The mutagenesis studies described here show that deletions removing entire hairpins in the RNA of RNase MRP result in no (ymP5, ymP6 and eP19) or very poor (ymP7 and ymP8) growth. Previous experiments demonstrated that hairpin P3 is also essential (25
). These results suggest that there is no extensive domain in the RNA that is dispensable for RNase MRP function. We were somewhat surprised that the termini of all six hairpins are dispensable [(25
), and this report], since terminal loops can contribute to protein binding (34
) or play an important role in three-dimensional folding of RNA (35
). Nevertheless, our finding that the termini of the hairpins are dispensable is consistent with the variability of sequence, length and helix irregularities in these portions of the RNase MRP RNAs from the Saccharomycetaceae family (24
The ymP5 hairpin is the most conserved hairpin in the RNase MRP RNAs of members of the Saccharomycetaceae family (Figure ). More than half of this hairpin is required, perhaps related to the well-conserved three nucleotides bulge in this hairpin. A mutation in this bulge (P5-149,150,151) results in very poor and temperature-sensitive growth. Furthermore, a G to A transition at position 122 in the base pair adjacent to the bulge generates the temperature-sensitive rrp2-2
). The bulge could play an important role in protein binding or in the formation of a long distance structural constraint in the RNA molecule, since helices distorted by this kind of irregularities can serve as protein-binding areas (34
). No equivalent of hairpin ymP5 with a three nucleotides bulge has been identified in the MRP RNA of Schizosaccharomyces pombe
or higher eukaryotes (24
). Interestingly, a 5 nt bulge implicated in binding of the Est1 telomerase protein to the telomerase RNA also appears to be conserved in yeast species closely related to S.cerevisiae
but not in higher eukaryotes (36
). Further experiments are needed to clarify the degree to which protein–RNA interactions have been conserved in these and perhaps other nuclear RNPs.
Figure 7 Consensus structure of hairpin ymP5 of the Saccharomycetaceae. The structure was based on the predicted structures of the members of the family described in (38). Bases conserved in all Saccharomycetaceae are shown by letters. Non-conserved nucleotides (more ...)
It is interesting that all of hairpin ymP7 or ymP8 could be deleted, and the cells could still grow, albeit slowly. Moreover, deletions removing almost all of those two hairpins or hairpin ymP6 had no detectable effect on cell growth or 5.8S rRNA processing. The requirement of just a few base pairs at the base of these hairpins suggests that they may contribute to proper folding of the RNase MRP RNA, but that they may not contain specific targets for any of the nine protein components of the enzyme.
More of hairpin eP19 was required, but, still, no more than 5 bp at the base were necessary for wild-type growth. Hairpin eP19 exhibits interesting characteristics. The growth of all other mutants was affected similarly on both glucose (fermentable) and glycerol (non-fermentable) media, but eP19-Δ1 grew on glycerol but not on glucose [this report; (25
)]. The mechanistic background for this effect of carbon source is not known, but may be related to assembly. For example, the relative rates of synthesis of components of RNase MRP in glycerol might be more conducive to assembly of eP19-Δ1 RNA-containing particles.
Having observed that the termini of MRP RNA are not essential when deleted individually, we combined deletions to determine the minimal sequence required for enzyme function. Mini1 MRP RNA, missing the termini of hairpins ymP5, ymP6, ymP7 and eP19 (66/339 bases = 19%), had an activity that was indistinguishable from the wild-type RNA. The Mini2 RNA mutant, lacking the termini removed in the Mini1 MRP RNA plus the terminus of ymP8 (82/339 bases = 24%), could support good growth at 30°C, although it could not grow at low (16°C) or high (37°C) temperatures.
Since neither the P8-Δ2T deletion nor the Mini1 collection of deletions by themselves display any defect, the temperature-sensitivity of the Mini2 cells must be caused by their additive effects. We hypothesize that, while not required individually, these hairpins cooperate to maintain the overall three-dimensional structure of the MRP RNA, which in turn is necessary for proper protein binding and/or enzymatic activity. This conclusion is corroborated by the finding that the functionality of the Mini2 RNase MRP is improved by manipulating RNase MRP protein genes. Although these experiments provide no direct information about the mechanism behind the suppressor effects, they support our conclusion that the central core remaining in the Mini2 RNA subunit is the only part of the RNA subunit required for substrate recognition and catalysis. The Mini2 RNA core structure is shown in Figure .
Figure 8 Predicted structure of RNase MRP RNA core remaining in the Mini2 mutant. Local RNA folding of the remaining portions of hairpins ymP5, ymP6, ymP7, ymP8 and yeP19 was predicted using mfold (39).
The Mini2 RNA contains only 275 nt (including bases introduced for the three tetra-loops), so one wonders how all nine proteins with a collective molecular weight of 2.8 × 105
could be arranged on such a small RNA molecule. It is very likely that some proteins are not bound directly to RNA, but rather, require protein–protein interactions. Indeed, two- and three-hybrid studies are consistent with this idea (37
When shifted to a non-permissive temperature, cells synthesizing only Mini2 MRP RNA showed a significant increase in the accumulation of the long form of 5.8S rRNA. Since the 5.8SL
rRNAs are both processing end products, the ratio between the short and long forms changes only as the rRNAs produced before
the temperature shift are diluted with products processed after
the temperature shift. For this reason, the change in the ratio of long to short 5.8S rRNAs already visible by 2 h after the shift to 37°C suggests that pre-existing Mini2 MRP RNA-containing enzyme is rapidly inactivated after a shift to non-permissive temperature. At 16°C the change in the 5.8S rRNA composition took longer, as expected since all biological processes are severalfold slower at this lower temperature. To our knowledge, this is the first RNase MRP mutation that exhibits a strong temperature-dependent defect in 5.8S rRNA processing. Other mutants that show temperature-sensitive growth exhibit a defect in 5.8SS
processing that is more or less independent of the growth temperature [Figure B and the rrp2-2
mutant described previously in (4