Receptor Diversity and Host Interaction of Bacteriophages Infecting Salmonella enterica Serovar Typhimurium

Background Salmonella enterica subspecies enterica serovar Typhimurium is a Gram-negative pathogen causing salmonellosis. Salmonella Typhimurium-targeting bacteriophages have been proposed as an alternative biocontrol agent to antibiotics. To further understand infection and interaction mechanisms between the host strains and the bacteriophages, the receptor diversity of these phages needs to be elucidated. Methodology/Principal Findings Twenty-five Salmonella phages were isolated and their receptors were identified by screening a Tn5 random mutant library of S. Typhimurium SL1344. Among them, three types of receptors were identified flagella (11 phages), vitamin B12 uptake outer membrane protein, BtuB (7 phages) and lipopolysaccharide-related O-antigen (7 phages). TEM observation revealed that the phages using flagella (group F) or BtuB (group B) as a receptor belong to Siphoviridae family, and the phages using O-antigen of LPS as a receptor (group L) belong to Podoviridae family. Interestingly, while some of group F phages (F-I) target FliC host receptor, others (F-II) target both FliC and FljB receptors, suggesting that two subgroups are present in group F phages. Cross-resistance assay of group B and L revealed that group L phages could not infect group B phage-resistant strains and reversely group B phages could not infect group L SPN9TCW-resistant strain. Conclusions/Significance In this report, three receptor groups of 25 newly isolated S. Typhimurium-targeting phages were determined. Among them, two subgroups of group F phages interact with their host receptors in different manner. In addition, the host receptors of group B or group L SPN9TCW phages hinder other group phage infection, probably due to interaction between receptors of their groups. This study provides novel insights into phage-host receptor interaction for Salmonella phages and will inform development of optimal phage therapy for protection against Salmonella.


Introduction
Emergence of antibiotic-resistant pathogens due to abuse of various antibiotics is driving the development of alternative approaches to pathogen control. Bacteriophages are considered a possible alternative biocontrol agent for bacterial pathogens [1,2]. This approach has advantages including narrow species specificity and safety for human applications [3]. As an example, a clinical trial in which volunteers were given oral doses of T4 phage indicated that it was safe [4]. In addition, a cocktail (ListShield TM , Intralytix, Inc., Baltimore, MD, USA) of Listeriainfecting bacteriophages was approved by the US Food and Drug Administration in 2006, gaining the status of ''Generally Recognized as Safe'' [2,5], supporting that bacteriophage may be a good candidate as a biocontrol agent for human applications.
Salmonella is a Gram-negative foodborne pathogen causing 1.4 million cases of salmonellosis including 17,000 hospitalization and 600 deaths every year in US [6,7]. S. enterica serovar Typhimurium is common serotype in human infection and is frequently isolated from clinical and non-clinical samples from chicken sources. A large proportion of S. Typhimurium strains are resistant to several antimicrobial drugs, for example the multi-drug resistant S. Typhimurium phage type DT104 [8]. Therefore, use of S. Typhimurium bacteriophages is now getting more attractive as an alternative approach in the treatment for antibiotic-resistant pathogens.
Recent reports have described the isolation of new Salmonella bacteriophages and evaluation of their bactericidal effects [9][10][11][12]. For example, Salmonella-specific phage st104a or st104b reduced the bacterial cell number by up to 2 logs within 1 h of each phage infection [12]. W25 phage reduced S. Typhimurium viable cell number up to 2.19 logs within 24 h [13,14]. However, the rapid emergence of phage-resistant Salmonella is an obstacle to effective biocontrol using phages [5,[14][15][16][17]. To reduce the impact of phage-resistance, phage cocktails have been developed and found to be effective in control of phage-resistant Salmonella [15,[18][19][20]. For example, a phage cocktail containing 45 different phages reduced Salmonella cell numbers up to 5 log in 2 h [15]. The phage cocktail approach can also broaden the host range, for example, one phage cocktail designed for serovar Typhimurium increased the host range to include S. enterica serovar Enteritidis and serovar Kentucky [21].
Because attachment of bacteriophage to the specific receptor of the host bacteria is the critical first step of phage infection [22], mutation of the receptor is the most frequent route to phageresistance of the host. To date, several Salmonella phage receptors are known including FhuA [23], TolC [24], BtuB [25,26], OmpC [27], Vi capsular antigen [28], lipopolysaccharide (LPS) [29], and flagella [13]. The study of phage receptors is expected to provide insight into the emergence of phage-resistance in Salmonella and guide optimization of phage cocktails for Salmonella control.
In this study, host receptors for 25 new Salmonella phages were determined using a phage-resistant Tn5-insertion random mutant library of S. Typhimurium. Cross infection studies with these phages and resistant strains revealed novel phage-host interactions and infection mechanisms. Further investigation of diversity of host receptors in Salmonella phages will increase our ability to circumvent phage resistance using phage cocktails and thus prevent food spoilage due to S. Typhimurium.

Bacterial Strains and Growth Conditions
Bacterial strains used in this study are listed in Table 1. All strains were grown in Luria-Bertani (LB) broth medium (Difco, Detroit, MI, USA) at 37uC with shaking for 12 h. Salmonella enterica serovar Typhimurium strain SL1344 was used for isolation of Salmonella-infecting phages from the collected samples, and prophage-free S. Typhimurium strain LT2C [30] (Cancer Research Center, Colombia, MO, USA) was used for purification of phages.

Bacteriophage Isolation and Propagation
Bacteriophages isolated in this study were listed in Table 2. No live chickens were used in this study. Seventy-six samples obtained from chicken feces collected from farms and commercially processed broiler skins obtained from markets were used as sources for isolation of Salmonella-specific bacteriophages. The chicken feces were collected with permissions from the farm owners and the samples were collected for the purpose of this research only. Twenty-five grams of each sample were mixed with 225 ml of sodium chloride-magnesium sulfate (SM) buffer (100 mM NaCl, 10 mM MgSO 4 ?7H 2 O and 50 mM Tris?HCl, pH 7.5) without gelatin in sterile bags. Twenty-five milliliters of each homogenized sample were mixed with 25 ml of 26 concentrated LB broth and incubated with shaking at 37uC for 12 h. After centrifugation (5,0006g for 10 min), the supernatant was filtered using 0.22-mm pore size filters (Millipore, Billerica, MA, USA). Ten milliliters of each filtered sample were mixed with 50 ml LB broth with 10 7 CFU/ml of overnight cultured S. Typhimurium SL1344. The mixture was incubated with shaking at 37uC for 12 h. After centrifugation (5,0006g for 10 min), the supernatant was filtrated using 0.22 mm pore size filters. The presence of phages was assessed using a plaque forming assay with molten 0.4% LB soft agar containing 10 7 CFU/ml of S. Typhimurium SL1344. After incubation at 37uC for 12 h, individual plaques were picked and eluted with 1 ml of SM buffer without gelatin. Plaque isolation and elution were repeated more than five times for pure isolation of individual phages in S. Typhimurium LT2C. One liter of exponentially growing S. Typhimurium LT2C (OD 600 nm = 1.0) was infected with each SPN phage at a multiplicity of infection (MOI) of approximately 1 and incubated with shaking at 37uC for 4 h. Cell debris was removed by centrifugation at 5,0006g for 10 min and filtered using 0.22-mm pore size filters. Phage particles were precipitated from the filtrate by addition of 10% polyethylene glycol 6,000 (Sigma, St. Louis, MO, USA). Finally, a stepped CsCl density ultracentrifugation (himac CP 100b, Hitachi, Japan) with step densities of 1.3, 1.45, 1.5, and 1.7 g/ml at 78,5006g for 2 h was conducted at 4uC. The bands of viral particles were withdrawn from the tube with a syringe and dialyzed using 1 L of SM buffer for 1 h and stored at 4uC.

Transposon Mutagenesis and Receptor Screening
Random insertion mutagenesis of S. Typhimurium SL1344 was performed using the EZ-Tn5 TM ,R6Kcori/KAN-2.Tnp Transposome kit (Epicentre, Madison, WI, USA) according to manufacturer's procedure. The transposon construct was obtained by treating pMOD3 with PvuII restriction enzyme, and gel purifying the fragment from a 1% agarose gel using the QIAquick Gel Extraction Kit (Qiagen, Valencia, CA, USA). Electrocompetent cells were prepared as follows: Fifty milliliter of LB broth inoculated at 1% with an overnight culture of S. Typhimurium SL1344 was incubated with shaking at 37uC for 1.5 h. Cells were then harvested by centrifugation at 5,0006g for 10 min, and the pellet was washed three times with 1 ml ice-cold molecular grade water and resuspended with 100 ml ice-cold molecular grade water. The complexes were electroporated into S. Typhimurium SL1344 using the Gene-Pulser Xcell system (Bio-Rad, Hercules, CA, USA) at 2.45 kV, 200 V and 25 mF in 2 mm electroporation cuvettes. Transformants were selected on LB agar plates containing kanamycin (Kan, 50 mg/ml), and individual colonies were cultured and stored in LB broth with Kan (50 mg/ml) containing 15% glycerol at 280uC.
To screen mutant strains resistant to each SPN phage, the mutant library was duplicated in the 96-well plates containing LB broth supplemented with 50 mg/ml of Kan. The SPN phage was added to one of those duplicated 96-well plates (MOI = 10) after shaking at 37uC for 1.5 h and growth inhibition was compared with the corresponding 96-well plate not inoculated with phage. All plates were incubated at 37uC for an additional 3 h. Phage resistance mutants were identified by comparing the optical density at 600 nm of the plates with and without phage using an iMark microplate absorbance reader (Bio-Rad). Rescue cloning of transposed genomic DNA and partial sequencing was conducted to identify the Tn5 transposon insertion sites according to the manufacturer's protocol.

Construction of Deletion Mutants and Complementation Plasmids
S. Typhimurium SL1344 derivatives with deletions of flgK, fliR, fliC, fljB, fliC/fljB or rfaL genes were constructed using the lambda red recombination method [31]. The kanamycin resistance (Kan R ) cassette from plasmid pKD13 was amplified using primers specific for each gene, for example, flgK-lamb-F1 and flgK -lamb-R1 for flgK. Sequences of all primers for these constructs are provided in Table S1. The polymerase chain reaction (PCR) products were transformed into S. Typhimurium SL1344 harboring pKD46 and integrated into the chromosomal flgK, fliR, fliC, fljB or rfaL genes.

Electron Microscopy
The morphology of CsCl-purified SPN phages was determined using transmission electron microscopy (TEM). Concentrated viral samples were diluted with SM buffer without gelatin and 5 ml of each phage sample was applied to the surface of carbon coated copper grids. Excess volume was removed by carefully touching the side of grid with filter paper and 5 ml 2% uranyl acetate (pH 4.0) was spotted on the grid for negative staining and removed after a short interval. The prepared samples were observed using TEM (LIBRA 120, Carl Zeiss, Switzerland) at 80 kV. Taxonomy of the SPN phages was determined according to the guidelines of the International Committee on Taxonomy of Viruses [35].

Host Range Determination by Spotting Assay
A 100-ml aliquot of bacterial culture was added to 6 ml molten 0.4% LB agar and then poured on a 1.5% LB agar plate. After solidification of the top agar, 10 ml serially diluted phage suspension ranging from 10 2 to 10 5 plaque forming unit/ ml was spotted on the top agar and the plates were incubated at 37uC for 12 h. After incubation, the sensitivity of indicator strains to the tested phages was determined by degrees of clearing in the spots or plaques ( Table 1). The plaque assay was performed in triplicate.

Isolation of Phage-resistant Strains and Determination of Cross Resistance
To investigate influence of resistance against one type of receptor to infection by phages using different receptors, resistance strains were developed against phages and then they were used for the infections by phages using different host receptors. Group L phage-resistant strains showing resistance to re-infection were isolated using high-titer overlay assay following modified Kim and Ryu's protocol [26]. In this modified protocol, separate colonies of group L phage-resistant strains were obtained by additional streaking on LB agar plate. Because high-titer overlay assay did not work for isolation of transiently phage-resistant strains by group F, group B, and group L SPN9TCW phage infections, they were isolated using high-titer broth assay to increase the yield of the phage-resistant strains and to maintain resistance in the presence of phages. For the high-titer broth assay, phages were added to an LB broth culture of S. Typhimurium LT2C Table 2. Characteristics of the isolated S. Typhimurium-specific bacteriophages and their identified receptors.  (OD 600 nm = 1.0) at MOI = 100, and the culture was incubated with shaking at 37uC until the OD 600 nm reached 1.0, again. To remove excessive phages, the phage-infected cells were then harvested by centrifugation at 5,0006g for 10 min, resuspended in 200 ml ice-cold molecular grade water and used as a host for the second infection of SPN phages using different host receptors. The host resistance to the second infection was monitored using plaquing assays as described above.

Lysogen Induction
All phage-resistant strains of S. Typhimurium LT2C were cultivated at 37uC until OD 600 nm reached to 1.0 and 0.5 mg/ml of mitomycin C (Sigma) was added to the cultures. Then, these cultures were additionally incubated at 37uC for 2 h. After incubation, the cells were removed by centrifugation and filtration and the supernatant was collected. The spotting assay of this supernatant with S. Typhimurium LT2C was conducted to confirm the lysogen formation. To confirm the unstable lysogen formation of group F phage-resistant strains, the group F, B, L-SPN9TCW phage-resistant strains were plate in green plate (Evans blue uranine agar plate, 0.5% NaCl, 1% Tryptone, 0.5% Yeast extract, 0.5% K 2 HPO 4 , 0.04 M Glucose, 0.04% Evans blue, 0.04% Uranine, and 1.5% Micro agar, final concentration) as the procedure developed by Chan et al. [19].

Real-time Reverse Transcription (RT)-PCR
Total RNA was isolated from S. Typhimurium using the RNeasy Mini Kit (Qiagen) and converted to cDNA using the Omniscript RT Kit (Qiagen) and random hexamers (Invitrogen, Carlsbad, CA, USA) using manufacturer's instructions. Quantitative real-time RT-PCR was performed as previously described [36] with primers listed in Table S1.

Bacteriophage Isolation
Between September and December 2009, 25 bacteriophages were newly isolated from 18 of 76 samples (23.7% phage recovery frequency). All phages in Table 2 were designated as Salmonella Phage Number (SPN) and specific numbers were used to indicate the sample sources and sometimes characters were used to differentiate the isolated phages from the same samples, respectively.

Grouping of Bacteriophages Based on their Receptors
To determine the host receptors for 25 phages, a Tn5 random mutant library of S. Typhimurium SL1344 containing thousands of Tn5-inserted mutants was constructed and mutant strains resistant to each phage via inactivation of host receptor genes by Tn5 insertions were screened. Interestingly, sequence analysis of the Tn5 insertions in the phage-resistant mutants revealed only three different types of genes: flagellar production genes, the gene encoding the vitamin B 12 uptake outer membrane protein, and genes involved in LPS-related O-antigen production. The Tn5 random mutant library screening results were confirmed by targeted inactivation of the identified genes by lambda red recombination method and complementation experiments. The host receptor genes deleted by the lambda red recombination method are indicated in the receptor gene clusters presented in Fig. 1. Based on the deleted specific genes for formation of phage receptors, 25 phages were grouped into group F, group B, and group L phages, respectively (Table 2 and Fig. 2A).
The deleted genes involved in flagella production included genes encoding the flagella hook-filament junction protein (flgK), a putative component of the type III flagella export apparatus (fliR), and the gene encoding flagellin (fliC). Because S. Typhimurium expresses either one of two flagellin genes, fliC or fljB [38], we tested whether the group F phages can use both flagellins by phage infection analysis of fliC and fljB deletion mutants. The group F phages can be categorized into two groups, F-I and F-II, as shown on Table 3 and Fig. 2B. The group F-I phage could not infect the fliC mutant or the fliC/fljB double mutant, but could infect the fljB mutant, suggesting that the group F-I phage can only use FliC as a receptor. Group F-II phage could not infect the fliC/fljB double mutant but could infect the fliC and the fljB single mutants, suggesting that the group F-II phage can use either FliC or FljB as a receptor (Table 3 and Fig. 2B).
One group of resistant mutants has deletion mutations in the btuB gene encoding the membrane transporter for vitamin B 12 , suggesting that BtuB is a group B phage receptor. The mutated genes in the O-antigen biosynthesis are O-antigen ligase (rfaL) and CDP glucose 4,6-dehydratase (rfbG). Complementation of the deleted genes with pACYC184 and pUHE21-lac q expression vectors containing the wild-type genes restored susceptibility, supporting that O-antigen is a receptor for group L phage infection. Overall, 11 out of 25 phages use flagella (group F phage), seven out of 25 use BtuB (group B phage), and another seven out of 25 use LPS-related O-antigen (group L phage) as receptors. Although the Salmonella outer membrane proteins (OMPs) such as TolC [24], FhuA [23] and OmpC [27] are known receptors for some phages, no phages using those receptors was present in this set of 25 phages. It is not clear why BtuB was the only OMP detected as a receptor in this study.

Morphology
All 25 phages could be categorized into three morphological groups (Fig. 3). Interestingly, this morphological grouping is correlated with the grouping of Salmonella-specific phages based on their receptors ( Table 2). All of the group F and B phages have isometric heads and non-contractile, cross-banded tails that are longer than the tails of the group L phages (Fig. 3). These phages can be classified into the B1 morphotype of the Siphoviridae family, although the group F phages have a single, long, kinky tail fiber structure and the group B phages have four or five L-shaped fibers ( Fig. 1A and 1B). The group L phages are classified as members of the Podoviridae family. They have isometric heads with very short tails that are distinct from other groups.

Host Range of the Isolated Bacteriophages
Infection analysis of 25 Salmonella-specific bacteriophages was conducted with 21 Salmonella strains, five E. coli strains and nine other species of Gram-negative and Gram-positive bacteria (summarized in Table 1). In general, SPN bacteriophages infect S. Typhimurium strains but not other Gram-negative and Gram-positive bacteria, suggesting that they are specific to S. Typhimurium. The group F phages also infect S. enterica subsp. arizonae and subsp. salamae, and the group B phages can infect several strains of E. coli, indicating that these bacteriophages have a host range that extends to other Salmonella subspecies or E. coli strains. In addition, group L phages and some group F phages infect S. Paratyphi. The group B phages have much broader host range than the group F and L phages. They can infect E. coli and Shigella flexneri strains as well as Salmonella ( Table 1), suggesting that the BtuB proteins of S. Typhimurium, E. coli, and Shigella flexneri are similar. Comparative sequence analysis of BtuB proteins in the group B phage susceptible strains S. Typhimurium LT2, E. coli MG1655 and Shigella flexineri 2a strain 2457T revealed .87% identity at the amino acid level between S. Typhimurium and the other two species, whereas S. Typhimurium and Vibrio fischeri ES-114, which is not susceptible to group B phage, share ,35% identity at the amino acid level. These findings support the hypothesis that the group B phage receptor motifs are shared among Salmonella, E. coli and Shigella. Comparative host range analysis of group F-I and F-II phages revealed that F-II phages infect a larger number of S. Typhimurium isolates (data not shown), probably because the F-II phages can use either the FliC or the FljB flagellin as a receptor while the F-I phages can only use FliC (Table 3 and Fig. 2B).

Lysogenization
It is intriguing that while the group F and L phages generally make turbid plaques, and thus may be temperate phages, the group B phages make clear plaques and may be virulent phages [39]. Induction experiments in which mitomycin C was used to induce lytic growth indicate that the group L phage-resistant S. Typhimurium LT2C strains carry an inducible prophage. In contrast, mitomycin C treatment of the group L SPN9TCW phage-resistant and group B phage-resistant Salmonella strains did not yield phage, indicating that these phages do not make lysogens in the LT2C strain consistent with the clear plaque morphology (Table 1 and 3). Most of the group F phage-resistant Salmonella strains did not yield phage after treatment with mitomycin C even though they make turbid plaques. Five percent of the group F phage-resistant Salmonella made phage in response to mitomycin C, but they also lose resistance easily upon subculturing in the absence of the phages, suggesting the possibility of formation of unstable lysogens or pseudolysogens [40] (Table 4). To confirm if transient resistance is due to unstable lysogeny or pseudolysogeny, we conducted the green plate (Evans blue uranine agar plate) experiment (Fig. S1). In the green plate experiment, only the cells lysed by phage induction make blue colonies due to pH change. It revealed that while the resistant strains to the virulent phages in group B and group L-SPN9TCW did not show any blue colony, the resistant strains to group F phages did show small number of blue colonies (approximately 5% of all colonies) in the green plates. These results indicate that a few of the strains resistant to group F phages were lysed, suggesting that group F phages do not form stable lysogens.

Cross-resistance of Phage-resistant Salmonella to the Different Receptor Group Phages
To understand the interaction of phage with the specific host receptors, derivatives of S. Typhimurium LT2C that are resistant to the group F, group B and group L phages described here were isolated and characterized. Group B phage-resistant Salmonella [26] were transiently resistant to re-infection with group B phages, and most group F phage-resistant and group L SPN9TCW phage-resistant strains also showed transient resistance to re-infection with phages from their own group. All group L phage-resistant strains, except for the strain resistant to phage SPN9TCW, showed stable phage resistance to group L phages. Interestingly, all group L phage-resistant strains except those resistant to SPN9TCW were lysogens, suggesting that the resistance for the group L phages is due to prevention of superinfection by a stable prophage (Table 4 and Fig. 4B) [41,42]. Cross-infection of group F phages on other phage-resistant strains showed sensitivity to these phages, suggesting no mutual influence between flagellin and other phage receptors on the sensitivity to the phages (Table 4 and Fig. 4). However, cross-infection of group B phages on group L phageresistant strains yielded two different patterns. While the group L phage-resistant strains are sensitive, the SPN9TCW phage-resistant strain is resistant to group B phage infection, depending on the formation of lysogen (Table 4 and Fig. 4BD). Furthermore, group L phages were not able to infect group B phage-resistant strains, suggesting a possible influence between the BtuB and LPS receptors on the sensitivity to the phages (Table 4 and Fig. 4C).

Discussion
The details of the molecular interactions between phages and their host receptors that determine host specificity are not fully understood yet. Here, we isolated 25 Salmonella-specific phages, identified their receptors, studied their host specificity, and examined cross-resistance among the phages and phage resistant hosts. Although this study on Salmonella phage receptors using 25 phages does not represent the complete range of infection mechanisms used by Salmonella-infecting bacteriophages, the  Table 3. Flagellin-targeting phages: receptor and sensitivity patterns based on specific gene mutation.   Table 4. Cross resistance of phage-resistant strains. results provide novel insights into general host-phage interactions of typical Salmonella-infecting bacteriophages. Bacteriophages tend to use structures exposed on the outer membrane of the host bacteria as a receptor because they are easily accessible. Unexpectedly, only three kinds of receptors were identified in this study, flagella, O-antigen, and the outer membrane protein BtuB, although several other outer membrane proteins, including FhuA, TolC, and OmpC, are characterized as phage receptors of Salmonella. One possible reason may be the complex nature of Salmonella Typhimurium LPS [43,44], which may block access of phage to some outer membrane proteins, making it more challenging to isolate those phages. In a seeming contradiction, LPS may help attachment of group B phage to BtuB on a S. Typhimurium host. T5 phage, for example, has both a receptor-binding tail protein Pb5 and a L-shaped tail fiber protein on the phage tail that targets the host outer membrane protein and LPS, respectively [45,46]. The Pb5 as a major host specificity protein is reported to mediate irreversible binding to a specific outer membrane protein and the L-shaped tail fiber protein as a helper protein provides reversible binding to LPS. Therefore, the L-shaped tail fiber protein increases the infection rate of T5 phage by stabilization of binding between phage and an outer membrane receptor via Pb5. Group B phages are in the same family as T5 phage, Siphoviridae, and have a similar tail structure containing an Pb5-like protein possibly targeting BtuB and an L-shaped tail fiber protein possibly targeting LPS (Fig. 3). Due to this similarity between T5 and group B phages, it is suggested that this tail fiber protein may help the binding of Pb5like protein to BtuB to increase the infection rate of group B phages, but this proposed binding mechanism needs further study.
Morphological characterization of the group F and group L bacteriophages also provided insight into the interaction of these phages with host receptors. The group F phages have relatively long, non-contractile and cross-banded tails with a single and twisted tail fiber structure. This structure is very similar to that of chi phage that infects E. coli, Salmonella and Serratia through flagella filament receptors [47,48], suggesting possible interaction between the group F twisted tail fibers and bacterial flagella. Group L phages are morphologically belong to Podoviridae and use LPS as a main receptor like other phages including e15 [49,50], P22 [51] and T7 [52], which all interact with LPS via major host specificity proteins. However, Salmonella phages of Myoviridae family were not isolated in this study. It is not clear why we failed to isolate phages of Myoviridae family. The standard phage isolation protocol employed in our study may not be suitable for isolation of Salmonella phages in Myoviridae family because the genome sizes of Salmonella phages of Myoviridae family except for Pedovirinae subfamily are generally bigger than those of other family phages [53][54][55][56][57]. Otherwise, there might be unknown bias in the phage isolation.
Host range analysis of group F-I phages showed that they successfully infected all S. Typhimurium strains tested except strain ATCC 14028. It is not clear why the group F-I phages that use only FliC as a receptor could not infect ATCC 14028 while the group F-II phages could. The complete genome sequence of strain ATCC 14028 (GenBank accession number CP001363) showed that it has a fliC gene and multiple sequence alignment with the fliC genes of susceptible Salmonella strains revealed no sequence differences. Real-time RT-PCR and SDS-PAGE analysis confirmed that the fliC gene is expressed and translated for flagella formation, indicating that FliC in the strain ATCC 14028 is functional (data not shown). The group B phages also infected all S. Typhimurium strains except one, strain DT104. Sequencing and real-time RT-PCR analyses of the btuB gene in strain DT104 showed that it is expressed (data not shown), so it is not clear why the strain is resistant. These two examples imply that there may be an additional unknown factor(s) that makes the phage and host receptor interaction more specific and complicated. However, only S. Typhimurium ATCC 14028 has Gifsy-3 prophage, so this prophage could cause superinfection exclusion to group F-I phages.
The phages that can make lysogen normally make turbid plaques but many other factors are involved in turbid plaque formation that it is not simple to distinguish lysogen formation based on plaque turbidity. While the lysogen generally resists superinfection by expression of the phage genes, acquisition of phage resistance by host mutation has been hardly found in lysogen.
Therefore, the aim of this lysogenization experiment was to test whether the resistance against phage infection was due to lysogenization or other factors related with a host receptor. To confirm their lysogen formation, mitomycin C was treated to induce the prophages. Group L phages except for SPN9TCW did yield phage after mitomycin C induction, substantiating the lysogen formation by most of group L phage infection. However, most of group F phage-resistant strains did not yield phage by mitomycin C induction, even though these phages make turbid plaques, suggesting that these phages may make unstable lysogen, resulting in very low frequency of mitomycin C induction (approximately 5%) in the resistant strains. It has been known that Mu-like prophages were generally not induced by mitomycin C treatment [58], suggesting the possibility that group F phage may be mitomycin C-insensitive phage. To verify this, we performed PCR detection of group F phage genomes in the resistant hosts and the green plate experiment. Recently, we sequenced completely the genomes of four phages in group F and phage-specific primers were designed. Using these phage-specific primers, PCR was conducted to detect group F phage genomes in the genomes of group F phage-resistant S. Typhimurium LT2C strains. Interestingly, very low number of group F phage-resistant strains (approximately 5%) showed the presence of group F phage genomes in the host genomes, suggesting the formation of very unstable lysogens, supporting our observation of low mitomycin C induction with group F phage-resistant strains (data not shown). And green plate experiments also showed that most of group F phage infection does not make stable lysogens. We still do not understand this very distinct feature of group F phages for unstable lysogen formation and it needs to be elucidated soon.
Analysis of cross resistance among phage-resistant strains revealed that group F phage-resistant strains are sensitive to group L or B phages (Table 4 and Fig. 4A). This sensitivity indicates that the host resistance of group F phages does not disrupt the interactions between these other phages and the cell surface receptors. In contrast, group B phage-resistant strains are resistant to their phages as well as group L phages (Table 4 and Fig. 4C). The concurrent resistance to both group B and group L phages imply that BtuB may influence the interaction between LPS and phage, as in the case of E. coli phage T5 [45,46]. The interaction between LPS and phage has been reported to accelerate adsorption of phage T5 to E. coli even though an outer membrane protein is the cell surface receptor.
All but one group L phage-resistant strain was sensitive to infection by group B phages (Table 4 and Fig. 4B). The group L SPN9TCW phage-resistant strain was resistant to group B phages (Table 4 and Fig. 4D), even though phage SPN9TCW uses LPS as a receptor and the resistant strain does not appear to be a lysogen. Therefore stable lysogen formation of group L phages may be a key to determine the host resistance against group B phages (Table 4). Further analyses of group B phage-resistant Salmonella strains are required to elucidate the mechanism of cross-resistance observed against the group B and L phages.
Many Salmonella-specific phages that use LPS as a receptor modify LPS as a mechanism to protect from superinfection when they lysogenize a host [59,60]. The recent complete genome sequence analysis of a lysogenic SPN1S phage in group L revealed that the phage genome encodes a GtrA and two copies of lipopolysaccharide modification acyltransferase, supporting this [61]. Even though the LPS modification protects the lysogen from a superinfection by other group L phages, the lysogen is still sensitive to phages that target other receptors such as flagella and BtuB (Table 4 and Fig. 4B). Furthermore, group F, group B and group L SPN9TCW phage-resistant strains showed transient resistance to re-infection of the same phages (Table 4). These resistant strains were collected for the cross resistant experiment after the bacterial growth resumed in the presence of the phage. In this case, these collected phage-resistant strains are probably not lysogens, suggesting that host defense mechanisms such as CRISPR [62,63] or restriction-modification systems [64,65] or still unknown host defense mechanisms are probably activated during growth recovery.

Author Contributions
Conceived and designed the experiments: HS JL SR. Performed the experiments: HS JL HK YC. Analyzed the data: HS JL SH SR. Contributed reagents/materials/analysis tools: JL SH. Wrote the paper: HS JL SR.