As obesity is a major risk factor for noninsulin-dependent diabetes mellitus, adipose tissue may generate a mediator that influences the activity of insulin on various target tissues. Recent evidence suggests that a cytokine, tumor necrosis factor-alpha (TNF alpha), may serve this role. This study investigates whether the expression of TNF alpha and its receptors is modulated during drug treatment to reduce insulin resistance. The effects of moderate weight loss by dietary restriction were also examined. We show here that a marked induction of TNF alpha mRNA occurs in adipose tissues from a mouse model of obesity-linked diabetes (KKAy) compared to that in nondiabetic mice (C57). Likewise, RNA transcripts encoding TNF R2 receptors (p75) were significantly increased in fat tissues of the obese diabetic animals. In muscle from these diabetic animals, RNA transcripts encoding both TNF R1 (p55) and R2 were significantly elevated, although R2 transcript abundance was less elevated than in fat. We also observed that the overexpression of mRNA for TNF alpha and both of its receptors could be at least partly normalized by treatment of the diabetic animals with the insulin-sensitizing agent pioglitazone. Treating of the obese diabetic animals by food restriction reduced the expression of mRNA for TNF R2 in muscle, but not fat. These results clearly indicate that gene expression for the TNF systems can be regulated by an insulin-sensitizing drug and reduction of body weight. Such findings support a role for this cytokine in the insulin-resistant diabetic state and show its modulation by therapies that reverse the disorder.

Insulin-stimulated glucose uptake into muscle and fat involves regulation of the subcellular distribution and the expression of a specific facilitative glucose transporter protein (GLUT4). Peripheral glucose uptake is lowered in diabetes, and the expression of GLUT4 is depressed in animals that have been made diabetic (i.e. insulin deficient) by destruction of the pancreatic beta-cells. In the present study we found that GLUT4 expression is also decreased in an animal model for type II diabetes mellitus (noninsulin-dependent diabetes mellitus), KKAY obese mice. These KKAY mice have elevated circulating insulin levels, but target cell resistance to the metabolic actions of insulin. Treatment of both types of diabetic animals with pioglitazone, a new antihyperglycemic compound, corrects deficits in glucose transport and GLUT4 mRNA and protein abundance. Such corrections are, however, more readily detected in fat than in muscle. Increases in GLUT4 mRNA and protein levels and glucose transport function by pioglitazone are dependent upon the presence of circulating insulin. Treatment with pioglitazone alone is sufficient for correction of glucose transport in hyperinsulinemic insulin-resistant animals, but hypoinsulinemic animals require insulin therapy along with pioglitazone treatment for similar corrections. In these insulin-deficient animals, neither treatment with the drug alone nor minimal insulin replacement therapy results in substantial correction. Since insulin and this antihyperglycemic agent seem to work synergistically, it is likely that pioglitazone acts to amplify cellular responses to insulin.

No abstract

Burn injury and endotoxin lead to the development of a systemic inflammatory response. Because tumor necrosis factor-alpha (TNF-alpha) is a component of the proinflammatory response, we have determined the effect of burn injury and endotoxin in a murine model of thermal on tissue specific TNF-alpha levels in the liver and lung. Male mice were divided into four groups and injected with endotoxin (ETX) (2.5 mg/kg intraperitoneally) or saline (CNTL) or subjected to a 16% full-thickness scald burn (B), or ETX administration 72 hours after burn injury (B+ETX). Animals were killed at 0 to 24 hours after ETX or CNTL, 0 to 72 hours after B, and 72 to 96 hours after B+ETX (ETX administration 72 hours after B). TNF-alpha mRNA by Northern blot and protein analysis by enzyme-linked immunosorbent assay were determined and protein expressed as nanogram per gram of tissue. Statistical analysis was performed using analysis of variance with significance at p < 0.05. Burn injury did not result in detectable levels of liver or lung TNF protein or mRNA. Endotoxin administration resulted in a near six-fold rise in liver TNF protein compared with controls at 1, 2, and 6 hours after ETX (p < 0.05 to p < 0.001). Liver mRNA remained elevated from 20 minutes to 24 hours after ETX versus CNTL (p < 0.05). Endotoxin injection produced a persistent lung TNF protein elevation reaching significance at 1 and 2 hours (p < 0.001) and a rise in mRNA at 40 minutes to 6 hours (p < 0.05) versus CNTL. The liver showed a trend of reduced mRNA after B+ETX versus ETX (p = NS), whereas protein levels were reduced by 50 to 60% at 1 and 2 hours (p < 0.01). Lung mRNA values after B+ETX were only 40% compared with ETX at nearly all time points (p < 0.001) but were 15 times above CNTL values at 2 hours (p < 0.05). Based on these results, we conclude that burn injury did not cause an increase in liver or lung tissue specific TNF-alpha. However, the presence of a preexisting burn injury dramatically altered the response to endotoxin and the primary point of regulation appears to be at the posttranscriptional level.