Diabetes Genome Anatomy Project |
|
Projects > Project 3
Project 3:
The Anatomy of Gene Expression in Insulin Resistant States
Primary Investigator: Michael P. Czech, Ph.D. (U. Mass Medical Center)
Summary and Specific Aims
Genomics is a particularly powerful approach to the problem of identifying genes involved in insulin signaling to glucose transport since this insulin effect is restricted to muscle and fat. Heterologous expression of insulin receptors and the insulin-regulated transporter GLUT4 in other cell types fails to restore insulin regulation of glucose transport to these cells, strongly indicating that additional fat- and muscle-specific gene products are involved. Furthermore, adipocytes during differentiation, maturation and enlargement undergo a dramatic conversion from an insulin-unresponsive, fibroblastic state to a highly insulin-sensitive state, and then to an insulin resistant state in which they are virtually unresponsive to insulin. Thus, identifying subsets of genes selectively expressed in adipocytes and muscle, as well as genes differentially expressed between insulin-sensitive and insulin-resistant states, should include potential insulin signaling components or modifiers of this signaling pathway. Such genes may in turn be candidates for susceptibility to type 2 diabetes because insulin resistance appears to be a key primary aspect of the etiology of the disease.
Based upon the above considerations, the overall objective of this project within DGAP is to generate genomics databases of difference states that reveal differentially expressed genes whose products may be involved in insulin signaling to GLUT4 or involved in other processes that regulate the degree of insulin sensitivity of target tissues. We will address the following issues in our studies:
Aim 1.
Identify and catalog genes that are selectively expressed in insulin-sensitive
mouse muscle cells and adipocytes, but not expressed or expressed at very low
levels in fibroblasts. In preliminary studies, we have created databases of
genes highly expressed in both 3T3-L1 adipocytes and mouse muscle but not
fibroblasts by a suppression subtractive hybridization strategy as well by
using Affymetrix GeneChips. The former method has the advantage of detecting
unknown genes not yet represented on genechips, but has the disadvantage of
often not detecting low abundance genes. We propose to generate additional
databases of such selectively expressed genes by probing Affymetrix Genechips
with cRNA probes derived from mouse primary white adipocytes, skeletal muscle,
and fibroblasts from the fasted (insulin-resistant) and fasted, refed
(insulin-sensitized) states. Preliminary studies related to genes encoding
mitochondrial proteins are presented in Mol Cell Biol (2003) 23:1085-1094.
Aim 2.
Define the gene and protein sets that vary in expression in 3T3-L1 cells as
they differentiate from insulin-unresponsive to insulin-sensitive adipocytes.
Affymetrix GeneChips will be used to screen probes prepared from 3T3-L1
fibroblasts after 2 days of differentiaton (insulin-unresponsive) and from
these cells after 4 and 6 days of differentiation (insulin-responsive and
highly insulin-responsive, respectively). Suppressive subtraction hybridization
will be used as an alternative approach to generate databases of genes that are
expressed between day 2 and 4. Proteins differentially expressed in adipocyte
cytosol, plasma membrane, mitochondria and GLUT4-containing membranes from the
2 day vs. 4 day 3T3-L1 cells will also be identified using the UMass Proteomics
Core.
Aim 3.
Define the gene and protein sets that vary in expression in fat and muscle from
lean vs. ob/ob mice that vary in insulin sensitivity. We reported many years
ago the high sensitivity to insulin action of small fat cells from 4 week ob/ob
mice as well as the expected severe insulin resistance of fat cells from these
mice at 26 weeks of age. Genechip analysis of white fat cells and skeletal
muscle derived from lean vs ob/ob 4 week, 8 week, 14 week and 26 week old mice,
respectively, will be performed. Adipocyte proteins differentially expressed
from 4 week ob/ob vs. 26 week ob/ob mice will also be identified using the
UMass Proteomics Core.
Initial Studies and Results
In addition to insulin, which acts through a PI 3-kinase signaling pathway, other stimuli such as exercise, osmotic shock and heterotrimeric G proteins have been reported to activate the translocation of GLUT4 via PI-3 kinase independent mechanisms. Insulin signaling to GLUT4 may also require PI 3-kinase-independent components. Taken together, these observations indicate that multiple pathways activate GLUT4 translocation, and that many of the components of these pathways have yet to be elucidated. In two sets of preliminary experiments, we have sought to discover genes that may encode such components by generating databases of genes highly expressed in both fat and muscle, but not (or poorly expressed) in insulin-insensitive fibroblasts.
In one set of experiments, we generated a Muscle-Adipocyte Union cDNA Library by modifying a suppression subtractive hybridization method. To make this library, digested 3T3-L1 adipocyte cDNA was ligated to one adaptor, and digested mouse muscle cDNA was ligated to a second adaptor. Both cDNAs were then hybridized to an excess of digested 3T3-L1 fibroblast cDNA in order to subtract out common sequences. The two hybridization reactions were then mixed to create hybrid molecules in which one strand originates from adipocytes while the second strand is from muscle. The final PCR products of the hybrid molecules were cloned into plasmid vector pCR2.1 to produce a library of about 104 clones, and 766 clones were sequenced. Each gene fragment was spotted onto a nylon array, and arrays were then hybridized to probes that were made by 32P labeling, first strand cDNA synthesis. This analysis demonstrated that most of clones in the Union library are in fact highly expressed in these tissues.
As a second approach to find genes highly expressed in muscle and adipocytes we probed the Affymetrix U74 series GeneChips with cRNA made from mouse muscle, 3T3-L1 fibroblasts and 3T3-L1 adipocytes. The three mouse U74 A,B and C GeneChip arrays containing probe sets representing known genes as well as EST clusters were screened with three independent cRNA preparations from each cell type. Using the difference call statistic method, 187 probe sets were found to show increased expression in all three muscle and adipocyte cRNA preparations compared to one of the fibroblast samples, but were unchanged in both of the other two fibroblast samples. Similar to the genes in the Union cDNA library, 108 of the 187 genes found by this method encode proteins involved in metabolism. About 14% of the genes discovered by this method encode signaling/regulatory proteins, and a many represent cytoskeletal or membrane trafficking proteins.
Protocols
- Isolation of Mouse or Rat Primary Fat Cells
- Glucose Uptake in Primary Fat Cells
- Assay for Free Fatty Acid Release from Isolated Fat Cells
- Assay for Conversion of U-14C-Glucose, 3H-Glycerol or 14C-pyruvate to Glyceride-Glycerol and Fatty Acids in Isolated Fat Cells
- RNA Isolation From Primary Fat Cells using Trizol from Invitrogen