Type 2 Diabetes Impairs Insulin Receptor Substrate-2-Mediated Phosphatidylinositol 3-Kinase Activity in Primary Macrophages to Induce a State of Cytokine Resistance to IL-4 in Association with Overexpression of Suppressor of Cytokine Signaling-3¹

Type 2 diabetes (T2D)³ and obesity (diabesity) are often associated with an inflammatory profile characterized by persistent subclinical chronic inflammation (1, 2, 3). In addition, accumulating evidence suggests that chronic inflammation contributes to the pathogenesis of T2D (1, 3, 4, 5, 6, 7, 8, 9, 10). A variety of proinflammatory markers and acute phase reactants (i.e., IL-1β, TNF-α, IL-6, sialic acid, serum amyloid, C-reactive protein, cortisol) are consistently elevated in individuals with T2D (6, 11, 12, 13, 14, 15) focusing research attention in this area, leading Crook (13) and Pickup (14) to advance the hypothesis that T2D is a disease of the innate immune system. Inflammation, however, is a homeostatic process that involves both pro- and anti-inflammatory systems. The relative role of anti-inflammatory processes in diabetes is less studied and less clear. Normal counter-regulation of proinflammation involves anti-inflammatory cytokines, such as insulin-like growth factor-1 (IGF-1), IL-10, IL-13, and IL-4. These anti-inflammatory agents reduce secretion of proinflammatory cytokines by macrophages and stimulate the production of a number of anti-inflammatory molecules including IL-1RA (16, 17), IL-1R2 (18), and soluble TNF receptors (19). We have reported that db/db mice exhibit impaired recovery from LPS-induced neuroinflammation (20), and that IGF-1 is not as effective in suppressing the acute inflammatory response induced by LPS in db/db mice (21). These findings suggest that suppression of inflammation by anti-inflammatory cytokines is impaired in T2D.

Macrophage activation is a central component to the inflammatory process resulting in the elaboration of proinflammatory cytokines and induction of the acute phase response. Macrophage activation has repeatedly been shown in both human and animal studies of T2D to be altered. In human patients with T2D, circulating monocytes have increased expression of CD14 (22), which is the coreceptor for LPS. The scavenger receptor, CD36, is also up-regulated in macrophages in diabetic conditions (23). We have shown that peritoneal macrophages from db/db mice elaborate more IL-1β in response to LPS and fail to appropriately up-regulate brain IL-1RA and IL-1R2 mRNA (20). Interestingly, both IL-1RA and IL-1R2 are inducible by IGF-1, IL-10 (18), and IL-4 (18, 23, 24). We have also shown that the human monocytic cell line (U937) cultured under chronic type 2 diabetic conditions of high glucose and insulin has impaired IL-4 signaling (25).

IL-4 is an important anti-inflammatory cytokine that directs macrophages toward a phenotype that is characterized by the elaboration of other anti-inflammatory molecules, like IL-10, IL-1RA, and IL-1R2 (26, 27). Additionally, IL-4 shifts the inflammatory balance by inhibiting the secretion of the proinflammatory cytokines IL-1β, TNF-α, and IL-6 from macrophages (28, 29, 30). Importantly, a disturbance of the anti-inflammatory response could be a critical component of the chronic inflammation found in T2D. The relative contribution, however, of disrupted anti-inflammatory cytokine function remains poorly understood. IL-4 binds its receptor, which is expressed on myeloid cells (33, 34). Upon binding, heterodimerization of IL-4 receptor α- and γc-chains occurs, resulting in activation of JAK1 and JAK3 (35). One or both of these kinases phosphorylate the IL-4 receptor α-chain on tyrosine residues, which allows the binding of various Src homology (SH)2 domain-containing proteins to interact with the receptor. The two primary signaling pathways activated by IL-4 are the JAK/STAT-6 and IRS-2/PI3K pathways. The importance of the IRS-2/PI3K pathway to IL-4 function is not well defined.

The suppressor of cytokine signaling (SOCS) family of proteins has emerged as a potential bridge between T2D and altered cytokine function. There are eight SOCS members, CIS and SOCS1–7 (36, 37, 38), and each contains a central SH2 domain, an N-terminal variable length sequence, and a C-terminal 40 aa SOCS box (37). SOCS proteins bind to tyrosine-phosphorylated receptors, including the insulin and IL-4 receptors, via their SH2 domain where they act as negative regulators of hormone/cytokine signaling (39). Mooney and colleagues (40) first identified that IL-6 induces the expression of SOCS proteins that block insulin signaling by disrupting the formation of signaling complexes at the insulin receptor. IL-6 is well known to be chronically elevated in T2D and has garnered interest for its potential predictive value in determining individuals likely to develop T2D. Importantly, SOCS-3 binding of the insulin receptor does not alter the phosphorylation state of the receptor or its kinase activity. Rather, this association prevents downstream signaling through IRS-1 and IRS-2 (39). Cai and colleagues (10) found that diet-induced obesity and genetic diabetes (ob/ob) in mice resulted in both insulin resistance and SOCS protein up-regulation in the liver. In this study we show, for the first time, that the IRS-2 signaling arm of the IL-4 pathway is required for IL-4-dependent up-regulation of IL-1RA in primary macrophages, and is disrupted by type 2 diabetic conditions. We show that IL-4-dependent expression of IL-1RA requires IRS-2-mediated PI3K activity, and the loss of IL-4 function in db/db mice is accompanied by chronic overexpression of SOCS-3.

All reagents and chemicals were purchased from Sigma-Aldrich except as noted below. ECL detection reagents, Hybond-ECL nitrocellulose, and protein G-Sepharose beads were purchased from Amersham Biosciences. Primer pairs (Qiagen); SYBR, green PCR master mix and MicroAmp optical 96-well reaction plates (PE Applied Biosystems); TRIzol (Invitrogen Life Technologies); Superscript III RNase H⁻ reverse transcriptase, 10 mM dNTP mix, and oligo (dT) primers (Invitrogen Life Technologies); RNasin Rnase inhibitor (Promega). Phosphotyrosine (pY) (catalog no. 05-321), IRS-2 (catalog no. 06-506), p-STAT-6 (catalog no. 06-937), JAK-1 (catalog no. 06-272), and PI3K p85 (catalog no. 06-497) Abs were purchased from Upstate Biotechnology. PE-conjugated (catalog no. 552509) and purified (catalog no. 551853) mAbs directed against the IL-4R were purchased from BD Pharmingen. SOCS-3 Ab (catalog no. sc-9023) was purchased from Santa Cruz Biotechnology. STAT-6 (catalog no. PA-ST6) and actin (catalog no. I-19) Abs were purchased from R&D Systems, and the Ab toward p-JAK-1 (catalog no. AB3850) was ordered from Chemicon International. FCS (0.05 ng/ml, 0.48 endotoxin unit per milliliter) and recombinant murine IL-4 were purchased from Atlanta Biologicals. IL-1RA and IL-1R2 ELISA standard, IL-1RA and IL-1R2 polyclonal Abs, and biotinylated IL-1RA and IL-1R2 Abs were purchased from R&D Systems. Tetramethylbenzidine substrate solution (n301) was purchased from Endogen. Maxisorp-coated 96-well ELISA plates were purchased from Nalge Nunc International. Raw 264.7 macrophages were purchased from American Type Culture Collection. Peptides Y608 (AAVALLPAVLLALLAPGYMPMSP) and pY608 (AAVALLPAVLLALLAPGpYMPMSP) were synthesized by the University of Illinois Biotechnology Center (Urbana, IL).

All animal care and use was conducted in accordance with the Guide for the Care and Use of Laboratory Animals as we described (20, 21). Eight to 12-wk-old B6.Cg-M^+/+Lepr^db (db/+) and B6.Cg-+Lepr^db/+Lepr^db (db/db) mice were bred in the Animal Care Facility at the University of Illinois from mice purchased from The Jackson Laboratory. Mice were housed in standard shoebox cages and allowed pelleted food (NIH 5K52, LabDiet; Purina Mills) and water ad libitum in a temperature (72°F) and humidity (45–55%) controlled environment with a 12-h light/dark cycle (7:00 a.m. to 7:00 p.m.).

Cytokine protein levels were assayed using either a multiplex mouse cytokine bead array system (Bio-Rad) according to manufacturer’s instructions or by ELISA. As we have described (20), 96-well microtiter plates were coated with purified anti-mouse IL-6, IL-1RA, and IL-1R2 polyclonal Ab (2 mg/ml in PBS) overnight at 25°C, and then blocked with 4% BSA containing PBS for 1 h. Standards, samples, and biotin-labeled anti-mouse IL-6, IL-1RA, and IL-1R2 Ab (62.5 ng/ml) were added to the appropriate wells and allowed to incubate at 25°C for 2 h. Plates were washed and resolved with HRP-conjugated streptavidin, 3,3′,5,5′-tetramethylbenzidine substrate solution (30 min at 25°C) and 0.18 M H₂SO₄ stop buffer. Absorbance was measured on an OPTImax tunable microplate reader (Molecular Devices) at 450 nm to 550 nm. The concentration of IL-6, IL-1RA, and IL-1R2 in the samples was determined by reference to a standard curve.

As we have described (20), mice were sacrificed by CO₂ asphyxiation, peritoneal lavage fluid was collected by washing (2×) the peritoneum with 5 ml of ice-cold growth medium (RPMI 1640 medium supplemented with 10% FCS, 1 g/L glucose, 2 g/L sodium bicarbonate, 110 mg/L sodium pyruvate, 62.1 mg/L penicillin, and 100 mg/L streptomycin, 10 mM HEPES (pH 7.4)). For ex vivo experiments, peritoneal cells were resuspended in 5 ml of hypertonic RBC lysis buffer (142 mM NaCl, 1 mM KHCO₃, and 118 mM Na-EDTA (pH 7.4)) at room temperature for 4 min. An equal volume of cold growth medium was added followed by cell washing and resuspension in 37°C growth medium. Cells were plated at 5 × 10⁵ cells/ml. After 30 min, nonadherent cells were removed by washing twice with growth medium. Remaining cells were at least 80% macrophages confirmed by CD11b staining and morphology (50).

As we have described (21), total RNA from isolated macrophage samples was extracted in TRIzol reagent. All reverse transcriptase reactions were conducted in a Stratagene Robocycler Gradient 96. All RNA samples from a single experimental group were reverse transcribed simultaneously to minimize interassay variation associated with the reverse transcription reaction.

Real-time PCR was performed as described (21). Primer Express software (PE Applied Biosystems) was used to design appropriate primer pairs. The primer sequences used were as follows: β-actin forward, GGCGCTTTTGACTCAGGATT; β-actin reverse, GGGATGTTTGCTCCAACCAA; IL-6 forward, CCAGAAACCGCTATGAAGTTCCT; IL-6 reverse, CACCAGCATCAGTCCCAAGA. Real-time PCR was performed on Applied Biosystems Prism 7700 (PE Applied Biosystems) by using the SYBR Green PCR Master Mix. To normalize gene expression, a parallel amplification of endogenous and target genes was performed with SYBR Green reagents. Reactions with no reverse transcription and no template were included as negative controls. Relative quantitative evaluation of target gene levels was performed by comparing ΔC_t, where C is the threshold concentration.

Macrophages were harvested as described above. Cells were washed once in wash buffer (Dulbecco’s PBS containing 0.5% BSA without calcium and magnesium). PE-conjugated IL-4R Ab at 10 μg/ml/test was added to 1 × 10⁶ cells in 100 μl of wash buffer then incubated on ice for 15 min followed by washing with wash buffer. Fluorescence was detected on an Epics XL flow cytometer (Beckman Coulter).

PI3K assays were performed as previously described (41). Cells were treated as indicated and lysed in ice-cold lysis buffer (1% Triton X-100, 100 mM NaCl, 50 mM NaF, 1 mM PMSF, 2 μg/ml aprotinin, 2 μg/ml leupeptin, 2 mM sodium orthovanadate, 50 mM okadaic acid, and 50 mM Tris (pH 7.4)). IRS-2-associated PI3K was immunoprecipitated from clarified lysates with anti-IRS-2. Kinase assays were performed in a buffer containing 0.33 mg/ml l-α-phosphatidylinositol, 7.5 mmol/L MgCl₂, 0.4 mmol/L EGTA, 0.4 mmol/L NaPO₄, 7.5 μmol/L (γ-³²P)ATP (0.48 MBq/nmol), and 20 mmol/L HEPES (pH 7.1) for 15 min. Kinase reactions were stopped by addition of 15 μl of 4 N HCl. Phospholipids were extracted with 1:1 chloroform/methanol and resolved on silica gel plates by thin layer chromatography in chloroform/methanol/ammonium hydroxide (4 mol/L) (75:58:17). γ-³²P was detected by phosphorimaging on a Typhoon PhosphorImager System (Molecular Dynamics using Molecular Dynamics PhosphorImage software for analysis.

Macrophages were lysed in ice-cold lysis buffer. Proteins were resolved by SDS-PAGE (250 μg/lane) under reducing conditions in 4–20% gradient gels and electrotransferred to nitrocellulose. Immunoreactive proteins were visualized by ECL, followed by autoradiography and densitometry. For blots requiring multianalysis, nitrocellulose membranes were incubated at 100°C for 10 min in stripping buffer (2% SDS, 0.704% (v/v) 2-ME, and 6.25 mmol/L Tris-HCl (pH 6.8)). Membranes were washed in 0.05% Tween 20 and Tris-buffered saline (pH 7.4). Western analysis was again performed after membranes were reblocked with 5% BSA and Tris-buffered saline (pH 7.4) for 1 h. For illustrative purposes and space consideration throughout the manuscript, a representative gel image from multiple independent (3, 4) experiments is presented. The numerical data reported in Results is based on the summary of densitometric analysis of all blots followed by statistical analysis. Statistical significance is indicated where appropriate at p < 0.05.

Data are presented as mean ± SEM. Where indicated, experimental data were analyzed either by Student’s t test for comparison of means, or by two-factor ANOVA using Excel (Microsoft). Statistical significance was denoted at p < 0.05.

T2D is associated with subacute chronic inflammation (1, 2, 3), and IL-6 is chronically elevated in both human T2D and rodent models of diabesity. Table I shows that increased expression of proinflammatory cytokines were detected in the peritoneal fluid of db/db mice compared with heterozygote control (db/+) animals. Basal levels of IL-1α, IL-17, IFN-γ, and TNF-α were not detected in db/+ mice but were significantly increased in db/db mice. IL-1β, IL-2, and IL-6 were increased 2.78-fold, 1.31-fold and 6.75-fold, respectively. In addition, IL-6 showed the greatest absolute increase (8809 pg/ml). To determine whether macrophages from db/db mice contributed directly to this higher basal level of IL-6, non-elicited resident peritoneal macrophages from db/+ and db/db mice were isolated and IL-6 mRNA levels were analyzed by real-time PCR. Fig. 1 A shows that the relative transcript expression of IL-6 mRNA was 2.6-fold greater in db/db macrophages when compared with db/+ controls (2.6 ± 0.33 vs 1 ± 0.21). In addition, IL-6 protein secretion was increased nearly 100% in db/db mouse macrophages when compared with db/+ controls (442 ± 68 vs 223 ± 28 pg/mg total protein). Taken together, these findings indicate that proinflammatory cytokine expression is augmented in the db/db mouse and IL-6 is a sizable component of that increase.

We have previously shown that db/db mice have reduced spleen and brain expression of IL-1RA and IL-1R2 in response to i.p. LPS (20). To determine whether macrophages from db/db mice failed to up-regulate IL-1RA and/or IL-1R2 in response to IL-4, non-elicited resident peritoneal macrophages from db/+ and db/db mice treated ex vivo with IL-4 (10 ng/ml) were examined. Fig. 2,A shows that in response to 2 h of exposure to IL-4 db/+ macrophages increased IL-1RA 2.8-fold (150 ± 12.2 vs 425.9 ± 87.6 pg/mg protein) and IL-1R2 2.5-fold (15.2 ± 2.6 vs 38.4 ± 3.9 pg/mg protein). Macrophages from db/db mice failed to increase elaboration of IL-1RA (68.5 ± 6.1 vs 107.2 ± 26.0 pg/mg protein) and IL-1R2 (11.1 ± 5.1 vs 17.6 ± 6.8 pg/mg protein). In addition, basal levels of IL-1RA produced by db/+ macrophages were 2.2-fold greater than those from db/db mice (150.0 ± 12.2 vs 68.5 ± 6.1 pg/mg protein). As a control, we examined IL-4 levels in peritoneal fluid from db/+ and db/db mice. Fig. 2 C demonstrates that peritoneal levels of IL-4 were not significantly different in db/+ and db/db mice (not detectable vs 4.64 ± 3.38 pg/ml). Additionally, in response to the innate immune stimulant, LPS (10 ng/ml, 2 h), IL-4 levels increased comparably in db/+ and db/db mice. Taken together, these findings indicate that IL-4-dependent up-regulation of the IL-1β antagonists IL-1RA and IL-1R2 is blocked in db/db mice and that this defect is not simply due to differential expression of IL-4, as intraperitoneal IL-4 levels are unaffected by the diabetic condition of these animals.

The role of the IRS-2/PI3K pathway in IL-4 signaling is ill-defined. To determine whether PI3K activity was required for IL-4-dependent expression of IL-1RA, PI3K inhibition studies were performed. Fig. 3,A shows that when resident peritoneal macrophages from nondiabetic, db/+, mice were incubated ex vivo with the PI3K inhibitor, wortmannin (1 μM), for 30 min before treatment with IL-4 (10 ng/ml for 2 h), wortmannin attenuated IL-4-dependent IL-1RA production by 43% (324.8 ± 36.9 vs 186.4 ± 28.8 pg/mg protein). Wortmannin, itself, had no impact on IL-1RA elaboration by macrophages unstimulated with IL-4. Fig. 3,B demonstrates that IL-4 (10 ng/ml) induced IRS-2-associated PI3K activity was inhibited completely by wortmannin as demonstrated in Fig. 3 A. Taken together these findings indicate that IL-4-dependent up-regulation of IL-1RA is dependent on PI3K activity.

Tyrosine phosphopeptides in the YMXM or YXXM motifs bind to the SH2 domains of the p85 subunit of PI3K and block its ability to bind to IRS-1/2 (42). To determine whether IRS-2 associated PI3K activity was required for IL-4-dependent IL-1RA production in macrophages, we designed a cell permeable tyrosine-phosphorylated peptide analog to the PI3K binding site Y608 of mouse IRS-1 and Y649 of mouse IRS-2 (43). The p85 SH2 domain reactive peptide sequence was GpYMPMSP (pY608), which we have shown in vitro to block IRS-1/PI3K association (42). The nonphosphorylated analog GYMPMSP (Y608) was used as a control. Fig. 4,A demonstrates that when isolated resident peritoneal macrophages from db/+ mice were pretreated with 10 μM pY608 for 2 h, IL-4-dependent IL-1RA production (as measured in Fig. 3) was inhibited (172.4 ± 5.96 vs 116.6 ± 8.33). Y608 peptide had no impact on IL-4-dependent production of IL-1RA (172.4 ± 5.96 vs 167.5 ± 5.59). Fig. 4,B demonstrates that IL-4-dependent IRS-2 associated PI3K activity was completely blocked by pY608 as used in Fig. 4 A. As expected, the pY608 and Y608 inhibitors had no effect IL-1RA expression or IRS-2-associated PI3K activity when used without IL-4 (data not shown). Taken together these findings indicate that IL-4-dependent up-regulation of IL-1RA is dependent on IRS-2/PI3K association.

We have shown in vitro that diabetic conditions can interfere with cytokines that signal through the IRS/PI3K pathway (25, 44). To confirm that interaction between IRS-2 and PI3K was reduced in macrophages isolated from db/db mice, macrophages were treated ex vivo with 10 ng/ml IL-4 for 15 min. Fig. 5,A (top panel) shows that p85 mass that coimmunoprecipitates with IRS-2 was reduced by 60% in db/db macrophages (p < 0.05 vs db/+ IL-4-treated macrophages). Analysis of IRS-2 mass (bottom panel) confirmed equal precipitation mass. To determine whether IL-4-dependent tyrosine phosphorylation of IRS-2 was blunted in db/db mice, resident peritoneal macrophages from db/db mice treated ex vivo with IL-4 (10 ng/ml, 15 min) were examined by Western analysis. Fig. 5,B (top panel) shows that in db/+ mice IL-4 induced tyrosine phosphorylation of IRS-2 and that in db/db mice IL-4-dependent tyrosine phosphorylation of IRS-2 was reduced by nearly 70% (p < 0.05 vs db/+ IL-4-treated macrophages). Fig. 5,B also demonstrates that whole cell lysates probed for JAK1 (middle panel) and STAT6 (bottom panel) tyrosine phosphorylation showed no noticeable difference in db/+ and db/db macrophages. Fig. 5,C shows that expression of IL-4α receptors is not significantly different in peritoneal macrophages from db/+ and db/db mice. In addition, total cellular mass of JAK1, IRS-2, and STAT6 were comparable between db/+ and db/db mice peritoneal macrophages (Fig. 5 D). JAK-3 was not expressed in consistently detectable amounts (data not shown). Taken together, these data indicate that resident peritoneal macrophages from db/db mice are resistant to the ability of IL-4 to induce IRS-2 tyrosine phosphorylation and subsequently bind PI3K.

As shown above, IL-4-dependent tyrosine phosphorylation of IRS-2 is inhibited in db/db mice. We have already demonstrated chronic IL-6 elevation (Table I) in db/db mice and SOCS-3 is known to be an IL-6-inducible protein that can impair insulin signaling. To determine whether SOCS-3 might be associated with disrupting IL-4 signaling in peritoneal macrophages from db/db mice, basal protein expression was determined. Fig. 6,A demonstrates that SOCS-3 basal expression was, in fact, increased 2.7-fold in db/db mice when compared with db/+ mice (p < 0.05 vs db/+ macrophages). To determine, whether IL-4-dependent receptor activation was altered in db/db macrophages, resident peritoneal macrophages from db/+ and db/db mice were treated ex vivo with IL-4 (10 ng/ml, for 15 min). Fig. 6,B (top panel) shows that IL-4-dependent tyrosine phosphorylation of the IL-4 receptor was not noticeably different in db/db and db/+ mice. Blots were stripped and reprobed for IL-4 receptor to confirm equal protein mass in immunoprecipitates (middle panel). Importantly, Fig. 6 B (bottom panel) shows that IL-4-dependent SOCS-3/IL-4 receptor association was increased by 2.5-fold in db/db mice when compared with db/+ mice (p < 0.05 vs db/+ IL-4-treated macrophages). Of note, we did not detect measurable SOCS-3 mass in IRS-2 or JAK-1 immunoprecipitates (data not shown), indicating that SOCS-3 specifically interacts with the tyrosine-phosphorylated IL-4 receptor, and SOCS-1 did not coprecipitate with the IL-4 receptor (data not shown). These findings are consistent with previous reports that SOCS-3 is elevated in diet-induced and ob/ob mouse models of T2D (10). Taken together, these data indicate SOCS-3 is chronically up-regulated in db/db mouse macrophages and that IL-4-dependent binding of SOCS-3 to the activated IL-4 receptor is increased. Therefore, SOCS-3 is a potential mediator of impaired IL-4 function in T2D.

T2D is associated with alterations in the innate immune system that appear to impact the complications and/or comorbid conditions associated with diabetes including delayed wound healing, accelerated atherosclerosis, and retinopathy. Prospective studies have indicated that alterations in inflammatory status are also seen before the onset of overt diabetes. For example, C-reactive protein and IL-6 predict the development of T2D in human subjects (15). IL-6 has recently received much attention for its relationship to insulin resistance. IL-6 is chronically elevated in both human T2D and in mouse models of diabetes, as we show in Table I. In 2002, Senn and colleagues (46) demonstrated that IL-6 induced insulin resistance in hepatocyte cell cultures, and in a subsequent study they showed the same effect in vivo, as infusion of exogenous IL-6 induced hepatic insulin resistance in mice (47). Conversely, IL-6 depletion in the ob/ob mouse model of T2D improved insulin action (48). Although work on the classically described proinflammatory cytokines in T2D has been vigorous, much less is known about the relative role of anti-inflammatory cytokines.

In vivo studies in mice demonstrate that the anti-inflammatory molecules IGF-1, IL-10, and IL-4 are able to attenuate the acute inflammatory responses induced by LPS (49, 50, 51). A critical function of IL-4 in relationship to LPS-induced proinflammation is to cause expression and secretion of IL-1RA from macrophages to counteract LPS-induced IL-1. We have previously reported that db/db mice fail to normally suppress IL-1β-dependent immunobehavioral responses induced by either IL-1β or LPS (25). In this study we show, in Fig. 2, that macrophages isolated from db/db mice do not elaborate IL-1RA or IL-1R2 in response to IL-4 in contrast to control mice. These data, for the first time, demonstrate that the main anti-inflammatory function of IL-4 toward IL-1 is absent in T2D. Fig. 2 also demonstrates that LPS-dependent IL-4 production was normal in db/db mice indicating that the delayed immunobehavioral recovery from LPS we observed in db/db mice previously (26) was due to deficient IL-1RA and IL-1R2 production and not failure to express IL-4.

The PI3K pathway is critical to insulin signaling, and in T2D, insulin-stimulated PI3K activity is reduced. IL-4 also activates PI3K, but its role in IL-4 action is not well understood. Furthermore, almost nothing is known about IL-4-dependent PI3K activation in macrophage function. Fig. 3 shows that inhibition of PI3K with wortmannin significantly inhibited IL-4-dependent elaboration of IL-1RA from peritoneal macrophages. Because pharmacologic inhibitors can lack specificity, and primary cells, especially macrophages, can be difficult to transfect, we developed another method to confirm that PI3K was required for IL-4-dependent production of IL-1RA. We synthesized a cell permeable tyrosine phosphopeptide homologous to tyrosine 608 of IRS-1 and 649 of IRS-2. The sequence of this peptide was AAVALLPAVLLALLAPGpYMPMSP where AAVALLPAVLLALLAP was analogous to the h-region of Kaposi’s fibroblast growth factor and provided cell permeability (52). The phosphotyrosine residue in the GpYMPMSP motif is known to interact strongly and specifically with the PI3K p85 subunit (53, 54) and block IRS/PI3K interaction (43). Fig. 4 shows that GpYMPMSP, but not GYMPMSP, significantly attenuated IL-4-dependent IL-1RA expression and blocked IL-4-dependent activation of PI3K, confirming the wortmannin experiments in Fig. 3. They also demonstrate the need for PI3K-SH2 domain interactions and IRS-2/PI3K association for IL-4-dependent production of IL-1RA in peritoneal macrophages. These findings are the first to show that a prevalent defect (impaired IRS-mediated signaling) of T2D has negative consequences on the functional anti-inflammatory role of IL-4.

To explore a potential mechanism by which T2D may block IL-4-dependent IL-1RA expression in db/db mice, the IL-4 signaling cascade was examined. Fig. 5,A shows that IL-4-activated IRS-2 PI3K p85 association was markedly reduced in isolated peritoneal macrophages from db/db mice. This finding was consistent with our previous work that found that IL-4-activated IRS-2-associated PI3K activity is decreased in diabetic mice (26). Fig. 5 B demonstrates that IL-4-dependent tyrosine phosphorylation of IRS-2 was markedly reduced whereas IL-4-dependent JAK1 activation and STAT-6 tyrosine phosphorylation were unimpaired by diabetes. Importantly, surface expression of the IL-4 receptor and mass of IRS-2, JAK1, and STAT-6 were similar in db/db and db/+ peritoneal macrophages. Taken together, these results demonstrate that the primary IL-4-associated signaling deficit present in db/db peritoneal macrophages was blunted tyrosine phosphorylation of IRS-2 and concomitant loss of IRS-2/PI3K association. These finding were consistent with our previous work (26), which showed that in an in vitro model of T2D using the human monocytic cell line U937 and high and low glucose and insulin conditions, IL-4-dependent tyrosine phosphorylation of IRS-2 was inhibited and consequent PI3K p85 association reduced. In this previous study, we also found that IL-4-dependent activation of the JAK/STAT pathway in U937 cells was unaffected by high glucose and insulin conditions (26). These findings expand on our previous study by identifying a relevant and important functional consequence of impaired IRS-2 mediated IL-4 signaling.

Finally, Table I demonstrates that IL-6 was significantly increased in the peritoneal fluid of db/db mice. Importantly, Fig. 1 indicates that resident peritoneal macrophages from db/db mice contribute to this increased basal IL-6 production. Chronic increased expression of IL-6 is linked to hepatic insulin resistance due to SOCS protein up-regulation (40, 46, 47, 52). The consensus SOCS-3 binding motif was identified by De Souza and colleagues (53) as pY-(SAVYF)-hydrophobic-hydrophobic, which is highly homologous to Y759 of gp130. This finding predicted that SOCS-3 binding sites existed on other tyrosine-phosphorylated receptors. In fact, SOCS-3 has now been shown to associate with a variety of cytokine and hormone receptors, including the insulin receptor and IL-4 receptor (36, 39). As predicted based on the basal up-regulation of SOCS-3, Fig. 6 shows that IL-4-activated IL-4 receptors in db/db mouse peritoneal macrophages bound more SOCS-3 than db/+ macrophages. In addition, db/db mouse macrophages expressed nearly 3 times the amount of SOCS-3 than macrophages isolated from db/+ mice. This finding is important because the SOCS proteins can bind to the IL-4 receptor and block subsequent receptor-mediated interactions (36). Therefore, like in IL-6-dependent SOCS-3-mediated blockade of insulin signaling (40), we suggest that SOCS-3 binds the activated IL-4 receptor preventing IRS-2 tyrosine phosphorylation and formation of IRS-2/PI3K p85 complexes. Our finding are consistent with Ueki and colleagues (40) because they found SOCS-3 inhibited downstream insulin signaling through IRS-1 and IRS-2 without altering insulin receptor kinase activity, and we found no impact of SOCS-3 on IL-4 receptor tyrosine phosphorylation or JAK1 activation. Therefore, like the findings of Ueki and colleagues, SOCS-3, in the IL-4 signaling system appears to mediate loss of IRS-2-associated PI3K activity that subsequently prevents IL-4 from fully inducing diabetic macrophage IL-1RA expression. It is important to note that SOCS-3 has been demonstrated to bind JAK proteins and inhibit subsequent activation of STAT proteins (54, 55). However, we did not detect any coassociation of SOCS-3 with either JAK-1 or STAT-6 (data not shown). This is not entirely surprising, however, given that SOCS inhibition is reported to be highly selective and cell-type specific (56). Also, Nicholson et al. (57) reported that SOCS-3 was unable to inhibit JAK kinase activity in vitro, and in a subsequent study, they showed that SOCS-3 displayed a high affinity for the gp130 receptor compared with weak or undetectable affinity for JAK or STAT proteins (58). Our data are consistent with these reports, as Fig. 6 B demonstrated, in our model, no change in JAK/STAT phosphorylation when complexed with SOCS-3 at the IL-4R. Finally, little data are available in SOCS-3 transgenic models, because gene deletion is embryonic lethal. SOCS-3 conditional KO mice are viable, and they are characterized by a number of inflammatory and metabolic problems. Interestingly, SOCS-3-deficient adipocytes do not develop TNF-induced insulin resistance (59), which is consistent with our findings that overexpression of SOCS-3, as a result of T2D conditions, appears to interfere with IL-4 receptor signaling. Although further investigation is needed to completely identify the precise mechanism by which this happens.

We thank Dr. Robert Mooney for expert advice and insight during the revision of this manuscript.

The authors have no financial conflict of interest.