Over the past five years, several new defects in the beta-thalassemias have been described from this laboratory using both restriction enzyme and sequencing analyses of cloned beta-thalassemia genes. The enzyme HphI has been shown to recognize a single nucleotide change at the 5' end of beta-IVS 2, and, using restriction enzyme analysis, demonstrated for the first time a specific defect associated with beta(0)-thalassemia. Cloning and sequencing of a beta-thalassemia gene have identified a single base change within IVS 2 at a position 705 nucleotides from the 5' end of IVS 2 that results in a beta(0)-thalassemia phenotype; no normal splicing occurs in this gene despite the fact that both the 5' and 3' ends of IVS 2 are unchanged. A unique and strong cryptic 3' acceptor splice site present in the normal gene at a position 580 nucleotides from the 5' end is used extensively in the mutant gene. Studies of this gene have indicated that there are sequences within IVS that are responsible for optimal expression of this gene; changes in these sequences can lead to markedly abnormal patterns of splicing. In addition, beta-globin gene expression has been evaluated in human erythroleukemia cells, K562 cells, and, although stable transformants with integrated beta-globin genes have been obtained, none of these transformants expressed the added beta-globin genes. This is presumably due to trans-acting factors or distal cis-acting effects that suppress the expression of these added beta-globin genes. In addition, a low epsilon-producing cell line, Bos cells, was used as a recipient for an exogenous epsilon-globin gene. A neomycin resistance gene was cotransfected into these cells, and a neomycin analogue (G418) was used to select cells containing both the neomycin resistance and epsilon-globin genes. Using Southern blotting, 10 of 11 stably transformed G418-resistant lines, which contain intact epsilon-globin genes, express epsilon-globin mRNA at much higher levels than the Bos cells into which they were transfected. Two of these lines express the epsilon-globin genes at a level comparable to that of wild-type K562 cells. These results indicate that the transfer and expression of human globin genes in human erythroid cells is feasible, and can occur at a high level.(ABSTRACT TRUNCATED AT 400 WORDS)