Highlights of the 15th Annual Meeting and Clinical Congress of the American Association of Clinical Endocrinologists

April 26-30, 2006

Zachary T. Bloomgarden, MD


August 08, 2006

In This Article

New Insights From Gene Chips

Attendees at the Plenary Session of the American Association of Clinical Endocrinology 15th Annual Meeting on April 27, 2006 in Chicago, Illinois, had the unusual opportunity to learn about developments in diabetes mellitus from the perspectives of molecular biology, clinical medicine, and social change.

Ronald Kahn, MD, Joslin Diabetes Center, Boston, Massachusetts, reviewed the findings of the Diabetes Genome Anatomy Project (DGAP) in a session titled "New Insights Into Diabetes and Obesity From DNA Chips."[1]

There are 21 million persons with diabetes in the United States, and this number increases by 1 million each year. The worldwide diabetes epidemic is part of the "much larger clinical problem" of the metabolic syndrome. "We all know that there's a metabolic syndrome," Dr. Kahn said, pointedly taking issue with the recent American Diabetes Association/European Association for the Study of Diabetes position paper questioning the syndrome.[2] Furthermore, Dr. Kahn continued, the metabolic syndrome is all "centered around insulin resistance," which causes a host of adverse outcomes, including dyslipidemia, nonalcoholic steatohepatitis, polycystic ovary syndrome (PCOS), cardiovascular disease (CVD), malignancy, and even Alzheimer's disease.

Insulin signaling constitutes 2 main pathways: (1) the metabolic pathway for which phosphatidylinositol-3-kinase is a central control point for cellular functions, including glucose transport, glycogen formation, and lipid synthesis; and (2) a number of control cycles known to respond to growth factors and environmental stress, including Ras protein and the mitogen-activated protein kinase pathways. Many additional pathways exist, all with effects on gene expression, thus allowing investigation of genetic (study of DNA), genomic (study of RNA), and proteomic (study of protein synthesis) patterns.

"Over the last 10 years there has been a real quantum leap in our ability to study...gene expression," Dr. Kahn said. The major technique is gene chip technologies, such as 1-cm2 microarrays in which up to hundreds of thousands of oligonucleotide cDNA probes are embedded and synthesized to match, partially or fully, one of the RNA sequences of a given individual genome.

The DGAP goal is to define the "normal anatomy" of insulin action on gene and protein expression in cells of mice and humans.[3] Dr. Kahn described a number of questions being addressed by the DGAP, including the characterization of differences between insulin action and insulin resistance; differences of insulin action between various cell types, such as muscle and fat; determination of the role of mitochondria in diabetes and insulin resistance, which Dr. Kahn described as a "very new and hot topic;" and the unraveling of new mechanisms of islet cell function.

In a streptozotocin (STZ) mouse type 1 diabetes model, analysis of 12,000 genes screened showed that expression of 513 genes increased or decreased significantly in uncontrolled diabetic animals. Many of the changes in gene expression are quite small (10% to 15%), leading Dr. Kahn to suggest that "a lot of small changes" occur in "coordinated abnormalities." One of the largest effects was in fatty acid (FA) metabolism. Levels of FA transporter 1 increased 28-fold, and the changes were reversed by administration of insulin, giving a genomic view of the dramatic increase in FA metabolism in uncontrolled diabetes. Genes for most of the mitochondrial electron transport chain proteins, in contrast, decrease, again with reversal (even to supra-normal levels) by insulin administration. These findings suggest the mechanisms of abnormal metabolism in diabetes, and reflect a combination of direct and indirect effects of insulin deficiency.

In mice with genetic knockouts (KO) of various parts of the insulin receptor system, such as the muscle insulin receptor (MIRKO), differences in gene expression can be analyzed to distinguish the lack of insulin action from the effects of insulin deficiency. Comparing MIRKO with SZT-diabetes, most of the gene expression changes are concordant. Assessment of discordant changes may reveal direct effects of diabetes rather than effects of lack of insulin action on muscle, or the converse. Of the 513 genes functioning abnormally in STZ-diabetes, approximately 40 appear to be directly insulin-regulated using this comparison. The mitochondrial electron transport genes decrease in response to STZ and are corrected by insulin, but show no systematic change in the MIRKO animal. If STZ is administered to a MIRKO animal, mitochondrial electron transport genes decrease but show little immediate (2 day) effect of subsequent insulin administration. In human diabetes, expression of mitochondrial oxidative metabolism genes decreases with insulin resistance and with type 2 diabetes. These findings confirm the growing understanding that the pathogenesis of diabetes involves an early abnormality in mitochondrial function, beginning with obesity and in relatives of persons with type 2 diabetes. The linkage of mitochondrial energy generation to diabetes suggests that peroxisome proliferator-activated receptor (PPAR) gamma coactivators 1 alpha and 1 beta, which appear to be "master regulators of mitochondrial metabolism" that bind to transcription factors such as PPAR gamma and nuclear respiratory factor (NRF)-1, lead to decreased oxidative phosphorylation and lipid oxidation early in the pathogenesis of the disease. Thiazolidinedione (TZD) administration may, then, "bypass to some extent" certain parts of the insulin resistance pathway.

"The brain turns out to also have insulin receptors and have insulin resistance," Dr. Kahn stated, which led him to speculate that there is a relationship between insulin resistance and food intake and other behaviors.

Beta cells also have insulin receptors, and exhibit the phenomenon of "glucose desensitization," in which hyperglycemia beyond a certain level decreases rather than increases insulin secretion. There must be an important role for insulin action in the beta cell, because a wide variety of both metabolic and mitogenic insulin actions can be demonstrated in this cell type. Mice not expressing beta-cell insulin receptors lose the first phase of glucose-stimulated insulin secretion and develop impaired glucose tolerance, despite (initially) normal insulin sensitivity in all other tissues.

Dr. Kahn asked whether a defect in insulin action can be demonstrated in the beta cell of humans with diabetes. Based on studies that compared human islets (isolated in a fashion similar to that used in islet transplantation) from type 2 diabetes patients not requiring insulin with the islets of nondiabetic persons, it appears likely that a number of transcription factors abnormal in the maturity-onset diabetes of youth (MODY) syndromes, such as hepatic nuclear factor (HNF)-1alpha, -4alpha, and pancreatic homeodomain protein (PDX)-1, insulin signaling genes, and glucose uptake and metabolism genes should play roles. Indeed, gene chip analysis shows that HNF-4alpha levels are decreased; glucose-6 phosphatase, phosphofructokinase, and other genes of glucose metabolism are decreased; and the insulin receptor substrate (IRS)1, IRS2, and protein kinase B (Akt) genes are decreased in expression, confirming that a number of genes that can cause type 2 diabetes in animal models are affected in human type 2 diabetes.

"The problem is," Dr. Kahn said, "assessing which if any of these abnormalities is primary." The gene ARNT (aryl hydrocarbon receptor nuclear translocator, also called hypoxia inducible factor 1 beta) appeared to show the greatest decrease in islets from persons with type 2 diabetes. ARNT has been shown to be a receptor for the potent environmental toxin dioxin, which is associated with chloracne, hepatotoxicity, malignancy, and perhaps with the development of type 2 diabetes. ARNT protein is readily detectable in beta-cell nuclei. Use of small inhibitory RNA to reduce ARNT gene function decreased glucose-stimulated insulin secretion, and decreased HNF-4alpha and HNF-1alpha, IRS-1 and IRS-2, and Akt. Dr. Kahn asked, "Could ARNT be a new type 2 diabetes gene in humans?" In an animal model not expressing ARNT, there was loss of first-phase insulin secretion and early development of glucose intolerance, which further supports this hypothesis.

Dr. Kahn concluded that the use of gene chips "can give ... completely new insights," with the potential that new diagnostic avenues and therapeutic targets may soon be discovered based on clues uncovered using the new technology.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.