Hypoxia-lnducible Factor 1: Transcriptional activation by HIF-1
Forced expression of recombinant HIF-la and HIF-1 p in tissue culture cells resulted in a marked increase in transcription of reporter genes containing a wild-type 50-bp hypoxia-response element from the human EPO gene in cells exposed to 20% 02 and a superinduction in cells exposed to 1% 02. A mutation in the HIF-1 binding site of the hypoxia-response element eliminated the transcriptional response to hypoxia and the response to recombinant HIF-1. Forced expression of HIF-la alone was sufficient to elicit maximal transactivation of reporter genes, indicating that HIF-lp was present in excess and that the levels of HIF-la determined the transcriptional response to hypoxia (Fig 3).
Truncation of HIF-la after amino acid 390 resulted in a specific loss of transcriptional activation under hypoxic conditions. In addition, the truncated protein was expressed at constitutively elevated levels compared with the full-length recombinant HIF-la that was present at much higher levels in hypoxic compared with nonhypoxic cells. These results suggested that the transcriptional activity of HIF-1 is regulated by changes in the steady-state levels of HIF-la as well as changes in its transcriptional activity. Subsequent studies have confirmed these hypotheses by the identification of two transcriptional activation domains within the carboxyl terminal half of HIF-la, the activity of which is regulated independently of the protein levels Reading here antibiotics online. Thus, the level of transcriptional activation mediated by HIF-1 is determined by the steady-state level of HIF-la protein and the specific activity of the HIF-la transactivation domains, both of which are modulated by cellular 02 tension.
In human Hep3B hepatoblastoma cells exposed to 20% 02, HIF-la protein was undetectable, but after exposure of cells to 1% 02, HIF-la was detectable in the nucleus within 30 min and levels peaked at 4 to 8 h of continuous hypoxia, similar to the kinetics of induction as determined previously for HIF-1 DNA-binding activity. When cells were exposed to 1% 02 for 4 h and then returned to 20% 02, HIF-la levels decayed rapidly with a half-life of <5 min, indicating that the protein is exquisitely unstable in posthypoxic cells. HIF-la contains two 20-amino-acid sequences (residues 499-518 and 581-600) that each contain 15 (75%) PEST (proline, glutamic acid, serine, threonine) residues that have been implicated in protein instability. HIF-1 (3 protein was present in low levels, primarily within the cytoplasm of nonhypoxic cells, and nuclear protein levels were modestly induced by hypoxia with kinetics of induction and decay similar to those of HIF-la, suggesting that HIF-1 p protein may be stabilized by heterodimerization with HIF-la.
Figure 3. Transcriptional activation by HIF-1. Top, A: plasmid constructs for transient transfection. Hep3B cells were transfected with an SV40-promoter (SV40 Pro) P-galactosidase (|3-gal) control plasmid and a reporter plasmid consisting of a 68-bp hypoxia-response element (HRE) from the ENOl gene cloned upstream of a basal SV40 promoter driving expression of luciferase coding sequences. In addition, the same cells were also transfected with CMV-promoter (CMV Pro) expression vectors containing either no insert (0) or HIF-la (a) or HIF-ip (P) cDNA coding sequences or both (a + (3). Bottom, B: results of transient expression assays. Cells were exposed to 20% 02 (open box) or 1% 02 (stippled box) for 24 h and then assayed for expression of luciferase and p-galactosidase. The ratio of luciferase/p-galactosidase was calculated to quantitate transcriptional activation mediated by the hypoxia response element.