The primary goal of this project was to study the interactions of the spacer regions of the seven highly homologous E. coli rRNA (rrn) operons. These operons comprise the largest reiterated gene family of E. coli. The spacers consist of tRNA genes (two classes) and an 85 base pair sequence of unknown function (rsl for ribosomal spacer loop). In the first part of this work, strains were constructed which contained altered ratios of tRNA spacer genes. This was a noteworthy accomplishment in that the arrangement of these genes has been very stable throughout evolution, implying a strong selection. It was seen that one particularly severe rearrangement of these spacers produced a very small colony morphology. However, cells with rrn recombinations which increased the copy number of the deficient spacer class produced normal colonies. Using this colony morphology assay, 195 of these "revertants" were examined by Southern analysis. Four of these strains were also seen to contain chromosomal inversions between rrn loci. By several lines of reasoning, it was shown with a greater than 99.9% probability that at least one of these reversions must have been caused by gene conversion. This represents the first clear demonstration of conversion of large tracts of non-homology in E. coli. This has implications with regard to the mechanisms which prokaryotes use to maintain sequence homogeneity in gene families. The final phase of the project concerned the rsl sequences. Southern analysis of strains from the E. coli pedigree revealed that E. coli K12 originally had the sequence present at both rrnG (rslG) and rrnB (rslB). However, strain W6, which was isolated as a spontaneous relA mutant of 58-161, contains rslG only. There was a strong suggestive relationship between these two mutations in that relA affects the amino acid-regulated control of rRNA synthesis whereas rsl is part of an rrn operon. Attempts were made to determine if rslB confers some selective disadvantage on relA strains and if rslG is an essential sequence. No selective disadvantage was found and the question of rslG being essential is still an open one although negative evidence suggests it may be essential.
