Virology Journal Genome Sequence of the Bacteroides Fragilis Phage Atcc 51477-b1

The genome of a fecal pollution indicator phage, Bacteroides fragilis ATCC 51477-B1, was sequenced and consisted of 44,929 bases with a G+C content of 38.7%. Forty-six putative open reading frames were identified and genes were organized into functional clusters for host specificity, lysis, replication and regulation, and packaging and structural proteins. Findings Bacteriophages infecting Escherichia coli and Bacteroides fragilis serve as fecal pollution indicators (1). Bacteroides phages are more attractive fecal indicators than E. coli because Bacteroides are more abundant in fecal matter, provide higher host specificity, and are anaerobic and thus less likely to reproduce in aquatic environments than E. coli [1-4]. Phage ATCC 700786-B1, infecting B. fragilis RYC2056, serves as an ISO standard reference phage, but the host is also susceptible to phages in other animal feces [5,6]. Phage ATCC 51477-B1 infects the host B. fragilis HSP40, which is reported to be only susceptible to phages in human feces and surface water polluted with munici-pal/septic wastewater [7]. A drawback to using Bacteroides phages as water quality indicators results from the requirement to cultivate them on an anaerobic host [6]. This can be circumvented with direct phage detection via PCR [8], but assay design is difficult because only one gene for one Bacteroides bacteri-ophage is publically available. The lack of ability to detect Bacteroides bacterial phage is startlingly inadequate because the host genera is dominant in the human gut and contains important antibiotic resistant pathogens [3,9]. In this study, the genome of phage ATCC 51477-B1 was sequenced to promote fecal source tracking assay development and aid construction of a B. fragilis phage bioreporter [10]. Phage ATCC 51477-B1 was propagated on B. fragilis HSP40 (ATCC 51477) at 37°C in anaerobic Bacteroides phage recovery medium (BPRM) amended with kanamy-cin, nalidixic acid, bile salts, and Oxyrase (Oxyrase; Mans-field, OH) [11,12]. Phage were purified and concentrated from lysate using polyethylene glycol [13] and displayed a non-elongated icosohedral head and non-contractile tail consistent with the family Siphoviridae and other B. frag-ilis phages isolated from municipal wastewater (Figure 1) [14]. Structural dimensions were similar to B. fragilis phage B40-8 with respect to head diameter (60 ± 4.0 μm) and tail length (162 ± 17.0 μm), but the tail diameter was which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Findings
Bacteriophages infecting Escherichia coli and Bacteroides fragilis serve as fecal pollution indicators (1). Bacteroides phages are more attractive fecal indicators than E. coli because Bacteroides are more abundant in fecal matter, provide higher host specificity, and are anaerobic and thus less likely to reproduce in aquatic environments than E. coli [1][2][3][4]. Phage ATCC 700786-B1, infecting B. fragilis RYC2056, serves as an ISO standard reference phage, but the host is also susceptible to phages in other animal feces [5,6]. Phage ATCC 51477-B1 infects the host B. fragilis HSP40, which is reported to be only susceptible to phages in human feces and surface water polluted with municipal/septic wastewater [7].
A drawback to using Bacteroides phages as water quality indicators results from the requirement to cultivate them on an anaerobic host [6]. This can be circumvented with direct phage detection via PCR [8], but assay design is difficult because only one gene for one Bacteroides bacteriophage is publically available. The lack of ability to detect Bacteroides bacterial phage is startlingly inadequate because the host genera is dominant in the human gut and contains important antibiotic resistant pathogens [3,9]. In this study, the genome of phage ATCC 51477-B1 was sequenced to promote fecal source tracking assay development and aid construction of a B. fragilis phage bioreporter [10].
Phage DNA was extracted with Lambda minipreps (Promega, Madison, WI) and digested with HindIII. Several HindIII restriction fragments were cloned into pUC19 [15] and sequenced using primer walking. PCR reactions bridging the cloned HindIII restriction fragments were cloned and sequenced to confirm fragment order. Nonredundant areas of the genome that were difficult to clone were directly sequenced. In total, dideoxynucleotide sequencing provided 2× coverage except at the 5' and 3' ends. Confirmation of sequence data was sought by GS FLX pyrosequencing (MWG Biotech, Inc., High Point, NC) which produced 23,263 reads, 58,730 base calls, and assembly of 62 contigs using the GS De Novo Assembler. Thirty eight of the contigs with more than 3 reads were coassembled with the dideoxynucleotide sequences using DNASTAR Lasergene 7.1 software suite (DNASTAR, Inc., Madison, WI). A total of 21,935 pyrosequencing reads in 27 contigs were maintained in the final assembly which aligned at 91% similarity for 100× coverage of the phage genome, including the 5' and 3' ends. The final assembled genome was 44,929 bp with a 38.7% G+C content ( Figure  2).
Open reading frames (ORFs) were identified using Gene-Mark [16,17] and the NCBI ORF Finder and examined for known protein functions, structures, and motifs using a conserved domain database [18] (Additional file 1, Figure  2). Ten of the 46 putative ORFs contained amino acid sequences with predictable functions or motifs (Additional file 1). The 5' ends of three ORFs (ORF39, ORF40, and ORF43) displayed translated similarity to previously published N-terminal amino acid sequences for B. fragilis phage B40-8 head (MP1 and MP3) and tail (MP2) proteins at 100, 80, and 90% similarity, respectively [13]. The B40-8 MP2 gene (the only B. fragilis phage gene present in GenBank), was present in ATCC 51477-B1 but contained a 119 bp insertion and was split into three misarranged sections, one of which was inverted ( Figure 2). This suggested the putative ATCC 51477-B1 tail protein was chimeric with respect to B40-8.
The ATCC 51477-B1 genome contained gene functional clusters for host specificity (tail fiber), lysis, replication and regulation, and packaging and structural proteins ( Figure 2) [19]. The host specificity region contained ORF7 with translated similarity to the tail fiber of Enterococcus phage phiEF24C (Additional file 1) and a large size (1,897 amino acids) suggesting it may be a tape measure gene [20]. A putative M-15 type peptidase (ORF10) was the only gene clearly linked with the lysis region. The replication and regulation region included phage anti-repressors (ORF22 and ORF26), DNA modifying enzymes (ORF15 and ORF16), and replication proteins (ORF22, ORF31, and ORF32). Within the packaging and structural cluster, an ORF with similarity to the TerL protein was identified (ORF 38), as well as ORFs with N-terminal sequences similar to the three phage B40-8 structural proteins previously mentioned (ORF39, ORF40, and ORF43) [13].
Four potential promoters were identified on the positive strand of the genome based on similarity to promoters found in B. fragilis [21]. Two were in the regulation and replication region, 5' of ORFs 17 and 22. The other two potential promoters were near the beginning of the genome, 5' of ORF2, and in the lysis region, 5' of ORF12. A tandem repeats finder [22] revealed a 23 bp repeat within an 83 bp segment having two perfect and two Transmission electron micrographs of phage ATCC 51477 degenerate repeats in the replication and regulation region. Another 96 bp perfect repeat sequence was identified at the beginning of the phage genome and again at the end of the replication and regulation region ( Figure  2).
Eight polymorphic regions occurring within ORFs and displaying sequence variability from 4% to 13% were identified with the extensive pyrosequencing data. These regions ranged in size from 200 bp to greater than 1,000 bp. ORF7, the putative tail fiber protein, displayed the most polymorphisms, including a pyrosequencing contig with a deletion relative to the final assembly. Tail fiber variability modifies the phage host range [23] and may, for this phage, reflect the antigenic variability of B. fragilis surface components [21,24].
The sequence for B. fragilis phage ATCC 51477-B1 was deposited in GenBank with accession number FJ008913.