{"paper_id":"aca60d49-dfca-4c58-a531-37b90c5e1fb8","body_text":"Revisiting typing systems for group B Streptococcus (GBS)\nprophages: an application in prophage detection and\nclassification in GBS isolates from Argentina\nShort title: Revisiting typing systems for GBS prophages in GBS from Argentina\nVerónica Kovacec 1, Sabrina Di Gregorio 1,2, Mario Pajón 1, Chiara Crestani 3, Tomás\nPoklepovich4, Josefina Campos 4,5, Uzma Basit Khan 6, Stephen D. Bentley 6, Dorota\nJamrozy6, Marta Mollerach 1,2, Laura Bonofiglio 1,2,*\n1 Universidad de Buenos Aires. Facultad de Farmacia y Bioquímica. Instituto de\nBacteriología y Virología Molecular. Buenos Aires, Argentina.\n2 CONICET, Buenos Aires, Argentina.\n3 Global Health Department, Institut Pasteur. Paris, France.\n4 Unidad Operativa Centro Nacional de Genómica y Bioinformática, ANLIS Dr. Carlos G.\nMalbrán, Buenos Aires, Argentina\n5 Current Address: International Pathogen Genomic Surveillance Network (IPSN). WHO\nBerlin Hub.\n6 Parasites and Microbes Programme, Wellcome Sanger Institute. Cambridgeshire, United\nKingdom.\n* Corresponding author:\nE-mail: lbonofi@ffyb.uba.ar (LB)\n1\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nORCID: 0000-0003-3855-0553\nVK, SDG, MM and LB conceived this study and LB supervised it. TP , JC, UBK, DJ and SB\nconducted and coordinated the whole genome sequencing. VK and SDG designed the\nbioinformatic analyses. VK, MP and SDG performed the acquisition of the data. VK and\nSDG conducted the data analysis. SDG guided and supervised the bioinformatic analysis. CC\nparticipated in the discussion of the results and gave critical insights for their analysis. VK\nwrote the original draft of the manuscript. All authors reviewed, edited and approved the final\nversion of the manuscript.\nAbstract\nGroup B Streptococcus (GBS) causes severe infections in neonates and adults with\ncomorbidities. Prophages have been reported to contribute to GBS evolution and\npathogenicity. However, no studies are available to date on the presence and diversity of\nprophages in GBS isolates from humans in South America. This study provides insights into\nthe prophage content of 365 GBS isolates collected from clinical samples in the context of an\nArgentinean multicentric study. Using whole genome sequence data, we implemented two\npreviously proposed methods for prophage typing: a PCR approach (carried out in silico)\ncoupled with a blastx-based method to classify prophages based on their prophage group and\nintegrase type, respectively. We manually searched the genomes and identified 325\nprophages. However, only 80% of prophages could be accurately categorised with the\nprevious approaches. Integration of phylogenetic analysis, prophage group and integrase type\nallowed for all to be classified into 19 prophage types, which correlated with GBS clonal\n2\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\ncomplex grouping. The revised prophage typing approach was additionally improved by\nusing a blastn search after enriching the database with 10 new genes for prophage group\nclassification combined with the existing integrase typing method. This modified and\nintegrated typing system was applied to the analysis of 615 GBS genomes (365 GBS from\nArgentina and 250 from public databases), which revealed 29 prophage types, including 2\nnovel integrase subtypes. Their characterization and comparative analysis revealed major\ndifferences in the lysogeny and replication modules. Genes related to bacterial fitness,\nvirulence or adaptation to stressful environments were detected in all prophage types.\nConsidering prophage prevalence, distribution and their association with bacterial virulence,\nit is important to study their role in GBS epidemiology. In this context, we propose the use of\nan improved and integrated prophage typing system suitable for rapid phage detection and\nclassification with little computational processing.\nAuthor summary\nBacteriophages, which are viruses that infect bacteria, exert a profound influence on\nmicrobial evolution when integrated into the bacterial genome, a state in which they are\ncalled prophages. It has been proposed that prophage acquisition played a role in the\nemergence of Streptococcus agalactiae (GBS) as a human pathogen in European countries.\nFurther study and characterization of prophages of GBS from around the world would\nprovide valuable insights into the mechanisms underlying GBS adaptation, evolution and\nepidemiology.\nUnfortunately, existing tools for prophage screening exhibit limitations in the detection and\nclassification of all prophages present in GBS genomes. To address this issue, in this work we\npropose a new prophage typing system that allows the detection and classification of GBS\n3\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nprophages based on both their phylogenetic lineage and integration site within the bacterial\ngenome. Using this methodology we were able to identify 29 prophage types in 615 GBS\nisolated globally. We further characterised these prophages and found that they carried genes\nthat could give an evolutionary advantage to their host and that different lineages of GBS\ncarried different prophage types.\nComprehensive exploration and characterization of prophages represent an indispensable\nendeavour, providing critical insights into microbial evolution, epidemiology, and potential\ntherapeutic interventions.\nIntroduction\nStreptococcus agalactiae (group B Streptococcus; GBS) is a commensal bacterium that\ncolonises the human intestinal and genitourinary tracts. GBS is a major cause of neonatal\nsepsis and other perinatal infections, such as meningitis and pneumonia, globally [1]]. In\nrecent decades, invasive infections caused by GBS in non-pregnant adults have been on the\nrise, especially in the elderly people and in those suffering from underlying medical\nconditions [[2–4]].\nProphages are important vehicles for horizontal gene transfer (HGT) [[5] and can constitute\nup to 20% of a bacterial genome. Prophages play a significant role in bacterial evolution by\nintroducing genes that enhance bacterial fitness and virulence [6,7]. Furthermore, pathogenic\nstrains tend to carry more phage-related genes than non-pathogenic strains [8–10], which was\nalso observed for GBS [11].\nGBS temperate bacteriophages (lysogenic prophages) were first described in 1969 in strains\nfrom bovine origin [12]. Recent studies on human GBS isolates revealed an association\n4\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nbetween certain prophages (some of possible animal origin) and the emergence of specific\npathogenic GBS clones among isolates recovered from neonates and adults in Europe\n[11,13–15]. Little is known about the epidemiology of GBS prophages and their impact on\npathogenicity in other geographical areas. To date, there are no reports of prophages in GBS\nisolates from South America [16].\nTwo approaches have been previously developed for the screening and classification of\nprophages in GBS genomes, one based on full-prophage sequence diversity [11] and the other\nbased on integrase typing [16]. However, both have limitations and may under or\noverestimate prophage presence.\nThis study aims to analyse the prophage content in GBS genomes from Argentina by\nintegrating and improving the existing methods for the detection and typing of GBS\nprophages, providing a novel strategy for a global surveillance of GBS prophage\nepidemiology and diversity.\nMaterials and Methods\nM1. Isolate collection.\nWe collected 450 GBS isolates from maternal carriage as well as invasive and urinary tract\ninfections as part of a national multicentric study that involved 40 health centres in 12\nprovinces of Argentina between 2014-2015. All isolates had been characterised\nphenotypically (antibiotic susceptibility and serotyping) and invasive isolates had also been\ncharacterised genotypically (PFGE) [4,17,18].\n5\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nM2. Whole genome sequencing and data processing.\nGenomic DNA of 450 GBS isolates was extracted using a QIACube HT protocol and\nsequenced at the Wellcome Sanger Institute on Illumina NovaSeq 6000 platform (as part of\nour collaboration with the Juno consortium, https://www.gbsgen.net/). For 10 GBS isolates,\ngenomic DNA was extracted using Wizard® Genomic DNA Purification Kit (Promega) and\nsequenced at the Malbrán Institute on Illumina MiSeq. Quality of the reads was assessed with\nFastQC v0.11.7 [19] and Kraken v0.10.6 [20]. De novo assemblies were obtained with\nSPAdes v3.12.0 [21] and quality checked with Quast v5.0.0 [22]. The 365/450 assemblies\nthat passed the quality controls were annotated with Prokka v1.12 [23]. MLSTs were\ndetermined with the software mlst v2.22.1 [24] and assigned to clonal complexes [CC] using\nthe PubMLST website [25,26] (https://pubmlst.org/organisms/streptococcus-agalactiae ).\nM3. Prophage detection and typing.\nThe methodology followed for the detection and typing of the GBS prophages is summarised\nin this section and Fig 1. For detailed information see M3 in S1 File.\n6\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 1. Summarised methodology used for GBS-prophage detection, typing and improvement of\nthe prophage typing system.\n7\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nIn the first instance, prophage sequences were detected and classified using the previous\nscreening methods for prophage groups [11] (here perfomed in silico) and integrase type [16].\nResults of the two methods were combined and the putative prophages were preliminarily\nclassified into prophage types according to prophage group and integrase type (Fig 1).\nProphage sequences within the assembled genomes were manually searched and extracted.\nProphages fragmented across contigs were reconstructed by de novo assembly against\nreference prophages. All prophage sequences were annotated. (Fig 1).\nThe extracted prophage sequences were aligned and a phylogenetic tree was reconstructed.\nProphages that could not be assigned to a prophage group during the initial screening stage\nwere classified with the same group as the prophages in their phylogenetic cluster.\nClassification by prophage type was updated accordingly (Fig 1).\nM4. Improvement of the prophage typing system\nThe methods followed for the improvement of the prophage typing system are summarised in\nthis section and Fig 1. For detailed information see M4 in S1 File.\nIn order to avoid false positive and false negative results obtained in the initial screening, we\nimproved the detection of prophage groups by performing a blastn search against a curated\ndatabase of prophage group-specific genes (Table S1 in S2 File). The methodology was tested\non the 22 prophages detected by van der Mee-Marquet et al. (2018), all prophages from\nArgentinean GBS genomes and 615 GBS complete genomes from Argentina and public\ndatabases (Table S2 in S2 File). A result was considered positive when at least one of the\n8\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\ngenes for the prophage group was detected with a minimum of 75% of identity and coverage.\nProphage types were then defined combining these results with those of integrase types.\nM5. Prophage characterization.\nThe steps followed for prophage characterization are summarised in this section and Fig 2.\nFor detailed information see M5 in S1 File and results section R2.\nFig 2. Summarised methodology used for GBS-prophage characterization.\n9\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nProphage sequences were searched for genetic determinants of virulence and antimicrobial\nresistance, as well as any genes potentially beneficial for the host bacteria. Genes coding for\nintegrase, helicase, terminase large subunit, major capsid protein and lysin, were used for the\nphylogenetic analysis of each prophage module.\nOne phage of each prophage type (n=29, see results section) was selected for further\ncharacterization. The morphology of the prophages was predicted, as well as the function of\nthe genes annotated as encoding hypothetical proteins and the catalytic domain of the\nputative integrases and lysins present. A comparative sequence analysis was performed to\nstudy the genetic differences between prophages of the same prophage group but different\nintegrase type and vice versa.\nTo provide a broader context to the prophages from Argentinian GBS, a phylogenetic\nanalysis of 764 prophages from GBS (isolated in Argentina and worldwide) and other 43\nstreptococcal species was performed (Table S3 in S2 File).\nM6. Integration of information.\nThe Microreact application [27] was used for an integral visualisation of the collected\ninformation.\nM7. Statistical analysis\nFisher’s exact test (two-tailed) was used to evaluate the association between prophage\npresence and GBS clonal complex. A p-value of ≤0.05 was considered to be significant.\n10\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nM8. Ethics statement.\nEthical approval for the ‘‘Argentinian multicentric study on infections due to Streptococcus\nagalactiae’’ was provided by the Ethics Committee of the Faculty of Pharmacy and\nBiochemistry, University of Buenos Aires, Res [D] RESCD-2022-400-E-UBA-DCT.\nResults\nR1. Prophage detection and typing in GBS genomes from\nArgentina.\nProphage screening based on prophage phylogenetic group and integrase type, [11,16]\ndetected 383 putative prophages in the 365 GBS genomes. A total of 200/383 (52%)\nprophages were grouped into 10 prophage types (prophage group + integrase type). In 60/383\n(16%) prophages only the integrase type was determined (the prophage group was not\ndetermined [ND]). A total of 123/383 prophages (32%) belonged to prophage group A,\nknown to lack the integrase gene [16,28].\nIn contrast, a manual search revealed 325 prophages among the 365 GBS genomes.\nForty-two out of the 325 prophages (13%) were fragmented in two or more contigs but were\nsuccessfully assembled with reference-based read mapping. The additional prophage\nsequences detected based on the screening approach (n=58) represented mostly groups A\n(36/58 lacked the lysis module) and F (19/58 integrase gene only present). In 3/58 cases the\nscreened prophages were disregarded due to high level of fragmentation.\n11\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nOnly the 325 manually detected prophage sequences were analysed further. Phylogenetic\nanalysis revealed clustering by the prophage group (Fig 3, Fig S1A in S3 File). Prophages\nthat could not be typed by group with the initial screening (ND prophages) were assigned a\nprophage group corresponding to their phylogenetic cluster. Prophage groups E and F were\nfurther divided in two subclusters each (100% bootstrap support). In both cases, prophages of\none subcluster had not been detected by the in silico PCR. Those that were detected by the in\nsilico PCR were reclassified as subgroups E1 and F1, respectively, while those not detected\nwere reclassified as subgroups E2 and F2, respectively (Fig S1B in S3 File). The\nreclassification of group F prophages into the subgroups F1 and F2 coincides with van der\nMee-Marquet’s classification of prophages with insertion sites F1 and F2, respectively [11].\n12\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 3. Phylogeny of 325 prophages found in 365 Argentinian Group B Streptococcus (GBS)\ngenomes. Maximum-likelihood phylogenetic tree, midpoint rooted, with nodes coloured by prophage\ntype (determined based on the combination of prophage group and integrase type). Support values\n(SH-aLRT/ Ultrafast Bootstrap) are shown as labels for selected nodes. Host GBS clonal complex is\nshown as coloured blocks. The scale bar represents the number of SNPs per variable site.\nhttps://microreact.org/project/philogeny-argentinean-gbs-prophages\n13\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nAs a result of this integrated analysis all prophages (n=325) were reclassified into 19\nprophage types (Fig 3). The majority of GBS genomes (54%) carried a single prophage.\nMultiple integrase types/subtypes were found within prophage groups and vice versa (Fig 3),\nas previously described [16], indicating that screening of GBS prophages by previous\nmethods alone is insufficient to accurately classify the prophages.\nR2. Improvement of the screening and typing system.\nIn order to improve the screening and typing system for GBS prophages, we developed a\ndatabase of phylogroup-specific prophage genes (Table S1 in S2 File, Data S1 in S3 File).\nThe genes were selected based on 12 PCR-amplified fragments using previously described\nprimers [11], and refined to avoid false positive results. Detection of group A prophages is\nbased on genes hhaI or clpP combined with a presence of either a holin or a lysin gene (Table\nS1 in S2 File). F1 prophage integrase gene (hin) was replaced by a gene coding for a\nterminase large subunit (Table S1 in S2 File). The hypothetical prophage gene representative\nof group D prophages was replaced by two new genes (Table S1 in S2 File). Finally, gene\nsequences for the detection of group E2 and F2 prophages were also added to the database\n(Table S1 in S2 File). These changes markedly improved the specificity and accuracy of\nprophage detection and typing in the 365 GBS genomes from Argentina (Fig 4).\n14\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 4. Prophage detection and typing in 365 Argentinian Group B Streptococcus (GBS)\ngenomes. Three methods are compared: the initial integrated screening of prophage group by in silico\nPCR (with primers designed by van der Mee-Marquet et al.) and integrase type by blastx search\nagainst a GBS-prophage integrase database (built by Crestani et al.); the manual search in the GBS\ngenomes of the screened prophages and their classification according to their phylogeny; the\nimproved screening and typing method by prophage group: blastn search against a database of the\nrepresentative genes for each group proposed by van der Mee-Marquet et al., and new genes proposed\nin this work, integrated with the integrase typing designed by Crestani et al.\nMore importantly, our improved method detected 10 additional prophage types when tested\non a collection of 615 globally-derived GBS genomes (including the 365 from Argentina),\ngiving a total of 29 distinct prophage types. The 10 prophage types not found among the GBS\ngenomes from Argentina included 2 new integrase subtypes: GBSInt6.3 and GBSInt8.2 and\ntheir sequences were added to the integrase database (Data S2 in S3 File).\nA graphical summary of the proposed improved screening and typing method for GBS\nprophages is shown in Fig 5.\n15\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 5. Methodology summary for the improved screening and typing system for GBS prophages.\nR3. Characterization of GBS prophages from Argentina.\nThe most prevalent prophage types in GBS isolates from Argentina were: A (87/325, 27%),\nE1/GBSInt3 (85/325, 26%), E2/GBSInt3 (42/325, 13%) and F1/GBSint11.2 (28/325, 9%),\n(Fig 4). Significant associations were found between most prophage types and CC\nassignment (p<0.05, Fig S2 in S3 File). Prophages of type A were associated with CC23 and\nCC1; B/GBSInt9.2 with CC17; C/GBSInt4 with CC17 and CC26; D/GBSInt1, D/GBSInt2.2\nand E2/GBSInt3 with CC19; D/GBSInt8.1 with CC103; E1/GBSInt3 with CC23 and CC452;\nE1/GBSInt8.1 with CC23; F1/GBSInt11.2 with CC12.\nThe phylogenetic analysis (Fig 3) revealed monophyletic clusters for most prophages within a\ngroup or subgroup with the exception of group D prophages, which were more diverse.\n16\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nThe insertion site and att sequences matched those described by Crestani et al. [16] for each\nintegrase type, with the exception of some prophages with minor changes in their att\nsequences (Table S4 in S2 File). One gene of each prophage module (integrase, helicase,\nterminase large subunit, major capsid protein, lysin) was selected for phylogenetic analysis of\ntheir nucleotide sequences, to determine whether similar modules could be found in different\nprophage types. These analyses showed, in general, similar clustering between phylogenies\nbased on individual prophage module genes and that observed based on alignment of full\nprophage sequences (Fig S3 in S3 File).\nSeveral genes potentially beneficial for the host bacteria were found exclusively in\nB/GBSInt9.2 prophages, including genes involved in carbohydrate or RNA metabolism,\nwhich were present in all prophages of this type. A single B/GBSInt9.2 prophage additionally\ncarried genes coding for phosphoenolpyruvate synthase, a multidrug resistance efflux pump,\ngenes involved in DNA metabolism and several genes encoding different types of permeases\n(Table S5 in S2 File).\nPhylogenetic analysis of 764 Streptococcus spp. prophages revealed that GBS phages from\nArgentina were related to other GBS prophages and with bacteriophages from other\nstreptococcal species (Fig 6). In particular, group B prophages were closely related to S.\npyogenes, S. iniae, S. oralis and S. pneumoniae phages, while group A and F prophages were\nrelated to phages from more than ten different streptococcal species. However, prophages\nfrom groups C, D, and E were more closely related to each other in this phylogeny (100%\nbootstrap support) and seem to be mainly restricted to GBS, with an occasional spread to S.\ndysgalactiae, S. equi, S. iniae and S. pyogenes (clustering with group C) and one single S.\nuberis prophage (clustering with group E1).\n17\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 6. Phylogenetic tree of prophages from streptococcal species. Maximum-likelihood\nphylogenetic tree of 764 prophages found in 580 genomes from 44 streptococcal species, including\nGBS. Nodes are coloured by GBS prophage type, where applicable. Host species are shown in the\ncoloured ring. Tree was rooted at midpoint. Support values (SH-aLRT/ Ultrafast Bootstrap) are shown\n18\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nas labels for selected nodes. The scale bar represents the number of SNPs per variable site.\nhttps://microreact.org/project/gbs-prophages-in-a-global-context\nR4. Comparative analysis of the 29 distinct prophage types\nMore than one integrase type or subtype was found in all prophage groups, with the exception\nof group A (prophages without integrase). Integrase types GBSInt1, GBSInt3, GBSInt4,\nGBSInt8.1 and GBSInt11.1 were found in more than one prophage group, mostly D and E.\nAlso, D and E2 prophages had the most diversity in integrase types (7 and 6 types,\nrespectively). Group B and F prophages had two subtypes of the same integrase type\n(GBSInt9 and GBSInt11, respectively).\nOne prophage sequence for each of the 29 types was selected for further characterization.\nBased on their predicted morphology, all phages were classified as siphovirus (former\nSiphoviridae family). Visual inspection of the annotated prophage sequences confirmed the\nmodular organisation following a specific order based on their function in the phage life\ncycle: lysogeny, replication, packaging, morphogenesis and host lysis (Fig 7). Genome sizes\nranged from 32.6 to 47.9 Kb, with group B prophages the smallest (likely due to shorter\ngenes and a shorter morphogenesis module) and the prophages of groups D and E1 the largest\n(likely due to higher gene content in the lysogeny and replication modules). Manual\ninspection of the sequences revealed that all prophage types contained open reading frames\n(ORFs) encoding putative proteins related to: bacterial fitness, defence mechanisms and/or\nvirulence (Table S4 in S2 File). ORFs coding for hypothetical proteins where no known\nconserved domain was found constituted from 29 to 59% of the coding sequences of each\nprophage. These genes were found throughout the entire prophage genome and not organised\n19\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nin a single module, although they were more frequent in the lysogeny and replication\nmodules.\n20\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nFig 7. Comparative analysis of representative sequences of 29 distinct prophage types found in\nthe analysed GBS. Colours represent groups of homologous genes. Genes with more than 40% of\nidentity are linked with grey-black strokes, as shown in the scale.\nThe amino acid sequences of each integrase type and lysins found in each of the 29 reference\nprophages, and lysins coded by genes from each phylogenetic cluster (Fig S3E in S3 File)\nwere analysed to determine their catalytic domains (Table S4 in S2 File). All integrase types\nhad a tyrosine recombinase domain, except for GBSInt11.1 and GBSInt11.2, which had a\nserine recombinase domain. The putative lysins were classified based on their cleavage site,\ninto three of the five major endolysin classes [29]: N-acetyl-β-D-glucosaminidase,\nN-acetyl-β-D-muramidase and N-acetylmuramoyl-L-alanine-amidase. Interestingly, lysins\nfrom all prophages contained the domain N-acetylmuramoyl-L-alanine-amidase and, in most\ncases, the lysins were bifunctional, as they also carried a second catalytic domain with a\ndifferent cleavage site, either N-acetyl-β-D-glucosaminidase or N-acetyl-β-D-muramidase.\nAll prophages from the same type had lysins of the same class, except for types A,\nC/GBSInt4, D/GBSInt3, E1/GBSInt3, E2/GBSInt4, E2/GBSInt11.1 and F1/GBSInt11.2, in\nwhich some lysins had only the N-acetylmuramoyl-L-alanine-amidase domain, while others\nwere bifunctional.\nComparative sequence analysis of the 29 prophage types (Fig 7) showed more than 60%\nidentity between the morphogenesis modules encoding most tail proteins of the groups D, E\nand F and around 45% identity between most genes in the replication modules of the type A,\nD/GBSInt4 and D/GBSInt2.2. Group F prophages shared more similarities in a modular level\nwith groups D and E than with phylogenetically closer groups B and C, which did not show\nidentity on a modular level with other prophage types.\n21\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nComparative analysis of prophages belonging to the same prophage group but carrying\ndifferent integrase types or subtypes (Fig S4 in S3 File) revealed that the major differences\nbetween the prophages within groups B, C, D and E and subgroup F2 resided in the lysogeny\nand replication modules. The F1prophage showed less than 60% identity with subgroup F2\nbetween all their modules except for the morphogenesis. In addition, prophages with the\nsame integrase type but belonging to different prophage groups (Fig S5 in S3 File) did not\nreveal any similarity, other than the integrase gene itself. The exceptions were prophages with\nintegrase type GBSInt4, where more than 90% of identity was found between the lysogeny\nmodule and the genes following the host lysis module (Fig S5C in S3 File).\nDiscussion\nProphage presence has been reported to impact GBS epidemiology in collections from\nEurope [11,13–15]. In particular, the acquisition of certain prophages has been linked with the\nemergence of GBS infections in neonates and adults in France [11] and it has been postulated\nthat the presence of prophages encoding virulence factors was responsible for the increased\nincidence of severe neonatal infections both in France [13,28] and the Netherlands [14].\nThis is the first report on the diversity of prophages in human GBS isolates from South\nAmerica. Employing two methods for GBS prophage typing [11,16] and subsequent manual\ninspection, we were able to detect 325 prophages among 365 GBS isolates collected in an\nArgentinian multicentric study.\nNew prophage typing system\n22\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nWe propose an enhanced method that combines and simplifies some screening steps of\npreviously developed GBS-prophage typing methods that allowed the screening and\nclassification of prophages by either phylogenetic group [11] or integrase type [16], which do\nnot often correlate [16, this work]. Our results show that using each method individually can\nlead to false negative or false positive results. Furthermore, it can be insufficient to classify\nGBS prophages accurately, as determining the prophage group provides information about\nthe genomic characteristics of the phage but not its insertion site. The latter can be\ndetermined based on integrase type, which in contrast does not offer insights into\ncomposition of full prophage sequence.\nThe new method increases detection of full prophage sequences, as well as prophages that are\nfragmented into multiple contigs. Additionally, it allows the classification of prophages based\non both their phylogenetic lineage and integration site within the bacterial genome. Using this\napproach, we were able to identify a total of 29 distinct prophage types, including 19\nprophage types in GBS from Argentina and 10 additional types in prophages from GBS\ncollected in other countries.\nOur results demonstrate that this improved integrated method is less likely to detect prophage\nremnants, allows identification of novel prophage integrases and offers fast detection of GBS\nprophages in a large genomic dataset, with little computational processing.\nEvolution of GBS prophages\nGenome mosaicism was observed in all prophage types, in accordance with the proposed\nmodular evolution of prophages [30–32]. Genes belonging to the packaging module, those\nencoding capsid proteins and a few coding for tail proteins appear to have been acquired as a\nblock, independently of the rest of the prophage genes. This is especially evident in groups\nwith a greater number of prophage types, D and E, since the clustering of phages into\n23\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nphylogenetic subclades (Fig 3) correlates with their grouping according to the homology in\nthe aforementioned gene region (Fig 7). The presence of homology between several genes\nencoding tail proteins in prophages D, E and F suggests that this group of genes share a\ncommon ancestor and that they might have been acquired in an independent recombination\nevent from the rest of the structural genes, which did not present homology between the\ndifferent phage groups (Fig 7). In general, prophages within the same group presented the\nsame classes of lysins (same catalytic domains, Table S4 in S2 File), even if they did not\nshare high homology in their lysis modules. However, there was no lysin class exclusive to\none prophage group, which could indicate an independent recombination of the lysin-coding\ngene from the rest of the lysis module.\nDivergence between prophages, even belonging to the same phylogenetic group, was\ntypically observed in the lysogeny and replication modules, where the majority of the genes\nencoding hypothetical proteins and genes potentially beneficial for GBS were located. This\nsuggests that these regions would be the most prone to suffer recombination events. This is\nalso supported by the lack of similarities in the lysogeny module of prophages with the same\nintegrase type but different prophage group, which implies that the prophage integration site\nin the host genome is dependent only on the integrase gene and att sequence. The general\nabsence of homology between prophages of groups A, B and C with other prophage types\nwould suggest that these prophage groups have little propensity for horizontal gene exchange\nwith GBS prophages from other groups.\nMost prophage groups and subgroups seem to be highly clonal, whereas group D showed\nmore divergence, demonstrating evidence of microevolution (Fig 3, Fig 6 and Fig S1A in S3\nFile). The reason for such variability is not clear yet, but future studies are needed to\nunderstand if it might be advantageous to the host.\n24\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nThe level of conservation of genes encoding tail proteins among various prophage groups\n(Fig 7) could mean that they are involved in the specific recognition between the phage and\nthe bacterial receptor, a process that has not yet been studied in GBS. If so, these phages\ncould have a similar host range. Interestingly, tail protein genes are also conserved in\nprophage types from other streptococcal species (S. pyogenes, S. pneumoniae) [33,34].\nFurther analysis of the tail protein sequence conservation can reveal new insights into the\nmechanisms of prophage sharing between Streptococcus spp.\nProphage presence in the context of GBS epidemiology\nThe prevalence of GBS isolates carrying at least one prophage and the average number of\nprophages per isolate found in our dataset correlates with previous reports on prophage\ncontent in GBS [11,13,15,35,36]. Our results also aligned with the integrase type and CC\nassociations reported previously by Crestani [16].\nThe reason for such association should be explored, but could be related to the presence of\nspecific restriction-modification systems, which may limit the recombination and horizontal\ntransfer of mobile genetic elements between GBS lineages, as seen in other bacterial species\n[37–39].\nIn line with previous reports, GBS prophages identified in our dataset carried genes\npotentially involved in bacterial fitness, host adaptation to stressful environments and\nvirulence [11,13,15,28,40]. However, most of these genes were not detected by the online\nprophage genome screening tools, which stresses how manual curation remains an important\npart of the prophage genome investigations. A single B/GBSInt9.2 prophage was found to\ncarry a gene encoding a multidrug and toxic compound extrusion (MA TE)-like protein,\nsimilar to MepA from Staphylococcus aureus [41]. It has been reported that expression of\nMA TE proteins in bacteria confers resistance to toxic dyes and multiple antibiotics [42,43].\n25\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nWhile carriage of antibiotic resistance genes (ARGs) is not common in prophages [44], there\nhave been reports of ARGs carriage in conjugative elements within prophages of\nstreptococcal species [45,46]. The fact that these elements were found in only a single GBS\nprophage analysed in this study, highlights the overall low prevalence of ARGs in GBS\nprophages. However, it is important to note that, on average, 50% of GBS prophage genes\ncode for hypothetical proteins lacking recognizable conserved domains. Hence, it is\nimportant to continue the study of these prophages to better understand their biology and\nimpact on the host cells, particularly in a context of bacteriophage therapy development [47].\nGBS carriage of prophages associated to those from other streptococcal species causing\ninfections in humans and animals has been extensively documented in the literature\n[11,15,35,48,49]. Our results further show that prophages detected in GBS isolates from\nArgentina are globally distributed and suggest that prophages belonging to groups A, B, and\nF might have evolved from prophages horizontally transferred between different species of\nstreptococci (Fig 6). In contrast, prophage groups C, D and E, that were found to share a\ncommon ancestor in this global phylogeny, seem to be mainly restricted to GBS (Fig 6).\nFurther experimental work is needed in order to confirm the prophage transfer between\nstreptococcal species, receptor specificity (tail proteins) and the restriction of C, D, and E\nprophages to GBS only. It is also crucial to investigate if the insertion sites of these\nprophages favour the mobilisation of genetic material and the potential for horizontal transfer\nof genes present in defective prophages or phage remnants.\nStudy limitations\nLimitations of this study include using GBS genome assemblies from short read data to detect\nand classify prophages. All prophage sequences identified after manual curation remain\ntheoretical, even when found in the same contig, due to possibility of assembly errors,\n26\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nparticularly in prophage modules with high sequence similarity between different phage types\n(e.g., structural modules). This could be addressed by performing long-read sequencing\nfollowed by a hybrid assembly.\nConclusions\nThis study performed a comprehensive comparative analysis of prophage genomes in GBS\nisolates from Argentina, and is a first report on GBS prophage diversity in Latin America. We\npropose the use of an improved and integrated prophage typing system suitable for rapid\nphage detection in GBS genomes and their classification with little computational processing.\nThe presence of prophages in most GBS isolates analysed here, association between\nprophage groups and GBS lineages as well as carriage of genes beneficial for the host\nbacteria reinforce the hypothesis that acquisition of prophages confers an evolutionary\nadvantage to GBS and may play an important role in its epidemiology. The diversity of\nprophage types found in GBS isolates from Argentina along with the observed lysin diversity\nis a promising finding, which can be explored further to identify novel lysins with activity\nagainst GBS as an alternative therapy for GBS infections. In the light of the global challenge\nposed by antimicrobial resistant bacteria it is imperative to advance our current knowledge of\nbacteriophage biology and applicability of lysins as an alternative treatment against bacterial\ninfections, to promote and expedite the approval and regulation of such therapies.\nAcknowledgments\nWe would like to thank everyone involved in the genome sequencing of our GBS collection\nat the sequence facilities of the Wellcome Sanger Institute and Instituto Malbrán.\n27\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nReferences\n1. Alotaibi NM, Alroqi S, Alharbi A, Almutiri B, Alshehry M, Almutairi R, et al. Clinical\nCharacteristics and Treatment Strategies for Group B Streptococcus (GBS) Infection in\nPediatrics: A Systematic Review. Medicina (Mex). 2023;59: 1279.\ndoi:10.3390/medicina59071279\n2. Björnsdóttir ES, Martins ER, Erlendsdóttir H, Haraldsson G, Melo-Cristino J, Kristinsson\nKG, et al. Changing epidemiology of group B streptococcal infections among adults in\nIceland: 1975–2014. Clin Microbiol Infect. 2016;22: 379.e9-379.e16.\ndoi:10.1016/j.cmi.2015.11.020\n3. Navarro-T orné A, Curcio D, Moïsi JC, Jodar L. Burden of invasive group B\nStreptococcus disease in non-pregnant adults: A systematic review and meta-analysis.\nMelo-Cristino J, editor. PLOS ONE. 2021;16: e0258030.\ndoi:10.1371/journal.pone.0258030\n4. Arias B, Kovacec V, Vigliarolo L, Suárez M, T ersigni C, Lopardo H, et al. Epidemiology\nof Invasive Infections Caused by Streptococcus agalactiae in Argentina. Microb Drug\nResist. 2022; mdr.2021.0071. doi:10.1089/mdr.2021.0071\n5. Juhas M. Horizontal gene transfer in human pathogens. Crit Rev Microbiol. 2015;41:\n101–108. doi:10.3109/1040841X.2013.804031\n6. Davies EV, Winstanley C, Fothergill JL, James CE. The role of temperate\nbacteriophages in bacterial infection. Millard A, editor. FEMS Microbiol Lett. 2016;363:\nfnw015. doi:10.1093/femsle/fnw015\n7. Khan A, Wahl LM. Quantifying the forces that maintain prophages in bacterial\ngenomes. Theor Popul Biol. 2020;133: 168–179. doi:10.1016/j.tpb.2019.11.003\n8. Hayashi T . Complete Genome Sequence of Enterohemorrhagic Eschelichia coli\nO157:H7 and Genomic Comparison with a Laboratory Strain K-12. DNA Res. 2001;8:\n11–22. doi:10.1093/dnares/8.1.11\n28\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\n9. Winstanley C, Langille MGI, Fothergill JL, Kukavica-Ibrulj I, Paradis-Bleau C,\nSanschagrin F , et al. Newly introduced genomic prophage islands are critical\ndeterminants of in vivo competitiveness in the Liverpool Epidemic Strain of\nPseudomonas aeruginosa. Genome Res. 2008;19: 12–23. doi:10.1101/gr.086082.108\n10. Busby B, Kristensen DM, Koonin EV. Contribution of phage-derived genomic islands to\nthe virulence of facultative bacterial pathogens: Genomics update. Environ Microbiol.\n2013;15: 307–312. doi:10.1111/j.1462-2920.2012.02886.x\n11. van der Mee-Marquet N, Diene SM, Barbera L, Courtier-Martinez L, Lafont L, Ouachée\nA, et al. Analysis of the prophages carried by human infecting isolates provides new\ninsight into the evolution of Group B Streptococcus species. Clin Microbiol Infect.\n2018;24: 514–521. doi:10.1016/j.cmi.2017.08.024\n12. Russell H, Norcross NL, Kahn DE. Isolation and Characterization of Streptococcus\nagalactiae Bacteriophage. J Gen Virol. 1969;5: 315–317.\ndoi:10.1099/0022-1317-5-2-315\n13. Renard A, Barbera L, Courtier-Martinez L, Dos Santos S, Valentin A-S, Mereghetti L, et\nal. phiD12-Like Livestock-Associated Prophages Are Associated With Novel\nSubpopulations of Streptococcus agalactiae Infecting Neonates. Front Cell Infect\nMicrobiol. 2019;9: 166. doi:10.3389/fcimb.2019.00166data\n14. Jamrozy D, Bijlsma MW, de Goffau MC, van de Beek D, Kuijpers TW, Parkhill J, et al.\nIncreasing incidence of group B Streptococcus neonatal infections in the Netherlands is\nassociated with clonal expansion of CC17 and CC23. Sci Rep. 2020;10: 9539.\ndoi:10.1038/s41598-020-66214-3\n15. Lichvariková A, Soltys K, Szemes T , Slobodnikova L, Bukovska G, Turna J, et al.\nCharacterization of Clinical and Carrier Streptococcus agalactiae and Prophage\nContribution to the Strain Variability. Viruses. 2020;12: 1323. doi:10.3390/v12111323\n16. Crestani C, Forde TL, Zadoks RN. Development and Application of a Prophage\nIntegrase Typing Scheme for Group B Streptococcus. Front Microbiol. 2020;11: 1993.\n29\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\ndoi:10.3389/fmicb.2020.01993\n17. Arias B, Kovacec V, Vigliarolo L, Suárez M, T ersigni C, Müller L, et al.\nFluoroquinolone-Resistant Streptococcus agalactiae Invasive Isolates Recovered in\nArgentina. Microb Drug Resist. 2019;25: 739–743. doi:10.1089/mdr.2018.0246\n18. Vigliarolo L, Arias B, Suárez M, Van Haute E, Kovacec V, Lopardo H, et al. Argentinian\nmulticenter study on urinary tract infections due to Streptococcus agalactiae in adult\npatients. J Infect Dev Ctries. 2019;13: 77–82. doi:10.3855/jidc.10503\n19. Andrews S. FastQC: A Quality Control T ool for High Throughput Sequence Data\n[Online]. Available online at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.\n2010.\n20. Davis MPA, van Dongen S, Abreu-Goodger C, Bartonicek N, Enright AJ. Kraken: A set\nof tools for quality control and analysis of high-throughput sequence data. Methods.\n2013;63: 41–49. doi:10.1016/j.ymeth.2013.06.027\n21. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. SPAdes: A\nNew Genome Assembly Algorithm and Its Applications to Single-Cell Sequencing. J\nComput Biol. 2012;19: 455–477. doi:10.1089/cmb.2012.0021\n22. Gurevich A, Saveliev V, Vyahhi N, T esler G. QUAST : quality assessment tool for\ngenome assemblies. Bioinformatics. 2013;29: 1072–1075.\ndoi:10.1093/bioinformatics/btt086\n23. Seemann T . Prokka: rapid prokaryotic genome annotation. Bioinforma Oxf Engl.\n2014;30: 2068–2069. doi:10.1093/bioinformatics/btu153\n24. Seemann T . mlst Github https://github.com/tseemann/mlst. 2016. Available:\nhttps://github.com/tseemann/mlst\n25. Jones N, Bohnsack JF , T akahashi S, Oliver KA, Chan M-S, Kunst F , et al. Multilocus\nSequence Typing System for Group B Streptococcus. J Clin Microbiol. 2003;41:\n2530–2536. doi:10.1128/JCM.41.6.2530-2536.2003\n26. Jolley KA, Bray JE, Maiden MCJ. Open-access bacterial population genomics: BIGSdb\nsoftware, the PubMLST .org website and their applications. Wellcome Open Res.\n30\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\n2018;3: 124. doi:10.12688/wellcomeopenres.14826.1\n27. Argimón S, Abudahab K, Goater RJE, Fedosejev A, Bhai J, Glasner C, et al.\nMicroreact: visualizing and sharing data for genomic epidemiology and\nphylogeography. Microb Genomics. 2016;2. doi:10.1099/mgen.0.000093\n28. Renard A, Diene SM, Courtier-Martinez L, Gaillard JB, Gbaguidi-Haore H, Mereghetti\nL, et al. 12/111phiA Prophage Domestication Is Associated with Autoaggregation and\nIncreased Ability to Produce Biofilm in Streptococcus agalactiae. Microorganisms.\n2021;9: 1112. doi:10.3390/microorganisms9061112\n29. Love MJ, Abeysekera GS, Muscroft-T aylor AC, Billington C, Dobson RCJ. On the\ncatalytic mechanism of bacteriophage endolysins: Opportunities for engineering.\nBiochim Biophys Acta BBA - Proteins Proteomics. 2020;1868: 140302.\ndoi:10.1016/j.bbapap.2019.140302\n30. Botstein D. A THEORY OF MODULAR EVOLUTION FOR BACTERIOPHAGES*. Ann\nN Y Acad Sci. 1980;354: 484–491. doi:10.1111/j.1749-6632.1980.tb27987.x\n31. Lima-Mendez G, T oussaint A, Leplae R. A modular view of the bacteriophage genomic\nspace: identification of host and lifestyle marker modules. Res Microbiol. 2011;162:\n737–746. doi:10.1016/j.resmic.2011.06.006\n32. Dion MB, Oechslin F , Moineau S. Phage diversity, genomics and phylogeny. Nat Rev\nMicrobiol. 2020;18: 125–138. doi:10.1038/s41579-019-0311-5\n33. Canchaya C, Proux C, Fournous G, Bruttin A, Brüssow H. Prophage Genomics.\nMicrobiol Mol Biol Rev. 2003;67: 238–276. doi:10.1128/MMBR.67.2.238-276.2003\n34. Garriss G, Henriques-Normark B. Lysogeny in Streptococcus pneumoniae.\nMicroorganisms. 2020;8: 1546. doi:10.3390/microorganisms8101546\n35. Crestani C, Seligsohn D, Forde TL, Zadoks RN. How GBS Got Its Hump: Genomic\nAnalysis of Group B Streptococcus from Camels Identifies Host Restriction as well as\nMobile Genetic Elements Shared across Hosts and Pathogens. Pathogens. 2022;11:\n1025. doi:10.3390/pathogens11091025\n36. Sirimanapong W, Phước NN, Crestani C, Chen S, Zadoks RN. Geographical, T emporal\n31\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nand Host-Species Distribution of Potentially Human-Pathogenic Group B Streptococcus\nin Aquaculture Species in Southeast Asia. Pathogens. 2023;12: 525.\ndoi:10.3390/pathogens12040525\n37. McCarthy AJ, Witney AA, Lindsay JodiA. Staphylococcus aureus T emperate\nBacteriophage: Carriage and Horizontal Gene Transfer is Lineage Associated. Front\nCell Infect Microbiol. 2012;2. doi:10.3389/fcimb.2012.00006\n38. Oliveira PH, T ouchon M, Rocha EPC. Regulation of genetic flux between bacteria by\nrestriction–modification systems. Proc Natl Acad Sci. 2016;113: 5658–5663.\ndoi:10.1073/pnas.1603257113\n39. DebRoy S, Shropshire WC, Tran CN, Hao H, Gohel M, Galloway-Peña J, et al.\nCharacterization of the Type I Restriction Modification System Broadly Conserved\namong Group A Streptococci. Fey PD, editor. mSphere. 2021;6: e00799-21.\ndoi:10.1128/mSphere.00799-21\n40. Furfaro LL, Payne MS, Chang BJ. Host range, morphological and genomic\ncharacterisation of bacteriophages with activity against clinical Streptococcus\nagalactiae isolates. Melo-Cristino J, editor. PLOS ONE. 2020;15: e0235002.\ndoi:10.1371/journal.pone.0235002\n41. McAleese F , Petersen P , Ruzin A, Dunman PM, Murphy E, Projan SJ, et al. A Novel\nMATE Family Efflux Pump Contributes to the Reduced Susceptibility of\nLaboratory-Derived Staphylococcus aureus Mutants to Tigecycline. Antimicrob Agents\nChemother. 2005;49: 1865–1871. doi:10.1128/AAC.49.5.1865-1871.2005\n42. Claxton DP , Jagessar KL, Mchaourab HS. Principles of Alternating Access in Multidrug\nand T oxin Extrusion (MATE) Transporters. J Mol Biol. 2021;433: 166959.\ndoi:10.1016/j.jmb.2021.166959\n43. Huang H, Wan P , Luo X, Lu Y , Li X, Xiong W, et al. Tigecycline Resistance-Associated\nMutations in the MepA Efflux Pump in Staphylococcus aureus. Wang H, editor.\nMicrobiol Spectr. 2023;11: e00634-23. doi:10.1128/spectrum.00634-23\n44. Enault F , Briet A, Bouteille L, Roux S, Sullivan MB, Petit M-A. Phages rarely encode\n32\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nantibiotic resistance genes: a cautionary tale for virome analyses. ISME J. 2017;11:\n237–247. doi:10.1038/ismej.2016.90\n45. Dai X, Sun J, Zhu B, Lv M, Chen L, Chen L, et al. Various Mobile Genetic Elements\nInvolved in the Dissemination of the Phenicol-Oxazolidinone Resistance Gene optrA in\nthe Zoonotic Pathogen Streptococcus suis: a Nonignorable Risk to Public Health.\nSchaufler K, editor. Microbiol Spectr. 2023;11: e04875-22.\ndoi:10.1128/spectrum.04875-22\n46. Santoro F , Pastore G, Fox V, Petit M-A, Iannelli F , Pozzi G. Streptococcus pyogenes\nΦ1207.3 Is a T emperate Bacteriophage Carrying the Macrolide Resistance Gene Pair\nmef(A)- msr(D) and Capable of Lysogenizing Different Streptococci. Visca P , editor.\nMicrobiol Spectr. 2023;11: e04211-22. doi:10.1128/spectrum.04211-22\n47. Monteiro R, Pires DP , Costa AR, Azeredo J. Phage Therapy: Going T emperate? Trends\nMicrobiol. 2019;27: 368–378. doi:10.1016/j.tim.2018.10.008\n48. Bai Q, Zhang W, Yang Y , T ang F , Nguyen X, Liu G, et al. Characterization and genome\nsequencing of a novel bacteriophage infecting Streptococcus agalactiae with high\nsimilarity to a phage from Streptococcus pyogenes. Arch Virol. 2013;158: 1733–1741.\ndoi:10.1007/s00705-013-1667-x\n49. Rezaei Javan R, Ramos-Sevillano E, Akter A, Brown J, Brueggemann AB. Prophages\nand satellite prophages are widespread in Streptococcus and may play a role in\npneumococcal pathogenesis. Nat Commun. 2019;10: 4852.\ndoi:10.1038/s41467-019-12825-y\nSupporting information\nS1 File. Detailed methodology\nS2 File. Supplementary tables\n33\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint \n\nS3 File. Supplementary figures and data\n34\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 10, 2024. ; https://doi.org/10.1101/2024.05.08.593127doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}