Competition assays define the interaction between the human beta-globin gene and its locus control region

The human beta-globin locus control region (beta-globin LCR) is a strong erythroid-specific enhancer that can protect globin transgenes from position effects. Since its discovery over a decade ago, it has been used as a prototype to dissect general functions of LCRs as well as enhancers. Currently, two theories have been debated, the scanning model and the looping model. In the scanning model, the beta-globin LCR is portrayed as a recruitment center for protein factors that move down the DNA to facilitate gene transcription along their path. According to the looping model, LCR-bound transcriptional activators may exert enhancement through physical interaction with gene-tethered factors. Both views have been employed to explain results obtained from two widely used systems, transgenic mice and transfected cells. In order to gain further insights into the mechanisms of the beta-globin LCR action, we set up a competition assay in transfected cells. Two adult human beta-globin genes, one wild type and the other marked, were cloned on either side of a 6.4-kb cassette of the beta-globin LCR alone or in conjunction. The constructs were introduced into MEL (murine erythroleukemia) cells and stable populations were selected for drug resistance. A competition between the two genes was observed. Unexpectedly, however, the beta-globin gene placed 5' of the LCR alone was expressed, on average, 4-fold lower than the downstream copy. In contrast, there was no notable difference in expression of both genes when the marker gene tkNeo was replaced by the pgkPuro and selected upon. Furthermore, the beta-globin gene upstream of the LCR was transcribed at normal levels and without position effects when co-transfected with pgkPuro, whereas co-transfection with tkNeo caused higher variation in expression among individual cell clones. The results also showed that the adverse effect of the tkNeo gene was related to its relative distance to the beta-globin LCR. Therefore, tkNeo, but not pgkPuro, competes with the beta-globin gene for LCR enhancement and is not an appropriate selectable gene in LCR function studies. On the other hand, co-transfection assays with pgkPuro as the drug selectable gene reiterate transgenic studies. The implication of this finding is discussed in an attempt to understand some contradictory results obtained with transgenic mice and transfected cells.;The pgkPuro co-transfection assays were subsequently used to examine gene competition. Various regions of the 3'-placed beta-globin gene were systematically deleted and expression of these truncated genes was studied alone or with the 5' wild type beta-globin gene also in cis. The constructs were cotransfected into MEL cells with a pgkPuro plasmid and individual cell clones were analyzed. We found that some transcriptionally incompetent mutant genes were able to compete strongly with the wild type beta-globin gene. Only the removal of the core promoter from the beta-globin gene abolished such competition. Thus, our data support the direct interaction notion and suggest that the core promoter region of the beta-globin gene could be primarily involved in the gene-LCR interaction. In addition, we propose that other regions of the beta-globin gene may play complementary roles in such interaction. Based on these observations and relevant publicized investigations, a detailed looping model is discussed.