TYPE Original Research PUBLISHED 19 August 2022 DOI 10.3389/fmicb.2022.918015 The transcriptional regulator OPEN ACCESS CtrA controls gene expression in Alphaproteobacteria phages: EDITED BY Ellie Harrison, The University of Sheffield, United Kingdom REVIEWED BY Alexander N. Ignatov, Evidence for a lytic deferment pathway Peoples’ Friendship University of Russia, Russia Catriona Thompson, John Innes Centre, United Kingdom Elia Mascolo 1†, Satish Adhikari 2†, Steven M. Caruso 1, *CORRESPONDENCE Ivan Erill Tagide deCarvalho 3, Anna Folch Salvador 4, erill@umbc.edu Joan Serra-Sagristà 4, Ry Young 5, Ivan Erill 1‡* and Patrick D. Curtis pdcurtis@olemiss.edu Patrick D. Curtis 2‡* † These authors have contributed equally to 1 Department of Biological Sciences, University of Maryland Baltimore County, Baltimore, MD, this work and share first authorship United States, 2 Department of Biology, University of Mississippi, Oxford, MS, United States, 3 Keith R. Porter Imaging Facility, College of Natural and Mathematical Sciences, University of Maryland ‡ These authors have contributed equally to Baltimore County (UMBC), Baltimore, MD, United States, 4 Universitat Autònoma de Barcelona, this work and share last authorship Cerdanyola del Vallès, Barcelona, Spain, 5 Center for Phage Technology, Texas A&M University, SPECIALTY SECTION College Station, TX, United States This article was submitted to Phage Biology, a section of the journal Pilitropic and flagellotropic phages adsorb to bacterial pili and flagella. These Frontiers in Microbiology phages have long been used to investigate multiple aspects of bacterial RECEIVED 11April 2022 physiology, such as the cell cycle control in the Caulobacterales. Targeting ACCEPTED 18 July 2022 PUBLISHED 19 August 2022 cellular appendages for adsorption effectively constrains the population of CITATION infectable hosts, suggesting that phages may have developed strategies to Mascolo E, Adhikari S, Caruso SM, maximize their infective yield. Brevundimonas phage vB_BsubS-Delta is a deCarvalho T, Folch Salvador A, recently characterized pilitropic phage infecting the Alphaproteobacterium Serra-Sagristà J, Young R, Erill I and Curtis PD (2022) The transcriptional Brevundimonas subvibrioides. Like other Caulobacterales, B. subvibrioides regulator CtrA controls gene expression in divides asymmetrically and its cell cycle is governed by multiple transcriptional Alphaproteobacteria phages: Evidence for a lytic deferment pathway. regulators, including the master regulator CtrA. Genomic characterization of Front. Microbiol. 13:918015. phage vB_BsubS-Delta identified the presence of a large intergenic region with doi: 10.3389/fmicb.2022.918015 an unusually high density of putative CtrA-binding sites. A systematic analysis COPYRIGHT of the positional distribution of predicted CtrA-binding sites in complete phage © 2022 Mascolo, Adhikari, Caruso, deCarvalho, Folch Salvador, Serra-Sagristà, genomes reveals that the highly skewed distribution of CtrA-binding sites Young, Erill and Curtis. This is an observed in vB_BsubS-Delta is an unequivocal genomic signature that extends open-access article distributed under the to other pilli- and flagellotropic phages infecting the Alphaproteobacteria. terms of the Creative Commons Attribution License (CC BY). The use, distribution or Moreover, putative CtrA-binding sites in these phage genomes localize reproduction in other forums is permitted, preferentially to promoter regions and have higher scores than those detected provided the original author(s) and the copyright owner(s) are credited and that in other phage genomes. Phylogenetic and comparative genomics analyses the original publication in this journal is show that this genomic signature has evolved independently in several phage cited, in accordance with accepted lineages, suggesting that it provides an adaptive advantage to pili/flagellotropic academic practice. No use, distribution or reproduction is permitted which does not phages infecting the Alphaproteobacteria. Experimental results demonstrate comply with these terms. that CtrA binds to predicted CtrA-binding sites in promoter regions and that it regulates transcription of phage genes in unrelated Alphaproteobacteria- infecting phages. We propose that this focused distribution of CtrA-binding sites reflects a fundamental new aspect of phage infection, which we term lytic deferment. Under this novel paradigm, pili- and flagellotropic phages Frontiers in Microbiology 01 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 exploit the CtrA transduction pathway to monitor the host cell cycle state and synchronize lysis with the presence of infectable cells. KEYWORDS bacteriophage, lysis, pseudolysogeny, cell cycle, regulation, transcription, CtrA, Caulobacter Introduction MacKrell, 2011; Greene et al., 2012; Zan et al., 2013, 1; Wang et al., 2014; Francez-Charlot et al., 2015; Hernández-Valle Following infection, virulent phages replicate and lyse the cell et al., 2020). to release virions. Temperate phages, on the other hand, have the Phages use multiple strategies to exert regulatory control additional option of lysogeny, in which phage gene expression is on the transcription of many bacterial genes (Yang et al., repressed and a prophage is established via integration into the 2014). In contrast, host transcriptional regulation of phage bacterial chromosome or as a plasmid (Howard-Varona et al., genes has not been so amply documented and is typically 2017). The phage infection process starts with the adsorption of associated with phage adaptation to cellular stress. A canonical phage particles to the bacterial cell. Adsorption is mediated example of this phenomenon is the activation of the lytic cycle through interactions between phage binding proteins and in temperate phages in response to DNA damage. Phage bacterial cell surface receptors, which are typically components of transcriptional regulators, epitomized by the Enterobacteria the bacterial cell wall (Bertozzi Silva et al., 2016). Some phages, phage lambda CI repressor, regulate the activity of lytic phase however, initiate adsorption through interactions with bacterial genes by competing for promoter occupancy with Cro-type pili or flagella (Bradley, 1972; Lagenaur et al., 1977; Frost, 1993). repressors (Ptashne, 1992). The presence of DNA damage, as Pili-and flagellotropic phages include several major bacteriophage a result of different stresses, results in the autocatalytic families and pilitropic phages are known to target both plasmid- cleavage of CI-like repressors, leading to the onset of the and chromosomally encoded pili, suggesting that pilus-mediated phage lytic cycle (Sauer et al., 1982). Recent work has shown adsorption has evolved independently in several phage lineages that phages can incorporate additional layers of logic to this (Ceyssens and Lavigne, 2010). decision making, using quorum sensing-like mechanisms to Pilitropic phages played an important role in the investigation reduce prophage induction at high phage concentrations of the cell cycle of the Alphaproteobacterium C. crescentus. This (Brady et al., 2021). bacterium has a dimorphic lifestyle, producing two different cell Several phages have independently evolved the ability to types after division: a surface-attached sessile “stalked” cell and a exploit a bacterial transcriptional regulator, the LexA repressor, to motile “swarmer” cell with a polar flagellum surrounded by polar trigger their lytic cycle (Fornelos et al., 2016). Like CI, LexA pili. Bacteriophage ΦCbk and related pilitropic phages target the undergoes RecA-mediated autocatalytic cleavage in the presence polar flagellum and pili of swarmer cells, and ΦCbk was of DNA damage, de-repressing its target genes and inducing the instrumental in unraveling the genetic regulation of pili bacterial SOS response (Erill et al., 2007). Induction of the lytic production (Skerker and Shapiro, 2000). cycle in these phages is mediated by the presence of LexA-binding Cell-cycle control in C. crescentus is orchestrated by a sites in the promoter region of antirepressor, co-repressor or lytic complex network of transcriptional, epigenetic and proteolytic phase genes in the phage chromosome (Fornelos et al., 2011). interactions involving several global regulators; including CtrA, Co-option of host transcription factors to regulate phage gene DnaA, GcrA, CcrM, CckA and DivK (Brilli et al., 2010). Among expression is not restricted to activation of the phage lytic cycle by these, CtrA, CcrM, DnaA and GcrA are widely conserved across LexA. The virulent Bacillus phage Φ29 contains multiple the Alphaproteobacteria that have not undergone massive promoter-associated binding sites for the master regulator of genomic reduction. CtrA is the master regulator of development sporulation initiation Spo0A. Binding of Spo0A to these sites in C. crescentus, having critical roles in the regulation of cell suspends lytic cycle development, enabling spore entrapment of division, chromosomal replication, pilus assembly and flagellar the Φ29 chromosome by the chromosome segregation protein biosynthesis (Laub et al., 2002). CtrA regulates transcription of Spo0J (Meijer et al., 2005). The phage can thus become dormant target genes by binding a conserved palindromic motif with during nutrient starvation and resume its lytic cycle following consensus sequence TTAA-N7-TTAA (Ouimet and Marczynski, spore germination. These two cases illustrate the ability of phages 2000). The conservation of this transcriptional regulator and its to exploit host-encoded signal transduction systems to control motif in multiple Alphaproteobacteria with simpler phage gene expression and modulate its lifestyle in response to morphogenetic life cycles suggested that its primary function cellular states. might not be cell-cycle regulation (Brilli et al., 2010). Subsequent Here we report on the genomic characterization of a novel work has shown that the ancestral regulatory role of CtrA is phage infecting the Alphaproteobacterium Brevundimonas likely the control of motility and chemotaxis (Bird and subvibrioides that harbors a large intergenic region encompassing Frontiers in Microbiology 02 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 multiple putative CtrA-binding sites. Comparative genomics computed using the ViPhy pipeline to identify homologous analyses reveal that most pilitropic phages infecting sequence segments, using a limiting BLAST e-value threshold of Alphaproteobacteria species contain genomic regions enriched in 10, and normalizing identity counts on homologous segments to putative CtrA-binding sites, which are preferentially located in the overall sequence length. promoter regions. Our results show that CtrA binds to these predicted sites to regulate the expression of genes in several unrelated phages that infect different Alphaproteobacteria genera, Identification of putative transcription suggesting that CtrA regulation of phage genes has evolved factor-binding sites independently and hence likely conveys a selective advantage to pilitropic phages. A collection of experimentally validated CtrA-binding sites was downloaded from the CollecTF database (CollecTF, RRID:SCR_014405; Kılıç et al., 2014). Binding sites for GcrA, SciP Materials and methods and DnaA were obtained from previously published work (Hottes et al., 2005; Tan et al., 2010; Gonzalez et al., 2014; Haakonsen et al., Genome data collection 2015). Collections of TF-binding sites were then used to generate position-specific scoring matrix (PSSM) models to scan complete Complete genome sequences for all bacteriophages bacteriophage genomes with a custom Python script,1 as described analyzed in this work were retrieved from the NCBI GenBank previously (Hobbs et al., 2016). Only positions with a score above (GenBank, RRID:SCR_002760) and RefSeq (RefSeq, the Patser threshold, satisfying the equality between the negative RRID:SCR_003496) databases (Benson et al., 2017; Haft et al., logarithm of the false positive rate and the information content of 2018), prioritizing RefSeq records for a given bacteriophage the TF-binding motif, were considered as putative TF-binding whenever both were available. When annotated, the 5′ instance sites (Hertz and Stormo, 1999). of Direct Terminal Repeat (DTR) regions was removed from the resulting genome objects to avoid double counting. Bacteriophage genomes potentially related to Brevundimonas Computation of normalized positional phage vB_BsubS-Delta were identified by serially querying the entropy and Gini coefficient NCBI GenBank database with tBLASTN (NCBI BLAST, RRID:SCR_004870), restricted to taxonomy ID 10239 To analyze the asymmetry in TF-binding site distribution (Viruses), using each of the annotated vB_BsubS-Delta along the genome, bacteriophage genomes were divided into 50 proteins as the query sequence (Altschul et al., 1997). Only equal-sized bins and the number of putative TF-binding sites tBLASTN hits with e-value lower than 10−10 and query identified in each bin tallied and normalized to obtain site coverage larger than 0.75 were considered, and the nucleotide frequencies per bin. The normalized entropy of the distribution sequence of the genome encoding the putative homologous S of TF-binding sites along the genome was computed protein was retrieved from NCBI GenBank/RefSeq. Identifiers as follows: for the genome sequences of bacteriophages infecting Alphaproteobacteria hosts were obtained from INPHARED  H ( S) H ( S) = (Jan 2021; Cook et al., 2021). Identifiers for the genome max n ( H ( S) ) sequences of pilitropic bacteriophages were obtained from a b (1) targeted search on the NCBI PubMed service (PubMed, H ( S) = − ∑f ( si ) · log 2 ( f ( si ) ) i =1 RRID:SCR_004846), followed by manual curation of reference articles. Genome sequences were subsequently downloaded from the NCBI GenBank/RefSeq databases. Only sequences at where f(si) is the relative frequency of putative TF-binding least 10,000 bp-long were used for analysis. sites in bin i, b is the number of bins, maxn(H(S)) is the largest possible entropy, assuming the most even distribution for the number of sites observed in the genome, and Ĥ(S) is the resulting Genomic sequence analyses normalized entropy. This normalization addresses small number effects arising from low TF-binding site counts, which impact the Phage lifestyle (temperate vs. virulent) was predicted using the maximum entropy attainable. The Gini coefficient (O’Neill and BACterioPHage LIfestyle Predictor (BACPHLIP) classifier Erill, 2016) of the distribution S of TF-binding sites along the (Hockenberry and Wilke, 2021) with default parameters. Multiple genome was computed as: sequence alignment of DTR regions was performed using progressive Mauve (Darling et al., 2004), following the extraction of DTRs from genome sequences using their GenBank annotation. Pairwise nucleotide sequence identity values for DTRs were 1 https://github.com/ErillLab/LPEG_phages Frontiers in Microbiology 03 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015  G ( S) − min n ( G ( S) ) asymmetry maps and distribution violin plots were generated with G ( S) = max n ( G ( S) ) − min n ( G ( S) ) Python scripts.3 Comparative genome plots were generated using b b (2) the BLAST Ring Image Generator (BRIG, RRID:SCR_007802; ∑ =i 1∑ s − si =j 1 j Alikhan et al., 2011). Genome maps were generated with G ( S) = 2·b 2 si DNAmaster (Pope and Jacobs-Sera, 2018). where si is the number of sites in bin i, b is the number of bins Electron microscopy and max(G(S)) and min(G(S)) are, respectively, the maximum and minimum Gini coefficients attainable given the number of Phage lysate (10 μl) was deposited onto a 200 mesh formvar TF-binding sites detected in the genome. To avoid binning and carbon-coated copper grids (EMS). The lysate was allowed to artifacts, bins were defined as starting both at genome position 1 set for 1 min, briefly rinsed with ultra-pure water and stained with and at 1 + L/2, where L is the length of the bin for a given genome, 2% uranyl acetate for 2 min. Imaging was performed with a with rollover. The maximum (entropy) or minimum (Gini) value Morgagni 268 Transmission Electron Microscope (FEI, Hillsboro, obtained for the genome was then reported. Only phage genomes IL, United States) equipped with an Orius CCD camera (Gatan with more than 5 putative TF-binding sites were considered Inc., Pleasanton, CA, United States). Head and tail measurements for analysis. were performed with the Fiji software package (Fiji, RRID:SCR_002285; Schindelin et al., 2012). Phylogenetic analysis Extraction of phage genomic DNA and For phylogenetic analysis, bacteriophage genomes potentially nuclease assays related to bacteriophage vB_BsubS-Delta were complemented with the genome sequences of known Alphaproteobacteria Phage genomic DNA was purified following (Summer, 2009), bacteriophages and reported pilitropic and flagellotropic using the Wizard DNA Clean-up Kit (Promega, bacteriophages. A genome-based phylogeny was generated with RRID:SCR_006724). For general nuclease digestion, RNase-free the ViPhy Python pipeline.2 Agglomerative clustering was DNase I and DNase-free RNase A were obtained from Thermo performed with scikit-learn (scikit-learn, RRID:SCR_002577) on Scientific. For each digestion, 1 μg phage genomic DNA was used the resulting tree using 50 target clusters (Pedregosa et al., 2011). in combination with the DNase I buffer provided by the Representative phage genomes for each cluster were selected and manufacturer. In the DNase reaction 1 U of enzyme was used, complemented with genomes of phages presenting skewed CtrA- while the RNase reaction used 10 μg of that enzyme. Digestions binding site distribution. The final set of 100 phage genomes was were incubated at 37°C for 2 h, followed by gel electrophoresis. used to generate a genome-based phylogeny with the VICTOR Double restriction digestion of phage genomic DNA with enzymes web service (Meier-Kolthoff and Göker, 2017). Inter-genomic AflII and PciI was carried out according to manufacturer’s protein sequence distances were computed with 100 pseudo- protocols (New England Biolabs, RRID:SCR_013517), followed bootstrap replicates using the Genome-BLAST Distance by gel electrophoresis. Phylogeny (GBDP) method optimized (distance formula d6) for prokaryotic viruses (Meier-Kolthoff et al., 2013), and a minimum evolution tree was computed with FASTME on the resulting inter- Growth conditions and strains genomic distances (Lefort et al., 2015). The resulting phylogenetic tree was annotated using different types of information (lifestyle, All strains, plasmids and primers used in this study are listed asymmetric distribution of CtrA-binding sites, known pili/ in Supplementary Table S1. All C. crescentus strains were grown flagellotropic nature) using the iTOL web service (iTOL, in PYE medium or plates (2 g/l peptone, 1 g/l yeast extract, 0.3 g/l RRID:SCR_018174; Letunic and Bork, 2019). MgSO4, 7H2O, 0.0735 g/l CaCl2, 2H2O, 1.5% agar for plates) at 30°C or 37°C. Antibiotics were added when necessary. Tetracycline was added at 1 μg/ml. Escherichia coli strains were Statistics and data visualization grown in LB media (10 g/l tryptone, 5 g/l yeast extract, 10 g/l NaCl, 1.5% agar for plates) at 37°C. Tetracycline was added at 12 μg/l Differences in the distributions of putative TF-binding site when required. scores and intergenic localization for selected datasets were analyzed using a one-sided Mann–Whitney U test. Positional 2 https://github.com/ErillLab/ViPhy 3 https://github.com/ErillLab/pilitropic_phages Frontiers in Microbiology 04 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 CtrA purification and electrophoretic Strain construction and mobility shift assay beta-galactosidase assays For CtrA purification, the coding region of ctrA was amplified The promoter region of Brevundimonas phage vB_BsubS- from C. crescentus using primers ctrApET28aF and Delta DELTA_120 gene was amplified using primers ctrApET28aR. The amplified region was cloned into pET28a delta120betagalF and delta120betagalR, and cloned into the plasmid using Gibson Assembly to create a 6X his tag on the N plac290 plasmid (Gober and Shapiro, 1992) using Gibson terminus of CtrA (pAR0001). This plasmid was introduced into Assembly to create plasmid pSA1002. The promoter region of E. coli BL21 (DE3) strain by electroporation. CtrA purification was genes CbK_gp005 and CbK_gp018 from Caulobacter phage performed by following the QIAexpressionist manual (QIAGEN, ΦCbk was amplified using primer-pairs phicbkgp5promoterF RRID:SCR_008539) under native conditions. DNA probes based and phicbkgp5promoterR, and cbkgp18plac290F and on phage sequences were synthesized (Supplementary Table S1), cbkgp18plac290R, respectively, and cloned into plac290 with the forward strand receiving 5′ biotinylation. Forward and plasmid using Gibson Assembly to create plasmids pSA1003 reverse strands were mixed, heated and gently cooled to create and pSA1004. The promoter region of Cp1R7AA1_019 from double-stranded probes. For DELTA_120, a probe was designed Mesorhizobium phage Cp1R7A-A1 was amplified using that lacked biotinylation. For other phage DNA probes, probes primers Mesopromoter019F and Mesopromoter019R and were designed where the CtrA binding sites were mutated such cloned into plac290 plasmid using Gibson Assembly to create that the TTAA-N7-TTAA CtrA-binding sites were replaced with plasmid pSA1005. Plasmids were introduced into C. crescentus GGCC-N7-GGCC. NA1000 (WT) and LS2195 (ctrA temperature sensitive) strains To perform the EMSA, 7.5 μl CtrA at a final concentration of using electroporation. 3.3 μm was phosphorylated by incubation with 1 μl ATP (1 mm For beta-galactosidase assays, cells were grown up to mid log final concentration) and 1 μl of 7.5 μm MBP-EnvZ (final stage. Z buffer (60 mm Na2HPO4.7H2O, 40 mm NaH2PO4.H2O, concentration 0.75 μm) in a buffer consisting of 50 mm Tris HCl pH 7.0, 10 mm KCl, 1 mm MgSO4.7H2O, 50 mm (pH 7.8), 50 mm KCl, 20 mm MgCl2, 1 mm DTT, at 30°C for β-mercaptoethanol) and 50 μl of chloroform was added to 200 μl 20 min. This is a modified protocol of (Reisenauer et al., 1999). of cells to a final volume of 850 μl and incubated at 30°C for 5 min. Double stranded biotin-labeled DNA (50 ng) was then mixed with This was followed by addition of 200 μl ONPG (4 mg/ml) and the 2 μl Binding buffer [50 mm KCl, 5 mm MgCl2, 20 mm Tris–HCl reaction was allowed to proceed until yellow coloration started to (pH 8), 100 μm EDTA, 1 mm DTT, 10% glycerol] and 1 μg Poly develop. This time duration for the appearance of yellow dI-dC. Phosphorylated CtrA (5 μl) was then added to the mixture coloration was noted and the reaction was stopped by addition of (final concentration 1.65 μm) making the total volume to 10 μl and 400 μl sodium carbonate (1 M). This was followed by centrifugation it was incubated for 25 min at RT. Loading buffer [2 μl, 100 mm (12,000 × g, 5 min) at room temperature and the absorbance at Tris–HCl (pH 7.8), 0.2% bromophenol blue] was then added to 420 nm of the supernatant was measured. Miller units for each the mix just before loading. A 5% native PAGE gel was prepared reaction were calculated using the Miller formula (A420 × 1,000)/ with the following composition: 7.5 ml of 1X TBE (pH 8), 5.25 ml [OD600 × time (min) × volume of cells used (ml)]. Reactions were water, 2.25 ml 40% polyacrylamide, 6 μl TEMED, 150 μl of 10% performed in triplicate and average and standard deviation APS. The gel was pre-run at 150 V for 5 min and then the samples were calculated. were loaded. Electrophoresis was carried out at 300 V for 30 min in 1X TBE buffer (pH 8). The gel, blotting paper and positively charged nylon membrane were then incubated in 0.5X TBE buffer Genome sequencing, assembly and (pH 8.3) for 20 min in a shaking platform at 4°C. Transfer was annotation carried out using semi-dry electrophoretic transfer cell (Biorad) at 25 V, 1A for 40 min. Post transfer, the DNA was crosslinked to Genome annotation was performed with DNA Master the membrane by exposure to UV short wave (254 nm) for 20 min (v5.23.3), using default settings (Pope and Jacobs-Sera, 2018). using a hand held UV device. Next, the membrane was incubated Automated gene calls from DNA Master were manually reviewed in blocking buffer (Tropix I-Block, Applied Biosystems) and further validated with ARAGORN (v1.2.38; Aragorn, supplemented with 0.2% Tween-20 on a shaking platform for 1 h RRID:SCR_015974), tRNAscan-SE (v2.0; tRNAscan-SE, at RT. The membrane was then incubated in blocking buffer with RRID:SCR_010835) and GeneMarkS (v3.25; GeneMark, 0.2% Tween-20 and 1.5 μl of 1.25 mg/ml HRP-conjugated RRID:SCR_011930; Besemer et al., 2001; Laslett and Canback, streptavidin at a dilution of 1:23000, on a shaking platform for 1 h 2004; Chan and Lowe, 2019). Functional annotations were at RT. Five consecutive washes were then carried out, each for completed using the NCBI BLASTP (NCBI BLAST, 5 min on a shaking platform at RT, with 1X TBST (0.2% Tween- RRID:SCR_004870) and HHPred (HHpred, RRID:SCR_010276) 20). The membrane was then incubated with 3 ml of Supersignal services with default parameters, following SEA-PHAGES West Pico PLUS Chemiluminescent Substrate for 10 min prior annotation guidelines (maximum BLASTP e-value: 10–7, to imaging. minimum HHpred probability: 90%) and independently curated Frontiers in Microbiology 05 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 by at least two annotators (Altschul et al., 1997; Söding et al., 2005; Genome sequencing resulted in an 87,225 bp-long linear Pope et al., 2014). genome assembly with distinct ends. To determine if the genome was circular, a restriction digest was carried out on the genome using enzymes AflII and PciI, which should define termini Results fragments of 2,938 and 3,793 bp in the assembled linear genome. Instead, two bands of approximately 4,000 bp were observed Morphology and genome architecture of corresponding to the internal digestion fragments of 3,448 and Brevundimonas phage vB_BsubS-Delta 4,497 bp, as well as a band of approximately 6,731 bp corresponding to the cumulative size of the two termini fragments (Figure 1B). Brevundimonas phage vB_BsubS-Delta was isolated from a This strongly suggests that vB_BsubS-Delta has a circularly freshwater pond in Lafayette County, Mississippi, and shown to permuted genome, in contrast to ΦCbk-like phages, which have have a narrow host range, infecting only four Brevundimonas linear genomes with large direct terminal repeats (DTR; Gill species and none of the Caulobacter species tested (Sperling et al., et al., 2012). 2019). Host mutants defective in pilus-biogenesis were shown to Functional annotation of the vB_BsubS-Delta genome be resistant to vB_BsubS-Delta, suggesting that this phage utilizes revealed a genome architecture distinct from that of ΦCbk-like pili to infect the cell (Sperling et al., 2019). Here we used phages (Figure 2). The vB_BsubS-Delta genome contains a well- transmission electron microscopy (TEM) and next-generation defined phage structural protein module (DELTA_1-DELTA_37) sequencing to assess the morphology and genomic architecture of encompassing tail and capsid components, the tape measure and vB_BsubS-Delta. portal proteins, and both terminase subunits located at the start TEM imaging of vB_BsubS-Delta reveals a canonical of the module. The structural module also contains a Lysin A siphophage morphology with long, non-contractile tails and an protein (DELTA_13). No independent lysis module was detected extremely prolate head (Figure 1A), reminiscent of the in the genome, but two of several predicted membrane proteins Caulobacter ΦCbk-like phages (Gill et al., 2012). Phage vB_ encoded at the end of the structural module (DELTA_34 and BsubS-Delta has a head width of 51.6 ± SD1.9 nm, consistent with DELTA_35) present homology to, respectively, phage holins and ΦCbk-like phages, but the head length is considerably shorter spanins, suggesting that they may constitute a separate lysis (91.4 ± SD3.0 nm) than that of ΦCbk-like phages (205–292 nm), module. The phage replication module (DELTA_38-DELTA_55) suggesting a smaller genome size. Extracted genomic nucleic acid contains a Pol I-type DNA polymerase, as well as DNA helicases, was subjected to digestion by either DNase I or RNase A. While a single-strand DNA-binding protein and a deoxynucleoside DNase led to complete destruction of genomic material, RNase monophosphate kinase. Additional genes involved in DNA had no effect, suggesting that the vB_BsubS-Delta genome synthesis and replication (DNA primase) are located in a small was DNA. divergently transcribed module (DELTA_56-DELTA_59) adjacent to the primary replication module. A separate module (DELTA_60-DELTA_119) divergently adjacent to the terminase-end of the structural module includes several genes encoding proteins likely involved in packaging- A B related DNA processing, including two HNH endonucleases, DNA topoisomerase, DNA helicase and DNA processing protein A, as well as a ThyX-like thymidylate synthase homolog. A large (1,312 bp) intergenic region separates this ancillary packaging module from the divergently transcribed structural module. Phylogenetic analysis of vB_BsubS-Delta Despite the morphological, receptor and host similarities between vB_BsubS-Delta and ΦCbk-like phages, the genomic organization of vB_BsubS-Delta, together with its smaller genome FIGURE 1 size, absence of tRNA genes and lower %GC content (62.49 %GC) Characterization of Brevundimonas phage vB_BsubS-Delta. suggested that this phage was not closely related to ΦCbk-like (A) TEM image of vB_BsubS-Delta from crude lysate. Scale bar=200 nm. Average head and tail length/width measurements phages. To investigate the possible relatedness of vB_BsubS-Delta from four independent TEM images. (B) Restriction digestion of to ΦCbk-like and other phages, we performed whole-proteome vB_BsubS-Delta genomic DNA with AflII and PciI to determine if the genome is linear or circular. Expected band sizes are phylogenetic inference. The analysis included 300 phage genomes indicated by arrows. partially related to vB_BsubS-Delta, identified via BLASTP with all vB_BsubS-Delta proteins as queries (Supplementary Table S2), Frontiers in Microbiology 06 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 2 Genome organization of Brevundimonas phage vB_BsubS-Delta. Predicted genes are represented as boxes above or below the genome, denoting, respectively, rightwards- and leftwards-transcription. representative complete sequences of 362 Alphaproteobacteria- Figure 4), which are known to disseminate laterally among phages infecting phages (Supplementary Table S3), and additional (Dwivedi et al., 2013). In contrast, hits to the Pseudomonas genomes for known pilitropic and flagellotropic phages YuA-like phages encompass primarily the major capsid and (Supplementary Table S4). A consensus phylogenetic tree for the terminase proteins (~10 kbp mark in Figure 4), supporting the 460 phage genomes was generated with ViPhy notion of an ancient relationship between vB_BsubS-Delta and (Supplementary Figure S1), and representative genomes for each Pseudomonas YuA-like phages as inferred in Figure 3. cluster identified via agglomerative clustering. A final set of 100 phage genome sequences (Supplementary Table S5) was used to perform whole-genome phylogenetic inference using the VICTOR Distribution of putative CtrA-binding web service. sites in phage genomes The consensus phylogenetic tree shown in Figure 3 (Supplementary Figure S1) reveals that phage vB_BsubS-Delta is The vB_BsubS-Delta genome contains a relatively large not directly related to any previously sequenced phages, placing it (1,312 bp) gap region separating divergently transcribed genes in a long branch within a well-supported small cluster (Figure 5). Visual inspection of this region revealed the presence encompassing phages targeting Roseobacter and Achromobacter of 30 TTAA repeats, 12 of which form canonical TTAA-N7- hosts. This small phage clade is loosely enclosed in a larger cluster TTAA CtrA-binding sites following the dyad-spacer CtrA- defined primarily by Pseudomonas YuA-like phages (Amgarten binding motif consensus (Quon et al., 1996). The TTAA-N7- et al., 2017), and far removed from the other two main TTAA CtrA-binding motif has been shown to be conserved and Alphaproteobacteria phage clusters identifiable in the tree: the functional across the Alphaproteobacteria, with the dyad HTVC019P-like Pelagibacter phages (Zhao et al., 2019) and the elements mediating the specific interaction with CtrA (Barnett ΦCbk-like Caulobacter phages (Gill et al., 2012). A genome-map et al., 2001; Brilli et al., 2010; Mercer et al., 2010; Greene et al., comparison of vB_BsubS-Delta with members of these three 2012; Poncin et al., 2018). Further analysis of this region clusters (Figure 4) shows that vB_BsubS-Delta shares only identified a previously undetected open reading frame (hereafter scattered and weak similarity with these disparate groups of referred to as DELTA_120) predicted to encode a 68 aa protein phages. Clear hits to Pelagibacter HTVC019P-like and and located immediately downstream of a cluster of CtrA- Caulobacter ΦCbk-like phages are restricted to the ribonucleoside- binding sites. Homology search with HHpred revealed weak diphosphate reductase subunits alpha and beta (~50 kbp mark in (probability 72–80%) structural similarity of DELTA_120 to Frontiers in Microbiology 07 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 3 Whole-genome phylogeny of vB_BsubS-Delta. Whole-proteome tree of complete bacteriophage genomes. The tree was generated with FASTML using BLAST-derived inter-genomic distances using VICTOR. Midpoint rooting was applied for circular visualization. Bootstrap support values are shown as branch widths, with the thickest branches corresponding to 100% bootstrap support. The branches corresponding to Alphaproteobacteria-infecting phages are colored in red. External filled circles and squares designate, respectively, known pilitropic and flagellotropic phages (Supplementary Table S4). Names for LPEG phages (Supplementary Table S7) are highlighted in yellow. External pie-charts indicate (green) the probability of temperateness assigned by the BACPHLIP algorithm (Supplementary Table S10). several thioltransferases, but no known function could To investigate whether this genomic feature was exclusive to be conclusively assigned to the DELTA_120 products. CtrA vB_BsubS-Delta, we performed a comparative analysis of CtrA- regulates cell-cycle control in the Caulobacterales and, more binding site distribution on phage genome sequences. This broadly, motility and chemotaxis in the Alphaproteobacteria. comparative analysis included 301 complete genome sequences of The presence of a genomic region with high concentration of phages partially related to vB_BsubS-Delta (Supplementary Table S2), CtrA-binding sites in a pilitropic phage therefore suggested that 362 complete sequences of Alphaproteobacteria-infecting phages these sites might play a functional role in the vB_BsubS- (Supplementary Table S3), and complete sequences from 93 known Delta genome. pilitropic and flagellotropic phages (Supplementary Table S5). Frontiers in Microbiology 08 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 4 Comparative genome map of representatives from major phylogenetic clusters using Brevundimonas phage vB_BsubS-Delta as a reference. The percent amino acid identities of each phage relative to Brevundimonas phage vB_BsubS-Delta, as identified by tBLASTX analysis, are shown as different hues. Concentric rings correspond to phages listed in the legend, with the innermost ring corresponding to Pseudomonas phage AAT-1 and the outermost ring corresponding to Pelagibacter phage HTVC105P. Segments with percent identity below 50% are not shown. The results shown in Figure 6 reveal that the genomes of a vB_BsubS-Delta and some other Alphaproteobacteria-infecting small subset of phages display marked asymmetry in the positional phages. Several of the LPEG phages branch together with ΦCbk- distribution of putative CtrA-binding sites, as measured by like phages in Figure 3 and are known to share substantial genomic positional entropy (Supplementary Tables S6, S7). The same and morphological similarities with them. In spite of their larger results are obtained when the Gini coefficient, a different measure size, Caulobacter phages CcrBL9, CcrBL10, CcrSC and CcrPW of distribution inequality, is applied to the positional distribution share a core proteome of 69 proteins with ΦCbk-like phages of putative CtrA-binding sites (Supplementary Figure S2). The (Wilson and Ely, 2019). The remaining LPEG phages present little plot clearly delineates a set of low positional entropy genome sequence similarity with previously sequenced phages (Figure 3). (LPEG) phages with remarkably low normalized positional Like vB_BsubS-Delta, Sphingomonas phage vB_StuS_MMDA13 entropy (Ĥ < 0.75) with regard to both the expected value given is a siphophage that has been reported to contain a large intergenic the identified number of sites and the rest of phage genomes gap separating divergently transcribed blocks of genes (Marmo analyzed. Identical analyses using the known TF-binding motifs et al., 2020, 13). In contrast, Rhizobium phage Pasto, presumed to of other transcription factors involved in Alphaproteobacteria be temperate, is a podophage with short DTRs (Manuel et al., cell-cycle regulation (GcrA, SciP and DnaA) yielded no evidence 2021). Rhizobium phage Palo is an unrelated podophage infecting of asymmetry in the distribution of putative TF-binding sites for Rhizobium phaseoli with short DTRs (Nabhani et al., 2021). any subset of phage genomes (Supplementary Figure S3). Ruegeria phage DSS3Φ1 and closely related phages (vB_RpoS-V7, The LPEG group of phages is dominated by pilitropic vB_RpoS-V11, vB_RpoS-V18, vB_RpoS-V16) have no large C. crescentus ΦCbk-like phages, but encompasses also intergenic gaps or reported DTRs and are closely related to Frontiers in Microbiology 09 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 5 Brevundimonas phage vB_BsubS-Delta gap region diagram. The positions of putative CtrA-binding sites and half-sites in Brevundimonas phage vB_BsubS-Delta gap region are displayed as green and orange boxes, respectively. Their sequences are reported, with consensus-matching bases in bold. The location of protein coding genes, including the predicted DELTA_120 open reading frame, is also shown. The 81bp probe upstream of DELTA_120 used in electromobility shift assays is underlined in red. flagellotropic phages of the Chivirus genus. These phages encode phages (Supplementary Figure S5), and indicate that the a repressor/integrase tandem and are presumed to be temperate CtrA-binding sites in LPEG phage genomes, and particularly phages (Zhan et al., 2018). in their DTR and gap genomic regions, may be under positive selection for CtrA binding. Putative CtrA-binding sites in LPEG phage genome Properties of CtrA-binding sites in LPEG sequences also localize primarily to intergenic regions (Mann– phages Whitney U test p < 10−5; Figure 7C), and this effect is mostly driven by the high-scoring sites in DTRs and gap regions To elucidate the possible role of the CtrA-binding sites in (Supplementary Figure S6). This suggests that, if bound by LPEG phages, we analyzed other properties of identified CtrA, the identified CtrA-binding sites may have regulatory CtrA-binding sites both within high-density regions and effects on nearby genes. To further characterize the prevalence overall in the genome. The PSSM score is a well-established of intergenic localization for CtrA-binding sites among LPEG indicator of binding specificity for most transcription factors phages, we analyzed the distribution of positional asymmetry (Zhao and Stormo, 2011). The average score for CtrA-binding of predicted CtrA-binding sites as a function of their intergenic sites in these phages is substantially larger than in control localization (Figure 8). phages (Mann–Whitney U test p < 10−5; Figure 7A). The results show that, when considered together, the genomic Furthermore, the majority of putative CtrA-binding sites in concentration and intergenic localization of predicted CtrA- the genome sequences of LPEG phages colocalizes to the binding sites define a clear separation between LPEG phages and distal ends of these phages, where DTRs or gap regions have the rest of phages in the dataset. The new analysis indicates that been primarily reported (Supplementary Figure S4), and the two additional Alphaproteobacteria-infecting phages also present average score of sites in DTRs and gap regions is larger than an unusual distribution of predicted CtrA-binding sites in the rest of the genome (Mann–Whitney U test p < 10−5; (Supplementary Table S7). Sphingobium phage Lacusarx Figure 7B). These differences are also observed when (Ĥ = 0.78) and Mesorhizobium phage Cp1R7A-A1 (Ĥ = 0.77) are comparing LPEG phages to Alphaproteobacteria-infecting both siphophages with previously reported morphological and Frontiers in Microbiology 10 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 6 Normalized Positional Entropy vs. number of identified CtrA-binding sites. Each filled circle indicates the normalized positional entropy of the genome as a function of the number of putative binding sites detected in its genome sequence. Uniform jitter was applied for clarity. Color coding in filled circles denotes, respectively, Caulobacter phages (green), other Alphaproteobacteria phages (red), vB_BsubS-Delta (purple), all the other analyzed phages (cyan) and the negative control (gray). The negative control was obtained by generating 10 random positional distributions for each possible number of sites and computing their normalized positional entropy (for a total of 5,890 data points). Open circles designate known pilitropic phages, and open squares denote known flagellotropic phages (Supplementary Table S4). genomic similarities with ΦCbk-like phages. These similarities predicted CtrA-binding sites in DTR regions, we performed encompass similar tail decorations, orthologous replication multiple sequence alignment of DTRs for representative module and tRNA clusters, the presence of large DTRs and a members of the different ΦCbk-like subclusters using Mauve postulated virulent lifestyle (Nielsen et al., 2017; Gunathilake and computed pairwise sequence identities with ViPhy. The et al., 2021). The presence of a distinct genomic signature in the resulting alignment (Supplementary Figure S7) revealed that predicted CtrA-binding site distribution of these phages, with the DTRs of Caulobacter phages Ccr29, ΦCbk, Rogue and Swift weaker colocalization than observed in LPEG phages but with present significant structural similarities and substantial strong preference for intergenic regions, suggests that the pairwise DNA sequence identity [ranging from 90 to 47% hypothetical transcriptional role of identified CtrA-binding sites (Supplementary Table S8)], as expected from previous genomic in Alphaproteobacteria-infecting phages may extend to analyses (Gill et al., 2012). The DTRs of these phages present non-LPEG phages. only low sequence identity to the Caulobacter phage CcrPW (12%) and Caulobacter phage CcrSC (8%) DTRs. The DTRs of Sphingobium phage Lacusarx and Mesorhizobium phage Properties of genomic regions with high Cp1R7A-A1 present marginal pairwise sequence identity (1%), density of putative CtrA-binding sites and similar residual similarity to Caulobacter phage CcrSC. Even among the closely related Caulobacter ΦCbk-like Sphingobium phage Lacusarx and Mesorhizobium phage phages (Ccr29, ΦCbk, Rogue and Swift), predicted CtrA- Cp1R7A-A1 define a well-supported phylogenetic cluster with binding sites are not fully conserved, and there is evidence of ΦCbk-like phages (Figure 3). These phages all present large gain/loss of CtrA-binding sites upstream of orthologous genes DTRs with a high concentration of predicted CtrA-binding sites following genomic rearrangements in DTRs (Figure 9). No and multiple protein coding genes with no known function. To known function could be conclusively assigned, based on investigate the evolutionary origin of the high-prevalence of homology search, to the product of any of the genes in these Frontiers in Microbiology 11 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 A B C FIGURE 7 Comparison of predicted CtrA-binding sites between LPEG and control phages. (A) Comparison of the distribution of the average PSSM score for predicted CtrA-binding sites over the entire genome, in LPEG phages and all the other analyzed phage genomes. (B) Comparison of the distribution of the average PSSM score for predicted CtrA-binding sites identified within and outside DTR/gap regions in LPEG phage genomes. (C) Comparison of the distribution of intergenic frequency for predicted CtrA-binding sites (defined as the proportion of predicted CtrA-binding sites identified in non-coding regions) in LPEG and all the other analyzed phage genomes. FIGURE 8 Normalized Positional Entropy vs. intergenic frequency of identified CtrA-binding sites. Each filled circle indicates the normalized positional entropy of the genome as a function of the intergenic frequency of putative CtrA-binding sites detected in its genome sequence. Uniform jitter was applied for clarity. Color coding in filled circles denotes, respectively, Caulobacter phages (green), phages infecting other Alphaproteobacteria (red), vB_BsubS-Delta phage (purple) and all the other analyzed phages (cyan). Open circles designate known pilitropic phages, and open squares denote known flagellotropic phages (Supplementary Table S4). DTRs presenting predicted CtrA-binding sites in their et al., 2021), and extend to the upstream region of the first Pasto upstream region. gene (CPT_Pasto_001), which encodes a Bacillus phage The remaining phages harboring regions with a high Φ29 single-stranded DNA-binding protein homolog concentration of CtrA-binding sites reveal unique configurations. (Supplementary Figure S8). In Rhizobium phage Palo, intergenic Predicted CtrA-binding sites in Rhizobium phage Pasto are CtrA-binding sites are found following its DTR, in the promoter primarily located toward the end of its DTR, which shows no region of genes CPT_Palo_014 and CPT_Palo_015, which have significant similarity to any other sequenced genomes (Manuel no known function (Supplementary Figure S8). Like Frontiers in Microbiology 12 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 vB_BsubS-Delta, Sphingomonas phage vB_StuS_MMDA13 Alphaproteobacteria-infecting hosts that encode a functional contains a ~ 1 Kb gap region separating divergently transcribed CtrA protein strongly suggests that the colocalization of CtrA- blocks of genes (Marmo et al., 2020). This region was shown to binding sites in these phage genomes has a functional role contain four instances of an imperfect 12 bp direct repeat motif involving CtrA binding. (5′-T3G2C4G4T4T4A4A4C4C4A3T4–3′) upstream of gene MMDA13_ gp54, which has no known function. These direct repeat motif instances overlap with the three putative CtrA-binding sites CtrA binds specifically to predicted identified here (Supplementary Figure S9). In contrast to other CtrA-binding sites in LPEG phages LPEG phages, Ruegeria phage DSS3Φ1 and closely related phages lack DTRs or gap regions, and present weak localization of CtrA- We performed Electrophoretic Mobility Shift Assays (EMSA) binding sites to intergenic regions (Supplementary Figure S10). to experimentally determine whether CtrA binds to predicted Predicted CtrA-binding sites also have consistently lower scores CtrA-binding sites or not. After unsuccessful attempts at purifying (10.07 ± SD 0.10) than those of other LPEG phages, and their B. subvibrioides CtrA, C. crescentus CtrA was purified from E. coli concentration in the genome is largely driven by three intragenic using an expression construct. B. subvibrioides and C. crescentus sites in gene DSS3P1_09 and its orthologs in the other are closely related, both belonging to the Caulobacterales order Ruegeria phages. (Curtis and Brun, 2014; Adhikari et al., 2021), and BLASTP The analysis of genomic regions with high density of predicted showed that these proteins are 87% identical. A multiple sequence CtrA-binding sites hence indicates that such regions have alignment revealed that representative Alphaproteobacteria CtrA independently evolved at least five times in unrelated homologs present an average amino acid identity of 74.44% ± SD Alphaproteobacteria-infecting phages (the ΦCbk-like cluster, 11.13, and confirmed that the B. subvibrioides CtrA contains all vB_BsubS-Delta, Sphingomonas phage vB_StuS_MMDA13, the domains and amino acid residues critical to its response Rhizobium phage Palo and Rhizobium phage Pasto). Moreover, regulator activity, including the well-documented C. crescentus the high concentration of CtrA-binding sites in these regions 51D phosphorylation site, the acidic pocket active site and the appears to have been maintained in the face of substantial genomic OmpR-like DNA-binding domain (Quon et al., 1996; Barnett divergence and reorganization within the diverse ΦCbk-cluster, et al., 2001; Supplementary Figure S11). Heterologous expression which encompasses phages infecting three different bacterial of CtrA orthologs from Alphaproteobacteria distantly related to genera (Caulobacter, Sphingobium and Mesorhizobium). C. crescentus [e.g., Sinorhizobium meliloti (82.25% identity), Together, the consistent concentration of putative CtrA-binding Rickettsia prowazekii (58.01% identity)] have been shown to fully sites in the specific genomic regions, their localization in complement a C. crescentus CtrA mutant (Barnett et al., 2001; intergenic space and the fact that this pattern is only observed in Brassinga et al., 2002), confirming that the CtrA phosphorelay FIGURE 9 Partial view of Mauve alignment of DTRs from Caulobacter ΦCbk-like phages with significant sequence identity. Sequence conservation is shown using the backbone color scheme, with mauve regions representing areas conserved among all displayed genomes. Open reading frames are shown as white boxes and predicted CtrA-binding sites as blue boxes. Superimposed red lines trace the conserved regions harboring predicted CtrA-binding sites. Colors in the locus tags of open reading frames designate orthologous genes. Frontiers in Microbiology 13 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 machinery and the CtrA-binding motif are conserved and the T401I mutation confers viability at 30°C but is lethal at 37°C functional across the Alphaproteobacteria (Brilli et al., 2010). due to CtrA inactivation (Quon et al., 1996). Given that CtrA is a response regulator that preferentially binds to DNA in C. crescentus CtrA had been shown to bind to the sequences of the its phosphorylated form. To overcome this issue, expression constructs (Figure 10), we examined the expression of we phosphorylated CtrA in vitro prior to the EMSA using a the expression constructs in C. crescentus NA1000 (WT) and modified form of a previously described methodology (Reisenauer LS2195 (temperature sensitive) strains (Figure 11). et al., 1999). In the WT C. crescentus strain, vB_BsubS-Delta DELTA_120 Binding of CtrA was assessed using DNA from three different construct expression was similar to the empty vector at both LPEG phages. For Brevundimonas phage vB_BsubS-Delta, a temperatures. However, when the construct was examined in the region of dsDNA (81 bp) upstream from the start codon of temperature-sensitive strain at the permissive temperature, DELTA_120 was used as a probe, as this was the open reading expression was >4.5-fold higher (Figure 11A). Under the frame with its 5′ end closest to the high-density region of predicted assumption of CtrA regulation, a shift is expected, since it is known CtrA-binding sites. This DNA probe included three CtrA-binding that the temperature-sensitive CtrA strain has compromised CtrA sites and two half-sites (Figure 5). For Caulobacter phage ΦCbk, function even at the permissive temperature, which then becomes 99 and 114 bp regions upstream of genes CbK_gp005 and CbK_ exacerbated to non-viability at the restrictive temperature (Quon gp018, both containing one predicted CtrA-binding site, were et al., 1996; Domian et al., 1999). This suggests that compromised used as probes. For Mesorhizobium phage Cp1R7A-A1, a 123 bp CtrA function leads to an increase in DELTA_120 gene expression. region upstream of gene Cp1R7AA1_019, containing a single When the strain was shifted to the restrictive temperature for 1 h, putative CtrA-binding site was used as a probe. No known expression further increased to almost 6-fold higher than WT function could be assigned to any of these genes via (Figure 11A), a shift that is larger than that observed for other CtrA- homology search. regulated genes (Domian et al., 1999). This data, combined with the As shown in Figure 10, the vB_BsubS-Delta DELTA_120 EMSA results (Figure 10), suggests that CtrA binds to the vB_ DNA probe by itself migrates near the bottom of the gel (lane 1), BsubS-Delta phage genome and represses DELTA_120 expression. but when the phosphorylated CtrA protein is added a significant The expression results for Caulobacter phage ΦCbk (CbK_ mobility shift is observed (lane 2). In addition, when CtrA was gp005 and CbK_gp018) and Mesorhizobium phage Cp1R7A-A1 incubated with unlabeled probe at 100-fold excess before (Cp1R7AA1_019) stand in stark contrast to those observed in incubating with labeled probe, most of the labeled probe was not vB_BsubS-Delta. The two Caulobacter phage ΦCbk constructs shifted (lane 3). Similarly, the ΦCbk CbK_gp005 and CbK_gp018 showed robust expression in WT C. crescentus (Figure 11B), with probes shift only in the presence of CtrA (lanes 5 and 8), but not the CbK_gp005 construct showing the highest expression. in its absence (lanes 4 and 7). In both cases, the CtrA-induced Conversely, both constructs showed virtually no expression in the retardation shift is abolished when a probe containing mutations LS2195 strain at both the permissive and restrictive temperatures. in the predicted CtrA-binding site is used. Identical results are Expression levels for the CbK_gp005 and CbK_gp018 constructs obtained for the Cp1R7A-A1 Cp1R7AA1_019 probe (lanes 10, 11 were, respectively, >29-fold and 5-fold lower in the LS2195 strain and 12). The lower molecular weight shift observed in lane 12 is compared to WT at the permissive temperature. Similar results likely the result of CtrA binding to two orphan TTAA half-sites were observed for the Cp1R7AA1_019 promoters from that were not mutated in that probe. Overall, these results show Mesorhizobium phage Cp1R7A-A1. Expression from this that CtrA binds specifically to predicted CtrA-binding sites in construct was almost 2-fold lower in the LS2195 strain at the non-orthologous regions from three different phages infecting permissive temperature compared to WT strain (Figure 11C). As distinct Alphaproteobacteria hosts. with the previous constructs, the expression levels were similar when the LS2195 strain was shifted to 37°C for 1 h. The fact that CtrA binds to predicted CtrA-binding sites in these promoters CtrA regulates gene expression in LPEG (Figure 10) and these results showing CtrA-dependent phages transcription suggest that CtrA acts as an activator for Caulobacter phage ΦCbk and Mesorhizobium phage Cp1R7A-A1 genes. To determine if CtrA binding affected expression of genes with upstream regions bound by CtrA, we created reporter constructs in which the promoter regions of DELTA_120 Discussion (Supplementary Figure S12), CbK_gp005 and CbK_gp018, and Cp1R7AA1_019 (Supplementary Figure S13) were fused with a Independent evolution of gene promoter-less beta-galactosidase gene in the plac290 plasmid regulation by host CtrA in multiple phage (Gober and Shapiro, 1992). Given that ctrA is essential in many genomes Alphaproteobacteria, including B. subvibrioides (Curtis and Brun, 2014), a ctrA null strain cannot be constructed. However, LS2195 A systematic analysis of the distribution of predicted CtrA- is a C. crescentus strain with a temperature-sensitive allele of CtrA; binding sites in 499 phage genome sequences uncovered an Frontiers in Microbiology 14 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 FIGURE 10 CtrA binds to LPEG phage DNA containing predicted CtrA-binding sites. EMSA assays using purified C. crescentus CtrA protein and DNA probes from different LPEG phages. Lane 1: Brevundimonas phage vB_BsubS-Delta DELTA_120 DNA. Lane 2: DELTA_120 DNA with phosphorylated CtrA. Lane 3: DELTA_120 DNA with phosphorylated CtrA and unlabeled DELTA_120 competitor DNA at 100-fold molar excess. Lane 4: Caulobacter phage ΦCbk CbK_gp005 DNA. Lane 5: CbK_gp005 DNA with phosphorylated CtrA. Lane 6: CbK_gp005 DNA with mutated CtrA binding site, with phosphorylated CtrA. Lane 7: Caulobacter phage ΦCbk CbK_gp018 DNA. Lane 8: CbK_gp018 DNA with phosphorylated CtrA. Lane 9: CbK_gp018 with mutated CtrA binding site, with phosphorylated CtrA. Lane 10: Mesorhizobium phage Cp1R7A-A1 Cp1R7AA1_019 DNA. Lane 11: Cp1R7AA1_019 DNA with phosphorylated CtrA. Lane 12: Cp1R7AA1_019 DNA with mutated primary CtrA binding site, with phosphorylated CtrA. unequivocal genomic signature in several Alphaproteobacteria- (vB_BsubS-Delta) or activate (ΦCbk & Cp1R7A-A1) transcription infecting phages, here termed LPEG phages (Figure 6). These of the associated genes. phages present marked positional asymmetry in the distribution Phylogenetic analysis of Alphaproteobacteria-infecting phages of predicted CtrA-binding sites, which localize primarily to either and phages potentially related to Brevundimonas phage vB_ intergenic gap regions or direct terminal repeats in these phage BsubS-Delta revealed that LPEG phages map to ostensibly genomes. Furthermore, putative CtrA-binding sites in LPEG unrelated phage groups, with vB_BsubS-Delta branching deep phages present both higher scores and preferential localization to within a cluster encompassing Pseudomonas YuA-like phages, intergenic regions, with both traits being driven largely by CtrA- Ruegeria phages defining a deep independent branch, binding sites in their DTRs or gap regions (Figure 7). The Sphingomonas phage vB_StuS_MMDA13 in an heterogeneous correlation between genomic positional asymmetry and cluster encompassing phages infecting different Proteobacteria intergenic localization (Figure 8) led to the identification of classes, and the rest of LPEG phages mapping to a large cluster of additional phages with unusual CtrA-binding site distribution. ΦCbk-like phages (Figure 3). The vast phylogenetic distance CtrA-binding sites often colocalize with DnaA- and SciP-binding between these disparate groups of phages, with no detectable sites in C. crescentus (Tan et al., 2010; Taylor et al., 2011), but no nucleotide sequence identity (Supplementary Table S9), indicates evidence of positional asymmetry was found in the distribution that the presence of colocalizing, predominantly intergenic CtrA- of predicted binding sites for DnaA, SciP or the functionally binding sites has evolved independently in all these phage lineages. related GcrA transcription factor, irrespective of the threshold Even among the Caulobacter ΦCbk-like phages, the DTRs used to identify putative binding sites (Supplementary Figure S3). harboring predicted CtrA-binding sites often share scant The functional role of predicted CtrA-binding sites was nucleotide sequence identity, and a multiple sequence alignment experimentally determined by EMSA and β-Gal assays in of DTR regions reveals substantial turnover in putative CtrA- Brevundimonas phage vB_BsubS-Delta, Caulobacter phage ΦCbk binding sites among closely related phages (Figure 9; and Mesorhizobium phage Cp1R7A-A1 (Figures 10, 11). Supplementary Figure S7). This, together with the experimental Phosphorylated CtrA was found to bind promoter regions observation of distinct modes of CtrA activity in promoters from containing putative CtrA-binding sites and to repress different phages (Figure 11), indicates that the prevalence of Frontiers in Microbiology 15 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 A B C FIGURE 11 CtrA regulates gene expression in multiple LPEG phages. Standardized amount of β-Gal activity for C. crescentus cells harboring a plac290 vector containing the promoter region of target genes fused with a promoter-less beta-galactosidase gene. Blue and orange denote, respectively, results for the WT and temperature-sensitive CtrA LS2195 C. crescentus strains. (A) Expression levels for Brevundimonas phage vB_BsubS-Delta DELTA_120 genes. (B) Expression levels for Caulobacter phage ΦCbk constructs for genes CbK_gp005 (gp5) and CbK_gp018 (gp18). (C) Expression levels for Mesorhizobium phage Cp1R7A-A1 Cp1R7AA1_019 (gp19) gene. Stars designate statistical significance (one-sided Mann– Whitney U test) in the difference between log-fold change (WT vs. LS2195) values at 30 and 37°C. CtrA-binding sites in these regions has been actively maintained lytic cycle leveraging the host transcriptional repressor LexA, in the face of substantial variation and rearrangement in which responds to DNA damage using the same molecular these regions. mechanism as phage repressors (Sassanfar and Roberts, 1990; A number of LPEG phages are known, or suspected based on Shearwin et al., 1998; Kimsey and Waldor, 2009; Kim and Ryu, substantial genomic similarity, to be pili/flagellotropic. The best 2013; Fornelos et al., 2015). Conversely, phages of the Φ29 family known example, Caulobacter phage ΦCbk, has been shown to exploit the Spo0A sporulation master regulator to halt the lytic target flagella for initial attachment, followed by subsequent cycle if the host transitions into the sporulation pathway (Castilla- adsorption at pilus portals (Guerrero-Ferreira et al., 2011). CtrA Llorente et al., 2009). This allows spore entrapment of the phage is best known for its role in cell-cycle progression in the chromosome, enabling the phage to attain a dormant state in the Rhizobiales, but experimental analyses indicate that the ancestral endospore (Meijer et al., 2005). These examples illustrate that the role of CtrA in the Alphaproteobacteria is the regulation of availability of a reliable signal transduced by a cellular motility, chemotaxis and some quorum sensing components (Bird transcription factor can be readily exploited by phages to control and MacKrell, 2011; Greene et al., 2012; Zan et al., 2013, 1; Wang their lytic cycle. et al., 2014; Francez-Charlot et al., 2015; Hernández-Valle et al., Our results indicate that several groups of pili/flagellotropic 2020), and a comparative genomics analysis of representatives Alphaproteobacteria-infecting phages have convergently from all Alphaproteobacteria orders supports this observation evolved the ability to use the cellular CtrA transcription factor (Supplementary Figure S14). The finding that CtrA regulates gene to regulate the expression of a small set of uncharacterized expression in several pili/flagellotropic phages raises the genes. This suggests that there is a clear adaptive advantage to tantalizing possibility that different groups of phages may have utilizing CtrA in phage decision-making. We hypothesize that independently evolved gene regulation mechanisms dependent pili/flagellotropic Alphaproteobacteria-infecting phages leverage upon the same transcriptional regulator that governs the the host CtrA to implement a form of lytic deferment, generation of the cellular appendages required for efficient conceptually similar to that observed in Bacillus Φ29-like phage infection. phages. We illustrate this hypothesis on the well-studied Caulobacter-ΦCbk system (Figure 12), as a representative of the Caulobacterales and Hyphomicrobiales orders. In these orders, CtrA-mediated lytic deferment in CtrA acquired a well-defined role as global regulator and key Alphaproteobacteria-infecting phages component of cell cycle progression, which subsequently tied flagellum and pilus synthesis to specific stages of the cell cycle. Decision-making in bacteriophages revolves predominantly In Caulobacter-like bacteria, flagella are synthesized in the around the control of the phage lytic cycle in response to external predivisional cell stage but rotation is not activated until cell stimuli. Enterobacteria phage Lambda and Salmonella phage P22 division; pilus filaments are also not synthesized until cell are canonical temperate phages encoding transcriptional division. Both processes are dependent on CtrA activity. Nearly repressors that activate the lytic cycle in response to DNA damage identical mechanisms are found in members of the (Phizicky and Roberts, 1980, 22), and several unrelated groups of Hyphomicrobiales, with cell cycle-driven CtrA-dependent phages have independently evolved the ability to activate their flagellum and/or pilus production seen in Agrobacterium Frontiers in Microbiology 16 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 tumefaciens (Kahng and Shapiro, 2001) and S. meliloti (Xue and Biondi, 2019). This means that for most organisms in the Caulobacterales and Hyphomicrobiales, a flagellotropic/ pilitropic phage will only infect swarmer cells. Moreover, in the infected swarmer cell, the CtrA-regulated phage genes would presumably be repressed, mirroring the silencing of host genes by CtrA in this phase. Thus, CtrA-regulated genes would need the cell cycle of the host to progress before expression could occur. For CtrA-repressed genes (as in Brevundimonas phage vB_BsubS-Delta) this would occur during the swarmer to stalked cell transition when CtrA is cleared from the cell. This event is coincident with host chromosome replication initiation (Quon et al., 1998). For CtrA-induced genes (such as those found in ΦCbk-like phages), the host cell cycle would need to progress through swarmer cell differentiation into the mid-to- FIGURE 12 Schematic representation of the proposed life cycle for ΦCbk- late predivisional cell stage, where CtrA is resynthesized and like phages. Phages infecting Caulobacter-like species require activated by phosphorylation; this is coincident with significant the pili/flagella present in swarmer cells for infection. Upon progression of chromosome replication (Domian et al., 1997). infection, cells encounter non-peak levels of activated CtrA and defer the lytic cycle. After the swarmer cell transitions to a stalked The timing of phage gene expression to host cell cycle cell, levels of activated CtrA start to rise, eventually triggering the progression could provide several advantages. First, members of expression of phage genes and the development of the lytic the Caulobacter clade are known to live in oligotrophic cycle in an environment where swarmer cells are being produced. environments where nutrients are scarce. Timing phage gene expression to the host cell cycle, particularly stages of the cell cycle that have chromosome replication, may ensure that the host has enough nutrients to synthesize new phage components. Secondly, is LD, where lysis inevitably occurs as part of the lytic members of the Caulobacter clade are known to often live in deferment pathway. multicellular arrangements, either attached to surfaces as biofilms We note that lytic deferment would provide time for the or as multi-cell clusters known as rosettes (Poindexter, 1964). swarmer cell to attach to a colony of growing cells, effectively Production of the infectious structures of flagella and pili occurs turning the infected swarmer cell into a Trojan Horse (Figure 12). on swarmer cells. Hence, it is in the phage’s best interest to Importantly, LD would be difficult to observe in laboratory be around cells that are producing new swarmer cells. This would conditions. In an actively growing culture, cells are constantly not be when a swarmer cell is swimming in bulk water, but when progressing through the cell cycle, and hence rapidly terminating a swarmer cell attaches to a colony of cells and differentiates. the lytic cycle deferment in the infecting phage. While these are We suggest that this CtrA-mandated temporary repression of only hypotheses, it is becoming increasingly evident that phage the phage infection cycle be termed lytic deferment (LD). infecting members of the Caulobacterales and Hyphomicrobiales Although gene regulation is involved, LD is completely distinct may have an interest in the cell cycle of their hosts. ΦCbk-like from the temperate phage pathway to lysogeny. In lysogeny, the phages infecting Caulobacter contain homologs of the cell cycle phage infection cycle undergoes a binary choice, toward repression regulator GcrA in their genomes (Gill et al., 2012), and some and establishment of the prophage state or toward productive Sinorhizobium phages carry analogs of the DNA methyltransferase lysis. This choice is influenced by physiological conditions and CcrM, which also participates in cell cycle-based gene expression stochastic events (Arkin et al., 1998), and the frequency of (Decewicz et al., 2017). lysogeny varies dramatically between different temperate phages Outside the Caulobacterales and Hyphomicrobiales, CtrA is under the same conditions. In contrast, we posit that every ΦCbk not known to be involved in cell-cycle regulation. However, CtrA infection enters the LD state, which lasts until the host transitions has an ancestral role as a regulator in motility, chemotaxis and into the stalked cell state. It is also possible that some phage genes quorum sensing systems across the Alphaproteobacteria (Bird and may accelerate the transition; indeed, ΦCbk encodes a homolog MacKrell, 2011; Greene et al., 2012; Zan et al., 2013, 1; Wang et al., of GcrA (Gill et al., 2012). We also would make the distinction 2014; Francez-Charlot et al., 2015; Hernández-Valle et al., 2020). between LD and the “pseudolysogeny” state that has been reported CtrA-mediated lytic deferment would hence provide a significant for various types of phage infections (Łoś and Węgrzyn, 2012). As adaptive advantage to pili/flagellotropic phages infecting the noted by Łoś et al., pseudolysogeny has not been formally defined Alphaproteobacteria, enabling phages to either infer the presence in terms of genes that are required for it in any system; rather, the of infectable cells in their environment or to sense a loss of motility claim for a pseudo-lysogenic state is usually based on the typically associated with high cellular density. Lytic deferment persistence of phage virions and/or gDNAs in a culture that does requires that the phage chromosome be capable of surviving for a not undergo lysis. This is clearly different from what we propose limited amount of time in the cell cytoplasm. A number of Frontiers in Microbiology 17 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 mechanisms, ranging from circularization to lysogeny, can PC, and IE: methodology. RY, PC, and IE: project administration. be implemented to increase chromosome survival. The specific SC, RY, PC, TC, and IE: resources. EM, AFS, and IE: software. mechanisms used by the different pili/flagellotropic phages to JS-S, RY, PC, and IE: supervision. EM and IE: visualization. EM, survive lytic deferment remain to be elucidated and are likely PC, SA, and IE: writing—original draft. EM, TC, SC, SA, JS-S, diverse, given the independent evolution of this strategy. In silico AFS, RY, PC, and IE: writing—review and editing. analyses did not detect the presence of plasmid-associated genes in LPEG phages (Pfeifer et al., 2021), but analysis of all analyzed phage genomes with the BACPHLIP algorithm strongly predicted several of the LPEG phages, including ΦCbk-like phages (p > 0.95), Funding as temperate (Figure 3; Supplementary Table S10; Supplementary Figure S15). Other LPEG phages, like This work was funded in part by the United States National Brevundimonas phage vB_BsubS-Delta, cluster deep within a Science Foundation CAREER program award 1552647 to PC, and large clade of phages shown to confer homoimmunity, but not to by grant NIH R35 GM136396 to RY. EM was supported by the integrate into host replicons (Williamson et al., 2001; Lohr et al., UMBC Department of Biological Sciences. 2005; Ceyssens and Lavigne, 2010). This suggests that some LPEG phages may be able to integrate into the host chromosome, while others adopt alternative mechanisms. This research clearly generates many more questions than it Acknowledgments answers, but it may be a first indication that phages infecting the Alphaproteobacteria, and particularly those relying on pili/flagella Phylogenetic inference computations were performed using to infect their hosts, may have co-opted the transcriptional the UMBC High Performance Computing Facility (HPCF). The activator that controls the development of these cellular facility is supported by the US National Science Foundation appendages to time their replication in a way that maximizes their through the MRI program (CNS-0821258, CNS-1228778 and chances of finding infectable cells and their dispersal across OAC-1726023) and the SCREMS program (DMS-08215311), scattered niches of high bacterial growth. In the absence of phage with additional support from the University of Maryland, mutants or homology-predicted functions for the phage genes Baltimore County (UMBC). Sequencing was carried out with shown to be CtrA-regulated, the precise molecular mechanisms support from the Center for Phage Technology, sponsored jointly underpinning lytic deferment cannot be determined and the by the Texas A&M AgriLife and the office of the Vice President proposal of such a pathway must remain speculative. However, for Research at Texas A&M University. Anurag Rijal provided while other interpretations for the computational and critical aid to the phage EMSA experiments. experimental results reported here are possible, the hypothesis of phages leveraging CtrA to implement lytic deferment provides a well-defined evolutionary rationale that is consistent with the Conflict of interest recent identification of quorum sensing mechanisms in temperate phages infecting the Bacillales, with the well-established control The authors declare that the research was conducted in the of Gene Transfer Agents (GTA) by CtrA (Lang et al., 2017), and absence of any commercial or financial relationships that could with the reported presence in different Alphaproteobacteria- be construed as a potential conflict of interest. infecting phages of homologs for cell-cycle regulators. Publisher’s note Data availability statement All claims expressed in this article are solely those of the The datasets and code used in this study can be found in the authors and do not necessarily represent those of their affiliated LPEG_phages repository (https://github.com/ErillLab/LPEG_ organizations, or those of the publisher, the editors and the phages). The Brevundimonas phage vB_BsubS-Delta complete reviewers. Any product that may be evaluated in this article, or genome sequence is available at the NCBI GenBank under claim that may be made by its manufacturer, is not guaranteed or accession number MN862068.1. endorsed by the publisher. Author contributions Supplementary material PC and IE: conceptualization. EM, SC, and IE: data curation. The Supplementary material for this article can be found online EM, SA, PC, and IE: formal analysis. PC and RY: funding at: https://www.frontiersin.org/articles/10.3389/fmicb.2022.918015/ acquisition. EM, RY, TC, SC, PC, SA, and IE: investigation. EM, full#supplementary-material Frontiers in Microbiology 18 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 References Adhikari, S., Erill, I., and Curtis, P. D. (2021). Transcriptional rewiring of the Domian, I. J., Reisenauer, A., and Shapiro, L. (1999). Feedback control of a master GcrA/CcrM bacterial epigenetic regulatory system in closely related bacteria. PLoS bacterial cell-cycle regulator. Proc. Natl. Acad. Sci. U. S. A. 96, 6648–6653. doi: Genet. 17:e1009433. doi: 10.1371/journal.pgen.1009433 10.1073/pnas.96.12.6648 Alikhan, N.-F., Petty, N. K., Ben Zakour, N. L., and Beatson, S. A. (2011). BLAST Dwivedi, B., Xue, B., Lundin, D., Edwards, R. A., and Breitbart, M. (2013). A ring image generator (BRIG): simple prokaryote genome comparisons. BMC bioinformatic analysis of ribonucleotide reductase genes in phage genomes and Genomics 12:402. doi: 10.1186/1471-2164-12-402 metagenomes. BMC Evol. Biol. 13:33. doi: 10.1186/1471-2148-13-33 Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., et al. Erill, I., Campoy, S., and Barbe, J. (2007). Aeons of distress: an evolutionary (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database perspective on the bacterial SOS response. FEMS Microbiol. Rev. 31, 637–656. doi: search programs. Nucleic Acids Res. 25, 3389–3402. doi: 10.1093/nar/25.17.3389 10.1111/j.1574-6976.2007.00082.x Amgarten, D., Martins, L. F., Lombardi, K. C., Antunes, L. P., de Souza, A. P. S., Fornelos, N., Bamford, J. K. H., and Mahillon, J. (2011). Phage-borne factors and Nicastro, G. G., et al. (2017). Three novel Pseudomonas phages isolated from host LexA regulate the lytic switch in phage GIL01▿. J. Bacteriol. 193, 6008–6019. composting provide insights into the evolution and diversity of tailed phages. BMC doi: 10.1128/JB.05618-11 Genomics 18:346. doi: 10.1186/s12864-017-3729-z Fornelos, N., Browning, D. F., and Butala, M. (2016). The use and abuse of LexA Arkin, A., Ross, J., and McAdams, H. H. (1998). Stochastic kinetic analysis of by Mobile genetic elements. Trends Microbiol. 24, 391–401. doi: 10.1016/j. developmental pathway bifurcation in phage lambda-infected Escherichia coli cells. tim.2016.02.009 Genetics 149, 1633–1648. doi: 10.1093/genetics/149.4.1633 Fornelos, N., Butala, M., Hodnik, V., Anderluh, G., Bamford, J. K., and Salas, M. Barnett, M. J., Hung, D. Y., Reisenauer, A., Shapiro, L., and Long, S. R. (2001). A (2015). Bacteriophage GIL01 gp7 interacts with host LexA repressor to enhance homolog of the CtrA cell cycle regulator is present and essential in Sinorhizobium DNA binding and inhibit RecA-mediated auto-cleavage. Nucleic Acids Res. 43, meliloti. J. Bacteriol. 183, 3204–3210. doi: 10.1128/JB.183.10.3204-3210.2001 7315–7329. doi: 10.1093/nar/gkv634 Benson, D. A., Cavanaugh, M., Clark, K., Karsch-Mizrachi, I., Lipman, D. J., Francez-Charlot, A., Kaczmarczyk, A., and Vorholt, J. A. (2015). The branched Ostell, J., et al. (2017). GenBank. Nucleic Acids Res. 45, D37–D42. doi: 10.1093/nar/ CcsA/CckA-ChpT-CtrA phosphorelay of Sphingomonas melonis controls motility gkw1070 and biofilm formation. Mol. Microbiol. 97, 47–63. doi: 10.1111/mmi.13011 Bertozzi Silva, J., Storms, Z., and Sauvageau, D. (2016). Host receptors for Frost, L. S. (1993). “Conjugative Pili and Pilus-specific phages,” in Bacterial bacteriophage adsorption. FEMS Microbiol. Lett. 363:fnw002. doi: 10.1093/femsle/ Conjugation. ed. D. B. Clewell (Boston, MA: Springer US), 189–221. fnw002 Gill, J. J., Berry, J. D., Russell, W. K., Lessor, L., Escobar-Garcia, D. A., Besemer, J., Lomsadze, A., and Borodovsky, M. (2001). GeneMarkS: a self-training Hernandez, D., et al. (2012). The Caulobacter crescentus phage phiCbK: genomics method for prediction of gene starts in microbial genomes. Implications for finding of a canonical phage. BMC Genomics 13:542. doi: 10.1186/1471-2164-13-542 sequence motifs in regulatory regions. Nucleic Acids Res. 29, 2607–2618. doi: Gober, J. W., and Shapiro, L. (1992). A developmentally regulated Caulobacter 10.1093/nar/29.12.2607 flagellar promoter is activated by 3′ enhancer and IHF binding elements. Mol. Biol. Bird, T. H., and MacKrell, A. (2011). A CtrA homolog affects swarming motility Cell 3, 913–926. doi: 10.1091/mbc.3.8.913 and encystment in Rhodospirillum centenum. Arch. Microbiol. 193, 451–459. doi: Gonzalez, D., Kozdon, J. B., McAdams, H. H., Shapiro, L., and Collier, J. (2014). 10.1007/s00203-011-0676-y The functions of DNA methylation by CcrM in Caulobacter crescentus: a global Bradley, D. E. (1972). Shortening of Pseudomonas aeruginosa pili after RNA-phage approach. Nucleic Acids Res. 42, 3720–3735. doi: 10.1093/nar/gkt1352 adsorption. J. Gen. Microbiol. 72, 303–319. doi: 10.1099/00221287-72-2-303 Greene, S. E., Brilli, M., Biondi, E. G., and Komeili, A. (2012). Analysis of the CtrA Brady, A., Quiles-Puchalt, N., del Sol, F. G., Zamora-Caballero, S., Felipe-Ruíz, A., pathway in Magnetospirillum reveals an ancestral role in motility in Val-Calvo, J., et al. (2021). The arbitrium system controls prophage induction. Curr. Alphaproteobacteria. J. Bacteriol. 194, 2973–2986. doi: 10.1128/JB.00170-12 Biol. 31, 5037–5045. doi: 10.1016/j.cub.2021.08.072 Guerrero-Ferreira, R. C., Viollier, P. H., Ely, B., Poindexter, J. S., Georgieva, M., Brassinga, A. K. C., Siam, R., McSween, W., Winkler, H., Wood, D., and Jensen, G. J., et al. (2011). Alternative mechanism for bacteriophage adsorption to Marczynski, G. T. (2002). Conserved response regulator CtrA and IHF binding sites the motile bacterium Caulobacter crescentus. Proc. Natl. Acad. Sci. U. S. A. 108, in the alpha-proteobacteria Caulobacter crescentus and Rickettsia prowazekii 9963–9968. doi: 10.1073/pnas.1012388108 chromosomal replication origins. J. Bacteriol. 184, 5789–5799. doi: 10.1128/ Gunathilake, K. M. D., Halmillawewa, A. P., MacKenzie, K. D., Perry, B. J., JB.184.20.5789-5799.2002 Yost, C. K., and Hynes, M. F. (2021). A bacteriophage infecting Mesorhizobium Brilli, M., Fondi, M., Fani, R., Mengoni, A., Ferri, L., Bazzicalupo, M., et al. (2010). species has a prolate capsid and shows similarities to a family of Caulobacter The diversity and evolution of cell cycle regulation in alpha-proteobacteria: a crescentus phages. Can. J. Microbiol. 67, 147–160. doi: 10.1139/cjm-2020-0281 comparative genomic analysis. BMC Syst. Biol. 4:52. doi: 10.1186/1752-0509-4-52 Haakonsen, D. L., Yuan, A. H., and Laub, M. T. (2015). The bacterial cell cycle Castilla-Llorente, V., Salas, M., and Meijer, W. J. J. (2009). Different responses to regulator GcrA is a σ70 cofactor that drives gene expression from a subset of Spo0A-mediated suppression of the related Bacillus subtilis phages Nf and phi29. methylated promoters. Genes Dev. 29, 2272–2286. doi: 10.1101/gad.270660.115 Environ. Microbiol. 11, 1137–1149. doi: 10.1111/j.1462-2920.2008.01845.x Haft, D. H., DiCuccio, M., Badretdin, A., Brover, V., Chetvernin, V., O’Neill, K., Ceyssens, P.-J., and Lavigne, R. (2010). Bacteriophages of Pseudomonas. Future et al. (2018). RefSeq: an update on prokaryotic genome annotation and curation. Microbiol. 5, 1041–1055. doi: 10.2217/fmb.10.66 Nucleic Acids Res. 46, D851–D860. doi: 10.1093/nar/gkx1068 Chan, P. P., and Lowe, T. M. (2019). tRNAscan-SE: searching for tRNA genes in Hernández-Valle, J., Sanchez-Flores, A., Poggio, S., Dreyfus, G., and Camarena, L. genomic sequences. Methods Mol. Biol. 1962, 1–14. doi: 10.1007/978-1- (2020). The CtrA Regulon of Rhodobacter sphaeroides favors adaptation to a 4939-9173-0_1 particular lifestyle. J. Bacteriol. 202, e00678–e00619. doi: 10.1128/JB.00678-19 Cook, R., Brown, N., Redgwell, T., Rihtman, B., Barnes, M., Clokie, M., et al. Hertz, G. Z., and Stormo, G. D. (1999). Identifying DNA and protein patterns with (2021). INfrastructure for a PHAge REference database: identification of large-scale statistically significant alignments of multiple sequences. Bioinformatics 15, biases in the current collection of cultured phage genomes. PHAGE. 2, 214–223. doi: 563–577. doi: 10.1093/bioinformatics/15.7.563 10.1089/phage.2021.0007 Hobbs, E. T., Pereira, T., O’Neill, P. K., and Erill, I. (2016). A Bayesian inference Curtis, P. D., and Brun, Y. V. (2014). Identification of essential alphaproteobacterial method for the analysis of transcriptional regulatory networks in metagenomic data. genes reveals operational variability in conserved developmental and cell cycle Algori. Molecular Bio. 11:19. doi: 10.1186/s13015-016-0082-8 systems. Mol. Microbiol. 93, 713–735. doi: 10.1111/mmi.12686 Hockenberry, A. J., and Wilke, C. O. (2021). BACPHLIP: predicting bacteriophage Darling, A. C. E., Mau, B., Blattner, F. R., and Perna, N. T. (2004). Mauve: multiple lifestyle from conserved protein domains. PeerJ 9:e11396. doi: 10.7717/peerj.11396 alignment of conserved genomic sequence with rearrangements. Genome Res. 14, Hottes, A. K., Shapiro, L., and McAdams, H. H. (2005). DnaA coordinates 1394–1403. doi: 10.1101/gr.2289704 replication initiation and cell cycle transcription in Caulobacter crescentus. Mol. Decewicz, P., Radlinska, M., and Dziewit, L. (2017). Characterization of Microbiol. 58, 1340–1353. doi: 10.1111/j.1365-2958.2005.04912.x Sinorhizobium sp. LM21 Prophages and virus-encoded DNA Methyltransferases in Howard-Varona, C., Hargreaves, K. R., Abedon, S. T., and Sullivan, M. B. (2017). the light of comparative genomic analyses of the Sinorhizobial Virome. Viruses Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J. 9:161. doi: 10.3390/v9070161 11, 1511–1520. doi: 10.1038/ismej.2017.16 Domian, I. J., Quon, K. C., and Shapiro, L. (1997). Cell type-specific Kahng, L. S., and Shapiro, L. (2001). The CcrM DNA methyltransferase of phosphorylation and proteolysis of a transcriptional regulator controls the G1-to-S Agrobacterium tumefaciens is essential, and its activity is cell cycle regulated. J. transition in a bacterial cell cycle. Cell 90, 415–424. doi: 10.1016/s0092-8674(00) Bacteriol. 183, 3065–3075. doi: 10.1128/JB.183.10.3065-3075.2001 80502-4 Frontiers in Microbiology 19 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 Kılıç, S., White, E. R., Sagitova, D. M., Cornish, J. P., and Erill, I. (2014). CollecTF: Poindexter, J. S. (1964). Biological properties and classification of the Caulobacter a database of experimentally validated transcription factor-binding sites in Bacteria. group. Bacteriol. Rev. 28, 231–295. doi: 10.1128/br.28.3.231-295.1964 Nucleic Acids Res. 42, D156–D160. doi: 10.1093/nar/gkt1123 Poncin, K., Gillet, S., and De Bolle, X. (2018). Learning from the master: targets Kim, M., and Ryu, S. (2013). Antirepression system associated with the life cycle and functions of the CtrA response regulator in Brucella abortus and other alpha- switch in the temperate podoviridae phage SPC32H. J. Virol. 87, 11775–11786. doi: proteobacteria. FEMS Microbiol. Rev. 42, 500–513. doi: 10.1093/femsre/fuy019 10.1128/JVI.02173-13 Pope, W. H., and Jacobs-Sera, D. (2018). Annotation of bacteriophage genome Kimsey, H. H., and Waldor, M. K. (2009). Vibrio cholerae LexA coordinates CTX sequences using DNA master: An overview. Methods Mol. Biol. 1681, 217–229. doi: prophage gene expression. J. Bacteriol. 191, 6788–6795. doi: 10.1128/JB.00682-09 10.1007/978-1-4939-7343-9_16 Lagenaur, C., Farmer, S., and Agabian, N. (1977). Adsorption properties of stage- Pope, W. H., Jacobs-Sera, D., Russell, D. A., Rubin, D. H. F., Kajee, A., Msibi, Z. N. specific Caulobacter phage phiCbK. Virology 77, 401–407. doi: 10.1016/0042- P., et al. (2014). Genomics and proteomics of mycobacteriophage patience, an 6822(77)90436-6 accidental tourist in the Mycobacterium neighborhood. mBio 5, e02145–e02114. doi: 10.1128/mBio.02145-14 Lang, A. S., Westbye, A. B., and Beatty, J. T. (2017). The distribution, evolution, and roles of gene transfer agents in prokaryotic genetic exchange. Annu Rev Virol 4, Ptashne, M. (1992). A genetic Switch: phage [lambda] and Higher Organisms. 2nd 87–104. doi: 10.1146/annurev-virology-101416-041624 ed. Cambridge, Mass.: Cell Press. Laslett, D., and Canback, B. (2004). ARAGORN, a program to detect tRNA genes Quon, K. C., Marczynski, G. T., and Shapiro, L. (1996). Cell cycle control by an and tmRNA genes in nucleotide sequences. Nucleic Acids Res. 32, 11–16. doi: essential bacterial two-component signal transduction protein. Cell 84, 83–93. doi: 10.1093/nar/gkh152 10.1016/s0092-8674(00)80995-2 Laub, M. T., Chen, S. L., Shapiro, L., and McAdams, H. H. (2002). Genes directly Quon, K. C., Yang, B., Domian, I. J., Shapiro, L., and Marczynski, G. T. (1998). controlled by CtrA, a master regulator of the Caulobacter cell cycle. Proc. Natl. Acad. Negative control of bacterial DNA replication by a cell cycle regulatory protein that Sci. U. S. A. 99, 4632–4637. doi: 10.1073/pnas.062065699 binds at the chromosome origin. Proc. Natl. Acad. Sci. U. S. A. 95, 120–125. doi: 10.1073/pnas.95.1.120 Lefort, V., Desper, R., and Gascuel, O. (2015). FastME 2.0: A comprehensive, accurate, and fast distance-based phylogeny inference program. Mol. Biol. Evol. 32, Reisenauer, A., Quon, K., and Shapiro, L. (1999). The CtrA response regulator 2798–2800. doi: 10.1093/molbev/msv150 mediates temporal control of gene expression during the Caulobacter cell cycle. J. Bacteriol. 181, 2430–2439. doi: 10.1128/JB.181.8.2430-2439.1999 Letunic, I., and Bork, P. (2019). Interactive tree Of life (iTOL) v4: recent updates and new developments. Nucleic Acids Res. 47, W256–W259. doi: 10.1093/nar/ Sassanfar, M., and Roberts, J. W. (1990). Nature of the SOS-inducing signal in gkz239 Escherichia coli. The involvement of DNA replication. J. Mol. Biol. 212, 79–96. doi: 10.1016/0022-2836(90)90306-7 Lohr, J. E., Chen, F., and Hill, R. T. (2005). Genomic analysis of bacteriophage PhiJL001: insights into its interaction with a sponge-associated alpha-proteobacterium. Sauer, R. T., Ross, M. J., and Ptashne, M. (1982). Cleavage of the lambda and P22 Appl. Environ. Microbiol. 71, 1598–1609. doi: 10.1128/AEM.71.3.1598-1609.2005 repressors by recA protein. J. Biol. Chem. 257, 4458–4462. doi: 10.1016/ S0021-9258(18)34744-6 Łoś, M., and Węgrzyn, G. (2012). Pseudolysogeny. Adv. Virus Res. 82, 339–349. doi: 10.1016/B978-0-12-394621-8.00019-4 Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., et al. (2012). Fiji: an open-source platform for biological-image analysis. Nat. Manuel, N., Rushing, L., Ravindran, A., Newkirk, H., Burrowes, B., Young, R., Methods 9, 676–682. doi: 10.1038/nmeth.2019 et al. (2021). Complete genome sequence of Rhizobium japonicum Podophage Pasto. Microbiol Resour Announc 10, e01442–e01420. doi: 10.1128/MRA.01442-20 Shearwin, K. E., Brumby, A. M., and Egan, J. B. (1998). The tum protein of coliphage 186 is an antirepressor. J. Biol. Chem. 273, 5708–5715. doi: 10.1074/ Marmo, P., Thaller, M. C., Di Lallo, G., Henrici De Angelis, L., Poerio, N., De jbc.273.10.5708 Santis, F., et al. (2020). Characterization of vB_StuS_MMDA13, a newly discovered bacteriophage infecting the agar-degrading species Sphingomonas turrisvirgatae. Skerker, J. M., and Shapiro, L. (2000). Identification and cell cycle control of a Viruses 12:E894. doi: 10.3390/v12080894 novel pilus system in Caulobacter crescentus. EMBO J. 19, 3223–3234. doi: 10.1093/ emboj/19.13.3223 Meier-Kolthoff, J. P., Auch, A. F., Klenk, H.-P., and Göker, M. (2013). Genome sequence-based species delimitation with confidence intervals and improved Söding, J., Biegert, A., and Lupas, A. N. (2005). The HHpred interactive server for distance functions. BMC Bioinfo. 14:60. doi: 10.1186/1471-2105-14-60 protein homology detection and structure prediction. Nucleic Acids Res. 33, W244– W248. doi: 10.1093/nar/gki408 Meier-Kolthoff, J. P., and Göker, M. (2017). VICTOR: genome-based phylogeny and classification of prokaryotic viruses. Bioinformatics 33, 3396–3404. doi: 10.1093/ Sperling, L., Mulero Alegría, M. D., Kaever, V., and Curtis, P. D. (2019). Analysis bioinformatics/btx440 of Brevundimonas subvibrioides developmental signaling systems reveals inconsistencies between phenotypes and c-di-GMP levels. J. Bacteriol. 201, e00447– Meijer, W. J. J., Castilla-Llorente, V., Villar, L., Murray, H., Errington, J., and e00419. doi: 10.1128/JB.00447-19 Salas, M. (2005). Molecular basis for the exploitation of spore formation as survival mechanism by virulent phage phi29. EMBO J. 24, 3647–3657. doi: 10.1038/sj. Summer, E. J. (2009). Preparation of a phage DNA fragment library for whole emboj.7600826 genome shotgun sequencing. Methods Mol. Biol. 502, 27–46. doi: 10.1007/978-1-60327-565-1_4 Mercer, R. G., Callister, S. J., Lipton, M. S., Pasa-Tolic, L., Strnad, H., Paces, V., et al. (2010). Loss of the response regulator CtrA causes pleiotropic effects on gene Tan, M. H., Kozdon, J. B., Shen, X., Shapiro, L., and McAdams, H. H. (2010). An expression but does not affect growth phase regulation in Rhodobacter capsulatus. essential transcription factor, SciP, enhances robustness of Caulobacter cell cycle J. Bacteriol. 192, 2701–2710. doi: 10.1128/JB.00160-10 regulation. Proc. Natl. Acad. Sci. U. S. A. 107, 18985–18990. doi: 10.1073/pnas. 1014395107 Nabhani, A., Rushing, L., Newkirk, H., Burrowes, B., Young, R., and Gonzalez, C. (2021). Complete genome sequence of Rhizobium phaseoli Podophage Palo. Taylor, J. A., Ouimet, M.-C., Wargachuk, R., and Marczynski, G. T. (2011). The Microbiol. Resource Announc. 10:e01443-20. doi: 10.1128/MRA.01443-20 Caulobacter crescentus chromosome replication origin evolved two classes of weak DnaA binding sites. Mol. Microbiol. 82, 312–326. doi: 10.1111/j.1365-2958.2011.07785.x Nielsen, T. K., Carstens, A. B., Browne, P., Lametsch, R., Neve, H., Kot, W., et al. (2017). The first characterized phage against a member of the ecologically important Wang, H., Ziesche, L., Frank, O., Michael, V., Martin, M., Petersen, J., et al. (2014). sphingomonads reveals high dissimilarity against all other known phages. Sci. Rep. The CtrA phosphorelay integrates differentiation and communication in the marine 7:13566. doi: 10.1038/s41598-017-13911-1 alphaproteobacterium Dinoroseobacter shibae. BMC Genomics 15:130. doi: 10.1186/1471-2164-15-130 O’Neill, P. K., and Erill, I. (2016). Parametric bootstrapping for biological sequence motifs. BMC Bioinfo. 17:406. doi: 10.1186/s12859-016-1246-8 Williamson, S. J., McLaughlin, M. R., and Paul, J. H. (2001). Interaction of the PhiHSIC virus with its host: lysogeny or pseudolysogeny? Appl. Environ. Microbiol. Ouimet, M. C., and Marczynski, G. T. (2000). Analysis of a cell-cycle promoter 67, 1682–1688. doi: 10.1128/AEM.67.4.1682-1688.2001 bound by a response regulator. J. Mol. Biol. 302, 761–775. doi: 10.1006/jmbi.2000.4500 Wilson, K., and Ely, B. (2019). Analyses of four new Caulobacter Phicbkviruses Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., et al. indicate independent lineages. J. Gen. Virol. 100, 321–331. doi: 10.1099/jgv.0.001218 (2011). Scikit-learn: machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830. doi: 10.5555/1953048.2078195 Xue, S., and Biondi, E. G. (2019). Coordination of symbiosis and cell cycle functions in Sinorhizobium meliloti. Biochim Biophys Acta Gene Regul Mech 1862, Pfeifer, E., Moura de Sousa, J. A., Touchon, M., and Rocha, E. P. C. (2021). Bacteria 691–696. doi: 10.1016/j.bbagrm.2018.05.003 have numerous distinctive groups of phage-plasmids with conserved phage and variable plasmid gene repertoires. Nucleic Acids Res. 49, 2655–2673. doi: 10.1093/ Yang, H., Ma, Y., Wang, Y., Yang, H., Shen, W., and Chen, X. (2014). Transcription nar/gkab064 regulation mechanisms of bacteriophages: recent advances and future prospects. Bioengineered 5, 300–304. doi: 10.4161/bioe.32110 Phizicky, E. M., and Roberts, J. W. (1980). Kinetics of RecA protein-directed inactivation of repressors of phage lambda and phage P22. J. Mol. Biol. 139, 319–328. Zan, J., Heindl, J. E., Liu, Y., Fuqua, C., and Hill, R. T. (2013). The CckA-ChpT- doi: 10.1016/0022-2836(80)90133-3 CtrA Phosphorelay system is regulated by quorum sensing and controls Flagellar Frontiers in Microbiology 20 frontiersin.org Mascolo et al. 10.3389/fmicb.2022.918015 motility in the marine sponge Symbiont Ruegeria sp. KLH11. PLoS One 8:e66346. Zhao, Y., Qin, F., Zhang, R., Giovannoni, S. J., Zhang, Z., Sun, J., et al. (2019). doi: 10.1371/journal.pone.0066346 Pelagiphages in the Podoviridae family integrate into host genomes. Environ. Microbiol. 21, 1989–2001. doi: 10.1111/1462-2920.14487 Zhan, Y., Huang, S., and Chen, F. (2018). Genome sequences of five bacteriophages infecting the marine Roseobacter bacterium Ruegeria pomeroyi Zhao, Y., and Stormo, G. D. (2011). Quantitative analysis demonstrates most DSS-3. Microbiol Resour Announc 7, e00959–e00918. doi: 10.1128/MRA. transcription factors require only simple models of specificity. Nat. Biotechnol. 29, 00959-18 480–483. doi: 10.1038/nbt.1893 Frontiers in Microbiology 21 frontiersin.org