Search

'Ribosome feedback' effects on the expression of the genes specifying tRNA and EF-Tu in E. coli have been studied at increased rrnB doses (rRNA gene doses). We confirm previous observations that the introduction into the cell of a multicopy plasmid carrying the rrnB operon reduces the cellular content of most tRNAs, including those encoded by the tRNA-tufB operon, but leaves the 5S rRNA content unaffected. Increase of the dosage of intact, but not of deleted rRNA genes, causes a slight drop in total EF-Tu that can be fully accounted for by a decrease in EF-TuB level. The drop in EF-TuB content (approx. 25%) is much smaller than that in tRNA content (approx. 80%). The synthesis rate of total EF-Tu is hardly affected, indicating that the turnover of EF-Tu has not changed. The ratio of tRNA over tuf RNA synthesis rates remains the same after elevation of rrnB dosage. Considering the large decrease in tRNA content this means that both RNA synthesis rates decrease to approximately the same extent. The relatively small drop in EF-Tu synthesis must be due, therefore, to an enhancement of the number of EF-Tu molecules synthesized per mRNA molecule. Apparently a post-transcriptional mechanism, regulating EF-Tu synthesis, becomes operative under these conditions. Growth-rate-dependent regulation of the tRNA-tufB operon has been studied using lysogens carrying tRNA':lacZ and tRNA-tufB':lacZ operon fusions and a tufB':lacZ' gene fusion. These experiments show that the cellular contents of tRNA, tufB RNA and EF-TuB vary in direct proportion to the growth rate. This indicates that growth rate control of tRNA-tufB operon transcription resembles that of stable RNA operons and not of r-protein operons, and that the read-through of the terminator at the end of the tRNA gene cluster remains unaltered.

The mechanism underlying feedback inhibition of tufB expression has been studied in vivo by gene-dosage experiments and by gene and operon fusions involving lacZ. Raising the cellular EF-Tu content, by introducing a multicopy plasmid encoding EF-TuA into the cell, repressed the level of EF-TuB but left the content of tRNA(Thr)3, encoded by the tRNA-tufB operon, unaffected. This indicates that autoregulation of chromosomal tufB expression does not occur by modulating transcription initiation at the promoter of the tRNA-tufB operon. This conclusion is further substantiated by experiments with a tRNA':lacZ operon fusion. The molecular ratio of chromosome-borne tufA and tufB transcripts also remained unaltered under conditions of excess EF-Tu, though experiments with a tRNA-tufB':lacZ operon fusion showed a decrease of tufB transcripts. Our data further exclude drastic effects of the autogenous repressor on processing of the contranscript of the operon into monocistronic tufB RNA and on alteration of EF-TuB turnover. Two possible mechanisms remain, which cannot yet be decided between. One is modulation of EF-Tu by transcription termination either directly or indirectly by affecting antitermination. The second is translational repression. In vitro translation of transcripts derived from SP6 clones did not reveal any feedback inhibition of EF-TuB synthesis. Surprisingly, addition of EF-Tu to a coupled transcription/translation systems was found to block transcription initiation at the primary promoter of the tRNA-tufB operon by over 90%. Although this in vitro effect of EF-Tu could not be demonstrated in vivo, possibly because of a difference in higher-order structure between plasmid-borne and chromosome-borne DNA, it indicates that under certain conditions EF-Tu binds very specifically to the tRNA-tufB operon promoter or its upstream region.