This study shows that PCR amplicon markers for the four lambdoid prophages found on the B. anthracis Ames strain chromosome are present in all of a large number of diverse B. anthracis strains tested. Genome sequence analysis could not predict the functionality of the phages since all prophages appear to have a full complement of phage genes. All prophages were demonstrated to be excision proficient and the prophage excision frequency reported here is comparable to that seen in another Gram-positive species, Lactobacillus lactis measured by real-time PCR. In that case, the frequency of spontaneous prophage induction varied from 10-7-10-1/cell depending on the conditions examined . In a limited analysis performed in this study, no viable phage production was observed upon mitomycin C induction as determined by plaque assays using nearly 100 diverse Bacillus spp strains as indicators. Although mitomycin C induction exhibited a significant growth inhibition, there was no obvious cell lysis characteristic of phage induction. Direct electron microscopic examination of the concentrated culture supernatants failed to reveal any intact or defective phage particle. This result is in contradiction with an earlier report that polyethylene glycol-concentrated culture supernatants of a B. anthracis Sterne strain showed defective phage particles/phage components under electron microscopic examination  and is probably due to strain differences or the method used for concentration of the culture supernatant.
There are several possible explanations for the apparent failure in detecting viable phage particles. 1) We have not found an appropriate phage sensitive strain in the panel of 96 strains used in this study for the plaque assay, which might be due to the absence of the phage receptor in these strains. Even if any of the strains possess the receptor, they may lack some other factor needed for phage growth as was shown in the case of γ phage infection of a B. thuringiensis strain that possesses the γ phage receptor, the gamR gene . Similarly, in another study six phages isolated from B. anthracis were screened for their host range and only 2 out of 64 non-anthracis strains tested were susceptible to two of the phages . 2) We have not found a repressor mutant that can overcome superinfection immunity and form plaques on 34F2. 3) The fact that all four prophages are simultaneously defective in their ability to produce viable phage particle or lyse the host cell may be indicative of a defective host factor that is involved in phage life cycle.
Although the failure of plaque formation even on closely related species does not necessarily indicate that phages are defective, further work is needed to unequivocally establish this fact. One approach to show that the phages are defective would be to test for plaque formation on the parental carrier strain cured of the respective prophages. Our future studies will focus on testing this possibility.
All four prophages contain genes encoding recombinases and terminal-repeat DNA motifs that may function as attachment (att) sites. However, the lambdaBa0l recombinase (GBAA3832) is a pseudogene owing to a frameshift mutation. We hypothesize that lambdaBa0l might recruit some other site-specific recombinase enzyme for its excision, or have a mechanism for frameshift correction during translation. Alternatively, prophage excision might occur via recA mediated general recombination pathway, in which case, the frequency of excision will increase with the increase in the length of the repeats and the products will be identical to the site-specific recombination products.
In Escherichia coli, prophage excision is classically triggered as a response to DNA damage . In this regard, B. anthracis possesses some of key components of an active SOS response system, including a lexA ortholog. Despite the fact that there is an intron in the B. anthracis recA gene, the encoded protein is still apparently functional . The results of mitomycin C induction experiment described here supports the role of DNA damage in induction of some prophages.
Excision proficient prophage sequences are generally not considered useful targets for bacterial identification because of their instability. However, the constant presence of all four prophages can be advantageous for the definitive discrimination of B. anthracis from all its neighbors. There are several schemes for molecular detection of B. anthracis DNA using unique markers present on virulence plasmids pXO1 and pXO2 [26–28] but there are relatively few unique chromosomal targets. Targets for confirmation of B. anthracis chromosomal DNA include using 16S rDNA sequences , rpoB [30, 31], gyrB , gyrA , spore structural protein gene sspE  and S-layer protein gene sap  for discrimination of B. anthracis from other Bacillus spp. Several multiplex PCR assays have also been described for B. anthracis detection and discrimination [34–36]. One multiplex PCR assay described earlier entailed amplification of unique fragments identified through suppression subtractive hybridization (SSH) and several of these were located on prophage regions . Because of the close relatedness of the B. anthracis to B. cereus group strains, many of the PCR based detection techniques are prone to false positive identification of non-anthracis strains. The prophages are unique to B. anthracis and are present in all B. anthracis strains examined so far. Failure to produce the B. anthracis specific signals in multiplex PCR would be the result of simultaneous excision of all four prophages, which is highly unlikely. Hence they may offer unique signatures for B. anthracis chromosome. From an evolutionary standpoint, it seems possible that the conservation of the phages in all B. anthracis may be due to the very recent emergence of the lineage.