Bacillus anthracis, B. cereus and B. thuringiensis are closely related organisms and represent a diverse range of pathogens as well as biotechnologically useful, non-pathogenic bacteria [1,2,3,4]. Bacteriophages and their components can be very useful tools for typing, detection and control of pathogenic bacteria in the food chain, biodefense measures, and for treatment of human infections. Thus, a better knowledge of the phages infecting Bacillus bacteria is highly desirable [5,6,7,8].

It is unfortunate that, over the last decades, the name TP21 has been assigned to three independently isolated phages infecting B. anthracis, B. thuringiensis and B. cereus [9,10,11]. However, it may be assumed that all of them are different, based on electron microscopy and/or partial genome sequences available [10]. The aim of this study was to determine the complete genome sequence of a specific TP21 phage isolate previously studied in our laboratory [10]. For reasons of clarity, we added the modifier "L" to the phage name, i.e., naming it TP21-L.

TP21-L was propagated using host HER1399 (ATCC 13472) in liquid culture [10], and purified by poly-ethylene glycol precipitation and density gradient centrifugation as described elsewhere [12]. Electron microscopy of purified negatively stained phage particles [13,14] (Figure 1), allowed assignment of TP21‑L to the Siphoviridae family, in the order of the Caudovirales [15]. The phage features an isometric head of 58.5 nm diameter and a long, non-contractile, flexible tail of 144.8 nm length and 11.0 nm diameter. A putative tail fiber could be observed (Figure 1, indicated by arrows), corresponding to a putative tail fiber encoding gene in the genome of TP21-L (see below).

Phage genomic DNA was extracted from purified phage as described elsewhere [12]. The TP21‑L genome sequence was determined using a shotgun cloning and sequencing approach, followed by manual gap closing using primer walking. After assembly of the complete genome with 5.6-fold average coverage, TP21‑L revealed a unit genome sequence of 37,456 base pairs (Figure 2), which is in good agreement with pulsed field gel electrophoresis analysis of full length phage DNA [13] (Figure 3). The G+C content is 37.8 mol%, slightly higher than that of the host bacteria (35.5 mol%, as determined from Genbank data of published host bacteria genomes (NC_003909, NC_004742, NC_005957, NC_003997, and NC_008600). The sequence of TP21-L has been deposited in the databases under accession number EU887664.

The TP21‑L DNA molecule features invariable, fixed ends, as revealed by the typical pattern of single fragments disappearing over time when Bal31-predigested DNA is subjected to restriction digestion [13] (Figure 4). Heating of restriction digests prior to electrophoresis did not reveal any change compared to non-heated samples (data not shown), suggesting the absence of cohesive ends [16].

Fifty-six putative open reading frames could be identified in the TP21‑L genome, but only 17 of them revealed homologies to proteins of other bacteriophages (Table 1). No indication for modified, hydroxy-methylated bases [17] was found by restriction profiling. Mass spectrometry based identification (MALDI-MS peptide fingerprinting) of SDS-PAGE-separated structural proteins (Figure 5) [13,18] led to the identification of ten distinct protein species: one could be allocated to a minor structural protein (gp17) of 192.6 kDa, and another could be allocated to the putative tape measure protein (gp15) of 99.9 kDa. The dimension of the phage tail is in good correlation with the size of the Tmp protein, as demonstrated for phage Lambda and others [19-21] (Table 1, Figure 2). The putative portal protein (gp3) was identified in a band of approx 67 kDa, not in agreement with the predicted mass, and a tail fiber protein (Figures 1, 2) (gp16) was identified in a band of approx. 58 kDa. A putative capsid protein (gp4) as well as the major capsid protein (gp7) were also identified (39 kDa and 30 kDa, respectively). Yet another band (approximately 21 kDa) apparently contained traces of the major capsid protein, probably due to proteolytic cleavage or posttranslational modification of Cps, a fact also observed in other phages [13,22]. The predicted gene products gp6, gp8, and gp11 could also be identified as structural proteins, but do not exhibit amino acid homology to any known protein from the databases. However, the predicted mass of gp6 did not match the observed mass (Figure 5), suggesting posttranslational processing of this protein.

From the genome sequence (Figure 2), the endolysin (gp19) [10,23], a putative holin (gp18; featuring a TMHMM-predicted transmembrane domain [24]), as well as the lysogeny control region could be deduced. The latter consists of only two genes (gp20 and gp21) (Figure 2), whose predicted products share homologies with a transcription repressor and a recombinase. This correlates well with the ability of TP21‑L to lysogenize its host (HER1399). In order to demonstrate this, we have isolated clones of HER1399 resistant to infection with TP21-L, sub-cultured the cells, treated them with UV light (254 nm, 120 mJ/cm², 2 min), and used the sterile-filtered supernatants for demonstration of lytic activity against HER1399 on pre-inoculated agar plates. As a control, a UV-induced culture of wild-type HER1399 did not release any phage. The lysogenized strains were homoimmune to superinfection with up to 10⁹ Pfu/ml TP21‑L. PCR using TP21‑L specific primers (see below) confirmed the presence of phage in the lysogenized strains (results not shown).

Unfortunately, the designation TP21 has coincidentally been assigned to a total of three phages in the past. In our attempt to identify the individual TP21 isolates and clarify their nomenclature, the origins and history of the isolates is briefly described. TP21‑L was made available to our lab by Hans-Wolfgang Ackermann, from the Félix d’Herelle Reference Center for Bacterial Viruses in 1994; it was described as a phage infecting Bacillus thuringiensis and designated TP21. Unfortunately, it was not possible to further elucidate the exact origin of this phage, and it was also not kept in the Félix d’Herelle Reference Center for Bacterial Viruses collection. However, since the genome sequence is now available, and it is also clear that TP21‑L is unique among the Bacillus phages, it should be re-deposited in the collection.

The second Bacillus phage named TP21 was isolated more than 20 years ago by Ruhfel and Thorne [11,25]. We propose that this phage should be renamed to TP21‑T. The phage is unusual since it was shown to persist as an autonomously replicating plasmidal prophage in B. thuringiensis kurstaki [25]. In order to determine that TP21‑L is different from TP21‑T, we designed and used a set of PCR primers to amplify unique sequences of 0.5 to 1.1 kb from the genomes of both phages (the primer sequences were generated based on the unpublished genome sequence of TP21‑T (R. Okinaka and P. Jackson, personal communication). Amplification with primers specific for TP21‑L phage do not yield detectable products using template DNA from B. thuringiensis kurstaki HD1-9 or strain DP 4848 / UM 101, a carrier of TP21‑T (Figure 6), although the primer combination TP21‑L2_Fw an TP21‑L2_R seems to produce an unspecific product both with HER1399 and DB 4848 as template (Figure 6, l and 2). Using primer combinations specific for TP21‑T produced products of the expected size with strain DB 4848 as template but not with HER1399 (Figure 6). Vice versa, using TP21‑L DNA as template did also not yield specific amplification products with TP21‑T-specific primers [11] (data not shown). We conclude that TP21‑L is not the same as TP21‑T from Ruhfel and Thorne [11,25].

The third TP21 isolate was isolated from a Chinese factory producing B. thuringiensis powder and described by He and coworkers [9], and thus renamed to TP21‑H. This phage apparently features an elongated head of 87 x 55 nm and a flexible tail of 140 x 8 nm in size (B2 morphotype), thus clearly distinguishing it from TP21‑L (Figure 1). Unfortunately, however, further details are not available, and the whereabouts of TP21‑H are unclear. It is not kept in the Félix d’Herelle Reference Center for Bacterial Viruses collection or any other collection, and has probably been lost (H.-W. Ackermann, personal communication). Even though, we propose to name this phage TP21‑H, should it ever be rediscovered or reisolated.

We present the complete genome sequence and molecular characterization of the temperate Bacillus phage TP21‑L. The TP21‑L unit genome is 37.5 kb in size and encodes 56 open reading frames, 17 of which could be assigned a putative function. Ten structural proteins of the virion were identified by peptide mass fingerprinting. We analyzed the relationship of TP21‑L to two Bacillus phages sharing the same designation. The three TP21 isolates analyzed here clearly represent different phages, as determined by genome sequencing of TP21‑L, PCR probing of TP21‑T and comparison of morphological data on TP21‑H and TP21‑L. Only two of them have been maintained, and they should be designated TP21‑L and TP21‑T. Unfortunately, the confusion in naming or re-naming bacteriophages is symptomatic for the confusing nomenclature of phages infecting many different organisms, and underlines the urgent need for a strong, reliable and unambiguous classification scheme which should be based primarily on molecular data. Most bacteriophage names do not provide immediate recognition of host and origin and often have dual meanings (e.g. TP can stand for ‘transducing phage’ or ‘B. thuringiensis phage’). Recently, a rational scheme for phage nomenclature has been proposed by Kropinski et al. [26], which would resolve ambiguities by a name prefix composed of virus type identifier, host abbreviation, virus family and specific designation. However, the TP21 phages represent the rare case where even such a robust system alone would fail, since all TP21 phages are from the same virus type, family and host organism. In these cases, the addition of a suffix, as proposed here, provides means of resolution of the naming disorder, even it if results in re-naming of the phages with unpredictable side-effects to citation of scientific literature.