Characterization of a Novel Group of Listeria Phages That Target Serotype 4b Listeria monocytogenes

Listeria monocytogenes serotype 4b strains are the most prevalent clinical isolates and are widely found in food processing environments. Bacteriophages are natural viral predators of bacteria and are a promising biocontrol agent for L. monocytogenes. The aims of this study were to characterize phages that specifically infect serotype 4b strains and to assess their ability to inhibit the growth of serotype 4b strains. Out of 120 wild Listeria phages, nine phages were selected based on their strong lytic activity against the model serotype 4b strain F2365. These nine phages can be divided into two groups based on their morphological characteristics and host range. Comparison to previously characterized phage genomes revealed one of these groups qualifies to be defined as a novel species. Phages LP-020, LP-027, and LP-094 were selected as representatives of these two groups of phages for further characterization through one-step growth curve and inhibition of serotype 4b L. monocytogenes experiments. Listeria phages that target serotype 4b showed an inhibitory effect on the growth of F2365 and other serotype 4 strains and may be useful for biocontrol of L.monocytogenes in food processing environments.


Introduction
Listeria monocytogenes is a Gram-positive foodborne pathogen that infects humans and animals [1]. L. monocytogenes is widely isolated from soil, agriculture environments, and urban environments, and can tolerate high salt concentrations and a broad range of temperatures and pH levels [2][3][4][5]. Contamination of the food processing environment with L. monocytogenes can lead to consumers ingesting the pathogen, which can cause the potentially fatal invasive disease listeriosis [6]. The global burden of listeriosis cases has been estimated at 23,150 annually, with a mortality rate of 26% [7]. Within the United States alone, there is an estimated 2518 annual cases, with a 20% mortality rate [8,9], and economic losses caused by listeriosis were more than $3.1 billion in 2018 [10].
Bacteriophages are natural viral predators of bacteria, which infect and lyse specific host strains [11]. High specificity, self-replication capability, and tolerance of a wide range of temperatures and pHs [12] make bacteriophages a promising candidate for biocontrol of L. monocytogenes in the food processing environment [13]. Listeria phage biocontrol products have been approved for use by the United States Food and Drug Administration since 2006. These products are marketed to control L. monocytogenes contamination on food and in food processing plants. Currently used Listeria phages have been characterized as belonging to the genus Pecentumvirus [14]. Pecentumvirus phages have been shown to utilize rhamnose and N-acetylglucosamine of wall teichoic acids as binding receptors during the adsorption step of infection [15]. Presence or absence of these sugars corresponds to the various serotypes of L. monocytogenes [16]; thus, Listeria phages show some level of serotype specificity [17].
Based on cell surface antigenic determinants, L. monocytogenes can be divided into at least 13 serotypes. Serotype 4b strains account for most clinal isolates from humans,

Transmission Electron Microscopy
A purified, high titer phage sample (~1 × 10 10 PFU/mL) was prepared for transmission electron microscopy (TEM) as previously described with modifications [26]. One milliliter of each phage sample was washed using a 0.1 M ammonium acetate solution (pH 7) and centrifuged at 21,000× g with a microcentrifuge (Thermo Fischer Scientific, Waltham, MA, USA). One drop of the phage sample was deposited onto a 150-200 mesh carbon-coated Formvar film copper grid (Electron Microscopy Sciences, Hatfield, PA, USA) and stained using 1% phosphotungstic acid (PTA; pH 7.4). Samples were imaged using a JEOL 1400 Flash transmission electron microscope at 120 kV. Images were analyzed using Fiji 3 v.2.0.0-rc-69/1.52p.

DNA Extraction and Genomic Analysis
Phage DNA were extracted by the phenol-chloroform method as previously described [26].

Efficiencies of Plaquing and Relative Phage Activity
All ten Listeria phages and sixteen L. monocytogenes strains (except MACK) listed in Table 1 were used to conduct efficiency of plaquing (EOP) and relative phage activity (RPA) assays as previously described [25,53]. In brief, bacterial lawns were prepared with the double agar overly method and allowed to solidify. All the phages were amplified from the original phage stock and diluted to 1 × 10 7 PFU/mL as a working stock. Serial dilutions were made from each working stock and spotted onto bacterial lawns. The EOP of each phage was determined from the highest dilution with countable plaques against the strain in question compared to the number of plaques against the phage propagation host strain. Similarly, the RPA of each phage was determined from the highest dilution with observable inhibitory activity against the strain in question compared to the phage propagation host strain. Inhibitory activity is defined as an observable inhibition of the growth of the bacterial lawn where the phage dilution was spotted either with or without the formation of any phage plaques. Three biological replicates were performed. EOP and RPA clustered heatmaps were generated using pheatmap in R [54].

One-Step Growth Curve
An exponential-phase culture of F2365 was infected with LP-020, LP-027, or LP-094 at a multiplicity of infection (MOI) of 0.1. The infected culture was incubated at 25 • C and 160 RPM for 3 h. To measure infected host cells and unabsorbed viable phages, two samples were taken at each time point. One sample was serially diluted and enumerated by the spot assay method immediately after collection, the other sample was treated with 5%(v/v) chloroform for 15 min and then serially diluted and enumerated by the spot assay method, and the plates were incubated at 25 • C for 12 h. Three biological replicates were performed.
2.6. Inhibition Growth Curve of Listeria monocytogenes F2365 by LP-020, LP-027, and LP-094 A measure of 2 mL of L. monocytogenes F2365 overnight culture was added into 100 mL LB-MOPS with 0.1% glucose, 1 mM CaCl 2 , and 1 mM MgCl 2 . The culture was incubated at 25 • C and 160 rpm until the OD 600nm grew to~0.1. The culture was then diluted 10-fold with fresh supplemented LB-MOPS. A measure of 7 mL of the diluted culture was added to twelve sterile 15 mL glass culture tubes. Each tube was infected with LP-020, LP-027, or LP-094 at MOI = 0.1, 1, 10 with SM buffer as a negative control and incubated at 25 • C and 160 RPM for 15 h. The OD 600nm of each tube was measured every half an hour for 15 h on a Genesys 30 Visible spectrophotometer (Thermo-Fisher Scientific, Waltham, MA, USA). Three biological replicates were performed. total of 7 mL of the diluted cocktail was added to eight sterile 15 mL glass culture tubes. Each tube was infected with LP-020 or LP-094 with MOI = 0.1, 1, 10 or SM buffer as a negative control, grown at 25 • C, and shaken at 160 RPM for 15 h. The OD 600nm of each tube was measured every half an hour for 15 h. Three biological replicates were performed.

Transmission Electron Microscopy Imaging of Wild Type Listeria Phages Revealed Two Distinct Morphologies
Based on morphological characteristics, all phages were classified within the Siphoviridae family ( Figure 1). One group that includes two phages, LP-024 and LP-027, were found to have icosahedral capsids with flexible, elongated tails ( Table 3). The second group also had icosahedral capsids; however, these phages were found to have short flexible tails (Table 3).

Transmission Electron Microscopy Imaging of Wild Type Listeria Phages Revealed Two Distinct Morphologies
Based on morphological characteristics, all phages were classified within the Siphoviridae family ( Figure 1). One group that includes two phages, LP-024 and LP-027, were found to have icosahedral capsids with flexible, elongated tails ( Table 3). The second group also had icosahedral capsids; however, these phages were found to have short flexible tails (Table 3).

Genomic Analysis
All phage reads assembled into complete, single-contig genomes; genome statistics are presented in Table 4. Based on genome statistics, two distinct groups were evident. LP-024 and LP-027 were 41.0-41.4 kb, with G+C contents of 36.5-36.6%, and each contained 74 coding sequences (CDS) and no RNAs. The other phage genomes were 35.6-36.0 kb, with G+C contents of 39.9-40.0%, and each contained 54-57 CDS and no RNAs. The two distinct groups were also supported by average nucleotide identity (ANI) values (Tables 5 and 6) and amino acid similarity ( Figure S1). LP-024 and LP-027 had an ANI of 100.00% over 95.51-96.63% of their genomes and were most similar to LP-030-3 (99.99-100.00% ANI over 95.92-96.63% of their genomes) ( Table 5), a putative temperate phage. LP-030-3 was previously classified as an Orthocluster IV siphovirus [41], with a 41.2 kb genome containing 73 predicted genes. Electron micrographs show that LP-030-3 has a long and rigid tail [41]. Morphology and genome features of LP-030-3 are consistent with the LP-027-like phages and are likely putative temperate phage [41]. These two phages are likely the same genus and species as LP-030-3, as they are above the 50% and 95% cutoffs for genus and species delineation. LP-030-3 is currently listed as a "unclassified Siphoviridae", with no genus classification, on NCBI (NCBI:txid1458852) and is not included in the most recent ICTV Master Species List 2019.v1. The other phages had an ANIs of 97.66-100.00%, over 90.63-99.24% of their genomes, and were most similar to P35 (79.51-80.36% ANI over only 83.03-87.67% of their genomes) (Table 6, Figure 2), a lytic phage [41]. P35 was previously classified as an Orthocluster II siphovirus [41], with a 35.8 kb genome containing 56 predicted genes. Electron micrographs show that P35 has a short tail [41,55]. Morphology and genome features of P35 are consistent with the LP-020-like phages and they are likely putative obligate lytic phages [41,55]. Given the >50% nucleotide similarity to P35, this second group of phages likely belongs to the same genus [56]. However, they clearly qualify as a novel species, as they are well below the 95% similarity cutoff to be considered the same species as P35. P35 is currently listed as a "unclassified Siphoviridae," with no genus classification, on NCBI (NCBI:txid330398) and is not included in the most recent ICTV Master Species List 2019.v1. LP-024 and LP-027 were confidently predicted as having a temperate lifestyle by PHACTS (Table S1) and both genomes contained a phage integrase gene. This was further confirmed through BLAST, which showed they have high similarity to L. monocytogenes genomes (up to 99.79% identity and 86% query coverage). LP-020 and LP-057 were nonconfidently predicted as having a lytic lifestyle and LP-021, LP-053, LP-054, LP-085, and LP-094 were non-confidently predicted as having a temperate lifestyle (Table S1). However, none of these genomes contained an integrase gene, and BLAST showed they did not have high similarity to any L. monocytogenes genomes (only up to 5% query coverage), indicating that they likely have a lytic lifestyle.   LP-024 and LP-027 were confidently predicted as having a temperate lifestyle by PHACTS (Table S1) and both genomes contained a phage integrase gene. This was further confirmed through BLAST, which showed they have high similarity to L. monocytogenes

Host Range Analysis
A511, which has been described as a broad host range phage [28], showed activity against 11 out of 17 L. monocytogenes ( Figure 3A, Figure S2), but only formed visible plaques on seven strains ( Figure 3B, Figure S2). Similarly, two previously characterized phages, LP-048 and LP-125, showed activity against a broad range of strains (thirteen and nine strains, respectively), compared to their ability to form plaques on the same strains (nine and four strains, respectively). All three of these Pecentumvirus phages were unable to infect all serotype 4b and 4a strains. nine strains, respectively), compared to their ability to form plaques on the same strains (nine and four strains, respectively). All three of these Pecentumvirus phages were unable to infect all serotype 4b and 4a strains.
The phages selected for this study that showed strong activity against serotype 4b strain F2365 can be divided into two groups, and these groupings are consistent with those based on observed morphological characteristics. The first group, comprised of LP-024 and LP-027, showed a very narrow host range, forming plaques on F2364 (4b), FSL F2-501 (4b), and FSL J1-208 (4a). The second group, comprised of LP-020, LP-021, LP-053, LP-054, LP-057, LP-085, and LP-094, showed activity to all L. monocytogenes serotype 4 strains, and formed visible plaques on nearly all L. monocytogenes serotype 4 strains. Interestingly, LP-020 is the only phage in this group that was able to form plaques on non-serotype 4 strains.
Based on the combination of their morphological characteristics and the results of their EOP and RPA, LP-020, LP-027, and LP-094 were selected as representative phages for further evaluation in this study.  The phages selected for this study that showed strong activity against serotype 4b strain F2365 can be divided into two groups, and these groupings are consistent with those based on observed morphological characteristics. The first group, comprised of LP-024 and LP-027, showed a very narrow host range, forming plaques on F2364 (4b), FSL F2-501 (4b), and FSL J1-208 (4a). The second group, comprised of LP-020, LP-021, LP-053, LP-054, LP-057, LP-085, and LP-094, showed activity to all L. monocytogenes serotype 4 strains, and formed visible plaques on nearly all L. monocytogenes serotype 4 strains. Interestingly, LP-020 is the only phage in this group that was able to form plaques on non-serotype 4 strains.
Based on the combination of their morphological characteristics and the results of their EOP and RPA, LP-020, LP-027, and LP-094 were selected as representative phages for further evaluation in this study.

One-Step Growth Curves
LP-020 was found to have a short adsorption time, in which 97.9% of LP-020 adsorbed to F2365 in 5 min. This is in contrast with LP-027 and LP-094, in which 30 min was required to attain adsorptions of 96.4% and 97.8%, respectively. One-step growth curve analysis of LP-020 showed a latent period of 60~90 min, an eclipse period of 5~15 min, and a burst size of~9.7 (SE, 2.9) PFU/cell (Table 7, Figure 4A). One-step growth curve analysis of LP-027 showed a latent period of 45~60 min, an eclipse periods were 30~45 min, and a burst size of~34.4 (SE, 5.8) PFU/cell (Table 7, Figure 4B). One-step growth curve analysis of LP-094 showed a latent period of 30~45 min, an eclipse period of 30~45 min, and a burst size of 28.3 (SE, 4.1) PFU/cell ( Table 7, Figure 4C).    However, the one-step growth curve of LP-020 showed a fluctuation in completed phage numbers between 5 min and 30 min ( Figure 4A). Interestingly, after 12 h of incubation, the sample infected with LP-094 at MOI = 1 has a higher OD 600nm than the sample infected with LP-094 at MOI = 0.1, possibly due to the lower MOI (0.1) providing an opportunity for host to continue during early infection, leading to greater levels of phage production during later infection ( Figure 5C).

Inhibition Growth Curve of Listeria monocytogenes Cocktail by LP-020 and LP-094
LP-020 was able to inhibit the growth of a cocktail of L. monocytogenes serotype 4 at all tested MOIs, while LP-094 only showed an inhibitory effect with the highest MOI ( Figure 6). For the first 8 h after LP-020 incubation, the growth curve of the cocktail showed no substantial difference. The growth curve of the cocktail treated with different dosages of LP-020 reached its peak at 11 h, 10 h, and 9 h, respectively. After reaching their peak, the bacterial cocktails treated with LP-020 showed a reduction in OD 600nm , with the 1 and 10 MOI infections reducing the OD 600nm to 0.06 and 0.08, respectively. Interestingly, the low MOI treatment group (MOI = 0.1) showed a large variation in OD 600nm , which may be due to the diversity of the host strains in the cocktail generating a more varied response to phage infection as the individual concentrations of each L. monocytogenes strain could vary from replicate to replicate due to complex growth interactions ( Figure 6A). For the LP-094 infection, low MOI (MOI = 0.1, 1) showed no major influence of the growth curve of the cocktail. However, LP-094 can keep the OD 600nm of the cocktail under 0.1 for 8 h with a high dosage (MOI = 10) ( Figure 6B). Compared with LP-094, LP-020 s ability to reduce the OD 600nm of a cocktail of serotype 4a, 4b, and 4c Listeria strains suggests it could be a promising phage for biocontrol applications in food processing environments. Viruses 2021, 13, x FOR PEER REVIEW 11 of 15

Conclusions
In this study we describe two groups of Listeria phages that showed high levels of infectivity against serotype 4 strains of L. monocytogenes. One of these groups of phages, LP-020-like phages showed a level of nucleotide dissimilarity with previously sequenced phages that is well above the 5% cutoff for qualifying as a novel species. Inhibition assays of L. monocytogenes against a cocktail of serotype 4 strains (4a, 4b, and 4c) confirmed that EOP and RPA assays were predictive of a phage's inhibitory potential. We identified LP-020 as a phage that may be particularly useful in biocontrol settings. It shows strong activity against all serotype 4a, 4b, and 4c strains tested here. Given that some types of phage resistance is specific to a single species or closely related group of phages, such as CRISPR-Cas mediated resistance, it is critical to consider using diverse cocktails of phages. LP-020 may serve together with Listeria phages from other groups, such as Pecentumviruses

Conclusions
In this study we describe two groups of Listeria phages that showed high levels of infectivity against serotype 4 strains of L. monocytogenes. One of these groups of phages, LP-020-like phages showed a level of nucleotide dissimilarity with previously sequenced phages that is well above the 5% cutoff for qualifying as a novel species. Inhibition assays of L. monocytogenes against a cocktail of serotype 4 strains (4a, 4b, and 4c) confirmed that EOP and RPA assays were predictive of a phage's inhibitory potential. We identified LP-020 as a phage that may be particularly useful in biocontrol settings. It shows strong activity against all serotype 4a, 4b, and 4c strains tested here. Given that some types of phage resistance is specific to a single species or closely related group of phages, such as CRISPR-Cas mediated resistance, it is critical to consider using diverse cocktails of phages. LP-020 may serve together with Listeria phages from other groups, such as Pecentumviruses [11,57,58] and Homburgviruses [26,59,60], to increase both the diversity of Listeria phage cocktails and their infectivity against the problematic serotype 4 strains that are often associated with human illness. Further, the temperate phages described here, which are similar to LP-030-3, may be useful as biocontrol agents if their integrases are knocked out, which has previously been shown for the temperate Listeria phage PSA [61].

Conflicts of Interest:
The authors declare no conflict of interest.