| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* The Lundberg Laboratory for Diabetes Research,
Department of Clinical Nutrition,
Department of Clinical Physiology, and
The Wallenberg Laboratory for Cardiovascular Research, Sahlgrenska Academy at Göteborg University, Sweden,
# Department of Internal Medicine and Molecular Science, Osaka University, Japan; and
¶ Department of Medicine, University of Helsinki, Finland
1Correspondence: The Lundberg Laboratory for Diabetes Research, Department of Internal Medicine, Sahlgrenska Academy at Göteborg University, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. E-mail: ulf.smith{at}medic.gu.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and lipoprotein lipase) in the group of individuals characterized by a low IRS-1 expression and insulin resistance. A low IRS-1 expression in fat cells is a marker of insulin resistance and risk for type 2 diabetes and is associated with evidence of early vascular complications. Impaired adipocyte differentiation, including low gene expression and circulating levels of adiponectin, can provide a link between the cellular marker and the in vivo phenotype.—Jansson, P.-A., Pellmé, F., Hammarstedt, A., Sandqvist, M., Brekke, H., Caidahl, K., Forsberg, M., Volkmann, R., Carvalho, E., Funahashi, T., Matsuzawa, Y., Wiklund, O., Yang, X., Taskinen, M.-R., Smith, U. A novel cellular marker of insulin resistance and early atherosclerosis in humans is related to impaired fat cell differentiation and low adiponectin.
Key Words: type 2 diabetes • insulin signaling • adipose tissue • macroangiopathy • insulin.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and is a serious global health problem. A major clinical problem with type 2 diabetes today is the two- to fourfold increased morbidity and mortality in cardiovascular disease (2)
. The mechanisms for this are multifactorial and complex. The associations between insulin resistance and a number of metabolic abnormalities, collectively named the Insulin Resistance (or Metabolic) Syndrome, are well established (3
, 4)
, but these conventional risk factors explain less than 50% of the increased cardiovascular morbidity in type 2 diabetes (5)
.
It is clear that insulin resistance and most of its metabolic consequences, including postprandial lipid intolerance (6)
, precede the development of type 2 diabetes (7
, 8)
. Indeed, insulin resistance increases the risk for cardiovascular disease three- to fivefold (9)
, and a relationship between insulin resistance and the common carotid artery intima media thickness (IMT), an early marker of atherosclerosis (10)
, has been shown in different populations (11)
. Thus, successful intervention needs to be early; preferably before diabetes becomes established (12)
. This, however, is hampered by the fact that the genetic causes of insulin resistance and type 2 diabetes are largely unknown; thus, distinct genetic markers are lacking.
We recently reported that subjects with insulin resistance and type 2 diabetes have a low protein expression (<50% of normal) of IRS-1, a key intracellular protein for insulin signaling and action, in fat cells (13
14
15)
. A low IRS-1 protein expression is seen in 
5% of an unselected group and 30% of individuals with a marked genetic predisposition for type 2 diabetes (14)
. Furthermore, low IRS-1 protein expression is associated with a low cellular gene and protein expression of the GLUT-4 glucose transporter, showing that other genes/proteins are differentially expressed in these cells (15)
.
In this study, we performed extensive characterization of individuals with a normal (NIRS) and low IRS-1 (LIRS) protein expression, including measurements of the intima media thickness of the carotid artery. We studied whether this cellular marker can more distinctly identify subjects at risk for type 2 diabetes than the most commonly used risk marker (a known genetic predisposition for the disease). We examined the expression of several genes in the fat cells from an additional group of individuals identified in the same way to identify mechanisms through which the adipose tissue can contribute to the phenotype.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Twenty-five subjects (22 FDR and 3 C) agreed to undergo a needle biopsy and were examined for IRS-1 protein content in abdominal subcutaneous fat cells; 14 had a marked reduction in adipocyte IRS-1 protein content (reduction of >50% compared with internal standards including rhIRS-1) whereas 11 had a normal IRS-1 protein content. Comparisons were made between these individuals and the group of healthy relatives and control subjects. All participants gave informed consent and the study was approved by the Ethical Committee of Göteborg University.
|
Gene expression in the fat cells was analyzed in two additional groups (low, n=12, vs. normal, n=10, IRS-1) identified in the same way (16)
. The individuals with a low IRS-1 protein expression exhibited a similar degree of insulin resistance relative to the control group (
40%) as found in the present study (16)
. All subjects underwent a 75 g oral glucose tolerance test after having fasted overnight. Blood pressure was measured with a sphygmomanometer in the supine position and waist and hip circumferences were measured. Baseline blood samples were drawn from an antecubital vein and stored at -80°C until analysis. Whole blood was stored at -40°C to determine the Apo E phenotype.
Insulin sensitivity, acute insulin response, and body fat
After an overnight fast, the acute insulin response and insulin sensitivity were measured with the Minimal Model technique described by Bergman et al. (17)
. Glucose was injected i.v. in an antecubital vein over a period of 60 s (0.3 g/kg body weight of 30% glucose) to measure the acute insulin response. After 20 min, insulin (Actrapid®, Novo Nordisk, Copenhagen, Denmark) was administered i.v. as a bolus of 0.03 U/kg. Blood samples were collected at 20 time points during the IVGTT. Insulin sensitivity index and acute insulin response (0–10 min) were calculated using the MINIMOD program (17)
.
Body fat was measured with the bioelectrical impedance method (BIA-103; RJL Systems, Detroit, MI, USA).
Carotid artery ultrasound examinations
Carotid vessel wall imaging was performed by a certified sonographer using high-resolution B-mode ultrasound with an 8-MHz linear transducer (Sequoia 512, Acuson Corp, Mountain View, CA, USA). To guarantee standardization of the vessel wall circumference, images were captured at the peak of the R-wave by simultaneous ECG recordings (lead II). IMT was defined as the distance from the leading edge of the lumen–intima interface to the leading edge of the media–adventitia interface of the far wall. Three scannings for each section were recorded on S-VHS videotapes and the mean value was measured in an automated analyzing system (18)
. The IMT values shown are the mean of the left and right carotid arteries. The intrasonographer coefficient of variation (CV) for repeated readings of 20 ultrasound images was 0.8%, Pearson’s correlation coefficient r = 0.99. Intersonographer variability for repeated readings using the same analyzing system as in this study showed a CV of 6.0%.
Analytical procedures
Glucose was analyzed in venous blood using an automatic glucose analyser (Yellow Springs Instrument, Yellow Springs, OH, USA) and plasma insulin was analyzed with a standard radioimmunoassay having 40% cross-reactivity with proinsulin (Pharmacia, Uppsala, Sweden). C-peptide was assessed by a radioimmunoassay and proinsulin was measured by the Mercodia® Proinsulin ELISA enzyme immunoassay (Mercodia AB, Uppsala, Sweden).
Lipid concentrations were determined with an automated Cobas Mira analyser (Hoffman-LaRoche, Basel, Switzerland) as previously reported (6)
and LDL cholesterol was calculated according to Friedewald’s formula (LDL cholesterol = total cholesterol – HDL cholesterol – 0.45 x triglyceride level (mmol/L)).
C-Reactive protein was measured with an immunoenzymometric assay for quantitative determination of human CRP (CRP IEMA Test, Medix Biochemica, Kauniainen, Finland). The Apo B concentration was measured by immunoturbidometric Cobas Fara II autoanalyzers (Unimate 2 APOA/APOB, Hoffmann-La Roche).
Fibrinogen and plasminogen activator inhibitor-1 were analyzed using standard methods and Apo E genotype was determined by a polymerase chain reaction (PCR) technique. Leptin was measured with a radioimmunoassay (Linco Research, St. Louis, MO, USA). Adiponectin was analyzed by a quantitative immunoblotting technique (19)
. All interassay coefficients of variation were <10%.
Cell lysates and immunoblotting
The adipose specimens obtained from the needle biopsies were digested with collagenase to obtain isolated fat cells (20)
. The cells were then lysed in the presence of protease inhibitors, followed by protein separation in 7.5% sodium dodecyl sulfate PAGE (SDS-PAGE) as previously reported (13)
. The protein content of IRS-1 was analyzed by immunoblotting with a carboxyl-terminal antibody (UBI; Lake Placid, NY, USA) and gels were scanned with a PhosphorImager. IRS-1 expression was related to internal standards including rhIRS-1 protein on each gel for quantitation. This method of quantitation was verified with 125I-protein A (13)
.
Gene analyses
RNA was extracted from the adipose tissue using guanidinium isothiocyanate as described (21)
. Gene expression was measured with real-time RT-PCR using cDNA as template (22)
and HPRT as endogenous control. Primers and probes were designed using the Gene Express program (ABI Prism 7700, Applied Biosystems, Foster City, CA, USA). Sequences used are available upon request.
Statistical analysis
The Mann-Whitney test was used for comparisons between the groups. Two-tailed values of statistical significance were evaluated and a P value <0.050 was considered significant. Logistic regression was used to adjust for a confounder when comparing groups. Correlations between IMT and other variables were determined with the nonparametric Spearman’s rank test with partial correlation coefficients (rs). Statview 4.5 (Abacus Concepts, Inc., Berkeley, CA, USA) was used for statistical calculations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. With the exception of one control subject (BMI 31.2 kg/m2) all subjects were non-obese (BMI <30 kg/m2) and both relatives (FDR) and control subjects (C) as well as the low (LIRS) and normal IRS-1 (NIRS) groups were of similar age and BMI. Waist/hip-ratio tended to be increased in the LIRS compared with the NIRS group. Heart rate, Apo E phenotype, and cigarette smoking were similar in the different groups (not shown). Abdominal fat cell size tended to be larger in LIRS (97.9±1.4 µm) compared with NIRS (92.3±3.6 µm), but this difference did not reach statistical significance.
Risk markers for atherosclerosis
All subjects had normal glucose tolerance and fasting glucose levels (not shown). Serum LDL cholesterol and Apo B were significantly increased and C-reactive protein tended to be increased in the LIRS group (Table 1)
. However, probably due to the exclusion of hypertriglyceridemic and obese subjects, the differences in HDL cholesterol, triglycerides, PAI-1, and fibrinogen did not reach statistical significance. There were no significant differences between the FDR and C in any of the circulating risk markers.
Fasting insulin and proinsulin levels were significantly increased in the LIRS vs. NIRS group, but not in FDR vs. C (Table 1)
.
The glucose-stimulated early insulin response was markedly enhanced in the LIRS vs. NIRS subjects, but this was not found in FDR vs. C (Fig. 1
a). This was also true when the insulin response was corrected for the prevailing glucose concentrations (data not shown).
|
The LIRS group was significantly more insulin resistant than the NIRS group. A trend for this was also found for the FDR vs. C, but this difference did not reach full statistical significance (Fig. 1b
).
Adiponectin
Adiponectin levels (Fig. 1c
) were significantly lower in the LIRS than in the NIRS group even when corrected for waist circumference, BMI, and body fat, whereas leptin levels tended to be higher (Table 1)
. Although a trend for a difference was found, adiponectin levels were not significantly different between FDR and C (Fig. 1c
). We then examined whether adiponectin levels adjusted for age were related to degree of insulin sensitivity in the LIRS and NIRS groups. Circulating adiponectin levels were positively correlated to insulin sensitivity (r=0.40, P=0.06) and negatively correlated to fasting insulin (r=-0.45, P<0.05), proinsulin (r=-0.65, P<0.001), PAI-1 (r=-0.51, P<0.05), and CRP levels (r=-0.35, P=0.09).
Taken together, these data show that adiponectin levels correlate with insulin sensitivity and, as a consequence, negatively with ambient insulin, proinsulin, and PAI-1.
IMT of the carotid artery and relationships with other variables
No difference in IMT was found in FDR compared with C. However, mean IMT of the carotid bulb was clearly increased in the LIRS vs. NIRS group (Fig. 1d
), indicating that IRS-1 expression in adipocytes is a biomarker for subsequent risk for cardiovascular disease. No difference in IMT of the common carotid artery was seen in any of the groups (data not shown).
In a correlation analysis adjusted for age of all subjects in the LIRS and NIRS groups, the strongest associations for IMT of the bulb were to insulin secretion rate (r=0.57, P<0.01), fasting insulin (r=0.43, P<0.05), insulin sensitivity index (r=–0.43, P<0.05), waist circumference (r=0.41, P<0.05), and PAI-1 levels (r=0.40, P=0.05). When adjustments were made for age, waist circumference, and insulin sensitivity, insulin secretion rate (r=0.47, P<0.05) showed the closest correlation with IMT of the bulb. Due to the small group we could not obtain a valid model for multiple regression analysis.
In the relatives, a correlation between IMT of the bulb and adiponectin (r=–0.37, P<0.05) was found whereas no significant correlations were found to any of the variables measured in the control group.
Gene expression analyses
We examined the expression of several genes related to differentiation and insulin action in adipose tissue biopsies from 22 additional individuals characterized for IRS-1 expression and where RNA could be extracted (NIRS n=10; LIRS n=12). Genes related to the differentiation of the fat cells like adiponectin, aP2, PPAR, and LPL were significantly decreased whereas the difference for C/EBP
reached borderline significance (P<0.1). However, other genes like IGFBP as well as another secreted protein, leptin, were unchanged in the LIRS (Table 2
).
|
There were generally strong correlations between the expression levels of IRS-1, PPAR, aP2, adiponectin, C/EBP, and LPL but not for leptin and IGFBP (Table 3
). Thus, low adipocyte IRS-1 protein is associated with an impaired expression of several genes related to fat cell differentiation (PPAR, aP2, adiponectin, and LPL) as well as insulin action (IRS-1, LPL, and GLUT4) (15)
. These results are consistent with the lower circulating adiponectin levels found in this study (Fig. 1c
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
14
15)
but also of individuals showing evidence of early atherosclerosis; and 3) this group of individuals exhibits a gene expression profile consistent with an impaired differentiation of the fat cells, including lower circulating adiponectin levels. Thus, low IRS-1 expression is not likely to be the cause of the phenotype described here, but rather the first protein we identified of several genes and proteins that are differentially expressed in the fat cells. This conclusion is documented by the lower gene expression of several key differentiation markers such as adiponectin, aP2, PPAR, and LPL. In contrast, leptin and IGFBP expression were not changed whereas C/EBP clearly plays a role in the development of certain insulin-responsive genes in adipocytes (23)
and shows a strong correlation with most of the differentiation markers; we found no correlation with IRS-1 mRNA levels.
aP2 and particularly adiponectin expression are markers of PPAR gene activation (24
, 25)
. This is supported by the finding that patients with PPAR mutations leading to loss of function have low circulating adiponectin levels and are unresponsive to the stimulating effect of the insulin-sensitizing PPAR ligands, the thiazolidinediones (25)
. Our findings are consistent with this concept and show that the expression of several genes related to differentiation and insulin responsiveness like IRS-1, PPAR, aP2, adiponectin, and LPL were closely intercorrelated. With the exception of IRS-1, this was also true for C/EBP.
Adiponectin expression may be a key link between the cellular marker (low IRS-1) and the in vivo findings of a marked insulin resistance and early signs of atherosclerosis. Adiponectin maps to a chromosomal location (3q27) which is associated with the Insulin Resistance (or Metabolic) Syndrome in Caucasians (26)
.
Experimental evidence in animal models of insulin resistance and type 2 diabetes has shown that administration of adiponectin improves insulin resistance (24
, 27)
. Several studies have shown that adiponectin can play an important role in atherogenesis by influencing the expression of adhesion molecules and scavenger receptors (28)
. Animals lacking adiponectin show early evidence of atherogenesis, such as increased neo-intimal formation (29
, 30)
, and adiponectin administration prevents the atherosclerotic development in animal models (31)
.
The underlying molecular mechanism(s) for the lower expression of genes related to differentiation of the adipocytes in individuals with low IRS-1 expression is currently unclear. Global gene profiling and genotyping studies that are under way may clarify this.
As can be expected, history of a genetic predisposition for type 2 diabetes is not a sufficiently sensitive marker to distinctly identify individuals at risk for type 2 diabetes or its consequences. Thus, the likelihood of identifying individuals with insulin resistance and its associated metabolic and other abnormalities related to risk for type 2 diabetes depends on how many individuals with a true genetic predisposition happen to be included in the group of relatives vs. the control group. The present data clearly document the importance of such a "dilution" effect. However, it should be emphasized that a selection bias exists in the group of relatives included in this study, since the individuals were non-obese, fairly young, and were selected to have normal triglyceride levels (
2.0 mmol). With less stringent inclusion criteria, it is possible that clearer differences in insulin sensitivity and different metabolic variables would have been found, as previously reported (6
, 32)
. Nevertheless, this selection bias does not alter the conclusion that low IRS-1 expression is a much more sensitive marker of individuals at risk than is known genetic predisposition.
The subjects with low IRS-1 were also the most insulin resistant and showed an increased carotid artery bulb IMT vs. those with normal IRS-1 protein expression. In keeping with this concept, recent studies (11)
, although not uniformly supported (33)
, have demonstrated that subjects with atherosclerotic manifestations are more insulin resistant than subjects without known atherosclerosis.
One reason for the discrepant findings could be that most previous studies measured IMT in the common carotid artery (CCA) and not in the bulb, where low shear stress and a prolonged residence time for circulating elements favor the formation of plaques (34)
. In fact, some investigators did not find an association between the IMT of the CCA and coronary atherosclerosis (35)
, although a recent study concluded that IMT of the carotid artery bulb can be the best predictor for cardiovascular disease (36)
.
In addition to insulin resistance, enhanced insulin secretion, and low adiponectin levels, a striking finding was the increased fasting proinsulin levels in the low IRS-1 individuals, suggesting the possibility of an impaired processing of insulin in the ß cells in these subjects. Several recent studies have reported that proinsulin correlates with different measures of atherosclerosis (37
, 38)
and is related to future risk of cardiovascular disease (39)
. It is currently unclear whether this is due to a direct effect of proinsulin or, more likely, that proinsulin is related to the degree of insulin resistance and/or insulin secretion capacity in both the present and previous studies (40)
.
In conclusion, low IRS-1 expression in fat cells is a more sensitive marker for insulin resistance and its consequences than a known genetic predisposition for this disease. The present data show that not only individuals at risk for type 2 diabetes are identified, but also those at risk for the vascular complications. Adiponectin is likely to be an important link between low adipocyte IRS-1 expression, insulin resistance, and its complications since both circulating adiponectin levels and cellular gene expression were reduced. Furthermore, gene expression analyses suggest that the differentiation of the adipocytes is impaired in this group. Identification of the molecular basis for this may provide new insight into the causes of type 2 diabetes and its complications.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 19, 2002. Accepted for publication April 13, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Gustafson, A. Hammarstedt, C. X. Andersson, and U. Smith Inflamed Adipose Tissue: A Culprit Underlying the Metabolic Syndrome and Atherosclerosis Arterioscler. Thromb. Vasc. Biol., November 1, 2007; 27(11): 2276 - 2283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Subauste and C. F. Burant Role of FoxO1 in FFA-induced oxidative stress in adipocytes Am J Physiol Endocrinol Metab, July 1, 2007; 293(1): E159 - E164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. E. Graham, Q. Yang, M. Bluher, A. Hammarstedt, T. P. Ciaraldi, R. R. Henry, C. J. Wason, A. Oberbach, P.-A. Jansson, U. Smith, et al. Retinol-binding protein 4 and insulin resistance in lean, obese, and diabetic subjects. N. Engl. J. Med., June 15, 2006; 354(24): 2552 - 2563. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Gustafson and U. Smith Cytokines Promote Wnt Signaling and Inflammation and Impair the Normal Differentiation and Lipid Accumulation in 3T3-L1 Preadipocytes J. Biol. Chem., April 7, 2006; 281(14): 9507 - 9516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hammarstedt, J. Pihlajamaki, V. Rotter Sopasakis, S. Gogg, P.-A. Jansson, M. Laakso, and U. Smith Visfatin Is an Adipokine, But It Is Not Regulated by Thiazolidinediones J. Clin. Endocrinol. Metab., March 1, 2006; 91(3): 1181 - 1184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Iglseder, V. Mackevics, A. Stadlmayer, G. Tasch, G. Ladurner, and B. Paulweber Plasma Adiponectin Levels and Sonographic Phenotypes of Subclinical Carotid Artery Atherosclerosis: Data From the SAPHIR Study Stroke, December 1, 2005; 36(12): 2577 - 2582. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Pilz, R. Horejsi, R. Moller, G. Almer, H. Scharnagl, T. Stojakovic, R. Dimitrova, G. Weihrauch, M. Borkenstein, W. Maerz, et al. Early Atherosclerosis in Obese Juveniles Is Associated with Low Serum Levels of Adiponectin J. Clin. Endocrinol. Metab., August 1, 2005; 90(8): 4792 - 4796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Sugita, M. Fujimoto, T. Yasukawa, N. Shimizu, M. Sugita, S. Yasuhara, J. A. J. Martyn, and M. Kaneki Inducible Nitric-oxide Synthase and NO Donor Induce Insulin Receptor Substrate-1 Degradation in Skeletal Muscle Cells J. Biol. Chem., April 8, 2005; 280(14): 14203 - 14211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-T. Berthier, A. Houde, M. Cote, A.-M. Paradis, P. Mauriege, J. Bergeron, D. Gaudet, J.-P. Despres, and M.-C. Vohl Impact of adiponectin gene polymorphisms on plasma lipoprotein and adiponectin concentrations of viscerally obese men J. Lipid Res., February 1, 2005; 46(2): 237 - 244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Verreth, D. De Keyzer, M. Pelat, P. Verhamme, J. Ganame, J. K. Bielicki, A. Mertens, R. Quarck, N. Benhabiles, G. Marguerie, et al. Weight Loss-Associated Induction of Peroxisome Proliferator-Activated Receptor-{alpha} and Peroxisome Proliferator-Activated Receptor-{gamma} Correlate With Reduced Atherosclerosis and Improved Cardiovascular Function in Obese Insulin-Resistant Mice Circulation, November 16, 2004; 110(20): 3259 - 3269. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. B. Tanko, J. M. Bruun, P. Alexandersen, Y. Z. Bagger, B. Richelsen, C. Christiansen, and P. J. Larsen Novel Associations Between Bioavailable Estradiol and Adipokines in Elderly Women With Different Phenotypes of Obesity: Implications for Atherogenesis Circulation, October 12, 2004; 110(15): 2246 - 2252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Rotter, I. Nagaev, and U. Smith Interleukin-6 (IL-6) Induces Insulin Resistance in 3T3-L1 Adipocytes and Is, Like IL-8 and Tumor Necrosis Factor-{alpha}, Overexpressed in Human Fat Cells from Insulin-resistant Subjects J. Biol. Chem., November 14, 2003; 278(46): 45777 - 45784. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
