Abstract
The fat-derived protein adiponectin enhances insulin action in rodents and its plasma levels are correlated with insulin sensitivity in humans. Adiponectin levels are reduced in obesity and type 2 diabetes mellitus (T2DM) and increased by thiazolidinediones (TZDs) in concert with improved insulin action. Two adiponectin receptors (AdipoR1 and AdipoR2) have recently been cloned. We examined the expression of these receptors in human adipose tissue and whether they are also regulated by TZDs. Gene expression was quantified by real-time rt-PCR in subcutaneous (Sub) and omental (Om) adipose tissue from 10 nondiabetic patients (7 F/3 M, age 38 x 3 years, BMI 31.6 ± 1.5 kg/m2, range 23-38) undergoing elective abdominal surgery. In a subgroup (n = 5), adipocytes and stromal cells were immediately separated by collagenase. RNA copy numbers (plasmid standard curves, normalized to GAPDH) of AdipoR1 were universally {223}10-fold higher than AdipoR2, and expression of AdipoR1 was similar in Sub and Om, while AdipoR2 expression was higher in Om. While adiponectin gene expression was specific to adipocytes, the receptors were expressed in both adipocytes and stromal cells (adipocytes:stromal cell ratio = 5:4). We then studied the effects of short-term pioglitazone (45 mg daily for 21 days) vs placebo on AdipoR1/R2 gene expression and insulin action (insulin clamp studies) in 11 T2DM subjects (9 M/2 F, age 49 ± 3 years, HbA1C 9.7 ± 0.7%, BMI 32 ± 2 kg/m2). There was no correlation between baseline insulin action and expression of AdipoR1/R2 in muscle or fat. While pioglitazone increased adiponectin expression 2-fold in Sub fat and improved both hepatic and peripheral insulin action, it did not affect expression of either receptor in fat or muscle. There were also no correlations between changes in insulin action and changes in receptor expression with pioglitazone in individual subjects. This was in contrast with a remarkably tight correlation (r2 = .93) between improved hepatic insulin action and increase in the active high-molecular-weight form of adiponectin.
Conclusions Both adiponectin receptor subtypes are expressed in human fat, with AdipoR1 being much more abundant. Unlike adiponectin, the receptors are expressed in both adipocytes and stromal adipose cells and are not affected by TZDs. Additionally; receptor expression did not correlate with insulin action at baseline or in response to pioglitazone. Further study is warranted to determine the physiologic significance of these receptors in humans.