The priB Gene of Klebsiella pneumoniae Encodes a 104-Amino Acid Protein That Is Similar in Structure and Function to Escherichia coli PriB

Primosome protein PriB is a single-stranded DNA-binding protein that serves as an accessory factor for PriA helicase-catalyzed origin-independent reinitiation of DNA replication in bacteria. A recent report describes the identification of a novel PriB protein in Klebsiella pneumoniae that is significantly shorter than most sequenced PriB homologs. The K. pneumoniae PriB protein is proposed to comprise 55 amino acid residues, in contrast to E. coli PriB which comprises 104 amino acid residues and has a length that is typical of most sequenced PriB homologs. Here, we report results of a sequence analysis that suggests that the priB gene of K. pneumoniae encodes a 104-amino acid PriB protein, akin to its E. coli counterpart. Furthermore, we have cloned the K. pneumoniae priB gene and purified the 104-amino acid K. pneumoniae PriB protein. Gel filtration experiments reveal that the K. pneumoniae PriB protein is a dimer, and equilibrium DNA binding experiments demonstrate that K. pneumoniae PriB's single-stranded DNA-binding activity is similar to that of E. coli PriB. These results indicate that the PriB homolog of K. pneumoniae is similar in structure and in function to that of E. coli.


Introduction
The survival of cellular organisms depends on complete and faithful duplication of their genetic material. Throughout the life of a cell, the process of DNA replication is challenged by environmental and chemical factors that can bring about damage to the DNA, which can disrupt the DNA replication machinery (replisome) [1]. Since failure to replicate the genome can result in cell death, microorganisms have adapted to these challenges by developing various mechanisms to recognize and repair DNA damage and ensure complete replication of their genetic information [2,3]. In bacteria, DNA replication restart pathways facilitate reactivation of replisomes that have been disrupted following encounters with DNA damage [4].
E. coli has proved to be an excellent model organism to investigate bacterial DNA replication restart pathways. In E. coli, DNA replication restart is catalyzed by primosome proteins, including PriA, PriB, PriC, DnaT, and DnaG, that collectively facilitate reloading of the replisome to allow DNA replication to continue [4]. The function of these primosome proteins involves coordinated protein and nucleic acid binding within a large nucleoprotein complex called the DNA replication restart primosome. PriA helicase is the initiator protein that binds to a repaired DNA replication fork and unwinds double-stranded DNA at the fork to produce a short tract of single-stranded DNA (ssDNA) [5,6,7]. PriB binds to PriA, stabilizes PriA on the DNA, and stimulates its helicase-activity [6,8]. The PriA:PriB:DNA ternary complex recruits DnaT to the DNA, which could lead to release of ssDNA from PriB [6,9]. The replicative helicase, DnaB/ C, is recruited to the fork where it unwinds the parental duplex DNA to stimulate priming by DnaG and reloading of the replicative polymerase.
Although studies using E. coli have revealed much about the mechanism of DNA replication restart, it cannot necessarily be uniformly applied to all prokaryotes because some organisms do not encode the full complement of primosome protein genes. Genome sequencing projects have revealed that priA genes are highly conserved among sequenced bacterial genomes, but priB, priC, and dnaT genes are not. The absence of one or more of these primosome genes from bacterial genomes suggests that there might be mechanistic differences in DNA replication restart pathways across diverse bacterial species.
In accordance with this hypothesis, Hsieh and Huang recently reported the identification of a novel PriB protein in Klebsiella pneumoniae [10]. According to their study, the K. pneumoniae PriB protein is only 55 amino acids in length, which is considerably shorter than E. coli PriB. The sequence of K. pneumoniae PriB that appears to be missing is analogous to the amino terminal region of E. coli PriB and includes amino acid residues important for dimerization and DNA binding. The authors also report that PriB proteins from Pectobacterium carotovorum, Yersinia ruckeri, and Salmonella enterica have considerably shorter amino acid sequences compared to E. coli PriB. The implications are that the PriB homologs from these bacterial species must be different in structure and in function from the well-studied E. coli PriB [10].
Here, we report that the PriB protein of K. pneumoniae is a fulllength PriB homolog whose sequence is the same length as E. coli PriB. Our sequence analysis of the other bacterial PriB proteins that have been proposed to be missing amino-terminal sequences reveals that they, too, are full-length PriB homologs whose lengths are comparable to E. coli PriB. Furthermore, we have cloned the full-length priB gene from K. pneumoniae, overexpressed and purified the recombinant K. pneumoniae PriB protein, and examined its quaternary structure and DNA binding activity. Our results indicate that the structure and function of K. pneumoniae PriB are highly similar to that of E. coli PriB. Thus, K. pneumoniae PriB does not likely represent a novel PriB homolog.

Sequence analysis
According to the genetic sequence database at the National Center for Biotechnical Information (NCBI), K. pneumoniae PriB protein (GenBank ID:YP_001338213) consists of 55 amino acids as predicted ab initio by Genemark 2.0. Given that the vast majority of PriB proteins have a sequence of approximately 104 amino acids, we found it striking that K. pneumoniae PriB would be shorter to such a significant degree. Therefore, we examined the genome of K. pneumoniae in the region upstream of the annotated priB gene and noticed that the start codon of the priB gene reported in the database is preceded by sequence that codes for a stretch of amino acids that is highly similar to the amino-terminal region of E. coli PriB. By including this additional upstream sequence, along with the annotated K. pneumoniae priB sequence, we were able to identify an open reading frame in the K. pneumoniae genome that codes for a 104-amino acid protein whose amino acid sequence is 95% identical to that of E. coli PriB ( Figure 1). Since the 55-amino acid K. pneumoniae PriB sequence in the NCBI database was predicted ab initio, we think it is likely that the ab initio gene search incorrectly assigned an internal ATG as the priB start codon, resulting in a truncated PriB amino acid sequence being reported in the NCBI database. This truncated PriB sequence appears to have formed the basis for the study by Hsieh and Huang. We propose that the actual amino acid sequence of K. pneumoniae PriB is 104-amino acids in length and is highly similar to that of E. coli PriB.
We also examined the sequences of the other PriB homologs reported by Hsieh and Huang to be shorter than would be expected based on the sequence of a typical PriB homolog [10]. We found that the amino acid sequence of Pectobacterium carotovorum PriB reported in the NCBI database is 106 amino acids in length, and the amino acid sequence of Salmonella enterica PriB is 104 amino acids in length. The amino acid sequence of Yersinia ruckeri PriB, as reported in the NCBI database, is 55 amino acids in length. Therefore, we analyzed the genome of Yersinia ruckeri in the region upstream of the priB gene in the same manner as we did for K. pneumoniae priB and found additional sequence upstream of the annotated priB start codon that codes for the missing amino-terminal region of Y. ruckeri PriB. Based on this sequence analysis, we propose that the priB genes of K. pneumoniae, P. carotovorum, Y. ruckeri, and S. enterica all encode proteins of comparable length to E. coli PriB (Figure 1).

Quaternary structure of K. pneumoniae PriB
A previous report by Hsieh and Huang describes a 55-amino acid variant of K. pneumoniae PriB as a monomeric protein [10], while E. coli PriB exists as a homodimer [11,12,13,14]. In E. coli, the dimerization interface of PriB is extensive and involves a large number of contacts between individual monomers. Among the interactions are hydrogen bonds that form between the aminoterminal b1 strand of one monomer and the amino-terminal b1 strand of the other monomer [12,13,14]. Since these b strands include amino acids 1-11, it is possible that a variant of PriB that lacks a portion of its amino-terminus could exist as a monomeric protein. This appears to be the case for the 55-amino acid variant of K. pneumoniae PriB that is missing residues analogous to amino acid residues 1-49 of E. coli PriB [10].
Since our sequence analysis of K. pneumoniae PriB suggests that it is a 104-amino acid protein whose sequence is highly similar to that of E. coli PriB, we hypothesized that the full-length, 104-amino acid K. pneumoniae PriB should exist as a homodimer. To test this hypothesis, we purified the recombinant E. coli PriB and K. pneumoniae PriB proteins and compared their quaternary structures using gel filtration chromatography. E. coli PriB and K. pneumoniae PriB each migrate through a sephacryl S-100 sizeexclusion chromatography column as a single peak with retention volumes of 62.61 ml and 62.86 ml, respectively ( Figure 2). Based on a calibration of the column using proteins of known molecular weight, we determined that E. coli PriB migrates as a dimer with a molecular weight of approximately 22.5 kDa and K. pneumoniae PriB migrates as a dimer with a molecular weight of approximately 22.4 kDa. These results indicate that E. coli PriB and K. pneumoniae PriB have highly similar quaternary structures under these experimental conditions.

DNA binding activity of K. pneumoniae PriB
Due to the high degree of similarity between K. pneumoniae PriB and E. coli PriB at the level of primary and quaternary structure, we hypothesized that the mechanism of ssDNA binding is similar between the two PriB homologs. To test this hypothesis, we used fluorescence polarization spectroscopy to measure the DNA binding activity of K. pneumoniae PriB to compare it with that of E. coli PriB. For these experiments, we measured the apparent dissociation constant for the interaction between K. pneumoniae PriB and fluorescein-labeled ssDNA oligonucleotides. The fluorescein tag on the ssDNAs allows us to measure PriB binding to the ssDNA due to the increase in fluorescence anisotropy of the PriB:ssDNA complex relative to the unbound ssDNA. K. pneumoniae PriB protein was serially diluted and incubated with 1 nM fluorescein-labeled ssDNA and the fluorescence anisotropy was measured. Apparent dissociation constants were obtained by determining the concentration of PriB needed to achieve 50% binding to each of the various ssDNA substrates.
When K. pneumoniae PriB was incubated with each of the fluorescein-labeled ssDNA oligonucleotides, we observed a PriBdependent increase in fluorescence anisotropy, indicating that K. pneumoniae PriB binds to the ssDNAs (Figure 3). The apparent dissociation constants for 15-base, 30-base, and 45-base fluorescein-labeled ssDNAs are 5063 nM, 4567 nM, and 62614 nM, respectively. As a comparison, E. coli PriB's apparent dissociation constant for the same 30-base fluorescein-labeled ssDNA, measured using the same instrument and under similar experimental conditions, is 34.667.7 nM [12]. These results indicate that the affinity of K. pneumoniae PriB for ssDNA is highly similar to that of E. coli PriB.
Overall, the results of our study support the hypothesis that K. pneumoniae PriB is a 104-amino acid ssDNA-binding protein whose structure and function mirrors that of E. coli PriB. Therefore,K. pneumoniae PriB does not likely represent a novel PriB homolog.

Cloning K. pneumoniae priB and E. coli priB
The priB gene of K. pneumoniae was amplified from strain MGH78578 genomic DNA by polymerase chain reaction (PCR) using primers oML292 (59-GCG TAT TCC ATA TGA CCA  ACC GTC TGG AGC TG) and oML293 (59-GTC ACG GAT  CCC TAG TCT CCA GAA TCT ATC AAT TC). The PCRamplified product was cloned into the pET28b expression vector (Novagen) using NdeI and BamHI restriction sites. The resulting plasmid contains a six-Histidine tag upstream of the complete coding sequence of the K. pneumoniae priB gene, which is under the control of a T7 promoter for overexpression in hosts harboring T7 polymerase controlled by the lacUV5 promoter. The cloning of the priB gene of E. coli was described previously [12]. The fidelity of the priB genes was confirmed by DNA sequencing. All plasmids were individually transformed into BL21(DE3) E. coli to allow recombinant protein overexpression following induction with isopropyl-b-D-thiogalactopyranoside (IPTG).

Equilibrium DNA binding assays
Fluorescence polarization spectroscopy was performed at 25uC with a Beacon 2000 fluorescence polarization system (Invitrogen). PriB proteins were diluted serially from 10,000 nM to 0.01 nM into 20 mM Tris-HCl pH 8, 50 mM NaCl, 4% (v/v) glycerol, 1 mM MgCl 2 , 1 mM b-mercaptoethanol, 0.1 mg/ml bovine serum albumin (BSA) and incubated with 1 nM 39-fluoresceinlabeled ssDNA oligonucleotides of varying lengths: 15-mer (59-TAG CAA TGT AAT CGT), 30-mer (59-GCG TGG GTA ATT GTG CTT CAA TGG ACT GAC), 45-mer (59-GCC GTG ATC ACC AAT GCA GAT TGA CGA ACC TTT GCT CCA GTA ACC) in a total volume of 100 ml. Apparent dissociation constants (K d,app ) were calculated by determining the concentration of PriB required to bind 50% of the fluorescein-labeled ssDNA. The unbound state is reported by the fluorescence anisotropy of the fluorescein-labeled ssDNA in the absence of PriB. The fully-bound state is reported by the fluorescence anisotropy of the fluoresceinlabeled ssDNA in the presence of a sufficient concentration of PriB to saturate the fluorescence anisotropy signal. Data are reported in triplicate and associated uncertainties are one standard deviation of the mean.