The conserved adenosine in helix 6 of Archaeoglobus fulgidus signal recognition particle RNA initiates SRP assembly

lix 6 of archaea and eukaryotes is essential for the binding of protein SRP19 and the assembly of a functional complex. The conserved adenosine at the third position of the tetraloop of helix 6 (A149) is crucial for the binding of protein SRP19 in the mammalian SRP. Here we investigated the significance of the equivalent adenosine residue at position 159 (A159) of Archaeoglobus fulgidus SRP RNA. The A159 of A. fulgidus and A149 of human SRP RNA were changed to C, G or U, and fragments containing helix 6 or helices 6 and 8 were synthesized by run-off transcription with T7 RNA polymerase. The ability of recombinant A. fulgidus and human SRP19 to form ribonucleoprotein complexes was measured in vitro. The simultaneous presence of A149 and helix 8 is required for the high-affinity binding of SRP19 to the human SRP RNA. In contrast, A. fulgidus SRP19 binds to the SRP RNA fragments with high affinity irrespective of the nature of the nucleotide, demonstrating that A159 does not directly participate in protein binding. Instead, as indicated by the resistance of the wild-type A. fulgidus RNA towards digestion by RNase A, this residue allows the formation of a tightly folded RNA molecule. The high affinity between A. fulgidus SRP19 and RNA molecules that contain both helices 6 and 8 suggests that A159 is likely to initiate archaeal SRP assembly by forming a conserved tertiary RNA– RNA interaction.


Introduction
The signal recognition particle (SRP) binds to the signal sequence on the surface of a translating ribosome and thus directs the secretory protein to the proper subcellular compartment (reviewed by Lütcke 1995, Eichler and Moll 2001, Keenan et al. 2001, Nagai et al. 2003, Zwieb 2003).The SRP is composed of an RNA molecule, the SRP RNA (named 4.5S RNA in bacteria) and at least one SRP protein, called SRP54 or Ffh.The majority of archaea and all eukaryotes also contain SRP19 (Rosenblad et al. 2003), a protein that initiates the assembly of SRP (Walter and Blobel 1983, Walker et al. 1995, Bhuiyan et al. 2000).The archaea and eukaryotes share similar SRP RNA secondary structures (Larsen and Zwieb 1991), suggesting a conservation of the assembly mechanism.During the assembly of mammalian SRP, SRP19 must first bind to SRP RNA with high affinity before SRP54 can interact.However, in the archaeal SRPs of Archaeoglobus fulgidus, Acidianus ambivalens and Haloferax volcanii, SRP54 is able to interact with SRP RNA in the absence of protein SRP19 (Bhuiyan et al. 2000, Tozik et al. 2002, Moll 2003).Biochemical (Siegel andWalter 1988, Diener andWilson 2000) and site-directed mutagenesis studies (Zwieb 1991(Zwieb , 1992(Zwieb , 1994) ) demonstrate that SRP19 binds strongly to the tetraloop of SRP RNA helix 6, but also interacts at a lower affinity with the distal portion of helix 8.
Significant progress has recently been made in understanding the interactions between SRP RNA and protein SRP19 at high resolution.X-Ray crystallographic studies of the SRP RNA helix 6-bound human SRP19 (h19) (Wild et al. 2001) reveal that the protein contains three β-strands and two α-helical regions, arranged in a βαββα topology, similar to the RNP motif observed in other RNA-binding proteins (Burd and Dreyfuss 1994).The triple-stranded antiparallel β-sheet and loop 1 of h19 bind to the tetraloop of SRP RNA helix 6, whereas loop 3 of h19 engages near the tetraloop of helix 8 (Hainzl et al. 2002, Oubridge et al. 2002).This RNA-bound conformation of h19 is similar to the solution structure of the free A. fulgidus SRP19 (Af19) (Pakhomova et al. 2002).The conserved adenosine residue at the third position of the tetraloop of helix 6 (A149 of human SRP RNA) is essential for binding of h19 (Zwieb 1992).In the complex composed of Methanococcus jannaschii SRP19 (Mj19) and the large (S) domain of human SRP RNA (hSR) (Oubridge et al. 2002), A149 participates in an RNA-RNA interaction with the adenosine at position 201 (A201) of helix 8 through a hydrogen bonded A-A base pair.A similar interaction is observed between the equiv-alent positions (A165 and A212) of M. jannaschii SRP RNA (MjSR) in the complex with Mj19 (Hainzl et al. 2002).Based on these structures, SRP19 has been postulated to act as a "clamp" (Oubridge et al. 2002) that promotes the formation of an A-A pair between the tetraloops of SRP RNA helices 6 and 8.In the ternary complex composed of the large domain of hSR, SRP19 and SRP54 (Kuglstatter et al. 2002), the asymmetric loop of helix 8 has collapsed to form a continuously stacked helix 8, and two A-minor motifs (Nissen et al. 2001) consolidate the interaction between helices 6 and 8.
To investigate the assembly mechanism suggested by the high-resolution structures, we compared the function of the conserved adenosine residue of the helix 6 tetraloop in A. fulgidus SRP RNA (AfSR) and hSR in solution.Wild-type and mutant SRP RNA fragments with single-nucleotide changes of the adenosine residue at position 159 (A159) in AfSR or A149 in hSR were introduced into RNA molecules containing helix 6 only or both helices 6 and 8.The ability of the various RNA fragments to bind recombinant A. fulgidus and human SRP19 proteins was determined, and the susceptibility of the RNAs to digestion was measured.In contrast to what appears to occur during the assembly of the mammalian SRP, the data suggest that the conserved A159 of AfSR allows the formation of a more compact RNA molecule without the participation of protein SRP19.

In vitro synthesis of A. fulgidus and human SRP RNA fragments
Deoxyribonucleic acid fragments were assembled de novo from synthetic oligonucleotides.We constructed DNA inserts corresponding to positions 141 to 181 of helix 6 of AfSR (AfRH6) according to the sequence determined by Bhuiyan et al. 2000 (GenBank accession X07237)  Equimolar amounts of the appropriate oligonucleotides were mixed and incubated in 10 mM Tris-HCl, pH 9.0, 10 mM MgCl 2 , 5 mM DTT, 40 µM ATP with 10 units of T4 polynucleotide kinase for 20 min at 37 °C.Two µl of 250 mM EDTA, pH 8, and 348 µl of a buffer containing 10 mM Tris-HCl, pH 8.0, and 100 mM NaCl were added to a final volume of 400 µl.Samples were heated for 5 min by placing the reaction tubes into a beaker with boiling water, and allowed to cool gradually to 10 °C over several hours by keeping the beaker at 4 °C.The annealed DNAs were extracted once with phenol and chloroform, and concentrated by adding three volumes of ethanol and 8 µl of 5 M NaCl to the recovered aqueous phase, followed by incubation at -70 °C for 30 min and centrifugation.The concentrations and quality of the assembled DNA fragments were determined by electrophoresis of sample aliquots on 2% agarose gels.
Using a 5-to 20-fold molar excess of insert, annealed DNAs were ligated to EcoRI and BamHI-digested pΔ35 vector DNA (Zwieb 1991).Competent E. coli DH5-α cells (Invitrogen, Carlsbad, CA) were transformed with the ligated DNA mixture, and transformants were selected on LB plates containing 100 µg ml -1 ampicillin by growth at 37 °C.Plasmid DNAs from individual colonies were screened by restriction mapping with PvuII, EcoRI and BamHI, and verified by sequencing.Plasmid DNAs were prepared by centrifugation through CsCl density gradients.
Run-off transcriptions were carried out in a volume of 20 µl for 3 h at 37 °C with components of the T7-MEGAshortscript (Ambion, Austin, TX) using BamHI-or DraI-digested plasmid DNAs from the various constructs.Transcription was terminated by adding 115 µl of water, 15 µl of 5 M ammonium acetate and two volumes of ethanol with incubation at -20 °C for 20 min.The RNAs were collected by centrifugation, washed with 80% ethanol and dissolved in water.Concentrations were determined from a standard curve obtained with known amounts of E. coli 5S ribosomal RNA (Sigma, St. Louis, MO) by electrophoresis of sample aliquots on 2% agarose gels followed by staining with ethidium bromide and densitometry.

Purification of A. fulgidus and human SRP19 proteins
Human SRP19 (h19; MW = 16,145 Da) was purified as described by Walker et al. (1995) with minor modifications.Briefly, competent E. coli BL21 (DE3) cells (Novagen, Madison, WI) were transformed with expression plasmid pET23d-19X, grown overnight and used to seed 20 l of medium in a fermenter (Bioflo IV, New Brunswick Scientific, Edison, NJ).
Protein expression was induced by the addition of 1 mM IPTG (Gold Biotechnologies, St. Louis, MO) and continuous incubation for 2 h at 37 °C.Cells were harvested by centrifugation and lysed by passing through a French press.The lysate was centrifuged and loaded onto a Biorex 70 (BioRad, Hercules, CA) cation exchange column.The SRP19 containing fractions that eluted at about 300 mM salt were concentrated with an Amicon Centricon 10 device and purified on a Superdex 75 (Amersham Biosciences, Piscataway, NJ) gel filtration column equilibrated in 25 mM Tris-Cl, pH 7.5, 150 mM NaCl, 5 mM DTT. Archaeoglobus fulgidus SRP19 (Af19; MW = 12,396 Da) was prepared as described previously (Bhuiyan et al. 2000).Protein concentrations were measured by SDS-PAGE of sample aliquots on 15% gels by co-electrophoresis with known amounts of lysozyme (Sigma) followed by staining with Coomassie blue and densitometry.

Protein binding activities of SRP RNA fragments
We added 100 pmol of purified Af19 or h19 protein at room temperature to 50 µl reactions containing up to 1.27 nmol of wild-type or mutant RNA in binding buffer (50 mM Tris-HCl, pH 7.9, 300 mM KOAc, 5 mM MgCl 2 , 1 mM DTT).Samples were incubated at 37 °C for 10 min and loaded onto 40-µl-bedvolume DEAE-Sepharose Fast Flow (Pfizer, New York, NY) columns prepared in aerosol-barrier tips (Continental Laboratory Products, San Diego, CA) and equilibrated in binding buffer.Samples were collected and combined with a 150-µl wash of the column using the binding buffer to generate the flowthrough fraction (F).Bound protein and RNA were eluted (E) with 200 µl of binding buffer but containing 1 M KOAc.Subsequently, 1 µg of BSA and TCA were added to each sample to a final concentration of 16%, followed by a 30-min incubation on ice and a 10-min centrifugation.The proteins in the pellets were analyzed by SDS-PAGE on 15% gels followed by staining with Coomassie blue.The dried gels were scanned using a G-710 densitometer and the number of pixels in each band was measured using Quantity One software (BioRad).

RNase susceptibility assay
One µg of wild-type and mutant AfSR fragments were incubated in 20 µl of water containing 0.5 pg of RNase A (Sigma) at 37 °C.Two-µl aliquots were removed after 10, 20, 30 and 60 min, and analyzed at room temperature by electrophoresis on 2% agarose gels.The RNA fragments were stained with ethidium bromide, pictures were taken with a UV transilluminator, and the number of pixels in each band was measured using the ImageJ program available at http://rsb.info.nih.gov/ij/.

Synthesis of wild-type and mutant SRP RNAs
Secondary structures of the generated SRP RNA fragments, as predicted by comparative sequence analysis (Larsen and Zwieb 1991), are shown in Figure 1A.For the large domain of hSR, the base pairings shown have been confirmed by X-ray crystallography (Wild et al. 1999, 2001, Kuglstatter et al. 2002, Oubridge et al. 2002).The primary and secondary structures of helix 8 are known to be exceptionally well conserved (Larsen andZwieb 1991, Rosenblad et al. 2003).In contrast, helix 6 is more variable in sequence and extensively base paired.
A tetraloop with the sequence GGAG is present in helix 6 of A. fulgidus and human SRP RNA.In this study, the conserved adenosine residue at the third position of this loop (shown circled in Figure 1A) was targeted by site-directed mutagenesis.To investigate the function of A159 in AfSR and of A149 in hSR, 12 mutant RNA derivatives were constructed containing point mutations from A to C, G or U in the context of helix 6 and the larger AfRH68 and hRH678 RNAs.A small number of residues at the termini of the fragments deviated from the native sequences and were introduced to achieve optimal run-off transcription using T7 RNA polymerase.As expected, these terminal alterations had no influence on the ability of the RNAs to interact with the SRP19 proteins (data not shown).

Structural comparison of human and A. fulgidus SRP19
Human and A. fulgidus SRP19 proteins used in this study were purified to homogeneity from overexpressing E. coli cells.A comparison of the known crystal structures of RNA-bound h19 protein (Wild et al. 2001, Kuglstatter et al. 2002) with the solution structure of Af19 (Pakhomova et al. 2002) revealed no major differences in the overall fold.However, as indicated in the alignment of the two sequences shown in Figure 1B, Af19 lacks several amino acid residues in loop 4 and near the positively charged C-terminus.Loop 3 evidently undergoes a binding-induced structural transition as it is found to be wellordered and in contact with SRP RNA helix 8 in the crystal structures (Kuglstatter et al. 2002), but disordered in solution (Pakhomova et al. 2002).

Formation of SRP RNA-protein complexes with human and A. fulgidus SRP19
Using SRP reconstitution conditions (Walter and Blobel 1983), purified Af19 or h19 was mixed with the wild-type and mutant SRP RNA fragments to assess their ability to form complexes.In initial experiments, complex formation was driven by incubating the protein with a 3-to 6-fold molar excess of RNA.The RNA binding activities of Af19 and h19 were determined by comparing the amounts of the polypeptides present in the flowthrough (F) and eluate (E) fractions after chromatography of complexes on small DEAE columns.
The Af19 bound successfully to the wild-type and mutant hRH678, AfRH68 and AfRH6 SRP RNA fragments (Figure 2A, left panel).Reduced binding activities of 25 to 40% were observed in heterologous complexes of Af19 with the wild-type and mutant forms of hRH6.Hence, at the relatively high RNA concentrations used in the preliminary experiments, helix 8 and the conserved adenosine in the helix 6 tetraloop are not required for Af19 binding, although they may promote binding.
Human SRP19 (h19) bound wild-type hRH678 and AfRH68 RNAs, but showed less pronounced affinity or none at all to wild-type hRH6 and the AfRH6 RNAs, clearly dem-onstrating that helix 8 is essential to establishing high-affinity binding of h19 (Figure 2A, right panel).The binding activities of the mutant derivatives of hRH678 and AfRH68 relative to the wild-type constructs show that the conserved adenosine (A.fulgidus A159 or human A149) is critical for high affinity binding of h19.The adenosine is presumably required to maintain the tertiary RNA-RNA interaction with helix 8 (Hainzl et al. 2002, Oubridge et al. 2002), and this tertiary interaction in turn is required for the high-affinity binding of SRP19.These results explain the conserved nature of the adenosine, and agree well with the demonstrated importance of both helix 8 and A149 for the binding of h19 in the context of full-length hSR (Zwieb 1992).
The binding activities of Af19 to the A. fulgidus and human SRP RNAs were investigated in greater detail by varying the RNA-protein molar ratio from 0.3 to 10. Owing to the limited RNA-binding capacity of DEAE Sepharose, the two larger hRH678 and AfRH68 fragments were used only up to a 5-fold molar excess.The results presented in Figure 2B reveal that about 50% of Af19 binds at a 3-fold molar excess of wild-type hRH6 RNA, and 75% of Af19 binds at a 10-fold molar ratio.The three mutant derivatives of the hRH6 RNA show only up to 30% of Af19 bound at the highest RNA-protein ratio tested.This degree of binding is comparable to the nonspecific binding observed between Af19 and yeast tRNA (not shown).The binding of Af19 to wild-type hRH6 at a level of 70% may be a result of the formation of ternary ribonucleoprotein complexes containing a second RNA molecule that mimics helix 8, as observed in the crystal structure of the complex between h19 and SRP RNA helix 6 (Wild et al. 2001).Consistent with such an arrangement, the wild-type RNA binds at detectable levels, whereas binding by the three mutant RNAs appeared to be nonspecific.
At a 3-fold molar excess of RNA, Af19 was completely bound to AfRH6.The mutant derivatives exhibited no significant differences relative to wild type (Figure 2B), confirming that A159 is not required for AfSR to bind Af19.Although AfRH6 and hRH6 possess the same GGAG tetraloop, Af19 binds significantly better to the AfRH6 RNA.A change of the closing base pair of the tetraloop from G146 -C151 in hRH6 to C156-G161 in AfRH6 (see Figure 1) as well as some diversity in the adjacent helical region might introduce subtle structural differences underlying the higher affinity of Af19 to its own RNA.Similarities in the binding affinities among wild-type and mutant AfRH6 RNAs and the demonstrated long-term stability of the Af19-AfRH6 complex in sucrose gradients Arrows indicate residues located within 2.7 Å of hSR helix 6 or helix 8 in the crystal structures of the ternary complex.(Bhuiyan et al. 2001) suggest that the binding of Af19 to AfRH6 does not require the formation of a ternary complex with another AfRH6 molecule.The Af19 bound with significantly higher affinity to hRH678 than to hRH6 RNA.Complete binding to hRH678 was achieved at a 3-fold RNA-protein ratio, indicating that helix 8 contributes favorably to the overall stability of the complex.The interaction responsible for the enhanced binding of Af19 to hRH678 RNA presumably involves loop 3 of Af19 and the tetraloop of helix 8, similar to that observed in crystal structures of the RNA-bound forms of the h19 and Mj19 protein (Wild et al. 2001, Hainzl et al. 2002, Oubridge et al. 2002).The finding that Af19 binds equally well to the human wild-type and three mutant hRH678 RNAs confirms that the protein does not directly interact with A149 as shown by the structural analyses.However, Af19 is able to simultaneously bind helices 6 and 8 even when the conserved adenosine is mutated.
No apparent differences in the affinities are observed among the wild-type and the C, G and U point mutations.Because h19 does not bind to hRH6 or AfRH6 (Figure 2A, right panel), it appears that the mammalian and archaeal SRP19 proteins differ in the way they interact with helix 6.As shown previously, h19 contains an extended irregularly structured region of approximately 25 residues, designated as loop 1, which fills the major groove of SRP RNA helix 6, and is the principal surface through which SRP19 binds to the RNA.Loop 1 of the free Af19 in solution exhibits remarkable structural similarity to that reported for the RNA-bound form of h19, suggesting that this region undergoes only minor structural rearrangements upon RNA binding (Pakhomova et al. 2002).We therefore propose that loop 1 of h19 is either completely unstructured or weakly structured, and thus unable to bind helix 6 in the absence of additional stabilizing interactions such as those between loop 3 and SRP RNA helix 8.
Complete binding occurs between Af19 and the AfRH68 RNA and its mutant variants at equimolar concentrations of protein and RNA.The high affinity of Af19 to its own SRP RNA is likely due to the opportunity to engage SRP RNA helix 8, combined with the availability of a proper helix 6 that accommodates the already properly folded Af19.

RNase susceptibility of wild-type and mutant A. fulgidus SRP RNA fragments
The finding that neither A159 nor helix 8 was required for binding of the Af19 protein does not satisfactorily explain why the adenosine is conserved among the SRP RNAs of the Archaea.Instead, the results suggest that A159 has a proteinindependent role.To investigate this possibility, AfRH68 and the corresponding mutant RNAs were incubated with RNase A under mild digestion conditions, and the RNase susceptibility was measured by gel electrophoresis.The data demonstrate that, after a 60-min incubation with RNase A, more than 70% of the wild-type AfRH68 RNA remains intact, whereas all three mutant RNAs are predominantly fragmented (Figure 3).Therefore, A159 is both necessary and sufficient for the formation of a compact RNase-resistant molecule, which according to the crystal structure is characterized by an interaction between helices 6 and 8 in a side-by-side fashion (Oubridge et al. 2002).Noticeably, this RNA-RNA interaction occurs independently of added Af19 protein.No significant RNase susceptibility differences were observed in human SRP RNA between the wild-type and mutant hRH678 fragments.Furthermore, the susceptibility of the four human SRP RNA fragments was found to be similar to that of the three mutant AfRH68 fragments (not shown).

Conclusions
The data presented here are consistent with the assembly mechanism proposed previously by the high-resolution structures.Three residues, corresponding to A159, G202 and A205 of AfSR, are conserved among the known eukaryotic and archaeal SRP RNAs (Rosenblad et al. 2003), suggesting that the triplet structure formed by these residues is shared between these two domains of life.It should be noted that even in certain plant SRP RNAs that lack the helix 8 tetraloop (Campos et al. 1989, Marshallsay et al. 1989, Riedel et al. 1995), a similar co-planar arrangement of the AAG-triple can easily be achieved in molecular modeling experiments (not shown).
The conserved A159 plays a crucial role in the assembly of the A. fulgidus SRP by allowing the formation of a compact RNA molecule and the high-affinity binding of SRP19 to the tetraloops of helices 6 and 8.This RNA-RNA interaction between the tetraloops occurs independently of added Af19 protein, suggesting that the initial step in the assembly of an archaeal SRP relies exclusively on the RNA moiety.The importance of RNA for this early step of archaeal SRP assembly is corroborated by the apparent impairment in SRP19 function in certain archaea such as Thermoplasma acidophilum (Ruepp et al. 2000, Zwieb andEichler 2001).In addition, the data suggest that, even in the mammalian system, RNA contributes significantly to the folding of SRP19 and to SRP assembly.

Figure 1 .
Figure 1.Structural features of the SRP components used in this study.(A)Secondary structures of A. fulgidus SRP RNA helix 6 (AfRH6) and a fragment comprising helices 6 and 8 (AfR-H68).The corresponding regions from the human SRP RNA (hSR) are represented by hRH6 and hRH678.Nucleotides are numbered in increments of ten according to the full-length SRP RNA molecules.Residues deviating from the native sequences are shown in gray.The conserved adenosines targeted by site-directed mutagenesis are circled.Arrowheads in hRH678 mark nucleotides that are within 2.7 Å of SRP19 in the crystal structure of the ternary complex (coordinates are from 1MFQ.pdb).The red dashed line indicates the tertiary interaction between A149 and A201.(B) Sequence alignment of A. fulgidus SRP19 (Af19) and human SRP19 (h19).Residues are numbered in increments of ten.Conserved residues are shown in red.In Af19, the α-helices, β-strands and loops are shown as in the solution structure of the free protein.Secondary structural features of the h19 are derived from 1MFQ.pdb and assigned as in the Protein Data Bank(Bourne et al. 2004).Arrows indicate residues located within 2.7 Å of hSR helix 6 or helix 8 in the crystal structures of the ternary complex.

Figure 2 .
Figure 2. Protein SRP19 binding activities of SRP RNA fragments.(A) SDS-PAGE of A. fulgidus SRP19 (Af19) (left panel) or human SRP19 (h19) (right panel) in the flowthrough (F) and high-salt eluate (E) of DEAE-columns.The names of the various RNA fragments used in the binding reactions are shown on the left of each panel, whereas the identity of the residue at the third position of the helix 6 tetraloop (A, C, G or U) is indicated at the bottom.(B) Quantitative analysis of Af19 binding activities.Binding to wildtype and mutant SRP RNA fragments is indicated by lines connected to symbols, where: ᮀ = wildtype RNA; ᭺ = A159C mutant; = A159G mutant; and ᭹ = A159U mutant.Dashed vertical lines correspond to the conditions used for the results shown in the left panel of Figure 2A.

Figure 3 .
Figure 3.The RNase susceptibility of wild-type and mutant A. fulgidus SRP RNA helices 6 and 8 (AfRH68) RNAs.(A) The nature of the nucleotide residue (A, C, G or U) at position 159 of A. fulgidus SRP RNA (AfSR) is indicated on top.Numbers on the left show the incubation time in minutes.Filled and open arrowheads indicate the undigested and RNase-cleaved AfRH68 products, respectively.(B) Quantitative analysis of the RNase susceptibility of wild-type AfRH68 (ᮀ), A159C mutant RNA (᭺), A159G mutant RNA () and A159U mutant RNA (᭹).