Production of the glycoprotein hormone erythropoietin (Epo) Erythropoietin (Epo) is the hormone which regulates the oxygen-carrying capacity of the blood by controlling erythropoiesis: hypoxia stimulates production by a mechanism which involves increased Epo gene transcription (1). Transient-transfection studies in cultured cells have defined cisacting control sequences responsible for this effect, the most powerful of which is an oxygen-regulated transcriptional enhancer located 3' to the poly(A)-addition site of the gene (2-6). Extensive screening of tissue culture cell lines has revealed only two (the hepatoma lines Hep 3B and Hep G2) that produce Epo in an oxygen-regulated manner (7), yet the Epo 3' enhancer, when coupled to broadly active promoters, confers responsiveness to hypoxia in most (possibly all) cultured cells (8, 9). Furthermore, hypoxia-inducible factor 1 (HIF-1), which binds to a functionally critical region of the Epo 3' enhancer, has been found both in cells which produce Epo and in those that do not (10, 11). These findings suggest that oxygen sensing, signal transduction, and gene activation mechanisms very similar, if not identical, to those present in Epo-producing cells must be widespread. The implication is that these mechanisms are involved in the regulation of other genes in cells which do not produce Epo.We report here that the human phosphoglycerate kinase 1 (PGK-1) and mouse lactate dehydrogenase A (LDH-A) genes are regulated by hypoxia in a manner which is strikingly similar to Epo gene regulation. Using transient transfection, we have located cis-acting control sequences responsible for hypoxia-inducible expression in the 5' flanking region ofeach gene. For the PGK-1 gene we have characterized an 18-bp element which is necessary for hypoxia-inducible expression and which has sequence similarity to a region within the Epo 3' enhancer. Oligonucleotides containing the functionally active PGK-1 and Epo sequences cross-compete for a hypoxia-inducible factor (or factors) in electrophoretic mobility-shift assays.
MATERIALS AND METHODSCell Lines and Culture Conditions. The cell lines used were Hep G2 (human hepatoma), HeLa (human cervical carcinoma), and L cells (mouse fibroblast) grown to %700% confluence. The medium was then replaced, and cells were subjected to the following conditions for [14][15][16] hr: (i) normoxia (20% 02/5% C02/75% N2); (ii) hypoxia (1% 02/5% C02/94% N2 in a Napco 7100 incubator); (iii) hypoxia with cycloheximide (100 LAM); (iv) normoxia with cobaltous chloride (50 pM); (v) normoxia with cyanide (100 KM); or (vi) hypoxia with cyanide (100 pM).Translent Transfection and RNA Analysis. In all experiments the test plasmid (10-100 pg), encoding either human al-globin or human growth hormone (GH) as a reporter, was cotransfected with a control plasmid (10-50 ,g) by electroporation (4). After electroporation, transfected cells were split equally and incubated in parallel for 14-16 hr under normoxic or hypoxic conditions or were exposed to chemical agents, as ...