The authors have declared that no competing interests exist.
Current address: Albany Molecular Research, Inc., Bothell, Washington, United States of America
Conceived and designed the experiments: ZW TH. Performed the experiments: ZW. Wrote the paper: ZW TH. Carried out the Western blot analyses on insulin-stimulated tissues: OS.
Three p160 family members, p/CIP, SRC1, and TIF2, have been identified as transcriptional coactivators for nuclear hormone receptors and other transcription factors in vitro. In a previous study, we reported initial characterization of the obesity-resistant phenotypes of p/CIP and SRC-1 double knockout (DKO) mice, which exhibit increased energy expenditure, and suggested that nuclear hormone receptor target genes were involved in these phenotypes. In this study, we demonstrate that p/CIP and SRC1 control insulin signaling in a cell-autonomous manner both in vitro and in vivo. Genetic deletion of p/CIP and SRC-1 increases glucose uptake and enhances insulin sensitivity in both regular chow- and high fat diet-fed DKO mice despite increased food intake. Interestingly, we discover that loss of p/CIP and SRC-1 results in resistance to age-related obesity and glucose intolerance. We show that expression levels of a key insulin signaling component, insulin receptor substrate 1 (IRS1), are significantly increased in two cell lines representing fat and muscle lineages with p/CIP and SRC-1 deletions and in white adipose tissue and skeletal muscle of DKO mice; this may account for increased glucose metabolism and insulin sensitivity. This is the first evidence that the p160 coactivators control insulin signaling and glucose metabolism through IRS1. Therefore, our studies indicate that p/CIP and SRC-1 are potential therapeutic targets not only for obesity but also for diabetes.
The p160 coactivator proteins interact with and activate liganded nuclear hormone receptors, which function as transcription factors regulating target gene expression crucial for homeostasis, development, and reproduction
Mouse knockouts of individual p160 coactivators have distinct phenotypes: p/CIP (also called AIB1/ACTR/RAC3/NCoA1/SRC3) knockout mice have a somatic growth defect
In the current study, we examined the effects of loss of p/CIP and SRC-1 on insulin signaling and glucose metabolism, and discovered that p/CIP and SRC-1 negatively regulate the levels of IRS1 and insulin sensitivity both in vitro and in vivo. Our studies show that these coactivators play a crucial role in insulin resistance in obesity models in addition to their adipogenic and lipogenic activities. Our results demonstrate that reducing the levels and activities of p/CIP and SRC-1 results in increased insulin sensitivity and resistance to obesity. Therefore, p/CIP and SRC-1 are potential therapeutic targets for obesity and diabetes.
In our initial report of the p/CIP and SRC-1 DKO mice, we observed a lean phenotype with regular chow, and resistance to obesity with a high fat diet
A and B, Three-month old male mice (5 for each genotype) were fasted for six h, and intraperitoneally (ip) injected with a bolus of glucose (1.5 mg glucose/g body weight for GTT in panel A) or insulin (0.75 mU/g body weight for ITT in panel B). Blood glucose levels of mice were measured before and after injection with a glucometer at 20 min intervals. All results are presented as means ± S.E.M. * p<0.05, ** p<0.005, and *** p<0.0005.
To investigate the mechanism underlying the increased glucose uptake and insulin sensitivity, we studied expression levels of several components (such as insulin receptor, IRS1 and 2, AKT, GSK3, and S6 kinase) of the insulin pathway in DKO mice. The only significant change we observed was an increase in IRS1 RNA and protein levels in both white fat and muscle of DKO mice (
Three-month old mice were fasted overnight, injected (ip) either with insulin or with saline (fasted control), and mice were euthanized for tissue harvest. Cell lysates or total RNAs were prepared from white fat and muscle. A. Total RNAs isolated from white fat and muscle of fasted mice was used for Northern blot analyses. B. Tissue lysates of white fat and muscle of fasted and insulin treated mice were subjected to immunoblot analyses. C. DKO mice exhibited higher blood adiponectin levels. * p<0.05, ** p<0.005, and *** p<0.0005. D. Immunoblot analyses with AMPK and phospho-AMPK antibodies were performed.
Adiponectin is produced and secreted from fat, and has been well established as an insulin-sensitizing adipokine
To determine effects of these coactivators on insulin signaling in vitro, we used lentiviral based shRNA expression constructs to stably reduce endogenous p/CIP and SRC-1 levels in an NIH3T3 derived fibroblast cell line F442A, which is a preadipocyte cell line capable of differentiation into adipocytes in vitro. We obtained efficient knockdowns for p/CIP (PKD), SRC-1 (SKD), or both coactivators (double knockdown-DKD,
A. Lentiviral shRNA constructs of p/CIP, SRC-1, or both were used to stably knockdown endogenous levels of these coactivators in the F442A fibroblast cell line. 10 μg of total RNAs were used in Northern blot analyses to detect p/CIP and SRC-1 levels. A β-actin probe served as loading control. B. 50 μg total proteins of cell lysates from p/CIP knockdown (PKD), SRC-1 knockdown (SKD), wild-type, and double knockdown (DKD) were resolved by SDS-PAGE gel prior to immunoblot analyses. IRS1 protein levels were significantly increased in the DKD cell line, based on the α-tubulin loading control. C. Immunoblot analyses with stable p/CIP/SRC-1 DKD and IRS1 overexpression F442A cell lines. D. Serum starved F442A cell lines were stimulated for 15 or 30 min with 200 nM insulin, and immunoblot analyses were performed with the cell lysates using the indicated antibodies. Lanes 1, 4, and 7 were wild-type control with a vector construct only, lanes 2, 5, and 8 DKD cell line, and lanes 3, 6, and 9 IRS1 overexpression cell line. E. Immunoblot analyses with C2C12 myoblast stable DKD cell line. Lanes 1, 3, and 5 vector control, and lanes 2, 4, and 6 DKD.
To determine whether increased IRS1 levels were responsible for elevated insulin signaling, we established a stable F442A cell line overexpressing IRS1 at approximately the same level as in the DKD line (
The high fat diet common in developed countries results in not only obesity, but also insulin resistance and diabetes. To study the effects of loss of p/CIP and SRC-1 in a diet-induced obesity model, we challenged adult DKO mice (two months old) with a high fat diet (43% fat) for ten weeks. The wild-type and single knockouts gained 20–30% body weight on this diet, whereas the DKO mice gained considerably less weight despite significantly higher fat diet consumption
A. After 10 weeks on the high fat diet, the male mice were fasted overnight, blood was drawn and sera were prepared for measurements of insulin with a radio-labeled immunoassay (5 each). The DKO mice had lower insulin levels with this diet than their littermate controls. B. The age-matched male mice were fasted for 6 h and glucose tolerance tests were performed after 10 week on the diet. Increased glucose uptake was observed in the DKO mice in GTT. * p<0.05, ** p<0.005, and *** p<0.0005.
We proceeded to investigate effects of p/CIP and SRC-1 on age-related obesity and glucose intolerance, and observed increased levels of p/CIP and SRC-1 in fat and liver of one year old wild-type male mice compared to those of 3 month old normal male mice (
A. Northern blot analyses were performed to detect changes in p160 coactivator levels from 3 to 12 month old wild-type male mice. B. DKO mice remain lean at 12 months of age while the littermates become obese. C. Body weight gain is presented based on initial body weight of 3 month old male mice. D. GTT was performed with 12 month-old and sex-matched mice (5 for each genotype). * p<0.05, ** p<0.005, and *** p<0.0005.
In the current study, we report that loss of the p/CIP and SRC-1 p160 coactivators increased insulin signaling and insulin sensitivity both in vitro and in vivo. We show that the effects are at least partially cell-autonomous, since both the preadipocyte F442A and myoblast C2C12 lines with double knockdowns of p/CIP and SRC-1 exhibited acutely increased insulin signaling. We demonstrated that the levels of IRS1 were elevated in the two knockdown cell lines, and that this change was responsible for increased insulin signaling in the F442A knockdown cell line, and potentially also in the DKO mice, which exhibit increased IRS1 mRNA and protein levels in white fat and muscle. We also showed that the DKO mice had increased glucose uptake and metabolism in high fat diet-induced and age-related obesity models. We believe that our discovery of the control exerted by the p160 transcriptional coactivators on IRS1 has significant implications for type 2 diabetes (T2D). Recent genome-wide association studies with human T2D patients have identified a genetic variant near the IRS1 gene associated with decreased IRS1 levels, increased T2D, insulin resistance and hyperinsulinemia
In our previous study, we reported that p/CIP and SRC-1 functioned as key coactivators for the PPAR γ pathway in adipogenesis and obesity, and these coactivators played redundant roles
In the p/CIP and SRC-1 DKO mice, some but not all the PPARγ target genes required p/CIP and SRC-1 for their expression
We observed significantly higher levels of activated AMPK kinase in the DKO mice. We believe there are two reasons for that: one is that these mice have increased adiponectin levels, which may result from the lean phenotype of the DKO mice. Previous studies have established a correlation between lean subjects and increased adiponectin levels
In summary, our results suggest that targeting p/CIP and SRC-1 may not only reduce obesity but also enhance glucose uptake and insulin sensitivity, which would offer benefits not obtained by using TZDs for treatment of diabetes. We propose that decreasing the levels and/or activities of these coactivators may provide better solutions for obesity and diabetes in the future.
The homozygous knockout mice (DKO) were generated as described previously
Three-month old male mice were fasted for six h on the day of the experiment, and injected with a bolus of glucose (1.5 mg glucose/g body weight) or insulin (0.75 mU/g body weight) in PBS intraperitoneally. Blood glucose levels were measured with a glucometer (InDuo) and OneTouch Ultra strip from Lifescan, from tail bleeds before and after injection at 20 min intervals. To study insulin signaling in vivo, mice were fasted overnight and insulin was injected as above. After 30 min, mice were euthanized and tissues were dissected, harvested, and flash-frozen in liquid nitrogen. Blood insulin and adiponectin levels of overnight fasted, three-month old male mice were measured with radioimmunoassay kits from Linco.
White adipose tissue and muscle frozen in liquid nitrogen were extracted using a polytron tissue homogenizer in 7 volumes of ice-cold lysis buffer [40 mM HEPES, pH 7.5, 120 mM NaCl, 1% NP-40 (v/v) 1 mM EDTA, 10 mM pyrophosphate, 10 mM β-glycerophosphate, 50 mM NaF, 1 mM DTT, 1 mM Na3VO4, 1 mM PMSF, 1 μg/ml leupeptin, and 1 μg/ml aprotinin]. Extracts were solubilized on a vortex mixer for an additional 30 min at 4°C. Lysates were clarified by centrifugation at 13,000 ×g for 15 min at 4°C. Protein concentrations were determined spectrophotometrically using a BioRad DC protein assay kit. 30–50 μg of proteins were separated by SDS-PAGE and subjected to standard immunoblotting. AMPK, pAMPK-[T172], Akt, pAkt-[S473] and -[T308], pGSK-[S9/21], and pS6K1-[T389] antibodies were purchased from Cell Signaling Technology. A monoclonal antibody against α-tubulin was purchased from Sigma. The IRS1 antibody was from Santa Cruz Biotechnology.
Total RNAs from white fat and muscle were isolated with either TRIzol reagents (Invitrogen) or Qiagen's RNeasy Mini Kit, according to the manufacturers' protocols. Northern blot analyses were performed according to Current Protocols in Molecular Biology. Briefly, 10 μg RNA samples were denatured in formamide and formaldehyde and separated on a 1.2% denaturing agarose gel with 3% formaldehyde and 0.5 μg/ml ethidium bromide. Pictures of the agarose gel were then taken, and RNAs were transferred to Hybond-N+ nylon membrane with 20x SSC by blotting overnight. Denatured 32P-labeled cDNA probes were then hybridized with the membranes in Stratagene's Quik-Hyb solution at 65°C for 1 h. The membranes were washed once with 2× SSC and 0.1% SDS, once with 0.5× SSC and 0.1% SDS, each at room temperature for 15 min, and then 0.1% SSC and 0.1% SDS at 65°C for 15 min. Membranes were exposed to X-ray film at −80°C overnight with a screen. After stripping of signals with boiling 0.5% SDS, subsequent hybridizations were performed on the same membranes with indicated probes.
293T and C2C12 cell lines were from ATCC, and F442A was a generous gift from Dr. Peter Tontonoz
All results are presented as means ± S.E.M. A non-paired student
We would like to thank Drs. Chao Qi and Janardan Reddy for providing us the SRC-1 single knockout mice. We want to thank our colleagues in the Hunter/Eckhart lab for discussions and suggestions.