Although a number of human genetic diseases have been defined as mutations in or near the human beta-globin gene locus, regulation of gene expression of the human beta-globin gene is not well understood. To study this regulation, we are using a mutant human erythroleukemia cell line, K562. These cells can synthesize embryonic and fetal globins but not adult beta-globin, although they contain a structurally normal beta-globin gene which can be induced to express in transient heterokaryons. Thus, the molecular defect in K562 cells is most likely due to differences in trans-acting factors between K562 cells and normal erythroid cells such as continuous synthesis of a repressor, lack of synthesis of an activator molecule, or both. If there is negative regulatory factor in K562 cells, deletion of its DNA binding site might then allow expression. On the other hand, deletion of DNA containg a binding site for a positive acting factor should be seen as decreased expression if the gene were active. In an attempt to understand regulation of expression of the human beta-globin gene we are first studying its 5' DNA sequences. We have fused the 5' flanking region to a heterologous gene, chloramphenicol acetyl transferase (CAT). Consistent with other reports from this laboratory, we found no expression of this gene in uninduced K562 cells. Our deletion analysis of this DNA suggest there are at least three regulatory regions 5' to the beta- globin gene, two negtive control regions (NCH) and one positive control region (PCR). Only the PCR appeared to be specific for K562 cells when these deletions were studied in a Chinese hamster and a mouse erythroleukemia cell line. Analysis of these plasmids in hemin-induced K562 cells showed expression of the original CAT plasmid for the first time. This plasmid could also be expressed in uninduced K562 cells in the presence of the SV40 enhancer. Preliminary experiments suggest there may be protein(s) binding to both NCR1 and NCR2, as well as to the region between them.
