SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Protein-primed DNA replication constitutes a strategy to initiate viral DNA synthesis in a variety of prokaryotic and eukaryotic organisms. Although the main function of viral terminal proteins (TPs) is to provide a free hydroxyl group to start initiation of DNA replication, there are compelling evidences that TPs can also play other biological roles. In the case of Bacillus subtilis bacteriophage ϕ29, the N-terminal domain of the TP organizes viral DNA replication at the bacterial nucleoid being essential for an efficient phage DNA replication, and it contains a nuclear localization signal (NLS) that is functional in eukaryotes. Here we provide information about the structural properties of the ϕ29 TP N-terminal domain, which possesses sequence-independent DNA-binding capacity, and dissect the amino acid residues important for its biological function. By mutating all the basic residues of the TP N-terminal domain we identify the amino acids responsible for its interaction with the B. subtilis genome, establishing a correlation between the capacity of DNA-binding and nucleoid localization of the protein. Significantly, these residues are important to recruit the DNA polymerase at the bacterial nucleoid and, subsequently, for an efficient phage DNA replication.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Protein-primed DNA replication constitutes one of the strategies used to initiate DNA synthesis of linear genomes. By this mechanism, the DNA polymerase initiates DNA replication using the hydroxyl group provided by a serine, threonine or tyrosine residue of a terminal protein (TP) as a primer, catalysing the formation of a phosphodiester bond between the hydroxyl group of the amino acid and the initiating nucleotide. Thus, the primer TP becomes covalently attached to the 5′ ends of the new DNA strand (parental TP). Protein-priming mechanisms are used by a wide variety of organisms such as prokaryotic and eukaryotic viruses (Salas, 1991; de Jong et al., 2003), and the presence of terminal proteins has also been described in viruses infecting Archaea (Bath et al., 2006; Peng et al., 2007), some Streptomyces spp. (Chang and Cohen, 1994), linear plasmids from bacteria, fungi and higher plants (Salas, 1991; Meinhardt et al., 1997; Chaconas and Chen, 2005), transposable elements (Kapitonov and Jurka, 2006) and mitochondrial DNA (Fricova et al., 2010).

The protein-priming mechanism of DNA replication has been studied in detail using the B. subtilis bacteriophage ϕ29 as a model system (Salas, 1991; 1999). Phage ϕ29 possesses a linear dsDNA genome of 19 285 bp with a TP covalently linked to each 5′ end. To initiate DNA replication, a heterodimer formed by the ϕ29 DNA polymerase and a free TP molecule (primer TP) recognizes the viral TP-containing DNA ends, which constitute the origins of replication (Blanco et al., 1987). Then, the DNA polymerase catalyses the incorporation of the first dAMP to the hydroxyl group provided by Ser232 of the primer TP (Blanco and Salas, 1984; Hermoso et al., 1985). After a transition step, the DNA polymerase dissociates and continues processive elongation from both DNA ends coupled to strand displacement to complete duplication of the parental strands (Salas, 1999).

The crystallographic structure of the ϕ29 DNA polymerase/TP heterodimer (Kamtekar et al., 2006) revealed that the ϕ29 TP (266 amino acids) has an elongated three-domain structure (see Fig. 1A). This structure consists of a disordered N-terminal domain (residues 1–73), an intermediate domain (residues 74–172) containing two long α-helices and a short β-turn that makes extensive contacts with the DNA polymerase and confers specificity to this interaction, and a C-terminal domain (residues 173–266) that is comprised of a four-helix bundle and contains the priming Ser232 residue.

figure

Figure 1. A. Crystallographic structure of ϕ29 TP. The TP N-terminal domain is modelled as two green alpha-helices (adapted from Kamtekar et al., 2006).

B. Multiple alignment of the TP N-terminal amino acid sequences of B. subtilis phages ϕ29, PZA, Nf, B103 and GA1, in order of sequence identity with ϕ29 TP N-terminal domain (numbers at the right of the sequences correspond to the percentages relative to the ϕ29 TP N-terminal domain). Red boxes enclose residues conserved in all the sequences compared, green boxes enclose residues that are not identical in all sequences but conserve charge, and blue boxes enclose residues conserved in all sequences except that of phage GA1. ϕ29 TP N-terminal domain residues that have been subjected to mutagenesis are indicated with its corresponding position number in the protein sequence. Black boxes enclose the residues predicted to form alpha-helices in the TP N-terminal domain of the ϕ29-like phages. The predicted secondary structure of the ϕ29 TP N-terminal domain using Phyre2 software is represented as green helices connected by a loop (black line).

C. Far-UV CD spectrum of the ϕ29 TP N-terminal domain at 25°C. The recorded spectrum has two minima, one at 208 nm and other at 222 nm, typical of a protein folded mainly in a α-helical structure. The estimated secondary structure content was 60% α-helix and 40% random coil.

Download figure to PowerPoint

Previous work showed that ϕ29 TP localizes at the host nucleoid independently of other phage-encoded proteins and recruits the phage DNA polymerase to the active site of viral DNA replication (Muñoz-Espín et al., 2010). The TP N-terminal domain, responsible for TP nucleoid localization, was shown to have sequence-independent DNA-binding capacity in vitro and to be necessary for an efficient viral DNA replication in vivo (Muñoz-Espín et al., 2010). Also, it has been recently shown that the TP N-terminal domain contains a functional eukaryotic nuclear localization signal (NLS) within residues 1–37 (Redrejo-Rodríguez et al., 2012). Here, we provide insights about the secondary structure of the TP N-terminal domain and by constructing single-, double- and triple-point mutants, we determine the residues important for nucleoid localization and DNA-binding functions. In addition, we show that the residues involved in these functions are important for the recruitment of the DNA polymerase at the bacterial nucleoid and for an efficient viral DNA replication.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Structural properties of the ϕ29 TP N-terminal domain

The crystallographic structure of the ϕ29 DNA polymerase/TP heterodimer showed a disordered TP N-terminal domain (Kamtekar et al., 2006) (Fig. 1A) within the crystal lattice, and although it was modelled as two poly-alanine helices, it remains unsolved. To obtain further insights about the structural properties of the ϕ29 TP N-terminal domain we carried out a secondary structure prediction search using Phyre2 (protein homology/analogy recognition engine) (Kelley and Sternberg, 2009) and the iterative threading assembly refinement (I-TASSER) servers (Zhang, 2008; Roy et al., 2010). Both Phyre2 and I-TASSER servers predicted two regions of α-helix connected by a loop, which are conserved among other ϕ29-related phages (Fig. 1B). The helical folding spans ϕ29 TP residues 14–34 and 48–70 in the case of the Phyre2 software and residues 14–34 and 48–65 in the case of the I-TASSER server.

To determine experimentally the secondary structure content of the ϕ29 TP N-terminal domain we used a TP fragment (TP-Nt) spanning amino acids 1–73 (Muñoz-Espín et al., 2010) and performed far-UV circular dichroism (CD) spectroscopic analyses (Fig. 1C). The CD spectrum of the TP N-terminal domain revealed a helical content of about 60%, very similar to that derived from secondary structure predictions (about 60% and 53% using Phyre2 and I-TASSER servers respectively), and about 40% of random coil. This result shows that the TP N-terminal domain has a substantial content in α-helix and it is therefore consistent with the prediction of two helical segments connected by a disordered loop.

TP N-terminal domain residues involved in nucleoid localization

The subcellular localization of the ϕ29 TP at the bacterial nucleoid has been proposed to depend on the binding of the TP N-terminal domain to genomic DNA (Muñoz-Espín et al., 2010). Since the capacity of the TP N-terminal domain to interact with DNA is not sequence-specific (Muñoz-Espín et al., 2010), electrostatic bonds between positively charged amino acids and DNA phosphate groups may play a role in such interaction. The N-terminal domain of the ϕ29 TP comprises 73 residues out of which 23% have a basic character. Figure 1B shows a comparison between the sequence of the ϕ29 TP N-terminal domain and that of the ϕ29-like bacteriophages PZA, Nf, B103 and GA1. The sequence alignment shows that most of the basic residues are identical among these phages and, when non-identical, some of them conserve the positive charge. To investigate a possible role of these residues in ϕ29 TP nucleoid localization we replaced each of them independently by alanine, thus eliminating the electrostatic character of these amino acids. A total of seventeen point mutants were designed, as well as a double and a triple mutant comprising two clusters of lysine residues (positions 25 and 27, and 32–34). All but four of the basic residues mutated lied in the predicted α-helix portion of the TP N-terminal domain (Fig. 1B). To analyse the subcellular localization of these mutant TPs, we engineered B. subtilis strains containing an inducible fusion of yfp to the ϕ29 wild-type or mutant variants of gene 3, coding for the TP. All mutant proteins were stable and most of them were expressed to levels comparable to those of the wild-type TP (Fig. S1). Figures 2A and S2 show the subcellular localization of these YFP fusions in B. subtilis cells. As internal controls, wild-type TP (YFP-wt TP) colocalized with the bacterial nucleoid, whereas YFP alone and a TP N-terminal deletion mutant (YFP-ΔN73) localized throughout the cell (Muñoz-Espín et al., 2010). Merged images of the YFP and DAPI fluorescent signals revealed a disrupted nucleoid localization of mutant protein YFP-K25A/K27A, which was distributed uniformly along the entire length of the bacterial cell. The single mutant YFP-K25A displayed a wild-type phenotype, colocalizing with the bacterial nucleoid, whereas the YFP-K27A mutant showed an intermediate phenotype exhibiting a fluorescent signal spread along the cell, but with some accumulation at the nucleoid. The rest of mutant TPs displayed nucleoid localization analogous to that of the wild-type protein (Figs 2A and S2).

figure

Figure 2. Subcellular localization of ϕ29 wild-type TP and N-terminal mutant TPs fusioned to YFP.

A and B. YFP, DAPI, phase contrast and merged images of B. subtilis cells expressing inducible YFP-TP fusion proteins analysed 30 min after xylose addition. As a control, YFP localization is shown. For clarity, YFP and DAPI signals are false coloured green and red respectively.

Download figure to PowerPoint

To estimate the minimum TP sequence necessary to confer bacterial nucleoid localization, we constructed YFP fusions of TP N-terminal fragments of different lengths and TP N-terminally deleted variants, which in most cases were expressed similarly to wild-type TP (Figs S3 and S1 respectively), and studied their subcellular localization in B. subtilis cells. TP N-terminal fragments 1–50 (YFP-NT50) and 1–60 (YFP-NT60) were not sufficient to target the bacterial nucleoid (Fig. S4). However, fragment 1–70 (YFP-NT70) gave a localization pattern with an accumulation of the fluorescent signal at the nucleoid close to that of the complete N-terminal domain (YFP-NT73). On the other hand, the TP N-terminally deleted variants YFP-ΔN10 and YFP-ΔN15 retained the TP wild-type localization at the bacterial nucleoid, whereas the YFP-ΔN20 truncated variant showed a dispersed pattern along the cell (Fig. 2B). This result may suggest the requirement of a correct folding of the TP N-terminal domain to localize at the bacterial nucleoid as it could be inferred from the prediction that shows that an α-helix would be preserved both in the ΔN10 and ΔN15 variants but altered in the ΔN20 truncated TP fusion (Fig. 1B). To further dissect the minimum functional TP N-terminal portion, we constructed YFP fusions of TP N-terminal fragments comprising amino acids 16–70 (YFP-NT16-70) and 16–73 (YFP-NT16-73) and studied their localization patterns in B. subtilis live cells. Figure S4 shows that both YFP fusions displayed a fluorescent signal throughout the entire length of the cell. Altogether, we conclude from these experiments that the minimum TP sequence necessary to provide a functional domain conferring bacterial nucleoid localization spans residues 1–70, and that only the first fifteen residues comprising the N-terminal domain are dispensable in a full-length TP version.

TP N-terminal domain residues involved in DNA-binding

To determine if the partial or complete loss of nucleoid localization of K27A and K25A/K27A mutant TPs, respectively, was due to a defect of these proteins in DNA-binding, we performed electrophoretic mobility shift assays (EMSA) using a 216 bp dsDNA fragment corresponding to gene yshC of the B. subtilis genome. For this, we purified a set of TP mutant proteins (see Supporting information for details): R19A, K25A, K27A, double K25A/K27A, triple K32A/K33A/K34A and ΔNt (TP truncated N-terminal domain variant spanning residues 74–266). Figure 3 shows that wild-type TP produced a complete shift of the DNA fragment at a concentration of 12.4 nM. At this concentration only the K25A and R19A mutants displayed a comparable DNA-binding capacity. However, at lower concentrations, the binding capacity of mutant R19A is reduced. K32A/K33A/K34A mutant protein displayed some decrease in DNA-binding, although a clear shift was observed at the highest concentration assayed. Importantly, the DNA-binding capacity of mutants K27A and K25A/K27A was severely affected as only a faint band, that was more evident in the case of the K27A mutant, was shifted at the highest concentration tested. In agreement with previous results, the ΔNt-truncated version did not produce a band shift even at the highest protein concentration analysed (Muñoz-Espín et al., 2010).

figure

Figure 3. In vitro DNA-binding of ϕ29 N-terminal mutant TPs compared with wild-type TP. Increasing amounts of wild-type and mutant TPs were incubated with 1 nM 32P end-labelled dsDNA fragment of 216 bp corresponding to the B. subtilisyshC gene and analysed by EMSA. The experiment was done in triplicate and densitometric quantification of the retarded bands is shown in the bar graph. Columns represent mean and bars standard deviation.

Download figure to PowerPoint

To investigate the importance of residue K27 and the combination of residues K25 and K27 in DNA-binding in vivo, we performed cross-linked chromatin immunoprecipitation (X-ChiP) assays. For this, B. subtilis cells expressing YFP, YFP-wt TP and YFP-K25A, YFP-K27A and YFP-K25A/K27A mutant proteins were grown in LB medium, subjected to protein induction during exponential growth and formaldehyde cross-linking (see Experimental procedures for details). Then, the binding of the fusion proteins to four regions ranging from 104 to 200 bp spread along the B. subtilis genome (corresponding to gyrB, ftsZ, mreB and ung genes: 0°, 136°, 244° and 333° respectively), was quantified by real-time PCR. Figure S5 shows that, under these conditions, the expression of the wild-type and the mutant proteins analysed by Western blot was similar. Figure 4 shows the immunoprecipitation coefficient (IC, see Experimental procedures) of YFP, YFP-wt TP and YFP-K25A, YFP-K27A and YFP-K25A/K27A mutant fusion proteins for each one of the four regions analysed. Cells expressing YFP had IC values of 1–2% of the wild-type TP and the fusion of YFP to wild-type TP did not preclude binding of the TP to the DNA in vivo. According to a sequence-independent DNA-binding mode, wild-type TP bound to the four regions of the B. subtilis genome analysed in a very similar way (Fig. S6). Mutation of residue K25 gave IC values ranging between 45% and 55% relative to those of the wild-type TP and, in agreement with the results obtained by EMSA, the mutation of residues K27 and K25/K27 strongly affected the DNA-binding capacity of the TP in vivo, giving IC values of 4–7% and 8–11%, respectively, of the wild-type TP.

figure

Figure 4. Binding of ϕ29 wild-type TP and N-terminal mutant TPs K25A, K27A and K25A/K27A fusioned to YFP to B. subtilis DNA regions gyrB, ftsZ, mreB and ungin vivo. B. subtilis strains expressing the indicated proteins were grown in LB medium at 30°C, cross-linked with formaldehyde and processed as described in Experimental procedures. YFP binding was carried out as a control. Protein binding is expressed as Immunoprecipitation Coefficient (IC, see Experimental procedures) relative to the wild-type TP. The experiment was done in triplicate. Columns represent mean and bars standard deviation.

Download figure to PowerPoint

Effect of TP N-terminal domain mutations in DNA polymerase recruitment at the bacterial nucleoid and in viral DNA replication in vivo

Since ϕ29 TP recruits the viral DNA polymerase to the bacterial nucleoid (Muñoz-Espín et al., 2010), we analysed the effect of the above TP N-terminal mutations on this functional property. For this, B. subtilis strains coexpressing YFP-DNA polymerase and CFP-wt or mutant TP fusions under xylose and IPTG-inducible promoters, respectively, were constructed. The level of expression of the different TPs was very similar between them and in all cases the DNA polymerase was expressed similarly (Fig. S7). Figure 5 shows that the strains expressing CFP-wt TP and CFP-R19A TP, CFP-K25A TP and CFP-K32A/K33A/K34A TP mutants displayed a YFP-DNA polymerase fluorescent signal mainly localized at the bacterial nucleoid. In contrast, strains expressing CFP-K25A/K27A TP and CFP-ΔNt TP mutants exhibited a YFP-DNA polymerase fluorescent signal distributed throughout the entire length of the cell, similarly to the pattern obtained when CFP alone was coexpressed with a YFP-DNA polymerase fusion. The strain expressing the CFP-K27A TP mutant fusion showed an intermediate phenotype of YFP-DNA polymerase localization, being spread along the cell but with some accumulation of the fluorescent signal at the nucleoid. These results show that TP mutant K27A and to a higher extent TP mutant K25A/K27A are impaired in TP-mediated DNA polymerase recruitment at the bacterial nucleoid.

figure

Figure 5. Effect of TP N-terminal domain mutations in the recruitment of ϕ29 DNA polymerase (p2) at the bacterial nucleoid. CFP, YFP, phase and merged images of B. subtilis cells coexpressing ϕ29 wild-type TP or N-terminal mutant TPs fusioned to CFP and ϕ29 DNA polymerase fusioned to YFP 30 min after IPTG and xylose addition. For clarity, CFP and YFP signals are false-coloured green and red respectively.

Download figure to PowerPoint

The TP N-terminal domain has been shown to be required for an efficient ϕ29 DNA replication in vivo (Muñoz-Espín et al., 2010). To study if the mutations introduced in the TP N-terminal domain have an effect in ϕ29 DNA replication in live cells, we performed complementation experiments using a sus3(91) mutant phage, unable to synthesize TP. Accumulation of viral DNA in strains expressing wild-type TP or R19A, K25A, K27A, K25A/K27A, K32A/K33A/K34A and ΔNt mutant TPs fusioned to YFP, infected with sus3(91) mutant phage, was analysed by agarose gel electrophoresis and real-time PCR (Fig. 6A and B). Western blot analyses (Fig. S8) showed that the amount of protein produced in the above strains was comparable. As shown in Fig. 6, YFP-R19A, YFP-K25A and YFP-K32A/K33A/K34A mutant proteins provided in trans complemented the sus3(91) infection, as the viral DNA accumulation in the strains producing these proteins was similar to that of cells producing YFP-wt TP. As an internal control, when YFP-ΔNt TP was expressed, hardly any accumulation of viral DNA was detected at 40 min post-infection. Importantly, viral DNA synthesis was substantially reduced in strains producing YFP-K27A and YFP-K25A/K27A mutant TPs.

figure

Figure 6. Complementation experiments of B. subtilis strains expressing ϕ29 wild-type TP or N-terminal mutant TPs fusioned to YFP infected with mutant phage sus3(91).

A. Agarose gel electrophoresis showing the viral DNA accumulation in the strains indicated above at the indicated times post-infection. Arrows indicate positions of B. subtilis and ϕ29 DNA.

B. The amount of intracellular viral DNA accumulated was analysed by real-time PCR. The experiment was carried out in triplicate. Bars represent Standard Error of the Mean values.

Download figure to PowerPoint

Altogether, we conclude that the substitution of TP residue K27 and to a higher extent the double substitution of residues K25 and K27 affect the functional properties of the ϕ29 TP: bacterial DNA-binding capacity and nucleoid localization, DNA polymerase recruitment, and efficient ϕ29 DNA replication.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Both eukaryotic and prokaryotic viruses are able to take advantage of the organizing structures of their hosts to compartmentalize fundamental processes such as transcription, replication and packaging (Schaack et al., 1990; Novoa et al., 2005; Radtke et al., 2006; Muñoz-Espín et al., 2010; Netherton and Wileman, 2011). The strategy of compartmentalizing certain viral processes inside the host cell allows concentrating viral components in a specific place, thus increasing the efficiency of the process. In the case of eukaryotic adenoviruses, which replicate their genomes by a protein-priming mechanism, the nuclear matrix is thought to function as a scaffold where the different replication proteins are recruited (de Jong and van der Vliet, 1999). The anchoring of the adenovirus TP to the nuclear matrix has been shown to be important for an efficient viral DNA replication and transcription (Schaack et al., 1990).

Bacteriophages share a strategy common to some eukaryotic viruses, as they frequently direct their genomes to the bacterial nucleoid to be transcribed and replicated. In this sense, it has been shown recently that the Pseudomonas chlororaphis phage 201ϕ2-1 encodes its own tubulin-like protein (PhuZ) that would position viral DNA at mid-cell to allow viral genomes to be efficiently replicated and/or packaged into the capsids (Kraemer et al., 2012). In the case of phage ϕ29, it was shown that the TP directs the viral DNA replication machinery at the bacterial nucleoid at early stages of the infection. The TP N-terminal domain has been suggested to be responsible for this process by means of a sequence-independent DNA-binding capacity (Muñoz-Espín et al., 2010). Although interaction with host nucleoid protein(s) cannot be ruled out, the ϕ29 TP does not interact with the nucleoid proteins HBsu, Noc, SMC and ScpA. Besides, it also localizes at the bacterial nucleoid when expressed in a distantly related bacterium such as Escherichia coli (Muñoz-Espín et al., 2010). Because non-specific DNA-binding usually involves basic residues, we have replaced all arginine and lysine residues of the TP N-terminal domain by alanine and assessed their role in DNA-binding. We report here a correlation between DNA-binding and nucleoid localization. Substitution of residues K27 and K25/K27 notably reduces the TP binding to DNA and, as a consequence, its nucleoid localization. However, the smaller reduction of the DNA-binding capacity of TP mutants R19A and K32A/K33A/K34A was not sufficient to impair the association with the bacterial chromosome, as they displayed a comparable nucleoid localization to the wild-type TP. Importantly, K27A and K25A/K27A TP mutants, affected in nucleoid localization, were impaired in recruiting the DNA polymerase to the bacterial nucleoid and did not complement the infection with a sus3(91) mutant phage, being defective in providing an efficient viral DNA replication in vivo. It is worth to mention that, when expressed in the non-host bacterium E. coli, TP N-terminal mutants R19A and K32/33/34A fusioned to YFP localized at the nucleoid in the same manner as the wild-type TP did, whereas K27A and K25/27A mutant TPs were found to localize in a spotted pattern at the nucleoid (Redrejo-Rodríguez et al., 2013).

In addition to the specific residues mediating TP DNA-binding, we gained insight into the secondary structural properties of the TP N-terminal domain and obtained evidence that it has a substantial content in α-helix, which is consistent with the prediction of two helical segments connected by a disordered loop. It was previously shown that the TP deletion mutant ΔN20 has strongly reduced capacity to bind DNA (Zaballos and Salas, 1989) and, accordingly, our results show that it has impaired nucleoid localization in B. subtilis. Besides, the minimal fragment sufficient for nucleoid localization was shown to be that corresponding to residues 1–70, in accordance with the view that a folded TP N-terminal domain is essential to maintain the DNA-binding capacity of the protein and, as a consequence, its nucleoid association.

Phage ϕ29 is a virulent phage that promotes a rapid lysis of B. subtilis cells after infection. As the presence of the parental TP at the ends of the ϕ29 genome (TP-DNA) precludes its integration in the host chromosome, anchoring of the TPs to the bacterial nucleoid would be an elegant strategy to segregate viral DNA in conjunction with B. subtilis DNA. The strategy of maintaining and partitioning extrachromosomal viral genomes by tethering them to cellular chromosomes is common among diverse eukaryotic DNA viruses like the gamma herpesviruses Epstein-Barr virus (EBV), Kaposi's sarcoma associated Herpesvirus (KSHV), Herpesvirus saimiri (HVS) and murine gamma herpesvirus-68 (MHV-68). These viruses encode DNA-binding proteins that bind specifically to repeated sites in the viral DNA and tether the genome to the cellular mitotic chromosomes (for reviews see McBride, 2008; McBride et al., 2012; Knipe et al., 2013). By analogy, phage ϕ29 TP could be directing and attaching the viral genome to the host nucleoid to take advantage of the chromosome dynamics. As long as the B. subtilis chromosome is replicated during vegetative growth, the two newly synthesized DNA copies are translocated to the cell poles in a helix-like structure (Berlatzky et al., 2008). Since ϕ29 DNA replication is relocalized to the bacterial membrane at late infection times in a helical configuration (Muñoz-Espín et al., 2009), it has been suggested that ϕ29 might use the motor-like force that provides bacterial chromosome segregation to the future daughter cells to redistribute its replication machinery at peripheral sites (Muñoz-Espín et al., 2012).

Besides the essential role of priming DNA replication, TPs can perform additional functions like nuclear matrix attachment of viral DNA (Schaack et al., 1990), enhancement of transcription (Schaack et al., 1990), transfection (Hirokawa, 1972; Ronda et al., 1983; Porter and Dyall-Smith, 2008), DNA packaging (Bjornsti et al., 1982) and nuclear targeting of prokaryotic TPs in eukaryotes (Tsai et al., 2008). Interestingly, and similarly to several TPs from Streptomyces spp. (Tsai et al., 2008), it has been shown recently that ϕ29 TP possesses a functional NLS within its N-terminal domain that directs the phage TP to the nucleus of eukaryotic cells, suggesting a possible role of the ϕ29 TP in horizontal gene transfer between prokaryotes and eukaryotes (Redrejo-Rodríguez et al., 2012). Most of the known NLSs and DNA-binding domains contain a high proportion of basic residues and it has been shown that, in many cases, both domains overlap within a given protein (LaCasse and Lefebvre, 1995; Cokol et al., 2000). In eukaryotic cells, the ϕ29 TP N-terminal fragment 1–37 was sufficient to localize at the nucleus, whereas the fragment necessary to localize at the bacterial nucleoid in B. subtilis is that corresponding to residues 1–70, longer than that required for the NLS function. Taking into account the secondary structure predictions, it appears that the DNA-binding capacity of the TP N-terminal domain is rather more stringent in terms of protein folding than the capacity of nuclear localization, which is usually more dependent on the protein primary sequence. It is however important to note that the TP N-terminal domain mutation K27A and the double mutation K25A/K27A, shown to affect DNA-binding and nucleoid localization, have also an effect in the NLS function, as these mutations also prevented TP nuclear localization in eukaryotic cells (Redrejo-Rodríguez et al., 2012). In this sense, it has been hypothesized that some NLSs could have evolved from fragments of DNA-binding domains enriched in basic residues to localize proteins at the eukaryotic nucleus, a novel compartment not existing in prokaryotic, preceding cells (Cokol et al., 2000).

We report here the first in vivo direct evidence of the sequence-independent DNA-binding capacity of the ϕ29 TP to the B. subtilis genome, and gain insight into the secondary structure of the TP N-terminal domain, responsible for bacterial nucleoid targeting. We also determine the TP residues important for nucleoid localization and DNA-binding functions. Moreover, we show that the residues involved in these functions are important for the recruitment of the DNA polymerase at the bacterial nucleoid and for an efficient viral DNA replication.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Phages, bacterial strains and growth conditions

Bacterial strains and phages used are listed in Tables S1 and S2 respectively. E. coli strain XL1-Blue, used for cloning, was grown in Luria–Bertani (LB) medium containing ampicillin (100 μg ml−1). B. subtilis strains were grown at 37°C in LB medium containing 5 mM MgSO4 and supplemented with spectinomycin (100 μg ml−1), erythromycin (1 μg ml−1) or kanamycin (5 μg ml−1), when required. Overnight cultures were diluted 1:100 in fresh medium and incubated at 37°C for 2–3 h to re-establish exponential growth before induction of fusion proteins. Ectopically expressed proteins were induced by addition of 0.5% (w/v) xylose or 1 mM IPTG (final concentration). Plasmids used are listed in Table S4.

Fluorescence microscopy

For live cell imaging, cultures were grown at 37°C to exponential phase and expression of YFP or CFP fusions induced with 0.5% (w/v) xylose or 1 mM IPTG (final concentration), respectively, at an OD600 of 0.3. Thirty minutes after induction, DAPI was added to the medium to a final concentration of 1 μg ml−1 and cells were immobilized on microscope slides covered with a thin film of 1% (w/v) agarose in water. Image acquisition was performed using a C9100-02 CCD camera (Hamamatsu) attached to a Zeiss Axiovert 200 M microscope. Images were processed using ImageJ software (Rasband, 1997 –2011).

Electrophoretic mobility shift assay

A 216 bp DNA fragment corresponding to the B. subtilis yshC gene was amplified by PCR using genomic B. subtilis DNA as template and primers yshC-R and yshC-L (Table S3). The PCR product was purified from agarose gel (Gel Extraction Kit, Qiagen) and 5′-labelled with [γ-32P] ATP and T4 polynucleotide kinase. The incubation mixture contained, in a final volume of 20 μl, 50 mM Tris-HCl, pH 7.5, 1 mM DTT, 4% (v/v) glycerol, 0.1 mg ml−1 bovine serum albumin, 1 nM of the 5′-labelled DNA and the indicated concentrations of the corresponding protein. After incubation for 5 min at 4°C, the samples were subjected to electrophoresis in 6% (w/v) polyacrylamide gels containing 40 mM Tris-acetate, pH 7.5, and 1 mM EDTA, and run at 4°C in the same buffer at 25 mA for 1.5 h. After autoradiography, the TP complexed with dsDNA was detected as a mobility shift in the migrating position of the labelled DNA and quantified by Quantity One software.

Formaldehyde cross-linked chromatin immunoprecipitation

Cultures were grown at 30°C to an OD600 of 0.4–0.45. Then, protein fusions were induced by the addition of 0.5% xylose (w/v) (final concentration). One hour post-induction, 1 ml of culture was withdrawn for Western blot analysis and the rest (20 ml) was treated with 1% formaldehyde for 5 min. X-ChiP was performed essentially as described (González-Huici et al., 2004) with slight modifications. Before phenol:chloroform extraction, DNA samples were treated with RNase A (final concentration 50 μg ml−1) for 15 min and Proteinase K (final concentration 50 μg ml−1) for 3 h at 37°C. Then, samples were subjected to ethanol precipitation and resuspended in 30 μl of DNase-free water. qPCR reactions were performed on a Bio-Rad CFX 384 instrument in a 10 μl final reaction volume with 4 μl of a 1/40 dilution of DNA sample, 1 μl of each forward and reverse primer at 2.5 μM, and 5 μl of Power Sybr Green reaction mix (Applied Biosystems). Cycling parameters were as follows: 10 min at 95°C followed by 40 two-step cycles at 95°C for 15 s + 64°C for 60 s and a melting curve analysis at 95°C for 15 s + 60°C for 15 s + 95°C for 15 s. TP binding was expressed as the immunoprecipitation coefficient (IC); IC = [(αTP − pi)/T], where T is total DNA, αTP the DNA immunoprecipitated with specific serum against TP and pi the DNA immunoprecipitated with pre-immune serum. In the case of the quantification of the immunoprecipitated copies of each B. subtilis region, a standard curve constructed with known amounts of the B. subtilis genome was used.

Analysis of viral DNA synthesis by gel electrophoresis and real time PCR

Analysis of viral DNA synthesis in vivo was carried out as described (Bravo et al., 1994). Aliquots of 1 ml of B. subtilis cultures were withdrawn at the indicated times after infection and total intracellular DNA was isolated and analysed in 0.6% agarose gels. 1/50 dilutions of the total DNA sample were analysed by real-time PCR essentially as described (Holguera et al., 2012). The primer sets R-25 and R-OUT-SUPER were used to amplify a 297 bp fragment corresponding to the right end of the ϕ29 genome (Table S3). The data obtained were interpolated to a standard curve constructed with known amounts of phage DNA.

CD spectroscopy

CD measurements were carried out using a Jasco-600 spectropolarimeter equipped with a NESLAB RTE-100 temperature control unit interfaced to a computer. The recorded far-UV spectra were the average of four scans obtained at 25°C at a rate of 50 nm min−1, a response time of 2 s, and a bandwidth of 1 nm. TP-Nt samples were prepared in 25 mM phosphate buffer, pH 7.5, and 200 mM NaCl at a concentration of 20 μM. Data were analysed using CCA (Convex Constraint Analysis) program (Perczel et al., 1991).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We are grateful to Laurentino Villar for purification of TP N-terminal mutant proteins and to Miguel Ángel Fuertes for his help with circular dichroism experiments. This investigation was supported by grants BFU2011-23645 from the Spanish Ministry of Economy and Competitiveness and Consolider-Ingenio CSD2007-00015 from the Spanish Ministry of Science and Innovation to MS and by an Institutional grant from Fundación Ramón Areces to the Centro de Biología Molecular ‘Severo Ochoa’. DM-E and MR-R were holders of a Consolider-Ingenio contract. IH was holder of a FPU fellowship from the Spanish Ministry of Education. All authors declared that they have no conflicts of interest.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
mmi12456-sup-0001-si.pdf11009KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.