CHARACTERIZATION OF TWO INSULIN-SENSITIVE DNA-BINDING PROTEINS 
Maria C. Alexander-Bridges, M.D., Ph.D., Assistant Investigator 
The initial interaction of insulin with its cell sur- 
face receptor alters flux through diverse metabolic 
pathways culminating in alterations in cell growth, 
differentiation, and energy storage. The ability of in- 
sulin to stimulate glucose storage in a tissue-spe- 
cific manner distinguishes this hormone from other 
anabolic peptide hormones and growth factors that 
alter growth-related processes. To describe fully the 
molecular mechanism of insulin's action on gene 
expression. Dr. Alexander-Bridges has been particu- 
larly interested in defining markers of insulin's 
metabolic effects, as well as markers of its growth- 
related effects on gene transcription. One gene 
encodes the glycolytic enzyme glyceraldehyde-3- 
phosphate dehydrogenase (GAPDH); the other, the 
early growth response gene (EGRl), encodes a zinc 
finger protein. Using these markers, Dr. Alexander- 
Bridges has made considerable progress toward the 
goal of elucidating the molecular mechanisms that 
underlie the tissue-specific inductive effect of insu- 
lin on expression of certain metabolic enzymes. The 
regulation of these two insulin-sensitive genes has 
been extensively characterized in vivo in animal 
models of diabetes, with nutritional manipulations 
such as fasting and refeeding, and in tissue culture 
during differentiation of fibroblasts to adipocytes. 
In ongoing studies aimed at elucidating the 
mechanism of insulin action on the expression of en- 
zymes involved in glucose utilization. Dr. Alexander- 
Bridges has identified a target gene, GAPDH, that is 
regulated by insulin in adipose tissue and liver. 
GAPDH mRNA levels are induced 10-fold by insulin 
in cultured 3T3 adipocytes and in the epididymal 
fat pads of rats fasted and refed a high-carbohy- 
drate, low-fat diet. Furthermore, expression of this 
gene is decreased to 30% of control levels in pri- 
mary adipocytes isolated from diabetic animals and 
is increased threefold over control levels upon re- 
placement of insulin. Because of the fidelity with 
which insulin regulation of GAPDH gene expres- 
sion reflects the metabolic effects of insulin in vivo, 
subsequent efforts were focused on uncovering the 
molecular mechanisms of insulin action on this 
gene. 
Because GAPDH mRNA levels are markedly in- 
creased by insulin in 3T3 adipocytes but not in 3T3 
preadipocytes. Dr. Alexander-Bridges was interested 
in determining the mechanism of this tissue-specific 
response to insulin. Since the effect was a conse- 
quence of changes in gene transcription, she asked 
whether tissue specificity of the insulin response 
reflects the induction of a unique complement of 
transcription factors that are activated by acute ex- 
posure of 3T3 adipocytes to insulin. 
To eliminate the possibility that acquisition of in- 
sulin-responsive gene expression during differentia- 
tion of 3T3 preadipocytes was due entirely to the 
presence of more insulin receptors on mature ad- 
ipocytes. Dr. Alexander-Bridges examined the regu- 
lation of growth-related genes in these cells, think- 
ing that the growth effects of insulin on gene 
expression in 3T3 fibroblasts might be detectable 
when metabolic effects of insulin were not. Dr. 
Perry J. Blackshear (HHMI, Duke University Medical 
Center) has shown that the proto-oncogene c-fos is 
regulated by insulin in 3T3 adipocytes through the 
well-described serum regulatory element (SRE), a 
factor that is present in fibroblasts. Although insu- 
lin can stimulate c-fos gene expression in 3T3 fibro- 
blasts, the effects are small. The effect of insulin on 
EGRl, another serum-regulated gene, was very 
clear in 3T3 preadipocytes, however, and has pro- 
vided a good marker for these studies. The fact that 
insulin could stimulate EGRl gene expression in 
preadipocytes indicated that the signal transduc- 
tion pathway involved in mediating the growth re- 
sponse of insulin was intact and eliminated the triv- 
ial possibility that there were too few insulin 
receptors on the 3T3 preadipocyte to initiate an in- 
sulin response. 
Dr. Alexander-Bridges then proceeded to deter- 
mine what components of the signal transduction 
pathway of insulin's action on metabolic genes 
might be missing from 3T3 adipocytes. The ap- 
proach was to work backward from an end effect of 
insulin on gene transcription by defining the fac- 
tors that mediated the effect and determining 
whether they were expressed in a tissue-specific 
manner. In stably transfected cell lines the expres- 
sion of a fusion gene containing the 5 '-flanking se- 
quences and promoter of the human GAPDH gene 
and the chloramphenicol acetyltransferase gene 
(HGAPDH-CAT) was stimulated threefold by insulin 
in 3T3 adipocytes and five- to eightfold in H35 
hepatoma cell lines. Thus it appeared reasonable to 
assume that tissue-specific regulation of GAPDH 
was mediated at the transcriptional level. Analysis 
of Bal 31 deletion mutants of the HGAPDH-CAT 
construct indicated that the five- to eightfold stimu- 
lation by insulin was due to two independent se- 
Continued 
13 
