THE EFFECT OF POSITION AND ORIENTATION IN THE EXPRESSION OF TWO GENES IN A DUAL GENE CONSTRUCT

Abstract

Promoter occlusion (Adhaya and Gottesman, 1982) has been suggested as an explanation for differential gene expression of two genes in the same recombinant construct (Kadesch and Berg, 1986). In this study, four recombinant plasmids were constructed containing two genes, the prokaryotic neo gene and the human ras 1 gene. The pSV2neo plasmid (Southern and Berg, 1982) contains the cis functions necessary for expression in eukaryotic cells: SV40 enhancer, promoter, and k polyadenylation signal surrounding the bacterial neomycin phosphotransferase gene. Gene expression confers neomycin resistance in bacteria and Geneticin (G418) resistance in eukaryotic cells. Detection and quantitation of gene expression is measured by G418 resistant colony formation. The second gene is a clone of the oncogenic human Harvey ras 1 gene which transforms murine NIH 3T3 cells, producing morphologically altered foci that are easily counted. Thus, the dual construct is capable of producing transformed foci and G418 resistant colonies in NIH 3T3 cells selected in G418 medium. The promoter occlusion hypothesis states that one of the two genes may be preferentially expressed and that transcription of one gene in the construct could obstruct equal expression of the second gene. The recombinant neo-ras constructs were made in parallel and antiparallel orientation and with 1 kilobase (kb) and 3 kb of intervening sequence between genes. These plasmids were linearized and transfected, using calcium phosphate precipitation, and grown on selective G418 medium for 12 days. Parallel assays were performed in which G418 resistant colonies were counted as neo expression, and separate ras induced foci were enumerated. A total of five replicate tests were conducted per plasmid. The G418 resistant colony data were normalized to the pSV2neo construct to correct for possible inter-replicate differences not due to treatment effects or second gene expression. Differences between constructs in the mean number of G418 resistant colonies and foci were determined using the a single factor analysis of variance (ANOVA) for parametric data and a Kruskal-Wallis test for non-parametric data. There were no significant differences in foci or G418 resistant colony counts among the recombinant constructs. This suggests that, for NIH 3T3 cells transfected by calcium phosphate precipitation and assayed by focus/G418 resistant colony counts, the position and orientation of the two transcriptional units do not affect the expression of the two genes.