Deficiency of Cbl-b Gene Enhances Infiltration and Activation of Macrophages in Adipose Tissue and Causes Peripheral Insulin Resistance in Mice

  1. Katsuya Hirasaka1,
  2. Shohei Kohno1,
  3. Jumpei Goto1,
  4. Harumi Furochi1,
  5. Kazuaki Mawatari2,
  6. Nagakatsu Harada2,
  7. Toshio Hosaka2,
  8. Yutaka Nakaya2,
  9. Kazumi Ishidoh3,
  10. Toshiyuki Obata4,
  11. Yousuke Ebina4,
  12. Hua Gu5,
  13. Shin'ichi Takeda6,
  14. Kyoichi Kishi1 and
  15. Takeshi Nikawa1
  1. 1Department of Nutritional Physiology, Institute of Health Biosciences, The University of Tokushima Graduate School, Tokushima, Japan
  2. 2Department of Nutrition and Metabolism, Institute of Health Biosciences, The University of Tokushima Graduate School, Tokushima, Japan
  3. 3Institute for Health Sciences, Tokushima Bunri University, Tokushima, Japan
  4. 4Division of Molecular Genetics, The Institute for Enzyme Research, The University of Tokushima, Tokushima, Japan
  5. 5Department of Microbiology, Columbia University, New York, New York
  6. 6National Institute of Neuroscience, National Center of Neurology and Psychiatry, Tokyo, Japan
  1. Address correspondence and reprint requests to Takeshi Nikawa, MD, PhD, Department of Nutritional Physiology, Institute of Health Biosciences, The University of Tokushima Graduate School, 3-18-15 Kuramoto-cho, Tokushima 770-8503, Japan. E-mail: nikawa{at}nutr.med.tokushima-u.ac.jp

Abstract

OBJECTIVE—c-Cbl plays an important role in whole-body fuel homeostasis by regulating insulin action. In the present study, we examined the role of Cbl-b, another member of the Cbl family, in insulin action.

RESEARCH DESIGN AND METHODS—C57BL/6 (Cbl-b+/+) or Cbl-b-deficient (Cbl-b−/−) mice were subjected to insulin and glucose tolerance tests and a hyperinsulinemic-euglycemic clamp test. Infiltration of macrophages into white adipose tissue (WAT) was assessed by immunohistochemistry and flow cytometry. We examined macrophage activation using co-cultures of 3T3-L1 adipocytes and peritoneal macrophages.

RESULTS—Elderly Cbl-b−/− mice developed glucose intolerance and peripheral insulin resistance; serum insulin concentrations after a glucose challenge were always higher in elderly Cbl-b−/− mice than age-matched Cbl-b+/+ mice. Deficiency of the Cbl-b gene significantly decreased the uptake of 2-deoxyglucose into WAT and glucose infusion rate, whereas fatty liver was apparent in elderly Cbl-b−/− mice. Cbl-b deficiency was associated with infiltration of macrophages into the WAT and expression of cytokines, such as tumor necrosis factor-α, interleukin-6, and monocyte chemoattractant protein (MCP)-1. Co-culture of Cbl-b−/− macrophages with 3T3-L1 adipocytes induced leptin expression and dephosphorylation of insulin receptor substrate 1, leading to impaired glucose uptake in adipocytes. Furthermore, Vav1, a key factor in macrophage activation, was highly phosphorylated in peritoneal Cbl-b−/− macrophages compared with Cbl-b+/+ macrophages. Treatment with a neutralizing anti–MCP-1 antibody improved peripheral insulin resistance and macrophage infiltration into WAT in elderly Cbl-b−/− mice.

CONCLUSIONS—Cbl-b is a negative regulator of macrophage infiltration and activation, and macrophage activation by Cbl-b deficiency contributes to the peripheral insulin resistance and glucose intolerance via cytokines secreted from macrophages.

Obesity is a major cause of insulin resistance (1) and is considered a chronic low-grade inflammatory disease (2). Substantial evidence has accumulated in recent years that chronic infiltration and activation of macrophages in white adipose tissue (WAT) underlie the obesity-related component of these insulin-resistant states (35). Infiltrating macrophages secrete proinflammatory cytokines and stimulate the secretion of adipokines from adipocytes (4,5). Tumor necrosis factor (TNF)-α and interleukin (IL)-6 are key factors that induce insulin resistance (6,7). Leptin and adiponectin are also suggested to act as hormones that regulate insulin resistance; hyperleptinemia in obese subjects is associated with insulin resistance in tissues such as liver, WAT, and skeletal muscle (8), whereas adiponectin increases insulin sensitivity of these tissues (9). However, the molecular mechanism of macrophage activation in WAT is still unknown.

Cbl-b is a unique ubiquitin ligase that is associated with maturation and activation of macrophages and T-cells (10,11). Expression of Cbl-b is upregulated by macrophage/monocyte differentiation of HL60 and U937 cell lines (10). Cbl-b influences CD28-dependent T-cell activation by selectively restraining T-cell receptor–mediated Vav1 activation (1214). A recent study reported that the COOH-terminal–truncated Cbl-b caused severe pancreatitis (diffused infiltration of immune cells) and resulted in reduced insulin secretion from β-cells in rats (15). Based on these findings, we hypothesized that Cbl-b is a negative regulator of macrophage function. Dysfunction of Cbl-b may stimulate infiltration and activation of macrophages in WAT, resulting in macrophage-mediated peripheral insulin resistance, which is a scheme that increased secretion of cytokines from macrophages on the basis of cross-talk between infiltrated macrophages and adipocytes, inducing peripheral insulin resistance.

The present study was designed to investigate insulin resistance and glucose uptake in skeletal muscle and WAT of Cbl-b–deficient mice (Cbl-b−/− mice). The results showed development of glucose intolerance and insulin resistance, which are the main features of type 2 diabetes, in elderly Cbl-b−/− mice (>20 weeks old). Furthermore, significant activation of macrophages was observed in WAT of elderly Cbl-b−/− mice but not in wild-type mice (Cbl-b+/+ mice). Our results suggest that Cbl-b is a candidate gene of type 2 diabetes, in addition to type 1 diabetes, and is a potentially suitable target for the development of therapeutic strategies for diabetes.

RESEARCH DESIGN AND METHODS

Cbl-b−/− mice and sample preparation.

We generated and maintained Cbl-b−/− mice, using the method described previously (12,16). The mice used in our experiments had been backcrossed eight times from the C57BL/6 strain. In these experiments, we used C57BL/6 mice (Japan SLC, Shizuoka, Japan) as Cbl-b+/+ mice. The mice were housed at 23 ± 2°C on a light (0800–2000 h)-dark (2000–0800 h) cycle and allowed free access to a laboratory animal diet MF (Oriental Yeast, Tokyo, Japan) and water. Cbl-b−/− mice were fertile and apparently normal and were housed under specific pathogen-free conditions (16). Stromal vascular (SV) and adipocyte fractions were separated from WAT of Cbl-b+/+ or Cbl-b−/− mice, as described previously (17). All protocols described in the present study were conducted according to the Guide for the Care and Use of Laboratory Animals at The University of Tokushima and were approved by the Committee of the Care and Use for Laboratory Animals at The University of Tokushima School of Medicine.

Assessment of insulin resistance in vivo.

After 12 h of fasting, mice at the indicated ages underwent an intraperitoneal insulin tolerance test (IPITT), intraperitoneal glucose tolerance test (IPGTT), or a hyperinsulinemic-euglycemic clamp test, as described previously (18). For a hyperinsulinemic-euglycemic clamp test, mice received intravenous infusion of insulin at 60 pmol · kg−1 body wt · min−1. Blood glucose concentration was maintained at 110 mg/dl by glucose infusion. Whole-body glucose uptake represents the glucose infusion rate. In a group of 20-week-old littermates, 25 μg anti–monocyte chemoattractant protein-(MCP)-1 antibody or nonimmune IgG (R&D systems, Minneapolis, MN) was intraperitoneally injected five times every 3 days, as described previously (19). One day after the last injection, mice were subjected to IPITT.

Flow cytometry.

Cells in the SV fraction were stained with a rat anti-F4/80 antibody (Serotec, Oxford, U.K.), followed by incubation with anti-rat IgG conjugated with Alexa488 (Molecular Probes, Eugene, OR). After staining with Hoechst 33342 (Dojindo, Osaka, Japan), the cells were analyzed by flow cytometry (Elite EPS; Beckman-Coulter, Fullerton, CA) (20).

Co-culture system of 3T3-L1 adipocytes and peritoneal macrophages.

3T3-L1 adipocytes were co-cultured with peritoneal macrophages using the method of Suganami et al. (21) with slight modifications. Briefly, mouse 3T3-L1 fibroblasts (DaiNippon Pharmaceutical, Osaka, Japan) were pretreated with 0.25 μmol/l dexamethasone, 5 μg/ml insulin, and 0.5 mmol/l 3-isobutyl-1-methylxanthine and then cultured for up to 10 days until differentiation into 3T3-L1 adipocytes. On the other hand, Cbl-b+/+ and Cbl-b−/− mice (8 weeks old) were injected intraperitoneally with 0.2 g/2 ml of proteose peptone to stimulate macrophages (22). Three days later, the mice underwent peritoneal lavage with endotoxin-free PBS, and cells were collected. The cells were plated on dishes and cultured with Dulbecco's modified Eagle's medium for 3 h. Nonadherent cells were removed by washing three times with PBS. Adherent cells were collected in PBS containing 10 mmol/l EDTA. The collected cells were macrophage rich (>95%), as confirmed by flow cytometry for F4/80 expression (data not shown). Isolated Cbl-b+/+ or Cbl-b−/− macrophages (1 × 104) were plated with fully differentiated 3T3-L1 adipocytes at macrophages:3T3-L1 adipocytes ratio of 1:100 and cultured for 3 days.

Glucose uptake.

Glucose uptake by 3T3-L1 adipocytes co-cultured as described above was determined using the method of Inoue et al. (23). The co-cultured cells were glucose starved by culturing with Krebs-Ringer buffer, pH 7.4, for 3 h. Then, the cells were treated with or without 100 nmol/l insulin for 30 min, followed by the addition of [3H]-2-deoxyglucose (2-DG) (final concentration of 0.1 mmol/l) to the media. After 5 min incubation, the cells were solubilized with 1% SDS, and the levels of radioisotopes in the homogenates were counted using a β-scintillation counter (model LSC-3500; Aloka, Tokyo, Japan). Glucose uptake by skeletal muscles and WAT was measured by the method of Terada et al. (24) with slight modifications.

Western blotting, immunoprecipitation, and immunohistochemistry.

Western blot analysis, immunoprecipitation, and immunohistochemistry were performed as described previously (2527). Oil-red staining of mouse liver was performed according to the method of Catalono and Lillie (28). In some cases, the sections were further incubated with Hoechst 33342 (Dojindo) or counterstained with hematoxylin and eosin (H&E).

Real-time RT-PCR.

Total RNA was subjected to real-time RT-PCR with SYBR Green dye by using an ABI7300 real-time PCR system (Applied Biosystems, Foster City, CA) as described previously (16). The oligonucleotide primers used for real-time PCR are shown in Table S1 (available in an online-only appendix at http://dx.doi.org/10.2337/db06-1768).

Measurements of other biochemical parameters.

Protein concentration was determined by the method of Lowry et al. (29) with BSA as the standard. Serum concentrations of total cholesterol (30), triglyceride (31), and free fatty acid (32) were measured with the respective kits as described previously. Concentrations of insulin, TNF-α, IL-6, glucagon, and MCP-1 were determined by using enzyme-linked immunosorbent assay (ELISA) kits (Morinaga Institute, Tokyo, Japan; Ray Biotech, Norcross, GA; Pierce, Rockford, IL; Yanaibara Institute, Fujimiya, Shizuoka, Japan; and R&D Systems, respectively) (33).

Statistical analysis.

All data were statistically evaluated by ANOVA using SPSS software (release 6.1; SPSS Japan, Tokyo, Japan) and were expressed as means ± SD (n = 3–10). Differences between two groups were assessed with Duncan's multiple range test. Differences were considered significant at P < 0.05.

RESULTS

Parameters of glucose metabolism in Cbl-b−/− mice.

To examine the effects of Cbl-b deficiency on glucose metabolism, we measured food intake, various tissue wet weights, and blood metabolic parameters (Table 1). There were no significant differences in these parameters between young and adult (4 and 10 weeks old, respectively) Cbl-b+/+ and Cbl-b−/− mice. However, the weights of epididymal fat and liver were higher, and that of quadriceps muscle was lower in 30-week-old Cbl-b−/− mice than those of Cbl-b+/+ mice. Hyperinsulinemia was detected in fasted elderly Cbl-b−/− mice, although the concentration of blood glucose in these mice was similar to that in Cbl-b+/+ control mice. Aging or deficiency of Cbl-b gene did not affect serum glucagon levels. With regard to blood parameters of lipid metabolism, the serum concentrations of nonesterified fatty acid and triglyceride of elderly Cbl-b−/− mice were significantly lower than those of Cbl-b+/+ mice of the same age. Deficiency of Cbl-b did not influence the level of serum cholesterol, even in aged mice. In addition, we did not observe any gross abnormalities in organs of elderly Cbl-b−/− mice (at least 1 year old).

Glucose intolerance and insulin resistance in Cbl-b−/− mice.

Fasting blood glucose concentration was similar in Cbl-b+/+ and Cbl-b−/− mice, even those aged >20 weeks (Fig. 1A), as mentioned above (Table 1). In 4-week-old Cbl-b+/+ mice, intraperitoneal injection of glucose rapidly increased glucose concentration in blood, reaching a peak level at 30 min after injection but returned to basal level at 60–120 min (Fig. 1A). Changes in blood glucose concentration in 4-week-old Cbl-b−/− mice after glucose injection were not different from those of Cbl-b+/+ mice. In contrast, blood glucose concentration of 20-week-old Cbl-b−/− mice or older was higher after glucose injection than that of same-age Cbl-b+/+ mice. Blood glucose level of 30-week-old Cbl-b−/− mice was significantly higher at every time point after glucose injection, compared with that of the Cbl-b+/+ mice (Fig. 1A).

Intraperitoneal insulin injection reduced blood glucose levels to a similar extent in 4-week-old Cbl-b+/+ and Cbl-b−/− mice (Fig. 1B). The dip was maximal at 30 min after the injection but gradually returned to basal levels. Interestingly, the decrease in blood glucose levels after insulin injection was delayed in 20- and 30-week-old Cbl-b−/− mice, although insulin injection produced a rapid decrease in blood glucose levels in Cbl-b+/+ mice of the same age (Fig. 1B). This tendency of insulin resistance was more apparent in 30- than in 20-week-old Cbl-b−/− mice.

Plasma insulin levels and histochemical analysis of Langerhans’ islands.

Plasma insulin concentration of 30-week-old Cbl-b−/− mice was higher than that of Cbl-b+/+ mice before and at every time point after glucose injection, and Cbl-b−/− mice had significantly higher blood glucose levels than Cbl-b+/+ mice (Figs. 1A and 2A). Changes in plasma insulin level of 4-week-old Cbl-b−/− mice before and after glucose injection were not different from those of Cbl-b+/+ mice of the same age (Table 1 and data not shown).

H&E staining showed that most Langerhans’ islands were normal in 20- and 30-week-old Cbl-b−/− mice (Fig. 2B). Although we did not find any diffuse infiltration of immune-related cells into Langerhans’ island, mononuclear cells infiltrated the neighboring areas of several Langerhans’ islands (Fig. 2B). These cells were CD4, CD8, and CD68 immunopositive. The number of such Langerhans’ islands tended to increase in parallel with aging in Cbl-b−/− mice (Fig. 2B), whereas no infiltrating mononuclear cells were observed in Langerhans’ islands of elderly Cbl-b+/+ mice (>20 weeks old). Thus, deficiency of the Cbl-b gene stimulated infiltration of T-cells and macrophages in neighboring areas of Langerhans’ islands in the pancreas (Fig. 2C). On the other hand, no infiltrating immune cells, such as macrophages and lymphoblastic cells, were detected in skeletal muscle of Cbl-b–deficient mice (data not shown).

Expression of Cbl-b protein and glucose uptake in WAT and skeletal muscle.

Because Cbl-b−/− mice exhibited a normal insulin secretion pattern (Fig. 2), we examined the contribution of insulin-sensitive tissues to glucose intolerance and insulin resistance in these mice. The expression of Cbl-b and c-Cbl, another member of the Cbl family (34), in mouse WAT was upregulated with age, whereas their expression levels in skeletal muscle did not change, even at more than 20 weeks of age (Fig. 3A). In parallel with changes in Cbl-b and c-Cbl, a macrophage-specific marker protein, F4/80, was also increased in WAT of aged mice (older than 20 weeks) (Fig. 3A). There was no increase in F4/80 in skeletal muscle, in which both Cbls were not increased. These findings suggest that the increases in Cbls seen in WAT are due to infiltration of macrophages into the tissue.

A significant disturbance of glucose uptake by WAT was noted in 30-week-old Cbl-b−/− mice but not in same-age Cbl-b+/+ mice (Fig. 3B), although the basal levels of glucose uptake were similar in Cbl-b+/+ and Cbl-b−/− mice. There was no difference in the uptake by skeletal muscle between Cbl-b+/+ and Cbl-b−/− mice (Fig. 3B). For measurement of glucose uptake, isolated skeletal muscles are vulnerable to glucose starvation when incubated for 30 min in the buffer. Furthermore, we cannot exclude possible washout of proinflammatory cytokines during such incubation of skeletal muscles, hence leading to false-negative results. Accordingly, we examined insulin resistance of skeletal muscle by using the hyperinsulinemic-euglycemic clamp test. The glucose infusion rate was significantly lower in Cbl-b−/− mice compared with Cbl-b+/+ (Fig. 3C), suggesting that skeletal muscles also contributed to insulin resistance in Cbl-b−/− mice.

We also examined the role of the liver in peripheral insulin resistance. Oil-red staining of liver sections showed fat deposition, but this was not accompanied by infiltration of immunocytes in Cbl-b−/− mice (Fig. 3D). This finding and the changes in fat and liver wet weights (Table 1) suggest that deficiency of the Cbl-b gene increases adiposity in elderly mice. Interestingly, expression of phospholipase A2-IB (PLA2-IB), a lipolytic enzyme associated with fatty liver (35), was significantly suppressed in liver of Cb-b−/− mice compared with that of Cbl-b+/+ mice, although Cbl-b deficiency did not change the expression of the gluconeogenic enzymes, PEPCK and glucose-6-phosphatase, and a transcriptional factor for lipid synthesis, sterol regulatory element binding protein-1c (SREBP-1c) (Fig. 3D). Consistent with no infiltration of macrophages into liver, decreased expression of MCP-1 was noted in liver of Cbl-b−/− mice. The level of Akt phosphorylation after insulin injection was normal in the liver of Cbl-b−/− mice, although it was inhibited in WAT of these mice compared with Cbl-b+/+ mice (Fig. 3E).

Enhanced infiltration of macrophages into WAT in Cbl-b−/− mice.

Consistent with the results of Western blotting (Fig. 3A), infiltrating mononuclear cells were noted in WAT of 20-week-old Cbl-b+/+ mice (Fig. 4A). Because these mononuclear cells were CD68+ (Fig. 4A), they were considered to be macrophages. Furthermore, CD68 macrophages expressed Cbl-b and c-Cbl proteins, suggesting that infiltration of macrophages contributed to the increases in Cbl-b and c-Cbl proteins in WAT of 20-week-old mice (Fig. 4A and data not shown). Deficiency of Cbl-b gene stimulated infiltration of macrophages into WAT (Fig. 4B).

Flow cytometric analysis of cells in the SV fraction from WAT showed a threefold increase in the ratio of F4/80 and Hoechst 33342 double-positive cells to total cells in fractions from Cbl-b−/− mice compared with that of Cbl-b+/+ mice (Fig. 4C). As expected, the expression levels of macrophage marker transcripts, such as CD68, F4/80, and the disintegrin-like and metalloproteinase 8 (ADAM8) mRNA, were also upregulated in SV fractions of 20-week-old Cbl-b+/+ mice compared with those of 4-week-old Cbl-b+/+ mice (Fig. 4D). A more prominent upregulation of these marker genes was noted in SV fractions of 20-week-old Cbl-b−/− mice, whereas no such change in expression was identified in SV fraction of young Cbl-b−/− mice (Fig. 4D).

Expression profile of cytokines and adipokines in WAT.

TNF-α, IL-6, and MCP-1 were hardly expressed in the adipocyte fractions of young and aged Cbl-b+/+ and Cbl-b−/− mice (Fig. 5A). Consistent with the increased infiltration of macrophages, the amounts of TNF-α and MCP-1 transcripts were significantly higher in the SV fractions from 20-week-old Cbl-b+/+ mice than those of young Cbl-b+/+ mice (Fig. 5A). Deficiency of the Cbl-b gene enhanced the expression of these cytokines in the SV fractions at 20 weeks of age but not in young age. After adjustment of expression in macrophages for the CD68 mRNA expression level in the SV fraction, the expression levels of TNF-α, IL-6, and MCP-1 mRNAs were significantly upregulated in Cbl-b−/− mice (Fig. 5B), indicating that Cbl-b deficiency induces the expression of cytokines in macrophages. In contrast, the expression of leptin and adiponectin was mainly seen in the adipocyte fraction (Fig. 5C). In elderly Cbl-b+/+ mice, low leptin expression was noted in adipocytes, while that of adiponectin was 2.5-fold that of young Cbl-b+/+ mice. Conversely, adipocytes of 20-week-old Cbl-b−/− mice expressed significantly higher levels of leptin transcripts and lower levels of adiponectin transcripts compared with those of Cbl-b+/+ mice (Fig. 5C).

Effects of peritoneal Cbl-b−/− macrophages on adipokine expression, glucose uptake, and insulin signaling of 3T3-L1 adipocytes.

To examine whether macrophages of Cbl-b+/+ and Cbl-b−/− mice are functionally different, we measured cytokines released from the proteose peptone-stimulated peritoneal macrophages. Stimulated peritoneal Cbl-b−/− macrophages (1 × 104 cells) expressed TNF-α, IL-6, and MCP-1 transcripts at about two-, three-, and twofold, respectively, higher than stimulated Cbl-b+/+ macrophages (1 × 104 cells) (Fig. 6A). Consistent with changes in IL-6 and MCP-1 transcripts, the amounts of IL-6 and MCP-1 released from Cbl-b−/− macrophages were significantly higher than those from Cbl-b+/+ macrophages (Fig. 6B). TNF-α was not detected by our ELISA system even in media containing cultured Cbl-b−/− macrophages (data not shown). We previously reported that Cbl-b is a negative regulator of CD28-dependent signaling in T-cells via Vav1 dephosphorylation (12). Therefore, we examined tyrosine phosphorylation of Vav1 in macrophages of Cbl-b−/− mice. Vav1 was only slightly phosphorylated in Cbl-b+/+ macrophages (Fig. 6C) but significantly enhanced in Cbl-b−/− macrophages.

Co-culture of 1 × 104 proteose peptone-stimulated Cbl-b−/− peritoneal macrophages, but not Cbl-b+/+ macrophages, with 1 × 106 3T3-L1 adipocytes significantly upregulated the expression levels of leptin transcripts in the latter cells (Fig. 6D). Co-culture of these adipocytes with Cbl-b+/+ or Cbl-b−/− macrophages downregulated adiponectin mRNA expression to a similar extent compared with 3T3-L1 adipocytes cultured alone. In contrast, co-culture with Cbl-b+/+ macrophages significantly suppressed glucose uptake by 3T3-L1 adipocytes, whereas co-culture with Cbl-b−/− macrophages further decreased glucose uptake (Fig. 6E). We ignored the uptake of 2-DG by macrophages because these co-cultured cells contained only ∼1% of macrophages relative to adipocytes.

Two insulin-mediated signaling pathways have been described for GLUT4 translocation in adipocytes: the insulin receptor substrate (IRS)/phosphatidylinositol 3 kinase–dependent and –independent insulin-signaling pathways (36). TNF-α and IL-6 inhibit insulin-stimulated tyrosine phosphorylation of IRS-1, resulting in impaired insulin signaling in adipocytes (6,7). Therefore, we also examined the effects of co-culture of 3T3-L1 adipocytes with macrophages on these insulin-signaling pathways. Insulin treatment induced tyrosine phosphorylation of IRS-1 in 3T3-L1 adipocytes co-cultured with Cbl-b+/+ macrophages, while co-culture with Cbl-b−/− macrophages significantly inhibited insulin-mediated tyrosine phosphorylation of IRS-1 (Fig. 6F). In the IRS/phosphatidylinositol 3 kinase–independent pathway, insulin-stimulated phosphorylation of c-Cbl in 3T3-L1 adipocytes co-cultured with both Cbl-b+/+ and Cbl-b−/− macrophages (Fig. 6F). We also measured the translocation of c-Cbl and C3G into raft in plasma membrane, which are key steps for IRS-1/phosphatidylinositol 3 kinase–independent GLUT4 translocation (37). Unexpectedly, we could not detect insulin-stimulated translocation of c-Cbl and C3G into the Triton-X–insoluble fraction in 3T3-L1 adipocytes co-cultured with macrophages (Fig. 6G). Furthermore, deficiency of the Cbl-b gene in macrophages did not affect their translocation.

Effect of anti–MCP-1 antibody on insulin resistance in Cbl-b−/− mice.

We found that serum MCP-1 levels of elderly Cbl-b−/− mice were higher than those of Cbl-b+/+ mice of similar age, whereas serum MCP-1 levels of young Cbl-b+/+ and Cbl-b−/− mice were similar (Fig. 7A). Therefore, we examined whether MCP-1 provides a link between the inflammatory state of macrophages in Cbl-b−/− mice and the peripheral insulin resistance. Interestingly, an anti–MCP-1 neutralizing antibody improved peripheral insulin resistance in Cbl-b−/− mice compared with a nonimmune antibody (Fig. 7B). Furthermore, treatment with an anti–MCP-1 antibody, but not a nonimmune antibody, prevented macrophage infiltration into WAT (Fig. 7C).

DISCUSSION

Macrophages that infiltrate the WAT play a major role in induction of insulin resistance (15). Enhanced activation of macrophages in elderly Cbl-b−/− mice, for example, increases the secretion of TNF-α, IL-6, and MCP-1. This is the most important finding in our study because these proinflammatory cytokines are known to induce peripheral insulin resistance (6,7). Recent studies identified Vav1 as the main regulatory GDP/GTP exchange factor in toll-like receptor signaling, which produces MCP-1 in macrophages (38,39). Interestingly, in peritoneal macrophages of Cbl-b−/− mice, phosphorylation (activation) of Vav1 was significantly enhanced, compared with that in Cbl-b+/+ mice, indicating that activation of macrophages and T-cells in Cbl-b deficiency was mediated through Vav1 phosphorylation. In addition, co-culture of Cbl-b−/− macrophages with 3T3-L1 adipocytes induced expression of leptin and dephosphorylation of IRS-1, leading to disturbed glucose uptake in adipocytes. These results are in agreement with the results of two recent studies (40,41). Phosphorylation of Vav1 increased the expression of IL-6 via activation of nuclear factor (NF)-IL-6 (40), and Vav1 deficiency reduced macrophage migration (41). Thus, inhibition of Vav1 in macrophages could be a potentially effective target to prevent their infiltration in mice with chronic inflammation in WAT.

Another interesting finding of this study was the infiltration of macrophages into WAT associated with Cbl-b deficiency. In addition to a recent study showing that bone marrow–derived mast cells in Cbl-b−/− mice produced large amounts of MCP-1 (42), we demonstrated that Cbl-b deficiency enhanced MCP-1 expression in peritoneal and SV fraction–derived macrophages of elderly mice. Serum MCP-1 levels were higher in elderly Cbl-b−/− mice than in age-matched Cbl-b+/+ mice. Furthermore, treatment with an anti–MCP-1 antibody significantly improved peripheral insulin resistance and macrophage infiltration into WAT observed in elderly Cbl-b−/− mice. Our results suggest that in Cbl-b−/− mice, increased secretion of MCP-1 and activation of macrophages may contribute, in a coordinated fashion, to the infiltration of these cells into WAT and Langerhans’ islands. Thus, MCP-1 provides a link between the inflammatory-state macrophages and peripheral insulin resistance in Cbl-b−/− mice.

Previous studies using RNA interference techniques showed that deletion of c-Cbl and Cbl-b had no effect on insulin-stimulated glucose uptake in adipocytes (43,44). In fact, we found that Cbl-b was expressed in these macrophages, not adipocytes. Macrophages in Cbl-b−/− mice were more activated and infiltrated into WAT compared with Cbl-b+/+ mice. In contrast, insulin-stimulated tyrosine phosphorylation of IRS-1 was decreased in adipocytes co-cultured with Cbl-b−/− macrophages, compared with Cbl-b+/+ macrophages, although insulin-stimulated phosphorylation of c-Cbl was similar in adipocytes co-cultured with Cbl-b+/+ and Cbl-b−/− macrophages. Insulin-stimulated c-Cbl or C3G recruitment into lipid rafts was not detected in adipocytes co-cultured with Cbl-b−/− macrophages. In this regard, Lesniewski et al. (45) reported recently that deletion of the bone marrow–specific Cbl-associated protein (CAP) gene protected against high-fat diet–induced insulin resistance by preventing infiltration of macrophages into WAT (45). Thus, it is likely that in this case, Cbl-b has a modulatory role in macrophage function but not in adipocytes.

A significant disturbance of glucose uptake in WAT was noted in elderly Cbl-b−/− mice (>20 weeks old) but not in aged-matched Cbl-b+/+ mice. The glucose infusion rate was significantly lower in elderly Cbl-b−/− mice compared with Cbl-b+/+ mice. Therefore, WAT and skeletal muscles are responsible for the aforementioned changes in Cbl-b−/− mice. In contrast, deficiency of Cbl-b also caused an increase in wet weight and fat deposition in the liver, indicating mild fatty liver, although infiltration of immune cells into the liver was hardly observed in Cbl-b−/− mice. However, based on a normal response of hepatic insulin signaling and no increased expression of hepatic gluconeogenic enzymes in Cbl-b−/− mice, it is unlikely that hepatic insulin resistance occurs in elderly Cbl-b−/− mice. Given that fatty liver is frequently associated with hepatic insulin resistance, we cannot exclude the possibility that the liver could also play a role in peripheral insulin resistance in Cbl-b−/− mice. The precise mechanism through which Cbl-b deficiency induces hepatic insulin resistance requires further investigation.

Fatty liver is noted in Cbl-b−/− mice fed a standard diet. In this regard, deficiency of the Cbl-b gene significantly suppressed expression of PLA2-IB, a fatty liver–associated lipolytic enzyme (35), in liver, whereas expression of SREBP-1c, an important transcription factor for lipid synthesis, was not changed. Based on these findings, Cbl-b deficiency may influence lipolysis rather than lipogenesis in liver. There is no report, to our knowledge, that MCP-1 downregulates expression of PLA2-IB, although MCP-1 has been reported to upregulate hepatic expression of SREBP-1c in its transgenic mice fed a high-fat diet (46). At present, we cannot determine whether MCP-1 mediates impaired expression of PLA2-IB in liver of Cbl-b−/− mice. Further examinations are necessary to elucidate the mechanism of inhibitory effect of Cbl-b deficiency on hepatic expression of PLA2-IB mRNA.

FIG. 1.

Glucose intolerance and insulin resistance in Cbl-b−/− mice. A: Results of an IPGTT for Cbl-b+/+ and Cbl-b−/− mice of the indicated ages. Mice fasted for 12 h were injected intraperitoneally with glucose (2 g/kg body wt). Blood samples were taken from the tail vein before and at the indicated time points after glucose injection for measurement of blood concentrations. Data are means ± SD (n = 7). *P < 0.05 compared with Cbl-b+/+ mice. B: Results of an IPITT for Cbl-b+/+ and Cbl-b−/− mice of the indicated ages and littermates of mice subjected to IPGTT. Mice fasted for 12 h were injected intraperitoneally with insulin (0.75 units/kg body wt). Blood samples were taken from the tail vein before and at the indicated time points after insulin injection, for measurement of blood glucose level. Data are means ± SD (n = 7). *P < 0.05 compared with Cbl-b+/+ mice.

FIG. 2.

Plasma insulin levels and histochemical analysis of Langerhans’ islands in Cbl-b−/− mice. A: Serum insulin concentrations were determined in 30-week-old Cbl-b+/+ and Cbl-b−/− mice subjected to IPGTT. Mice underwent IPGTT, and blood samples were taken from the tail vein, as described in Fig. 1, for measurement of serum insulin levels by ELISA. Data are means ± SD (n = 7). *P < 0.05 compared with Cbl-b+/+ mice. B: The pancreas was dissected out from mice subjected to IPGTT, as described in the legend of Fig. 2A. Sections (3 μm thickness) from the pancreas of Cbl-b+/+ (top left panel) and Cbl-b−/− (top right panel) mice (30 weeks old) were stained with H&E. The number of Langerhans’ islands (indicated by the large arrow) infiltrated with mononuclear cells (indicated by arrowheads) in Cbl-b−/− pancreas was counted under high magnification (bottom left panel). The ratio of Langerhans’ islands infiltrated with mononuclear cells to normal ones (n = 100) was calculated in Cbl-b−/− mice at the indicated ages (bottom right panel). Similar results were obtained in three separate experiments. Data are means ± SD (n = 4/group/experiment). *P < 0.05 compared with 4-week-old Cbl-b−/− mice. Scale = 100 μm. C: Serial sections of the pancreas were immunostained with an anti-CD4, anti-CD8, and anti-CD68 antibody and Hoechst 33342, as indicated.

FIG. 3.

Expression of Cbl-b protein and glucose uptake in white adipose tissue and skeletal muscle. A: Epididymal WAT and gastrocnemius muscle were prepared from Cbl-b+/+ mice of the indicated ages. Homogenized proteins (40 μg/lane) were subjected to SDS–8% PAGE, and Western blotting was performed. The ratio of Cbl-b, c-Cbl, and F4/80 protein to β-actin was calculated by densitometric analysis. Data are means ± SD (n = 4). *P < 0.05 compared with the value of WAT of 4-week-old mice. B: Epididymal WAT and gastrocnemius muscle were isolated from 30-week-old Cbl-b+/+ or Cbl-b−/− mice. The uptake of 2-DG into the tissues was measured and calculated in the absence or presence of insulin (100 nmol/l). Glucose transport activity is expressed in nanomoles of 2-DG per gram of DNA per 1 h. Data are means ± SD (n = 4). *P < 0.05 compared with the values of absence of insulin, #P < 0.05 compared with the value of Cbl-b+/+ mice. C: Glucose infusion rate was determined by using the euglycemic-hyperinsulinemic glucose clamp test. Insulin-mediated whole-body glucose uptake was measured in anesthetized Cbl-b+/+ or Cbl-b−/− mice aged 20–25 weeks. Data are means ± SD (n = 6). *P < 0.05 compared with Cbl-b+/+ mice. D: Liver of Cbl-b+/+ and Cbl-b−/− mice (20 weeks old) was prepared at 0800 h and stained with oil-red and H&E (top). Similar results were obtained from four separate experiments. Scale = 100 μm. The abundance of hepatic PEPCK, glucose-6-phosphatase (G6Pase), SREBP-1c, PLA2-IB, MCP-1, and β-actin (internal standard) were quantified by real-time RT-PCR. Results are expressed relative to the corresponding value for Cbl-b+/+ mice (bottom). Data are means ± SD (n = 4). E: Cbl-b+/+ and Cbl-b−/− mice (20 weeks old) fasted for 12 h were injected intraperitoneally with 0.75 units/kg body wt of insulin or vehicle (PBS). Liver and WAT were isolated 15 min after the injection. Phosphorylation of Akt was assessed by Western blotting. Similar results were obtained from three separate experiments. BW, body weight; GIR, glucose infusion rate.

FIG. 4.

Infiltration of macrophages into WAT. A and B: Sections (3 μm thickness) from WAT of 20-week-old Cbl-b+/+ (A) and Cbl-b−/− (B) mice were stained with H&E. Serial sections were immunostained with anti–Cbl-b and anti-CD68 antibodies. Fluorescence microscopic images were merged. Similar results were obtained in four separate experiments, and the best results are shown as representative images. Scale = 50 μm. C: The ratio of F4/80 and Hoechst 33342 double-positive cells to total cells in SV fractions from WAT of Cbl-b+/+ and Cbl-b−/− mice was measured by flow cytometric analysis. Data are means ± SD (n = 3). *P < 0.05 compared with Cbl-b+/+ mice. D: Expression of macrophage marker transcripts in SV and adipocyte factions (ADIs) separated from epididymal WAT was assessed by real-time RT-PCR analysis. The fluorescence ratio of target gene cDNA to ribosomal protein S3 (Rps3), a housekeeping gene, was calculated. Data are means ± SD (n = 4). *P < 0.05, #P < 0.05 compared with the values of 4- and 20-week-old Cbl-b+/+ mice, respectively.

FIG. 5.

Expression of cytokines and adipokines in white adipose tissue. AC: Expression levels of cytokines (A and B) and adipokine (C) transcripts in SV and adipocyte factions (ADIs) separated from epididymal WAT were assessed by real-time RT-PCR analysis. The fluorescence ratio of target gene cDNA to ribosomal protein S3 (Rps3) (A), a housekeeping gene, or CD68 (B), a macrophage-specific gene, was calculated. Data are means ± SD (n = 4). *P < 0.05, #P < 0.05 compared with the values of 4- and 20-week-old Cbl-b+/+ mice, respectively.

FIG. 6.

Effects of cultured Cbl-b−/− peritoneal macrophages on cytokine expression, glucose uptake, and insulin signaling in 3T3-L1 adipocytes. AC: Peritoneal macrophages (MΦ) (1 × 104) from 8-week-old proteose peptone-treated Cbl-b+/+ and Cbl-b−/− mice were collected and cultured for 3 days. The amounts of cytokine transcripts (A) and proteins (B) were measured by real-time RT-PCR and ELISA, respectively. The fluorescence ratio of the target gene cDNA to Rps3, a housekeeping gene, was calculated. Phosphorylation of Vav1 was assessed by immunoprecipitation and Western blotting (C). Data are means ± SD (n = 4 for RT-PCR, n = 10 for ELISA, n = 4 for Western blotting). *P < 0.05 compared with Cbl-b+/+ mice. IP, immunoprecipitation; WB, Western blotting. D: 3T3-L1 adipocytes (1 × 106) were co-cultured with peritoneal Cbl-b+/+ or Cbl-b−/− macrophages (1 × 104), which were prepared as described in A. Three days later, the expression of adipokine transcripts was measured by real-time RT-PCR. The fluorescence ratio of the target gene cDNA to 36B4, a housekeeping gene of adipocytes, was calculated. Data are means ± SD (n = 4). *P < 0.05 compared with adipocytes cultured alone. E: 3T3-L1 adipocytes were co-cultured without or with Cbl-b+/+ and Cbl-b−/− macrophages for 3 days. The uptake of 2-DG by 3T3-L1 adipocytes was measured before and 30 min after treatment with 100 nmol/l insulin. Data are means ± SD (n = 4). *P < 0.05, #P < 0.05 compared with adipocytes cultured alone and co-cultured with Cbl-b+/+ macrophages, respectively. F: Phosphorylation of insulin signal molecules before or 5 min after treatment with 100 nmol/l insulin was assessed by immunoprecipitation and/or Western blotting. The ratio of phosphorylated IRS-1 or c-Cbl to respective total protein was calculated by densitometric analysis. Data are means ± SD (n = 4). *P < 0.05, #P < 0.05 compared with before insulin treatment and co-cultures with Cbl-b+/+ macrophages, respectively. G: 3T3-L1 adipocytes were co-cultured without or with Cbl-b+/+ and Cbl-b−/− macrophages for 3 days. Before or 5 min after treatment with 100 nmol/l insulin, the translocation of C3G and c-Cbl into lipid rafts in 3T3-L1 adipocytes was measured by Western blotting. Samples were separated into Triton X-100–soluble and –insoluble fractions, respectively. Similar results were obtained from three separate experiments.

FIG. 7.

Effects of anti–MCP-1 antibody on insulin resistance in Cbl-b−/− mice. A: Blood samples were prepared from Cbl-b+/+ and Cbl-b−/− mice of the indicated ages. Serum MCP-1 levels were determined by ELISA. Data are means ± SD (n = 7). *P < 0.05 compared with Cbl-b+/+ mice. B: Cbl-b−/− mice, intraperitoneally injected with an anti–MCP-1 IgG or nonimmune antibody, underwent IPITT. Blood glucose was measured at the indicated time after insulin injection. Data are means ± SD (n = 7). *P < 0.05 compared with nonimmune IgG treatment. C: Sections (3 μm thickness) from WAT of Cbl-b−/− mice injected with an anti–MCP-1 IgG or nonimmune antibody were stained with H&E. Similar results were obtained in four separate experiments, and representative images are shown. Scale = 50 μm.

TABLE 1

Metabolic parameters in Cbl-b−/− mice

Acknowledgments

This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan, to T.N. (number 17500430), and Grants-in-Aid of “Ground Research Announcement for Space Utilization” promoted by the Japan Aerospace Exploration Agency and Japan Space Forum, to T.N.

Footnotes

  • Published ahead of print at http://diabetes.diabetesjournals.org on 29 June 2007. DOI: 10.2337/db06-1768.

  • Additional information for this article can be found in an online appendix at http://dx.doi.org/10.2337/db06-1768.

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

    • Accepted June 27, 2007.
    • Received December 24, 2006.

REFERENCES

| Table of Contents

This Article

  1. Diabetes vol. 56 no. 10 2511-2522
  1. Online-Only Appendix
  2. All Versions of this Article:
    1. db06-1768v1
    2. 56/10/2511 most recent