Insulin resistance and obesity

Article metrics

Type 2 diabetes mellitus is a serious health problem in the Western world. It arises when resistance to the glucose-lowering effects of insulin combines with impaired insulin secretion to raise the levels of glucose in the blood beyond the normal range. Studies into the molecular basis of insulin resistance have focused on the peroxisome proliferator-activated receptor gamma (PPARγ). This molecule, a member of the nuclear-hormone-receptor family, is the cellular target of thiazolidinedione drugs, which are used to treat diabetes by increasing sensitivity to insulin.

What are the endogenous ligands for PPARγ? How does it promote the insulin-stimulated uptake of glucose? And is this effect essential for the normal action of insulin? The answer to the last of these questions may be nearer thanks to a study by Barroso et al.1 on page 880 of this issue. They report the identification of two loss-of-function mutations of PPARγ that are associated with severe insulin resistance and type 2 diabetes mellitus in humans. Although such mutations are rare — detected in just three of 85 insulin-resistant people, and none of 314 controls — the implication that PPARγ is required for normal insulin sensitivity in humans is an important advance.

Found in the nucleus of many cells, particularly fat cells, PPARγ is both a receptor and a transcription factor. When PPARγ is bound by ligand, such as a thiazolidinedione, it becomes activated and binds to specific DNA sequences in gene promoters. Then, in complex with another transcription factor known as the retinoid X receptor (RXR), it activates the transcription of specific genes2 (Fig. 1, overleaf). One of the best-studied effects of activated PPARγ is its ability to induce differentiation of fibroblasts or other undifferentiated cells into mature fat cells2. Signalling by the PPARγ–RXR complex is also implicated in the synthesis of biologically active compounds by vascular endothelial cells3 and circulating immune cells4. Mutations in PPARγ may contribute to cancer5, and increased PPARγ signalling (owing to a mutation that increases its intrinsic activity) is also associated with human obesity6.

Figure 1: Molecular complex formed by dimerization of PPARγ and the retinoid X receptor (RXR).

Binding of ligand such as thiazolidinedione to either transcription factor can activate the complex, allowing the DNA-binding domain to bind the promoter region of target genes and activate transcription. Barroso et al.1 have shown that mutations that impair ligand binding (X1) disrupt this process and are associated with insulin resistance and normal body weight in humans. Serine phosphorylation (Ser–PO4) of PPARγ inhibits its activity, and mutation of an adjacent amino acid (X2) blocks this site. This increases PPARγ signalling, and is associated with obesity6.

Barroso et al.1 now show that the people affected by loss-of-function PPARγ mutations (one affects a mother and her son; the other affects an unrelated woman) share common elements of the ‘insulin resistance syndrome’. Symptoms include insulin resistance, diabetes, high blood pressure, dyslipidaemia (an abnormal plasma-lipid profile) and a skin-pigmentation disorder known as acanthosis nigricans. But a cardinal feature of the insulin-resistance syndrome that these people do not show is obesity. Reduced PPARγ signalling therefore seems to cause insulin resistance in the absence of obesity.

This observation contrasts sharply with the symptoms of gain-of-function mutation of PPARγ, reported last year by Ristow et al.6. In their study, obesity was associated with relatively low levels of insulin, suggesting an increased sensitivity to insulin. However, neither report1,6 includes a measurement of insulin sensitivity. Moreover, a third mutation in PPARγ has variable effects on body weight and insulin sensitivity2,7,8. Nevertheless, all of these findings indicate that by increasing PPARγ function it may be possible to prevent insulin resistance from occurring when normally it would (for example, in the obese state). Conversely, mutations in PPARγ that cause reduced function could lead to insulin resistance in lean people, in whom it would not normally occur.

Although disease-causing mutations of PPARγ are rare, might the insulin resistance associated with human obesity result from impaired PPARγ signalling in the absence of a mutation? Insulin resistance is especially likely to occur when excess fat is deposited within the abdominal cavity. This reduces the insulin sensitivity of fat cells and also of other tissues including skeletal muscle and liver. But how might expanding adipose stores causes insulin resistance? One explanation is that increased release of free fatty acids from triglyceride-laden fat cells provides an alternative metabolic substrate, which decreases the need for glucose as a fuel. As a result, insulin-stimulated glucose clearance from the blood is reduced, an effect that is manifest as insulin resistance.

However, because some free fatty acids may be PPARγ ligands2, an alternative explanation presents itself. If obesity alters the availability of these fatty acids, it could reduce PPARγ signalling and produce insulin resistance. But there are other ways to regulate the function of PPARγ. For example, phosphorylation of PPARγ on serine residues reduces its function, even in the presence of thiazolidinediones2,9. This is another potential mechanism whereby expanding fat stores might impair PPARγ function.

The two PPARγ mutations reported by Barroso et al.1 lead to amino-acid substitutions in regions of the molecule involved in ligand binding. As a result, these changes impair the activation of PPARγ by thiazolidinediones. By contrast, the obesity-inducing PPARγ mutation reported by Ristow et al.6 results in an amino-acid substitution adjacent to the serine phosphorylation site. This mutation impairs phosphorylation, thereby increasing PPARγ function. In each case, affected patients have one mutant and one normal allele, suggesting that the mutant PPARγ molecule dominates functionally over the normal protein. Indeed, Barroso and colleagues found that the function of normal PPARγ was impaired when they co-expressed it in tissue culture with either of the mutant proteins. Such ‘dominant-negative’ mutations are well documented in other nuclear-receptor systems, and they help to explain how a single mutant allele can cause disease.

The study of PPARγ mutations is expanding what we know about the involvement of this molecule in human health and disease. However, the demonstration of clinical abnormalities in a small number of patients who have a mutation is not proof that the mutation caused the symptoms. Determining how these patients respond to treatment with thiazolidinediones would provide important additional information. Moreover, neither insulin sensitivity nor insulin secretion were quantified in people with PPARγ mutations, yet the interaction of these two parameters is critical for glucose homeostasis10. The importance of taking these measurements is highlighted by the presence of type 2 diabetes in three of four obese people with the gain-of-function PPARγ mutation6. Perhaps the low insulin levels in these people reflect impaired insulin secretion rather than increased insulin sensitivity.

We need more information before we can conclude that too much PPARγ causes obesity without the expected metabolic consequences, whereas too little PPARγ elicits the metabolic consequences without obesity. But continued study of this important molecule could yield new approaches to the treatment of diseases such as obesity and diabetes, which take an enormous toll on human health.


  1. 1

    Barroso, I. et al. Nature 402, 880–883 (1999).

  2. 2

    Auwerx, J. Diabetologia 42, 1033–1049 (1999).

  3. 3

    Satoh, H. et al. Biochem. Biophys. Res. Commun. 254, 757–763 (1999).

  4. 4

    Spiegelman, B. M. Cell 93, 153–155 (1998).

  5. 5

    Sarraf, P. et al. Mol. Cell 3, 799–804 (1999).

  6. 6

    Ristow, M., Müller-Wieland, D., Pfeiffer, A., Krone, W. & Kahn, C. N. Engl. J. Med. 339, 953–959 (1998).

  7. 7

    Deeb, S. S. et al. Nature Genet. 20, 284–287 (1998).

  8. 8

    Valve, R. et al. J. Clin. Endocrinol. Metab. 84, 3708–3712 (1999).

  9. 9

    Hu, E., Kim, J. B., Sarraf, P. & Spiegelman, B. M. Science 274, 2100–2103 (1996).

  10. 10

    Kahn, S. E. et al. Diabetes 42, 1663–1672 (1993).

Download references

Author information

Correspondence to Michael W. Schwartz.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Schwartz, M., Kahn, S. Insulin resistance and obesity. Nature 402, 860–861 (1999) doi:10.1038/47209

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.