![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Previous Article | Next Article 
Journal of Bacteriology, December 2002, p. 6893-6905, Vol. 184, No. 24
0021-9193/02/$04.00+0 DOI: 10.1128/JB.184.24.6893-6905.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Bacteriophage HP2 of Haemophilus influenzae
Bryan J. Williams,1 Miriam Golomb,2 Thomas Phillips,2 Joshua Brownlee,1 Maynard V. Olson,3 and Arnold L. Smith1*
Department of Molecular Microbiology & Immunology,1
Department of Biological Sciences, University of Missouri—Columbia, Columbia, Missouri 65212,2
Genome Center, University of Washington, Seattle, Washington 981953
Received 16 May 2001/
Accepted 1 August 2002
 |
ABSTRACT
|
|---|
Temperate bacteriophages effect chromosomal evolution of their bacterial hosts, mediating rearrangements and the acquisition of novel genes from other taxa. Although the Haemophilus influenzae genome shows evidence of past phage-mediated lateral transfer, the phages presumed responsible have not been identified. To date, six different H. influenzae phages are known; of these, only the HP1/S2 group, which lyosogenizes exclusively Rd strains (which were originally encapsulated serotype d), is well characterized. Phages in this group are genetically very similar, with a highly conserved set of genes. Because the majority of H. influenzae strains are nonencapsulated (nontypeable), it is important to characterize phages infecting this larger, genetically more diverse group of respiratory pathogens. We have identified and sequenced HP2, a bacteriophage of nontypeable H. influenzae. Although related to the fully sequenced HP1 (and even more so to the partially sequenced S2) and similar in genetic organization, HP2 has a few novel genes and differs in host range; HP2 will not infect or lysogenize Rd strains. Genomic comparisons between HP1/S2 and HP2 suggest recent divergence, with new genes completely replacing old ones at certain loci. Sequence comparisons suggest that H. influenzae phages evolve by recombinational exchange of genes with each other, with cryptic prophages, and with the host chromosome.
|
INTRODUCTION
|
|---|
The host range of temperate bacteriophages is determined by multiple factors. Since phages require cellular components for replication, they become specialized to a compatible bacterial species. Within a host species, divergent restriction systems and surface receptors create barriers to interstrain transmission. Phages may overcome host range barriers by evolving DNA methylation systems (24) or by varying the structure of tail fiber proteins used for adsorption (31). The temperate bacteriophages of nonencapsulated (e.g., nontypeable) Haemophilus influenzae (NTHI) face adaptive challenges because of the unusually high genetic diversity of these bacteria (43). Here we describe a new temperate prophage, HP2, found in NTHI strains that are associated with unusual virulence.
Although six H. influenzae phages are described in the literature (HP1, S2A, B, C, N3, and 
flu), only HP1 and three types of S2 have been described in detail (4, 19, 21, 30, 43). Both HP1 and S2 infect H. influenzae Rd strains (4, 19), which were originally derived from an encapsulated serotype d (Sd) strain, but do not possess the genes for capsular biosynthesis (2, 15). We have discovered a new member of the HP1/S2 family that occurs as a prophage in the chromosome of strain R2866, a nontypeable invasive H. influenzae isolate (26).
DNA sequence analysis of HP1 and S2 types A, B, and C shows that these phages are closely related. Closer examination shows that type C is probably the original HP1 (37). Type A has many similarities to type C, but differences in the structures of the early promoter region suggest a different regulation of the lytic-versus-lysogeny decision. The type B variety appears to be a chimera between types A and C. The original host of the HP1/S2 bacteriophages is unknown, but UV-induced mixed-culture filtrates lysogenized an Rd derivative. All HP1 and S2 type phages have similar morphologies when viewed with an electron microscope. The N3 bacteriophage has a similar head structure, but a longer tail. The N3 phage is found only in particular NTHI strains, and on restriction analysis, it has a pattern distinct from HP1 (43). No other information or sequence data on N3 are available. flu is an incomplete phage found in the Rd KW20 genome and has genes homologous to ones in HP1 (21).
HP1, with its 32-kb genome, belongs to the family of bacteriophages represented by Escherichia coli P2. Historically HP1 was used to elucidate the mechanism of natural transformation in H. influenzae (6, 27, 33, 41, 42). HP1 is a temperate phage capable of either a lytic infection or lysogeny of the host. The promoters controlling the lysis-versus-lysogeny decision are located near the 5' end of the genome (9): one leftward and two rightward promoters transcribe cI and cox, which have genetic and functional homology to transcriptional regulators in lambda. In vitro HP1 cI, cox, and int function similarly to their counterparts in lambda. In HP1, the majority of the genes downstream from these regulators appear to encode proteins that are part of phage structure and assembly apparatus. The function of these downstream genes is inferred on the basis of homology to genes in other phages.
The S2 phages also appear capable of a temperate life cycle in Rd hosts. The 5' 5.6 kb of this phage was sequenced for comparison to HP1 (36). Major sequence differences between S2 and HP1 are interspersed with regions of high homology.
While investigating a previously described invasive NTHI strain (26, 46), we found a prophage whose range was limited to this strain and a few other NTHI strains. To elucidate whether the phage provided clues to the unusual virulence of this strain, we sequenced its chromosome and found a close relationship to HP1 and S2. H. influenzae Sd strains and Rd derivatives are not lysogenic for HP2; however, HP2 can lysogenize a phage-deleted form of its original host.
|
MATERIALS AND METHODS
|
|---|
Bacteria and media.
The bacteria used in this study are described in Table 1. Strain R2866, originally described as Int1, is a biotype V, nontypeable H. influenzae strain isolated from the blood of an immunocompetent child with signs of meningitis (26). This strain is serum resistant and harbors a 54-kb conjugal plasmid that encodes a ß-lactamase. sBHI broth was made up of brain heart infusion (BHI) medium (Difco, Becton Dickinson, Sparks, Md.) supplemented with 10 µg (each) of hemin-HCl (Sigma, St. Louis, Mo.), L-histidine (Sigma), and ß-NAD (Sigma) per ml. The heme solution was prepared by mixing 100 mg of hemin-HCl and L-histidine in 100 ml of 50°C water, to which 0.4 ml of 10 N NaOH (Sigma) is added. The solution was filter sterilized with a 0.22-µm-pore-diameter filter and stored at 4°C in a lightproof container for no more than 3 weeks. ß-NAD was dissolved in water to a concentration of 1 mg/ml, filter sterilized, and stored at 4°C. One volume of these solutions was aseptically added to 100 volumes of BHI broth prior to use. Chocolate agar was prepared as described by Difco with GC Media base. To avoid contamination with gram-positive organisms, bacitracin was added to all solid H. influenzae growth media at a final concentration of 500 U/liter (10 µg/ml), and all incubations were done at 37°C in air. Luria-Bertani (LB) agar and broth (Difco) were used for E. coli.
|
Phage induction.
|
|---|
To induce bacteriophage from the lysogens, a 100-ml sBHI culture was grown with shaking at 1,200 rpm at 37°C to an A600 of 0.15 to 0.2. Mitomycin C (Sigma) was added to a final concentration of 35 ng/ml, and the culture was shaken at 50 to 100 rpm. Bacterial replication continued to an A600 of 1.0, after which the optical density decreased, presumably due to phage-mediated lysis. When the optical density reached its minimum, generally 4 to 6 h after the addition of mitomycin C, the cells were pelleted by centrifugation at 20,000 x g for 15 min at 4°C in a Beckman J21 centrifuge. The supernatant was removed, and the centrifugation step was repeated to remove residual intact cells. The resulting supernatant was passed through a 0.22-µm-pore-diameter filter, after which chloroform was added (20 µl/100 ml). Phage-containing supernatant was stored at 4°C until further use. To concentrate the bacteriophage particles, the supernatant was centrifuged in 33-ml ultracentrifuge tubes at 40,000 rpm in a Ti 50.2 rotor for 3 h at 15°C. The resulting pellet was resuspended in a minimal amount of phosphate-buffered saline (PBS) overnight at 4°C with gentle shaking.
|
Plaque assay.
|
|---|
Bacteria were grown in sBHI broth to an A600 of 0.2 and then mixed with 1:5 to 1:100,000 dilutions (in PBS) of culture supernatant prepared as described above for phage induction. Soft agar consisted of 5 ml of 0.7% sBHI agar layered on a standard sBHI agar plate. The target strain was grown in sBHI broth to an A600 of 0.2 and diluted 1:100 in the same medium, an aliquot was added to the phage preparation, and 0.1 ml was spread over the surface of the soft agar. After overnight incubation at 37°C, clear plaques were counted (44).
|
Electron microscopy.
|
|---|
Concentrated bacteriophage stocks were stained in uranyl acetate (5) and visualized by T.P. with a JEOL 1200 EX transmission electron microscope at the Electron Microscopy Core at the University of Missouri—Columbia.
|
DNA isolation.
|
|---|
To purify phage for sequencing, 200 µl of resuspended phage pellet in sBHI was treated with 0.1 U of DNase I (Gibco-BRL, Rockville, Md.) for 30 min at 37°C. After DNase treatment, the phage preparation was extracted with an equal volume of Tris-saturated phenol (pH 8.0)-chloroform-isoamyl alcohol in proportions of 25:24:1. The aqueous layer was removed, and the extraction was repeated with an equal volume of fresh phenol solution. The DNA was precipitated from the aqueous layer by addition of 1/10 volume of 3 M sodium acetate (pH 4.0) and 2.5 volumes of absolute ethanol at -20°C and concentrated by centrifugation, and the pellet was washed with 1 ml of 70% ethanol at room temperature. After centrifugation at 12,000 x g for 15 min at 4°C, the ethanol solution was aspirated, and the pellet was allowed to air dry before resuspension in 50 µl of water or PBS.
|
Sequencing.
|
|---|
Sequencing was performed at the University of Washington Genome Center as described by Stover et al. (40). Phage DNA was cloned into pUC19, and the insert was sequenced with primers synthesized by that unit. The data set involved 462 dye-terminator and 123 dye-primer sequencing reads, sampled at random from the phage genome. The average number of q20 bases per read was 408. A q20 base is a base call with an estimated error rate of 1% as calculated by the PHRED base-calling software (11, 12). The redundancy of the data, in terms of q20 bases, was 7.6. Low-quality regions were resolved by a combination of manual and automated finishing procedures as described previously (17). An estimate of the number of remaining errors in the sequence based on quality scores was calculated with the phrap assembly software (16), which can be accessed at http://www.phrap.org. The expected number of residual errors in this 31.5-kb sequence was 0.16. In our experience, sequence with less than one predicted error usually has no errors. In addition, both strands of the first 10 kb of HP2 from attP to orf10 were independently sequenced at the University of Missouri DNA Core by using the same vector and method, and no differences were observed.
|
DNA analysis.
|
|---|
Open reading frames (ORFs) were identified using the ORF finder function in the OMIGA software program (Oxford Molecular). The ribosome binding sites in the HP2 ORFs were compared to the previously determined HP1 and S2 sequences to verify the most likely start codons. Similarity plots were obtained with the GCG software program available (Wisconsin Genome Center).
|
H. influenzae transformation.
|
|---|
H. influenzae was transformed by the M-IV technique (39). Gel-purified PCR fragments or linearized plasmid DNA was added to competent H. influenzae, and dilutions were plated on chocolate agar plates containing either ribostamycin (Sigma) at 30 µg/ml (for the TSTE cassette) or chloramphenicol at 5 µg/ml (for the cat cassette).
|
Southern analysis.
|
|---|
DNA was transferred from agarose gels to nylon membranes (Osmonics, Inc., Minnetonka, Minn.) by using a vacuum-assisted apparatus (Hoeffer Scientific). Agarose gels were depurinated in 0.25 M HCl for 1 h, followed by denaturation in 1.5 M NaCl containing 0.5 M NaOH for 30 min (29). Transfers were performed for 
3 h in 20x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), after which the membrane was treated with UV light to cross-link the DNA. Chemiluminescent detection was performed with which digoxigenin-labeled oligonucleotide probes or double-stranded PCR products as recommended by the manufacturers (Roche, Indianapolis, Ind.).
|
PCR amplification.
|
|---|
Table 2 lists the primers used for PCR amplification of selected portions of the HP1 and HP2 prophages. The locations of these primers on the HP2 genome map are shown in Fig. 1. For PCR amplification of fragments shorter than 2 kb, standard Taq polymerase was used according to the manufacturer's instructions (Perkin-Elmer, Boston, Mass.). An Eppendorf thermocycler (model, Mastercycler) was set to run 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s in that order. Occasional primer sets required adjustment of the annealing temperature. For larger products of up to 18 kb, the long-range PCR kit from Roche (GeneAmp XL) was used.
View larger version (26K):
[in this window]
[in a new window]
|
FIG. 1. Genetic map of HP2. Straight arrows indicate the approximate position of each ORF. The scale is in kilobases. Bent arrows indicate the locations of RNA polymerase promoter elements. Ball-and-stick figures indicate the locations of transcription terminators. Arrows above the map indicate locations of primers used in this work. The length of the HP2 chromosome is 31,508 bp. Primer 1 is homologous to coordinates 3088 to 3072 of the H. influenzae Rd KW20 genome (section 9 of 163).
|
|
|
Construction of an HP2 host.
|
|---|
The plasmids used in the HP2 host construct are described in Table 3. Using PCR primers 1 and 2, a 1.7-kb fragment of HP2 DNA containing int and attP was amplified and ligated to pTrcHisB restricted with BamHI and XhoI (pBJ102). PCR primers 3 and 4 were used to amplify a 2.0-kb downstream portion of the HP2 prophage that was subsequently ligated into pBJ102 digested with BglII and EcoRI (pBJ102.2). A BamHI-restricted TSTE cassette was ligated into BglII-digested pBJ102.2 to create pBJ102.3. The TSTE cassette contains the aph(3')I gene flanked by H. influenzae-specific uptake (hUS) sequences (34). The TSTE cassette confers ribostamycin resistance to H. influenzae and kanamycin resistance to E. coli. Plasmid pBJ102.3 was digested with BamHI and EcoRI and used to transform competent H. influenzae strain R2866 with selection for ribostamycin resistance. Of 12 ribostamycin-resistant transformants, 2 were shown to be devoid of most of the prophage genome by Southern blotting and to lack phage production after mitomycin C treatment, as assessed by electron microscope observation and infection assays (data not shown). One such mutant was designated R3420 (Fig. 2). R3422 is a derivative of R3420 with a chloramphenicol acetyltransferase cassette replacing the aph(3')I gene, inserted between two HincII sites.
View larger version (45K):
[in this window]
[in a new window]
|
FIG. 2. Genetic maps of strains used in this work. Straight arrows indicate ORFs and their orientation. Bent arrows indicate the locations of the transcription promoters. Diagonal striped boxes indicate the boundaries of host DNA. DNA originating from the HP1 or HP2 host is shown in dark or light gray, respectively. The TSTE box indicates the location of the antibiotic cassette used for genetic manipulations. These maps are not to scale but approximate the total numbers of genes and their relative sizes.
|
|
|
Construction of hybrid lysogens.
|
|---|
Early experiments indicated that HP2 would not form plaques on any of the Rd derivatives or strain R3420, its original host, from which HP2 was isolated. To identify the genetic regions determining the host range of HP2, we created hybrid lysogens of HP1 and HP2 (Fig. 2). This was accomplished by first cloning a 7.5-kb HindIII prophage fragment containing the HP2 immunity genes from strain R2866 into the HindIII site of pUC18. This plasmid, designated pBJ100.1, contains a portion of a threonine synthetase gene and a BamHI site in an intergenic region 5' to the prophage. After cloning TSTE into this BamHI site, the plasmid (pBJ100.2) was linearized and transformed into competent R3152 selecting for ribostamycin resistance. One transformant (designated HP1/HP2P [strain R3403]) of 12 examined acquired the HP2 immunity region as indicated by PCR. The chromosomal DNA of another transformant that retained the HP1 immunity region was digested and transformed into R2866. One transformant of the 12 that acquired HP1 immunity region was designated HP2/HP1P (strain R3404) (Table 1). To verify the construction of the hybrid phages, we performed a Southern analysis of BglI-restricted DNA harvested from phage preparations of HP1, HP2, HP1/HP2P (R3403), and HP2/HP1P (R3404) by using a digoxigenin-labeled PCR product generated from primers 5 and 6 as a probe. HP2 contains a 2.0-kb fragment, while in HP1, the hybridizing fragment is smaller, as predicted. HP1/HP2P has the 2-kb BglI fragment, while HP2/HP1P has the smaller fragment.
|
Construction of marked HP2 derivative.
|
|---|
To assess the ability of HP2 to lysogenize a host, we created a prophage mutant in R2866 with the TSTE cassette marking the phage. The insertion of TSTE into the phage dam gene resulted in no detectable phenotypic changes in growth rate or phage yields. This insertion was created by first cloning a portion of the prophage with primers 7 and 8 to amplify a 7-kb segment of the phage containing most of the genes driven by the pR promoters. This PCR product was digested with HindIII and EcoRI and ligated into pKS to create pBJ105. pBJ105 was digested with NcoI, which cuts this plasmid uniquely in the dam gene, and was treated with T4 polymerase. A BamHI-digested, T4 polymerase-treated TSTE cassette was ligated to this plasmid to yield pBJ105.2. This plasmid was digested with HindIII and EcoRI and transformed into R2866 with subsequent selection of ribostamycin-resistant colonies. Southern analysis of chromosomal DNA and phage extract DNA from eight colonies revealed one mutant, R3435, which contained the TSTE cassette in the dam gene of the HP2 prophage (data not shown).
|
Nucleotide sequence accession number.
|
|---|
The HP2 sequence has been deposited in GenBank under accession no. AY027935.
|
RESULTS AND DISCUSSION
|
|---|
The HP2 genome.
The HP2 chromosome consists of 31,508 bp, similar to the size of S2 phage types A and B based on restriction mapping (28). The molar percentage of adenine and thymidine (A+T%) in the HP2 chromosome is 60.04%, a value similar to that in the Rd KW20 chromosome (61.86%) (15). The frequency of the triplet base combinations, coding and noncoding, in HP2 is also very similar to that in Rd KW20 (data not shown), which suggests that this bacteriophage was not recently introduced into H. influenzae.
The organization of the HP2 genome is shown in Fig. 1; cohesive ends are similar to those in HP1 (data not shown). HP2 appears to contain five transcriptional units, with the control of each of these units directing or repressing bacteriophage replication. As in HP1, the pR1, pR2, and pL1 promoters of HP2 adjoin the early regulatory elements. Flanking these promoters are elements believed to control the lysis-versus-lysogeny decision (13). If the products of the pL1 promoter dominate, lysogeny is maintained, repressing all other bacteriophage gene expression. If the pR1 and pR2 promoters are activated, the lytic cycle will ensue. Products of the pR1- and pR2-activated transcript should control bacteriophage DNA replication and presumably activation of the downstream genes through hypothetical promoter elements between orf16 and orf17. Genes responsible for bacteriophage particle production and host lysis reside in these diverging transcripts, one of which contains orf15 and orf16, while the other contains orf17 through orf35. Many of the ORFs in the latter transcript show homology to structural proteins of P2 and other phages. As in HP1, orf14 appears to have its own promoter and terminator. The role of this gene in HP1 and HP2 is unknown. It is unique in being the only gene in these phages that appears capable of independent control.
|
HP2 regulatory elements.
|
|---|
The pR and pL promoters controlling the lysis-versus-lysogeny decision differ among HP1, HP2, and S2 phages. Analysis of these regions indicates that both of the pR promoters are maintained in HP2, whereas the pR1 promoter, and its corresponding cI-coded protein binding site, is missing from S2 (Fig. 3). The nucleotide sequences of these promoter regions differ at numerous sites: areas that are conserved are the -10, -35, and cI- and cox-coded protein binding sites. This suggests HP2 has retained a functional control unit for phage induction and repression. As in S2, the cox homologue of HP1 is absent. Whereas the orf2 genes of HP2 and S2 are similar to cox (see below), the finding of intact Cox protein binding sites suggests that a Cox-like protein performs this function. The spacing between the -10 and -35 sites of pR1 in HP2 is 16 or 17 bp, depending on which thymidine residue is considered the start of the -10 site. The pR1 promoter may be functionally redundant, since S2 lacks pR1, yet appears fully capable of controlling lysis versus lysogeny in H
Top
Abstract
Introduction
