|Home | About | Journals | Submit | Contact Us | Français|
Maturation of tRNA precursors into functional tRNA molecules requires trimming of the primary transcript at both the 5′ and 3′ ends. Cleavage of nucleotides from the 3′ stem of tRNA precursors, releasing nucleotide diphosphates, is accomplished in Bacillus by a phosphate-dependent exoribonuclease, Rph. The crystal structure of this enzyme from B. anthracis has been solved by molecular replacement to a resolution of 1.7 Å and refined to an R factor of 19.3%. There is one molecule in the asymmetric unit; the crystal packing reveals the assembly of the protein into a hexamer arranged as a trimer of dimers. The structure shows two sulfate ions bound in the active-site pocket, probably mimicking the phosphate substrate and the phosphate of the 3′-terminal nucleotide of the tRNA precursor. Three other bound sulfate ions point to likely RNA-binding sites.
Transfer RNA molecules are transcribed as precursors with additional nucleotides at both their 5′ and 3′ ends. These flanking sequences must be cleaved off to reveal the mature functional tRNA species. In Bacillus, as in most organisms, endonucleolytic cleavage of the 5′ extension is performed by RNase P, a heterotetrameric ribonucleoprotein complex (Stams et al., 1998 ; Fang et al., 2001 ). However, the manner in which the 3′ nucleotides are removed varies from species to species. In Escherichia coli, the most studied system, the 3′ flanking sequence is first trimmed by an endoribonuclease such as RNase E and this is followed by exoribonucleolytic degradation by one or more of a number of enzymes including RNase II, RNase BN, RNase T and RNase PH. In Bacillus just two enzymes are known to process the 3′ termini: an endoribonuclease called RNase Z and the exoribonuclease RNase PH (Pellegrini et al., 2003 ; Wen et al., 2005 ).
RNase PH is a member of the phosphate-dependent 3′→5′ exoribonuclease (PDX) family (Zuo & Deutscher, 2001 ; Mian, 1997 ). This family includes polynucleotide phosphorylase (PNPase) and a shared characteristic is the utilization of inorganic orthophosphate in the cleavage reactions (Deutscher et al., 1988 ) to generate nucleoside diphosphates rather than the more common nucleoside monophosphates formed by the hydrolytic ribonucleases.
B. anthracis, the causative agent of anthrax, is a soil-dwelling Gram-positive endospore-forming bacterium. Its genome has been sequenced (Read et al., 2003 ), revealing the presence of an open reading frame that shares a high degree of identity to the well characterized RNases PH from B. subtilis, E. coli and other bacteria. Here, we report the determination of the structure of Rph from B. anthracis Ames, a target in a high-throughput structural genomics study.
The rph coding sequence was amplified by PCR from B. anthracis Ames genomic DNA using BA4715F (5′-CACCACCACCACATGCGAGTAGATGGTAGAGAGAAAA-3′) as a forward primer and BA4715R (5′-GAGGAGAAGGCGCGTTACTACTCTATATGAGATACGATGTCACCTAAC-3′) as a reverse primer and cloned using a ligation-independent cloning method (Alzari et al., 2006 ; Au et al., 2006 ; Fogg & Wilkinson, 2008 ). The resulting fragment was treated with T4 DNA polymerase in the presence of dATP to generate single-stranded DNA overhangs complementary to those present on a modified pET28a vector cut with BseRI and then similarly treated with T4 DNA polymerase in the presence of dTTP. The vector and PCR products were incubated together for 20 min to allow annealing of the single-stranded overhang regions (Aslanidis & de Jong, 1990 ) and the mixture was directly added to E. coli Novablue competent cells (Novagen), which were plated and grown overnight on antibiotic-containing media. Kanamycin-resistant colonies were picked the following day and used to inoculate overnight cultures, from which recombinant plasmids were purified. The resulting plasmid pETBaRph contains the coding sequence of rph fused to the coding sequence for an N-terminal MGSSHHHHHH tag. pETBaRph was introduced into E. coli BL21 (DE3) for overproduction of recombinant Rph.
N-terminally hexahistidine-tagged Rph was expressed from E. coli BL21 (DE3)/pETBaRph by overnight growth in kanamycin-supplemented autoinduction media (Studier, 2005 ) at 310 K. Cells from 0.5 l culture were harvested by centrifugation at 5000 rev min−1 for 15 min. Pelletted cells were resuspended in buffer A (20 mM Na2HPO4, 0.5 M NaCl, 10 mM imidazole pH 7.5) and lysed by sonication. Further centrifugation at 15 000 rev min−1 generated a cleared lysate which was loaded onto a 5 ml HiTrap nickel-chelation column, from which Rph was eluted following the application of a linear gradient of 10 to 500 mM imidazole in buffer A. Peak fractions were automatically directed onto a HiLoad 16/60 Superdex 200 prep-grade gel-filtration column (GE Healthcare) previously equilibrated with buffer B (50 mM Tris–HCl pH 7.5, 150 mM NaCl). The protein eluted with high purity and by comparison with the elution profiles of standard proteins the oligomerization state of Rph was determined to be hexameric. The 37.6 mg of purified protein obtained was concentrated to 20 mg ml−1 in buffer B, aliquoted and stored at 193 K.
A Mosquito nanolitre pipetting robot was used to screen Rph against different crystallization conditions using sitting-drop vapour diffusion in a 96-well plate format. 150 nl drops were equilibrated against 80 µl reservoir solution at 293 K. Small crystals were obtained in the Clear Strategy Screen (Brzozowski & Walton, 2001 ) with a reservoir solution comprising 2.7 M ammonium sulfate and 0.1 M bis-tris pH 5.5. These conditions were altered systematically, including a change to hanging-drop vapour diffusion in 24-well plates in order to optimize the crystal quality. Large single crystals appeared in 2 µl drops made up from 1 µl 2 M ammonium sulfate and 0.2 M potassium/sodium tartrate and 1 µl protein solution at a concentration of 20 mg ml−1.
A single crystal was mounted in a cryo-loop (Hampton) and flash-cooled in liquid nitrogen following brief immersion in a cryoprotectant composed of reservoir solution supplemented with 30%(v/v) glycerol.
Diffraction data were collected on beamline ID14.3 at the European Synchrotron Radiation Facility (ESRF). The crystal was maintained at a constant 100 K throughout the experiment and was exposed to radiation of wavelength 0.931 Å. Images were collected on an ADS Q4R detector. Indexing of the diffraction patterns with DENZO from the HKL-2000 program suite (Otwinowski & Minor, 1997 ) showed that the crystal belonged to space group R32, with unit-cell parameters a = b = 86.87, c = 179.31 Å. The diffraction images revealed that the low-resolution spots were surrounded by regions of diffuse scattering arising from a semi-ordered water–ice structure. Integration and scaling of the data were performed using XDS (Kabsch, 1988 ) and SCALA, respectively (Collaborative Computational Project, Number 4, 1994 ). The data statistics are presented in Table 1 .
MOLREP (Vagin & Teplyakov, 1997 ; Murshudov et al., 1997 ) was used to solve the structure by molecular replacement using a single subunit of the RNase PH structure from B. subtilis (which shares 69% sequence identity with Rph; PDB code 1oyr; Harlow et al., 2004 ) as the search model. After performing rotation and translation functions, a clear solution was found with an initial R factor of 49.1%; the next best solution had an R factor of 63.4%. The solution shows a single molecule in the asymmetric unit, indicating a solvent content of 48.8%. Rigid-body and TLS refinement were performed using REFMAC (Murshudov et al., 1997 ). At this stage, changes in sequence between the B. subtilis Rph search model and the B. anthracis Rph structure were apparent in the electron-density maps (Fig. 1 ). Structure refinement continued with successive rounds of model building in Coot (Emsley & Cowtan, 2004 ) followed by REFMAC refinement. The final R factor for the structure consisting of 2090 atoms and 168 water molecules was 19.26% (R free = 24.07%). Protein-structure and ligand analysis was performed using the PDBSUM server (Laskowski, 2001 ).
The structure of Rph from B. anthracis was solved to 1.7 Å resolution with a single protein molecule in the asymmetric unit. The protein chain assumes the βαβα topology (Fig. 2 a), as observed in other RNase PH structures (Harlow et al., 2004 ), consisting of nine β-strands and five α-helices. It can be thought of as a double sandwich with strands 1–5 and 6–9 surrounding helices 1–3 and with helices 1–3 and 4–5 surrounding strands 6–9. Analysis of the molecular packing revealed Rph to be a hexamer (Fig. 2 b) arranged as a trimer of dimers as indicated by the crystal symmetry. The β9 strands from a pair of subunits come together so that an eight-stranded intersubunit β-sheet is formed featuring strands 6–9 from the two adjacent subunits. The dimer has a buried surface area of 1523 Å2, 10% of which is accounted for by Arg212 present on strand β9, which also makes four hydrogen bonds across the interface. The trimer is formed predominantly through contacts between the loops linking strands 1–5, with 1465 Å2 buried at this surface. His23, Arg68, Arg76, Arg73, Asp115, Asp117 and Gln120, all of which are conserved residues, collectively contribute over 40% of this surface area and 70% of the hydrogen-bond interactions.
The loop region between strand β5 and helix α2, which is found at the centre of the hexamer, was poorly defined in the electron-density maps; as a result, the lysine at position 80 was omitted from the refined model. It has been suggested that this loop, with its arginine-rich character, is prone to proteolytic degradation and that the hexamer may be a storage mechanism to protect the loops (Harlow et al., 2004 ).
Rph from B. anthracis shares a high degree of sequence similarity to its orthologues from B. subtilis and E. coli, with 69% and 56% identity, respectively (Fig. 3 ). The five sulfate ions bound to B. anthracis Rph (Table 2 , Figs. 2 a and 4 ) make contacts with residues that are conserved, possibly indicating a role as tRNA phosphate mimics; two of them are located in the active-site pocket (Fig. 5 ). The first sulfate (Sul1), a likely mimic of the phosphate substrate, makes contacts with the highly conserved residues Arg86, Gly124, Thr125 and Arg126 (situated on helices 1 and 3), which have been shown to play important roles in the phosphorolytic reaction. This sulfate is located in a position equivalent to that of the bound tungstate in PNPase from Streptomyces antibioticus (Symmons et al., 2000 ), the phosphate bound in the structure of Rph from Aquifex aeolicus (Ishii et al., 2003 ) and the sulfate found in the active site of RNase PH from B. subtilis (Harlow et al., 2004 ). The second sulfate ion (Sul2) is bound to residues Arg86 and Arg126, as well as Arg92 from the neighbouring dimer-forming subunit. It is probable that this sulfate mimics the phosphate of the 3′-terminal phosphodiester bond found on the tRNA substrate. The presence of Sul2 in the active site may help to order the Arg86 side chain, which was poorly defined in the B. subtilis structure that featured just a single sulfate ion. The S atoms of Sul1 and Sul2 are 6.3 Å apart (Fig. 5 ). It appears that Arg86 and Arg126, which make contacts with both sulfates, ensure that the substrate tRNA and the phosphate nucleophile assume the appropriate juxtaposition within the active site. Attack of the nucleophilic oxygen of the phosphate at the phosphorus centre of the 3′-terminal nucleotide, thought to progress by an SN2-type mechanism, would result in the transient formation of a pentacoordinate phosphorus species (Mildvan, 1997 ). The conserved arginine residues in the active site may help to stabilize this negatively charged transition state.
The third sulfate ion (Sul3), which is in contact with Arg99, has also been observed in the structures of RNase PH from B. subtilis and A. aeolicus, where additional interactions with Trp58 and Thr60 are made (Harlow et al., 2004 ; Ishii et al., 2003 ). It is possible that the dimerization of two Rph subunits allows the insertion of the 3′ terminus of the tRNA precursor into the active-site pocket of molecule 1, with phosphates present elsewhere in the substrate interacting with residues 99, 58 and 60 of molecule 2. Sulfate 4 is hydrogen bonded to Arg73, a conserved residue from the motif RX 4RX 2R beginning at residue 68 and located at the trimer interface. A fifth sulfate (Sul5) forms hydrogen bonds to Lys156, Leu104, Val102 and Lys56 and is also observed in the structure of RNase PH from A. aeolicus (Ishii et al., 2003 ).
The conserved N-terminal motif RX 3RX 5R beginning at Arg2 (Zuo & Deutscher, 2001 ), which in this structure is associated with the binding of a fifth sulfate (Sul5) at Arg6, is unusually followed by three regularly spaced histidines at positions 13, 15 and 17 (Fig. 1 . If the RX 3RX 5R sequence is involved in interactions with the RNA, as has been proposed for the corresponding motif in PNPase (Symmons et al., 2002 ), then the histidines may form a base-stacking interaction with RNA bases, although these histidine residues are not strongly conserved in other members of the PDX family.
In conclusion, among the RNase PH structures, the structure of Rph from B. anthracis is distinct in exhibiting two sulfate ligands within the active site, possibly mimicking the 3′-terminal phosphate of both the tRNA substrate and the phosphate nucleophile. Other bound sulfates indicate likely tRNA-binding sites. The conserved motif RX 5RX 3RX 2R beginning at Arg86 and found in helix α1 makes contacts with Sul1, Sul2 and Sul3 as well as Sul2 of the adjacent subunit through interaction with residue Arg92. This suggests that the dimer may be required for catalysis.
The work described here was funded by the European Commission as SPINE contract No. QLG2-CT-2002-00988 under the RTD programme ‘Quality of Life and Management of Living Resources’. AER was funded by the BBSRC and VL by the Wellcome Trust. The authors would like to thank Sam Hart and Dr Tracey Gloster of the York Structural Biology Laboratory for their help with data collection at the ESRF and Dr Garib Murshudov for useful advice and assistance with model refinement.