Crystal structure of RecX: A potent regulatory protein of RecA fromXanthomonas campestris

Abstract

Bacterial RecA is a multifunctional protein,1-3 participating in a variety of activities such as ATP hydrolysis (as an ATPase), LexA repressor cleavage (as a coprotease), homologous DNA strand exchange (as a recombinase), and SOS responses when assembled on ssDNA and forms a helical nucleoprotein filament. Although homologous recombination plays important roles in maintaining structural and functional integrity of a genome, proteins associated with this activity need to be tightly regulated to prevent unnecessary DNA exchanges that may be lethal.4 Recently, a plethora of proteins exhibiting regulatory function against RecA activities, including RecBCD,5 RecF,6 RecO,7, 8 RecR,9, 10 DinI,11, 12 RdgC,13 and UvrD proteins, and so forth, have been discovered. But regulatory proteins affecting the RecA functions are not completely understood yet, and new members are being added to this regulatory network steadily.4 RecX is a recent addition to this RecA regulatory network. It is first described in 1993 in Pseudomonas aeruginosa as an open reading frame (ORF) located downstream of recA.14 Since then, many recX homologs have been detected in other microorganisms, including Escherichia coli,15-18 Deinococcus radiodurans,19, 20 Neisseria gonorrhoeae,21 Mycobacterium tuberculosis,22 Streptomyces lividans,23 Herbaspirillum seropedicae,24 Thermus thermophilus,25 Thiobacillus ferrooxidans,26 Xanthomonas campestris,27, 28 and X. oryzae.29 Sequence alignments between these RecX proteins indicated that they are moderately conserved with sequence identity values of roughly 30%, having a molecular weight of ∼20 kDa and an alkaline pI ranging from 8.8 to 10.7.30 RecX has been shown to serve different roles in different organisms. For instance, although it is shown to be a potent inhibitor protein against the various RecA functions,15, 16, 31 it is found to be required for UV resistance in E. coli and for recombination in N. gonorrhoeae.21 Overexpression of RecA is also lethal in the recX deletion mutants of P. aeruginosa,14 X. oryzae,27 and S. lividans.23 Furthermore, RecX is also found to repress the induction of antioxidant enzymes in D. readiodurans,19 suggesting that RecX can act as a repressor in regulating the expression of certain genes. These results indicate that RecX protein can be considered as a multifunctional protein but needs more studies to elucidate its various roles. To date, the structural information of RecX is still limited, although RecX has been shown to function as a potent inhibitor against the various RecA activities, such as the ATP hydrolysis,15, 31 LexA repressor cleavage,16 homologous DNA strand exchange,15, 16, 31 and RecA filament capping32 in vitro, among the others. To better understand the structure and function of RecX from X. campestris, we have, in this manuscript, determined its crystal structure to a high resolution of 1.5 Å. The final structure reveals that XcRecX adopts a novel tandem repeats of three-helix bundle. Preliminary docking study of XcRecX with XcRecA indicates that XcRecX can fit into the groove of the XcRecA filament, implying possible inhibition of XcRecX against the various activities of the XcRecA filament. The recX gene was PCR amplified directly from the plant pathogen X. campestris pv. campestris str. 17 (Xcc) using a forward primer 5′-TACTTCCAATCCAATGCTAT GAGTGAGCAAGCGCCCGCACC and a backward primer 5′-TTATCCACTTCCAATGTCAGTCCTCAAGGTCGAAGC GTGTTGCCA, for forming the fragment of required length. The PCR fragment has correct size in a SDS-PAGE experiment and was confirmed by DNA sequencing. A ligation-independent cloning (LIC) approach33, 34 was used to obtain the desired constructs. The final construct codes for a N-terminal His6 tag, a 17 amino acid linker, and the XcRecX target under the control of a T7 promoter. Overexpression of the Hig6-tag target protein was induced by the addition of 0.5 mM IPTG at 293 K for 20 h. The target protein was purified by immobilized metal affinity chromatography (IMAC) on a nickel column (Sigma). The His6-tag and linker was cleaved from XcRecX by tobacco etch virus (TEV) protease at 277 K for 16 h. For crystallization, XcRecX protein was further purified on a Superdex 200 column (AKTA, Pharmacia). The final fresh target protein exhibits purity greater than 99% and contains only an extra tripeptide (SNA) at the N-terminal end. Se-Met-labeled XcRecX was prepared in a similar way and was produced using an E. coli strain BL21 (DE3) as the host in the absence of methionine but with ample amounts of Se-met (100 mg/L). The M9 medium consists of 1 g of NH4Cl, 3 g of KH2PO4, and 6 g of Na2HPO4 supplemented with 20% (W/V) of glucose, 0.3% (W/V) of MgSO4, and 10 mg of FeSO4 in 1 L of double-distilled water. The induction was conducted at 293 K for 24 h by the addition of 0.5 mM IPTG. Purification of the Se-Met-labeled XcRecX protein was performed using the protocols as established for the native proteins. For crystallization, the native protein was concentrated to 8 mg/mL in 40 mM Tris-HCl, 500 mM NaCl using an Amicon Ultra-10 (Millipore). Screening for crystallization condition was performed by using a sitting-drop vapor diffusion method in 96-well plates (Hampton Research) at 277° K by mixing 0.5 μL protein solution with 0.5 μL reagent solution. Initial screens including the Hampton Clear Strategy Screen 1, the Structure Screens 1 and 2, a systematic PEG-pH screen, and a PEG/Ion screen were performed using the Gilson C240 crystallization workstation. Needle-like crystals appeared in 1 week from a reservoir solution comprising 0.1M Tris (pH 8.5), 0.3M Na(OAc), 15% PEG4000. Crystals suitable for diffraction experiments were grown by mixing 1.5 μL protein solution with 1.5 μL reagent solution at 277° K and reached dimensions of 0.1 mm × 0.1 mm × 0.4 mm after 1 week. Se-Met-labeled XcRecX was crystallized in the same way. Crystal was flash-cooled at 100 K under a stream of cold nitrogen. X-ray diffraction data was collected using the National Synchrotron Radiation Research Center (NSRRC) beamline 13B1 in Taiwan. A two-wavelength MAD data setup to 1.5 Å resolution were obtained. The data were indexed and integrated using the HKL2000 processing software,35 giving a data set that is ∼99% complete with overall Rmerge of 5.0–5.6% on intensities. The refinement of selenium atom positions, phase calculation, and density modification were performed using the program SOLVE/RESOLVE.36 The model was manually adjusted using the XtalView/Xfit package. CNS37 was then used for refinement to a final Rcryst of 18.7% and Rfree of 22.2%, respectively. The crystals belong to the P43 space group. The data collection and refinement statistics are summarized in Table I. The coordinates and structural factors of the XcRecX monomer have been deposited in the Protein Data Bank (accession number 3DDFG). Xc2845 was annotated as a RecX protein of the Pfam02631 family with an E value of 8e−09. It is located in a lexA-recA-recX gene cluster that contains a discretepromoter for each gene.27 Its direct interaction with XcRecA was confirmed by an immunoprecipitation experiment in vitro (using anti-XcRecX as a bait to pull down the XcRecX and XcRecA proteins, which was confirmed in a PAGE experiment by the anti-XcRecX and anti-XcRecA antibodies, respectively) and by a yeast two hybrid assay system in vivo (data not shown).16 Furthermore, XcRecX was also found to inhibit the XcRecA-mediated ATPase and coprotease activities (data not shown).15, 16 Having demonstrated that XcRecA exhibits ATPase and coprotease activities and that XcRecX inhibits XcRecA-mediated enzyme activities, we further determined the XcRecX tertiary structure to get a more thorough understanding of its inhibition mechanisms. After extensive tries, we are finally able to obtain needle-like crystals of XcRecX when grown in 0.3M sodium acetate, 15% PEG4000, 0.1M Tris buffer (pH 8.5) by a sitting-drop vapor diffusion method (data not shown). The phases and initial structure were determined by the multiple anomalous dispersion (MAD) approach to a resolution of 2.0 Å using a single crystal of Se-Met-substituted protein. The final structure was obtained by repeated refinement to a resolution of 1.5 Å, with final Rcrys and Rfree values of 18.7% and 22.2%, respectively. The XcRecX crystals belong to the space group of P43 and contain one molecule in each asymmetric unit. The data collection and refinement statistics are shown in Table I. The overall geometry of XcRecX is very good, with no backbone torsional angle deviating from the mostly favorable region in the Ramachandran plot. The final model comprises 142 amino acids, including residues from Gln17 to Phe158, and numerous well-defined water molecules, 256 in total. XcRecX forms a monomer in solution, as clearly demonstrated by the gel filtration and analytical ultracentrifugal experiments (data not shown). It contains 162 amino acids [Fig. 1(a)] and is an all-helix protein, consisting of 69.7% of α helix and 30.3% of random coil. It adopts a unique architecture comprising tandem repeats of three-helix bundle [Fig. 1(a,b)]. However, the geometry between the repeats is not uniform, with the axes between the R2-R3 repeats connected in a more or less collinear way, whereas those between the R1-R2 repeats in a more orthogonal way. The XcRecX thus adopts an “L-shaped” structure, with a short arm comprising the R1 repeat of 46 residues from amino acid Gln17 to Gly62 and a long arm comprising the R2 and R3 repeats of 47 residues and 49 residues, starting from amino acid Trp63 to Glu109 and Gly110 to Phe158, respectively [Fig. 1(a)]. The primary sequence and tertiary structure of XcRecX. (a) The primary sequence and secondary structural elements. Helices of the first helical bundle repeat are drawn above the sequence in blue columns, second repeat in green columns, and third repeat in red columns. (b) The stereo picture of the XcRecX tertiary structure drawn in ribbon. The R1 repeat was colored blue, R2 repeat colored green, and R3 repeat colored red. The axis of the R1 helical bundle forms a nearly 90° angle with the approximately collinear R2-R3 axes, causing the overall structure to form a “L-shaped” conformation. (c) The XcRecX tertiary structure drawn in electrostatic plot. Left figure shows the plot viewed from the concave side, whereas right figure viewed from the convex side. Positive charge is shown in blue and negative charge in red. Distinct bipolar nature is obvious from this figure, which indicates that the convex side is the likely place for interacting with nucleic acid molecules. Tandem repeats are usually consisted of small module, but multiple copies of these modules can be packed in unique ways to form elongated and curved structures for recognizing long nucleic acid sequences.38, 39 This strategy has indeed been incorporated by nature to form three-helix bundle repeats for recognizing various dsDNA.38-41 Similarly, XcRecX also comprises three-helix bundle repeats and contains an apparent bipolar electrostatic surface [Fig. 1(c)] as other three-helix bundles, which enables XcRecX to bind ssDNA and dsDNA (data not shown) through its positive charge-enriched surface. The multiple sequence and structural alignments of the XcRecX three-helix bundles are shown in Figure 2(a). They can superimpose very well with each other except for the loops connecting helix 1 to helix 2 [Fig. 2(b)]. Structural alignment using the MUSTANG program (http://www.cs.mu.oz.au/∼arun/mustang/)42 gives a r.m.s.d of 1.33 Å for 26 Cαs out of 40 residues between R1 and R2, and 0.75 Å for 27 Cαs out of 40 residues between R2 and R3 [Fig. 2(a)]. However, sequence identities between the repeats are not very high; that between R1 and R2 is only 17.6%, and that between R2 and R3 is only 19.2%. This phenomenon is similar to other HTH motifs, in which large sequence diversity can be tolerated within the HTH motif.43 Similar to other helix bundle proteins,44 nonpolar residues from every helix of the XcRecX bundles interdigitiate to form a hydrophobic core, as shown in a typical repeat of R1 in Figure 2(c). Comparison of three-helix bundles. (a) The multiple sequence and structural alignments of the H1, H2, and H3 helices of the R1, R2, and R3 repeats of the XcRecX three-helix bundle. The helices are drawn in columns above the sequences and shown in blue for the first repeat, green for the second repeat, and red for the third repeat, respectively. The highly conserved residues are shown in red, and conserved hydrophobic residues in the core region are marked with gray circles below the sequences. (b) Superimposition of the three repeats of XcRecX in stereo. The first repeat is drawn in blue, second repeat in green, and third in pink. The three-helix bundles in XcRecx can superimpose well with each other, except for the loops connecting helix 1 to helix 2. (c) A typical hydrophobic core of the three-helix bundle in XcRecX. Hydrophobic residues from each helix interdigitate to form a stable hydrophobic core for the three-helix bundle. (d) Superimposition in stereo of the XcRecX R1 three-helix bundle with a typical ds-DNA binding three-helix bundle of the c-Myb protein. The first, second, and third helices of the XcRecX R1 bundle are marked in H1/H2/H3 and colored blue, green, and red, whereas those of the c-Myb marked in h1/h2/h3, and colored light blue, pale green, and pink, respectively. It is clear that while the H3 and h3 helices of the two three-helix bundles superimpose extremely well, the H2/h2 and H1/h1 helices deviate to a significant extent. The helical appositions of the XcRecX three-helix bundles are very different from other three-helix bundles, such as that of the c-Myb tandem repeat (1H88).41 When the two three-helix bundles are superimposed, only one of the helices (H3 or h3) in the bundles can be superimposed well [Fig. 2(d)], whereas both H1 and H2 helices is found to exhibit significant deviations from the h1 and h3 helices, respectively. The c-Myb three-helix bundle repeat structure is also more curved than XcRecX. Thus tandem repeat modules have evolved to exhibit considerable flexibility in binding different DNA sequences.38-41 As stated earlier, the axes of the R1/R2 repeats are more linear, whereas those of the R2/R3 repeats are more orthogonal, resulting in an “L-shaped” architecture for the XcRecX tandem repeats. The angles are well defined by the strong interactions between the R1/R2 and R2/R3 repeats, as shown in Figure 3(a,b), respectively. Plenty of H-bonds, salt bridges, and hydrophobic interactions are extensively employed in this tandem repeat architecture to form the unique angles between the three-helix bundles. In the R1/R2 repeats, four H-bonds are found between the side chain atoms of Arg68, Gln64, Asp66, and Lys35, and the backbone oxygen atoms of Trp63, His32, Lys34, and Thr93, respectively [Fig. 3(a)]. Interestingly, the Cα and Cβ carbons of Ser33 are found to stack very well with the phenyl ring of Phe69 [connected by a dotted gray line in Fig. 3(a)], which, along with the side chains of Val73 and His94, form a hydrophobic core between the R1 and R2 repeats. In the R2/R3 repeats, one H-bond between the backbone atoms of G146 and Leu85, and two H-bonds between the side chain atom and backbone atom of Arg145 and the backbone atoms of Gly81 and Gly83 are detected [Fig. 3(b)]. Besides, three hydrophobic residues, Phe108, Trp112, and Pro84, each from a different helix of the bundle, are found to form interesting interdigitated hydrophobic interactions [connected by dotted gray lines in Fig. 3(b)]. A perpendicular Gly83-Hα and Phe147-phenyl ring interaction is also found [Fig. 3(b)]. Such a special perpendicular CH/π interaction is believed to play a role in stabilizing protein conformation.45 Interactions between the XcRecX three-helical bundle repeats. The R1/R2 and R2/R3 repeat interfaces are shown in stereo in Figure 3(a, b), respectively. The R1 repeat is drawn in blue, R2 in green, and R3 in pink. Residues involved in interfacial interactions are shown in ball-and-stick, with those participating in H-bonds connected in dotted orange lines, and those in hydrophobic interactions in dotted gray lines. Oxygen atoms are shown in red, nitrogen atoms in blue, and carbon atoms in gray. Interesting interdigitated hydrophobic interactions among the Phe108, Trp112, and Pro84 residues, each from a different helix, is observed. A perpendicular CαH-phenyl ring interaction between the Gly83 and Phe147 residues is also found. Interestingly, this “L-shaped” XcRecX structure turns out to fit into the notch of the XcRecA dimer and the deep groove of the XcRecA filament very well (data not shown). Three-helix bundle proteins comprise HTH motif, which is extensively employed in proteins with nucleic acid-binding capability.38, 40, 41, 46 In searches for structural homologs of XcRecX using the DALI program,47 with all three tandem repeats as a search model, no hit with Z-score greater than 5 was found; most returned structures display similarity with only one of the repeats. The other typical three-helix bundle repeat protein, such as the c-Myb,41, 48 is much more curved, exhibiting very different apposition of the component helices, and could not be overlapped with XcRecX at all [Fig. 2(d)]. In this respect, the tandem repeats of the XcRecX three-helix bundle can be considered unique, structurally speaking. In this manuscript, we report the crystal structure of RecX protein from the plant pathogen X. campestris that has been determined to a high resolution of 1.50 Å using X-ray crystallography. It adopts a three-helix bundle tandem repeat that bears significant difference to other reported three-helix bundle repeats. A modeled XcRecA/XcRecX filament structure indicates that XcRecX can fit into the notch and the deep groove of the XcRecA filament very well (data not shown), which can partially explain the inhibition mechanisms of the various XcRecA activities. The coordinates of two other RecX protein structures from E. coli (3C1D, determined to a resolution of 1.80 Å)49 and L. reuteri (3D5L, determined to a resolution of 2.35 Å) are recently released. They also adopt a similar structure of three-helix bundle repeats as XcRecX, except that the structure from L. reuteri contains an extra N-terminal domain comprising five β-strands and one helix. XcRecX can superimpose well to the EcRecX and LrRecX structures, with a r.m.s.d. of 1.52 and 1.40 Å for the superimposition of 94 and 95 Cαs out of 142 Cαs of XcRecX, respectively. But the angles between the short and long arms of the “L” shape tandem repeats are distinctive for each RecX protein. Whether such structural differences are related to their functions remain to be elucidated. The authors thank the Core Facilities for Protein X-ray Crystallography in the Academia Sinica, Taiwan, for help in crystal screening, and the National Synchrotron Radiation Research Center (NSRRC) in Taiwan, and the SPring-8 Synchrotron facility in Japan for assistance of X-ray data collection. The National Synchrotron Radiation Research Center is a user facility supported by the National Science Council, Taiwan, Republic of China, and the Protein Crystallography Facility is supported by the National Research Program for Genomic Medicine, Taiwan, Republic of China.