Abstract
Pseudouridine synthases are enzymes responsible for modifying uridines to pseudouridines in a site-specific and energy-independent manner. There are 5 families of these synthases, named after the first member of each family to be characterized in Escherichia coli : RluA, RsuA, TruA, TruB and TruD. The 23S ribosomal RNA in E. coli contains 10 pseudouridine modifications made by 6 specific synthases named RluA-RluF. These modifications cluster around important functional regions of the ribosome such as the peptidyl transferase center, the tRNA binding sites, and the inter-subunit bridge regions. My research focuses on understanding the mechanisms of substrate selection by pseudouridine synthases and the roles of these modifications in ribosome biogenesis and function. The main aims of my research were: a) to examine the substrate specificity determinants of RluD, an important E. coli synthase; b) to characterize a mutant strain of E. coli lacking a majority of the pseudouridines on 23S rRNA; and c) to determine the activity of RluA from Vibrio cholerae, which has two closely related RluA paralogs rather than just one, as seen in most organisms. Pseudouridine modifications in the stem-loop of helix 69 (H69) in domain IV of 23S ribosomal RNA are highly conserved in all phyla. The three pseudouridines in H69 in E. coli have been shown to play an important role in 50S subunit assembly and its association with the 30S subunit. These three modifications are made by the pseudouridine synthase, RluD. Previous work showed that RluD is required for normal ribosomal assembly and function, and is the only pseudouridine synthase required for normal growth in E. coli. Here, we show that RluD is far more efficient in modifying H69 in structured 50S subunits, rather than in naked or synthetic 23S rRNA. We suggest that pseudouridine modifications in H69 are made late in the assembly of 23 rRNA into the mature 50S subunit. This is the first reported observation of a pseudouridine synthase being able to modify a structured ribonucleoprotein particle, and may constitute an important late step in the maturation of 50S ribosomal subunits. Deletion of RluD results in aberrant ribosome assembly and impaired translation termination leading to severe growth defects. However, single deletion strains of the remaining five 23S rRNA synthases do not display an altered phenotype. In an effort to identify possible roles for the remaining seven pseudouridines, we constructed a strain (Delta 5 mutant) lacking all 23S rRNA synthases except RluD. Surprisingly, this strain does not exhibit a significant growth defect at 37C in rich or minimal media. However, it does display a slower growth rate at 20C compared to wild-type. When grown in competition with the wild-type strain at 37C, a strong selection against the mutant strain was observed. In order to evaluate the structure of the mutant ribosomes, we determined the effect of various antibiotics that target the 50S subunit. The mutant strain is significantly more sensitive than wild-type to antibiotics targeting the 50S subunit such as chloramphenicol, hygromycin, clindamycin and tiamulin but these effects can be attributed to the loss of the RluC modification at U2504 by itself. In phenotypic microarray tests, we observed that the Delta 5 mutant grew much poorer than wild-type when cultured in a medium containing 6% NaCl. Taken together, the data suggest that these pseudouridines may play an important role in maintaining the structural integrity of the ribosome. In E. coli, the pseudouridine synthase RluA is a dual specificity synthase capable of modifying U746 on 23S rRNA and U32 on 4 cytoplasmic tRNAs. Surprisingly, the Vibrio cholerae genome encodes not one, but two closely related RluA proteins. In order to examine the possible activities of these two proteins, we complemented an rluA deletion in E. coli with plasmid-borne Vibrio rluA1 and rluA2 constructs. Interestingly, only one of these two RluA proteins (Vibrio RluA1) was able to modify E. coli 23S rRNA at U746. In order to determine the structural basis for this difference between the closely related RluA1 and RluA2, we constructed homology models using the structure of E. coli RluA in complex with an RNA stem-loop (PDB ID: 2I82) as a template. These models implicated two possible three amino acid (GVF or FAL) inserts present near the catalytic aspartate in Vibrio RluA2 as the likely cause of the differential activity. We hypothesize that this insert may sterically occlude the binding of substrate RNA to the enzyme, thereby preventing a productive modification reaction.