Abstract
Phages infecting bacteria of the genus Staphylococcus play an important role in their host’s ecology and evolution. On one hand, horizontal gene transfer from phage can encourage the rapid adaptation of pathogenic Staphylococcus enabling them to escape host immunity or access novel environments. On the other hand, lytic phages are promising agents for the treatment of bacterial infections, especially those resistant to antibiotics. As part of an ongoing effort to gain novel insights into bacteriophage diversity, we characterized the complete genome of the Staphylococcus bacteriophage Metroid, a cluster C phage with a genome size of 151kb, encompassing 254 predicted protein-coding genes as well as 4 tRNAs. A comparative genomic analysis highlights strong similarities – including a conservation of the lysis cassette – with other Staphylococcus cluster C bacteriophages, several of which were previously characterized for therapeutic applications.
Pathogens of the genus Staphylococcus, known for their ability to evade the human immune system, are an important public health concern causing a multitude of community-acquired infections ranging from food poisoning to skin lesions and life-threatening sepsis (Pollitt et al. 2018). As Staphylococcus largely reproduces clonally, much of the genetic diversity among strains stems from horizontal gene transfer through bacteriophages. Thereby, the acquisition of novel genes may not only aid adaptation of a bacterial strain to novel environments (Xia and Wolz 2014), but it can also increase pathogenicity. Bacteriophages play an important role in bacterial pathogenesis (Deghorain et al. 2012) as they encode for many known staphylococcal virulence factors (see review by Malachowa and DeLeo 2010). Moreover, bacteriophages can mediate the mobilization and transfer of genomic pathogenicity islands (Xia and Wolz 2014). On the other hand, virulent bacteriophages, which lyse their host cell after successful reproduction, also represent promising new avenues for the treatment of antibiotic-resistant Staphylococcus infections through phage therapy (Moller et al. 2019).
Approximately 1030 bacteriophages are estimated to exist on our planet (Rohwer 2003), however much of their diversity remains under-sampled and therefore uncharacterized. Several Staphylococcus phages (order: Caudovirales; i.e., tailed dsDNA phages) have been isolated and sequenced (e.g., Kwan et al. 2005; Deghorain et al. 2012; Oliveira et al. 2019). Historically, Staphylococcus phages were grouped according to their lytic activity and serology; specifically, their reaction to (among others) polyclonal antiserum (Rountree 1949; Rippon 1952, 1956). In contrast, modern phage classification systems are based on either: 1) morphology (determined using transmission electron microscopy), categorizing Myoviridae (long, contractile tail; group A), Siphoviridae (long, non-contractile tail; group B), and Podoviridae (short tail; group C) (Ackermann 1975; Brandis and Lenz 1984); 2) genome size, categorizing class I (<20kb), class II (∼40kp), and class III (>125kb) (Kwan et al. 2005); or 3) gene homology (Goerke et al. 2009; Kahánková et al. 2010; McCarthy et al. 2012), with phages of like category generally being more closely related to one another (Kwan et al. 2005). In one of the largest Staphylococcus phage genomic studies published to date, Oliveira et al. (2019) used a comparative evolutionary approach to group Staphylococcus phages according to their gene content: cluster A (morphologically Podoviridae; genome size: 16-18kb), cluster B (a diverse cluster consisting of mostly temperate phages; genome size: 39-48kb), cluster C (morphologically Myoviridae; genome size: 127-152kb), and cluster D (morphologically Siphoviridae; genome size: 89-93kb). Based on predicted sequence similarities of protein families (phams), the authors further subdivided Staphylococcus phages into 27 subclusters (A1-A2, B1-B17, C1-C6, and D1-D2), members of which exhibit similar morphology and genomic features (i.e., genome size, GC-content, and number of genes; Oliveira et al. 2019). In contrast to the usually temperate Siphoviridae, most Myoviridae and Podoviridae experimentally characterized to date exhibit a lytic life cycle. Lytic phages destroy their host cells, making them interesting candidates for phage therapy (Xia and Wolz 2014).
Here, we report the complete genome sequence of the Staphylococcus bacteriophage Metroid, a Myoviridae sequenced as part of HHMI’s SEA-PHAGES program – an ongoing effort to systematically characterize bacteriophages and their relationship to their (often pathogenic) bacterial hosts. A comparative genomic analysis highlights strong similarities with other Staphylococcus cluster C bacteriophages, several of which were previously characterized for therapeutic applications (Vandersteegen et al. 2011; Gill 2014; Leskinen et al. 2017; Ajuebor et al. 2018; Philipson et al. 2018).
Materials And Methods
Sample collection, isolation, purification, amplification, and phage characterization followed the HHMI SEA-PHAGES Phage Discovery Guide (https://seaphagesphagediscoveryguide.helpdocsonline.com/home; last accessed 2020/04/30), with modifications indicated below. Media and reagent preparation followed the HHMI SEA-PHAGES recipe cards in the Phage Discovery Instructors Guide (https://phagediscoveryinstructorguide.helpdocsonline.com/appendix-b-recipe-cards; last accessed 2019/11/30). Library preparation, sequencing, assembly, and gene annotation followed the HHMI SEA-PHAGES Phage Genomics Guide (https://seaphagesbioinformatics.helpdocsonline.com/home; last accessed 2020/04/30).
Sample collection and isolation
To locate phage, ∼50 soil samples were collected from various locations in Arizona and plaque assays were performed on the sample filtrates. Most samples did not produce phage that could infect the host bacteria. The sample that produced Metroid was collected from a shaded and well-irrigated garden on Arizona State University’s Tempe campus (33.417708N, 111.935974W; ambient temperature 37.7°). The soil was loosely packed into half of a 15 mL conical tube and stored at 4° until phage isolation and a plaque assay were performed. In order to isolate bacteriophages, the sample was submerged in 10 mL PYCa liquid media (1 g/L of yeast extract, 15 g/L of tryptone, 4.5 mM CaCl2, 0.1% dextrose, 10 μg/mL cycloheximide), vortexed for one minute, and placed in a shaking incubator at room temperature for 30 min. This sample was then centrifuged at 4500 rpm for four minutes and filter-sterilized with a 0.22 μm syringe filter. A 250 μL sample of this filtrate was mixed with 250 μL of host bacteria. The host bacteria was isolated as a contaminant from frozen cultures of Arthrobacter globiformis. We suspect it to be of the genus Staphylococcus given that it possesses phage known to reside in this genus. Before mixing with the filtered soil sample, the host bacteria had been grown to saturation in PYCa and stored at 4°. After a ten minute incubation at room temperature, the 500 μL of phage plus bacteria was added to 4.5 mL molten PYCa top agar (60°) and immediately plated on a PYCa agar plate which was incubated for 48 hr at 37°.
Purification and amplification
Clear plaques appeared on the PYCa plates after 48 hr and were ∼3 mm in diameter. One plaque was picked with a sterile pipette tip, and phage were resuspended in phage buffer (10 mM Tris, 10 mM MgSO4, 68 mM NaCl, ddH2O, 1 mM CaCl2), and a series of six 10-fold serial dilutions were performed. Each dilution was inoculated with 250 µL of host bacteria and incubated at room temperature for ten minutes. Each dilution was plated with 4.5 mL PYCa top agar and incubated at 37° for 48 hr. A plaque from the plate representing the 10−2 dilution was selected to complete two additional rounds of purification through subsequent dilutions and plaque assays. For each purification, we chose to pick plaques from a ‘countable’ plate, on which plaques were separated enough to suggest that each grew from a single phage particle (typically a countable plate had 30 to 300 plaques).
Once purified, we amplified the phage to obtain a titer greater than 1x109 PFU/mL which would provide enough DNA for genome sequencing. A plate containing numerous purified phage plaques was flooded with 8 mL of phage buffer and set at room temperature for an hour to yield a phage lysate. The lysate was collected in a 15 mL tube and centrifuged at 8000 rpm for four minutes then filtered through a 3 mL syringe with a 0.22 µL filter. 10-fold serial dilutions were made with the collected lysate for amplification. A spot titer was made with the undiluted lysate as well as 10−1 to 10−10 lysate dilutions. Based on counting the number of plaques formed by each lysate in the spot titer assay, the 10−8 dilution was selected as the best candidate to produce a countable plate. A full titer plate was prepared with the 10−7, 10-8,, and 10−9 dilutions. The titer calculated from the full titer assay was 2.65x1010 PFU/mL.
Phage characterization – DNA extraction
DNA extraction was performed on the phage lysate using the Wizard DNA Clean-Up kit (Promega) with minor modifications. 5 µL of nuclease mix (150 mM NaCl, ddH2O, 0.25 mg/mL DNase 1, 0.25 mg/mL RNase A, 50% glycerol) was added to 1 µL of lysate and mixed by inversion. The solution was incubated at 37° for ten minutes. 15 µL of 0.5 M EDTA and 1 µL of 20 mg/mL Proteinase K were added to the solution and incubated at 37° for 20 min. 2 mL of Wizard DNA Clean-Up resin (Promega) was added to the solution and mixed by inversion for two minutes. The solution was syringed-filtered through two Wizard Genomic DNA columns (Promega) and then washed three times with 80% isopropanol. The columns were twice spun in a centrifuge at top speed for two minutes and then placed in a 90° heat block for one minute. 50 µL of ddH2O was used for elution. Final elutes were combined for 100 µL of total DNA extract. A Nanodrop ND 1000 was used to determine a DNA concentration of 114.9 ng/µL.
Phage characterization – Transmission Electron Microscopy
A high-titer lysate was made up for Transmission Electron Microscopy (TEM) by spinning 100 µL of phage lysate in a 4° Centrifuge at top speed for 22 min. The supernatant was removed and the pellet was resuspended in 10 µL of phage buffer. The high-titer lysate then underwent TEM preparation by negatively staining the virus particles. Specifically, isolated particles were adhered to a 300-mesh carbon-formvar grid for one minute, followed by staining with 1% aqueous uranyl acetate for 30 sec. Images were acquired using a Philips CM12 TEM operated at 80kV and equipped with a Gatan model 791 CCD camera.
Library preparation, sequencing, and de novo assembly
A sequencing library was prepared from genomic DNA by using a NEB Ultra II FS kit with dual-indexed barcoding and sequenced on an Illumina MiSeq, yielding a total of 901,246 single-end 150bp reads (>895X coverage). Quality control checks using FastQC v.0.11.7 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc; last accessed 2020/04/30) indicated that the data were of high quality (i.e., no adapters were present and base quality scores were >30, equivalent to an error rate of <0.1%; Figure S1). Consequently, no adapter contaminations were trimmed off of the 3′-end of the reads by scythe v.0.991 (a Naive Bayesian approach to detect and remove contamination) and sickle v.1.33 (a tool for quality-based read trimming) flagged only ∼0.1% of the reads for containing base qualities <20. With no significant changes to our dataset, additional read processing prior to assembly was thus deemed unnecessary. Following Russell (2018), reads were de novo assembled using Newbler v2.9, resulting in a single linear contig of size 150,935bp, which was checked for completeness, accuracy, and phage genomic termini using Consed v.29 (Gordon et al. 1998). All software was executed using default settings.
Genome annotation
Annotation was performed using DNA Master v.5.23.3 (http://cobamide2.bio.pitt.edu; last accessed 2020/04/30). Putative protein-encoding open reading frames (genes) were identified using Glimmer v.3.0 (Delcher et al. 1999) and GeneMark v.2.5 (Lukashin and Borodovsky 1998) with AUG (methionine), UUG and CUG (leucine), GUG (valine), and AUA (isoleucine) as start codons. Using annotated bacteriophage sequences from public databases, functional assignments were made with Blastp v.2.9 (Altschul et al. 1990) – both within the DNA Master environment and within NCBI to take advantage of the Conserved Domain Database (Marchler-Bauer et al. 2015) – as well as with HHPred (Söding et al. 2005) which, in addition to sequence similarity, also compares putative three-dimensional protein structures. TMHMM2 (Krogh et al. 2001) and SOSUI (Hirokawa et al. 1998) were used to identify membrane proteins. tRNAs were annotated using Aragon v.1.1 (included in DNA Master) and v.1.2.38 (Laslett and Canback 2004) as well as tRNAscan-SE v.2.0 (Lowe and Eddy 1997). All software was executed using default settings.
Comparative genomics analysis
Due to their similar length, number of genes and tRNAs, as well as GC-content, the genomes of the phages IME-SA1, IME-SA2, ISP (Vandersteegen et al. 2011), JA1 (Ajuebor et al. 2018), K (Gill 2014), vB_SauM_0414_108 (Philipson et al. 2018), and vB_SauM-fRuSau02 (Leskinen et al. 2017) were downloaded from GenBank (Table 1) to create a database of Staphylococcus cluster C1 phages (Oliveira et al. 2019) using PhamDB (Lamine et al. 2016). This custom database was used for all subsequent comparative analyses. First, a multiple sequence alignment was performed utilizing Kalign v.1.04 (Lassmann and Sonnhammer 2005) to produce a neighbor-joining tree. Second, dotplots, comparing the relatedness of different nucleotide sequences, were generated in 10bp sliding windows using Gepard v.1.40 (Krumsiek et al. 2007). Lastly, the database was loaded into Phamerator (Cresawn et al. 2011) to visually compare phage genomes.
Data availability
Figure S1 depicts the quality control checks of the raw read data using FastQC. Whole genome sequencing data are available through NCBI’s Sequence Read Archive (BioProject accession number PRJNA640949) and the annotated genome assembly is available through GenBank (accession number MT411892.1). Supplemental material available at figshare: https://doi.org/10.25387/g3.12585056.
Results And Discussion
The complete genome sequence of the Staphylococcus bacteriophage Metroid was sequenced and annotated (see “Materials and Methods” for details). The Myoviridae morphology (i.e., an icosahedral capsid [diameter: 100nm] enclosing the double-stranded DNA attached to a long, contractile tail [length: 108nm]; Figure 1a) as well as the genome size of 151kb (including the ∼10kb terminal repeat) suggests that Metroid belongs to the Staphylococcus phage cluster C. Metroid’s genome has a GC-content of 30.40%, similar to those of previously published Staphylococcus phages (27.98–34.96%) (Kwan et al. 2005; Deghorain et al. 2012; Oliveira et al. 2019). The tightly-packed genome contains 254 predicted protein-coding genes as well as 4 tRNAs, most of which are transcribed on the forward strand (Figure 1b). This corresponds to a gene density of 1.68 genes/kb – on the upper end of the range previously reported for cluster C phages (164-249 genes; 0-5 tRNAs; 1.25-1.64 genes/kb) (Oliveira et al. 2019). Although the overall gene coding potential of Metroid is 89.42%, only 26 of the 254 predicted proteins could be assigned a putative function. The majority of predicted proteins are either conserved but of no known function (170 out of 254), membrane proteins (22), or unique (i.e., without a match to any of the queried databases; 1). As previously observed in other Staphylococcus phages (Kwan et al. 2005), functionally related genes are organized into distinct modules (e.g., distinct head and tail modules connected by a head-to-tail adapter; Figure 1b), the respective order of which is largely conserved across phages of the same category.
CHARACTERIZATION OF METROID AND ITS RELATEDNESS TO OTHER STAPHYLOCOCCUS CLUSTER C PHAGES. a) Transmission electron microscopy image showing Metroid’s morphology. b) Metroid’s genome contains 254 predicted protein-coding genes as well as 4 tRNAs; total genome size: 151kb including the ∼10kb terminal repeat. The majority of genes are transcribed on the forward strand as shown in pink; genes transcribed on the reverse strand are highlighted in orange; tRNAs in blue. Functionally related genes are organized into distinct modules (highlighted in gray). c) Neighbor-joining tree and d) dotplot of Metroid and seven previously described Staphylococcus bacteriophages (Table 1). e) Genes in the lysis cassettes as well as in the packaging module show a strong conservation between Metroid and two closely-related Staphylococcus phages, K (Gill 2014) and vB_SauM_0414_108 (Philipson et al. 2018). Genes are labeled with their putative function, with genes belonging to the same protein family (pham) depicted in the same color. Purple coloring between genomes highlights regions of high nucleotide similarity (i.e., a BLAST e-value of 0).
Complementing the classification by morphology and genome size, comparative genomic analysis with seven Staphylococcus subcluster C1 phages highlights a strong relatedness on the sequence level (Figures 1c,d) and thus, provides additional evidence for the assignment of Metroid to cluster C. Metroid is most closely related to vB_SauM_0414_108 (Figures 1c,d) – a phage discovered as part of a recent effort proposing a guideline and standardized workflow to submit phages to the Federal Drug Administration to be considered as potential future treatments of bacterial infections (Philipson et al. 2018). More generally, genes in the lysis cassettes show a strong conservation between Metroid and the closely-related Staphylococcus cluster C phages, including phages K (Gill 2014) and vB_SauM_0414_108 (Philipson et al. 2018), which both share 99% amino acid identity with Metroid for endolysin and >97% amino acid identity for holin (Figure 3e). Both phages were previously characterized for therapeutic research, suggesting that Metroid might be a suitable candidate for future phage therapies.
Acknowledgments
This study was supported by the Howard Hughes Medical Institute SEA-PHAGES program and Arizona State University’s School of Life Sciences. DNA concentration was determined in the Arizona State University DNA Shared Resource Facility. Library preparation and sequencing was performed at the University of Pittsburgh. Computations were partially performed at Arizona State University’s High Performance Computing facility. We are grateful to David Lowry for transmission electron microscopy imaging, Suhail Ghafoor for IT support, Billy Biederman, Graham Hatfull, Deborah Jacobs-Sera, Welkin Pope, Daniel Russell, and Vic Sivanathan for library preparation, sequencing, and assembly as well as providing faculty training for the SEA-PHAGES program and guidance with our genome annotation.
Footnotes
Supplemental material available at figshare: https://doi.org/10.25387/g3.12585056.
Communicating editor: A. Carleton Wong
- Received June 29, 2020.
- Accepted July 26, 2020.
- Copyright © 2020 Crane et al.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
