The R388 plasmid-encoded drug-resistant type II dihydrofolate reductase gene (R . dhfr) was expressed in Saccharomyces cerevisiae by fusing the R . dhfr coding sequence to the yeast TRP5 promoter. Yeast cells harboring these recombinant plasmids grew in media with 10 micrograms of methotrexate per ml and 5 mg of sulfanilamide per ml, a condition which inhibits the growth of wild-type cells. Addition of a 390-base-pair fragment from the 3'-noncoding region of TRP5 downstream from R . dhfr increased expression. Presumably, the added segment promoted termination or polyadenylation or both of the R . dhfr transcript. The activity of the plasmid-encoded dihydrofolate reductase and the copy number of the R . dhfr plasmid in cells grown in drug-selective media were higher by one order of magnitude than those grown in nutrition-selective media. Plasmid copy number, as well as the plasmid-encoded enzyme level, decreased when cells were selected for prototrophy. In drug-selective media, the plasmid-encoded enzyme level and the content of R . dhfr transcripts were nearly constant in cells harboring R . dhfr plasmids containing different yeast promoters. In contrast, the plasmid copy number and beta-lactamase activity encoded in cis by plasmids were much higher when R . dhfr was associated with the weak TRP5 promoter than when it was fused to the strong ADC1 promoter. These results indicate that plasmid copy number, i.e., gene dosage of R . dhfr, correlates inversely with the strength of the promoter associated with R . dhfr, and cells with a higher plasmid copy number were enriched in drug-selective media. The transformation efficiency of R . dhfr fused to the ADC1 promoter was almost the same on drug-selective plates as on nutrition-selective plates, indicating that R . dhfr is suitable as a dominant selective transformation marker in S. cerevisiae.
Structure-Based Design of Dimeric Bisbenzimidazole Inhibitors to an Emergent Trimethoprim-Resistant Type II Dihydrofolate Reductase Guides the Design of Monomeric Analogues