Overexpression of the epidermal growth factor receptor (EGFR) in many human cancers and cancer cell lines is due to gene amplification and/or increased gene transcription. Our goal is to understand the mechanisms and factors involved in increased EGFR expression. Our approach to determine the mechanisms of upregulation has been to identify and characterize transcription factors that regulate EGFR gene expression. In our studies of EGFR gene regulation, we demonstrated that early growth response factor 1 (Egr-1) which is induced by hypoxia activated the basal transcriptional activity of the EGFR promoter. Egr-1 not only transactivated EGFR promoter activity but also enhanced endogenous EGFR expression. Using a series of EGFR promoter deletion mutants, we showed that the region between -484 and -389, which contains a putative Egr-1 consensus motif, was crucial for EGFR transactivation by Egr-1. Electrophoretic mobility shift assays (EMSA) revealed that Egr-1 binds to the oligonucleotide containing this Egr-1 motif. Also, introduction of an antisense oligonucleotide for Egr-1 diminished EGFR expression during hypoxia, indicating the upregulation of EGFR by hypoxia is mediated through Egr-1. Our results provide evidence that regulation of EGFR promoter activity by Egr-1 represents a mechanism for epidermal cell growth during hypoxia. We have also analyzed a potential role of NF-kappaB family members in the regulation of the EGFR transcription. EMSAs demonstrated that the p50 and p49, subunit proteins of the NF-kappaB, bound to the EGFR promoter at four out of five of these sites. However, we further demonstrated that NF-kappaB could not transactivate the EGFR by cotransfection experiments with each NF-kappaB subunit, using p50, p65 and c-Rel and an EGFR promoter luciferase reporter. Treatment of cells with tumor necrosis factor (TNF-alpha), which degrades the I-kappaB and then result in translocation of NF-kappaB to nucleus, did not enhance EGFR promoter reporter gene transcription. In addition, TNF-alpha did not induce EGFR expression at the protein level. These results indicate that even though purified NF-kappaB can bind to the putative sites, there is no evidence that NF-kappaB transactivates the EGFR promoter region. Another factor that we have shown to interact with the EGFR promoter region is the transcriptional repressor GCF2. The GCF2 gene was localized to chromosome 3q27. We have recently determined that GCF2 can repress EGFR promoter activity as well as TAT-mediated activation of the HIV-LTR. We have also determined that GCF2 is able to repress the activity of a minimal thymidine kinase promoter construct. In cases of the EGFR promoter and HIV-LTR we were able to demonstrate binding of GCF2 to the promoter regions. However, GCF2 did not bind to TATA DNA. Utilizing immunoprecipitation and pulldown experiments, we were able to show that GCF2 interacts with TFIID, which does bind TATA DNA. Thus, GCF2 appears to have separate domains for DNA and protein interactions which leads to different mechanisms of transcriptional repression. In collaboration with Howard Cedar, Hebrew University, Jerusalem, Israel, we have shown that GCF2 is able to bind to a repressor element in the upstream region of the IGF2 gene. GCF2 binding and repressor activity are abrogated by DNA methylation of the site in the maternal IGF2 allele. These results established a role of GCF2 in the imprinting of the IGF2 gene.
