Virus Genes DOI 10.1007/s11262-014-1053-0
Characterization of the RstB2 protein, the DNA-binding protein of CTX/ phage from Vibrio cholerae Alina Falero • Karen Marrero • Sonia Trigueros Rafael Fando
•
Received: 4 November 2013 / Accepted: 28 February 2014 Ó Springer Science+Business Media New York 2014
Abstract The low abundant protein RstB2, encoded in the RS2 region of CTX/, is essential for prophage formation. However, the only biochemical activity so far described is the single/double-stranded DNA-binding capacity of that protein. In this paper, a recombinant RstB2 (rRstB2) protein was overexpressed in E. coli with a yield of 58.4 mg l-1 in shaken cultures, LB broth. The protein, purified to homogeneity, showed an identity with rRstB2 by peptide mass fingerprinting. The apparent molecular weight of the RstB2 native protein suggests that occurs mostly as a monomer in solution. The monomers were able of reacting immediately upon exposure to DNA molecules. After a year of storage at -20 °C, the protein remains biologically active. Bioinformatics analysis of the amino acid sequence of RstB2 predicts the C-end of this protein to be disordered and highly flexible, like in many other single-stranded DNA-binding proteins. When compared with the gVp of M13, conserved amino acids are found at structurally or functionally important relative positions. These results pave the way for additional studies of structure and molecular function of RstB2 for the biology of CTX/. Keywords RstB2 protein Vibrio cholerae CTX/ phage DNA-binding protein Single-stranded DNA Double-stranded DNA A. Falero (&) K. Marrero R. Fando National Center for Scientific Research, Ave 25 and 158, Cubanaca´n, Playa, PO Box 6214, Havana, Cuba e-mail: [email protected] S. Trigueros Department of Physics, James Martin Institute of Nanoscience for Medicine, University of Oxford, Parks Road, Oxford OX1 3PU, UK
Introduction The main virulence factor of Vibrio cholerae, the cholera toxin, is encoded in the genome of the filamentous bacteriophage CTX/ [1]. This unusual filamentous phage can either replicate as a plasmid or site-specifically integrate into the larger of the two V. cholerae chromosomes by hijacking two host-encoded tyrosine recombinases XerC and XerD [2, 3]. However, the efficiency of integration of CTX/ is very low due to the sole attachment site that harbors its circular single-stranded DNA molecule [4], which becomes extremely efficient in the presence of RS region of the phage [5]. In this region, three proteins are encoded, required for repression (RstR), replication (RstA), and phage integration (RstB) [2]. Although this integration event has been extensively investigated [3, 4, 6], the role played by RstB in CTX/ integration is still unclear, since the predicted amino acid sequence of RstB shows no homology to any protein of known function in databases. The only biochemical function so far described for RstB is the single/double-stranded DNA-binding activity [7]. However, there is no clear connection between the DNA-binding activity of RstB and its requirement for CTX/ integration. On the other hand, it has been speculated that the single-stranded DNA-binding protein of RstB may play a role for stabilizing the integration substrate of CTX/ [8]. In this sense, further experiments must be designed and carried out, just to evaluate the contribution of RstB to the integrative process. Thus, enough quantities of the pure protein as well as preliminary knowledge about this protein are needed. In this paper, we obtained a high yield of the homogeneous rRstB2 protein. Moreover, their oligomeric state, isoelectric point, stability as well as the kinetics complex formation with ss and dsDNA were described. Furthermore, the bioinformatics
123
Virus Genes
analysis of its amino acid sequence identified some residues, likely to play an important role in biochemical function, which must be validated experimentally in a forthcoming paper.
Materials and methods Bacterial strains, plasmids Escherichia coli Top 10 transformed with pBAD/Myc– HisC–rstB 9, encoding a recombinant RstB2 protein with a C-terminal 6X-His-tail, or pBAD/Myc–HisC-derivative plasmid, encoding for a tag-less RstB2 variant, were cultured and induced for protein expression as described [7].
Protein stability assay The purified rRstB2–His protein was concentrated up to 3 mg ml-1 and filter-sterilized through a 0.22 lm poresize low binding protein membrane. Sample in 20 mM sodium phosphate, pH 8.0 was divided into aliquots with different additives: 0.5 M Tris, 0.4 M ammonium sulfate, 0.4 M NaCl, 10 mM triton X-100, or 50 % glycerol. An aliquot without additives was used as a control. The final protein concentration in all cases was 1.5 mg ml-1. All samples were stored at -20 °C, and 20 lg of rRstB2–His was assayed every month by SDS-PAGE and EMSA for evaluating its integrity and ssDNA-binding activity, respectively, as described above. Bioinformatics analysis
Overexpression and purification of the recombinant proteins RstB2 and the tag-less RstB2 The recombinant proteins were overexpressed and purified as described before [7]. Identification of rRstB2–His protein by mass spectrometry An aliquot with *5 lg of the recombinant rRstB2–His protein was in-gel digested with trypsine [9]. Peptides recovered with ZipTips were analyzed by low-energy ESI– MS in a QTof-2 mass spectrometer (Micromass, UK). The signals observed in the ESI–MS spectrum were processed with MassLynx v 3.5 software (Micromass, UK) and analyzed using Peptide Mass Fingerprinting. Several of the most intense signals were sequenced by analyzing their MS/MS spectra with the MS/MS ion search option of the MASCOT program (http://www.matrixscience.com). Isoelectric focusing and high performance liquid chromatography-size exclusion chromatography (HPLC-SEC) Both assays were carried out as described previously for the single-stranded DNA-binding protein of VGJ/ [10]. Kinetics analysis of RstB2–His–DNA complex formation The kinetics of ss and dsDNA complex formation of the rRstB2–His product was carried out by electrophoretic mobility shift assay (EMSA) in 0.5 % agarose gel electrophoresis as described previously for the single-stranded DNA-binding (SSB) protein of VGJ/ [10].
123
The amino acid sequence of RstB2 protein was submitted via web to the PredictProtein server, an Internet service that searches up-to-date public sequence databases, creates alignment, and predicts aspects of protein structure and function [11]. This server uses specialized software to predict residue solvent accessibility (PROFacc) and secondary structure (PROFsec) for reliably scored residues only [12]. Moreover, the program ISIS (Interaction Sites Identified from Sequence) predicts specifically residues involved in external protein–protein interactions [13]. Programs like PROFbval, Ucon, and MD predict different aspects of disorder: PROFbval is a neural network-based method that predicts flexible and rigid residues in proteins [14]. Ucon combines predictions for protein specific contacts with a generic pair wise potential for unstructured regions [15], and MD is a novel META-Disorder prediction method that molds various sources of information, predominantly obtained from orthogonal prediction methods, to improve significantly performance over its constituents [15, 16].
Results Expression, purification, and identification of the recombinant RstB2–His protein High expression of rRstB2–His protein was achieved from pBAD/Myc–HisC–rstB 9 plasmid in E. coli Top 10 with a yield of 58.4 mg l-1 of culture. As judged from SDSPAGE, protein preparation was homogeneous. A single band with an apparent molecular weight of *16 kDa was detected closely matching the predicted mass of 16.8 kDa for the recombinant RstB2–His protein. The identity of the purified protein, determined by mass spectrometry analyses of tryptic peptides, revealed several protein features. The
Virus Genes
Fig. 1 MS identification of the purified RstB2-His protein. Deconvoluted ESI–MS spectrum from the tryptic digestion of the purified RstB2-His protein. In parenthesis, the theoretical molecular mass of each signal in the MS spectrum. The predicted amino acidic sequence
Table 1 Sequenced peptides after automatic identification and manual interpretation of the MS/MS spectra
Protein
Peptide sequence
RstB2 protein [Vibrio cholerae]
LWVINMK FVSSDAMTR
[gi|147671922]
VFGASHSEGVSKa TGAPYLIPVLFVGKPIR
1,203.61 1,840.10
a
Peptides automatically identified considering one unspecific cleavage of trypsin. In these cases, the mass spectra were manually analyzed
of RstB2-His protein appears at the top of the figure. In bold and underlined, unidentified residues. With an asterisk, peptides automatically identified, considering one unspecific cleavage of trypsin
C-terminal Histagged
Experimental mass (Da)
Expected mass (Da)
Z
902.55
902.50
2
1,012.51
1,012.46
2
1,203.59 1,840.10
2 3
NDKGQCLTFGLQHQEVK
2,015.03
2,014.99
3
FVVFGASHSEGVSKa
1,449.75
1,449.73
2
TGAPYLIPVLFVGKa
1,473.89
1,473.86
2
GQCLTFGLQHQEVK
1,657.85
1,657.82
2
LEQTAFPVLVTFDNEPDPEDPSR
2,615.18
2,615.24
3
KLEQTAFPVLVTFDNEPDPEDPSR
2,743.40
2,743.33
3
NLVIDYQVVCSLFDNVPGGKPLDKPQPIK
3,266.75
3,266.74
3
NLVIDYQVVCSLFDNVPGGKa
2,250.16
2,250.14
2
SEEDLNSAVDHHHHHHa
1,899.86
1,899.80
3
LISEEDLNSAVDHHHHHH
2,126.00
2,125.97
3
analysis of the ESI–MS spectrum (Fig. 1) by Peptide Mass Fingerprinting identified the purified protein as RstB2 from V. cholerae, with a 92 % of sequence coverage. In addition, several of the most intense signals seen in the mass spectrum were fragmented to obtain their sequence information. The ESI–MS/MS spectra were automatically and in some cases, manually analyzed. The sequence of the identified peptides matched to RstB2’s peptides (Table 1). Moreover, the sequence of the peptide containing the Histail matched to that one, predicted from the nucleotide sequence coding for the recombinant protein (Table 1). Thus, all of the MS data confirmed the identity of the recombinant protein as the RstB2 protein from V. cholerae [gi|147671922], which supports the authenticity of the purified protein.
Characterization of the recombinant RstB2–His purified protein To ascertain its oligomeric structure, 100 lg of rRstB2–His protein was applied in three independent experiments on a gel filtration column (Superdex 200 HR 10/30 column), using as mobile phase 150 mM NaCl, 50 mM phosphate buffer, pH 7.0, flowing at 0.5 ml min-1. The column was initially calibrated with the protein size markers: conalbumin (75 kDa), ovoalbumin (43 kDa), carbonic anhydrase (29 kDa), ribonuclease A (13.7 kDa), and aprotinin (6.5 kDa) (GE Healthcare, UK). From their individual elution volumes, a curve of Kav versus the log10 transformed molecular masses of the protein standards was plotted (Fig. 2a). Under these specific conditions, the
123
Virus Genes
Fig. 2 Physico-chemical characterization of the recombinant RstB2 and RstB2-His proteins. a Molecular weight determination of the recombinant RstB2-His and RstB2 proteins by native HPLC-SEC. Gel filtration was performed as described under the ‘‘Materials and methods’’ with 100 lg of RstB2 or RstB2-His protein, loaded on a Superdex 75 column in three independent experiments. The standard curve was generated by plotting Kav versus log Mr for known
molecular weight standards (Aprotinin, Ribonuclease A, Carbonic anhydrase, Ovoalbumin, and Conalbumin). The positions of the rRstB2 and rRstB2-His on the standard curve are noted. b SDS-PAGE profile on 15 % polyacrylamide gels. Lane 1 broad-range protein molecular mass markers (Promega), lane 2 the recombinant RstB2 purified protein
rRstB–His protein eluted after conalbumin with a Kav = 0.55. This value corresponds to an estimated molecular mass of 67.57 kDa, which is consistent with the predicted 67.24 kDa mass of tetrameric RstB–His. This fact suggests that RstB–His adopts homotetrameric isoforms in solution. Since previous reports have shown that the C-terminal His-tail may promote oligomerization of some recombinant proteins [17–20], the oligomeric state of a tag-less rRstB2 variant was also evaluated. For this purpose, the rRstB2 protein was highly purified as described [7] and evaluated by Coomassie Blue staining after SDS-PAGE (Fig. 2b). The HPLC-SEC elution profile of approximately 100 lg of the purified protein was done under the same running conditions as the His-tagged variant. The rRstB2 protein eluted prior to Ribonuclease A with a Kav = 0.73. This value corresponds to an estimated molecular mass of 14.2 kDa (Fig. 2a), which is consistent with the predicted monomeric isoform of RstB2 (14.07 kDa). These data suggest a predominant monomeric state for the native protein instead of the tetrameric species, obtained for the His-tagged variant. These results reinforce the need of removing the His-tag before any structural characterization of recombinant proteins, which is otherwise in contrast to that one reported for N-terminal His-tagged variants of other four SSB proteins [21]. Isoelectrofocusing of the recombinants RstB2–His and RstB2 proteins showed strong main bands at pH 6.47 and 8.52, matching well to the NTI vector suite 6 program
(InforMax, Inc.) predicted pI of 6.46 and 8.6, respectively. The experimentally determined pI of the His-tagged protein is lower than the value predicted for the native V. cholerae RstB2, due to the presence of five acidic amino acids in the extra 23 C-terminal residues. RstB2 is a strongly alkaline protein, since the predicted isoelectric point equals 8.6, with a net positive charge at physiological pH, thus contributing to DNA binding through electrostatic interactions.
123
Kinetics of RstB2–His–DNA complex formation Since rRstB2–His binds both ssDNA and dsDNA, preferentially to ssDNA with the His-tail playing no role in DNA binding [7], the kinetics of rRstB–His complex assembly with both substrates was studied. In that paper, it was reported that the DNA retard increased for higher rRstB2– His protein concentrations [7], and the same result remained after a year of storage at -20 °C (Fig. 3a). Purified rRstB2–His protein bound ssDNA from an effective concentration of 0.02 lg ll-1 up to an apparent saturation of 0.51 lg ll-1 and no retardation was observed when the DNA–protein mixture was inactivated with 1:1 (vol/vol) phenol–chloroform, indicating that the binding activity is intrinsic to the purified rRstB–His protein.Therefore, 0.51 lg ll-1 of rRstB2–His was used in gel retardation assays to evaluate the reaction times in the formation of rRstB2–His–DNA complexes. Almost instant binding of rRstB–His to ssDNA (Fig. 3b) or dsDNA
Virus Genes
Fig. 3 Kinetics of rRstB2-His-DNA complex formation. rRstB2-HisssDNA binding versus increasing doses of rRstB2-His protein using 0.5 lg of control VGJU-ssDNA. Lanes 1–11: ssDNA plus 0, 0.01, 0.02, 0.07, 0.15, 0.29, 0.37, 0.44, 0.51, 0.59, and 0.66 lg ll-1 of rRstB2-His protein, respectively; lane 12, same as lane 9 treated with
phenol–chloroform (a). Binding reaction times to onset rRstB2-HisDNA complexes formation with both, ssDNA (b) and dsDNA (c). Lane 1: 0.5 lg control VGJU-ssDNA or VGJU-dsDNA, respectively, lanes 2–5, same as lane 1 plus 0.51 lg ll-1 of rRstB2-His protein incubated for binding reaction times of 0, 5, 10, and 20 min
Fig. 4 RstB2-His protein stability according to SDS-PAGE and EMSA. a SDS-PAGE profile of 20 lg RstB2-His protein after 12 months, stored at -20 °C for each of the six conditions studied. Lanes 1–6: 1 rRstB2-His purified protein without additives, 2 lane 1 plus 0.5 M Tris, 3 lane 1 plus 0.4 M (NH4)2SO4, 4 lane 1 plus 0.4 M NaCl, 5 lane 1 plus 10 mM triton X-100 and 6 lane 1 plus 50 %
glycerol. b ssDNA-binding activity of 20 lg rRstB2-His protein, evaluated by EMSA at 12 months, after storage at -20 °C. Lane 1 control of 0.5 lg of control VGJU-ssDNA; lane 2 same as lane 1 plus 20 lg of rRstB2-His protein; lanes 3–7 same as lane 2 plus each of the five stabilizers studied, respectively
(Fig. 3c) was demonstrated. These results indicate that not a noticeable difference in binding to both substrates by rRstB–His could be demonstrated, at least under these experimental conditions. The ability of the protein to bind dsDNA may suggest a wider role for RstB2 in the life cycle of CTX/. After the protein–DNA complexes were assembled, they remained apparently unchanged upon 20 min. Similar results were found before for ssDNA complex formation by pVVGJU protein [10], the singlestranded DNA-binding protein of phage VGJU.
test protein integrity and DNA-binding activity, respectively. The SDS-PAGE profile showed no differences among the purified protein without additive and after addition of all stabilizers, either in pre-storage sampling (Fig. 4a) or after a year of storing at -20 °C (not shown). In fact, unexpected bands indicating protein degradation were not observed. Moreover, rRstB2–His protein remained biologically active after a year stored at -20 °C, according to EMSA (Fig. 4b), even in the absence of additives. Thus, rRstB2 withstood storage at -20 °C for a year and retained its ssDNA-binding ability, which agrees with other results showing high stability for several SSB proteins [21]. In addition, rRstB2 seems to be more stable than pVVGJU, which partially lost ssDNA-binding activity when stored under the same conditions [10].
Stability of rRstB2–His protein upon frozen storage The stability over time of the recombinant rRstB2–His protein was evaluated by SDS-PAGE and EMSA, just to
123
Virus Genes Table 2 Summary of amino acidic composition of SSB proteins from filamentous vibriophages SSB proteins
Charged (%) (DERKHYC)
Polar (%) (NCQSTY)
Hydrophobic (%) (AILFWV)
VGJ
32
22
35
VSKK
32
21
37.6
VSK-intVf12
32.7 32
24 19
33 36
Vf33
32
19
36
F237
31
31
34.3
VF03K6
31
31
34.3
27
24
34
Parahaemolyticus RstB2
Fig. 5 Analysis of the amino acid sequence of the protein RstB2 (gi|147671922). a Visual output of the results from Predict protein web-interface [11]. The C-terminal region is predicted as disordered according to Ucon, MD, and PRObval methods and PROFacc predict it with high residue solvent accessibility (67.4 % of residues with more than 16 % of their surface exposed) and 32.5 % as buried. PROFacc profile-based neural network prediction of solvent accessibility, predictions range from highly accessible (blue) to fully buried (yellow); PROFsec profile-based neural network prediction of secondary structure: yellow rectangles represent predicted strands;
123
Structural analysis of RstB2 protein by bioinformatics methods For understanding the mode of action at the molecular level of RstB, a deep knowledge of the protein’s functional site is required, but experimental determination of protein structure remains expensive and time consuming [22]. Thus, we decide to use bioinformatics analyses that initially let us identify the nature and location of functionally important sites and residues, likely to play a key role in the biochemical function of this protein from its sequence. PredictProtein and I-TASSER are among the most widely used public servers for structure prediction. According to PredictProtein analysis, the RstB2 monomer (gi|147671922) contains 126 aa, being 27 % of them
red smaller rectangles represent a-helices; ISIS method that identifies interacting residues from sequence; PROFbval a web-based interface for prediction of flexibility/rigidity in proteins; Ucon a method for the prediction of unstructured regions; MD META-disorder prediction method; predictions range the higher reliability values of the prediction are shown in red. b Alignment of the amino acid residues of three SSB proteins. Protein sequence of filamentous phage Ff, Pf3, and RstB2 was aligned according to ClustalW2 program. The conserved residues of the DNA-binding domain were shown (Color figure online)
Virus Genes
charged, 23.8 % polar, and 34.1 % hydrophobic. Many of the last ones are found surrounded by positively charged amino acids, the same as all SSB proteins [23]. This amino acidic composition is similar to that of SSB proteins of other filamentous vibriophages (Table 2). A further analysis of the amino acidic composition of RstB2 by composition profiler software (http://www.cprofiler.org/help. html) against the standard protein dataset (SwissProt51) revealed that RstB2 is a proline-rich protein with 8.7 %. According to PredictProtein server, the visual output of the methods predicting the protein structure and function of the RstB2 are depicted (Fig. 5). PROFacc predicted with a high level of reliability (scored C7, scale 0–9) that RstB2 protein has a high residue solvent accessibility (68.25 % of residues with more than 16 % of their surface exposed). The predicted regions with higher residue solvent accessibility are mainly located into the loops, the a-helix, and the C-terminus. Regions predicted as buried are mostly positioned in the predicted b-sheets. The secondary structure predicted (PROFsec) for RstB2 contains 49.2 % b-sheets (interspersed with 44.4 % loop regions and 6.3 % a-helix). It is noteworthy the presence of proline and glycine residues in loop regions, which are frequently found in b-turns, a type of secondary structure that promotes the formation of antiparallel b-sheets [24]. On the other hand, ISIS predicted some residues involved in external protein–protein interactions, being most of them clustered in the region from residue 45–64. The majority of these residues are located in predicted flexible and unstructured regions. It is generally thought that regions involved in protein–protein interactions with multiple partners are flexible and unstructured, in order to fix the different requirements for the interacting partners [25]. The predicted flexible residues in this protein, according to PROFbval, are widespread throughout the whole aminoacidic sequence and coincide with those ones, predicted as exposed. On the other hand, the unstructured regions in this protein predicted by Ucon are found from residue 21, being the largest from residues 86 to 95 and 113 to 126, just located the last one in the C-terminus. Some charged residues (K, D, E, Y), glycine, and proline but few hydrophobic ones are found in these regions, as described for this kind of regions [26] and other unstructured regions of DNA-binding proteins [15]. Moreover, MD predicted some signal for the presence of disordered regions just in N-terminus and from residues 55–73 and 85–92, but the higher output values are found in the C-terminus (the last 17 residues). To gain an insight into the structure of the RstB2 protein, we have investigated its three-dimensional structure by using the I-TASSER server. I-Tasser utilizes a robust metathreading approach to identify potential templates. I-Tasser identified as the best ranked template the structure
of the SSB protein from the phage Pf3 from Pseudomonas aeruginosa (1pfs) and yielded five low quality models, according to analysis with the Q-Mean software. Prediction of RstB2 structure with other modeling servers (LOMETS, MUSTER, QUARTS, SEGMER, 3D-Jigsaw, and Raptor) yielded no models or low quality models. Then, taking into account the identification of 1pfs as template, the amino acid sequence of RstB2 was compared with the amino acid sequence of the well-characterized SSB proteins Pf3 and Ff GVP (Fig. 5b), which have very similar three-dimensional structures [27]. The alignment revealed that the three proteins have 8 identical and 15 similar residues, showing the poor sequence conservation among SSB proteins. Interestingly, among them, there were found some conserved residues that coincide with those involved in the structural and biological function of the GVP of M13 phage [27]. Hydrophobic conserved amino acids (V4, I6, L32, V43, I47, V63, I78, and V84) in the GVP of M13 have their corresponding residues (I7, M9, L57, V81, F85, I100, L118, and I124) in the RstB2 protein. Furthermore, two positively charged amino acids (K7 and R82) in the GVP of M13 were conserved in the RstB2 protein (K10 and Q122). In addition, several residues involved in DNA binding in the GVP of M13 (G18, S20, G23, Y26, and L28) have been conserved in RstB2 protein (G23, S25, G28, Y31, and L53), and the Y41 of M13 is replaced by F79 in RstB2 protein.
Discussion In the current study, we have made further biochemical characterization of rRstB2, to identify putative structural determinants of the function of this protein. First, we determined the oligomeric status of rRstB2 protein and found that rRstB2 appeared to be monomeric at concentrations &1 mg ml-1, unlike SSB proteins of other filamentous phages, which occurs as dimmers (Ff gVp [28, 29], M13 gVp [29], Cf gVp [30], and pVVGJU protein [10]. The behavior of rRstB2 in solution is similar to that one of the monomer T4 gene 32 protein [31]. This result does not exclude the possibility of assembling in higher order structures under other physiological conditions or upon DNA binding, as has been shown for the monomeric SSB protein of Bacillus anthracis, which appears to form a transient tetrameric structure with the ssDNA [32]. On the other hand, the recombinant RstB2 protein has been previously shown to be capable of preferentially binding ssDNA in vitro and accordingly, it was proposed to be renamed pVCTX/ [7]. Here, we have shown that the kinetic of reaction in the protein–DNA complex formation with both ss and dsDNA is similar, which remained after a year of storing at
123
Virus Genes
-20 °C. These results, along with those ones, previously found by our team, suggest that RstB2 may be a noncanonical filamentous phage SSB protein of an unusual filamentous phage such as CTX/. Although CTX/ is a filamentous phage, it is known that its ssDNA, substrate of integration, must be folded into a stem-loop structure, thus creating a small region of duplex DNA, which is the target of the site-specific recombinases [4]. However, it is still unknown how and when the attP site folds inside the cell, even the identity of the factors influencing this process. On the other hand, the preferential in vitro binding of ssDNA by rRstB2 [7] supports its proposed role in protecting the CTX/ ssDNA for integration [8]. Moreover, dsDNA binding may be physiologically important during CTX/ integration by stabilizing the secondary structure (stem loop) of the substrate integration, in a way not yet identified. Besides, it is interesting to know that two distinct binding modes have been previously reported for the SSB protein of two well-studied models, bacteriophages T7, and T4. Firstly, they bind trough weak non-electrostatic interactions to the backbone of the dsDNA; secondly, they can bind strongly to exposed ssDNA regions through electrostatic and non-electrostatic interactions. Both proteins show strong preferential ssDNA binding and weak dsDNA binding, which are essential for their ability to search dsDNA in one dimension, just to find available ssDNA-binding sites at the replication fork. These properties are likely essential for the ability of these proteins to effectively act as part of the DNA replication process [33]. In a similar way, RstB2 might dislodge the Vibrio SSB protein coating the displaced ssDNA molecules, formed in the first cycles of the replication by the rolling circle mechanism (through its ssDNA affinity) and may promote the formation or stabilization of the secondary structure (through its dsDNA affinity), needed for integration. Binding of ss and dsDNA molecules by RstB2 may demand some structural features, such as a big enough DNA-binding surface to fit the smaller single-stranded DNA and the much wider duplex as well as a highly flexible structure, being capable of suffer a deformation, in order to accommodate the relatively inflexible duplex onto the binding surface. On the other hand, RstB2 may undergo structural transformations upon binding ssDNA, which may support its preferential binding to ssDNA. In addition, the bioinformatics results predicted that RstB2 has seven b-sheets and one a-helix interspersed with loop regions. This secondary structure is suggestive of an OB-fold that in the case of RstB2, the formation of this central conserved OB-fold may be favored by the proline and glycine residues, found in the loop regions, which may promote the formation of antiparallel b-sheets [24]. However, the three-dimensional structure prediction failed in
123
producing a high quality model, which may be related to the known difficulty for identifying the OB-fold by sequence analysis, due to the poor sequence conservation found. This is reinforced in the case of phage SSB proteins, whose structures vary greater that those ones from cellular organisms, according to the SCOP database. Nevertheless, some conserved amino acid residues were found (Fig. 5b), which could be involved in structural (hydrophobic core residues) and DNA-binding functions. Thus, these results suggest that some structural or functional similarity between RstB2 and the SSB protein of Pf3 phage could be considered. Furthermore, analysis by PROFbval, Ucon, and MD predicted that RstB2 protein has a disordered, highly flexible C-terminus with no regular secondary structure, as expected for this kind of regions. Moreover, it has a proline residue involved in mediating protein–protein interactions, similar to other prokaryotic SSBs [25, 31, 34, 35]. In fact, the charged and proline residues found could contribute to the high flexibility, predicted, and needed for that region. It is known that the disordered regions are flexible, dynamic and can be partially or completely extended in solution. It is noteworthy that the aromatic residue (phenylalanine) found in this flexible unstructured region, which could be critical in binding ssDNA. On the other hand, it has been experimentally shown that disordered regions are involved in molecular recognition, in particular DNA binding, in order to facilitate processes such as transcription, transposition, packing, repair, and replication [26]. Therefore, prediction of such regions would provide the first step in methods for functionally identifying relevant ones. Furthermore, there are some proteins that require a conformational change, previous to DNA binding where C-terminal regions are involved. For instance, the T7 gene 2.5 protein and T4 gene 32 protein, which have a C-terminus functioning as a two-way switch that coordinates and modulates the ssDNA-binding and the protein–protein interactions [25, 36]. This mechanism is used by abundant cellular proteins as a tool for the coordination of nucleic acids and protein–protein interactions within the macromolecular machinery that ensures the integrity of the genome and its replication. We consider that the C-terminus disordered and flexible found in RstB2 protein could be functionally important, as described previously for other SSBs. This region is enriched in charged and polar residues that favor the interaction with water and other residues (glycine and proline), which may alter structured helices or strands and have high net charges [37]. These residues could contribute, on one hand, to the required conformational change before binding DNA and on the other hand, to increase DNA–protein and protein–protein interactions that occur upon binding to ssDNA. It is known that the cooperative binding described for SSBs involves protein–protein interactions between the core of one ssDNA-
Virus Genes
bound protein and the N-domain of the adjacent bound protein [31].
Conclusions The characterization of rRstB2 revealed that this protein occurs as a monomeric species in solution and its C-terminal His-tail promoted oligomerization and increased its isoelectric point. Moreover, it was confirmed that although it preferentially binds ssDNA, it is also capable of interacting with dsDNA, following a similar kinetics. Like other SSBs, the protein is predicted to have a disordered, flexible, and exposed C-terminus that may be physiological relevant during the chromosomal integration of the CTX/ genome. The protein remained biologically active after a year of storage at -20 °C. Acknowledgments The authors would like to thank Prof. Celso Perez for his valuable assistance in English revision.
References 1. M.K. Waldor, J.J. Mekalanos, Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272, 1910–1914 (1996) 2. M.K. Waldor, E.J. Rubin, G.D. Pearson, H. Kimsey, J.J. Mekalanos, Regulation, replication, and integration functions of the Vibrio cholerae CTXphi are encoded by region RS2. Mol. Microbiol. 24, 917–926 (1997) 3. K.E. Huber, M.K. Waldor, Filamentous phage integration requires the host recombinases XerC and XerD. Nature 417, 656–659 (2002) 4. M.E. Val, M. Bouvier, J. Campos, D. Sherratt, F. Cornet, D. Mazel, F.X. Barre, The single-stranded genome of phage CTX is the form used for integration into the genome of Vibrio cholerae. Mol. Cell 19, 559–566 (2005) 5. B. Das, J. Bischerour, M.E. Val, F.X. Barre, Molecular keys of the tropism of integration of the cholera toxin phage. Proc Natl Acad Sci USA 107, 4377–4382 (2010) 6. B. Das, J. Bischerour, F.X. Barre, VGJphi integration and excision mechanisms contribute to the genetic diversity of Vibrio cholerae epidemic strains. Proc Natl Acad Sci USA 108, 2516–2521 (2011) 7. A. Falero, A. Caballero, B. Ferran, Y. Izquierdo, R. Fando, J. Campos, DNA binding proteins of the filamentous phages CTXphi and VGJphi of Vibrio cholerae. J. Bacteriol. 191, 5873–5876 (2009) 8. B. Das, J. Bischerour, F.X. Barre, Molecular mechanism of acquisition of the cholera toxin genes. Indian J. Med. Res. 133, 195–200 (2011) 9. R. Weistermeier, T. Naven, Proteomics in Practice. A Laboratory Manual of Proteome Analysis (Wiley-VCH, Weinheim, 2002) 10. A. Falero, A. Caballero, S. Trigueros, C. Perez, J. Campos, K. Marrero, R. Fando, Characterization of the single-stranded DNA binding protein pV(VGJPhi) of VGJPhi phage from Vibrio cholerae. Biochim. Biophys. Acta 2011, 1107–1112 (1814) 11. B. Rost, G. Yachdav, J. Liu, The PredictProtein server. Nucleic Acids Res. 32, W321–W326 (2004)
12. B. Rost, How to Use Protein 1D Structure Predicted by PROFphd (Humana, Totowa, 2005) 13. Y. Ofran, B. Rost, ISIS: interaction sites identified from sequence. Bioinformatics 23, e13–e16 (2007) 14. A. Schlessinger, B. Rost, Protein flexibility and rigidity predicted from sequence. Proteins 61, 115–126 (2005) 15. A. Schlessinger, M. Punta, B. Rost, Natively unstructured regions in proteins identified from contact predictions. Bioinformatics 23, 2376–2384 (2007) 16. A. Schlessinger, M. Punta, G. Yachdav, L. Kajan, B. Rost, Improved disorder prediction by combination of orthogonal approaches. PLoS ONE 4, e4433 (2009) 17. M. Amor-Mahjoub, J.P. Suppini, N. Gomez-Vrielyunck, M. Ladjimi, The effect of the hexahistidine-tag in the oligomerization of HSC70 constructs. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 844, 328–334 (2006) 18. P. Perron-Savard, G. De Crescenzo, H. Le Moual, Dimerization and DNA binding of the Salmonella enterica PhoP response regulator are phosphorylation independent. Microbiology 151, 3979–3987 (2005) 19. K.L. Cheung, J. Huen, Y. Kakihara, W.A. Houry, J. Ortega, Alternative oligomeric states of the yeast Rvb1/Rvb2 complex induced by histidine tags. J. Mol. Biol. 404, 478–492 (2010) 20. T.S. Kaoud, A.K. Devkota, R. Harris, M.S. Rana, O. Abramczyk, M. Warthaka, S. Lee, M.E. Girvin, A.F. Riggs, K.N. Dalby, Activated ERK2 is a monomer in vitro with or without divalent cations and when complexed to the cytoplasmic scaffold PEA-15. Biochemistry 50, 4568–4578 (2011) 21. H. Shi, Y. Zhang, G. Zhang, J. Guo, X. Zhang, H. Song, J. Lv, J. Gao, Y. Wang, L. Chen, Systematic functional comparative analysis of four single-stranded DNA-binding proteins and their affection on viral RNA metabolism. PLoS ONE 8, e55076 (2013) 22. J. Cheng, A.Z. Randall, M.J. Sweredoski, P. Baldi, SCRATCH: a protein structure and structural feature prediction server. Nucl. Acids Res. 33, W72–W76 (2005) 23. B. Marintcheva, A. Marintchev, G. Wagner, C.C. Richardson, Acidic C-terminal tail of the ssDNA-binding protein of bacteriophage T7 and ssDNA compete for the same binding surface. Proc. Natl. Acad. Sci. USA 105, 1855–1860 (2008) 24. S.R. Trevino, S. Schaefer, J.M. Scholtz, C.N. Pace, Increasing protein conformational stability by optimizing beta-turn sequence. J. Mol. Biol. 373, 211–218 (2007) 25. B. Marintcheva, S.M. Hamdan, S.J. Lee, C.C. Richardson, Essential residues in the C terminus of the bacteriophage T7 gene 2.5 single-stranded DNA-binding protein. J. Biol. Chem. 281, 25831–25840 (2006) 26. V.N. Uversky, C.J. Oldfield, A.K. Dunker, Showing your ID: intrinsic disorder as an ID for recognition, regulation and cell signaling. J. Mol. Recognit. 18, 343–384 (2005) 27. R.H. Folmer, M. Nilges, R.N. Konings, C.W. Hilbers, Solution structure of the single-stranded DNA binding protein of the filamentous Pseudomonas phage Pf3: similarity to other proteins binding to single-stranded nucleic acids. EMBO J. 14, 4132–4142 (1995) 28. T.C. Mou, M.C. Shen, T.C. Terwilliger, D.M. Gray, Binding and reversible denaturation of double-stranded DNA by Ff gene 5 protein. Biopolymers 70, 637–648 (2003) 29. N.C. Alma, B.J. Harmsen, J.H. van Boom, G. van der Marel, C.W. Hilbers, 1H NMR studies of the binding of bacteriophageM13-encoded gene-5 protein to oligo(deoxyadenylic acid)s of varying length. Eur. J. Biochem. 122, 319–326 (1982) 30. J.D. Wen, D.M. Gray, Selection of genomic sequences that bind tightly to Ff gene 5 protein: primer-free genomic SELEX. Nucl. Acids Res. 32, e182 (2004) 31. K. Pant, R.L. Karpel, I. Rouzina, M.C. Williams, Salt dependent binding of T4 gene 32 protein to single and double-stranded
123
Virus Genes DNA: single molecule force spectroscopy measurements. J. Mol. Biol. 349, 317–330 (2005) 32. E.E. Biswas-Fiss, J. Kukiratirat, S.B. Biswas, Thermodynamic analysis of DNA binding by a Bacillus single stranded DNA binding protein. BMC Biochem. 13, 10 (2012) 33. R.D. Shereda, A.G. Kozlov, T.M. Lohman, M.M. Cox, J.L. Keck, SSB as an organizer/mobilizer of genome maintenance complexes. Crit. Rev. Biochem. Mol. Biol. 43, 289–318 (2008) 34. T. Hollis, J.M. Stattel, D.S. Walther, C.C. Richardson, T. Ellenberger, Structure of the gene 2.5 protein, a single-stranded DNA binding protein encoded by bacteriophage T7. Proc. Natl. Acad. Sci. USA 98, 9557–9562 (2001)
123
35. S.N. Savvides, S. Raghunathan, K. Futterer, A.G. Kozlov, T.M. Lohman, G. Waksman, The C-terminal domain of full-length E. coli SSB is disordered even when bound to DNA. Protein Sci. 13, 1942–1947 (2004) 36. L. Shokri, I. Rouzina, M.C. Williams, Interaction of bacteriophage T4 and T7 single-stranded DNA-binding proteins with DNA. Phys. Biol. 6, 025002 (2009) 37. M. Fuxreiter, I. Simon, P. Friedrich, P. Tompa, Preformed structural elements feature in partner recognition by intrinsically unstructured proteins. J. Mol. Biol. 338, 1015–1026 (2004)
Characterization of the RstB2 protein, the DNA-binding protein of CTX/ phage from Vibrio cholerae Alina Falero • Karen Marrero • Sonia Trigueros Rafael Fando
•
Received: 4 November 2013 / Accepted: 28 February 2014 Ó Springer Science+Business Media New York 2014
Abstract The low abundant protein RstB2, encoded in the RS2 region of CTX/, is essential for prophage formation. However, the only biochemical activity so far described is the single/double-stranded DNA-binding capacity of that protein. In this paper, a recombinant RstB2 (rRstB2) protein was overexpressed in E. coli with a yield of 58.4 mg l-1 in shaken cultures, LB broth. The protein, purified to homogeneity, showed an identity with rRstB2 by peptide mass fingerprinting. The apparent molecular weight of the RstB2 native protein suggests that occurs mostly as a monomer in solution. The monomers were able of reacting immediately upon exposure to DNA molecules. After a year of storage at -20 °C, the protein remains biologically active. Bioinformatics analysis of the amino acid sequence of RstB2 predicts the C-end of this protein to be disordered and highly flexible, like in many other single-stranded DNA-binding proteins. When compared with the gVp of M13, conserved amino acids are found at structurally or functionally important relative positions. These results pave the way for additional studies of structure and molecular function of RstB2 for the biology of CTX/. Keywords RstB2 protein Vibrio cholerae CTX/ phage DNA-binding protein Single-stranded DNA Double-stranded DNA A. Falero (&) K. Marrero R. Fando National Center for Scientific Research, Ave 25 and 158, Cubanaca´n, Playa, PO Box 6214, Havana, Cuba e-mail: [email protected] S. Trigueros Department of Physics, James Martin Institute of Nanoscience for Medicine, University of Oxford, Parks Road, Oxford OX1 3PU, UK
Introduction The main virulence factor of Vibrio cholerae, the cholera toxin, is encoded in the genome of the filamentous bacteriophage CTX/ [1]. This unusual filamentous phage can either replicate as a plasmid or site-specifically integrate into the larger of the two V. cholerae chromosomes by hijacking two host-encoded tyrosine recombinases XerC and XerD [2, 3]. However, the efficiency of integration of CTX/ is very low due to the sole attachment site that harbors its circular single-stranded DNA molecule [4], which becomes extremely efficient in the presence of RS region of the phage [5]. In this region, three proteins are encoded, required for repression (RstR), replication (RstA), and phage integration (RstB) [2]. Although this integration event has been extensively investigated [3, 4, 6], the role played by RstB in CTX/ integration is still unclear, since the predicted amino acid sequence of RstB shows no homology to any protein of known function in databases. The only biochemical function so far described for RstB is the single/double-stranded DNA-binding activity [7]. However, there is no clear connection between the DNA-binding activity of RstB and its requirement for CTX/ integration. On the other hand, it has been speculated that the single-stranded DNA-binding protein of RstB may play a role for stabilizing the integration substrate of CTX/ [8]. In this sense, further experiments must be designed and carried out, just to evaluate the contribution of RstB to the integrative process. Thus, enough quantities of the pure protein as well as preliminary knowledge about this protein are needed. In this paper, we obtained a high yield of the homogeneous rRstB2 protein. Moreover, their oligomeric state, isoelectric point, stability as well as the kinetics complex formation with ss and dsDNA were described. Furthermore, the bioinformatics
123
Virus Genes
analysis of its amino acid sequence identified some residues, likely to play an important role in biochemical function, which must be validated experimentally in a forthcoming paper.
Materials and methods Bacterial strains, plasmids Escherichia coli Top 10 transformed with pBAD/Myc– HisC–rstB 9, encoding a recombinant RstB2 protein with a C-terminal 6X-His-tail, or pBAD/Myc–HisC-derivative plasmid, encoding for a tag-less RstB2 variant, were cultured and induced for protein expression as described [7].
Protein stability assay The purified rRstB2–His protein was concentrated up to 3 mg ml-1 and filter-sterilized through a 0.22 lm poresize low binding protein membrane. Sample in 20 mM sodium phosphate, pH 8.0 was divided into aliquots with different additives: 0.5 M Tris, 0.4 M ammonium sulfate, 0.4 M NaCl, 10 mM triton X-100, or 50 % glycerol. An aliquot without additives was used as a control. The final protein concentration in all cases was 1.5 mg ml-1. All samples were stored at -20 °C, and 20 lg of rRstB2–His was assayed every month by SDS-PAGE and EMSA for evaluating its integrity and ssDNA-binding activity, respectively, as described above. Bioinformatics analysis
Overexpression and purification of the recombinant proteins RstB2 and the tag-less RstB2 The recombinant proteins were overexpressed and purified as described before [7]. Identification of rRstB2–His protein by mass spectrometry An aliquot with *5 lg of the recombinant rRstB2–His protein was in-gel digested with trypsine [9]. Peptides recovered with ZipTips were analyzed by low-energy ESI– MS in a QTof-2 mass spectrometer (Micromass, UK). The signals observed in the ESI–MS spectrum were processed with MassLynx v 3.5 software (Micromass, UK) and analyzed using Peptide Mass Fingerprinting. Several of the most intense signals were sequenced by analyzing their MS/MS spectra with the MS/MS ion search option of the MASCOT program (http://www.matrixscience.com). Isoelectric focusing and high performance liquid chromatography-size exclusion chromatography (HPLC-SEC) Both assays were carried out as described previously for the single-stranded DNA-binding protein of VGJ/ [10]. Kinetics analysis of RstB2–His–DNA complex formation The kinetics of ss and dsDNA complex formation of the rRstB2–His product was carried out by electrophoretic mobility shift assay (EMSA) in 0.5 % agarose gel electrophoresis as described previously for the single-stranded DNA-binding (SSB) protein of VGJ/ [10].
123
The amino acid sequence of RstB2 protein was submitted via web to the PredictProtein server, an Internet service that searches up-to-date public sequence databases, creates alignment, and predicts aspects of protein structure and function [11]. This server uses specialized software to predict residue solvent accessibility (PROFacc) and secondary structure (PROFsec) for reliably scored residues only [12]. Moreover, the program ISIS (Interaction Sites Identified from Sequence) predicts specifically residues involved in external protein–protein interactions [13]. Programs like PROFbval, Ucon, and MD predict different aspects of disorder: PROFbval is a neural network-based method that predicts flexible and rigid residues in proteins [14]. Ucon combines predictions for protein specific contacts with a generic pair wise potential for unstructured regions [15], and MD is a novel META-Disorder prediction method that molds various sources of information, predominantly obtained from orthogonal prediction methods, to improve significantly performance over its constituents [15, 16].
Results Expression, purification, and identification of the recombinant RstB2–His protein High expression of rRstB2–His protein was achieved from pBAD/Myc–HisC–rstB 9 plasmid in E. coli Top 10 with a yield of 58.4 mg l-1 of culture. As judged from SDSPAGE, protein preparation was homogeneous. A single band with an apparent molecular weight of *16 kDa was detected closely matching the predicted mass of 16.8 kDa for the recombinant RstB2–His protein. The identity of the purified protein, determined by mass spectrometry analyses of tryptic peptides, revealed several protein features. The
Virus Genes
Fig. 1 MS identification of the purified RstB2-His protein. Deconvoluted ESI–MS spectrum from the tryptic digestion of the purified RstB2-His protein. In parenthesis, the theoretical molecular mass of each signal in the MS spectrum. The predicted amino acidic sequence
Table 1 Sequenced peptides after automatic identification and manual interpretation of the MS/MS spectra
Protein
Peptide sequence
RstB2 protein [Vibrio cholerae]
LWVINMK FVSSDAMTR
[gi|147671922]
VFGASHSEGVSKa TGAPYLIPVLFVGKPIR
1,203.61 1,840.10
a
Peptides automatically identified considering one unspecific cleavage of trypsin. In these cases, the mass spectra were manually analyzed
of RstB2-His protein appears at the top of the figure. In bold and underlined, unidentified residues. With an asterisk, peptides automatically identified, considering one unspecific cleavage of trypsin
C-terminal Histagged
Experimental mass (Da)
Expected mass (Da)
Z
902.55
902.50
2
1,012.51
1,012.46
2
1,203.59 1,840.10
2 3
NDKGQCLTFGLQHQEVK
2,015.03
2,014.99
3
FVVFGASHSEGVSKa
1,449.75
1,449.73
2
TGAPYLIPVLFVGKa
1,473.89
1,473.86
2
GQCLTFGLQHQEVK
1,657.85
1,657.82
2
LEQTAFPVLVTFDNEPDPEDPSR
2,615.18
2,615.24
3
KLEQTAFPVLVTFDNEPDPEDPSR
2,743.40
2,743.33
3
NLVIDYQVVCSLFDNVPGGKPLDKPQPIK
3,266.75
3,266.74
3
NLVIDYQVVCSLFDNVPGGKa
2,250.16
2,250.14
2
SEEDLNSAVDHHHHHHa
1,899.86
1,899.80
3
LISEEDLNSAVDHHHHHH
2,126.00
2,125.97
3
analysis of the ESI–MS spectrum (Fig. 1) by Peptide Mass Fingerprinting identified the purified protein as RstB2 from V. cholerae, with a 92 % of sequence coverage. In addition, several of the most intense signals seen in the mass spectrum were fragmented to obtain their sequence information. The ESI–MS/MS spectra were automatically and in some cases, manually analyzed. The sequence of the identified peptides matched to RstB2’s peptides (Table 1). Moreover, the sequence of the peptide containing the Histail matched to that one, predicted from the nucleotide sequence coding for the recombinant protein (Table 1). Thus, all of the MS data confirmed the identity of the recombinant protein as the RstB2 protein from V. cholerae [gi|147671922], which supports the authenticity of the purified protein.
Characterization of the recombinant RstB2–His purified protein To ascertain its oligomeric structure, 100 lg of rRstB2–His protein was applied in three independent experiments on a gel filtration column (Superdex 200 HR 10/30 column), using as mobile phase 150 mM NaCl, 50 mM phosphate buffer, pH 7.0, flowing at 0.5 ml min-1. The column was initially calibrated with the protein size markers: conalbumin (75 kDa), ovoalbumin (43 kDa), carbonic anhydrase (29 kDa), ribonuclease A (13.7 kDa), and aprotinin (6.5 kDa) (GE Healthcare, UK). From their individual elution volumes, a curve of Kav versus the log10 transformed molecular masses of the protein standards was plotted (Fig. 2a). Under these specific conditions, the
123
Virus Genes
Fig. 2 Physico-chemical characterization of the recombinant RstB2 and RstB2-His proteins. a Molecular weight determination of the recombinant RstB2-His and RstB2 proteins by native HPLC-SEC. Gel filtration was performed as described under the ‘‘Materials and methods’’ with 100 lg of RstB2 or RstB2-His protein, loaded on a Superdex 75 column in three independent experiments. The standard curve was generated by plotting Kav versus log Mr for known
molecular weight standards (Aprotinin, Ribonuclease A, Carbonic anhydrase, Ovoalbumin, and Conalbumin). The positions of the rRstB2 and rRstB2-His on the standard curve are noted. b SDS-PAGE profile on 15 % polyacrylamide gels. Lane 1 broad-range protein molecular mass markers (Promega), lane 2 the recombinant RstB2 purified protein
rRstB–His protein eluted after conalbumin with a Kav = 0.55. This value corresponds to an estimated molecular mass of 67.57 kDa, which is consistent with the predicted 67.24 kDa mass of tetrameric RstB–His. This fact suggests that RstB–His adopts homotetrameric isoforms in solution. Since previous reports have shown that the C-terminal His-tail may promote oligomerization of some recombinant proteins [17–20], the oligomeric state of a tag-less rRstB2 variant was also evaluated. For this purpose, the rRstB2 protein was highly purified as described [7] and evaluated by Coomassie Blue staining after SDS-PAGE (Fig. 2b). The HPLC-SEC elution profile of approximately 100 lg of the purified protein was done under the same running conditions as the His-tagged variant. The rRstB2 protein eluted prior to Ribonuclease A with a Kav = 0.73. This value corresponds to an estimated molecular mass of 14.2 kDa (Fig. 2a), which is consistent with the predicted monomeric isoform of RstB2 (14.07 kDa). These data suggest a predominant monomeric state for the native protein instead of the tetrameric species, obtained for the His-tagged variant. These results reinforce the need of removing the His-tag before any structural characterization of recombinant proteins, which is otherwise in contrast to that one reported for N-terminal His-tagged variants of other four SSB proteins [21]. Isoelectrofocusing of the recombinants RstB2–His and RstB2 proteins showed strong main bands at pH 6.47 and 8.52, matching well to the NTI vector suite 6 program
(InforMax, Inc.) predicted pI of 6.46 and 8.6, respectively. The experimentally determined pI of the His-tagged protein is lower than the value predicted for the native V. cholerae RstB2, due to the presence of five acidic amino acids in the extra 23 C-terminal residues. RstB2 is a strongly alkaline protein, since the predicted isoelectric point equals 8.6, with a net positive charge at physiological pH, thus contributing to DNA binding through electrostatic interactions.
123
Kinetics of RstB2–His–DNA complex formation Since rRstB2–His binds both ssDNA and dsDNA, preferentially to ssDNA with the His-tail playing no role in DNA binding [7], the kinetics of rRstB–His complex assembly with both substrates was studied. In that paper, it was reported that the DNA retard increased for higher rRstB2– His protein concentrations [7], and the same result remained after a year of storage at -20 °C (Fig. 3a). Purified rRstB2–His protein bound ssDNA from an effective concentration of 0.02 lg ll-1 up to an apparent saturation of 0.51 lg ll-1 and no retardation was observed when the DNA–protein mixture was inactivated with 1:1 (vol/vol) phenol–chloroform, indicating that the binding activity is intrinsic to the purified rRstB–His protein.Therefore, 0.51 lg ll-1 of rRstB2–His was used in gel retardation assays to evaluate the reaction times in the formation of rRstB2–His–DNA complexes. Almost instant binding of rRstB–His to ssDNA (Fig. 3b) or dsDNA
Virus Genes
Fig. 3 Kinetics of rRstB2-His-DNA complex formation. rRstB2-HisssDNA binding versus increasing doses of rRstB2-His protein using 0.5 lg of control VGJU-ssDNA. Lanes 1–11: ssDNA plus 0, 0.01, 0.02, 0.07, 0.15, 0.29, 0.37, 0.44, 0.51, 0.59, and 0.66 lg ll-1 of rRstB2-His protein, respectively; lane 12, same as lane 9 treated with
phenol–chloroform (a). Binding reaction times to onset rRstB2-HisDNA complexes formation with both, ssDNA (b) and dsDNA (c). Lane 1: 0.5 lg control VGJU-ssDNA or VGJU-dsDNA, respectively, lanes 2–5, same as lane 1 plus 0.51 lg ll-1 of rRstB2-His protein incubated for binding reaction times of 0, 5, 10, and 20 min
Fig. 4 RstB2-His protein stability according to SDS-PAGE and EMSA. a SDS-PAGE profile of 20 lg RstB2-His protein after 12 months, stored at -20 °C for each of the six conditions studied. Lanes 1–6: 1 rRstB2-His purified protein without additives, 2 lane 1 plus 0.5 M Tris, 3 lane 1 plus 0.4 M (NH4)2SO4, 4 lane 1 plus 0.4 M NaCl, 5 lane 1 plus 10 mM triton X-100 and 6 lane 1 plus 50 %
glycerol. b ssDNA-binding activity of 20 lg rRstB2-His protein, evaluated by EMSA at 12 months, after storage at -20 °C. Lane 1 control of 0.5 lg of control VGJU-ssDNA; lane 2 same as lane 1 plus 20 lg of rRstB2-His protein; lanes 3–7 same as lane 2 plus each of the five stabilizers studied, respectively
(Fig. 3c) was demonstrated. These results indicate that not a noticeable difference in binding to both substrates by rRstB–His could be demonstrated, at least under these experimental conditions. The ability of the protein to bind dsDNA may suggest a wider role for RstB2 in the life cycle of CTX/. After the protein–DNA complexes were assembled, they remained apparently unchanged upon 20 min. Similar results were found before for ssDNA complex formation by pVVGJU protein [10], the singlestranded DNA-binding protein of phage VGJU.
test protein integrity and DNA-binding activity, respectively. The SDS-PAGE profile showed no differences among the purified protein without additive and after addition of all stabilizers, either in pre-storage sampling (Fig. 4a) or after a year of storing at -20 °C (not shown). In fact, unexpected bands indicating protein degradation were not observed. Moreover, rRstB2–His protein remained biologically active after a year stored at -20 °C, according to EMSA (Fig. 4b), even in the absence of additives. Thus, rRstB2 withstood storage at -20 °C for a year and retained its ssDNA-binding ability, which agrees with other results showing high stability for several SSB proteins [21]. In addition, rRstB2 seems to be more stable than pVVGJU, which partially lost ssDNA-binding activity when stored under the same conditions [10].
Stability of rRstB2–His protein upon frozen storage The stability over time of the recombinant rRstB2–His protein was evaluated by SDS-PAGE and EMSA, just to
123
Virus Genes Table 2 Summary of amino acidic composition of SSB proteins from filamentous vibriophages SSB proteins
Charged (%) (DERKHYC)
Polar (%) (NCQSTY)
Hydrophobic (%) (AILFWV)
VGJ
32
22
35
VSKK
32
21
37.6
VSK-intVf12
32.7 32
24 19
33 36
Vf33
32
19
36
F237
31
31
34.3
VF03K6
31
31
34.3
27
24
34
Parahaemolyticus RstB2
Fig. 5 Analysis of the amino acid sequence of the protein RstB2 (gi|147671922). a Visual output of the results from Predict protein web-interface [11]. The C-terminal region is predicted as disordered according to Ucon, MD, and PRObval methods and PROFacc predict it with high residue solvent accessibility (67.4 % of residues with more than 16 % of their surface exposed) and 32.5 % as buried. PROFacc profile-based neural network prediction of solvent accessibility, predictions range from highly accessible (blue) to fully buried (yellow); PROFsec profile-based neural network prediction of secondary structure: yellow rectangles represent predicted strands;
123
Structural analysis of RstB2 protein by bioinformatics methods For understanding the mode of action at the molecular level of RstB, a deep knowledge of the protein’s functional site is required, but experimental determination of protein structure remains expensive and time consuming [22]. Thus, we decide to use bioinformatics analyses that initially let us identify the nature and location of functionally important sites and residues, likely to play a key role in the biochemical function of this protein from its sequence. PredictProtein and I-TASSER are among the most widely used public servers for structure prediction. According to PredictProtein analysis, the RstB2 monomer (gi|147671922) contains 126 aa, being 27 % of them
red smaller rectangles represent a-helices; ISIS method that identifies interacting residues from sequence; PROFbval a web-based interface for prediction of flexibility/rigidity in proteins; Ucon a method for the prediction of unstructured regions; MD META-disorder prediction method; predictions range the higher reliability values of the prediction are shown in red. b Alignment of the amino acid residues of three SSB proteins. Protein sequence of filamentous phage Ff, Pf3, and RstB2 was aligned according to ClustalW2 program. The conserved residues of the DNA-binding domain were shown (Color figure online)
Virus Genes
charged, 23.8 % polar, and 34.1 % hydrophobic. Many of the last ones are found surrounded by positively charged amino acids, the same as all SSB proteins [23]. This amino acidic composition is similar to that of SSB proteins of other filamentous vibriophages (Table 2). A further analysis of the amino acidic composition of RstB2 by composition profiler software (http://www.cprofiler.org/help. html) against the standard protein dataset (SwissProt51) revealed that RstB2 is a proline-rich protein with 8.7 %. According to PredictProtein server, the visual output of the methods predicting the protein structure and function of the RstB2 are depicted (Fig. 5). PROFacc predicted with a high level of reliability (scored C7, scale 0–9) that RstB2 protein has a high residue solvent accessibility (68.25 % of residues with more than 16 % of their surface exposed). The predicted regions with higher residue solvent accessibility are mainly located into the loops, the a-helix, and the C-terminus. Regions predicted as buried are mostly positioned in the predicted b-sheets. The secondary structure predicted (PROFsec) for RstB2 contains 49.2 % b-sheets (interspersed with 44.4 % loop regions and 6.3 % a-helix). It is noteworthy the presence of proline and glycine residues in loop regions, which are frequently found in b-turns, a type of secondary structure that promotes the formation of antiparallel b-sheets [24]. On the other hand, ISIS predicted some residues involved in external protein–protein interactions, being most of them clustered in the region from residue 45–64. The majority of these residues are located in predicted flexible and unstructured regions. It is generally thought that regions involved in protein–protein interactions with multiple partners are flexible and unstructured, in order to fix the different requirements for the interacting partners [25]. The predicted flexible residues in this protein, according to PROFbval, are widespread throughout the whole aminoacidic sequence and coincide with those ones, predicted as exposed. On the other hand, the unstructured regions in this protein predicted by Ucon are found from residue 21, being the largest from residues 86 to 95 and 113 to 126, just located the last one in the C-terminus. Some charged residues (K, D, E, Y), glycine, and proline but few hydrophobic ones are found in these regions, as described for this kind of regions [26] and other unstructured regions of DNA-binding proteins [15]. Moreover, MD predicted some signal for the presence of disordered regions just in N-terminus and from residues 55–73 and 85–92, but the higher output values are found in the C-terminus (the last 17 residues). To gain an insight into the structure of the RstB2 protein, we have investigated its three-dimensional structure by using the I-TASSER server. I-Tasser utilizes a robust metathreading approach to identify potential templates. I-Tasser identified as the best ranked template the structure
of the SSB protein from the phage Pf3 from Pseudomonas aeruginosa (1pfs) and yielded five low quality models, according to analysis with the Q-Mean software. Prediction of RstB2 structure with other modeling servers (LOMETS, MUSTER, QUARTS, SEGMER, 3D-Jigsaw, and Raptor) yielded no models or low quality models. Then, taking into account the identification of 1pfs as template, the amino acid sequence of RstB2 was compared with the amino acid sequence of the well-characterized SSB proteins Pf3 and Ff GVP (Fig. 5b), which have very similar three-dimensional structures [27]. The alignment revealed that the three proteins have 8 identical and 15 similar residues, showing the poor sequence conservation among SSB proteins. Interestingly, among them, there were found some conserved residues that coincide with those involved in the structural and biological function of the GVP of M13 phage [27]. Hydrophobic conserved amino acids (V4, I6, L32, V43, I47, V63, I78, and V84) in the GVP of M13 have their corresponding residues (I7, M9, L57, V81, F85, I100, L118, and I124) in the RstB2 protein. Furthermore, two positively charged amino acids (K7 and R82) in the GVP of M13 were conserved in the RstB2 protein (K10 and Q122). In addition, several residues involved in DNA binding in the GVP of M13 (G18, S20, G23, Y26, and L28) have been conserved in RstB2 protein (G23, S25, G28, Y31, and L53), and the Y41 of M13 is replaced by F79 in RstB2 protein.
Discussion In the current study, we have made further biochemical characterization of rRstB2, to identify putative structural determinants of the function of this protein. First, we determined the oligomeric status of rRstB2 protein and found that rRstB2 appeared to be monomeric at concentrations &1 mg ml-1, unlike SSB proteins of other filamentous phages, which occurs as dimmers (Ff gVp [28, 29], M13 gVp [29], Cf gVp [30], and pVVGJU protein [10]. The behavior of rRstB2 in solution is similar to that one of the monomer T4 gene 32 protein [31]. This result does not exclude the possibility of assembling in higher order structures under other physiological conditions or upon DNA binding, as has been shown for the monomeric SSB protein of Bacillus anthracis, which appears to form a transient tetrameric structure with the ssDNA [32]. On the other hand, the recombinant RstB2 protein has been previously shown to be capable of preferentially binding ssDNA in vitro and accordingly, it was proposed to be renamed pVCTX/ [7]. Here, we have shown that the kinetic of reaction in the protein–DNA complex formation with both ss and dsDNA is similar, which remained after a year of storing at
123
Virus Genes
-20 °C. These results, along with those ones, previously found by our team, suggest that RstB2 may be a noncanonical filamentous phage SSB protein of an unusual filamentous phage such as CTX/. Although CTX/ is a filamentous phage, it is known that its ssDNA, substrate of integration, must be folded into a stem-loop structure, thus creating a small region of duplex DNA, which is the target of the site-specific recombinases [4]. However, it is still unknown how and when the attP site folds inside the cell, even the identity of the factors influencing this process. On the other hand, the preferential in vitro binding of ssDNA by rRstB2 [7] supports its proposed role in protecting the CTX/ ssDNA for integration [8]. Moreover, dsDNA binding may be physiologically important during CTX/ integration by stabilizing the secondary structure (stem loop) of the substrate integration, in a way not yet identified. Besides, it is interesting to know that two distinct binding modes have been previously reported for the SSB protein of two well-studied models, bacteriophages T7, and T4. Firstly, they bind trough weak non-electrostatic interactions to the backbone of the dsDNA; secondly, they can bind strongly to exposed ssDNA regions through electrostatic and non-electrostatic interactions. Both proteins show strong preferential ssDNA binding and weak dsDNA binding, which are essential for their ability to search dsDNA in one dimension, just to find available ssDNA-binding sites at the replication fork. These properties are likely essential for the ability of these proteins to effectively act as part of the DNA replication process [33]. In a similar way, RstB2 might dislodge the Vibrio SSB protein coating the displaced ssDNA molecules, formed in the first cycles of the replication by the rolling circle mechanism (through its ssDNA affinity) and may promote the formation or stabilization of the secondary structure (through its dsDNA affinity), needed for integration. Binding of ss and dsDNA molecules by RstB2 may demand some structural features, such as a big enough DNA-binding surface to fit the smaller single-stranded DNA and the much wider duplex as well as a highly flexible structure, being capable of suffer a deformation, in order to accommodate the relatively inflexible duplex onto the binding surface. On the other hand, RstB2 may undergo structural transformations upon binding ssDNA, which may support its preferential binding to ssDNA. In addition, the bioinformatics results predicted that RstB2 has seven b-sheets and one a-helix interspersed with loop regions. This secondary structure is suggestive of an OB-fold that in the case of RstB2, the formation of this central conserved OB-fold may be favored by the proline and glycine residues, found in the loop regions, which may promote the formation of antiparallel b-sheets [24]. However, the three-dimensional structure prediction failed in
123
producing a high quality model, which may be related to the known difficulty for identifying the OB-fold by sequence analysis, due to the poor sequence conservation found. This is reinforced in the case of phage SSB proteins, whose structures vary greater that those ones from cellular organisms, according to the SCOP database. Nevertheless, some conserved amino acid residues were found (Fig. 5b), which could be involved in structural (hydrophobic core residues) and DNA-binding functions. Thus, these results suggest that some structural or functional similarity between RstB2 and the SSB protein of Pf3 phage could be considered. Furthermore, analysis by PROFbval, Ucon, and MD predicted that RstB2 protein has a disordered, highly flexible C-terminus with no regular secondary structure, as expected for this kind of regions. Moreover, it has a proline residue involved in mediating protein–protein interactions, similar to other prokaryotic SSBs [25, 31, 34, 35]. In fact, the charged and proline residues found could contribute to the high flexibility, predicted, and needed for that region. It is known that the disordered regions are flexible, dynamic and can be partially or completely extended in solution. It is noteworthy that the aromatic residue (phenylalanine) found in this flexible unstructured region, which could be critical in binding ssDNA. On the other hand, it has been experimentally shown that disordered regions are involved in molecular recognition, in particular DNA binding, in order to facilitate processes such as transcription, transposition, packing, repair, and replication [26]. Therefore, prediction of such regions would provide the first step in methods for functionally identifying relevant ones. Furthermore, there are some proteins that require a conformational change, previous to DNA binding where C-terminal regions are involved. For instance, the T7 gene 2.5 protein and T4 gene 32 protein, which have a C-terminus functioning as a two-way switch that coordinates and modulates the ssDNA-binding and the protein–protein interactions [25, 36]. This mechanism is used by abundant cellular proteins as a tool for the coordination of nucleic acids and protein–protein interactions within the macromolecular machinery that ensures the integrity of the genome and its replication. We consider that the C-terminus disordered and flexible found in RstB2 protein could be functionally important, as described previously for other SSBs. This region is enriched in charged and polar residues that favor the interaction with water and other residues (glycine and proline), which may alter structured helices or strands and have high net charges [37]. These residues could contribute, on one hand, to the required conformational change before binding DNA and on the other hand, to increase DNA–protein and protein–protein interactions that occur upon binding to ssDNA. It is known that the cooperative binding described for SSBs involves protein–protein interactions between the core of one ssDNA-
Virus Genes
bound protein and the N-domain of the adjacent bound protein [31].
Conclusions The characterization of rRstB2 revealed that this protein occurs as a monomeric species in solution and its C-terminal His-tail promoted oligomerization and increased its isoelectric point. Moreover, it was confirmed that although it preferentially binds ssDNA, it is also capable of interacting with dsDNA, following a similar kinetics. Like other SSBs, the protein is predicted to have a disordered, flexible, and exposed C-terminus that may be physiological relevant during the chromosomal integration of the CTX/ genome. The protein remained biologically active after a year of storage at -20 °C. Acknowledgments The authors would like to thank Prof. Celso Perez for his valuable assistance in English revision.
References 1. M.K. Waldor, J.J. Mekalanos, Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272, 1910–1914 (1996) 2. M.K. Waldor, E.J. Rubin, G.D. Pearson, H. Kimsey, J.J. Mekalanos, Regulation, replication, and integration functions of the Vibrio cholerae CTXphi are encoded by region RS2. Mol. Microbiol. 24, 917–926 (1997) 3. K.E. Huber, M.K. Waldor, Filamentous phage integration requires the host recombinases XerC and XerD. Nature 417, 656–659 (2002) 4. M.E. Val, M. Bouvier, J. Campos, D. Sherratt, F. Cornet, D. Mazel, F.X. Barre, The single-stranded genome of phage CTX is the form used for integration into the genome of Vibrio cholerae. Mol. Cell 19, 559–566 (2005) 5. B. Das, J. Bischerour, M.E. Val, F.X. Barre, Molecular keys of the tropism of integration of the cholera toxin phage. Proc Natl Acad Sci USA 107, 4377–4382 (2010) 6. B. Das, J. Bischerour, F.X. Barre, VGJphi integration and excision mechanisms contribute to the genetic diversity of Vibrio cholerae epidemic strains. Proc Natl Acad Sci USA 108, 2516–2521 (2011) 7. A. Falero, A. Caballero, B. Ferran, Y. Izquierdo, R. Fando, J. Campos, DNA binding proteins of the filamentous phages CTXphi and VGJphi of Vibrio cholerae. J. Bacteriol. 191, 5873–5876 (2009) 8. B. Das, J. Bischerour, F.X. Barre, Molecular mechanism of acquisition of the cholera toxin genes. Indian J. Med. Res. 133, 195–200 (2011) 9. R. Weistermeier, T. Naven, Proteomics in Practice. A Laboratory Manual of Proteome Analysis (Wiley-VCH, Weinheim, 2002) 10. A. Falero, A. Caballero, S. Trigueros, C. Perez, J. Campos, K. Marrero, R. Fando, Characterization of the single-stranded DNA binding protein pV(VGJPhi) of VGJPhi phage from Vibrio cholerae. Biochim. Biophys. Acta 2011, 1107–1112 (1814) 11. B. Rost, G. Yachdav, J. Liu, The PredictProtein server. Nucleic Acids Res. 32, W321–W326 (2004)
12. B. Rost, How to Use Protein 1D Structure Predicted by PROFphd (Humana, Totowa, 2005) 13. Y. Ofran, B. Rost, ISIS: interaction sites identified from sequence. Bioinformatics 23, e13–e16 (2007) 14. A. Schlessinger, B. Rost, Protein flexibility and rigidity predicted from sequence. Proteins 61, 115–126 (2005) 15. A. Schlessinger, M. Punta, B. Rost, Natively unstructured regions in proteins identified from contact predictions. Bioinformatics 23, 2376–2384 (2007) 16. A. Schlessinger, M. Punta, G. Yachdav, L. Kajan, B. Rost, Improved disorder prediction by combination of orthogonal approaches. PLoS ONE 4, e4433 (2009) 17. M. Amor-Mahjoub, J.P. Suppini, N. Gomez-Vrielyunck, M. Ladjimi, The effect of the hexahistidine-tag in the oligomerization of HSC70 constructs. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 844, 328–334 (2006) 18. P. Perron-Savard, G. De Crescenzo, H. Le Moual, Dimerization and DNA binding of the Salmonella enterica PhoP response regulator are phosphorylation independent. Microbiology 151, 3979–3987 (2005) 19. K.L. Cheung, J. Huen, Y. Kakihara, W.A. Houry, J. Ortega, Alternative oligomeric states of the yeast Rvb1/Rvb2 complex induced by histidine tags. J. Mol. Biol. 404, 478–492 (2010) 20. T.S. Kaoud, A.K. Devkota, R. Harris, M.S. Rana, O. Abramczyk, M. Warthaka, S. Lee, M.E. Girvin, A.F. Riggs, K.N. Dalby, Activated ERK2 is a monomer in vitro with or without divalent cations and when complexed to the cytoplasmic scaffold PEA-15. Biochemistry 50, 4568–4578 (2011) 21. H. Shi, Y. Zhang, G. Zhang, J. Guo, X. Zhang, H. Song, J. Lv, J. Gao, Y. Wang, L. Chen, Systematic functional comparative analysis of four single-stranded DNA-binding proteins and their affection on viral RNA metabolism. PLoS ONE 8, e55076 (2013) 22. J. Cheng, A.Z. Randall, M.J. Sweredoski, P. Baldi, SCRATCH: a protein structure and structural feature prediction server. Nucl. Acids Res. 33, W72–W76 (2005) 23. B. Marintcheva, A. Marintchev, G. Wagner, C.C. Richardson, Acidic C-terminal tail of the ssDNA-binding protein of bacteriophage T7 and ssDNA compete for the same binding surface. Proc. Natl. Acad. Sci. USA 105, 1855–1860 (2008) 24. S.R. Trevino, S. Schaefer, J.M. Scholtz, C.N. Pace, Increasing protein conformational stability by optimizing beta-turn sequence. J. Mol. Biol. 373, 211–218 (2007) 25. B. Marintcheva, S.M. Hamdan, S.J. Lee, C.C. Richardson, Essential residues in the C terminus of the bacteriophage T7 gene 2.5 single-stranded DNA-binding protein. J. Biol. Chem. 281, 25831–25840 (2006) 26. V.N. Uversky, C.J. Oldfield, A.K. Dunker, Showing your ID: intrinsic disorder as an ID for recognition, regulation and cell signaling. J. Mol. Recognit. 18, 343–384 (2005) 27. R.H. Folmer, M. Nilges, R.N. Konings, C.W. Hilbers, Solution structure of the single-stranded DNA binding protein of the filamentous Pseudomonas phage Pf3: similarity to other proteins binding to single-stranded nucleic acids. EMBO J. 14, 4132–4142 (1995) 28. T.C. Mou, M.C. Shen, T.C. Terwilliger, D.M. Gray, Binding and reversible denaturation of double-stranded DNA by Ff gene 5 protein. Biopolymers 70, 637–648 (2003) 29. N.C. Alma, B.J. Harmsen, J.H. van Boom, G. van der Marel, C.W. Hilbers, 1H NMR studies of the binding of bacteriophageM13-encoded gene-5 protein to oligo(deoxyadenylic acid)s of varying length. Eur. J. Biochem. 122, 319–326 (1982) 30. J.D. Wen, D.M. Gray, Selection of genomic sequences that bind tightly to Ff gene 5 protein: primer-free genomic SELEX. Nucl. Acids Res. 32, e182 (2004) 31. K. Pant, R.L. Karpel, I. Rouzina, M.C. Williams, Salt dependent binding of T4 gene 32 protein to single and double-stranded
123
Virus Genes DNA: single molecule force spectroscopy measurements. J. Mol. Biol. 349, 317–330 (2005) 32. E.E. Biswas-Fiss, J. Kukiratirat, S.B. Biswas, Thermodynamic analysis of DNA binding by a Bacillus single stranded DNA binding protein. BMC Biochem. 13, 10 (2012) 33. R.D. Shereda, A.G. Kozlov, T.M. Lohman, M.M. Cox, J.L. Keck, SSB as an organizer/mobilizer of genome maintenance complexes. Crit. Rev. Biochem. Mol. Biol. 43, 289–318 (2008) 34. T. Hollis, J.M. Stattel, D.S. Walther, C.C. Richardson, T. Ellenberger, Structure of the gene 2.5 protein, a single-stranded DNA binding protein encoded by bacteriophage T7. Proc. Natl. Acad. Sci. USA 98, 9557–9562 (2001)
123
35. S.N. Savvides, S. Raghunathan, K. Futterer, A.G. Kozlov, T.M. Lohman, G. Waksman, The C-terminal domain of full-length E. coli SSB is disordered even when bound to DNA. Protein Sci. 13, 1942–1947 (2004) 36. L. Shokri, I. Rouzina, M.C. Williams, Interaction of bacteriophage T4 and T7 single-stranded DNA-binding proteins with DNA. Phys. Biol. 6, 025002 (2009) 37. M. Fuxreiter, I. Simon, P. Friedrich, P. Tompa, Preformed structural elements feature in partner recognition by intrinsically unstructured proteins. J. Mol. Biol. 338, 1015–1026 (2004)
