Using gene chip technology, Blattner and co-workers have shown that RrmJ’s mRNA level increases up to 20 fold upon heat shock induction (Richmond et al., 1999). We have now provided some preliminary evidence that RrmJ’s heat shock function might involve the methylation of a second RNA substrate, which appears to be independent from the ribosome.
Therefore, it is likely that the putative second substrate of RrmJ is located free in the cytosol.
RrmJ’s cytosolic yeast homologue, Trm7p, has been shown to be responsible for the 2’-O-ribose methylation at the positions 32 and 34 of certain tRNAs (Pintard et al., 2002b).
Moreover, it has been shown by Hans Bügl that RrmJ is able to recognize tRNA as methylation substrate in vitro (Bügl et al., 2000). Therefore, it is not unreasonable to assume that tRNA might serve as the possible second methylation target for RrmJ. In order to identify the RNAs, that might be modified by RrmJ under heat shock conditions, a combination of HPLC and mass spectrometry of rrmJ wild type and deletion strain grown at heat shock temperatures should be performed. While in strains grown under normal conditions, all tRNA modifications have been identified successfully with this method (Rozenski et al., 1999; Sprinzl and Vassilenko, 2005), RNA modifications in strains that have been subjected to heat shock have not yet been analyzed. However, there have been studies that reveal the importance of tRNA modifications in regard to heat shock. Most importantly, studies with the pseudouridine synthetase TruB as well as the tRNA methyltransferase Trm1 in E. coli have demonstrated that tRNA nucleoside modifications in the elbow region of tRNA is important for the thermal stability of tRNA in vivo and, therefore, crucial for the survival of the cell upon heat shock (Kinghorn et al., 2002). Furthermore, unmodified in vitro transcribed tRNA has been shown to be thermal instable (Kintanar et al., 1994) and susceptible for nucleases (Davanloo et al., 1979). Therefore, RrmJ’s heat shock function might be important for tRNA stability at elevated temperatures. If RrmJ, however, methylates only very few distinct tRNAs like its cytosolic homologue Trm7p in yeast, this difference in modification pattern between the rrmJ wild type and rrmJ deletion strain may not be detectable by HPLC/mass spectrometry. In that case, thin layer chromatography could be performed. This approach has been successfully used by Pintard and coworkers in order to identify the tRNA substrates for Trm7p (Pintard et al., 2002b). This method is based on the RNase T2 digestion of total tRNA extracted from strains grown in the presence of
[32P]orthophosphate. RNase T2 cleaves all phosphodiester linkages between nucleotides, except those that are 2’-O-methylated. The result of the RNase T2 digestion can then be analyzed on thin layer chromatography and the methylated nucleotides can be identified. Due to their 3’-neighboring nucleotide, it should be possible to assign these identified methylated nucleotides to distinct tRNAs. In vitro methylation assays with total tRNA prepared from the rrmJ wild type and rrmJ deletion strain that have been grown under heat shock conditions could serve as another possibility to test if RrmJ’s heat shock function involves methylating tRNAs. If free RrmJ methylates certain tRNAs, these should be modified in the rrmJ wild type strain when subjected to heat stress and RrmJ is not bound to the ribosome. They should, however, be unmethylated in the rrmJ deletion strain and serve as in vitro substrate.
RrmJ has been shown to act as ribosome assembly protein by binding tightly to the 50S ribosomal subunit. However, the exact location of RrmJ on the surface of the 50S ribosomal subunit has not yet been identified. Although it was possible to crosslink RrmJ to the 50S ribosomal subunit using glutaraldehyde as chemical crosslinker it was not possible to upscale this crosslinking reaction in quantities, which is necessary to identify of the crosslinked complexes by mass spectrometry. Therefore, it may be useful to test different crosslinking reagents and techniques. The Trigger Factor (TF), for example, could be successfully crosslinked to the ribosomal proteins L23 and L29 using the crosslinking reagent benzophenone-4-iodoacetamide (BPIA) (Kramer et al., 2002). The crosslinking was done by substituting an aspartate adjacent to the putative ribosomal binding site of TF with a cysteine.
Because TF lacks additional cysteine residues, it was then possible to couple the thiol-specific ultraviolet-activatable crosslinker BPIA thiol-specifically to this site of the TF mutant protein. This renders a nice strategy that would be also possible for RrmJ. RrmJ has two cysteine residues in its amino acid sequence. The 3D structure of RrmJ, however, reveals that both of them are buried inside the protein. Therefore, in analogy to the study with TF, it would be possible to mutate an amino acid, which has been shown to be in a region that is important for the binding of RrmJ to the 50S ribosomal subunit. However, this substitution should not influence the binding, therefore, it is important to choose a non-conserved amino acid for this mutagenesis. I showed that the N-terminus of RrmJ is crucial for 50S binding.
Therefore, it would be reasonable to substitute an amino acid within this region with a cysteine and couple BPIA specifically to this residue. Then, RrmJ could be crosslinked to ribosomal proteins via the crosslinker BPIA by ultraviolet irradiation and the bands could be
made visible on a one dimensional SDS PAGE. Subsequently the crosslinked complexes could be identified by mass spectrometry.