To find additional homologs of our cloned S. acidocaldarius sRNA genes, we initially ran BLAST [Altschul et al., 1997] on each cDNA clone against the non-redundant nucleotide database, which included four completely sequenced archaeal genomes (Methanococcus jannaschii, Methanobacterium thermoautotrophicum, Archaeoglobus fulgidis and Pyrococcus horikoshii). From our previous experience with eukaryotic snoRNAs, we knew generalized similarity search methods like BLAST were usually only effective at finding homologs in closely related species. Although we did not receive any hits to the distantly related complete archaeal genomes, we did recover two weak hits against sequences in other Sulfolobus species: Sac sR03 had a hit near the Sulfolobus shibatae top6B topoisomerase II gene (opposite strand of 3' UTR, BLAST score 40.1 bits, expect value 0.038), and Sac sR01 had a hit near the Sulfolobus solfataricus aspartate aminotransferase gene (3' UTR; BLAST score 38.2 bits, expect value 0.15). Normally we would not have considered these weak hits to be legitimate (both against only the 3' half of the respective sRNAs); however, manual examination of the upstream sequences revealed proper C boxes. Intriguingly, the hit against Sac sR01, when extended upstream, overlapped by 21 nucleotides into the C-terminal end of the aminotransferase coding region. Because the aspartate aminotransferase protein has clear homologs in several other archaeal genomes (with similarity extending up to but not including sRNA overlap), we believe the protein is real with an accurately predicted C-terminal end. Overlap of a snoRNA gene and a protein coding region is unprecedented, and because no other hits were found for the other sRNA clones, it was unclear whether these were true sRNA homologs.
Because of the uncertainty, I carried out primer extension analysis to test for the presence of the Sac sR1 homolog, dubbed Sso sR1, using total S. solfataricus RNA as template. A product with a length similar to that of Sac sR01 was detected (Figure 5.2c; gel pictured is from a reproduced experiment from the Dennis lab). Moreover, using the dNTP concentration-dependent reaction, both the Dennis lab and I were able to verify that the predicted U52 position in S. solfataricus small subunit (SSU) rRNA is likely to carry a methyl modification (Figure 5.3a). The presence of functionally related sRNAs in two distinct species that apparently guide methylation to the same U52 position in 16S rRNA provides additional support for the existence of eukaryotic-like C/D snoRNAs in the Archaea.