Resistin Knockout Mice Exhibit Impaired Adipocyte Glucose-Dependent Insulinotropic Polypeptide Receptor (GIPR) Expression

Retn^+/− mice were provided by Dr. Mitchell Lazar, University of Pennsylvania School of Medicine. Breeding and genotyping were performed at the University of British Columbia (UBC). Animal experiments were conducted in accordance with guidelines of the UBC Committee on Animal Care and Canadian Council on Animal Care.

Primary adipocytes were isolated (18) from male Retn^+/+, Retn^+/−, and Retn^−/− mice (10–12 weeks old) and plated using Matrigel (BD Biosciences). After GIP plus insulin treatment, adipocytes were recovered and LPL activity was measured, as described (6). For the fatty acid uptake assay (19), adipocytes were incubated with 4,4-difluoro-5-methyl-4-bora-3a, 4a-diaza-s-indacene-3-dodecanoic acid (C₁-BODIPY500/510-C₁₂, Molecular Probes) and sorted on a BD LSRII flow cytometer.

Total RNA was extracted using an RNeasy Mini Kit (Qiagen) and Gipr expression measured, as described previously (10). RT² Profiler PCR Arrays were performed to quantify mRNAs of multiple genes, including housekeeping genes, according to the manufacturer’s protocol.

Protein samples were immunoprecipitated using Dynabead protein A or protein G (Invitrogen) and c-Jun or phospho-serine/threonine antibody (Cell Signaling; ECM Biosciences). Precipitated products or protein samples were resolved by SDS-PAGE and probed with the antibodies indicated in the Figure legends.

c-Jun was immunoprecipitated from optimally sheared chromatin, thus preserving specific protein/DNA complexes of Retn^+/+ and Retn^−/− adipocytes, using Dynabead protein G and c-Jun antibody. Bound DNA was quantified by SYBR green PCR using oligonucleotides flanking 12-O-tetradecanoylphorbol-13-acetate (TPA) response element (TRE)-I, TRE-II, or TRE-III of the mouse Gipr promoter.

Significance was tested using ANOVA with the Newman-Keuls post hoc test or Student t test (P < 0.05), as indicated in the Figure legends.

Mating of heterozygous Retn^+/− mice resulted in expected Retn^+/+, Retn^+/−, and Retn^−/− genotypes (14). The mice did not differ significantly in body weight, glucose tolerance, or epididymal white adipose tissue (eWAT) adipocyte size (Supplementary Fig. 1A–E). Immunoreactive resistin was undetectable in Retn^−/− mice plasma and eWAT extracts, whereas plasma and fat levels in Retn^+/− mice were <25% of Retn^+/+ levels (Supplementary Fig. 1F and G).

To investigate the role of resistin in GIP-mediated increases in LPL activity, responses of Retn^+/+, Retn^+/−, and Retn^−/− adipocytes to GIP were examined. Incubation of Retn^+/+ adipocytes with GIP plus insulin (1 nmol/L) resulted in concentration-dependent stimulation of LPL activity (Fig. 1A), increased levels of phospho-PKB (Ser473), and decreased levels of phospho-LKB1 (Ser428) and phospho-AMPK (Thr172), but was without effect on Retn^+/− or Retn^−/− adipocytes (Supplementary Fig. 2A–C). In addition, treatment of Retn^+/+ adipocytes with GIP plus insulin (1 nmol/L) resulted in significantly increased fatty acid uptake, but was without effect on Retn^+/− or Retn^−/− adipocytes (Fig. 1B). The complete lack of Retn^+/− adipocyte responses was surprising and suggested a critical resistin secretion level was required for GIP-mediated stimulation of LPL activity and/or adipocyte responsiveness to GIP was compromised.

The possibility that resistin secretion from Retn^+/− adipocytes was reduced to ineffective levels for activating the PKB/LKB1/AMPK pathway was first tested. Incubation of Retn^+/+ adipocytes with GIP plus insulin (1 nmol/L) for 2 h resulted in concentration-dependent stimulation of resistin secretion (Supplementary Fig. 3A). Basal secretion of resistin by Retn^+/− adipocytes was significantly reduced, and secretion by Retn^−/− adipocytes was undetectable. Resistin secretion by Retn^+/− adipocytes was unresponsive to GIP plus insulin (1 nmol/L). In addition, insulin alone had no effect on resistin secretion by Retn^+/+ adipocytes (Supplementary Fig. 3B). These results indicated that the inability of GIP to activate the PKB/LKB1/AMPK/LPL pathway in Retn^+/− adipocytes involved severely blunted resistin secretion.

To examine the involvement of altered receptor expression, GIPR protein and mRNA levels were determined, and Retn^+/− and Retn^−/− adipocytes demonstrated severe downregulation (Fig. 1C and D). We next examined whether resistin acted as a direct regulator of GIPR expression. Treatment of Retn^+/+ adipocytes with resistin (100 nmol/L) resulted in a delayed increase in GIPR expression, after 72-h incubation (Fig. 2A), indicating chronic effects on expression, and 72-h treatment with resistin rescued GIP-induced LPL activity in Retn^+/− and Retn^−/− adipocytes (Fig. 2B).

Intriguingly, GIPR protein was also downregulated in brown adipose tissue and the stromal vascular fraction of Retn^+/− and Retn^−/− mice (Supplementary Fig. 4A and B), whereas no significant differences were noted in skeletal muscle or pancreatic islet GIPR protein expression among the groups (Supplementary Fig 4C and D). In addition, treatment of Retn^+/+ islets with resistin (100 nmol/L) had no effect on GIPR expression (Supplementary Fig. 4E), implying that regulation was tissue-specific.

With a view to identifying proteins involved in regulating the adipocyte Gipr, a PCR array was used to screen a panel of 84 genes associated with obesity, insulin resistance, and early onset of diabetes (RT² Profiler) in eWAT from Retn^+/+ and Retn^−/− mice. This revealed large reductions in Retn^−/− fat expression of genes coding for proteins involved in signaling (ectonucleotide pyrophosphatase/phosphodiesterase 1 [Enpp1]), intercellular adhesion (intracellular adhesion molecule [Icam1]), GLUT-4 trafficking (syntaxin 4a [Stx4a]), and glyceroneogenesis (soluble phosphoenolpyruvate carboxykinase 1 [Pck1]) (Fig. 3A). Of particular interest, genes encoding tumor necrosis factor (TNF [Tnf]), tumor necrosis factor receptor 2 (TNFR2 [Tnfrsf1b]), and the downstream signaling protein(s) stress-activated protein kinase (SAPK)/Jun NH₂-terminal kinase (JNK) (Mapk8), were decreased in Retn^−/− fat tissue. In contrast, Gsk3b (glycogen synthase kinase 3 [GSK3]-β) expression, a negative regulator of c-Jun (20) was increased in Retn^−/− fat (Fig. 3B). Levels of phospho-SAPK/JNK (Thr183/Tyr185) and phospho-c-Jun, were also decreased (Fig. 3C and D). Together, these results suggested that decreased adipocyte GIPR expression involved reduced signaling via TNF activation of SAPK/JNK and c-Jun.

In view of the decreased Gipr expression and c-Jun phosphorylation in Retn^−/− fat tissue, we focused on establishing a functional link. Analysis of mouse Gipr promoter sequences using MatInspector (Abteling Genetik, Germany) revealed three potential TREs for c-Jun binding, residing between -3583 and -3577, -2936 and -2930, and -298 and -292, that we termed TRE-I, -II, and -III, respectively. TRE-I (TGATTCA), TRE-II (TGAGCCA), and the inverted TRE-III (TGAGTAA) exhibit 85.7% homology to the consensus TRE (TGAG/CTCA) (Fig. 4A). To identify functional TRE(s), c-Jun binding to the Gipr promoter was determined by measuring TRE occupancy using ChIP assays. Significantly increased c-Jun occupancy in Retn^+/+ over Retn^−/− adipocytes was detectable only in TRE-III (Fig. 4B), indicating it is the functional cis-acting element for c-Jun–mediated regulation of the mouse Gipr promoter. Resistin (100 nmol/L) treatment of Retn^+/+ adipocytes significantly increased c-Jun occupancy in Gipr promoter TRE-III (Fig. 4C) and Gipr mRNA levels (Fig. 4D) by 24 h, with a further increase up to 72 h. c-Jun occupancy and Gipr mRNA levels at 72 h after resistin treatment were both greatly reduced by the SAPK/JNK inhibitor (JNKi). Resistin-mediated GIPR protein expression was also greatly attenuated by JNKi (Fig. 4E), strongly supporting the involvement of a SAPK/JNK/c-Jun–mediated pathway in the regulation of GIPR expression.

We previously demonstrated that GIP stimulated LPL activity in 3T3-L1 cells and human adipocytes through a pathway involving resistin-mediated increases in phosphorylation of PKB and decreases in LKB1 and AMPK phosphorylation (6–8); a similar role was confirmed in Retn^+/+ adipocytes (Fig. 1A and Supplementary Fig. 2). However, although resistin content of Retn^+/− adipocytes was ∼25% of that in Retn^+/+ mice, GIP responses were almost completely ablated, suggesting that a critical level of resistin signaling was essential for GIP action and/or that GIPR responsiveness was greatly reduced. Subsequently, GIPR protein and mRNA were both found to be downregulated in Retn^+/− and Retn^−/− adipocytes, indicating that attenuated adipocyte GIPR expression was the major contributor to blunted GIP responsiveness (Fig. 1B and C). In addition, resistin treatment of Retn^+/+ adipocytes for 72 h increased GIPR expression in Retn^+/+ adipocytes (Fig. 2A) and rescued GIP responsiveness of LPL activity in Retn^+/− and Retn^−/− adipocytes (Fig. 2B), supporting a role in its long-term regulation. Although resistin treatment of Retn^+/+ adipocytes resulted in a delayed increase in GIPR protein expression after 72-h incubation (Fig. 2A), increases in Gipr mRNA (Fig. 4D) and c-Jun Gipr promoter occupancy (Fig. 4C) were detectable after 24-h incubation. These results suggest involvement of complex post-translational regulation in resistin-mediated GIPR protein expression. The receptor is glycosylated, and the islet GIPR was downregulated in response to elevated glucose by a process involving ubiquitination (21), although whether these processes are involved in the delayed expression is unclear.

The PCR array study revealed Retn^−/− adipose tissue downregulation of genes encoding TNF (Tnf), TNFR2 (Tnfrsf1b), and the downstream signaling protein SAPK/JNK (Mapk8), whereas expression of Gsk3b was greatly increased (Fig. 3A and B). Resistin stimulates TNF-α gene expression in human and mouse macrophages (22), and TNF-α increases resistin protein and mRNA expression via a JNK pathway in human macrophages (23). Because activation of SAPK/JNK/c-Jun by TNF-α is an established pathway for regulating gene transcription (24), we considered it possible that its downregulation may contribute to the reduced Gipr expression in Retn^−/− eWAT. Supporting this, phosphorylated levels of c-Jun and SAPK/JNK (Thr183/Tyr185) were decreased in Retn^−/− eWAT (Fig. 3C and D). In addition, using ChIP assays, we established that TRE-III in the mouse Gipr promoter was functional and responsible for c-Jun–mediated transcriptional activation, which was downregulated in Retn^−/− adipocytes (Fig. 4B). Treatment of Retn^+/+ adipocytes with a SAPK/JNK inhibitor greatly reduced resistin-induced increases in c-Jun occupancy in TRE-III of the Gipr promoter (Fig. 4C) and Gipr mRNA and protein expression (Fig. 4D and E), strongly supporting functional involvement of a SAPK/JNK/c-Jun–mediated pathway in GIPR expression regulation. The elevation in Gsk3b would also be expected to result in phosphorylation-induced reduction in transcriptional activity of c-Jun (20). These results strongly support involvement of c-Jun in regulating Gipr expression; however, other factors could clearly be involved.

Although a strong link between resistin and TNFα has been identified in humans (12), the context in which resistin normally stimulates the TNF/TNFR2 pathway in adipocytes is unclear. TNFR1 and TNFR2 are the primary receptors for soluble and membrane-bound TNF, respectively (24). Most cell types constitutively express TNFR1, whereas TNFR2 expression is regulated by multiple factors. These receptors activate several cellular signaling modules, including SAPK/JNK, and GIP stimulates resistin secretion via activation of SAPK/JNK- and p38 MAPK-mediated pathways (8). Therefore, a complex interplay exists among GIP, resistin, and TNF-α.

In summary, effects of GIP on PKB/LKB1/AMPK/LPL in Retn^+/+ adipocytes were almost completely abolished in Retn^+/− and Retn^−/− adipocytes largely due to greatly reduced GIPR expression, associated with downregulated TNFα signaling and c-Jun–mediated transcriptional activation of the Gipr. In view of the use of incretin-related compounds for diabetes treatment, it will be important to establish the physiological and pathophysiological significance of the strong interaction between GIP and resistin actions on the adipocyte, especially in humans.

These studies were generously supported by funding from the Canadian Diabetes Association and Canadian Institutes of Health Research. The authors thank Dr. Mitchell Lazar, University of Pennsylvania School of Medicine, for generously providing the Retn^+/− mice that were developed with funding from National Institutes of Health Grant P01 DK-49210.

No potential conflicts of interest relevant to this article were reported.

S.-J.K. researched data, contributed to discussion, and wrote, reviewed, and edited the manuscript. C.N. researched data. C.H.S.M. contributed to discussion and wrote, reviewed, and edited the manuscript. C.H.S.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Parts of this study were presented at the 72nd Scientific Sessions of the American Diabetes Association, Philadelphia, Pennsylvania, 8–12 June 2012.