Bioprospecting Staphylococcus Phages with Therapeutic and Bio-Control Potential

Emergence of antibiotic-resistant bacteria is a serious threat to the public health. This is also true for Staphylococcus aureus and other staphylococci. Staphylococcus phages Stab20, Stab21, Stab22, and Stab23, were isolated in Albania. Based on genomic and phylogenetic analysis, they were classified to genus Kayvirus of the subfamily Twortvirinae. In this work, we describe the in-depth characterization of the phages that electron microscopy confirmed to be myoviruses. These phages showed tolerance to pH range of 5.4 to 9.4, to maximum UV radiation energy of 25 µJ/cm2, to temperatures up to 45 °C, and to ethanol concentrations up to 25%, and complete resistance to chloroform. The adsorption rate constants of the phages ranged between 1.0 × 10−9 mL/min and 4.7 × 10−9 mL/min, and the burst size was from 42 to 130 plaque-forming units. The phages Stab20, 21, 22, and 23, originally isolated using Staphylococcus xylosus as a host, demonstrated varied host ranges among different Staphylococcus strains suggesting that they could be included in cocktail formulations for therapeutic or bio-control purpose. Phage particle proteomes, consisting on average of ca 60–70 gene products, revealed, in addition to straight-forward structural proteins, also the presence of enzymes such DNA polymerase, helicases, recombinases, exonucleases, and RNA ligase polymer. They are likely to be injected into the bacteria along with the genomic DNA to take over the host metabolism as soon as possible after infection.


Introduction
The WHO considers the rapid emergency of Multi-Drug-Resistant Bacteria (MDRB) as a threat to global public health [1]. It is also predicted that MDRB-related infections will cause about 10 million deaths annually by 2050 [2]. In addition, MDRB are predicted to become a major burden to the global economy as the World Bank estimates that by the year 2050 the bacteria might result in an annual gross domestic production (GDP) loss worth 120 trillion US dollars [3]. The predicted reduction in GDP will be due to increased morbidity and mortality among the workforce, and livestock loss caused by infectious MDRB pathogens such as Staphylococcus aureus. S. aureus is a pathogen that is an etiological agent of bacteremia, soft skin and tissue infections, osteomyelitis, endocarditis, meningitis, haematogenous organ infections, food poisoning, and toxic shock syndrome in humans [4,5]. In the livestock and pet industry, staphylococci are associated with arthritis and comb necrosis in of staphylococci and phage isolations were performed at 37 • C using Lysogeny Broth (LB) [33]. The bacteriological agar (Lab M limited, Lancashire, UK) content was 0.3% (w/v) for soft agar and 1.5% (w/v) for solid medium.

Phage Isolation and Purification
The isolation of phages Stab20, Stab21, Stab22, and Stab23 from Albanian sewage and river water samples using S. xylosus strain DD-34 as host bacteria was described in [31]. The accession numbers of the DNA sequences of the genomes are LR215718, LR215719, LR215720 and LR215721, respectively. The phages were propagated as described previously [21]. Semi-confluent solid media propagation method was used to prepare high titer stocks. Briefly, 200 µL of 90 min cultured host bacteria at OD 600 1-1.5, and 50 µL of appropriately diluted phage suspension were added to 3.0 mL LB overlay medium (0.3% agar) supplemented with 5 mM CaCl 2 . The phages were recovered from the overlay agar as described [34]. The phages were resuspended into SM-buffer with 8% sucrose after concentrating and washing thrice, and stored at +4 • C.

Transmission Electron Microscopy (TEM)
Phage suspensions (>10 7 pfu/mL) were centrifuged 90 min at full speed in Eppendorf microfuge and the phages resuspended into 200 µL of 0.1 M ammonium acetate. Three µL of phage suspension was pipetted on carbon-coated copper grids, and after 60 sec adsorption the grids were stained with 2% uranyl acetate (pH 7.4) for 15 s. The grids were then observed using the JEOL JEM-1400 TEM (Jeol Ltd., Tokyo, Japan) fitted with a bottom-mounted Gatan Orius SC 1000B camera (Gatan Inc., Pleasanton, CA, USA). The specimens were inspected at 80 KV beam voltage with 80,000× and 150,000× magnifications at Electron Microscopy Unit (Institute of Biotechnology, University of Helsinki-Finland, Helsinki, Finland). The dimensions of five to ten virions were determined and the measurements used to calculate the averages and standard errors.

Adsorption Rate Experiment
In the adsorption rate experiment log-phase host bacteria and pre-determined number of phages were used. Briefly, S. xylosus DD-34 bacteria were sub-cultured to 5 mL of LB and incubated at 37 • C to OD 600 of 0.5 to 1.0. Then the bacteria were pelleted by centrifugation at 4500× g for 20 min and resuspended into 0.9 mL of fresh LB. Thereafter, 100 µL of phage suspension (4.5 × 10 6 pfu) was added to experimental tube (A) and to a control tube (B) without bacteria. The two tubes were then incubated at 37 • C, 120 rpm for 10 min and sampling done at an interval of 5 min from tubes A and B. 50 µL was picked at each interval and dispensed into pre-chilled Eppendorf tubes. The samples were briefly vortexed then centrifuged at 16,100× g at +4 • C for 10 min. The numbers of free phages in 50 µL of the supernatant were determined by double-layer plaque assay. Plaques were counted from all the plates and number of plaques recorded at their respective time points (from 0 min to 10 min). Plaque counts from control tubes (tube B) were used as time point 0 min reference points. The values were normalized by having the average PFU of tube B representing 100%. The adsorption rate constants (k-values) were calculated for 5 min time points as described [36].

One Step Growth Curves
One step growth curve experiments were carried out as described elsewhere [37]. The plaques were counted from each plate and recorded as per corresponding time points (5,10,15,20, and every 10 min until 60 min). The experiment was repeated five to ten times for each phage (Stab20, Stab21, Stab22, and Stab23) on different days.

SDS-PAGE and LC/MS-MS Analysis
The Stab phages were concentrated by centrifugation for 30 min at 4 • C and 5000 rpm using 100,000 kDa cutoff Vivaspin concentrator ® 20 [38], and further purified by a 5/40% glycerol step gradient centrifugation as described elsewhere [39]. After suspension of the pelleted phages each had a titer > 6 × 10 10 pfu/mL. The concentrated phage stocks were diluted appropriately with RNAse free H 2 O and 20 µL mixed with 20 µL of 2× Laemmli buffer. The mixtures were heated at 100 • C for 5 min and then cooled on ice. Ten µL aliquots of the samples were analyzed by 10% SDS-PAGE. The protein bands were stained for 3 hr using the InstantBlue™ ready-to-use Coomassie protein stain. The excess stain was washed with milli-Q water and the gel image taken with the Bio-Rad XR+ gel documentation system.
Phage particle proteomes were analyzed by liquid chromatography coupled with mass spectrometry (LC-MS/MS) at the Proteomics Unit, Institute of Biotechnology, University of Helsinki. Stab phages with a titer > 6 × 10 11 pfu/mL were used for the analysis. Prior to digestion of proteins to peptides with trypsin, the proteins in the samples were reduced with tris (2-carboxyethyl) phosphine (TCEP) and alkylated with iodoacetamide. Tryptic peptide digests were purified by C18 reversed-phase chromatography columns [40] and the mass spectrometry (MS) analysis was performed on an Orbitrap Elite Electron-Transfer Dissociation (ETD) mass spectrometer (Thermo Scientific, Waltham, MA, USA), using Xcalibur version 2.2, coupled to a Thermo Scientific nLC1000 nanoflow High Pressure Liquid Chromatography (HPLC) system. Peak extraction and subsequent protein identification were achieved using Proteome Discoverer 1.4 software (Thermo Scientific). Calibrated peak files were searched against the Stab20, Stab21, Stab22 and Stab23, and Staphylococcus aureus subsp. aureus ST398 proteins (ASM188707v1, NCBI) by a SEQUEST search engine. Error tolerances on the precursor and fragment ions were ± 15 ppm and ± 0.8 Da, respectively. For peptide identification, a stringent cut-off (0.05 false discovery rate or 5%) was used.

Host Range Testing
The host ranges of the Stab phages were determined using 100 Staphylococcus strains representing S. aureus, S. epidermidis, S. saprophyticus, and S. haemolyticus (Table S1). Strains to be screened for susceptibility were grown in LB medium for 90 min at 37 • C, 120 rpm to an OD 600 of 1-1.5. Warm 3.0 mL soft agar (0.3%) LB cooled to 50 • C was mixed with 0.15 mL of the bacterial suspension, poured evenly on pre-warmed 1.5% LB agar plates and allowed to solidify. The Stab phage stocks were serially diluted and 4.0 µL of each dilution was administered onto the solidified soft agar. The plates were incubated at overnight and the strains giving a positive spot assay result were tested for relative efficiency of plating (REOP). Briefly, similar aliquots of 10 −5 to 10 −6 dilution of each Stab phage were parallelly plated in the soft agar with the S. xylosus DD-34 indicator bacteria and the test strain. After overnight incubation at 37 • C, REOPs were calculated by dividing the resulting plaque counts of the test strains with those of the indicator bacteria.

Statistical Analysis
The physico-chemical (thermal, pH and U.V stability), adsorption and one step growth curve experiment data were analyzed by Prism GraphPad statistical tools [41]. The comparative analysis on the stability of Stab phages was carried out using the 2way ANOVA accompanied with Bonferroni post-tests at 95% and 99% confidence intervals.

Results
Our rationale to characterize the Stab phages originally isolated using S. xylosus as enrichment host [31] was based on the facts that Staphylococcus phages generally show wide host spectra, that new S. aureus specific phages are not easy to encounter, and that every phage able to infect clinical S. aureus isolates would be a welcome addition to our collection of potential therapeutic phages.

Morphology
In our previous study, genome sequence and phylogenetic analyses of phages Stab20, Stab21, Stab22, and Stab23, established that the phages are members of the genus Kayvirus which exclusively contains Staphylococcus phages [31]. Electron microscopy revealed that these phages possess icosahedral heads, long contractile tails with baseplates at the end, and tail fibers extending from the baseplates (Figure 1). The dimensions of the phage particles were close to each other but clearly distinct ( Table 1). The dimensions of phages Stab20, Stab21, Stab22 and Stab23 resemble those of other genus Kayvirus members [36,42,43].

Physico-Chemical Stability
The stability of the phages varied a little when exposed to the different environmental conditions including ultra-violet (UV) irradiation, temperature, pH and exposure to organic solvents (ethanol and chloroform). There was significant reduction of phage titer (p < 0.0001) when phages were exposed to 75 µJ/cm 2 of UV-irradiation or incubated at temperatures above 45 • C. Increase in acidity or alkalinity had negative impact on the phage viability. Each phage was inactivated below pH 5.4 or above pH 9.4 (Figure 2a-c). Ethanol concentration above 25% vol/vol was enough to inactivate all four phages while they all were resistant to chloroform (Table S2). The phages were stable in 100% chloroform indicating the absence of lipids in the phage particles.

Adsorption Rate and Growth Curves
The Stab phages further displayed their distinct nature through their growth curves that reflect varied adsorption rates and burst sizes. The adsorption curves represent the rates at which the phages attach to its host, also known as adsorption kinetics [44]. Of the phages, Stab21 adsorbed rapidly, ca 90% was adsorbed in 5 min while only 40, 60 and 70% of Stab20, Stab22 and Stab23, respectively, had adsorbed in the same time ( Figure 3). The adsorption rate constants calculated for the 5 min time point did not differ significantly between the phages. Each phage also had its unique one step growth curve characterized by varied latent and lag phase periods. The apparent latency periods were 25-30 min. The burst sizes varied between 42 and 130 ( Figure 4).

Analysis of Stab Phage Particle Proteomes
Previous in silico genome analysis of Stab phages predicted the presence of several structural proteins [31]. SDS-PAGE and mass spectrometry (LC-MS/MS) were used to identify the structural and phage particle-associated proteins of Stab20, Stab21, Stab22 and Stab23. The banding patterns of the phage particle proteins in SDS-PAGE analysis were very similar reflecting well the genomic similarities ( Figure 5). In SDS-PAGE the prominent bands represent the major proteins with highest copy numbers in the phages such as capsid and tail sheath proteins. Bands in the 50-55 kDa range very likely represent the predicted major capsid proteins of Stab20, Stab21, Stab22 and Stab23 that have calculated molecular masses of about 51.5 kDa. The phages also have putative 50.4 kDa capsid and scaffold proteins that would co-migrate with the major capsid protein. Indeed, close inspection of the 50-55 kDa bands of Stab20 and Stab23 in the gel reveals that they might be formed of two overlapping bands. The predicted tail tape measure proteins of the Stab phages are uniform with calculated molecular weights of 143.1 to 143.9 kDa ( Figure 5) with almost identical amino acid sequences (Table S3). The tail tape measure protein determines the length of the tails, but also controls the injection of phage DNA into the host bacteria and may possess muramidase for piercing the bacterial cell-wall. The Staphylococcus phage K has a 1351 amino acid residue long tape measure protein, only two residues longer than that of Stab22.
The identities of the proteins were further studied using the LC-MS/MS analysis that revealed, based on the inclusion criteria of at least two identified tryptic peptides, the presence of altogether 46 -100 phage particle-associated proteins. In Table S3, to facilitate comparisons between the phages, we have listed all the identified Stab phage proteins, with orthologues in the same rows, in parallel columns based on their decreasing molecular masses. The different total numbers of particle-associated proteins identified for the Stab phages likely reflects differences in the quantities of phage particles in the LC-MS/MS samples, with that of Stab23 being lowest. Table S3 reveals that proteins >20 kDa are most reliably identified from all phages while smaller proteins tend to be absent from some phages. Based on the results we can assign, based on experimental evidence, particle-associated/structural protein functions to 41 proteins originally annotated as hypothetical proteins (Table S4). The genomic locations of the genes encoding the phage structural and particle-associated proteins identified in the LC-MS/MS analysis are shown for phage Stab21 in Figure 6. The structural genes appear to be organized in several operon-like clusters.
In addition to proteins already annotated as structural proteins, several enzymes or proteins with other functions were also detected by LC-MS/MS (Tables S3 and S4). These included DNA polymerase (Stab21 gene g145, Figure 6), glycerophosphoryl diester phosphodiesterase (g193) likely involved in cell wall teichoic acid hydrolysis [45,46], nicotinamide phosphosribosyltransferase (g207) that augments nucleotide synthesis [47], ribonucleotide reductases (g138-g140) that are also involved in dNTP synthesis [48], endo-and exonucleases (g130, g132) playing role in host DNA degradation and provision of nucleoside 5 -monophosphate precursors for synthesis of phage DNA progeny [49]. The sigma factor (g151) will quide the host RNA-polymerase to start transcription from phage promoters. Also present were DNA helicases (g127, g129) that are significant in DNA synthesis and are involves in the unwinding of the double stranded DNA to create templates for replication of DNA [50], and ribose-phosphate pyrophosphokinase (g205) that is engaged in nucleotide salvage needed for phage replication [51]. The RNA ligase (g064c) was also present in the phage particles, likely involved in the repair, splicing, and editing pathways that either reseal broken RNAs or alter their primary structure [52]. The role of the PhoH-related protein (g066c) or PhoH-predicted ATPase is not as obvious, as they are reported to regulate phosphate uptake and metabolism under phosphate limitation [53]. GTP cyclohydrolase II has been found to be significant in riboflavin metabolism as a catalyst [54]. Furthermore, the metallophosphoesterases present in the phage particles could be involved in the catabolism of phosphodiester bonds of the host DNA to scavenge for phosphates needed in early protein biosynthesis of the phage. The lipase acylhydrolase domain protein (g038c) was only present in Stab21, encoded by the g038c gene. While no orthologous genes were present in the other Stab phages, Blastp search identifies several other Staphylococcus phages carrying the gene. The role of the enzyme could be in peptidoglycan/cell wall degradation thus facilitating the entry of phage DNA into the bacterial cell [55]. The proteomic data presented for the Stab phages is comparable to that of other Staphylococcus phages studied earlier, such as fRuSau02 for which 78 phage particle-associated proteins, including similar enzymes as for the Stab phages, were identified [30].

Host Range Analysis
The host ranges of the Stab phages were determined as relative efficiency of plating (REOP) when compared to the isolation host S. xylosus DD-34 (Table S1). The different Staphylococcus strains (n = 100) were isolates from both human and animal sources including MRSA (N = 46; 30 from pigs and 16 from humans), MSSA (n = 38), S. intermedius (n = 3), S. epidermidis (n = 4), S. saprophyticus (n = 5) and S. haemolyticus (n = 5). Phages Stab20 and Stab21 showed broadest host ranges. They infected 41 and 40 of the 100 strains, respectively, including both human and pig MRSA and MSSA strains and a few S. haemolyticus, S. epidermidis, and S. saprophyticus strains (Table S1). In contrast, Stab22 and Stab23 had very narrow host ranges, as both phages infected only one S. saprophyticus strain, and Stab 23, additionally, one S. aureus strain (Table S1). Interestingly, Stab20 and Stab21 propagated best in some S. aureus strains in which they showed REOPs of 2.4 and 2.1, respectively.

Conclusions
In this work, we have presented detailed characterization of four Staphylococcus phages that our previous report identified, based on genome analysis, as novel myoviruses free of any unwanted potentially harmful (toxin, antibiotic resistance, or virulence factor encoding) genes [31]. The fact that these phages can be propagated in an approved food fermenter host, S. xylosus strain DD-34 [32] is advantageous as such host bacteria would be safe to use to propagate phages for therapeutic or bio-control purposes [56]. A potential problem with clinical strains as propagation hosts is that they are likely to contain prophages and may produce enterotoxins, therefore, it is safer to propagate the phages in a food-quality S. xylosus. Even though the Stab phages had distinct host ranges, with Stab20 and Stab21 having a wide and Stab22 and Stab23 a narrow host range, their genomes proved to be closely related to other phages of therapeutic and biocontrol significance such as phage Sb-1 [31,57]. Proteomic analysis depicted uniformity among some of their structural proteins with the Staphylococcus phage K and other members of Kayvirus genus [58]. These Stab phages were stable at temperatures below 40 • C, between pH 5.4-9.4, and could resist U.V energy up to 25 µJ/cm 2 . In addition, they were resistant to chloroform but stable up to 25% ethanol.
The ancient battle for survival between bacteria and antibiotics or bacteriophages is never going to end. Continuing the use of old antibiotics and phages will not end the bacterial menace since the microbes can mutate and develop resistance to both agents [59]. However, the commercial development of novel antibiotics is both an expensive and time-consuming process often with limited scope for industrial profitability [14]. Subsequently, the search for new phages with therapeutic potential against MDR bacteria has turned into a promising alternative. Compared to discovery, development, and production of antibiotics, therapeutic phages are easy and cheap to produce [60]. In summary, these findings suggest that the Stab phages reported here may be useful candidates for phage cocktails of therapeutic or biocontrol significance. Indeed, phage Stab21 was the only phage able to infect an S. aureus strain recently isolated from a patient with chronic sinusitis.