RESEARCH ARTICLE Bleriot et al., Microbial Genomics 2020;6 DOI 10.1099/mgen.0.000369 Genomic analysis of 40 prophages located in the genomes of 16 carbapenemase-­producing clinical strains of Klebsiella pneumoniae Ines Bleriot1,2, Rocío Trastoy1,2, Lucia Blasco1,2, Felipe Fernández-­Cuenca2,3,4, Antón Ambroa1,2, Laura Fernández-­ García1,2, Olga Pacios1,2, Elena Perez-­Nadales4,5, Julian Torre-­Cisneros4,5, Jesús Oteo-­Iglesias2,4,6, Ferran Navarro2,7, Elisenda Miró2,7, Alvaro Pascual2,3,4, German Bou1,2,4, Luis Martínez-­Martínez2,4,5† and Maria Tomas1,2,4,*,† Abstract Klebsiella pneumoniae is the clinically most important species within the genus Klebsiella and, as a result of the continuous emer- gence of multi-­drug resistant (MDR) strains, the cause of severe nosocomial infections. The decline in the effectiveness of antibiotic treatments for infections caused by MDR bacteria has generated particular interest in the study of bacteriophages. In this study, we characterized a total of 40 temperate bacteriophages (prophages) with a genome range of 11.454–84.199 kb, predicted from 16 carbapenemase-­producing clinical strains of K. pneumoniae belonging to different sequence types, previously identified by mul- tilocus sequence typing. These prophages were grouped into the three families in the order Caudovirales (27 prophages belonging to the family Myoviridae, 10 prophages belonging to the family Siphoviridae and 3 prophages belonging to the family Podoviridae). Genomic comparison of the 40 prophage genomes led to the identification of four prophages isolated from different strains and of genome sizes of around 33.3, 36.1, 39.6 and 42.6 kb. These prophages showed sequence similarities (query cover >90 %, identity >99.9 %) with international Microbe Versus Phage (MVP) (http://​mvp.​medgenius.​info/​home) clusters 4762, 4901, 3499 and 4280, respectively. Phylogenetic analysis revealed the evolutionary proximity among the members of the four groups of the most fre- quently identified prophages in the bacterial genomes studied (33.3, 36.1, 39.6 and 42.6 kb), with bootstrap values of 100%. This allowed the prophages to be classified into three clusters: A, B and C. Interestingly, these temperate bacteriophages did not infect the highest number of strains as indicated by a host-­range assay, these results could be explained by the development of superin- fection exclusion mechanisms. In addition, bioinformatic analysis of the 40 identified prophages revealed the presence of 2363 pro- teins. In total, 59.7% of the proteins identified had a predicted function, mainly involving viral structure, transcription, replication and regulation (lysogenic/lysis). Interestingly, some proteins had putative functions associated with bacterial virulence (toxin expression and efflux pump regulators), phage defence profiles such as toxin–antitoxin modules, an anti-­CRISPR/Cas9 protein, TerB protein (from terZABCDE operon) and methyltransferase proteins. Data Summary (prophages) of the clinical strains of Klebsiella pneumoniae The genome sequences of the temperate bacteriophages (Spanish Network for Research in Infectious Diseases, REIPI) Received 21 October 2019; Accepted 31 March 2020; Published 29 April 2020 Author affiliations: 1Microbiology Department, Research Institute Biomedical A Coruña (INIBIC), Hospital A Coruña (CHUAC), University of A Coruña (UDC), A Coruña, Spain; 2Study Group on Mechanisms of Action and Resistance to Antimicrobials (GEMARA), Spanish Society of Infectious Diseases and Clinical Microbiology (SEIMC), Madrid; 3Clinical Unit for Infectious Diseases, Microbiology and Preventive Medicine, Hospital Universitario Virgen Macarena. Deparment of Microbiology and Medicine, University of Seville, Seville, Spain; 4Spanish Network for the Research in Infectious Diseases, REIPI, Seville, Spain; 5Microbiology Unit, Maimonides Biomedical Research Institute of Cordoba (IMIBIC), University Hospital Reina Sofía, University of Córdoba, Cordoba, Spain; 6Reference and Research Laboratory for Antibiotic Resistance and Health Care Infections, National Centre for Microbiology, Institute of Health Carlos III, Majadahonda, Madrid, Spain; 7Microbiology Department, Sant Pau Hospital, Autonomous University of Barcelona (Bellaterra), Barcelona, Spain. *Correspondence: Maria Tomas, ​MA.​del.​mar.​tomas.​carmona@​sergas.​es Keywords: Klebsiella pneumoniae; prophages; bioinformatics; genomic analysis; comparative genomics; phylogeny. Abbreviations: HGT, horizontal gene transfer; MDR, multi-­drug resistant; MLST, multilocus sequence typing; MVP, microbe versus phage; NCBI, National Center for Biotechnology Information; SEIMC, Spanish Society of Infectious Diseases and Clinical Microbiology; ST, sequence type; TA, toxin– antitoxin; TEM, transmission electron microscopy. †These authors contributed equally to this work Data statement: All supporting data, code and protocols have been provided within the article or through supplementary data files. Supplementary material is available with the online version of this article. 000369 © 2020 The Authors This is an open-­access article distributed under the terms of the Creative Commons Attribution License. 1 Bleriot et al., Microbial Genomics 2020;6 have been deposited in GenBank under BioProject accession number PRJNA565865 (http://www.​ncbi.​nlm.​nih.​gov/​biopro- Impact Statement ject/​565865). The genome and ORF data for the bacteriophages Klebsiella pneumoniae is a successful multi-­drug resistant are included in this BioProject record. human pathogen and an important source of hospital infections associated with high morbidity and mortality due INTRODUCTION to multiple factors. Temperate bacteriophages (prophages), Bacteriophages, i.e. viruses that infect bacteria, are the most located in the genome of clinical strains of K. pneumoniae, abundant biological entities on Earth [1–3]. They are found in act as mobile elements of horizontal gene transfer, being all environmental niches colonized by bacteria, with an esti- able to modulate the behaviour of bacteria by providing mated global population of 1031 viral particles [4, 5]. Bacterio- virulence and phage defence genes. Therefore, the char- phages have very diverse genomes and have been suggested to acterization of these prophages can lead to the discovery represent the largest source of gene diversity in the environment, of new molecular targets for the development of innovative as highlighted by the large number of novel genes of unknown function revealed by genome and metavirome sequencing [6]. antibacterial treatments. Detailed comparative analysis has been made of the bacterio- phages that infect various hosts and environments, including Mycobacterium [7], Acinetobacter [8], Pseudomonas [9], Bacillus and the various clinical, industrial and biotechnological applica- [10], Lactococcus [11], marine cyanobacterium [12], Salmonella tions [18]. [13] and Vibrio species [14] and also members of the family Enterobacteriaceae [15]. However, the analysis of the families Genomic comparative analysis has highlighted the mosaicism of all sequenced bacteriophages has illustrated how little of present in the genomes of all the dsDNA tailed bacteriophages. the global phage population has been genomically sampled The term mosaicism refers to the fact that bacteriophages [16, 17]. With an almost endless supply of diverse phages readily harbour in their genomes different regions with different accessible for isolation and analysis, research programmes will evolutionary history due to the phenomenon of horizontal continue to play substantial roles in analysing and describing gene transfer (HGT) [16, 17]. The phylogeny and evolutionary functional proteins in the phages to understand their ecology relationships between bacteriophages isolated from bacteria Table 1. Characteristics of the K. pneumoniae clinical strains and their origins (all isolated in Spain between November 2013 and April 2014) Strain MLST* Carbapenemase Origin GenBank accession no.† Genome size (bp) G+C (mol%) ST405-­OXA48 ST-405 OXA48 Wound WRXJ00000000 5 774 094 57.0 ST15-­VIM1 ST-15 VIM1 Blood WRXI00000000 5 122 717 57.0 ST11-­OXA245 ST-11 OXA245 Wound WRXH00000000 5 667 817 57.1 ST437-­OXA245 ST-437 OXA245 Rectal WRXG00000000 5 653 384 57.2 ST16-­OXA48 ST-16 OXA48 Urine WRXF00000000 5 447 971 57.3 ST101-­KPC2 ST-101 KPC2 Rectal WRXE00000000 5 523 997 57.2 ST147-­VIM1 ST-147 VIM1 Rectal WRXD00000000 5 695 918 56.8 ST11-­VIM1 ST-11 VIM1 Respiratory WRXC00000000 5 594 690 57.2 ST846-­OXA48 ST-846 OXA48 Sputum WRXB00000000 5 548 215 57.3 ST340-­VIM1 ST-340 VIM1 Rectal WRXA00000000 5 539 743 57.2 ST13-­OXA48 ST-13 OXA48 Rectal WRWZ00000000 5 585 239 57.0 ST512-­KPC3 ST-512 KPC3 Axillary smear WRWY00000000 5 650 413 57.2 ST15-­OXA48 ST-15 OXA48 Axillary smear WRWX00000000 5 188 264 57.3 ST11-­OXA48 ST-11 OXA48 Urine WRWW00000000 5 509 234 57.3 ST258-­KPC3 ST-258 KPC3 Urine WRWV00000000 5 564 025 57.3 ST974-­OXA48 ST-974 OXA48 Urine WRWT00000000 5 414 512 57.3 *Institute Pasteur MLST database (https://pubmlst.org/kpneumoniae). †GenBank accession numbers for the genomes of the 16 K. pneumoniae clinical strains, all belonging to BioProject accession number PRJNA565865 (https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA565865). 2 Bleriot et al., Microbial Genomics 2020;6 Table 2. Characteristics of the genome sequences of the 40 prophages found in 16 clinical strains of carbapenemase-­producing K. pneumoniae (BioProject accession number PRJNA565865; https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA565865) Bacteriophage Family Accession no. Genome size No. of ORFs Hypothetical G+C (mol%) tRNA (bp) protein (%) Klebsiella pneumoniae ST405-­OXA48 ST405-­OXA48phi1.1 Myoviridae MK388859.1 35.232 52 48.1 53.8 0 ST405-­OXA48phi1.2 Myoviridae MK416007.1 40.495 67 59.7 52.1 0 ST405-­OXA48phi1.3 Siphoviridae MK416008.1 32.010 47 66.0 50.9 0 Klebsiella pneumoniae ST15-­VIM1 ST15VIM1phi2 Myoviridae MK448228.1 46.342 73 54.8 53.6 0 Klebsiella pneumoniae ST11-­OXA245 ST11-­OXA245phi3.1 Myoviridae MK416009.1 33.326 44 20.5 51.5 0 ST11-­OXA245phi3.2 Podoviridae MK416010.1 60.118 72 37.5 55.6 0 Klebsiella pneumoniae ST437-­OXA245 ST437-­OXA245phi4.1 Myoviridae MK416011.1 39.642 53 26.4 52.8 0 ST437-­OXA245phi4.2 Myoviridae MK416012.1 18.281 27 25.9 50.5 0 Klebsiella pneumoniae ST16-­OXA48 ST16-­OXA48phi5.1 Siphoviridae MK416013.1 57.025 80 50.0 53.6 0 ST16-­OXA48phi5.2 Myoviridae MK448230.1 47.305 79 54.8 53.8 4 ST16-­OXA48phi5.3 Siphoviridae MK416014.1 29.301 51 49.0 50.2 0 ST16-­OXA48phi5.4 Myoviridae MK416015.1 38.322 46 28.3 50.1 0 Klebsiella pneumoniae ST101-­KPC2 ST101-­KPC2phi6.1 Myoviridae MK448231.1 48.131 75 53.2 52.6 4 ST101-­KPC2phi6.2 Myoviridae MK416016.1 11.454 17 11.8 59.0 0 ST101-­KPC2phi6.3 Siphoviridae MK416017.1 43.942 62 43.1 52.2 3 Klebsiella pneumoniae ST147-­VIM1 ST147-­VIM1phi7.1 Myoviridae MK416018.1 34.141 43 20.5 53.0 1 ST147-­VIM1phi7.2 Podoviridae MK448232.1 34.200 45 38.6 40.0 0 Klebsiella pneumoniae ST11-­VIM1 ST11-­VIM1phi8.1 Myoviridae MK448233.1 42.666 67 37.0 50.7 2 ST11-­VIM1phi8.2 Podoviridae MK448234.1 48.230 68 41.7 52.9 0 ST11-­VIM1phi8.3 Myoviridae MK416019.1 39.953 62 45.1 51.6 0 ST11-­VIM1phi8.4 Myoviridae MK416020.1 33.016 43 20.9 51.6 0 Klebsiella pneumoniae ST846-­OXA48 ST846-­OXA48phi9.1 Siphoviridae MK416021.1 38.370 38 36.8 53.3 0 ST846-­OXA48phi9.2 Myoviridae MK416022.1 57.402 91 63.7 51.6 0 Klebsiella pneumoniae ST340-­VIM15 ST340-­VIM1phi10.1 Myoviridae MK422455.1 36.124 58 58.6 51.7 0 ST340-­VIMphi10.2 Myoviridae MK422454.1 33.326 45 24.4 51.5 0 Continued 3 Bleriot et al., Microbial Genomics 2020;6 Table 2. Continued Bacteriophage Family Accession no. Genome size No. of ORFs Hypothetical G+C (mol%) tRNA (bp) protein (%) Klebsiella pneumoniae ST13-­OXA48 ST13-­OXA48phi12.1 Myoviridae MK422453.1 39.086 50 19.6 53.5 1 ST13-­OXA48phi12.2 Siphoviridae MK422452.1 34.141 51 49.0 50.5 0 ST13-­OXA48phi12.3 Siphoviridae MK422451.1 84.199 93 47.7 49.9 0 ST13-­OXA48phi12.4 Siphoviridae MK422450.1 59.049 86 57.5 52.1 0 ST13-­OXA48phi12.5 Siphoviridae MK714353.1 44.913 72 58.9 51.8 2 Klebsiella pneumoniae ST512-­KPC3 ST512-­KPC3phi13.1 Myoviridae MK448235.1 42.666 67 47.8 52.9 2 ST512-­KPC3phi13.2 Myoviridae MK422449.1 32.302 44 25.0 51.9 0 ST512-­KPC3phi13.5 Myoviridae MN166823.1 25.624 45 55.6 52.0 0 ST512-­KPC3phi13.6 Myoviridae MK433577.1 39.643 52 28.8 52.8 0 Klebsiella pneumoniae ST15-­OXA48 ST15-­OXA48phi14 Myoviridae MK433578.1 33.839 47 21.3 52.7 0 Klebsiella pneumoniae ST11-­OXA48 ST11-­OXA48phi15.1 Myoviridae MK433579.1 36.137 58 55.2 51.7 0 ST11-­OXA48phi15.3 Myoviridae MK433580.1 33.326 44 20.5 51.5 0 Klebsiella pneumoniae ST258-­KPC3 ST258-­KPC3phi16.1 Myoviridae MK433581.1 39.643 53 32.1 52.8 0 ST258-­KPC3phi16.2 Myoviridae MK433582.1 33.326 45 22.2 51.5 0 Klebsiella pneumoniae ST974-­OXA48 ST974-­OXA48phi18 Siphoviridae MK448237.1 51.967 80 54.3 52.9 1 have been analysed in several studies [19, 20], confirming their acquired through HGT, such as vagCD genes, which encode a importance. The phenomenon of HGT, as well as bacteriophages functional broad-­spectrum TA system and are conserved on themselves, act as drivers of the evolution and diversification of the large multiple-­antibiotic-­resistance-­conferring plasmids in bacteria, allowing them to acquire virulence factors (e.g. diph- this species [29]. theria or botulinum toxins) and/or genes related to metabolism, antibiotic resistance (e.g. β-lactamases) and adaptation to new In this study, we analysed the genomic characteristics of 40 environmental niches [21–23]. genomes of prophages identified in 16 clinical isolates of carbapenem-­producing K. pneumoniae. Moreover, we used Klebsiella pneumoniae, which belongs to the family Enterobac- comparative bioinformatic tools and microscopic techniques teriaceae, is a Gram-­negative opportunistic bacterial pathogen to investigate the phylogeny and evolutionary relationships of associated with a wide range of diseases, such as urinary tract the prophages. infections, pneumonia and septicaemia, as well as infection of wounds and soft tissues [24]. In recent years, different strains of this species have acquired genes for resistance to antibiotics, METHODS especially genes that encode enzymes capable of breaking down most β-lactamases [25]. This has led to the generation Origin of the K. pneumoniae genomes of multi-­drug resistant (MDR) pathogens, which constitute a In this study, we analysed 16 isolates of K. pneumoniae serious public-­health problem [26]. The success of K. pneumo- belonging to different sequence types (STs), previously niae as a nosocomial pathogen is due to its intrinsic virulence, determined by multilocus sequence typing (MLST), isolated attributed to its ability to cause invasive infection via fimbrial from clinical samples and harbouring the most prevalent adhesins [27] and to the presence of a thick capsule, which acts carbapenemase genes (OXA 48, VIM and KPC β-lactamases) as a possible antiphagocytic factor [28], and to toxin–antitoxin (Table 1). The STs and carbapenemase β-lactamases were (TA) systems associated with the stability of mobile elements previously determined by Esteban-­Cantos et al., according 4 Bleriot et al., Microbial Genomics 2020;6 Table 3. Comparison by sequence of prophage genomes and bacterial genomes Prophage size (bp)* Prophage Bacterial strain MPV international cluster‡ Name Genome accession Name Genome accession no.† no.† 33 226 ST11-­OXA245phi3.1 MK416009.1 ST11-­OXA245 WRXH00000000 4762 ST437-­OXA245phi4.2 MK416012.1 ST437-­OXA245 WRXG00000000 ST11-­VIM1phi8.4 MK416020.1 ST11-­VIM1 WRXC00000000 ST340-­VIM1phi10.2 MK422454.1 ST340-­VIM1 WRXA00000000 ST512-­KPC3phi13.2 MK422449.1 ST512-­KPC3 WRWY00000000 ST11-­OXA48phi15.3 MK433580.1 ST11-­OXA48 WRWW00000000 ST258-­KPC3phi16.2 MK433582.1 ST258-­KPC3 WRWV00000000 36 124 ST340-­VIM1phi10.1 MK422455.1 ST340-­VIM1 WRXA00000000 4901 ST11-­OXA48phi15.1 MK433579.1 ST11-­OXA48 WRWW00000000 39 643 ST437-­OXA48phi4.1 MK416011.1 ST437-­OXA48 WRXG00000000 3499 ST512-­KPC3phi13.6 MK433577.1 ST512-­KPC3 WRWY00000000 ST258-­KPC3phi16.1 MK4333581.1 ST258-­KPC3 WRWV00000000 42 666 ST11-­VIM1phi8.1 MK448233.1 ST11-­VIM1 WRXC00000000 4280 ST512-­KPC3phi13.1 MK448235.1 ST512-­KPC3 WRWY00000000 *Prophage genomes located in the genomes of different K. pneumoniae clinical strains have been placed into four size groups (33.3, 36.1, 39.6 and 42.6 kb). †Accession numbers of genomes from BioProject accession no. PRJNA565865. ‡MVP international clusters 4762, 4901, 3499 and 4280 from the MVP database (https://mvp.medgenius.info). to the methods proposed by the Pasteur Institute (https://​ the bacteriophage regions and the sequences of the assembled bigsdb.​pasteur.​fr/​klebsiella/​klebsiella.​html) [30]. The study bacterial genome. Focusing on bacteriophage clusters of the highlighted that K. pneumoniae ST258 and ST15 are high-­risk Microbe Versus Phage (MVP) database (http://​mvp.​medge- clones in the worldwide spread of carbapenemases. nius.​info) in K. pneumoniae strains, we searched for possible homologies between these groups and the prophages under Genomic DNA was isolated from the strains with a Wizard study. For this purpose, we used the Position-­Specific Iterative genomic DNA kit (Promega), following the manufac- Basic Local Alignment Search Tool (psi-­blast) to analyse the turer’s protocol for extraction, for subsequent sequencing complete prophage genome against the representative sequence of the different genomes with the MiSeq system (Illumina). Sequences of 250 bp paired-­end reads of each isolate were (i.e. the longest) of the cluster and applied the following cut off assembled ‘de novo’ with Velvet v.1.2.10 (https://www.​ebi.​ values: query coverage >60 %, identity >89 %. ac.​uk/~​zerbino/​velvet/). All the assembly attributes were The MVP database can be used to study phage–host relation- included in GenBank BioProject PRJNA565865 (https://www.​ ships [32], as it provides the user with an important list of ncbi.​nlm.​nih.​gov/​bioproject/?​term=​PRJNA565865). interactions in addition to international clusters (sets of viral sequences grouped according to their sequence similarities). Identification of prophages in the genome of K. We searched the database for possible sequence homologies pneumoniae isolates between the 40 prophages under study and the international Prophages in the genomes of different strains of K. pneumo- clusters. We observed homology with 25 international clusters niae were identified with the phaster (Phage Search Tool of K. pneumoniae (549, 934, 1013, 1319, 1690, 1920, 2329, –Enhanced Release) bioinformatics tool (http://​phaster.​ca/). 2808, 3207, 3429, 3546, 3969, 4280, 4331, 4575, 4762, 4864, Only those temperate bacteriophages identified by the program 4868, 4901, 5263, 8737, 9608, 9912, 10255, 11441), by applying as intact (score >90) were used for analysis in the present study. the aforementioned cut-­off values. The comparative genomic analysis of the four K. pneumoniae prophages found in different isolates of K. pneumoniae was Integration of prophages carried out with the blast Ring Image Generator (brig) In the final step of the study, the integration sites of the program [31], which uses the sequence similarity between prophages were identified with the Standard Nucleotide 5 Bleriot et al., Microbial Genomics 2020;6 Fig. 1. Comparative genomic analysis of K. pneumoniae temperate bacteriophages displaying the highest sequence identity with cluster 4762 (query cover >99.9 %; identity >99.8 %) from the MVP database (https://mvp.medgenius.info), constructed with the brig program. The sequence (ST11-­OXA245phi3.1) is located on the innermost side and is indicated in black. The double ring adjacent to the reference sequence represents the G+C content (black) and the G+C skew (purple and green). The other rings are as indicated in the key. The white parts of the rings represent absent or divergent content relative to the unknown sequence. Most of the prophages were approximately 33.3 kb in size (ST11-­VIM1phi8.4, ST340-­VIM1phi10.2, ST512-­KPC3phi13.2, ST11-­OXA48phi15.3 and ST258-­KPC3phi16.2). Basic Local Alignment Search Tool (blastn) by searching gaps/missing data were treated as complete deletions; (iii) the for nucleotide sequence similarity between the K. pneumo- bootstrap consensus tree was inferred from 1000 replicates niae prophages and all sequences in the National Center [36]; and (iv) the condensed tree was displayed with a value for Biotechnology Information (NCBI) database. Groups of of 100 %. proteins located before the integrase proteins were considered integration sites. Isolation of the bacteriophages and transmission electron microscopy (TEM) studies Phylogenic relationships among the 40 temperate Mitomycin C was used to induce bacteriophage production bacteriophages predicted to belong to the order following the protocol described by López et al. [8]. For this Caudovirales purpose, 15 ml Luria–Bertani (LB) broth was inoculated with The dot plot alignment of the nucleotide sequences of the 40 150 µl overnight culture of strains and incubated with shaking temperate bacteriophages was inferred using Genome Pair (180 r.p.m.) at 37 °C, until the optical density measured at Rapid Dotter (gepard) [33], and a fasta file was constructed wavelength 600 nm (OD600) reached 0.5. Mitomycin C was with all the sequences. Bacteriophages lack a universal then added (10 µg ml−1) and the culture was allowed to grow, marker gene for phylogenetic analysis; however, the use of with shaking (180 r.p.m.) at 37 °C, until lysis occurred, i.e. the terminase large subunit of the tail protein or the major when the culture appeared clear (after approximately 2–4 h). capsid protein is often reported [34]. In the present study, a The lysates were centrifuged at 3500 r.p.m. for 10 min, and the phylogenetic analysis of the sequences of the terminase large supernatant was filtered through a 0.22 nm filter (Millipore subunit was performed with the ClustalW program (http:// Express PES membrane; Merck). NaCl was added (to a final www.​ebi.​ac.​uk/​clustalw/) in mega x software [35]. A tree was concentration of 0.5 M), and the suspensions were then mixed generated by multiple alignments, with the neighbour-­joining and left on ice for 1 h. The suspensions were centrifuged at method. The following parameters were applied to produce 3500 r.p.m. for 40 min at 4 °C, and the supernatants were the tree: (i) the number of differences was established; (ii) transferred to sterile tubes. PEG 6000 (10 %, w/v) was added 6 Bleriot et al., Microbial Genomics 2020;6 Fig. 2. Comparative genomic analysis of K. pneumoniae temperate bacteriophages displaying the highest sequence identity with cluster 4901 (% query cover >98.9; % identity >99.9) of the MVP database [https://mvp.medgenius.info], constructed with the BRIG program. The sequence (ST340-­VIM1phi10.1) is located on the innermost side and is indicated in black. The double ring adjacent to the reference sequence represents the GC content (black) and the GC skew (purple and green). The white part of the rings represents absent or divergent content relative to the unknown sequence. All prophages were of size 36.1 kb (ST11-­OXA48phi15.1). and dissolved by rocking the tubes at room temperature for bacteriophage mix solutions were added to the TA agar plates. 1 h and incubating them overnight at 4 °C. Bacteriophages A negative control consisting of SM buffer was included for were then precipitated at 3500 r.p.m. for 40 min at 4 °C and each plate. resuspended in SM buffer (0.1 M NaCl, 1 mM MgSO4, 0.2 M Tris-­HCl, pH 7.5) [37]. Finally, the samples were stored at ORF annotation 4 °C until processed for TEM on a JEOL JEM-1011 electron microscope. The putative functions of the ORFs were determined by sequence identity with the Rapid Annotation by Subsystem Induced bacteriophage mix spot test Technology (rast) server (http://​rast.​nmpdr.​org/) and the blast-­protein tool psi-­blast developed by the NCBI (https://​ Spot tests were carried out using the mixed induced samples blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). In addition, we used the of bacteriophages obtained from the purified PEG prepara- HHpred tool in the MPI informatics toolkit (https://​toolkit.​ tion used for TEM. All of them contained representative members of each cluster (A, B, C, D and E). The spot test tuebingen.​mpg.​de/#​hhpred/) [39], which predicts functions method was a modified version of the protocol described through protein structure with a model accuracy that is by Raya and Hebert [38]. When the OD600 reached 0.6 nm, competitive with that of the best servers in CASP8. The cut-­ 200 µl culture from the 16 K. pneumoniae isolates was mixed off E value for annotation of proteins was <10−5 and all ORFs with 4.5 ml soft agar (0.5 % NaCl, 1 % tryptone and 0.4 % agar) with E values >105 were annotated as hypothetical proteins and poured onto TA agar plates (0.5 % NaCl, 1 % tryptone (Critical Assessment of Technique for Protein Structure and 1.5 % agar). Once the soft medium had solidified, 15 µl Prediction) [40]. 7 Bleriot et al., Microbial Genomics 2020;6 Fig. 3. Comparative genomic analysis of K. pneumoniae temperate bacteriophages displaying the highest sequence identity with cluster 3499 (% query cover >92; % identity>99.9) of the MVP database [https://mvp.medgenius.info], constructed with the BRIG program. The sequence (ST258-­KPC3phi16.1) is located on the innermost side and is indicated in black. The double ring adjacent to the reference sequence represents the GC content (black) and the GC skew (purple and green). The whitepart of the rings represents absent or divergent content relative to the unknown sequence. All prophages wereof size 39.6 kb (ST437-­OXA245phi4.1 and ST512-­KPC3phi13.6). RESULTS prophages (ST11-­OXA245phi3.1, ST340-­VIM1phi10.2, Predicted prophage genomes ST11-­OXA48phi15.3 and ST258-­KPC3phi16.2). The genome sequences also displayed a 100% sequence identity with MVP Whole-­ genome sequencing of the 16 carbapenemase-­ database cluster 4762, which has the highest host range (67 producing clinical strains of K. pneumoniae (Table 2) and the bacterial strains) of all phage clusters included in the MVP use of the phaster bioinformatics tool (http://​phaster.​ca/) revealed the presence of a total of 40 prophages considered database and is also known to infect K. pneumoniae, among intact (score >90), all belonging to the order Caudovirales. other hosts (Fig. 1). Moreover, the 33.3 kb genomes displayed Most were members of the family Myoviridae (27 prophages), only partial sequence identity with genomes of the other two and the others belonged to the families Siphoviridae (10 prophages (32.302 kb ST512-­KPC3phi13.2 and 33.016 kb ST11-­ prophages) and Podoviridae (3 prophages). The families were VIM1phi8.4), possibly due to their shorter size. We believe that assigned by sequence homologies with the most common prophage ST437-­OXA245phi4.2, also belonging to the 33.3 kb bacteriophage indicated by phaster in the Virus-­Host Data- cluster, is an incomplete prophage, as its genome is about half the base (https://www.​genome.​jp/​virushostdb/). size of that of the 33.3 kb genomes (18.281 kb versus 33.326 kb). The sequence of the 36.1 kb prophage genome was identical in After annotation of the prophage genomes (see the Supple- two prophages (ST340-­VIM1phi10.1 and ST11-­OXA48phi15.1) mentary Material, available with the online version of this and it was also identical to cluster 4901 (which showed a article) with rast, blastp and HHpred, it was evident that four of them (33.3, 36.1, 39.6 and 42.6 kb) were represented in host range of 57 bacterial strains) (Fig. 2). The sequence of different strains of K. pneumoniae (in seven, two, three and two the 39.6 kb prophage was identical to the genomes of three strains, respectively) (Table 3). Comparative genomic analysis prophages (ST437-­OXA48phi4.1, ST512-­KPC3phi13.6 and of these prophages was performed with the brig program ST258-­KPC3phi16.1). In addition, cluster 3499 (which showed according to their sequence similarities (Figs 1–4). Comparison a 19 strain host range) had partial sequence identity with the of the sequences with different prophage sequences and clus- three prophages (Fig. 3). Finally, the sequence of the 42.6 kb ters from the MVP database (see Table 3, Figs 1–4) showed temperate bacteriophage (prophage) (ST512-­KPC3phi13.1) had that the 33.3 kb prophage genome sequences derived from an identical sequence in the prophage (ST11-­VIM1phi8.1) and different K. pneumoniae isolates were identical in multiple the cluster 4280 (Fig. 4). 8 Bleriot et al., Microbial Genomics 2020;6 Fig. 4. Comparative genomic analysis of K. pneumoniae temperate bacteriophages displaying the highest sequence identity with cluster 4280 (query cover >98.9 %; identity >99.9 %) from the MVP database (https://mvp.medgenius.info) analysed with the brig program. The sequence (ST512-­KPC3phi13.1) is located on the innermost side and is indicated in black. The double ring adjacent to the reference sequence represents the G+C content (black) and the G+C skew (purple and green). The other rings are as indicated in the key. The white parts of the rings represent absent or divergent content relative to the unknown sequence. All prophages were of size 42.6 kb (ST11-­ VIM1phi8.1). Integration sites of prophages the disruption of the adjacent gene. (iv) Four prophages were In all prophages, the attachment sites (attL and attR) were integrated within the SapABCDEF operon that encodes an identified by phaster. Moreover, by studying the integra- ATP-­ binding cassette (ABC) transporter, more precisely tion sites of the prophages, we observed that these were between sapB and sapC intact gene (ST16-­OXA48phi5.1, ST11-­ integrated at various sites, as detailed in the following text. VIM1phi8.2, ST846-­OXA48phi9.2 and ST974-­OXA48phi18). (i) Fourteen prophages were integrated before or after intact (v) Five prophages were integrated immediately after an intact host tRNA (ST405-­ OXA48phi1.2, ST405-­ OXA48phi1.3, protease (ST16-­ OXA48phi5.3, ST11-­ VIM1phi8.3, ST340-­ ST15-­VIM1phi2, ST437-­OXA245phi4.1, ST101-­KPC2phi6.1, VIM1phi10.1, ST512-­KPC3phi13.5 and ST11-­OXA48phi15.1). ST101-­KPC2phi6.2, ST147-­VIM1phi7.2, ST11-­VIM1phi8.1, (vi) Five prophages were integrated next to a protein of unknown ST13-­OXA48phi12.1, ST13-­OXA48phi12.2, ST13-­ function (ST101-­ KPC2phi6.3, ST147-­ VIM1phi7.1, ST13-­ OXA48phi12.5, ST512-­KPC3phi13.1, ST512-­KPC3phi13.6 and OXA48phi12.3, ST13-­OXA48phi12.4 and ST15-­OXA48phi14). ST258-­KPC3phi16.1), with tRNA-­arg being the most common In the case of ST15-­OXA48phi14, the phage integration trun- host tRNA present before prophages in six cases. (ii) Seven cated the protein of unknown function. (vii) One prophage was prophages were integrated between the TerT transcriptional integrated after a sensor domain-­containing diguanylate cyclase, regulator intact genes and transporter intact genes (all bacterio- truncating the latter (ST16-­OXA48phi5.4). phages 33.3 kb: ST11-­OXA245phi3.1, ST437-­OXA245phi4.2, ST11-­VIM1phi8.4, ST340-­VIM1phi10.2, ST512-­KPC3phi13.2, Phylogenetic study ST11-­OXA48phi15.3 and ST258-­KPC3phi16.2), integration of To observe the phylogenetic relationships between the the prophages did not truncate these genes. (iii) Four prophages temperate bacteriophages under study and the clusters were integrated next to the host transcriptional regulator (ST405-­ included in the MVP database, we reconstructed a neighbour-­ OXA48phi1.1, ST11-­OXA245phi3.2, ST16-­OXA48phi5.2 and joining tree by the dot plot alignment of nucleotide sequences, ST846-­OXA48phi9.1). In some cases (ST16-­OXA48phi5.2, constructed from a fasta file with all genomes of the 40 ST846-­OXA48phi9.1), the integration of the prophage involved prophages studied with the gepard program. This enabled 9 Bleriot et al., Microbial Genomics 2020;6 Fig. 5. (a) Dot plot alignment of nucleotide sequences from 40 K. pneumoniae temperate bacteriophage genomes inferred using the gepard tool. For this purpose, a fasta file was constructed with all sequences. Black diagonal lines parallel to the main diagonal indicate strong sequence similarity, while grey lines indicate lower sequence similarity. The differently coloured boxes illustrate the assignment of clusters A (A1; A2), B, C, D and E, and are represented in red, blue, yellow, pink and green, respectively. The bacteriophage families are shown on the horizontal axis (M, Myoviridae; P, Podoviridae; S, Siphoviridae). (b) The evolutionary history of the large terminase subunit protein of 40 K. pneumoniae temperate bacteriophages and the representative MVP clusters were inferred using the neighbour-­joining method. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. The evolutionary distances were computed using the number of differences method. All positions containing gaps and missing data were eliminated. Evolutionary analyses were conducted in mega x [35]. us to define five clusters: A, B, C, D and E (Fig. 5a). Cluster bacteriophages (Fig. 6a–e), a long, flexible tail in the Sipho- A was subdivided because it included two different groups of viridae bacteriophages (Fig. 6g, h) and a small tail in the temperate bacteriophages (A1 and A2). Podoviridae bacteriophages (Fig. 6f, i). Moreover, we reconstructed a neighbour-­joining tree (Fig. 5b) with a packaging protein, the terminase large subunit, in the Induced bacteriophage mix spot test tree with bootstrap values (100 %) [41]. The four temperate The lytic potential of induced bacteriophages mixes with one bacteriophages, which were included in different strains phage representative of each cluster (previously observed considered in the study (33.3, 36.1, 39.6 and 42.6 kb, together by TEM) was assayed by the spot test in all the isolates of with their international clusters from MVP database, 4762, K. pneumoniae (Table 4). A spot was observed in one strain 4901, 3499 and 4280), were included in clusters and subclus- for all the bacteriophage mixes almost. The induced bacte- ters: A1, C, A2 and B, respectively (Table 3, Fig. 5b). riophage mix 3 derived from the strain ST101-­KPC2, which harbours ST101-­KPC2phi6.1 (48 131 bp; cluster B), ST101-­ TEM studies KPC2phi6.2 (11 454 bp; not included in any cluster) and TEM images revealed the presence of the three families of ST101-­KPC2phi6.3 (43 942 bp; not included in any cluster), bacteriophages in the order Caudovirales. Representative was able to produce a halo with five clinical isolates. Despite images of the bacteriophages in each cluster are shown a large amount of halo, it is not possible to know whether it in Fig. 6. The typical morphology of the different families is due to the lytic capacity of bacteriophages or due to the was observed, including a long, rigid tail in the Myoviridae phenomenon of lysis from without. 10 Bleriot et al., Microbial Genomics 2020;6 Fig. 6. TEM images showing the different families of prophages present in different clusters. (a–e) The family Myoviridae obtained from ST11-­OXA245phi3.1, ST13-­OXA48phi12.1, ST101-­KPC2phi6.1, ST405-­OXA48phi1.2 and ST846-­OXA48phi9.2 prophages belonging to A1, A2, B, C and D clusters, respectively. (f and i) The family Podoviridae obtained from ST11-­VIM1phi8.2 and ST147-­VIM1phi7.2 prophages belonging to D and E clusters, respectively. (g and h) The family Siphoviridae obtained from ST974-­OXA48phi18 and ST13-­OXA48phi 12.2 prophages belonging to D and E clusters, respectively. Annotation of the prophages find lysogenic genes, such as integrase, in all of the prophages Using rast, psi-­blast and HHprep tools, we obtained studied. In fact, 40.47 % of the studied prophages lacked inte- functional information for 59.7 % of the prophage proteins grase. Intending to study the insertion site of all the bacte- analysed (Fig. 7, Supplementary Material). riophages included in this study, we used the blastn tool to identify the region of the bacterial genome where this integra- The most abundant functional category found in the prophage tion took place. This analysis revealed that the integrase was genomes involved genes that constitute the minimum basic located upstream from what phaster considered to be the unit (universally presented genetic markers) of a Caudovi- initial site in those bacteriophages that appeared to lack inte- rales bacteriophage, i.e. the genes related to structure and grase, because this program is not able to identify sequences packaging, lysis, lysogenesis, transcription, replication and from two different contigs. In addition, the prophage genomes regulation (Fig. 7). One of the most important functions of were also composed of genes (ORFs) involved in the phage– the bacteriophage tail, and more precisely of the tail fibre and host interaction: virulence factors and phage defence genes the baseplate protein, is to recognize the host surface recep- tors. In addition to these structural genes, the minimum unit (Fig. 8) [42]. of a bacteriophage was also composed of genes involved in The following virulence factors were identified in the collec- packaging. The packaging machine of most bacteriophages tion of temperate bacteriophages of the K. pneumoniae clinical contains two proteins called the terminase small subunit and strains: (i) invasion-­associated protein B T4SS (type IV secre- the terminase large subunit [34]. In this study, the terminase tion system) [43] (GenBank accession number QBQ71533.1; large subunit (approx. 1700 bp) was identified and annotated E value 1.60×10−11) from prophage ST512-­KPC3phi13.1; (ii) in all genomes of the temperate bacteriophages; however, the viral coat protein (transferase-­kinase) [44, 45] (GenBank acces- same did not occur with the small subunit (approx. 600 bp), sion numbers QBP27909.1, QBP08028.1 and QBP27756.1; which was only identified and annotated in 28 temperate E values 4×10−37, 4×10−37 and 4×10−37, respectively) in bacteriophage genomes. prophages ST258-­KPC3phi16.1, ST437-­OXA245phi4.1 We also detected lysis genes involved in the lysis of the bacte- and ST512-­KPC3phi13.6, respectively; and (iii) the MarR rial cell, such as the holin gene and lysozyme-­encoding gene, family of transcriptional regulators [46] involved in regu- in all viral genomes [42]. However, we did not consistently lating cellular processes such as antibiotic resistance and the 11 Bleriot et al., Microbial Genomics 2020;6 Table 4. Induced bacteriophage mix spot test Spot test assay of seven induced bacteriophage mixes isolated from seven clinical K. pneumoniae isolates. The infectivity of these cocktails was tested in the 16 K. pneumoniae strains studied. Bacteriophage mix 1, induced from ST405-­OXA48, is composed of ST405-­OXA48phi1.1, ST405-­OXA48phi1.2 and ST405-­OXA48phi1.3; bacteriophage mix 2, induced from ST11-­OXA245, is composed of ST11-­OXA245phi3.1 and ST11-­OXA245phi3.2; bacteriophage mix 3, induced from ST101-­KPC2, contains the bacteriophages ST101-­KPC2phi6.1, ST101-­KPC2phi6.2 and ST101-­KPC2phi6.3; bacteriophage mix 4, induced from ST147-­VIM1, contains ST147-­VIM1phi7.1 and ST147phi7.2; bacteriophage mix 5, induced from ST11-­VIM1, is composed of ST11-­ VIM1phi8.1, ST11-­VIM1phi8.2, ST11-­VIM1phi8.3 and ST11-­VIM1phi8.4; bacteriophage mix 6, induced from ST846-­OXA48, contains ST846-­OXA48phi9.1 and ST846-­OXA48phi9.2; and bacteriophage mix 7, induced from ST13-­OXA48, contains ST13-­OXA48phi12.1, ST13-­OXA48phi12.2, ST13-­OXA48phi12.3, ST13-­OXA48phi12.4 and ST13-­OXA48phi12.5. The induced bacteriophage mix 3 was able to produce halos in the highest number of strains, causing a halo in five different strains in five clinical isolates. −, Absence of lysis plaques; +, presence of lysis plaques. Strain Control (SM buffer) Bacterial bacteriophage mix 1 2 3 4 5 6 7 ST405-­OXA48 − + − + − − − + ST15-­VIM1 − − − − − − − − ST11-­OXA245 − − − − − − − − ST437-­OXA245 − − − + − − − − ST16-­OXA48 − − − + − − − − ST101-­KPC2 − − − − − − − − ST147-­VIM1 − − − − − − − − ST11-­VIM1 − − + − + − − − ST846-­OXA48 − − − − − − − − ST340-­VIM1 − − − − − − − − ST13-­OXA48 − − − − − − − − ST512-­KPC3 − − − + − + + + ST15-­OXA48 − − − − − − − − ST11-­OXA48 − + + − + + + + ST258-­KPC3 − − − + − − − − ST974-­OXA48 − − − − − − − − expression of virulence factors in several members of the QBP07854.1 and QBP07855.1, E value 5.4×10−4; and family Enterobacteriaceae. This MarR-­like protein regulator GenBank accession numbers QBP07684.1 and QBP07685.1, (GenBank accession numbers QEA09493.1, QBP27467.1, E value 1.2×10−8). (ii) The presence of two CRISPR-­associated QBP28293.1, QBQ71610.1, QBP28507.1 and QBP28244.1: E endoribonuclease Cas2 located in prophages ST512-­ values 2.00×10−68, 4.60×10−12, 9.00×10−69, 1.6×10−14, 9.00×10−69 KPC3phi13.6 and ST437-­OXA245phi4.1 (GenBank accession and 3.00×10−68, respectively) was located in prophages ST13-­ numbers QBP27775.1 and QBP08047.1; E values 4.9×10-14 OXA48phi12.5, ST13-­OXA48phi12.3, ST16-­OXA48phi5.2, and 6.2×10-14, respectively) [48]. (iii) The presence of one ST101-­KPC2phi6.3, ST405-­OXA48phi1.3, ST11-­VIM1phi8.2 putative anti-­CRISPR/Cas9 protein, AcrIIC3-­like, located in and ST15-­VIM1phi2.1. prophage ST846-­OXA48phi9.2 (GenBank accession number QBP0798.1; E value 3.7×10−8) [48]. (iv) The presence of TerB Regarding phage defence against attack from bacteria as well protein [49] (GenBank accession number QBP27498.1; E as other viruses, we highlight the following. (i) The presence value 6.00×10−93) in prophage ST13-­OXA48phi12.3 from of four TA modules, two of which had partial sequence iden- the operon terZABCDEF, which confers tellurite resistance tity with RelBE-­like TA proteins [47, 48] (GenBank accession and has been related to the inhibition of bacteriophages. (v) numbers QBP08163.1 and QBP08164.1, E value 8.3×10−4; and The presence of 32 ORFs of methyltransferase proteins [44] GenBank accession numbers QBP07972.1 and QBP07973.1, in 25 prophages from these K. pneumoniae clinical isolates E value 1.2×10−7) located in prophages ST405-­OXA48phi1.2 (Table 5). and ST16-­OXA48phi5.3, respectively. The other two TA systems showed sequence identity with HigBA-­ like TA Finally, we observed the presence of tRNA in the genome of modules located in the prophages ST11-­VIM1phi8.3 and eight prophages (ST16-­OXA48phi5.2, ST101-­KPC2phi6.1, ST846-­ OXA48phi9.2 [48] (GenBank accession numbers ST147-­VIM1phi7.1, ST11-­VIM1phi8.1, ST13-­OXA48phi12.1, 12 Bleriot et al., Microbial Genomics 2020;6 Fig. 7. Numbers of ORFs carried by the prophages of each strain analysed, with associated functional categories. ST13- OXA48phi12.5, ST512-­KPC3phi13.1 and identity with international bacteriophage clusters 4762, 4901, ST974-­OXA48). In most cases (ST16-­OXA48phi5.2, ST101-­ 3499 and 4280 (MVP database), respectively, which are the KPC2phi6.1, ST11-­VIM1phi8.1, ST13-­OXA48phi12.5, most prevalent clusters in clinical strains of K. pneumoniae. ST512-­KPC3phi13.1 and ST974-­OXA48), the tRNAs were Interestingly, these temperate bacteriophages were included located between the phage antitermination protein and lysis-­ by phylogenetic relationships in three main clusters (A, B and related proteins, such as holin. By contrast, in prophages C). These findings indicate the high frequency of temperate ST147-­VIM1phi7.1 and ST13-­OXA48phi12.5, the tRNAs bacteriophages in clinical populations of K. pneumoniae were located between the G/U mismatch-­specific uracil DNA [52, 53], as well as in other clinical pathogens such as Acine- glycosylase and a hypothetical protein [50]. tobacter baumannii [8] and Pseudomonas aeruginosa [9, 54]. Interestingly, analysis of spot tests revealed that the most DISCUSSION frequent temperate bacteriophages (33.3, 36.1, 39.6 and In this study, we identified and annotated the genomes of 42.6 kb) do not produce halos in the highest number of 40 prophages (mainly belonging to the order Caudovirales strains. This is probably due to the superinfection exclusion family Myoviridae) from 16 clinical strains of carbapenemase-­ systems promoted by the prophages in the bacterial genome. producing K. pneumoniae and determined their phylogenetic Thus, the bacteria become immune to subsequent infection relationships. According to the available data about the size by other bacteriophages that are the same or very similar to of the phage genomes, we observed diverse sizes of temperate the integrated prophages [55]. bacteriophages, ranging from 11.445 kb (prophage ST101-­ KPC2phi6.2) to 84.199 kb (prophage ST13-­OXA48phi12.3) The integration of the bacteriophages into the bacterial and corresponding to the size of dsDNA viruses (i.e. 18 to genome is a crucial step in the lysogenic cycle [56]. This event 500 kb) [51]. Interestingly, prophages of sizes 33.3, 36.1, is mediated by the integrase protein, a DNA recombinase 39.6 and 42.6 kb were present in several clinical strains of encoded by bacteriophages, at a specific bacterial genome K. pneumoniae, possibly indicating important roles in these attachment site (attB), which is identical to an attachment isolates. Moreover, these prophages had partial sequence site (attP) of the bacteriophage genome [56]. According to 13 Bleriot et al., Microbial Genomics 2020;6 Fig. 8. Illustration representative of transmission of virulence factor and phage defence genes by prophages in the bacterial chromosome of the K. pneumoniae clinical isolates considered in this study. previous reports, there have been indications that prophages function of 59.7 % of proteins, by using different bioinfor- are not randomly distributed in genomes. Indeed, it has been matics tools. Identification of protein functions is a challenge observed that those prophages that encode tyrosine integrase that must be addressed to improve knowledge of bacterio- are usually integrated next to a host tRNA [57]. One possible phages and the level of security of their applications [36]. It explanation for this finding is the affinity that the temperate should be noted that in other studies a better understanding bacteriophage may have for palindromic structures, enabling of the bacteriophage's protein function allowed bacteriophage integration [57]. These data corroborate our findings, whereby engineering to be undertaken. In fact, knowledge of bacte- 35 % of the prophages under study were integrated close to riophage regulatory proteins has allowed the conversion of host tRNA, specifically tRNA-­ARG, which coincides with the a lysogenic bacteriophage into a lytic one in A. baumannii, findings of Roszniowski et al. for a Burkholderia cenocepacia which presented activity against various clinical strains of prophage sequence [58]. Additionally, the subsequent integra- A. baumannii [64]. In addition, a cocktail of natural lytic tion sites that we identified are also consistent with those most bacteriophages along with engineered bacteriophages has frequently observed by these researchers (ABC transporter recently been successfully used to treat an infection caused by genes and transcriptional regulators) [58]. a drug-­resistant Mycobacterium abscessus strain in a patient with cystic fibrosis [65]. Most of the genes that encode bacteriophages (50–80 %) lack a described function to date and, therefore, are currently depos- We subsequently observed that the next most abundant ited in public databases, such as the NCBI, as hypothetical genes in the prophage genomes were those that constitute proteins [59–61]. Thus, several studies have revealed a high the minimum unit of a Caudovirales bacteriophage, i.e. genes percentage of hypothetical proteins in the genomes of specific related to structure, packaging, lysis, lysogenesis, transcrip- bacteriophages of Salmonella [36], K. pneumoniae [62] and tion, replication and regulation [66, 67]. In the present Escherichia coli [63]. In the present study, we predicted the study, we predicted virulence factors belonging to secretion 14 Bleriot et al., Microbial Genomics 2020;6 Table 5. GenBank accession numbers and protein IDs for the predicted contributes to resistance to oxidative stress, bacterial survival methyltransferase ORFs carried by the prophages of the K. pneumoniae within macrophages and also bacterial survival in infection clinical isolates models. In another example, RovA is a member of the MarR/ Prophage GenBank Protein ID E value SlyA family, which regulates expression of inv (an adhesion accession no. and invasion factor) in the enteric pathogens Yersinia entero- colitica and Yersinia pseudotuberculosis, as well as expression ST405-­OXA48phi1.2 MK416007.1 QBP08186.1 2.00E–164 of the psa locus of Yersinia pestis, the causative agent of ST11-­OXA245phi3.1 MK416009.1 QBP08076.1 0.00E+00 bubonic and pneumonic plague. Members of this transcrip- ST11-­OXA245phi3.2 MK416010.1 QBQ71801.1 3.00E–147 tional regulator family are also thought to be important for the adaptation of K. pneumoniae in the mammalian host [70]. ST437-­OXA245phi4.1 MK416011.1 QBP08055.1 0.00E+00 Interestingly, several mechanisms associated with phage ST437-­OXA245phi4.2 MK416012.1 QBQ71739.1 0.00E+00 defence against other viruses or bacteria have been detected. ST16-­OXA48phi5.1 MK416013.1 QBQ71677.1 2.00E–52 We must highlight the prediction of modules of TA systems in the genome of some prophages from this study. TA systems QBQ71678.1 5.00E–54 are small modules consisting of a stable toxin and its unstable ST16-­OXA48phi5.2 MK448230.1 QBP28277.1 0.00E+00 cognate antitoxin [71]. These systems are involved in abor- QBP28278.1 2.00E–126 tive infection (bacteriophage immunity through altruistic suicide), which not only protects bacteria from bacteriophage ST16-­OXA48phi5.4 MK416015.1 QBP07929.1 0.00E+00 infection in the cultures grown, but also affects the use of ST101-­KPC2phi6.1 MK448231.1 QBP28349.1 0.00E+00 bacteriophage therapy [48]. The use of bioinformatics tools has led to the detection of four TA modules. Two of these QBP28350.1 5.00E–160 modules displayed partial sequence identity with RelBE-­like ST101-­KPC2phi6.3 MK416017.1 QBQ71614.1 0.00E+00 proteins, and the other two displayed partial sequence iden- ST147-­VIM1phi7.2 MK448232.1 QBP28455.1 0.00E+00 tity with HigBA TA proteins [47, 48]. ST11-­VIM1phi8.1 MK448233.1 QBQ71410.1 0.00E+00 We also detected a putative anti-­ CRISPR/Cas9 protein, AcrIIC3. The CRISPR/Cas system is known to play an QBQ71411.1 5.00E–160 important role in protecting bacteria against invasion by ST11-­VIM1phi8.4 MK416020.1 QBP07832.1 0.00E+00 bacteriophages and other mobile genetic elements. Neverthe- less, bacteriophages have evolved different means of evading ST846-­OXA48phi9.2 MK416022.1 QBP07693.1 3.00E–117 CRISPR/Cas defence systems, such as point mutations and ST340-­VIM1phi10.2 MK422454.1 QBP27648.1 0.00E+00 the actions of small proteins that directly interact with the ST13-­OXA48phi12.1 MK422453.1 QBP27632.1 0.00E+00 CRISPR/Cas system and shut it down. The first anti-­CRISPR protein was discovered in 2013, in P. aeruginosa prophages ST13-­OXA48phi12.2 MK422452.1 QBP27559.1 7.00E–160 [72, 73], and since then multiple anti-­CRISPR/Cas proteins ST13-­OXA48phi12.3 MK422451.1 QBP27503.1 2.00E–141 have been discovered in bacteriophages of various bacterial species, e.g. Moraxella bovoculi, Sulfolobus islandicus, Listeria ST13-­OXA48phi12.4 MK422450.1 QBQ71947.1 2.00E–116 monocytogenes and Streptococcus thermophilus [74]. The anti-­ ST13-­OXA48phi12.5 MK714353.1 QEA09480.1 0.00E+00 CRISPR/Cas9 AcrIIC3 was discovered in Neisseria meningitis and acts by inhibiting the Cas9 protein, blocking the DNA QEA09481.1 1.00E–159 loading step through binding to a non-­conserved surface of ST512-­KPC3phi13.1 MK448235.1 QBQ71476.1 0.00E+00 the HNH domain and interacting with the rec lobe of Cas9 QBQ71477.1 5.00E–160 leading to dimerization of the ArcII3-­Cas9 complex [75, 76]. We also detected a CRISPR-­associated endoribonuclease ST512-­KPC3phi13.2 MK422449.1 QBP27409.1 0.00E+00 Cas2 in the genomes of two prophages. Bacteriophages are ST512-­KPC3phi13.6 MK433577.1 QBP27783.1 0.00E+00 thought to use such proteins to evade host immunity in an as yet uncharacterized way [77]. ST15-­OXA48phi14 MK433578.1 QBQ71986.1 0.00E+00 Also related to phage infection immunity, we found the TerB ST258-­KPC3phi16.1 MK433581.1 QBP27938.1 5.00E–179 protein, despite it being involved in resistance to tellurite ST258-­KPC3phi16.2 MK433582.1 QBP27988.1 0.00E+00 when it is part of terZABCDEF operon in bacterial patho- gens such as E. coli and K. pneumoniae [78]. Furthermore, TerB has been found to be involved in resistance to infection systems such as T4SS (type IV secretion system) (involved in by various bacteriophages, such as T1 and T5, a mechanism bacterial competence) [68], viral coat proteins [69] and the known as phage inhibition (Phi), as well as related with resist- MarR family regulator of efflux pumps (which are associated ance to pore-­forming colicins (PacB) [79]. Finally, numerous with virulence factors in many bacteria [46]). For example, predicted methyltransferase enzymes were located in 62.5 % of SlyA regulates Salmonella pathogenicity island-2 genes and the prophages from K. pneumoniae clinical strains, and were 15 Bleriot et al., Microbial Genomics 2020;6 reported to be involved in protecting the viral genome against 3. Wommack KE, Colwell RR. Virioplankton: viruses in aquatic the host restriction enzymes from the bacteria [53, 80]. ecosystems. Microbiol Mol Biol Rev 2000;64:69–114. 4. Taylor VL, Fitzpatrick AD, Islam Z, Maxwell KL. The diverse impacts of phage morons on bacterial fitness and virulence. Adv Virus Res Conclusion 2019;103:1–31. This study characterized 40 prophages (order Caudovirales) in 5. Breitbart M, Rohwer F. Here a virus, there a virus, everywhere the same virus? Trends Microbiol 2005;13:278–284. the genomes of 16 clinical strains of K. pneumoniae belonging to different STs (MLST). The analysis revealed the presence of 6. Brister JR, Le Mercier P, Hu JC. Microbial virus genome annotation — mustering the troops to fight the sequence onslaught. Virology four size groups of prophages (33.3, 36.1, 39.6 and 42.6 kb), 2012;434:175–180. whose sequences were similar to those of the international 7. Pope WH, Bowman CA, Russell DA, Jacobs-­Sera D, Asai DJ et al. host-­bacteriophage clusters registered in the MVP database, Whole genome comparison of a large collection of mycobacte- 4762, 4901, 3499 and 4280, respectively. Moreover, these riophages reveals a continuum of phage genetic diversity. eLife 2015;4:e06416. 33.3, 36.1, 39.6 and 42.6 kb prophages were included in three main clusters (A, B and C) by phylogenetic relation- 8. López M, Rueda A, Florido JP, Blasco L, Fernández-­García L et al. Evolution of the Quorum network and the mobilome (plasmids ships. The annotation of prophages has revealed 40.3 % of and bacteriophages) in clinical strains of Acinetobacter baumannii prophage genomes encoded genes of unknown function. during a decade. Sci Rep 2018;8:2523. However, this annotation has also revealed that the second 9. Kwan T, Liu J, DuBow M, Gros P, Pelletier J. Comparative genomic most important group of genes constituted the minimum unit analysis of 18 Pseudomonas aeruginosa bacteriophages. J Bacteriol 2006;188:1184–1187. of Caudovirales bacteriophages. Interestingly, we observed virulence factors (compounds of secretion systems or regu- 10. Grose JH, Jensen GL, Burnett SH, Breakwell DP. Genomic compar- ison of 93 Bacillus phages reveals 12 clusters, 14 singletons and lators) and components of phage defence against bacteria remarkable diversity. BMC Genomics 2014;15:855. and also against other bacteriophages (TA modules, anti-­ 11. Deveau H, Labrie SJ, Chopin M-­C, Moineau S. Biodiversity and CRISPR/Cas9, TerB protein and methyltransferase proteins) classification of lactococcal phages. Appl Environ Microbiol in prophage genomes. Future lines of research should focus on 2006;72:4338–4346. obtaining more information about genes of unknown func- 12. Gregory AC, Solonenko SA, Ignacio-­Espinoza JC, LaButti K, Cope- tion to provide a better understanding of phage genomes for land A et al. Genomic differentiation among wild cyanophages despite widespread horizontal gene transfer. BMC Genomics possible therapeutic applications of bacteriophages, such as 2016;17:930. phage-­derived protein, as well as for engineering phages with 13. Switt AI, Sulakvelidze A, Wiedmann M, Kropinski AM, Wishart DS activity against MDR pathogens. et al. Salmonella phages and prophages: genomics, taxonomy, and applied aspects. Methods Mol Biol 2015;1225:237–287. 14. Kalatzis PG, Rørbo NI, Castillo D, Mauritzen JJ, Jørgensen J et al. Funding information Stumbling across the same phage: comparative genomics of wide- This study was funded by grant PI16/01163 awarded to M.T. within spread temperate phages infecting the fish pathogen Vibrio anguil- the State Plan for R+D+I 2013–2016 (National Plan for Scientific larum. Viruses 2017;9:122. Research, Technological Development and Innovation 2008–2011), 15. Grose JH, Casjens SR. Understanding the enormous diversity of and co-­financed by the ISCIII-­Deputy General Directorate for Evalu- bacteriophages: the tailed phages that infect the bacterial family ation and Promotion of Research – European Regional Development Enterobacteriaceae. Virology 2014;468-470:421–443. Fund ‘A Way of Making Europe’ and Instituto de Salud Carlos III FEDER, 16. Hendrix RW, Smith MCM, Burns RN, Ford ME, Hatfull GF. Evolutionary Spanish Network for the Research in Infectious Diseases (REIPI) relationships among diverse bacteriophages and prophages: all (RD16/0016/0001, RD16/0016/0006 and RD16/0016/0008) and by the the world's a phage. Proc Natl Acad Sci USA 1999;96:2192–2197. Study Group on Mechanisms of Action and Resistance to Antimicrobials (GEMARA) (Spanish Society of Infectious Diseases and Clinical Microbi- 17. Dion MB, Oechslin F, Moineau S. Phage diversity, genomics and ology – SEIMC; http://www.​seimc.​org/). R.T. and L.F.-G. were financially phylogeny. Nat Rev Microbiol 2020;18:125–138. supported by grants from the SEIMC and the Deputacion Provincial da 18. Hatfull GF. Dark matter of the biosphere: the amazing world of Coruña (Xunta de Galicia), respectively. bacteriophage diversity. J Virol 2015;89:8107–8110. 19. de Jonge PA, Nobrega FL, Brouns SJJ, Dutilh BE. Molecular and Conflicts of interest evolutionary determinants of bacteriophage host range. Trends The authors declare that there are no conflicts of interest. Microbiol 2019;27:51–63. 20. Argov T, Azulay G, Pasechnek A, Stadnyuk O, Ran-­Sapir S et al. Data Bibliography Temperate bacteriophages as regulators of host behavior. Curr 1. Forty-­two bacteriophages isolated from 17 clinical strains of Kleb- Opin Microbiol 2017;38:81–87. siella: genome sequencing and assembly (REIPI – Spanish Network for Research in Infectious Diseases). Bleriot I, Trastoy R, Blasco L, 21. Pratama AA, van Elsas JD. Gene mobility in microbiomes of the Fernández-­Cuenca F, Ambroa A, Fernández-­García L, Pacios O, Perez-­ mycosphere and mycorrhizosphere – role of plasmids and bacte- Nadales E, Torre-­Cisneros J, Oteo-­Iglesias J, Navarro F, Miró E, Pascual riophages. FEMS Microbiol Ecol 2019;95:fiz053. A, Martínez-­ Martínez L, Tomás M. BioProject accession number 22. Wiedenbeck J, Cohan FM. Origins of bacterial diversity through PRJNA565865 (https://www.​ncbi.​nlm.​nih.​gov/​bioproject/?​term=​ horizontal genetic transfer and adaptation to new ecological PRJNA565865) (2019). niches. FEMS Microbiol Rev 2011;35:957–976. 23. Brüssow H, Canchaya C, Hardt W-­D. Phages and the evolution of References bacterial pathogens: from genomic rearrangements to lysogenic 1. Shabbir MAB, Hao H, Shabbir MZ, Wu Q, Sattar A et al. Bacteria conversion. Microbiol Mol Biol Rev 2004;68:560–602. vs. bacteriophages: parallel evolution of immune arsenals. Front 24. Gorrie CL, Mirceta M, Wick RR, Judd LM, Wyres KL et al. Microbiol 2016;7:1292. Antimicrobial-­resistant Klebsiella pneumoniae carriage and infec- 2. Bergh Ø, Børsheim KY, Bratbak G, Heldal M. High abundance of tion in specialized geriatric care wards linked to acquisition in the viruses found in aquatic environments. Nature 1989;340:467–468. referring hospital. Clin Infect Dis 2018;67:161–170. 16 Bleriot et al., Microbial Genomics 2020;6 25. Pitout JDD, Peirano G, Kock MM, Strydom K-­A, Matsumura Y. The 48. Harms A, Brodersen DE, Mitarai N, Gerdes K. Toxins, targets, global ascendency of OXA-48-­type carbapenemases. Clin Microbiol and triggers: an overview of toxin-­ antitoxin biology. Mol Cell Rev 2019;33:e00102-19. 2018;70:768–784. 26. Pendleton JN, Gorman SP, Gilmore BF. Clinical relevance of the 49. Chiang S-­K, Lou Y-­C, Chen C. NMR solution structure of KP-­TerB, a ESKAPE pathogens. Expert Rev Anti Infect Ther 2013;11:297–308. tellurite-­resistance protein from Klebsiella pneumoniae. Protein Sci 27. Vuotto C, Longo F, Balice M, Donelli G, Varaldo P. Antibiotic resist- 2008;17:785–789. ance related to biofilm formation in Klebsiella pneumoniae. Patho- 50. Bailly-­Bechet M, Vergassola M, Rocha E. Causes for the intriguing gens 2014;3:743–758. presence of tRNAs in phages. Genome Res 2007;17:1486–1495. 28. Bengoechea JA, Sa Pessoa J. Klebsiella pneumoniae infection 51. Casjens SR. Comparative genomics and evolution of the tailed-­ biology: living to counteract host defences. FEMS Microbiol Rev bacteriophages. Curr Opin Microbiol 2005;8:451–458. 2019;43:123–144. 52. Adriaenssens E, Brister JR. How to name and classify your phage: 29. Duprilot M, Decre D, Genel N, Drieux L, Sougakoff W et al. Diver- an informal guide. Viruses 2017;9:70. sity and functionality of plasmid-­borne VagCD toxin-­ antitoxin 53. Maciejewska B, Roszniowski B, Espaillat A, Kęsik-­ systems of Klebsiella pneumoniae. J Antimicrob Chemother Szeloch A, Majkowska-­Skrobek G et al. Klebsiella phages 2017;72:1320–1326. representing a novel clade of viruses with an unknown DNA modi- 30. Esteban-­Cantos A, Aracil B, Bautista V, Ortega A, Lara N et al. The fication and biotechnologically interesting enzymes. Appl Microbiol carbapenemase-­producing Klebsiella pneumoniae population is Biotechnol 2017;101:673–684. distinct and more clonal than the carbapenem-­susceptible popu- 54. López M, Rueda A, Florido JP, Blasco L, Gato E et al. Genomic lation. Antimicrob Agents Chemother 2017;61:e02520-16. evolution of two Acinetobacter baumannii clinical strains from 31. Alikhan N-­F, Petty NK, Ben Zakour NL, Beatson SA. BLAST Ring ST-2 clones isolated in 2000 and 2010 (ST-2_clon_2000 and ST-2_ Image Generator (BRIG): simple prokaryote genome comparisons. clon_2010). Genome Announc 2016;4:e01182-16. BMC Genomics 2011;12:402. 55. Folimonova SY. Superinfection exclusion is an active virus-­ 32. Gao NL, Zhang C, Zhang Z, Hu S, Lercher MJ et al. MVP: a microbe– controlled function that requires a specific viral protein. J Virol phage interaction database. Nucleic Acids Res 2018;46:D700–D707. 2012;86:5554–5561. 33. Krumsiek J, Arnold R, Rattei T. Gepard: a rapid and sensi- 56. Feiner R, Argov T, Rabinovich L, Sigal N, Borovok I et al. A new tive tool for creating dotplots on genome scale. Bioinformatics perspective on lysogeny: prophages as active regulatory switches 2007;23:1026–1028. of bacteria. Nat Rev Microbiol 2015;13:641–650. 34. Black LW. Old, new, and widely true: the bacteriophage T4 DNA 57. Bobay L-­M, Rocha EPC, Touchon M. The adaptation of packaging mechanism. Virology 2015;479-480:650–656. temperate bacteriophages to their host genomes. Mol Biol Evol 2013;30:737–751. 35. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol 58. Roszniowski B, McClean S, Drulis-­Kawa Z. Burkholderia cenoce- Biol Evol 2018;35:1547–1549. pacia prophages - prevalence, chromosome location and major genes involved. Viruses 2018;10:297. 36. Bardina C, Colom J, Spricigo DA, Otero J, Sánchez-­Osuna M et al. Genomics of three new bacteriophages useful in the biocontrol of 59. Hatfull GF, Hendrix RW. Bacteriophages and their genomes. Curr Salmonella. Front Microbiol 2016;7:545. Opin Virol 2011;1:298–303. 37. Zabarovskiĭ ER, Turina OV. Rapid isolation of phage lambda DNA. 60. Van den Bossche A, Ceyssens P-­J, De Smet J, Hendrix H, Mol Biol 1988;22:1451–1455. Bellon H et al. Systematic identification of hypothetical bacterio- phage proteins targeting key protein complexes of Pseudomonas 38. Raya RR, H'bert EM. Isolation of phage via induction of lysogens. aeruginosa. J Proteome Res 2014;13:4446–4456. Methods Mol Biol 2009;501:23–32. 61. Klumpp J, Fouts DE, Sozhamannan S. Bacteriophage functional 39. Zimmermann L, Stephens A, Nam S-­Z, Rau D, Kübler J et al. A genomics and its role in bacterial pathogen detection. Brief Funct completely reimplemented MPI bioinformatics toolkit with a new Genomics 2013;12:354–365. HHpred server at its core. J Mol Biol 2018;430:2237–2243. 62. Tabassum R, Shafique M, Khawaja KA, Alvi IA, Rehman Y et al. 40. Hildebrand A, Remmert M, Biegert A, Söding J. Fast and accu- Complete genome analysis of a Siphoviridae phage TSK1 showing rate automatic structure prediction with HHpred. Proteins biofilm removal potential against Klebsiella pneumoniae. Sci Rep 2009;77:128–132. 2018;8:17904. 41. Meier-­Kolthoff JP, Göker M. VICTOR: genome-­ based phylogeny 63. Xu Y, Yu X, Gu Y, Huang X, Liu G et al. Characterization and genomic and classification of prokaryotic viruses. Bioinformatics study of phage vB_EcoS-­B2 infecting multidrug-­resistant Escherichia 2017;33:3396–3404. coli. Front Microbiol 2018;9:793. 42. Cahill J, Young R. Phage lysis: multiple genes for multiple barriers. 64. Blasco L, Ambroa A, Lopez M, Fernandez-­Garcia L, Bleriot I et al. Adv Virus Res 2019;103:33–70. Combined use of the Ab105-2φΔCI lytic mutant phage and different 43. Rivera-­Calzada A, Fronzes R, Savva CG, Chandran V, Lian PW et al. antibiotics in clinical isolates of multi-­ resistant Acinetobacter Structure of a bacterial type IV secretion core complex at subnano- baumannii. Microorganisms 2019;7:556. metre resolution. EMBO J 2013;32:1195–1204. 65. Dedrick RM, Guerrero-­Bustamante CA, Garlena RA, Russell DA, 44. Nguyen KB, Sreelatha A, Durrant ES, Lopez-­Garrido J, Musze- Ford K et al. Engineered bacteriophages for treatment of a patient wska A et al. Phosphorylation of spore coat proteins by a with a disseminated drug-­resistant Mycobacterium abscessus. Nat family of atypical protein kinases. Proc Natl Acad Sci USA Med 2019;25:730–733. 2016;113:E3482–E3491. 66. Fokine A, Rossmann MG. Molecular architecture of tailed double-­ 45. Kim DJ, Park K-­S, Kim J-­H, Yang S-­H, Yoon JY et al. Helico- stranded DNA phages. Bacteriophage 2014;4:e28281. bacter pylori proinflammatory protein up-­ regulates NF-­ kappaB 67. Feiss M, Rao VB. The bacteriophage DNA packaging machine. Adv as a cell-­translocating Ser/Thr kinase. Proc Natl Acad Sci USA Exp Med Biol 2012;726:489–509. 2010;107:21418–21423. 68. Boyd EF, Carpenter MR, Chowdhury N. Mobile effector proteins on 46. Deochand DK, Grove A. MarR family transcription factors: dynamic phage genomes. Bacteriophage 2012;2:139–148. variations on a common scaffold. Crit Rev Biochem Mol Biol 69. Liekniņa I, Kalniņš G, Akopjana I, Bogans J, Šišovs M et al. Produc- 2017;52:595–613. tion and characterization of novel ssRNA bacteriophage virus-­like 47. Yao J, Guo Y, Wang P, Zeng Z, Li B et al. Type II toxin/antitoxin particles from metagenomic sequencing data. J Nanobiotechnol system ParE. Environ Microbiol 2018;20:1224–1239. 2019;17:61. 17 Bleriot et al., Microbial Genomics 2020;6 70. Palacios M, Miner TA, Frederick DR, Sepulveda VE, Quinn JD et al. 76. Borges AL, Davidson AR, Bondy-­Denomy J. The discovery, mecha- Identification of two regulators of virulence that are conserved in nisms, and evolutionary impact of anti-­CRISPRs. Annu Rev Virol Klebsiella pneumoniae classical and hypervirulent strains. mBio 2017;4:37–59. 2018;9:e01443-18. 77. Hudaiberdiev S, Shmakov S, Wolf YI, Terns MP, Makarova KS 71. Yang QE, Walsh TR. Toxin–antitoxin systems and their role in et al. Phylogenomics of Cas4 family nucleases. BMC Evol Biol disseminating and maintaining antimicrobial resistance. FEMS 2017;17:232. Microbiol Rev 2017;41:343–353. 78. Chen Y-­T, Chang H-­Y, Lai Y-­C, Pan C-­C, Tsai S-­F et al. Sequencing 72. Bondy-­Denomy J, Pawluk A, Maxwell KL, Davidson AR. Bacterio- and analysis of the large virulence plasmid pLVPK of Klebsiella phage genes that inactivate the CRISPR/Cas bacterial immune pneumoniae CG43. Gene 2004;337:189–198. system. Nature 2013;493:429–432. 79. Whelan KF, Colleran E, Taylor DE. Phage inhibition, colicin 73. Maxwell KL. The anti-­CRISPR story: a battle for survival. Mol Cell resistance, and tellurite resistance are encoded by a single 2017;68:8–14. cluster of genes on the IncHI2 plasmid R478. J Bacteriol 74. Hwang S, Maxwell KL. Meet the anti-­CRISPRs: widespread protein 1995;177:5016–5027. inhibitors of CRISPR-­Cas systems. CRISPR J 2019;2:23–30. 80. Murphy J, Mahony J, Ainsworth S, Nauta A, van Sinderen D. Bacte- 75. Zhu Y, Gao A, Zhan Q, Wang Y, Feng H et al. Diverse mechanisms riophage orphan DNA methyltransferases: insights from their of CRISPR-­Cas9 inhibition by type IIC anti-­CRISPR proteins. Mol Cell bacterial origin, function, and occurrence. Appl Environ Microbiol 2019;74:296–309. 2013;79:7547–7555. Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review – average time to first decision is 4–6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 80% of our authors rate our submission process as ‘excellent’ or ‘very good’. 5. Your article will be published on an interactive journal platform with advanced metrics. Find out more and submit your article at microbiologyresearch.org. 18