Determinants of a protein-induced RNA switch in the large domain of signal recognition particle identified by systematic-site directed mutagenesis
Determinants of a protein-induced RNA switch in the large domain of signal recognition particle identified by systematic-site directed mutagenesisKrishne Gowda and Christian Zwieb*
Department of Molecular Biology, The University of Texas Health Science Center at Tyler, Highway 271 N, PO Box 2003, Tyler, TX 75710, USA
Received March 27, 1997;Revised and Accepted June 4, 1997
ABSTRACT
Signal recognition particle (SRP) is a ribonucleoprotein complex that associates with ribosomes to promote the co-translational translocation of proteins across biological membranes. Human SRP RNA molecules exist in two distinct conformations, SR-A and SR-B, which may exchange during the assembly of the particle or could play a functional role in the SRP cycle. We have used systematic site-directed mutagenesis of the SRP RNA to determine the electrophoretic mobilities of altered RNA molecules, and we have identified the nucleotides that avert the formation of the two conformers. The conformer behavior of the various RNAs was addressed quantitatively by calculating a value [zeta] as an indicator of conformational variability. Single loose A-like forms were induced by changes in helix 5 [nucleotides (nt) at positions 111-128 or 222-231], helix 6 (nt at positions 141-150) and helix 7 (nt at position 169 and 170), whereas other mutations in helix 6 and helix 8 preserved the conformational variability of the mutant RNA molecules. The more compact B-like form was induced only when a small region (129-CAAUAU-134), located in the 5'-proximal portion of helix 6, was altered. Since this region is part of the binding site for SRP19, we suggest that protein SRP19 uses nucleotides at 129-134 to trigger the formation of the favored SR-B-form, and thus directs an early step in the hierarchical assembly of the large SRP domain.
INTRODUCTION
Conformational differences in RNA molecules reflect their functional importance in a wide variety of cellular processes (1 -3 ). Because of its small size, signal recognition particle (SRP), a cytosolic ribonucleoprotein complex involved in the translation of secretory proteins and their translocation across lipid bilayers (reviewed in 4 ), serves as a convenient model system for resolving molecular details of how conformational switching can occur in the RNA. The SRP has been characterized best in the mammalian system where it is an elongated particle with a small and a large domain separated by a variable adapter region (5 ,6 ). Chemical and enzymatic modification experiments, electron microscopy, comparative sequence analysis and model building were among the methods used to determine the structural and functional properties of the SRP components (5 -11 ).
SRP contains six proteins, of which SRP9, SRP14, SRP19, SRP54 and SRP68, are in direct contact with the SRP RNA (10 ,12 ). Of special interest is protein SRP54 as it requires SRP19 for assembly. Since no interactions between the free SRP19 and SRP54 proteins were detected, a conformational change in the large domain of the SRP RNA was postulated to accompany the sequential binding of these two polypeptides (13 ,14 ).
Experimental support for conformational changes was provided by the discovery of two distinct major SRP RNA forms, SR-A and SR-B, on non-denaturing polyacrylamide gels (15 ,16 ). The faster-migrating B-form was preferred by protein SRP19 over the slow-migrating A-form, and association of SRP19 with SRP RNA resulted in the reconstitution of a uniform complex (17 ). However, the location of the SRP RNA switch remained unknown. Furthermore, the role of protein SRP19 in the triggering of the RNA switch was unresolved.
The detailed molecular analysis of RNA switches is hampered by the fact that they tend to form the core of a folded RNA molecule. This is in contrast to the analysis of RNA-protein interactions where nucleotides may be accessible on the surface and therefore are amenable to enzymatic and chemical modifications. Therefore, to identify the sites in SRP RNA that influence conformation, we chose a systematic site-directed mutagenesis approach for introducing changes throughout the large domain of the human SRP RNA (18 ) as this approach had the advantage that all sites could be analyzed, including those that were hidden from external probes. Introducing mutations systematically assured that the results were unbiased towards the functional role of a particular site. A small region was identified in the SRP RNA that triggers the formation of SR-B. This trigger sequence may be used by protein SRP19 to induce the preferred B-conformer.
MATERIALS AND METHODS
Construction and synthesis of mutant SRP RNAs
Plasmids coding for authentic human SRP RNA or for mutant RNAs that lacked individual RNA helices were described earlier (14 ). Mutant [Delta]H67 was generated by PCR site-directed mutagenesis (19 ) using oligonucleotide 5'-CGGTTCACCCCTTGCCGAACTTAGTG-3' as a mutagenic primer. Details of the construction of other mutations in the large domain of the SRP were communicated previously (18 ). All mutations were confirmed by DNA sequencing using Sequenase (US Biochemicals).
Analysis of conformers
Display of the large SRP RNA domain
The three-dimensional model of the human SRP RNA structure (11 ) was obtained from the SRP database (23 ) in the PDB file-format (24 ) at the Internet address http://pegasus.uthct.edu/SRPDB/SRPDB.html. The model was displayed on an Silicon Graphics Indigo 2 Extreme workstation using the program DRAWNA (25 ). The model was accepted as provided with no additional structural constraints. Experimental data were incorporated by color-coding mutated sections of the SRP RNA that had significant effects on the formation of the A or B-conformers as described in Results and Figure 4 .
Compensates central segment of helix 60.05 +- 0.01A
157-ACCACC-162 -> CAGUGG
C3
141-CUCCCG-146 -> CGGGGG
Compensates distal segment of helix 60.03 +- 0.001A
151-CGGGGG-156 -> GCCCUC
Mutants are listed in the left column; changes relative to unmutated human SRP RNA (hSR) with the expected effects on RNA structure are tabulated in the second and third columns and. [zeta]-values, indicating the degree of the conformational variability (see Materials and Methods), are shown in the fourth column. Data are compiled from at least three independent experiments with the standard error of the means shown. In the right column, the mobilities of dominant conformers (0 <= [zeta] <= 0.1) relative to the A- or B-form of unmutated SRP RNA is indicated; for 0.1 < [zeta] <= 0.3, the dominant conformer is listed in parentheses. ND, not determined.
RNA of mutant 5A separated into two conformers of nearly identical mobility: athe mobilities of the large deletion mutants [Delta]H6 and [Delta]H8 relative to SR-A and SR-B were determined earlier (17); bRNAs from mutants [Delta]H67 and [Delta]H8 co-migrated on non-denaturing polyacrylamide gels (not shown) indicating that [Delta]H67 corresponds to the faster migrating B-form.
RESULTS
Systematic site-directed mutagenesis in the large SRP domain
The RNA secondary structure corresponding the large domain of the human SRP is shown in Figure 1 and includes a large portion of helix 5, as well as helices 6, 7 and 8. The majority of the mutations in this region were used earlier to characterize the binding sites of proteins SRP19 (18 ) and SRP54 (26 ). Alterations included a deletion of helix 6 (mutant [Delta]H6) or helix 8 (mutant [Delta]H8). PCR site-directed mutagenesis (18 ,19 ; and Materials and Methods) was used to construct mutant [Delta]H67, which lacks helices 6 and 7. The multiple residue changes of the other mutant RNAs were expected to disrupt base pairing in various helical sections as indicated in Figure 1 and listed in Table 0 . There were also three mutations in helix 6 that retained base pairing by changing both RNA strands in helix 6, and two tetranucleotide (tetraloop) mutants (6T and 8T). All RNAs were synthesized in vitro by run-off transcription with T7 RNA polymerase (Materials and Methods) using linearized derivatives of plasmid phSR as templates (14 ).
Analysis of SRP RNA conformers
RNAs were activated by incubation in SRP reconstitution buffer at 60oC followed by slow cooling to room temperature (Materials and Methods). For each mutant RNA, conformers were analyzed at room temperature by electrophoresis on native polyacrylamide gels. Human SRP RNA and E.coli 5S ribosomal RNA were used to confirm the appropriate separation of the conformers. After electrophoresis, the ethidium bromide-stained gels were photographed (Fig. 2 ) and profiles of the lanes were generated using Image software (22 , Fig. 3 ). To quantitate the degree of conformational variability, a value [zeta] was calculated for each RNA (see Materials and Methods). ([zeta]was zero when a single conformer was identified, and had a theoretical value of one when there were two or more equally-represented conformers.) The two highest [zeta]-values were observed for E.coli 5S rRNA (0.80 +- 0.05) and the unmutated human SRP RNA (0.73 +- 0.06). All conformer profiles were highly reproducible as indicated by the low standard deviation of the means determined from at least three independent experiments (Table 0 ).
Localization of conformer-sensitive sites in 3D
To visualize better the effects of the mutations on SRP RNA conformation, we incorporated the results from the site-directed mutagenesis experiments into the current model of the three- dimensional structure of the SRP RNA (11 ). This hypothetical model is characterized by a relatively `straight' helix 6 and by a `bent' helix 8. Figure 4 shows a view of the large domain with color-coded regions in the RNA to highlight strong or mild influences on the formation of SR-A or SR-B. The calculated [zeta]-values were used to group the conformer behavior of each mutant RNA as described in the legend to Figure 4 . Shown in red are those RNA regions that, when altered, favor A-like forms. Such sites were located throughout helices 5, 6 and 7, but were not found in helix 8. Only the 5'-proximal portion of helix 6, shown in green, constrained the RNA to a single B-conformer. Mild effects on SR-A or SR-B (shown in yellow and orange, respectively) were caused by mutations in the 3'-half of helix 6 and also by some sites in helix 8.
DISCUSSION
SRP RNA conformers
Assembly of SRP19 and SRP54 the large SRP domain
We consider it significant that out of the 149 residues altered, only six induced the single B-form. Protein SRP19 is the primary assembly protein of the SRP and binds alongside the 5'-portion of helix 6 (18 ). The affinity of protein SRP19 to the more compact B-conformer was ~3.5 times greater than to the looser A-conformer (17 ). Since the sequence 129-CAAUAU-134 (Fig. 4 ) is located within the SRP19 binding site, we propose that it is a trigger site which is activated upon SRP19 binding to helix 6. This view is consistent with the idea that the tetraloop of helix 6 serves as a nucleation site in SRP assembly (26 ), and that the region at the proximal portion of helix 6, when touched by SRP19, triggers the formation of SR-B. Nevertheless, we do not exclude the possibility that conformational switching may occur in the fully-assembled SRP (8 ).
In mammals, protein SRP54 interacts with the SRP RNA only after an SRP19-RNA complex has been formed (10 ). However, induction of SR-B by SRP19 is insufficient for the association with SRP54. As we have shown recently (26 ), the M-domain of human SRP54 (SRP54M) will not interact with a mixture of SRP RNA molecules that contains a substantial amount of SR-B. Furthermore, when SRP19 was absent, neither the pure B-form mutant 6A nor C1 bound even trace amounts of SRP54M. Thus, after the induction of the SR-B conformer, additional structural changes may occur in the RNA or in the protein during the organization of the SRP19-RNA complex. Furthermore, since SRP19 is an integrated part of the SRP, protein-protein interactions between RNA-bound SRP19 and protein SRP54 may be required to complete the assembly.
ACKNOWLEDGEMENTS
We thank Shaun D. Black for critical reading of the manuscript. This work was supported by NIH grant GM-49034 to C.Z.
REFERENCES
1 Cech, T., Zaug, A., and Grabowski, P. (1981) Cell, 27, 487-496.MEDLINE Abstract
2 Guerrier-Takada, C., Gardiner, K., Marsh, T., Pace, N., and Altman, S. (1983) Cell, 35, 849-857.MEDLINE Abstract
3 Noller, H., Hoffarth, V., and Zimniack, L. (1992) Science, 256, 1416-1419.MEDLINE Abstract
11 Zwieb, C., Müller, F., and Larsen, N. (1996) Folding Des., 1, 315-324.
12 Lütcke, H., Prehn, S., Ashford, A., Remus, M., Frank, R., and Dobberstein, B. (1993) J. Cell Biol., 121, 977-985.MEDLINE Abstract
13 Lingelbach, K., Zwieb, C., Webb, J. R., Marshallsay, C., Hoben, P. J., Walter, P., and Dobberstein, B. (1988) Nucleic Acids Res., 16, 9431-9442.MEDLINE Abstract