Brief Critical Reviews
Leptin deficiency, found in transgenic lipodystrophic mice and in obese (ob/ob) mice, was shown to cause increased lipogenesis in liver, through action of the sterol regulatory elementbinding protein- c, and increased liver gluconeogenesis, through a decline in the insulin receptor substrate-2. The resulting stimulation of insulin secretion by the pancreas owing to high blood glucose initiates a vicious cycle of insulin resistance.
Noninsulin-dependent diabetes mellitus (NIDDM), or type 2 diabetes, is associated with insulin resistance consisting of hyperglycemia and hyperinsulinemia. Persons with NIDDM also tend to be obese, suggesting a connection between fat metabolism and insulin resistance. Many causes of insulin resistance have been suggested although the underlying etiology is far from being understood.
An important animal model for insulin resistance was recently developed by Shimomura et al.1 Underlying this model was the discovery of the family of sterol regulatory element-binding proteins (SREBPs).2 The SREBPs are membrane-bound proteins. Their activity is regulated by proteases releasing protein fragments that can enter the nucleus and bind to and activate promoters of genes involved in cholesterol and fatty acid synthesis. The specific protease becomes active and releases the fragment when the cellular cholesterol level is low. The protein fragment activates multiple genes leading to synthesis of cholesterol as well as the low-density lipoprotein receptor active in cholesterol uptake. When cellular cholesterol rises, the protease becomes inactive and the SREBP fragment remains membrane-bound. The isomeric member of the SREBP family that activates cholesterol synthesis is SREBP-2; those members that activate fatty acid synthesis are SREBP-la and SREBP- 1c. The SREBP-1a isomer occurs in growing cells. The SREBP-1c isomer is found in most organs in the adult, particularly in liver and adipose tissue.3
Under normal circumstances, insulin suppresses gluconeogenesis by decreasing expression of the genes for the gluconeogenic enzymes (phosphoenolpyruvate carboxykinase, fructose-1,6-bisphosphatase), and by decreasing the activity of the enzyme (glucose-6-phosphatase) that controls the release of glucose into the circulation. Normally, insulin stimulates the expression of the genes for the lipogenic enzymes (acetyl CoA carboxylase, fatty acid synthase), as well as for the enzyme that promotes the conversion of glucose to fatty acids (glucokinase). Insulin resistance in NIDDM consists of the failure of insulin to suppress hepatic gluconeogenesis and glucose secretion, which leads to hyperglycemia. The resulting high level of blood glucose stimulates the pancreas to secrete increasing levels of insulin, thus causing hyperinsulinemia.
In liver, insulin binds to the insulin receptor on hepatocytes and initiates a cascade of phosphorylation reactions that activate the components of the cascade. Each component phosphorylates and activates the next enzyme downstream. When insulin binds to its receptor, the insulin receptor substrates 1 and 2 (IRS-1 and IRS-2) become phosphorylated. They initiate the phosphorylation of a kinase called Akt, which then activates the particular insulin-regulated enzymes.
An important discovery by Shimomura et al.3 was the production of transgenic mice that overexpressed the active protein fragment derived from SREBP-1c: nSREBP-1c. In these mice nSREBP-Ic expression was placed under the control of the promoter of the gene for the fatty acidbinding protein aP2, directed specifically to adipocytes. Because the SREBP-1 members of the SREBP family are involved in the activation of lipogenesis, it was expected that the overexpression of nSREBP-1c in adipose tissue would stimulate fat synthesis. Paradoxically, for reasons not understood, the nSREBP-1c-overexpressing mice suffered from lipodystrophy owing to blocked differentiation of adipose tissue; their fat stores were much reduced. These mice were insulin resistant, had hyperglycemia and hyperinsulinemia, and developed fatty livers. A possible explanation for this reduction in adipose tissue was the four- to fivefold induction by the nSREBP-1c of the protein Pref-1, a known inhibitor of adipocyte differentiation.4 Whereas in the work described above, Shimomura et al.3 studied the effect of nSREBP-1c overexpression on adipose tissue, a more recent investigation by Shimomura et al.1 addressed the effect on liver and derived from resuits a hypothesis to explain insulin resistance associated with NIDDM.
The authors1 cite the following facts already known from their earlier work to support a function of SREBP-1c in mediating insulin action in liver: 1) insulin increases SREBP-1c levels in liver; 2) overexpression of SREBP-1c in livers of transgenic mice, bypassing insulin action, activates the genes normally activated by insulin; 3) a dominant-negative version of nSREBP-1c expressed in hepatocytes blocks the stimulation of lipogenic genes by insulin; 4) low levels of insulin, produced by fasting, suppress SREBP-1c in liver, whereas high levels, produced by refeeding, stimulate it. It therefore seems clear that SREBP1c mediates insulin action in liver.
In order to delineate the mechanism whereby SREBP1c activates the downstream genes normally responding to insulin, the authors' employed two model systems: the ob/ob mouse that, as a result of a defect in the appetitecontrolling hormone leptin, overeats and becomes obese, insulin resistant, and hyperglycemic, and develops fatty liver. The second model was the transgenic mouse of Shimomura et al.3 that overexpresses nSREBP-lc in its adipose tissue. As already mentioned, this condition caused a block in adipocyte differentiation, which led to a scarcity of adipose tissue, a condition also known as lipodystrophy. Consequently, the adipocyte-specific leptin declined radically, leading to overeating (without obesity), insulin resistance, hyperglycemia, and fatty livers. The two model systems, the ob/ob mouse and the nSREBP-1c transgenic mouse, are therefore similar to each other and could be restored to normality by injection of leptin.
Analysis of Northern blots confirmed that liver SREBP-1 mRNA of both the ob/ob and the transgenic mice were elevated and that this rise could be brought down to wild-type control levels by injection of leptin. The expression of fatty acid-synthesizing and gluconeogenic enzymes was also elevated and subsequently decreased by leptin. Investigating the posttranscriptional action of insulin, the authors1 showed that, though the insulin receptors of the liver cells themselves were not affected, the IRS-2 mRNA and the IRS-2 enzyme were suppressed in the ob/ob and the transgenic mice and restored to wildtype levels by leptin. Furthermore, the insulin-stimulated phosphorylation of the last member of the insulin cascade, Akt, was suppressed in the transgenic mice.
It was clear that leptin depletion, caused by defective leptin in the ob/ob mice or by lipodystrophy in the nSREBP-1c transgenic mice, showed equal stimulation of gluconeogenesis and lipogenesis in their livers, reversible in both cases by leptin injection. It was found that IRS-2, a component of the insulin cascade on the pathway between insulin binding to liver cells and its ultimate activation of the insulin-responsive enzymes, had declined precipitously. The authors1 asked the question, "was this decline caused by the leptin deficiency or by the hyperinsulinemia occurring in both the ob/ob mice and the transgenic mice?"
To answer this question the investigators' subjected normal wild-type mice to a 24-hour fast, followed by refeeding; insulin declined upon fasting and increased 20-fold upon refeeding. Northern blots showed that mRNAs of SREBP-1 and of the gluconeogenic and lipogenic enzymes of liver varied in proportion to insulin levels, but IRS-2 mRNA varied inversely with insulin levels. Rats made diabetic by treatment with streptozotocin had high liver IRS-2 mRNA, which could be suppressed by insulin injection. Parallel results were obtained with freshly isolated rat hepatocytes. Removal of insulin from the media raised their IRS-2 mRNA levels dramatically; levels fell precipitously after adding 100 nM insulin. In these hepatocytes, insulin raised the level of SREBP-1c mRNA in a time- and dose-dependent manner, maximally approximately 20-fold by 10 nM insulin in 12 hours.
The final experiments by Shimomura et al.1 concerned the effect of insulin on elements downstream from the insulin receptor. Primary hepatocytes, when treated with 100 nM insulin, showed reduced IRS-2 and little phosphorylated Akt. When insulin-starved cells with high levels of IRS-2 were treated with insulin, Akt became phosphorylated. These experiments demonstrated that the protein directly affected by insulin levels was neither the insulin receptor nor the final effector of insulin action (Akt), but rather IRS-2, the intermediate in the cascade.
Owing to these experiments it became clear that liver insulin levels regulated the intermediate insulin effector IRS-2. Insulin treatment leading to the decline in IRS-2 and subsequent high levels of the mRNAs of the gluconeogenic enzymes mimics the hyperinsulinemia accompanying insulin resistance in the livers of leptin-deficient animals. On the other hand, the pathway from the insulin receptor to the lipogenic enzymes (acetyl CoA carboxylase, fatty acid synthase) in liver by SREBP- 1 showed no abnormality in hyperinsulinemia; the report1 clearly demonstrated an increase in SREBP-1c mRNA and protein with increasing insulin levels, giving rise to fatty acid synthesis and fatty liver, and curable by leptin injection.
The defective leptin in the ob/ob mice and the absent leptin in the transgenic lipodystrophic mice apparently caused the hyperinsulinemia, which resulted in the normal response to insulin (increased lipogenesis via SREBP1) and the abnormal response of the liver (increased gluconeogenesis and glucose secretion, through a decline in IRS-2) (Figure 1). The authors1 suggest the possibility that leptin regulates insulin secretion from the pancreas: a low level of leptin leading to oversecretion of insulin.5 However, other possibilities exist whereby leptin depletion, through hyperphagia, might stimulate the sympathetic nervous system and cause increased insulin production.
In conclusion, Shimomura et al.1 proposed the following hypothesis: when leptin deficiency produces hyperinsulinemia, IRS-2 is downregulated, resulting in resistance to the insulin effect on gluconeogenesis in liver and glucose secretion from liver. On the other hand, SREBP-1 remains unaffected and is a normally upregulated factor, leading to overproduction of fatty acids in liver. High levels of circulating glucose and fatty acids lead to further increases in insulin secretion and aggravated hyperinsulinemia, thus initiating a vicious cycle (Figure
1). Though still hypothetical, this scheme may explain the insulin resistance associated with NIDDM. Insulin resistance, also presenting with increased glucose production and lipogenesis, has been observed under quite different conditions such as in patients experiencing severe trauma.6 The underlying molecular mechanism may be the same as in the leptin-deficient mice.
1. Shimomura I, Matsuda M, Hammer RE, et al. Decreased IRS-2 and increased SREBP-1c lead to mixed insulin resistance and sensitivity in livers of lipodystrophic and ob/ob mice. Mol Cell 2000;6:7786
2. Brown MS, Goldstein JL. The SREBP pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor. Cell 1997;89:331-40
3. Shimomura I, Hammer RE, Richardson JA, et al. Insulin resistance and diabetes mellitus in transgenic mice expressing nuclear SREBP-lc in adipose tissue: model for congenital generalized lipodystrophy. Genes Dev 1998;12:3182-94
4. Smas CM, Sul HS. Pref-1, a protein containing EGFlike repeats, inhibits adipocyte differentiation. Cell 1993;73:725-34
5. Seufert J, Kieffer TJ, Habener JF. Leptin inhibits insulin gene transcription and reverses hyperinsulinemia in leptin-deficient ob/ob mice. Proc Natl Acad Sci U S A 1999;96:674-9
6. Schwarz J-M, Chiolero R, Revelly J-P, et al. Effects of enteral carbohydrates on de novo lipogenesis in critically ill patients. Am J Clin Nutr 2000;72:9405
This review was prepared by George Wolf, D.Phil., Department of Nutritional Sciences, University of California, Berkeley, CA 94720-3104, USA. Please address reprint requests to the Nutrition Reviews Editorial Office, 711 Washington Street, Boston, MA 02111, USA.
Copyright International Life Sciences Institute and Nutrition Foundation Jun 2001
Provided by ProQuest Information and Learning Company. All rights Reserved