scientific (Rockford, IL). Anti-phosphoIRS1 (Ser636/639), anti-phospho-IRS1 (Ser307), anti-phospho-Akt (Ser473), anti-phospho-AS160 (Thr642), anti-phospho-JNK (Thr183/Tyr185) and anti-phospho-S6K (Thr389) were from Cell Signaling Technology (Danvers, MA). Horseradish peroxidase (HRP)bound goat anti-mouse, goat anti-rabbit IgG antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA). The Immobilon Western Chemiluminescent HRP Substrate was purchased from Millipore (Billerica, MA).
supplemented with 10% FBS (v/v). Wild type C2C12 myoblasts were generated and differentiated into myotubes as described [13]. The cells were used in 5 days after differentiation. 3T3-L1 fibroblasts were generated and differentiated into adipocytes as described [8]. RAW 264.7 murine macrophages were passaged as described [8]. 2.4. Conditioned media preparation Conditioned media (CM) from primary adipocytes: Adipocytes from lean and obese mice were incubated under 21% O2 (normoxic condition) or 1% O2 (hypoxic condition), respectively. Culture media were collected after 24 h incubation following centrifugation at 1000 rpm for 5 min to remove any cell debris. Cells on the top layer, the lower solution was CM. CMs from lean or obese adipocytes under normoxic or hypoxic condition were named as L-NF, L-HF, OeNF, OeHF. CMs were used freshly to muscle cells or stored at 80 C for cytokine analysis. CM from macroghages: RAW264.7 macrophages were incubated for 16 h with different adipocyte-derived CMs (L-NF, L-HF, OeNF, OeHF), then washed twice with Hanks following addition of DMEM with 5% HS. Culture media were collected after 8 h and centrifuged at 1000 rpm for 5 min. The supernatants were CM from macroghages, named as LeNFeM, LeHFeM, OeNF-M, OeHF-M. Macroghage-derived CMs were used freshly on muscle cells or stored at 80 C for cytokine analysis. CM from co-cultured cells: CMs from 3T3-L1 adipocytes, RAW264.7 macrophages and co-cultured of these 2 types cells cultured under normoxia or hypoxia were collected as NF, HF, NM, HM, NF/M and HF/MM. All CMs from adipocytes were collected at day 8 after initial exposure to a differentiation cocktail. For normoxia and hypoxia treatment, the medium of cells was replaced with DMEM containing 5% HS CMs were collected after 24 h and centrifuged at 1000 rpm for 5 min. The supernatant was used freshly or stored at 80 C for cytokine analysis. Muscle cells were incubated with different CMs for 16 h. Muscle cells regular media was used as control. 2.5. Cytokine measurement
2.2. Animals Twenty-week old male wild-type C57BL/6 mice were housed in vented cages in a temperature-controlled room with a 12-h light/ dark cycle and free access to food and water. This research was approved by the Tianjin Medical University Animal Care and Use Committee under the guidelines of the Chinese Academy of Sciences. The mice were randomly divided into control diet (n ¼ 10) and HFD (n ¼ 10) groups, then fed with chow diet and fat-enriched diet consisting of 60% kcal% fat (Research Diets, New Brunswick, NJ) for 16 weeks, respectively. 2.3. Cell culture Primary adipocytes isolated from C57BL/6 mice epididymal fat. In a laminar flow hood, excised adipose tissue was transferred to a sterile 100 mm dish, wash twice in HBSS, and cut into approximately 1-mm2 pieces using sterile scissors and thumb forceps. Pipette sterile 0.1% collagenase solution to each tube 2 ml/g and incubated at 37 C, shacked the dish every 5 min for 30 min. The tissue slurry was transferred to 50 ml centrifuge tubes. Addition of DMEM containing 4.5 g/l glucose supplemented with 10% FBS (v/v) 20ml/1 ml collagenase solution was followed by centrifugation for 15 min at 2000g. Afterwards, the fatty layer was transferred into a new 100 mm dish then maintained in a humidified atmosphere of air and 5% CO2 at 37 C with DMEM containing 4.5 g/l glucose
The concentrations of TNFa and MCP-1 in different CMs were measured with ELISA kits, respectively. 2.6. Cell lysates and immunoblotting Cell lysates and immunoblotting of C2C12 wild type myotubes were taken as described [13]. 2.7. Statistical analysis Results were represented as means ± S.E. Statistical analyses were carried out using 3.0 software (San Diego, CA). Two groups were compared using Student’s unpaired t-test and more than two groups were compared using analysis of variance (ANOVA) with Tukey’s post hoc analysis. P < 0.05 is considered statistically significant. 3. Results 3.1. CM from adipocytes induced macrophages to release inflammatory cytokines Macrophages in obese adipose tissue are the main sources of inflammatory cytokines [4e7]. Obese adipose tissue has hypoxia and adipocytes under hypoxia condition released more
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
inflammatory cytokines which induced insulin resistance in skeletal muscle cells [8]. Here, we detected the effect of CMs from adipocytes of lean or obese mice on macrophages by measuring TNFa and MCP1 released from macrophages. RAW 264.7 macrophages treated with L-NF, L-HF, OeNF or OeHF. The CMs were collected named as LeNFeM, LeHFeM, OeNF-M and OeHF-M, respectively. All these macrophages CMs had significantly higher levels of TNFa and MCP-1 than the CM from macrophages treated with regular adipocyte growth medium (RM). The fold increases of TNFa were 3.9 ± 0.3fold, 11.3 ± 0.2-fold, 11.7 ± 0.4-fold and 11.0 ± 0.5-fold, respectively (p < 0.001 vs RM). The fold increases of MCP-1 were 5.0 ± 0.3-fold, 8.6 ± 0.3-fold, 9.7 ± 0.4-fold, 9.4 ± 0.5-fold (p < 0.01, p < 0.001, p < 0.001, p < 0.001, vs RM), respectively (Fig. 1A and B). In addition, levels of TNFa and MCP-1 in LeHFeM, OeNF-M and OeHF-M are significantly higher than LeNFeM. The fold increases of TNFa were 2.8 ± 0.4-fold, 2.9 ± 0.3-fold, 2.7 ± 0.4-fold, respectively (p < 0.001 vs LeNFeM). The fold increases of MCP-1 were 1.7 ± 0.3-fold, 1.9 ± 0.3, 1.8 ± 0.5, respectively (p < 0.001 all vs LeNFeM, Fig. 1A and B).
3.2. CM from macrophages-treated with adipocyte-derived CM induced insulin resistance by activating stress and inflammatory signals in skeletal muscle cells We next detected if the increased secretion of inflammatory
3
cytokines from macrophages treated with adipocyte-derived CM affect insulin signal in skeletal muscle cells. Macrophage CMs (LeNFeM, LeHFeM, OeNF-M, OeHF-M) were applied to C2C12 myotubes for 16 h following treatment with 100 nM insulin in the last 10 min. The phosphorylation of insulin signaling molecules Akt and AS160 (also known as TBC1D4) were detected. All these Macrophage CMs significantly impaired insulin-stimulated Akt Ser473 phosphorylation by 47 ± 4%, 78 ± 9%, 79 ± 13, 85% ± 11% reduction, respectively (p < 0.001 vs insulin in RM) and AS160 Thr642 phosphorylation by 34 ± 6%, 47 ± 10%, 49 ± 5%, 51 ± 7% reduction, respectively (p < 0.05 vs insulin in RM, Fig. 2A and B). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2A and B). This is consistent with the marked increased inflammatory cytokines in macrophage-derived CMs. IRS-1 serine phosphorylation reduces its insulin-dependent tyrosine phosphorylation and downstream signaling leading to insulin resistance in C2C12 myotubes [14,15]. We examined phosphorylation of IRS1 Ser307 and Ser636/639 in myotube treated with macrophages CMs. All macrophage CMs significantly elevated both basal IRS1 pSer307 phosphorylation by 1.32 ± 0.05-fold, 1.79 ± 0.08-fold, 1.83 ± 0.05-fold, 1.81 ± 0.07-fold, respectively (p < 0.05 vs basal in RM) and insulin-stimulated phosphorylation by 1.26 ± 004-fold, 1.71 ± 0.07-fold, 1.77 ± 0.05-fold, 164 ± 0.09fold, respectively (p < 0.05 vs insulin in RM, Fig. 2C). LeNFeM, LeHFeM, OeNF-M and OeHF-M significantly elevated both basal IRS1 pSer636/639 phosphorylation by 1.31 ± 0.03-fold, 1.65 ± 0.1fold, 1.69 ± 0.05-fold, 1.68 ± 0.07-fold, respectively (p < 0.05, p < 0.05, p < 0.001, p < 0.001 vs basal in RM) and insulin-stimulated phosphorylation by 1.39 ± 0.05-fold, 1.63 ± 0.11-fold, 1.66 ± 0.07fold, 1.69 ± 0.1-fold, respectively (p < 0.05 vs insulin in RM, Fig. 2D). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2C and D). JNK and S6K can phosphorylate IRS1 Ser307 and Ser636/639, respectively [16,17], which mediate the inhibitory effects of TNFa on insulin signal. Indeed, we found phosphorylation of JNK and S6K was increased in myotubes treated with all macrophage CMs. LeNFeM, LeHFeM, OeNF-M and OeHF-M significantly elevated both basal JNK pThr183/185 phosphorylation by 1.33 ± 0.07-fold, 1.77 ± 0.12-fold, 1.85 ± 0.09-fold, 1.83 ± 0.11-fold, respectively (p < 0.05 vs basal in RM) and insulin-stimulated JNK pThr183/185 phosphorylation by 1.21 ± 0.05-fold, 1.67 ± 0.1-fold, 1.74 ± 0.1-fold, 1.81 ± 0.13-fold, respectively (p < 0.05 vs basal in RM) (Fig. 2E). Similarly, all macrophage CMs significantly elevated both basal S6K pThr389 phosphorylation by 1.47 ± 0.06-fold, 1.93 ± 0.14-fold, 1.86 ± 0.11-fold, 1.83 ± 0.12-fold, respectively (p < 0.01, p < 0.001, p < 0.001, p < 0.001 vs basal in RM) and insulin-stimulated S6K pThr389 phosphorylation by 1.46 ± 0.07-fold, 2.17 ± 0.14-fold, 2.23 ± 0.13-fold, 2.27 ± 0.15-fold, respectively (p < 0.05 vs basal in RM, Fig. 2F). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2E and F). The trends of phospho-JNK and phospho-S6K are consistent with the heightened phosphorylation of IRS1 Ser307 and Ser636/639 (Fig. 2C and D). 3.3. CM from co-cultured adipocytes and macrophages secretes more inflammatory cytokines
Fig. 1. Adipocytes induce inflammatory response in macrophages. Concentrations of TNFa (A) and MCP-1 (B) in conditioned media from macrophages were measured by ELISA, respectively. Results are means ± S.E. of 4 independent experiments. ***p < 0.001.
Macrophages within the hypoxic obese adipose tissue are exposed to adipocytes-derived adipokines and cytokines [18,19]. We wondered how hypoxia specifically affected macrophages and how the two types of cells affected each other. We cultured adipocytes and macrophages under normoxic or hypoxic condition, respectively. Meanwhile, we co-cultured adipocytes and macrophages in a transwell system under normoxic or hypoxic condition, respectively. Consistent with our previous results [8], hypoxia elevated the levels of TNFa (3.84 ± 0.4-fold, p < 0.001) and MCP-1
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
4
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
5
(5.53 ± 0.5-fold, p < 0.05) in HF compared with NF (Fig. 3 and B, vs NF), but did not affect the amounts of TNFa and MCP-1 in NM and HM cultures (Fig. 3A and B). Interestingly, co-cultured adipocytes and macrophages markedly increased the releases of TNFa and MCP-1 in both normoxia and hypoxia conditions. NF/M had 11.3 ± 0.6-fold of TNFa and 9.7 ± 0.4-fold of MCP-1 compared to NF (p < 0.001). HF/M had 11.54 ± 0.5-fold of TNFa and 10.6 ± 0.6-fold of MCP-1 compared to NF (p < 0.001, Fig. 3A and B). Moreover, the levels of TNFa and MCP-1 in NF/M and HF/M were higher than that in HF. The fold increases for TNFa and MCP-1 were 2.9 ± 0.5-fold, and 1.8 ± 0.5-fold (p < 0.001, NF/M vs HF, Fig. 3A and B), respectively. The fold increases for TNFa and MCP-1 were 3.0 ± 0.6-fold and 2.0 ± 0.6-fold (p < 0.001, HF/M vs HF, Fig. 3A and B), respectively. 3.4. CM from co-cultured adipocytes and macrophages activate stress and inflammatory signals and evoke insulin resistance in skeletal muscle cells We detected the effects of CMs from co-cultured adipocytes and macrophages on Akt and AS160 in muscle cells. Compared to Akt phosphorylation in control RM group, NF, NM and HM did not change basal or insulin-stimulated Akt and AS160 phosphorylation (Fig. 4A and B). Similar to our previous report, HF did not change basal levels, but blunted insulin-stimulated Akt and AS160 phosphorylation by 43 ± 5% (p < 0.01 vs insulin in RM, Fig. 4A) and 26 ± 3% (p < 0.001 vs insulin in RM, Fig. 4B). Moreover, both NF/M and HF/M did not affect basal, but markedly reduced insulinstimulated Akt phosphorylation by 59 ± 4% and 61 ± 7% (p < 0.001 vs insulin in RM, Fig. 4A) and AS160 phosphorylation by 41 ± 5%, 42 ± 5% (p < 0.001 vs insulin in RM, Fig. 4B). In addition, NF/M and HF/M induced more severe insulin resistance of Akt and AS160 than HF (p < 0.001 vs insulin in HF, Fig. 4A and B). We next examined IRS1 S307 and S636/639 phosphorylation. NF, NM and HM had no effect on IRS1 serine phosphorylations (Fig. 4C and D). However, HF, NF/M and HF/M raised basal phosphorylations by 1.76 ± 0.07-fold, 1.99 ± 0.11-fold, 1.95 ± 0.09-fold for S307 (p < 0.05 vs basal in RM, Fig. 4C), 1.89 ± 0.13-fold, 3.16 ± 0.11-fold, 3.95 ± 0.19-fold for S636/639 (p < 0.01, p < 0.001, p < 0.001 vs basal in RM, Fig. 4D). Moreover, NF/M and HF/M were more effective than HF on IRS1 S307 (p < 0.05 vs. HF, Fig. 4C) and S636/639 (p < 0.01, p < 0.001 vs HF, Fig. 4D). The patterns of JNK and S6K phosphorylations were consistent with IRS1 serine phosphorylations. NF, NM and HM had no effect on JNK and S6K phosphorylations (Fig. 4E and F). HF, NF/M and HF/M raised JNK phosphorylation by 1.87 ± 0.09-fold, 3.44 ± 0.14-fold, 3.49 ± 0.12fold (p < 0.01, p < 0.001, p < 0.001 vs basal in RM, Fig. 4E) and S6K phosphorylation by 2.45 ± 0.15-fold, 3.21 ± 0.1-fold, 3.65 ± 0.13fold (p < 0.001 vs basal in RM, Fig. 4F). In addition, NF/M and HF/ M were more effective than HF on JNK (p < 0.001 vs. HF, Fig. 4E) and S6K (p < 0.05 vs HF, Fig. 4F). 4. Discussion Massive expansion of adipose tissue is the hallmark of obesity and excess of adipose tissue is attributed to hypertrophy and hyperplasia of adipocytes. Adipose tissues of obese individuals manifest insulin resistance associated with macrophages infiltration and local inflammation. Recent studies shows that hypoxia in
Fig. 3. Co-cultured of adipocytes with macrophages exacerbate inflammatory response than each cell type in both normoxic and hypoxic condition. TNFa (A) and MCP-1 (B) in conditioned media of 3T3-L1 adipocytes, RAW264.7 macrophages and cocultured of these 2 types cells were measured by ELISA, respectively. Results are means ± S.E. of 4 independent experiments. ***p < 0.001.
adipose tissues is an early phenomenon in obesity. We previously reported that hypoxia-treated 3T3-L1 adipocytes secreted increased levels of inflammatory cytokines. CM from hypoxic adipocytes induced insulin resistance of muscle cells. MCP-1 played a role in this mechanism [8]. Macrophages infiltrated in adipose tissue are the main sources of inflammatory cytokines of obese adipose tissue. The effects of hypoxia on macrophages and on the interaction of adipocytes and macrophages are not clear. The present study was undertaken to illustrate the effect of hypoxia on the interaction between adipocytes and macrophages. The effect of the factors from macrophages on insulin signal in muscle cells was also explored. To this end, we prepared CM from macrophages incubated with conditioned medium from normoxic or hypoxic adipocytes, from cultured macrophages alone or from macrophageadipocyte co-cultures under normoxic or hypoxic conditions,
Fig. 2. Conditioned media from RAW 264.7 macrophages reduces insulin-stimulated Akt and AS160 phosphorylation and increases IRS1 serine, JNK, S6K phosphorylation in C2C12 myotubes. C2C12 myotubes were treated with conditioned media from macrophages for 16 h, followed by 100 nM insulin treatment in the last 10 min prior to cell lysis and immunoblotting for phospho-Akt S473 (A), phospho-AS160 T642 (B), phospho-IRS1 Ser307 (C), phospho-Ser636/639 (D), phospho-JNK Thr183/Tyr185 (E) and phospho-S6K Thr389 (F). Shown are representative blots. The densitometry mean ± S.E. from 4 independent experiments is plotted as the fold change relative to RM group or as indicated. *p < 0.05, **p < 0.01, ***p < 0.001.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
6
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 4. Conditioned media from co-cultured adipocytes and macrophages exacerbate muscle cells insulin resistance than CMs from each cell type. C2C12 myotubes were treated with indicated conditioned medium for 16 h, followed by 100 nM insulin treatment in the last 10 min prior to cell lysis and immunoblotting for phospho-Akt S473 (A), phosphoAS160 T642 (B), phospho-IRS1 Ser307 (C), phospho-Ser636/639 (D), phospho-JNK Thr183/Tyr185 (E) and phospho-S6K Thr389 (F). Shown are representative blots. The densitometry mean ± S.E. from 4 independent experiments is plotted as the fold change relative to RM group or as indicated. *p < 0.05, **p < 0.01, ***p < 0.001.
respectively. Then the effects of CM from macrophages on insulin signal in muscle cells were determined. Macrophages inside the obese adipose tissue contribute to the release of inflammatory cytokines. Here we studied the direct and indirect effects of hypoxia on the release of cytokines from macrophages. Obese adipocytes and hypoxia-treated adipocytes released more TNFa and MCP-1 (data not shown). CMs from
adipocytes induced macrophages to secrete TNFa and MCP-1. Moreover, CMs from hypoxia-treated lean mice adipocytes, normoxia- and hypoxia-treated obese mice adipocytes (L-HF, OeNF, OeHF) stimulated macrophages to produce more TNFa and MCP1 than CM from normoxia-treated lean mice adipocytes (L-NF). Interestingly, although the levels of TNFa and MCP-1 in LeHFeM, OeNF-M, OeHF-M are different, the ability of these CM to induce
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
macrophages to produce TNFa and MCP-1 were the same. These data implicate that hypoxic adipocytes can release factors to activate macrophages. TNFa plays a role, but other factors in the CM may also be involved. To investigate the direct effect of macrophages on muscle cells insulin signal, we incubated muscle cells with CM from macrophages treated with different adipocyte CM. Consistent with the levels of TNFa and MCP-1 in the adipocytes CMs, LeNFeM, LeHFeM, OeNF-M and OeHF-M all lowered insulin-stimulated phosphorylation of Akt and AS160. Moreover, phosphorylation of IRS1 S307 and S636/639 were elevated, which are associated with impaired propagation of insulin signals from IRS1 to Akt and AS160 [15]. Phosphorylation of IRS1 S307 and S636/639 are regulated by JNK and S6K, respectively. Accordingly, macrophage CMs activated JNK and S6K. The trends of the phosphorylations of Akt, AS160, IRS1 serine, JNK and S6K correlated with the changes in the levels of TNFa and MCP-1. We previously reported that hypoxia-treated 3T3L1 adipocytes diminished insulin-stimulated phosphorylation of Akt and AS160 and activated JNK and S6K in muscle cells. MCP-1 played a role in the mechanism. In addition, TNFa activated JNK and S6K in muscle tissue [17]. MCP-1 and TNFa may involve in the activation of JNK and S6K by adipocyte-macrophage CMs. Macrophages infiltrate obese white adipose tissue. Hypoxia affects the release of factors from adipocytes, which further affect macrophages. We explored if hypoxia has direct effect on macrophages. We also tested the interaction between adipocytes and macrophages. Our data showed that hypoxia had no effect on macrophages to produce TNFa and MCP-1. Obese adipose tissues expend faster than angiogenesis and adipocytes in obese white adipose tissue are hypertrophic. Thus, oxygen has a lower probability to reach the inside of adipocytes in obese adipose tissue leading to hypoxia responses. On the other hand, macrophages escape from blood vessel to adipose tissue. Macrophages are close to blood vessel and the adipocytes are not hypertrophic in these areas. Therefore, hypoxia may not be the reason for macrophages to secrete inflammatory cytokines. However, co-cultured adipocytes and macrophages secreted much more TNFa and MCP-1 than each cell type, which confirmed the effect of hypoxic adipocytes on macrophages. Interestingly, co-culture adipocytes and macrophages under nomoxic condition also induced TNFa and MCP-1 secretion. One caveat is that the ratio of macrophages to adipocytes is higher than that in vivo. Too many macrophages may crosstalk with adipocytes and in turn to induce the release of TNFa and MCP-1 from macrophages. Some details may require further study. Accordingly, insulin-stimulated Akt and AS160 phosphorylation were reduced by HF, NF/M and HF/M. The phosphorylation of IRS1 S307 and S36/639 were increased. Again, HF, NF/M and HF/M activated JNK and S6K as the same trend of IRS1 S307 and S36/639, respectively. MCP-1 and TNFa may involve in the mechanism. In summary, hypoxia or obese adipocytes and co-culture with adipocytes induce macrophages to secrete TNFa and MCP-1. CMs from these macrophages reduce insulin signal in muscle cells. This may be related to the activation of JNK and S6K. TNFa and MCP-1 may play a role in this mechanism.
Author contributions CZ and XL conducted the majority of the data with equal contributions. DZ, BD, JT and MZ analysis the data and made the figures. LC and HD contributed to discussion and reviewed the manuscript. WN contributed to the study design, reviewed the data and wrote the paper.
7
Declaration of competing interest All authors declare no conflicts of interest. Acknowledgments This work was supported by grants from National Natural Science Foundation of China (#81670731, #81870547, #81170740 and #81161120545), the Tianjin Municipal Science and Technology Commission (#15JCZDJC35500) and the Tianjin Health and Family Planning Commission (#15KG102) to W Niu. Deng Bangli and Zhang Mo were supported by Scientific Research Funding of Tianjin Medical University Chu Hsien-I Memorial Hospital (#2018RC03, #2014DX07). Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2019.10.162. References [1] B.A. Swinburn, I. Caterson, J.C. Seidell, W.P. James, Diet, nutrition and the prevention of excess weight gain and obesity, Public Health Nutr. 7 (2004) 123e146. [2] M. Krekoukia, G.P. Nassis, G. Psarra, K. Skenderi, G.P. Chrousos, L.S. Sidossis, Elevated total and central adiposity and low physical activity are associated with insulin resistance in children, Metabolism 56 (2007) 206e213. [3] G.S. Hotamisligil, Inflammation and metabolic disorders, Nature 444 (2006) 860e867. [4] R.M. Pirzgalska, A.I. Domingos, Macrophages in obesity, Cell. Immunol. 330 (2018) 183e187. [5] J.C. McNelis, J.M. Olefsky, Macrophages, immunity, and metabolic disease, Immunity 41 (2014) 36e48. [6] J.M. Olefsky, C.K. Glass, Macrophages, inflammation, and insulin resistance, Annu. Rev. Physiol. 72 (2010) 219e246. [7] R.W. O’Rourke, A.E. White, M.D. Metcalf, A.S. Olivas, P. Mitra, W.G. Larison, E.C. Cheang, O. Varlamov, C.L. Corless, C.T. Roberts Jr., D.L. Marks, Hypoxiainduced inflammatory cytokine secretion in human adipose tissue stromovascular cells, Diabetologia 54 (2011) 1480e1490. [8] J. Yu, L. Shi, H. Wang, P.J. Bilan, Z. Yao, M.C. Samaan, Q. He, A. Klip, W. Niu, Conditioned medium from hypoxia-treated adipocytes renders muscle cells insulin resistant, Eur. J. Cell Biol. 90 (2011) 1000e1015. [9] F.S. Thong, C.B. Dugani, A. Klip, Turning signals on and off: GLUT4 traffic in the insulin-signaling highway, Physiology 20 (2005) 271e284. [10] S.H. Um, D. D’Alessio, G. Thomas, Nutrient overload, insulin resistance, and ribosomal protein S6 kinase 1, S6K1, Cell Metabol. 3 (2006) 393e402. [11] K.D. Copps, M.F. White, Regulation of insulin sensitivity by serine/threonine phosphorylation of insulin receptor substrate proteins IRS1 and IRS2, Diabetologia 55 (2012) 2565e2582. [12] Y.H. Lee, J. Giraud, R.J. Davis, M.F. White, c-Jun N-terminal kinase (JNK) mediates feedback inhibition of the insulin signaling cascade, J. Biol. Chem. 278 (2003) 2896e2902. [13] W. Niu, P.J. Bilan, S. Ishikura, J.D. Schertzer, A. Contreras-Ferrat, Z. Fu, J. Liu, S. Boguslavsky, K.P. Foley, Z. Liu, J. Li, G. Chu, T. Panakkezhum, G.D. Lopaschuk, S. Lavandero, Z. Yao, A. Klip, Contraction-related stimuli regulate GLUT4 traffic in C2C12-GLUT4myc skeletal muscle cells, Am. J. Physiol. Endocrinol. Metab. 298 (2010) E1058eE1071. [14] J.H. Lim, J.I. Lee, Y.H. Suh, W. Kim, J.H. Song, M.H. Jung, Mitochondrial dysfunction induces aberrant insulin signalling and glucose utilisation in murine C2C12 myotube cells, Diabetologia 49 (2006) 1924e1936. [15] J. Boucher, A. Kleinridders, C.R. Kahn, Insulin receptor signaling in normal and insulin-resistant states, Cold Spring Harb. Perspect. Biol. 6 (2014). [16] V. Aguirre, T. Uchida, L. Yenush, R. Davis, M.F. White, The c-Jun NH(2)terminal kinase promotes insulin resistance during association with insulin receptor substrate-1 and phosphorylation of Ser(307), J. Biol. Chem. 275 (2000) 9047e9054. [17] J. Zhang, Z. Gao, J. Yin, M.J. Quon, J. Ye, S6K directly phosphorylates IRS-1 on Ser-270 to promote insulin resistance in response to TNF-(alpha) signaling through IKK2, J. Biol. Chem. 283 (2008) 35375e35382. [18] S.U. Amano, J.L. Cohen, P. Vangala, M. Tencerova, S.M. Nicoloro, J.C. Yawe, Y. Shen, M.P. Czech, M. Aouadi, Local proliferation of macrophages contributes to obesityassociated adipose tissue inflammation, Cell Metabol. 19 (2014) 162e171. [19] Y. Kawano, J. Nakae, N. Watanabe, T. Kikuchi, S. Tateya, Y. Tamori, M. Kaneko, T. Abe, M. Onodera, H. Itoh, Colonic pro-inflammatory macrophages cause insulin resistance in an intestinal ccl2/ccr2-dependent manner, Cell Metabol. 24 (2016) 295e310.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant Chang Zhang a, b, 1, Xianjun Luo a, 1, Da Zhang c, Bangli Deng a, Jingkai Tong d, Mo Zhang a, Liming Chen a, Hongquan Duan b, Wenyan Niu a, * a
Department of Immunomlogy, Key Laboratory of Immune Microenvironment and Disease (Ministry of Education), NHC Key Laboratory of Hormones and Development (Tianjin Medical University), Tianjin Key Laboratory of Metabolic Diseases, Tianjin Medical University Chu Hsien-I Memorial Hospital & Tianjin Institute of Endocrinology, Tianjin Medical University, Tianjin, 300070, China b School of Pharmacy, Research Center of Basic Medical Science, Tianjin Medical University, Tianjin, 300070, China c Department of Ultrasound, Tianjin Hospital, Tianjin, 300211, China d Department of Endocrinology, Tianjin First Center Hospital, Tianjin, 300192, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 October 2019 Accepted 24 October 2019 Available online xxx
Adipose tissue hypoxia occurs early in obesity and is associated with increased tissue macrophages and systemic inflammation that impacts muscle insulin responsiveness. We investigated how hypoxia interacted with adipocyte-macrophage crosstalk and inflammatory cytokine release, using co-culture and conditioned media (CM). Murine primary adipocytes from lean or obese mice were cultured under normoxic (21% O2) or hypoxic (1% O2) conditions. RAW264.7 macrophages were incubated under normoxic or hypoxic conditions with or without adipocyte conditioned media. Macrophage and adipocyte-macrophage co-culture CM were also collected. We found hypoxia did not elicit direct cytokine release from macrophages. However, adipocyte CM or adipocyte co-culture, synergistically stimulated TNFa and MCP-1 release from macrophages that was not further impacted by hypoxia. Exposure of muscle cells to elevated cytokines led to reduced insulin and muscle stress/inflammatory signaling. We conclude hypoxia or obesity induces release of inflammatory TNFa and MCP-1 from mice primary adipocytes but the two environmental conditions do not synergize to worsen macrophage signal transduction or insulin responsiveness. © 2019 Elsevier Inc. All rights reserved.
Keywords: Obesity Hypoxia Insulin signaling Conditioned medium Inflammation
1. Introduction Obesity is a “modern” disease with sedentary lifestyle and energy-dense diets having major roles in its development [1,2]. Obesity-associated chronic low-grade systemic inflammation is a consequence that can impact many metabolic diseases such as insulin resistance, atherosclerosis and fatty liver disease [3], by contributing to the onset of insulin resistance in adipose tissue, liver and skeletal muscle. An excess of adipose tissue is attributed to hypertrophy of adipocytes, meanwhile hypertrophic adipocytes endure less than adequate oxygen supply. Adipose tissue of genetic mice, diet-induced obese mice and obese human were found
* Corresponding author. Department of Immunology, Key Laboratory of Immune Microenvironment and Disease (Ministry of Education), Tianjin Medical University, Tianjin, 300070, China. E-mail address: [email protected] (W. Niu). 1 These authors contribute equally.
hypoxia [4,5]. Adipose inflammation is characterized with macrophage infiltration into the tissue [4,5]. Obese adipose tissue secretes fatty acids and inflammatory cytokines. Lipotoxicity and inflammation in other metabolic tissues are two main factors of obesity-induced the whole-body insulin resistance. Specifically, inflammatory cytokines released from obese adipose tissue are thought to contribute whole body insulin resistance by affecting insulin signaling in muscle and liver. Adipose tissue macrophages are a major source of inflammatory cytokines, such as TNFa and MCP-1 in the adipose tissue [6,7]. We previously showed that hypoxic adipocytes secrete inflammatory cytokines to evoke muscle cell insulin resistance [8]. However, the effect of factors released by macrophages under different conditions on insulin signaling is not clear. Skeletal muscles play an important role in the regulation of whole-body glucose metabolism. The majority of glucose is taken by skeletal muscle upon insulin stimulation after a meal. Skeletal muscle insulin resistance contributes to type 2 diabetes (T2DM)
https://doi.org/10.1016/j.bbrc.2019.10.162 0006-291X/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
2
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
and accompanies the development of T2DM. Upon binding to its receptor, insulin elicits a cascade of intracellular signals through IRS1, PI3K, Akt and AS160, leading GLUT4 translocation to the cell surface to transport more glucose into the cell [9]. Impaired insulin signal transduction leads to insulin resistance and at the molecular level, insulin resistance in skeletal muscle may manifest blunted sensitivity and responsiveness of the insulin signaling pathway that is associated with elevated serine phosphorylation of IRS1 via activation of the stress kinase c-Jun NH2-terminal kinase (JNK) and/ or the negative feed-back of p70S6-kinase (S6K) [10e12]. Systemic low-level inflammation caused by obesity induces insulin resistance in skeletal muscle due to secreted factors by hypoxic adipose tissue [8]. Here, we hypothesize that obese or hypoxic adipocytes affect adipose tissue macrophages to release inflammatory factors which may affect insulin signaling in muscle cells. 2. Material and methods 2.1. Materials Dulbecco’s modified Eagle’s medium (DMEM), horse serum (HS), and trypsin-EDTA were obtained from Invitrogen (Carlsbad, CA). Fetal bovine serum (FBS) was from BioInd (Israel). Dexamethasone, 3-isobutyl-1-methylxanthine, porcine insulin, collagenase, protease inhibitor cocktail, anti-a-actinin-1 and all other chemicals unless otherwise noted were purchased from Sigma Chemical (St. Louis, MO). Human insulin (Humulin R) was from Eli Lilly Canada (Toronto, ON, Canada). ELISA kits for TNFa, MCP-1 were obtained from R&D Systems, Inc. (USA). The Pierce BCA protein detection kit was purchased from Thermo Fisher Scientific (Rockford, IL). Anti-phosphoIRS1 (Ser636/639), anti-phospho-IRS1 (Ser307), anti-phospho-Akt (Ser473), anti-phospho-AS160 (Thr642), anti-phospho-JNK (Thr183/Tyr185) and anti-phospho-S6K (Thr389) were from Cell Signaling Technology (Danvers, MA). Horseradish peroxidase (HRP)bound goat anti-mouse, goat anti-rabbit IgG antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA). The Immobilon Western Chemiluminescent HRP Substrate was purchased from Millipore (Billerica, MA).
supplemented with 10% FBS (v/v). Wild type C2C12 myoblasts were generated and differentiated into myotubes as described [13]. The cells were used in 5 days after differentiation. 3T3-L1 fibroblasts were generated and differentiated into adipocytes as described [8]. RAW 264.7 murine macrophages were passaged as described [8]. 2.4. Conditioned media preparation Conditioned media (CM) from primary adipocytes: Adipocytes from lean and obese mice were incubated under 21% O2 (normoxic condition) or 1% O2 (hypoxic condition), respectively. Culture media were collected after 24 h incubation following centrifugation at 1000 rpm for 5 min to remove any cell debris. Cells on the top layer, the lower solution was CM. CMs from lean or obese adipocytes under normoxic or hypoxic condition were named as L-NF, L-HF, OeNF, OeHF. CMs were used freshly to muscle cells or stored at 80 C for cytokine analysis. CM from macroghages: RAW264.7 macrophages were incubated for 16 h with different adipocyte-derived CMs (L-NF, L-HF, OeNF, OeHF), then washed twice with Hanks following addition of DMEM with 5% HS. Culture media were collected after 8 h and centrifuged at 1000 rpm for 5 min. The supernatants were CM from macroghages, named as LeNFeM, LeHFeM, OeNF-M, OeHF-M. Macroghage-derived CMs were used freshly on muscle cells or stored at 80 C for cytokine analysis. CM from co-cultured cells: CMs from 3T3-L1 adipocytes, RAW264.7 macrophages and co-cultured of these 2 types cells cultured under normoxia or hypoxia were collected as NF, HF, NM, HM, NF/M and HF/MM. All CMs from adipocytes were collected at day 8 after initial exposure to a differentiation cocktail. For normoxia and hypoxia treatment, the medium of cells was replaced with DMEM containing 5% HS CMs were collected after 24 h and centrifuged at 1000 rpm for 5 min. The supernatant was used freshly or stored at 80 C for cytokine analysis. Muscle cells were incubated with different CMs for 16 h. Muscle cells regular media was used as control. 2.5. Cytokine measurement
2.2. Animals Twenty-week old male wild-type C57BL/6 mice were housed in vented cages in a temperature-controlled room with a 12-h light/ dark cycle and free access to food and water. This research was approved by the Tianjin Medical University Animal Care and Use Committee under the guidelines of the Chinese Academy of Sciences. The mice were randomly divided into control diet (n ¼ 10) and HFD (n ¼ 10) groups, then fed with chow diet and fat-enriched diet consisting of 60% kcal% fat (Research Diets, New Brunswick, NJ) for 16 weeks, respectively. 2.3. Cell culture Primary adipocytes isolated from C57BL/6 mice epididymal fat. In a laminar flow hood, excised adipose tissue was transferred to a sterile 100 mm dish, wash twice in HBSS, and cut into approximately 1-mm2 pieces using sterile scissors and thumb forceps. Pipette sterile 0.1% collagenase solution to each tube 2 ml/g and incubated at 37 C, shacked the dish every 5 min for 30 min. The tissue slurry was transferred to 50 ml centrifuge tubes. Addition of DMEM containing 4.5 g/l glucose supplemented with 10% FBS (v/v) 20ml/1 ml collagenase solution was followed by centrifugation for 15 min at 2000g. Afterwards, the fatty layer was transferred into a new 100 mm dish then maintained in a humidified atmosphere of air and 5% CO2 at 37 C with DMEM containing 4.5 g/l glucose
The concentrations of TNFa and MCP-1 in different CMs were measured with ELISA kits, respectively. 2.6. Cell lysates and immunoblotting Cell lysates and immunoblotting of C2C12 wild type myotubes were taken as described [13]. 2.7. Statistical analysis Results were represented as means ± S.E. Statistical analyses were carried out using 3.0 software (San Diego, CA). Two groups were compared using Student’s unpaired t-test and more than two groups were compared using analysis of variance (ANOVA) with Tukey’s post hoc analysis. P < 0.05 is considered statistically significant. 3. Results 3.1. CM from adipocytes induced macrophages to release inflammatory cytokines Macrophages in obese adipose tissue are the main sources of inflammatory cytokines [4e7]. Obese adipose tissue has hypoxia and adipocytes under hypoxia condition released more
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
inflammatory cytokines which induced insulin resistance in skeletal muscle cells [8]. Here, we detected the effect of CMs from adipocytes of lean or obese mice on macrophages by measuring TNFa and MCP1 released from macrophages. RAW 264.7 macrophages treated with L-NF, L-HF, OeNF or OeHF. The CMs were collected named as LeNFeM, LeHFeM, OeNF-M and OeHF-M, respectively. All these macrophages CMs had significantly higher levels of TNFa and MCP-1 than the CM from macrophages treated with regular adipocyte growth medium (RM). The fold increases of TNFa were 3.9 ± 0.3fold, 11.3 ± 0.2-fold, 11.7 ± 0.4-fold and 11.0 ± 0.5-fold, respectively (p < 0.001 vs RM). The fold increases of MCP-1 were 5.0 ± 0.3-fold, 8.6 ± 0.3-fold, 9.7 ± 0.4-fold, 9.4 ± 0.5-fold (p < 0.01, p < 0.001, p < 0.001, p < 0.001, vs RM), respectively (Fig. 1A and B). In addition, levels of TNFa and MCP-1 in LeHFeM, OeNF-M and OeHF-M are significantly higher than LeNFeM. The fold increases of TNFa were 2.8 ± 0.4-fold, 2.9 ± 0.3-fold, 2.7 ± 0.4-fold, respectively (p < 0.001 vs LeNFeM). The fold increases of MCP-1 were 1.7 ± 0.3-fold, 1.9 ± 0.3, 1.8 ± 0.5, respectively (p < 0.001 all vs LeNFeM, Fig. 1A and B).
3.2. CM from macrophages-treated with adipocyte-derived CM induced insulin resistance by activating stress and inflammatory signals in skeletal muscle cells We next detected if the increased secretion of inflammatory
3
cytokines from macrophages treated with adipocyte-derived CM affect insulin signal in skeletal muscle cells. Macrophage CMs (LeNFeM, LeHFeM, OeNF-M, OeHF-M) were applied to C2C12 myotubes for 16 h following treatment with 100 nM insulin in the last 10 min. The phosphorylation of insulin signaling molecules Akt and AS160 (also known as TBC1D4) were detected. All these Macrophage CMs significantly impaired insulin-stimulated Akt Ser473 phosphorylation by 47 ± 4%, 78 ± 9%, 79 ± 13, 85% ± 11% reduction, respectively (p < 0.001 vs insulin in RM) and AS160 Thr642 phosphorylation by 34 ± 6%, 47 ± 10%, 49 ± 5%, 51 ± 7% reduction, respectively (p < 0.05 vs insulin in RM, Fig. 2A and B). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2A and B). This is consistent with the marked increased inflammatory cytokines in macrophage-derived CMs. IRS-1 serine phosphorylation reduces its insulin-dependent tyrosine phosphorylation and downstream signaling leading to insulin resistance in C2C12 myotubes [14,15]. We examined phosphorylation of IRS1 Ser307 and Ser636/639 in myotube treated with macrophages CMs. All macrophage CMs significantly elevated both basal IRS1 pSer307 phosphorylation by 1.32 ± 0.05-fold, 1.79 ± 0.08-fold, 1.83 ± 0.05-fold, 1.81 ± 0.07-fold, respectively (p < 0.05 vs basal in RM) and insulin-stimulated phosphorylation by 1.26 ± 004-fold, 1.71 ± 0.07-fold, 1.77 ± 0.05-fold, 164 ± 0.09fold, respectively (p < 0.05 vs insulin in RM, Fig. 2C). LeNFeM, LeHFeM, OeNF-M and OeHF-M significantly elevated both basal IRS1 pSer636/639 phosphorylation by 1.31 ± 0.03-fold, 1.65 ± 0.1fold, 1.69 ± 0.05-fold, 1.68 ± 0.07-fold, respectively (p < 0.05, p < 0.05, p < 0.001, p < 0.001 vs basal in RM) and insulin-stimulated phosphorylation by 1.39 ± 0.05-fold, 1.63 ± 0.11-fold, 1.66 ± 0.07fold, 1.69 ± 0.1-fold, respectively (p < 0.05 vs insulin in RM, Fig. 2D). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2C and D). JNK and S6K can phosphorylate IRS1 Ser307 and Ser636/639, respectively [16,17], which mediate the inhibitory effects of TNFa on insulin signal. Indeed, we found phosphorylation of JNK and S6K was increased in myotubes treated with all macrophage CMs. LeNFeM, LeHFeM, OeNF-M and OeHF-M significantly elevated both basal JNK pThr183/185 phosphorylation by 1.33 ± 0.07-fold, 1.77 ± 0.12-fold, 1.85 ± 0.09-fold, 1.83 ± 0.11-fold, respectively (p < 0.05 vs basal in RM) and insulin-stimulated JNK pThr183/185 phosphorylation by 1.21 ± 0.05-fold, 1.67 ± 0.1-fold, 1.74 ± 0.1-fold, 1.81 ± 0.13-fold, respectively (p < 0.05 vs basal in RM) (Fig. 2E). Similarly, all macrophage CMs significantly elevated both basal S6K pThr389 phosphorylation by 1.47 ± 0.06-fold, 1.93 ± 0.14-fold, 1.86 ± 0.11-fold, 1.83 ± 0.12-fold, respectively (p < 0.01, p < 0.001, p < 0.001, p < 0.001 vs basal in RM) and insulin-stimulated S6K pThr389 phosphorylation by 1.46 ± 0.07-fold, 2.17 ± 0.14-fold, 2.23 ± 0.13-fold, 2.27 ± 0.15-fold, respectively (p < 0.05 vs basal in RM, Fig. 2F). LeHFeM, OeNF-M and OeHF-M were more effective than LeNFeM (Fig. 2E and F). The trends of phospho-JNK and phospho-S6K are consistent with the heightened phosphorylation of IRS1 Ser307 and Ser636/639 (Fig. 2C and D). 3.3. CM from co-cultured adipocytes and macrophages secretes more inflammatory cytokines
Fig. 1. Adipocytes induce inflammatory response in macrophages. Concentrations of TNFa (A) and MCP-1 (B) in conditioned media from macrophages were measured by ELISA, respectively. Results are means ± S.E. of 4 independent experiments. ***p < 0.001.
Macrophages within the hypoxic obese adipose tissue are exposed to adipocytes-derived adipokines and cytokines [18,19]. We wondered how hypoxia specifically affected macrophages and how the two types of cells affected each other. We cultured adipocytes and macrophages under normoxic or hypoxic condition, respectively. Meanwhile, we co-cultured adipocytes and macrophages in a transwell system under normoxic or hypoxic condition, respectively. Consistent with our previous results [8], hypoxia elevated the levels of TNFa (3.84 ± 0.4-fold, p < 0.001) and MCP-1
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
4
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
5
(5.53 ± 0.5-fold, p < 0.05) in HF compared with NF (Fig. 3 and B, vs NF), but did not affect the amounts of TNFa and MCP-1 in NM and HM cultures (Fig. 3A and B). Interestingly, co-cultured adipocytes and macrophages markedly increased the releases of TNFa and MCP-1 in both normoxia and hypoxia conditions. NF/M had 11.3 ± 0.6-fold of TNFa and 9.7 ± 0.4-fold of MCP-1 compared to NF (p < 0.001). HF/M had 11.54 ± 0.5-fold of TNFa and 10.6 ± 0.6-fold of MCP-1 compared to NF (p < 0.001, Fig. 3A and B). Moreover, the levels of TNFa and MCP-1 in NF/M and HF/M were higher than that in HF. The fold increases for TNFa and MCP-1 were 2.9 ± 0.5-fold, and 1.8 ± 0.5-fold (p < 0.001, NF/M vs HF, Fig. 3A and B), respectively. The fold increases for TNFa and MCP-1 were 3.0 ± 0.6-fold and 2.0 ± 0.6-fold (p < 0.001, HF/M vs HF, Fig. 3A and B), respectively. 3.4. CM from co-cultured adipocytes and macrophages activate stress and inflammatory signals and evoke insulin resistance in skeletal muscle cells We detected the effects of CMs from co-cultured adipocytes and macrophages on Akt and AS160 in muscle cells. Compared to Akt phosphorylation in control RM group, NF, NM and HM did not change basal or insulin-stimulated Akt and AS160 phosphorylation (Fig. 4A and B). Similar to our previous report, HF did not change basal levels, but blunted insulin-stimulated Akt and AS160 phosphorylation by 43 ± 5% (p < 0.01 vs insulin in RM, Fig. 4A) and 26 ± 3% (p < 0.001 vs insulin in RM, Fig. 4B). Moreover, both NF/M and HF/M did not affect basal, but markedly reduced insulinstimulated Akt phosphorylation by 59 ± 4% and 61 ± 7% (p < 0.001 vs insulin in RM, Fig. 4A) and AS160 phosphorylation by 41 ± 5%, 42 ± 5% (p < 0.001 vs insulin in RM, Fig. 4B). In addition, NF/M and HF/M induced more severe insulin resistance of Akt and AS160 than HF (p < 0.001 vs insulin in HF, Fig. 4A and B). We next examined IRS1 S307 and S636/639 phosphorylation. NF, NM and HM had no effect on IRS1 serine phosphorylations (Fig. 4C and D). However, HF, NF/M and HF/M raised basal phosphorylations by 1.76 ± 0.07-fold, 1.99 ± 0.11-fold, 1.95 ± 0.09-fold for S307 (p < 0.05 vs basal in RM, Fig. 4C), 1.89 ± 0.13-fold, 3.16 ± 0.11-fold, 3.95 ± 0.19-fold for S636/639 (p < 0.01, p < 0.001, p < 0.001 vs basal in RM, Fig. 4D). Moreover, NF/M and HF/M were more effective than HF on IRS1 S307 (p < 0.05 vs. HF, Fig. 4C) and S636/639 (p < 0.01, p < 0.001 vs HF, Fig. 4D). The patterns of JNK and S6K phosphorylations were consistent with IRS1 serine phosphorylations. NF, NM and HM had no effect on JNK and S6K phosphorylations (Fig. 4E and F). HF, NF/M and HF/M raised JNK phosphorylation by 1.87 ± 0.09-fold, 3.44 ± 0.14-fold, 3.49 ± 0.12fold (p < 0.01, p < 0.001, p < 0.001 vs basal in RM, Fig. 4E) and S6K phosphorylation by 2.45 ± 0.15-fold, 3.21 ± 0.1-fold, 3.65 ± 0.13fold (p < 0.001 vs basal in RM, Fig. 4F). In addition, NF/M and HF/ M were more effective than HF on JNK (p < 0.001 vs. HF, Fig. 4E) and S6K (p < 0.05 vs HF, Fig. 4F). 4. Discussion Massive expansion of adipose tissue is the hallmark of obesity and excess of adipose tissue is attributed to hypertrophy and hyperplasia of adipocytes. Adipose tissues of obese individuals manifest insulin resistance associated with macrophages infiltration and local inflammation. Recent studies shows that hypoxia in
Fig. 3. Co-cultured of adipocytes with macrophages exacerbate inflammatory response than each cell type in both normoxic and hypoxic condition. TNFa (A) and MCP-1 (B) in conditioned media of 3T3-L1 adipocytes, RAW264.7 macrophages and cocultured of these 2 types cells were measured by ELISA, respectively. Results are means ± S.E. of 4 independent experiments. ***p < 0.001.
adipose tissues is an early phenomenon in obesity. We previously reported that hypoxia-treated 3T3-L1 adipocytes secreted increased levels of inflammatory cytokines. CM from hypoxic adipocytes induced insulin resistance of muscle cells. MCP-1 played a role in this mechanism [8]. Macrophages infiltrated in adipose tissue are the main sources of inflammatory cytokines of obese adipose tissue. The effects of hypoxia on macrophages and on the interaction of adipocytes and macrophages are not clear. The present study was undertaken to illustrate the effect of hypoxia on the interaction between adipocytes and macrophages. The effect of the factors from macrophages on insulin signal in muscle cells was also explored. To this end, we prepared CM from macrophages incubated with conditioned medium from normoxic or hypoxic adipocytes, from cultured macrophages alone or from macrophageadipocyte co-cultures under normoxic or hypoxic conditions,
Fig. 2. Conditioned media from RAW 264.7 macrophages reduces insulin-stimulated Akt and AS160 phosphorylation and increases IRS1 serine, JNK, S6K phosphorylation in C2C12 myotubes. C2C12 myotubes were treated with conditioned media from macrophages for 16 h, followed by 100 nM insulin treatment in the last 10 min prior to cell lysis and immunoblotting for phospho-Akt S473 (A), phospho-AS160 T642 (B), phospho-IRS1 Ser307 (C), phospho-Ser636/639 (D), phospho-JNK Thr183/Tyr185 (E) and phospho-S6K Thr389 (F). Shown are representative blots. The densitometry mean ± S.E. from 4 independent experiments is plotted as the fold change relative to RM group or as indicated. *p < 0.05, **p < 0.01, ***p < 0.001.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
6
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Fig. 4. Conditioned media from co-cultured adipocytes and macrophages exacerbate muscle cells insulin resistance than CMs from each cell type. C2C12 myotubes were treated with indicated conditioned medium for 16 h, followed by 100 nM insulin treatment in the last 10 min prior to cell lysis and immunoblotting for phospho-Akt S473 (A), phosphoAS160 T642 (B), phospho-IRS1 Ser307 (C), phospho-Ser636/639 (D), phospho-JNK Thr183/Tyr185 (E) and phospho-S6K Thr389 (F). Shown are representative blots. The densitometry mean ± S.E. from 4 independent experiments is plotted as the fold change relative to RM group or as indicated. *p < 0.05, **p < 0.01, ***p < 0.001.
respectively. Then the effects of CM from macrophages on insulin signal in muscle cells were determined. Macrophages inside the obese adipose tissue contribute to the release of inflammatory cytokines. Here we studied the direct and indirect effects of hypoxia on the release of cytokines from macrophages. Obese adipocytes and hypoxia-treated adipocytes released more TNFa and MCP-1 (data not shown). CMs from
adipocytes induced macrophages to secrete TNFa and MCP-1. Moreover, CMs from hypoxia-treated lean mice adipocytes, normoxia- and hypoxia-treated obese mice adipocytes (L-HF, OeNF, OeHF) stimulated macrophages to produce more TNFa and MCP1 than CM from normoxia-treated lean mice adipocytes (L-NF). Interestingly, although the levels of TNFa and MCP-1 in LeHFeM, OeNF-M, OeHF-M are different, the ability of these CM to induce
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162
C. Zhang et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
macrophages to produce TNFa and MCP-1 were the same. These data implicate that hypoxic adipocytes can release factors to activate macrophages. TNFa plays a role, but other factors in the CM may also be involved. To investigate the direct effect of macrophages on muscle cells insulin signal, we incubated muscle cells with CM from macrophages treated with different adipocyte CM. Consistent with the levels of TNFa and MCP-1 in the adipocytes CMs, LeNFeM, LeHFeM, OeNF-M and OeHF-M all lowered insulin-stimulated phosphorylation of Akt and AS160. Moreover, phosphorylation of IRS1 S307 and S636/639 were elevated, which are associated with impaired propagation of insulin signals from IRS1 to Akt and AS160 [15]. Phosphorylation of IRS1 S307 and S636/639 are regulated by JNK and S6K, respectively. Accordingly, macrophage CMs activated JNK and S6K. The trends of the phosphorylations of Akt, AS160, IRS1 serine, JNK and S6K correlated with the changes in the levels of TNFa and MCP-1. We previously reported that hypoxia-treated 3T3L1 adipocytes diminished insulin-stimulated phosphorylation of Akt and AS160 and activated JNK and S6K in muscle cells. MCP-1 played a role in the mechanism. In addition, TNFa activated JNK and S6K in muscle tissue [17]. MCP-1 and TNFa may involve in the activation of JNK and S6K by adipocyte-macrophage CMs. Macrophages infiltrate obese white adipose tissue. Hypoxia affects the release of factors from adipocytes, which further affect macrophages. We explored if hypoxia has direct effect on macrophages. We also tested the interaction between adipocytes and macrophages. Our data showed that hypoxia had no effect on macrophages to produce TNFa and MCP-1. Obese adipose tissues expend faster than angiogenesis and adipocytes in obese white adipose tissue are hypertrophic. Thus, oxygen has a lower probability to reach the inside of adipocytes in obese adipose tissue leading to hypoxia responses. On the other hand, macrophages escape from blood vessel to adipose tissue. Macrophages are close to blood vessel and the adipocytes are not hypertrophic in these areas. Therefore, hypoxia may not be the reason for macrophages to secrete inflammatory cytokines. However, co-cultured adipocytes and macrophages secreted much more TNFa and MCP-1 than each cell type, which confirmed the effect of hypoxic adipocytes on macrophages. Interestingly, co-culture adipocytes and macrophages under nomoxic condition also induced TNFa and MCP-1 secretion. One caveat is that the ratio of macrophages to adipocytes is higher than that in vivo. Too many macrophages may crosstalk with adipocytes and in turn to induce the release of TNFa and MCP-1 from macrophages. Some details may require further study. Accordingly, insulin-stimulated Akt and AS160 phosphorylation were reduced by HF, NF/M and HF/M. The phosphorylation of IRS1 S307 and S36/639 were increased. Again, HF, NF/M and HF/M activated JNK and S6K as the same trend of IRS1 S307 and S36/639, respectively. MCP-1 and TNFa may involve in the mechanism. In summary, hypoxia or obese adipocytes and co-culture with adipocytes induce macrophages to secrete TNFa and MCP-1. CMs from these macrophages reduce insulin signal in muscle cells. This may be related to the activation of JNK and S6K. TNFa and MCP-1 may play a role in this mechanism.
Author contributions CZ and XL conducted the majority of the data with equal contributions. DZ, BD, JT and MZ analysis the data and made the figures. LC and HD contributed to discussion and reviewed the manuscript. WN contributed to the study design, reviewed the data and wrote the paper.
7
Declaration of competing interest All authors declare no conflicts of interest. Acknowledgments This work was supported by grants from National Natural Science Foundation of China (#81670731, #81870547, #81170740 and #81161120545), the Tianjin Municipal Science and Technology Commission (#15JCZDJC35500) and the Tianjin Health and Family Planning Commission (#15KG102) to W Niu. Deng Bangli and Zhang Mo were supported by Scientific Research Funding of Tianjin Medical University Chu Hsien-I Memorial Hospital (#2018RC03, #2014DX07). Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2019.10.162. References [1] B.A. Swinburn, I. Caterson, J.C. Seidell, W.P. James, Diet, nutrition and the prevention of excess weight gain and obesity, Public Health Nutr. 7 (2004) 123e146. [2] M. Krekoukia, G.P. Nassis, G. Psarra, K. Skenderi, G.P. Chrousos, L.S. Sidossis, Elevated total and central adiposity and low physical activity are associated with insulin resistance in children, Metabolism 56 (2007) 206e213. [3] G.S. Hotamisligil, Inflammation and metabolic disorders, Nature 444 (2006) 860e867. [4] R.M. Pirzgalska, A.I. Domingos, Macrophages in obesity, Cell. Immunol. 330 (2018) 183e187. [5] J.C. McNelis, J.M. Olefsky, Macrophages, immunity, and metabolic disease, Immunity 41 (2014) 36e48. [6] J.M. Olefsky, C.K. Glass, Macrophages, inflammation, and insulin resistance, Annu. Rev. Physiol. 72 (2010) 219e246. [7] R.W. O’Rourke, A.E. White, M.D. Metcalf, A.S. Olivas, P. Mitra, W.G. Larison, E.C. Cheang, O. Varlamov, C.L. Corless, C.T. Roberts Jr., D.L. Marks, Hypoxiainduced inflammatory cytokine secretion in human adipose tissue stromovascular cells, Diabetologia 54 (2011) 1480e1490. [8] J. Yu, L. Shi, H. Wang, P.J. Bilan, Z. Yao, M.C. Samaan, Q. He, A. Klip, W. Niu, Conditioned medium from hypoxia-treated adipocytes renders muscle cells insulin resistant, Eur. J. Cell Biol. 90 (2011) 1000e1015. [9] F.S. Thong, C.B. Dugani, A. Klip, Turning signals on and off: GLUT4 traffic in the insulin-signaling highway, Physiology 20 (2005) 271e284. [10] S.H. Um, D. D’Alessio, G. Thomas, Nutrient overload, insulin resistance, and ribosomal protein S6 kinase 1, S6K1, Cell Metabol. 3 (2006) 393e402. [11] K.D. Copps, M.F. White, Regulation of insulin sensitivity by serine/threonine phosphorylation of insulin receptor substrate proteins IRS1 and IRS2, Diabetologia 55 (2012) 2565e2582. [12] Y.H. Lee, J. Giraud, R.J. Davis, M.F. White, c-Jun N-terminal kinase (JNK) mediates feedback inhibition of the insulin signaling cascade, J. Biol. Chem. 278 (2003) 2896e2902. [13] W. Niu, P.J. Bilan, S. Ishikura, J.D. Schertzer, A. Contreras-Ferrat, Z. Fu, J. Liu, S. Boguslavsky, K.P. Foley, Z. Liu, J. Li, G. Chu, T. Panakkezhum, G.D. Lopaschuk, S. Lavandero, Z. Yao, A. Klip, Contraction-related stimuli regulate GLUT4 traffic in C2C12-GLUT4myc skeletal muscle cells, Am. J. Physiol. Endocrinol. Metab. 298 (2010) E1058eE1071. [14] J.H. Lim, J.I. Lee, Y.H. Suh, W. Kim, J.H. Song, M.H. Jung, Mitochondrial dysfunction induces aberrant insulin signalling and glucose utilisation in murine C2C12 myotube cells, Diabetologia 49 (2006) 1924e1936. [15] J. Boucher, A. Kleinridders, C.R. Kahn, Insulin receptor signaling in normal and insulin-resistant states, Cold Spring Harb. Perspect. Biol. 6 (2014). [16] V. Aguirre, T. Uchida, L. Yenush, R. Davis, M.F. White, The c-Jun NH(2)terminal kinase promotes insulin resistance during association with insulin receptor substrate-1 and phosphorylation of Ser(307), J. Biol. Chem. 275 (2000) 9047e9054. [17] J. Zhang, Z. Gao, J. Yin, M.J. Quon, J. Ye, S6K directly phosphorylates IRS-1 on Ser-270 to promote insulin resistance in response to TNF-(alpha) signaling through IKK2, J. Biol. Chem. 283 (2008) 35375e35382. [18] S.U. Amano, J.L. Cohen, P. Vangala, M. Tencerova, S.M. Nicoloro, J.C. Yawe, Y. Shen, M.P. Czech, M. Aouadi, Local proliferation of macrophages contributes to obesityassociated adipose tissue inflammation, Cell Metabol. 19 (2014) 162e171. [19] Y. Kawano, J. Nakae, N. Watanabe, T. Kikuchi, S. Tateya, Y. Tamori, M. Kaneko, T. Abe, M. Onodera, H. Itoh, Colonic pro-inflammatory macrophages cause insulin resistance in an intestinal ccl2/ccr2-dependent manner, Cell Metabol. 24 (2016) 295e310.
Please cite this article as: C. Zhang et al., Hypoxic adipocytes induce macrophages to release inflammatory cytokines that render skeletal muscle cells insulin resistant, Biochemical and Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2019.10.162