Molecular Characterization of a Variant of Bacillus anthracis-Specific Phage AP50 with Improved Bacteriolytic Activity

The genome sequence of a Bacillus anthracis-specific clear plaquemutant phage, AP50c, contains 31 open reading frames spanning14,398 bp, has two mutations compared to wild-type AP50t, andhas a colinear genome architecture highly similar to that ofgram-positive Tectiviridae phages. Spontaneous AP50c-resistantB. anthracis mutants exhibit a mucoid colony phenotype.

INTRODUCTION

Top

INTRODUCTION

Bacillus anthracis, a category A biothreat agent, is a spore-forminggram-positive bacterium of the Bacillus cereus sensu lato group.It is a zoonotic soil bacterium that infects animals and occasionallyhumans, causing the disease anthrax. The standard diagnostictests for suspected B. anthracis recommended by the Centersfor Disease Control and Prevention (CDC) include testing for ${gamma}$ phage sensitivity (5). Recently, the inherent binding specificityand lytic action of the ${gamma}$ phage-encoded lysin enzyme have beenexploited for rapid detection and killing of B. anthracis (18).Although ${gamma}$ phage exhibits a fairly narrow host range, severalB. cereus strains (e.g., ATCC 4342) have been shown to be sensitiveto infection by this phage (1, 4, 6, 18).

In this study, we have characterized a mutant B. anthracis-specificphage of the family Tectiviridae, AP50, isolated from soil usingB. anthracis Sterne as the host in 1972 (2, 11). AP50 possessesa typical capsid architecture of the family Tectiviridae, i.e.,isometric nucleocapsids with icosahedral symmetry and a capsidshell composed of two layers (Fig. 1a) (15). Originally it wasthought to be an RNA phage but was later shown to contain double-strandedDNA and phospholipid (15). The wild-type AP50 phage, designatedAP50t, was shown to have a narrow host range; only one-thirdof the 34 B. anthracis strains and none of the 52 strains belongingto six different Bacillus spp. were susceptible to infectionby AP50 (13). Nine major structural proteins were identifiedon sodium dodecyl sulfate-polyacrylamide gel electrophoresisgels. The molecular mass of the phage DNA was estimated to be9 x 10⁶ daltons (14). Treatment with organic solvents such aschloroform (5%) and ether (25%) for 30 min inactivated the phageto a survival of about 1 x 10^–4 PFU/ml (12).

View larger version (79K):
[in this window]
[in a new window]

FIG. 1. Transmission electron micrograph of AP50c phage particles. (a) Uranyl acetate staining at a magnification of 297,000. Scale bar, 100 nm. (b) Plaque morphology of turbid and clear plaques in a mixed lysate.

Isolation of AP50c and determination of its burst size.

Top

Isolation of AP50c and...

AP50t produces turbid plaques characteristic of lysogenic phages,and spontaneous clear plaque variants designated AP50c werefound in lysates after repeated propagation (Fig. 1b). The clearplaques were larger after incubation at 25°C than afterincubation at 37°C and larger in phage assay agar (20) thanin Luria-Bertani or blood agar plates (data not shown). Theburst sizes at 25°C and 37°C reached about 200 to 300particles/cell compared to about 3 to 5 particles/cell at 42°C.

Comparison of host ranges of phages AP50c and ${gamma}$ .

Top

Comparison of host ranges...

AP50c infected 111 of the 115 B. anthracis strains ( $~$ 97%) andnone of the 100 B. cereus sensu lato strains (16) tested inthis study, indicating a narrow host range. A side by side comparisonof the infectivities of B. anthracis strains by phages AP50cand ${gamma}$ indicated that only 4 out of 115 strains were resistantto AP50c whereas 10 out of 115 strains were resistant to ${gamma}$ phage.Nine of the 10 strains that were resistant to infection by ${gamma}$ phage were susceptible to infection by AP50c, 3 of the 4 AP50c-resistantstrains were sensitive to ${gamma}$ , and one strain was resistant toboth AP50c and ${gamma}$ phages (see Table S1 in the supplemental material).

AP50c genome sequencing and comparative analyses to other gram-positive bacterium-specific tectiviral genomes.

Top

AP50c genome sequencing and...

The genome sequence of phage AP50c was determined by pyrosequencingtechnology using a Roche/454 Life Sciences GS20 sequencer (10).AP50c DNA isolated from purified phage particles was used tosequence to an average depth of coverage of 175 reads per base.The final assembled sequence of the AP50c genome was determinedto be 14,398 bp.

The G+C content of AP50c is comparable to that of its host,i.e., 38% compared to 35% for the B. anthracis chromosome and32% and 33% for plasmids pXO1 and pXO2, respectively. The AP50cgenome has short (28-bp) almost perfect inverted terminal repeats,19 of which are identical, and contains 31 putative open readingframes (ORFs) (Fig. 2a). AP50c exhibits limited similarity atthe DNA level to other sequenced tectiviral phages. Bacillusthuringiensis phages Bam35, pGil01, and pGil16c are closelyrelated to each other (86 to 99% identity), and B. cereus elementpBclin15 is more closely related to AP50c (61%) than to theB. thuringiensis elements (59%) (Fig. 2b).

View larger version (45K):
[in this window]
[in a new window]

FIG. 2. (a) Genome map of AP50. Three gene clusters based on functional grouping and similarities to other Tectiviridae phages infecting gram-positive bacteria are shown. ORF boxes are color coded to indicate the degree of amino acid identity with proteins of other tectiviral phages. The color of the ORFs and percentage identities are as follows: green, 60 to 80%; red, 40 to 60%; purple, 15 to 40%; blue, <15%. ITR, inverted terminal repeat; HVR, highly variable region. Open arrowheads indicate the locations of the mutations in AP50c phages. (b) Visualization summary of whole-genome nucleotide alignments of gram-positive tectiviral phages. The ClustalW alignment file generated from Multifasta alignment was visualized in Base-By-Base (3). In this type of alignment, if two sequences have insertions or deletions relative to one another, the output looks different depending on which of the two sequences is used as the base sequence. White, perfect nucleotide homology; blue, single nucleotide polymorphism; red, deletions in the indicated phage; green, insertions in the indicated phage. The GenBank accession numbers for the sequences used in the alignment are as follows: Bam35c, NC_005258; pBth35646, NZ_AAJM00000000; Gil01, AJ536073; Gil16c, AY701338; AP50, EU408779; pBclin15, AE01878. (c) Sequence changes in AP50c and AP50t genomes. The mutation in the noncoding region just upstream of ORF 1 at nucleotide position 164 is indicated. The second mutation is in ORF 28 at position 12881 and changes amino acid residue 91 (an isoleucine in AP50c to a valine in AP50t).

AP50c proteins have an overall percent identity to other gram-positivebacterium-specific tectiviral phage genomes of $~$ 53%, and extensivesimilarities exist (15% to 77%) throughout the genome. All theputative proteins of AP50c and the locations of their codingsequences in comparison to the Gil16, Bam35, and pBclin15 proteincoding sequences are listed in Table 1. The AP50c genome iscolinear with other known gram-positive Tectiviridae phage genomes(7, 17, 19, 21, 22) and has a cassette-like organization. Theleft end of the genome contains genes involved in genome replication(ORFs 2 to 5) and lysogeny control, and the middle part containsvirion structural and DNA packaging genes (ORFs 11 to 27). Atleast five capsid proteins can be predicted based on similaritiesto other Tectiviridae phages: ORFs 11, 17, 23, 26, and 27. Theright end contains the host recognition and lysis genes. Similarlyto other Tectiviridae phages, 21 of the AP50c ORFs were predictedto encode soluble proteins, while 10 of the proteins had oneor two transmembrane regions.

View this table:
[in this window]
[in a new window]

TABLE 1. Comparison of predicted ORFs and their functions in the AP50c genome to those of other Tectiviridae genomes

Several ORFs of AP50c encoded proteins that showed no significantsimilarity to any proteins in the databases, and they have beendesignated as highly variable regions. The ORFs encoding theseproteins are located in two distinct regions of the genome:the first between bp 4827 and 5242 (ORFs 8 to 10) and the secondbetween bp 12611 and 13531 (ORFs 28 to 30). Except for theseORFs, all the ORFs have orthologs in other Tectiviridae phages.

There are two ORFs in the AP50c genome that encode putativeregulatory proteins. ORF 1 encodes a MerR-like protein, andORF 6 encodes a LexA-like repressor protein. AP50c ORF 1 issimilar to ORF 1 of the Bam35, Gil16c, and pBClin15 genomes,which exhibited similarities to truncated MerR-like regulatoryproteins. A role for the LexA-like repressor (ORF 6 product)in lysogeny maintenance in Gil01 and Gil16 phages has been determined(21). Clear plaque variants of these phages possessed mutationsin the ORF 6-ORF 7 region.

ORF 31 of AP50c encodes a putative endolysin. A protein alignmentof ORF 31 to similar ORFs in other tectiviral phage genomesrevealed several interesting features. ORF 31 of the AP50c genomeis very similar to ORF 31 of Gil16c: the encoded proteins are65% identical and 79% positive over a 186-amino-acid region.They possess the conserved endolysin domain of the amidase-3family (N-acetylmuramoyl-L-alanine amidase), involved in degradationof bacterial cell walls. Also, the putative endolysins of Bam35(encoded by ORF 30) and pBclin15 (encoded by ORF 28) are similarto each other (67% identical/77% positive over a 226-amino-acidregion). They possess an Acm (1,4-β-N-acetylmuramidase)/glycosylhydrolase family 25 domain. As expected, neither the AP50c northe Gil16c ORF, encoding an amidase, is similar to the Bam35or pBclin15 ORF, encoding a muramidase, despite the fact thatoverall Gil16c is genetically closer to Bam35 than it is toAP50 (Fig. 2b). Therefore, it seems possible that there hasbeen an interspecies exchange of endolysin domains between AP50/Gil16cand pBclin15/Bam35, resulting in divergent endolysins in otherwiseclosely related B. thuringiensis phages (Bam35 and Gil16c).Such nonorthologous replacements have been found to be commonamong tailed phages (8). Interestingly, the carboxy-terminalregions of endolysins, involved in cell wall binding and hostrecognition (9), of Gil16 and Bam35 endolysins are somewhatsimilar (37% identical/58% positive over 70 amino acids at theC terminus), as they share the same host, B. thuringiensis.Also, the carboxyl termini of AP50, pBclin15, and Bam35/Gil16endolysins are different, probably due to differing host specificities.

Identification of the genetic changes in turbid/clear plaque variants.

Top

Identification of the genetic...

DNA sequencing of the entire genomes of five turbid plaque phagesrevealed two changes in the genome: a single base substitution(T to C) at position 154 in a noncoding region at the left endof the AP50 genome and a nonsynonymous change at position 12881within ORF 28 of unknown function which alters an isoleucineat position 91 in the AP50c protein to leucine in the AP50tphage protein (Fig. 2c).

AP50c exhibits very high bacterial killing efficiency.

Top

AP50c exhibits very high...

Bacterial survivors of phage AP50t infection arose at a veryhigh frequency (50% of the initial CFU prior to infection),and the colonies were morphologically different from the uninfectedcells: flat and wrinkled (Fig. 3a and b). These survivors werethe result of lysogen formation. On the other hand, AP50c eliciteda very high killing efficiency (5- to 6-log decrease in bacterialviability) since it propagates exclusively lytically. The survivorsof AP50c infection appeared at a frequency of $~$ 1 x 10^–6CFU/ml. These AP50c^r cells exhibited other phenotypes: whilewild-type 34F₂ cells settled and formed a pellet at the bottomof the tube upon overnight incubation statically at room temperature,AP50c^r cells formed a weak pellet and the colonies were mucoid.Furthermore, AP50c^r cells appeared to secrete an extracellularflocculent material. Indeed, scanning electron microscopic analysisof AP50c^r cells revealed the presence of a thin coating of anextracellular material on the outer cell surface (Fig. 3c and d)and the absence of AP50c particle attachment. The AP50c^r cellswere sensitive to infection by ${gamma}$ phage.

View larger version (39K):
[in this window]
[in a new window]

FIG. 3. Colony morphologies of B. anthracis Sterne strain 34F₂ after infection with AP50t and morphology of AP50c-resistant 34F₂ mutants. (a) Uninfected 34F₂ cells diluted and plated on phage assay agar plates. (b) AP50t-infected culture, diluted and plated. The smooth and round white colonies are uninfected nonlysogens (solid arrowhead), and the flat wrinked colonies (open arrowhead) are lysogens. (c) Scanning electron micrograph of a wild-type 34F₂ cell infected with AP50c. The arrowheads indicate AP50c particles attached to the outer surface of the bacterium (magnification, 60,000; Dennis Kunkel Microscopy, Inc.). (d) Scanning electron micrograph of 34F₂ AP50c^r mutant infected with AP50c showing the presence of extracellular material coating the outer cell surface and the absence of attached phage particles (magnification, 60,000; Dennis Kunkel Microscopy, Inc.).

Thus, AP50c, as described here, exhibits specificity and sensitivitycomparable to those of ${gamma}$ . Phages AP50c and ${gamma}$ probably have differenttarget bacterial receptors since some of the naturally occurringand laboratory-isolated strains resistant to ${gamma}$ infection aresensitive to AP50 and vice versa (see Table S1 in the supplementalmaterial; S. Sozhamannan, unpublished data). Thus, a combinationof these two phages might be a better alternative for phage-baseddiagnostics, therapeutics for B. anthracis infection, and environmentalclean-up of B. anthracis-contaminated areas.

Nucleotide sequence accession number.

Top

Nucleotide sequence accession...

The nucleotide sequence of phage AP50c reported here has beendeposited in the NCBI GenBank database under accession numberEU408779.

ACKNOWLEDGMENTS

This work was supported by funds from the Defense Threat ReductionAgency, Department of Defense of the U.S. government (BDRD authors;grant no. 8.10084_08_NM_B), and by Finnish Centre of ExcellenceProgram (2006-2011) grant 1213467 (J.K.H.B.).

The views expressed in this article are those of the authorsand do not necessarily reflect the official policy or positionof the Department of the Navy, Department of Defense, or theU.S. government.

FOOTNOTES

* Corresponding author. Mailing address: Genomics Department, Biological Defense Research Directorate Annex, Naval Medical Research Center, 12300 Washington Avenue, Rockville, MD 20852. Phone: (301) 231-6713. Fax: (301) 231-6799. E-mail: Shanmuga.Sozhamannan{at}med.navy.mil

Published ahead of print on 12 September 2008.

Supplemental material for this article may be found at http://aem.asm.org/.

REFERENCES

Top