Adiponectin Is a Negative Regulator of NK Cell Cytotoxicity¹ Free

Adipose tissue is currently considered to be a hormonally active system, because adipokines that are secreted from adipocytes are transported into the bloodstream and are associated with various aspects of metabolic control (1). These adipokines, which include IL-6, TNF-α, leptin, resistin, and adiponectin, play important roles in both the immune system and metabolism. IL-6 and TNF-α were first studied as components of the immune system and were later reported to act as adipokines, whereas leptin and adiponectin were isolated in adipocytes, and their roles in the immune system were investigated later (2, 3, 4, 5). The role of leptin in the regulation of the immune system has been extensively examined. Leptin enhances GM-CSF production in peritoneal macrophages (6) and mediates both proliferative and antiapoptotic activities in a variety of cell types, including T cells (7) and monocytes (8). Leptin has been proposed to serve as a potential linker between obesity and chronic inflammation (9).

Adiponectin is produced exclusively by white adipocytes, but, paradoxically, its level is lower in obese than in lean individuals (10). It is known that a decrease in adiponectin levels is associated with diabetes, and that recovery of adiponectin level ameliorates insulin resistance through an increase in free fatty acid oxidation and glucose uptake (11). There is a growing body of evidence to suggest that adiponectin also has anti-inflammatory activity. It negatively regulates the growth of myelomonocytic progenitors and the functions of macrophages (12) and inhibits the endothelial NF-κB signaling pathway (13), foam cell formation by reducing the accumulation of lipids in human macrophages (5), and vascular inflammation in adiponectin knockout mice (14). Adiponectin also causes an increase in IL-10 mRNA expression in human macrophages (15). Although its role as an immune regulator has been studied, little is known concerning the function of adiponectin in NK cell activity. The findings presented in this report show that adiponectin suppresses IL-2-induced NK cell cytotoxicity by inhibiting the NF-κB signaling pathway and down-regulating IFN-γ, Fas ligand (FasL),³ and TRAIL expression. As a result, we report a new function of adiponectin as a regulator of NK activity.

Human NK 92 cells were cultured in α-MEM supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, 100 U/ml streptomycin, 10 mM HEPES, and 100 U/ml IL-2, and primary mouse NK cells were maintained with RPMI 1640 medium containing 10% FBS, l-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin, and 10 mM HEPES (Invitrogen Life Technologies) at 37°C in a humidified 5% CO₂ incubator.

Escherichia coli strain BL21 (λDE3) was transformed with pET28-adiponectin (Novagen) containing adiponectin cDNA without a signal sequence coding region. The expression of adiponectin was induced by treatment with 0.5 mM isopropyl-1-thio-β-d-galactopyranoside for 3 h at 37°C. After sonication and centrifugation, the inclusion body was solubilized in guanidine buffer (50 mM Tris (pH 8.0) and 7 M guanidine). The refolding conditions for adiponectin were determined according to the Pro-Matrix protein refolding guide (Pierce). Refolded adiponectin in refolding buffer (55 mM Tris, 21 mM NaCl, 0.88 mM KCl, 1 mM DETA, 2 mM reduced glutathione, and 0.2 mM glutathione disulfide (pH 8.2)) was dialyzed against 20 mM Tris-HCl (pH 8.0), filtered, loaded onto a Q-Sepharose column, then eluted with 0.5 M NaCl in a stepwise manner. Fractions containing pure adiponectin were dialyzed against 20 mM Tris-HCl (pH 8.0), and endotoxin contaminants were removed by extraction with Triton X-114. The level of endotoxin was determined with a Limulus amebocyte lysate QCL-1000 kit (Cambrex). The purity of the adiponectin exceeded 95%, as evidenced by SDS-polyacrylamide gel staining, and the level of endotoxin was <0.128 endotoxin units/mg in the purified adiponectin.

For NK cell isolation, female C57BL/6 mice were purchased from Polars International. An NK cell isolation kit (Miltenyi Biotec) was used to separate NK cells from splenocytes. Briefly, isolated splenocytes were washed twice with BSA-PBS (0.5% BSA and 0.2 mM EDTA in PBS), incubated with a biotin-Ab mixture (anti-CD4, -CD5, -CD8a, -CD19, -Ly-6G, and -Ter¹¹⁹) at 4°C for 10 min, then incubated with the microbeads conjugated to a monoclonal anti-biotin Ab at 4°C for 15 min. After washing twice, the labeled cells were applied to a MACS column (CS⁺; Miltenyi Biotec). The column was rinsed with 30 ml of BSA-PBS to flush out unbound NK cells. For analysis of the purity of the cells using flow cytometry, negatively isolated NK cells were stained with goat anti-mouse NK1.1-FITC. The purity of the NK cells exceeded 95%.

After the stimulation period, NK cells were harvested and washed twice. Viable NK cells were counted using the trypan blue dye exclusion method, and an equal number of viable NK cells was used as effectors in a ⁵¹Cr release assay. K562 and YAC-1 target cells were incubated with 1.5 μCi of ⁵¹Cr for 1 h at 37°C, allowing ⁵¹Cr to enter target cells. The labeled target cells were then washed twice and dispensed at a concentration of 5000 target cells/well in a round-bottom, 96-well plate. The effectors were distributed in triplicate at E:T cell ratios from 200:1 to 1:1 and incubated for 4 h at 37°C in 5% CO₂. After the incubation, the supernatants were harvested, and the released radioactivity was counted with a scintillation counter (RACTOBETA; LKB Instruments). The percentage of specific activity was calculated using the following equation: % specific release = ((experimental release − spontaneous release)/(maximum release − spontaneous release)) × 100%.

The dsDNA oligonucleotides containing the NF-κB element (5′-agttgaggggactttcccaggc-3′) were labeled with [γ-³²P]ATP (ICN) using T4 polynucleotide kinase (New England Biolabs) and purified using a Sephadex G-50 column (Roche). Ten micrograms of nuclear extracts was incubated with ³²P-labeled NF-κB oligonucleotides in the binding buffer (100 mM KCl, 30 mM HEPES, 1.5 mM MgCl₂, 0.3 mM EDTA, 10% glycerol, 1 mM DTT, 1 mM PMSF, 1 μg/ml aprotinin, and 1 μg/ml leupeptin) for 15 min at 4°C. To reduce nonspecific bands, 1 μg of poly(dI-dC) was added to the reaction. DNA binding activity was observed by separating the unbound probe on a 4.8% polyacrylamide gel in 0.5× TBE (250 mM Tris (pH 8.5), 1.92 M boric acid, and 10 mM EDTA). After the electrophoresis run, the gel was dried, then exposed to an x-ray film.

The pNF-κB-luciferase (pNF-κB-luc) gene reporter vector was purchased from BD Clontech. The pCMV-Renilla luciferase (pCMV-RL) vector was purchased from Promega. Constitutively active AMP-activated protein kinase-1α (AMPK-α1) and dominant negative AMPK-α1 (DN-AMPK-α1) expression plasmids were gifts from Dr. D. Carling (Medical Research Council Clinical Sciences Center, London, U.K.). Human embryonic kidney 293T cells were plated in 12-well plates 1 day before the experiment, and transient transfection of the indicated genes was conducted using the calcium phosphate method. The next day, the cells were treated with adiponectin and 5-aminoimidazole-4-carboxamide-1β-4-ribofuranoside (AICAR), purchased from Toronto Research Chemicals, for a period of 24 h. The constitutively active pCMV-Renilla plasmid, which expresses Renilla luciferase, was used to normalize for transfection efficiency. The cells were washed twice with PBS and lysed in 300 μl of reporter lysis buffer (Promega). The two luciferase activities were measured using a dual luciferase assay kit (Promega) on a microplate luminometer (Tuner Designs).

Total cell lysates were prepared 24 h after treatment, separated on a 12% SDS-polyacrylamide gel, then transferred to a polyvinylidene difluoride membrane (Millipore). The membrane was blocked with 5% skim milk in TBS (20 mM Tris-HCl (pH 7.5) and 137 mM NaCl) for 30 min, then incubated with phospho-AMPK and phospho-IκB Abs (Cell Signaling Technology) overnight. After washing three times with TBST (0.05% Tween 20, 20 mM Tris-HCl (pH 7.5), and 137 mM NaCl) for 15 min, the membrane was incubated with the anti-rabbit IgG secondary Ab conjugated with HRP. The signal was detected using the ECL system (Amersham Biosciences).

Total RNAs from splenic NK and human NK 92 cells were isolated using RNAzol B (Tel-Test). For the real-time PCR analysis, 3 μg of total RNA was reverse transcribed using Moloney murine leukemia virus reverse transcriptase (Promega) at 37°C for 1 h. PCR for the amplification of various cytokines was then performed. Each amplification cycle consisted of 10 s of denaturation at 94°C, 15 s of annealing at 55°C, and 20 s of extension at 72°C, which was performed on a Rotor-Gene 3000 real-time cycler (Corbett Research). β-Actin was used as the control in the comparative cycle threshold method.

IL-2-, adiponectin-, AICAR-, and compound C- (Calbiochem) treated primary NK cells were stained with anti-FasL and anti-NK1.1 (PK136) mAbs (BD Pharmingen) for 20 min at 4°C. NK cell viability was assessed using an annexin V kit (BD Pharmingen). Briefly, NK cells were washed twice with PBS and stained with annexin V-FITC for 15 min at room temperature in the dark. The stained cells were analyzed by flow cytometry. The resulting data were analyzed using the WinMDI software program.

Mouse IFN-γ production was determined using a mouse IFN-γ ELISA kit (Pierce). Briefly, purified NK cells were treated with the indicated reagents, and the resulting supernatant was used according to the manufacturer’s instructions.

The data are presented as the mean ± SEM and were statistically analyzed using an unpaired t test. A value of p < 0.05 was considered a significant difference.

In the present study the role of adiponectin in NK cytotoxicity was examined. NK cells, purified from splenocytes, were treated with adiponectin and IL-2 for 24 h, washed twice, and cultured with ⁵¹Cr-labeled YAC-1 target cells for 4 h, and the released ⁵¹Cr activity was determined. Interestingly, adiponectin dramatically suppressed IL-2-enhanced NK cytotoxicity in an E:T cell ratio-dependent manner, but did not affect basal spontaneous NK cytotoxicity in the absence of IL-2 (Fig. 1,A). To determine whether adiponectin inhibits the cytotoxicity of NK cells that had been pretreated with IL-2, purified NK cells were pretreated with IL-2 for 4 h, followed by washing twice, and then were incubated with adiponectin alone for an additional 16 h. At the end of the culture period, the amount of ⁵¹Cr released into the culture medium was determined. Although IL-2 was removed 4 h after treatment, adiponectin inhibited IL-2-induced NK cytotoxicity (Fig. 1,B), suggesting that IL-2-triggered signal transduction pathways are blocked by adiponectin. We next examined the issue of whether adiponectin affects IL-2-enhanced NK cell viability, because IL-2 enhances rodent NK cell viability. IL-2 increased NK cell viability, as assessed by annexin V staining. Adiponectin treatment decreased IL-2-increased NK cell viability, but had no effect on basal NK cell viability (Fig. 1 C). Therefore, an equal number of viable NK cells was used to avoid the effect of dead NK cells in the NK cytotoxicity assay.

To examine the dose-dependent effect of adiponectin on the suppression of IL-2-mediated NK cytotoxicity, total splenocytes were treated with various concentrations of adiponectin in the presence or the absence of IL-2 for 24 h (Fig. 1,D). NK cytotoxicity was suppressed in the physiological concentration range of adiponectin. The cytotoxicity of human NK 92 cells was also suppressed by treatment with human recombinant adiponectin (data not shown). It is possible that adiponectin inhibits IL-2-induced NK cell cytotoxicity through binding to IL-2. To test this possibility, rIL-2 and adiponectin were mixed, incubated for 1 h, then precipitated with anti-IL-2 Ab. The resulting precipitates were subjected to Western blot assay using anti-IL-2 and anti-adiponectin Abs, respectively. No binding was observed (Fig. 1 E). We also examined adiponectin receptor expression in NK cells using RT-PCR. The expression of adiponectin receptor 1 was higher than that of adiponectin receptor 2 (data not shown). These results indicate that adiponectin inhibits IL-2-induced NK cell cytotoxicity through binding to its specific receptors without interacting with IL-2.

IL-2 increases NK cell cytotoxicity and proliferation through activation of the NF-κB signaling pathway. We investigated whether adiponectin has an effect on the IL-2-activated NF-κB signaling pathway in NK cells. Purified NK cells were treated with adiponectin in the presence or the absence of IL-2 for 24 h, and nuclear extracts were isolated. To measure the DNA-binding activity of NF-κB to its respective recognition element, a gel EMSA was performed. Adiponectin inhibited IL-2-mediated NF-κB activation (Fig. 2,A). The specificity of the shifted NF-κB bands was verified by means of a supershift assay. The shifted NF-κB bands were supershifted with anti-p65 Ab and anti-p50 Ab, indicating that NF-κB bands are composed of p65 and p50. To confirm that the decreased NF-κB activity was associated with the suppression of NF-κB target gene expression, a pNF-κB-luc vector was transfected into 293 cells. Luciferase activity was significantly decreased as the result of adiponectin treatment (Fig. 2 B). It is known that AMPK is an important signaling component in the adiponectin signal pathway and inhibits NF-κB activation in cultured HUVECs (16). Thus, the effects of constitutively active AMPK, DN-AMPK, and AICAR, which is an AMPK activator, on the suppression of NF-κB activity by adiponectin were examined. As expected, constitutively active AMPK and AICAR inhibited NF-κB luciferase activity, as did adiponectin, whereas DN-AMPK alone increased NF-κB activity through suppression of endogenous AMPK activity. Treatment of DN-AMPK transfectants with adiponectin indicated the control level of NF-κB luciferase activity, because DN-AMPK prevented the adiponectin-mediated suppression of NF-κB luciferase activity. These data indicate that adiponectin inhibits NF-κB activation through the activation of AMPK.

NF-κB activity is regulated by IκB. Upon phosphorylation, IκB proteins are degraded via the proteasome pathway, resulting in the nuclear translocation of NF-κB and, ultimately, the activation of target genes. We examined the effect of adiponectin on the phosphorylation of IκB. IL-2 treatment increased the level of phosphorylation of IκB, but the phosphorylation of IκB was inhibited by cotreatment with adiponectin (Fig. 3,A). Because it is well known that adiponectin activates AMPK (11), we examined the effect of AICAR, an AMPK activator, instead of adiponectin. As expected, AICAR blocked the IL-2-induced phosphorylation of IκB. To confirm the role of AMPK in the regulation of IκB phosphorylation, compound C, an AMPK inhibitor, was used. The adiponectin-mediated suppression of IκB phosphorylation was recovered in the presence of compound C and IL-2. To determine whether adiponectin and these two compounds affected AMPK signaling pathway, the phosphorylation of AMPK was examined (Fig. 3 A). The results indicate that adiponectin inhibits NK cell cytotoxicity by suppressing the IL-2-mediated phosphorylation of IκB.

We next performed a gel mobility shift assay to examine the nuclear translocation of NF-κB. Adiponectin inhibited the IL-2-induced nuclear translocation of NF-κB, and the AMPK inhibitor, compound C, reversed the effect of adiponectin on the nuclear translocation of NF-κB, whereas AICAR mimicked adiponectin function (Fig. 3,B). We next examined whether the AMPK inhibitor and activator had the ability to regulate NK cell cytotoxicity. Purified NK cells were treated for 24 h as indicated in Fig. 3 C, and the released ⁵¹Cr activity was determined. As expected, the AMPK activator inhibited IL-2-mediated NK cell cytotoxicity, similar to adiponectin, and the AMPK inhibitor recovered the NK cell cytotoxicity that had been inhibited by adiponectin.

IL-2 stimulates NK cells to produce IFN-γ (17), and the production of IFN-γ is also regulated by the NF-κB element on the promoter of IFN-γ (18). Meanwhile, IFN-γ enhances NK cell cytotoxicity via the induction of FasL and TRAIL expression in NK cells. Thus, we determined whether adiponectin regulates IL-2-enhanced IFN-γ production in NK cells. Purified NK cells were treated with adiponectin in the presence or the absence of IL-2 for 24 h, and the supernatant was used to measure IFN-γ production. Treatment with adiponectin abrogated IL-2-induced IFN-γ production, whereas adiponectin alone had no effect on IFN-γ production (Fig. 4 A). AICAR, an AMPK activator, also inhibited IFN-γ production, although its effect was not much stronger than that of adiponectin. This suggests that adiponectin exerts its NK inhibitory effect via the suppression of IFN-γ production.

Because adiponectin suppressed IL-2-induced IFN-γ production in NK cells, it is possible that adiponectin reduces FasL and TRAIL expression by down-regulating IFN-γ production. To confirm this, FasL and TRAIL mRNA expressions were measured by RT-PCR after the various treatments indicated in Fig. 4,B. As expected, adiponectin dramatically suppressed IL-2-induced FasL and TRAIL mRNA expressions. We also examined FasL protein expression using flow cytometry. Purified NK cells were treated as indicated in Fig. 4 C. Cotreatment of adiponectin or AMPK activator with IL-2 suppressed the expression of FasL protein, and cotreatment of adiponectin with the AMPK inhibitor in the presence of IL-2 reversed the adiponectin-mediated suppression of FasL protein expression. Collectively, these data indicate that adiponectin inhibits FasL and TRAIL expression via the down-regulation of IFN-γ production induced by IL-2.

Previous studies have demonstrated that NK cells are activated by a number of cytokines, including IL-2, IL-12, IL-15, and type I IFNs (19, 20, 21), and are inhibited by glucocorticoids at physiological concentrations and by nutrients (22, 23, 24). The findings we report demonstrate that adiponectin inhibits IL-2-increased NK cell cytotoxicity without the effect of basal NK cell cytotoxicity and decreases IL-2-induced NK cell viability without the effect of basal NK cell viability. The findings also show that the adiponectin-mediated inhibition of NF-κB activity is an important molecular mechanism by which adiponectin suppresses NK cell cytotoxicity. Adiponectin activated AMPK through its receptor, and the activation of AMPK suppressed IL-2-enhanced NF-κB activity, which is critical in the increase in NK cell cytotoxicity by IL-2 (Fig. 3), consistent with previous reports indicating that AMPK inhibits fatty acid-enhanced NF-κB activity and NF-κB-mediated gene expression by TNF-α in HUVECs (16). This is also supported by previous findings showing that adiponectin inhibits endothelial NF-κB signaling through a cAMP-dependent pathway (13), and cAMP is able to activate AMPK (25). Experiments involving the use of inhibitors or activators of AMPK confirmed that AMPK plays a role in the adiponectin-mediated suppression of NK cell cytotoxicity. However, the AMPK inhibitor failed to fully recover the adiponectin-mediated inhibition of NK cytotoxicity and NF-κB activity (Fig. 3, B and C). Thus, the existence of other pathways that regulate NK cell activity or NF-κB activity by adiponectin cannot be ruled out. IL-2 has been shown to activate ERK1/2 and Akt to induce human telomerase reverse transcriptase expression in a human NK cell line NK 92 (26), and the ERK pathway plays a critically positive role in IL-2-induced NK cell activation (27). Therefore, ERK1/2 may be an alternative pathway. In contrast, adiponectin has been reported to suppress ERK1/2 activity (28), leading to the activation of NF-κB (29). Thus, it is possible that IL-2-stimulated NF-κB activation through ERK1/2 activation might be a target signaling pathway that can be blocked by adiponectin.

Leptin, an adipocyte-secreted adipokine, plays an important role in regulation of the immune system. Leptin enhances the cytotoxicity of splenic NK cells (30) at physiological concentrations and stimulates proliferation and cytokine synthesis in peripheral human lymphocytes (4, 31). Leptin levels are highly correlated with body fat mass in both lean and obese subjects. Adiponectin also plays a role in the immune system. It inhibits myelomonocytic activity, phagocytic activity, and TNF-α production (12); suppresses lipid accumulation in human monocyte-derived macrophages (5); and prevents the activation of NF-κB in aortic endothelial cells, porcine macrophages, and 3T3-L1 adipocytes (32). In addition to these known anti-inflammatory activities of adiponectin, the findings we present show that adiponectin suppresses cytokine-enhanced NK cell cytotoxicity, indicating that adiponectin may act as a negative regulator, whereas leptin functions as a positive regulator, in the immune system. Meanwhile, contrary to leptin, adiponectin levels are negatively correlated with body fat mass in obese subjects. This might be related to the opposite roles of these two adipokines.

It has been proposed that obesity is a type of low grade chronic inflammation (33), because a higher mass of adipose tissue secretes increased amounts of proinflammatory cytokines, including IL-1, IL-6, TNF-α, and acute phase proteins (34, 35), and macrophages infiltrate the fat mass in obese individuals (36). In addition, the production of adiponectin, which has anti-inflammatory properties, is decreased in obese individuals. Thus, a reduced level of adiponectin may contribute to produce a microenvironment for chronic inflammation in obesity. This chronic inflammation status may be indirectly associated with the development of diabetes.

Several reports have appeared, demonstrating that obesity is associated with altered NK function. For example, the NK cell population among total T lymphocytes in epididymal and inguinal fat tissues is greater than that in the lymph node, and NK cell numbers are decreased in obese mice fed a high fat diet (37). IL-2-induced NK cell-dependent cytotoxicity in PBL of obese individuals is increased with fat and carbohydrate intake (38). Collectively, these results suggest that adipose tissue may elaborate an immune system, including NK cells. Although the physiological role of adiponectin in the regulation of NK cell activity remains to be elucidated, the findings reported in this study demonstrate that adiponectin, an anti-inflammatory adipokine, plays a negative role in the regulation of NK cell activity.

The authors have no financial conflict of interest.