ABSTRACT. Erythropoietin (Epo) is a glycoprotein hormone that is the primary regulator of red blood cell production. Epo production increases in response to tissue hypoxia. This increase occurs primarily at the transcriptional level. Hypoxia inducible factor (HIF-1) is a DNA binding protein that binds to a hypoxia inducible enhancer in the 3' flanking sequence of the Epo gene. HIF-1 is a heterodimer that consists of an alpha and beta subunit. HIF-1 DNA binding activity is induced in response to hypoxia. In order to determine if one or both HIF-1 subunits is capable of ligand binding, subsequently leading to Epo production we performed co-transactivation experiments. Transfections were performed in Hep 3B, an Epo producing human hepatoma cell line and Cos-7, a non-Epo producing monkey kidney cell line. Cells were co-transfected with the 38 bp Epo enhancer fragment bearing the HIF-1 binding motif, subcloned in the luciferase reporter plasmid and either the HIF-1alpha cDNA, HIF- 1beta cDNA, HIF-1alpha and HIF-1beta cDNAs or pREP-4 respectively. Cells were incubated in an hypoxic [1%O(2)] or normoxic [21%O(2)] environment and assayed for luciferase activity. Epo levels were measured in the culture media from the transfected plates by an ELISA assay. Under hypoxic conditions Hep 3B cells transfected with the HIF-1alpha cDNA alone showed a 2.2 fold increase in luciferase activity, HIF-1beta showed a 3.4 fold increase and cells transfected with HIF-1 alpha and beta showed a 6.9 fold increase in activity over cells transfected with pREP-4. The baseline luciferase activity in transfected 3B cells incubated in normoxia was very low. However, a similar fold increase in luciferase activity in cells transfected with both HIF-1alpha and beta was noted. Under normoxic or hypoxic conditions in Cos-7 cells, a 1.5 fold increase was obtained with the HIF-1alpha and beta constructs transfected independently and a 3.5 fold increase was noted in cells transfected with both constructs. Epo levels increased several fold in all Hep 3B cells that were incubated in hypoxic conditions. However, there was no additional increase in Epo levels in transfected Hep 3B cells. We therefore conclude that although the HIF-1alpha and beta subunits can independently co-transactivate the Epo enhancer, binding of both subunits and a hypoxic environment is necessary for maximal transactivation. Overexpression of the HIF-1 protein alone in normoxic or hypoxic conditions is insufficient for an increase in Epo secretion. Activation/inactivation and interaction of other tissue specific factors is necessary for an increase in Epo gene expression in response to hypoxia.

Keywords: Co-transactivation, hypoxia, erythropoietin.

Reprint requests to: Ashima Madan, M.D., Division of Neonatology, 750 Welch Road, Suite 315, Palo Alto, CA 94304, phone; (415) 498-5641, fax: (415) 725-7419, e-mail: mn.axm.forsythe.stanford.edu.

ABSTRACT. Pyrroloquinoline quinone (PQQ), a novel cofactor of biological redox processes, is ubiquitous in animal cells. We have examined the effects of PQQ on protein synthesis. PQQ inhibits protein synthesis in hemin-supplemented rabbit reticulocyte lysates. This inhibition is characterized by increased phosphorylation of eIF-2alpha and by diminished guanine nucleotide exchange activity of eIF-2B. The increased eIF-2alpha phosphorylation is the result of activation by PQQ of the heme- regulated eIF-2alpha kinase (HRI). The addition of 10 microM PQQ completely inhibits the increase in protein synthesis that occurs on the addition of hemin (20 microM) to heme-deficient lysates, whereas a lower concentration of PQQ (100 nM) causes a very slight stimulation of protein synthesis. The increased eIF- 2alpha phosphorylation that occurs at high concentrations of PQQ inhibits eIF-2B activity, presumably due to formation of a 15S complex [eIF-2(alphaP).eIF-2B] in which eIF-2B becomes non- functional. Low concentrations of PQQ (0.1-1 microM) do not affect eIF-2alpha phosphorylation, but rather enhance the guanine nucleotide exchange activity of eIF-2B in reticulocyte lysates. In Chinese hamster ovary cell extract which is devoid of significant eIF-2alpha kinase activity, addition of both low and high concentrations of PQQ results in an increase in eIF-2B activity. The addition of PQQ to reticulocyte lysates activates HRI whereas addition of PQQ to purified HRI in vitro inhibits the autokinase and eIF-2alpha kinase activity of the HRI; the inhibition of purified HRI by PQQ is observed both in the presence and absence of hemin. These findings suggest that PQQ inhibits purified HRI by acting as an oxidant whereas in lysates in which PQQ is readily reduced, the PQQ acts as a reductant and increases the activities of both HRI and eIF-2B.

Keywords: Reticulocyte lysates, Chinese hamster ovary cell extract, eIF-2alpha phosphorylation, guanine nucleotide exchange activity, oxidation-reduction.

Reprint requests to: Irving M. London, M.D., Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, 77 Massachusetts Avenue, E25-551, Cambridge, MA 02139, phone: (617) 253-4305, fax: (617) 253-3459, e-mail: imlondon@mit.edu.

ABSTRACT. Regulatory elements linked to the beta-globin gene cluster modulates gamma-globin gene expression. The location of all of these elements and their mechanisms of action are still incompletely defined. Phylogenetically conserved DNA within the beta-globin gene cluster locus control region (LCR), but outside the core sequences of its hypersensitive sites (HS), were identified and we searched for any differences between HS 3 and HS 2, and HS 2 and HS 1, among patients with sickle cell anemia with different levels of Hb F who were homozygous for the common haplotypes. DNA was amplified with and without GC clamps, digested with restriction endonucleases, and examined by denaturing gradient gel electrophoresis (DGGE). We found limited fragment size diversity. However, the type of differences found and their distribution among haplotypes did not suggest that they represented distinctive changes that might explain the differential expression of the gamma-globin genes in sickle cell anemia.

Keywords: Fetal hemoglobin, gene expression, gamma globin gene.

Reprint requests to: Martin H. Steinberg, M.D., Associate Chief of Staff for Research, Professor of Medicine, VA Medical Center, 1500 E. Woodrow Wilson Drive, Jackson, MS 39216, phone: (601) 364-1315, fax: (601) 364-1390, e-mail: mhs@fiona.umsmed.edu.