Prophage-encoded chitinase gene supports growth of its bacterial host isolated from deep-sea sediments

Abstract

Auxiliary metabolic genes encoded by bacteriophages can influence host metabolic function during infection. In temperate phages, auxiliary metabolic genes (AMGs) may increase host fitness when integrated as prophages into the host genome. However, little is known about the contribution of prophage-encoded AMGs to host metabolic properties. In this study, we examined a temperate bacteriophage, and its piezotolerant Pseudomonas sp. host obtained from sediment samples collected from the Kermadec Trench at ~10 000 m water depth. Both the phage and host were present throughout the sediment profiles from the surface to 30 cm into the sediment, covering large gradients of environmental conditions. The host and phage each carried one chitinase gene, which differed from each other, suggesting that chitin degradation plays a role in their substrate supply. We demonstrated that prophage-encoded chitinase supported host chitin degradation and growth in the presence of chitin. Furthermore, prophage induction dynamics were strongly substrate-dependent, suggesting that the host controls the lysis-lysogeny switch in response to the presence of chitin, thus optimizing the trade-off between the loss of cells from prophage induction and prophage enhancement of host performance. Overall, the results demonstrate prophage-encoded AMGs as collaborative goods for their hosts and emphasize the potential role of phage-host interactions in benthic biogeochemical cycling, as well as for the capability of deep-sea bacteria to efficiently adapt and thrive at a wide range of environmental conditions.

Introduction

Viruses that infect prokaryotes (phages) are key components of microbial communities in marine sediments, where they influence the mortality, diversity, and biogeochemical cycling of prokaryotes [1, 2]. Through lytic infection and subsequent cell lysis, phages promote the release of bioavailable organic matter, thereby contributing to the recycling of carbon and nutrients [3]. In a cross-oceanic study [4], the relative importance of phages in prokaryote mortality and carbon cycling was shown to increase with water depth as the flux of labile organic matter by sinking particles decrease. At water depths >1000 m, phages were estimated to account for ~90% of prokaryotic mortality in marine sediments, sustaining up to 35% of the bacterial carbon demand [4]. These early values may not be representative for the deep ocean in general [5, 6], but they reflect the potential importance of phages for benthic biogeochemical cycling.

In addition to the lysis-driven effects of phages on biogeochemical cycling, phage-encoded auxiliary metabolic genes (AMGs) influence the metabolic function of their host. AMGs are found in both lytic and temperate phages and they have the potential to be utilized at any stage of host infection. Some of the most studied examples (e.g. photosynthesis [7], sulfur oxidation [8], and nucleotide biosynthesis [9]) are AMGs in lytic phages, which contribute to or alter key host metabolic functions to optimize viral proliferation during infection. Although viral reprogramming of host metabolic machinery through AMGs during the lytic infection phase has been studied in several virus-host systems, the role of AMGs in temperate phages during the lysogenic phase is less known.

Temperate phages spend part of their life cycle as prophages lying dormant inside their hosts, and it has been suggested that their AMGs increase host fitness rather than that of the phage itself [10, 11]. This has led to the suggestion that phage life cycle strategies and host habitat select for the role and composition of AMGs in marine phages [12]. One example is AMGs encoding virulence factors (e.g. [10, 13, 14]), which have been experimentally demonstrated to increase host virulence. However, the majority of studies (e.g. [12, 15]) into the potential role of AMGs involved with metabolic function are entirely based on in silico identification and annotation of functional genes in prophage genomes, without experimental evidence to support gene function prediction and gene use. Therefore, studying AMG utilization in temperate phage-host systems experimentally is important for validating in silico predictions and obtaining insights into the function and regulation of AMGs and how they influence host fitness and ecology.

Although prophages represent a hereditary biosynthetic burden and a latent genetic bomb, it has been suggested that efficient dispersal of phages combined with mutualistic interactions between temperate phages and their hosts contributes to the stable coexistence of free phages, prophages, and hosts [16]. Furthermore, environmental conditions also seem to play an important role in the overall prevalence of lysogenized populations [17]. Little is known about the prophage induction frequency and regulation, especially in environmental systems. Known prophage induction signals have traditionally been associated with host DNA damage, activation of the SOS response, and poor metabolic state of the host cell [18–20]. More recently [21], quorum sensing signaling has been shown to regulate prophage induction in response to cell density changes, suggesting that bacterial hosts may take control of the lysis-lysogeny switch to promote host fitness at high cell densities, further emphasizing the potential of prophages to provide functional advantages to their host. Therefore, both the direct effects of prophage manipulation of host genetics and the indirect effects of viral-induced cell lysis in the marine environment may play an important role in catalyzing biogeochemical cycling on a global scale.

Hadal trenches (>6000 m depths) represent the deepest parts of the global ocean. These extreme environments have been shown to be hotspots of microbial activity driven by seismic-related mass-wasting and downslope funneling of organic matter, compared with the adjacent abyssal plain sediments (4000–6000 m depths), which are characterized by low input of organic material and low microbial activity [22–24]. Recently, analyses [25–27] of microbial metagenomes from hadal sediments documented a large diversity of viruses in these environments and found indications of a variety of AMGs potentially supporting host metabolic functions, suggesting that prophages influence biogeochemical cycling in these environments [21]. However, these studies are mostly based on sequence derived data, and experimental evidence of AMGs supporting host and/or phage performance in the deep sea is limited [28, 29]. In the current study, we isolated and characterized a temperate bacteriophage and its Pseudomonas sp. host obtained from sediment samples collected along the axis of the Kermadec Trench at a depth of ~10 000 m, as well as from an adjacent abyssal site. The phage encodes a chitinase gene, and we showed that the prophage significantly stimulated host performance in the presence of chitin, and provided evidence to suggest that the host was able to control the lysis-lysogeny switch. We hypothesize that prophage AMGs play important roles in the functional properties and adaptations of deep-sea bacteria to variable environmental conditions, thus influencing benthic biogeochemical cycling.

Materials and methods

Sampling and isolation

Kermadec Trench is ~1500 km long and ~ 60 km wide, with a maximum depth of 10 047 m, and is located in a relatively oligotrophic region of the southwest Pacific [30]. We sampled two hadal sites (K4:31°08.41′S, 176°48.48′W, 9300 m and K5:31°56.14′S, 177°17.48′W, 10010 m) in the trench and one reference site at the abyssal plain adjacent to the trench (K7:32°11.22′S, 176°33.66′W, 6080 m). Sediment cores were collected by subsampling in a box corer (50 × 50 cm).

Samples were retrieved and processed as previously described [31]. In brief, samples used for bacterial isolation were collected from sediment slurries prepared with virus-free bottom water prefiltered trough a 30 kD Vivaflow 200 polyethersulfone membrane (Sartorius, Göttingen, Germany) and addition of 50 mmol L⁻¹ tetrasodium pyrophosphate (Merck, Darmstadt, Germany). Triplicate samples of 1.8 ml were added sterilized molecular-grade glycerol (Sigma-Aldrich, CA, USA) to a final concentration of 15% before storage at −80°C until further use.

For bacterial isolation, 100 μl of the cryopreserved samples were spread on thiosulfate–citrate–bile salts–sucrose agar plates (Sigma-Aldrich, CA, USA), which selects for pseudomonads and vibrios, and incubated at room temperature for 3 days. A single colony morphotype appeared on all plates, and from each of the five stations and depths, ~2–3 colonies were picked and transferred to 96-well plates with Marine Broth (MB) made from 5 g Bacto Peptone (Gibco, ThermoFisher Scientific, MA, USA), 1 g yeast extract (Sigma-Aldrich, CA, USA), and 20 g sea salts (Sigma-Aldrich, CA, USA) in 1000 ml MilliQ water and autoclaved at 121°C for 20 min. A total of 96 isolates were cultured and purified three times by spreading on MB agar plates (1 L of MB amended with 15 g agar (Sigma-Aldrich, CA, USA)) incubated at room temperature, and single colonies were isolated (Table S1). Cryostocks in 15% glycerol final concentration were prepared for each isolate and stored at −80°C.

For phage isolation, 10 of the 96 bacterial isolates were selected for use in the double agar overlay method [32]. In brief, bacteria were grown overnight in MB before being diluted with MB to an OD₆₀₀ of 0.1. Bacterial culture (300 μl) was mixed with 100 μl of 0.22 μm filtered (Qmax syringe filter; Frisenette, Denmark) cryopreserved slurry from the same sample as the bacteria were isolated from, before adding it to 4 ml molten soft agar (MB amended with 0.4% agar, melted, and kept at 43°C). The mixture was poured onto MB agar plates to produce double agar overlay plates. The plates were incubated for 3 days at room temperature and inspected for clearing zones (plaques) as an indication of phage infection in the bacterial lawn [33].

Plaques were obtained on a bacterial isolate originating from station K4 inside the Kermadec Trench [31] at 2–4 cm depth in the sediment, and this susceptible host was subsequently named KT_2_4. This species was identified as a Pseudomonas sp. by subsequent genomic analyses (see below). Individual plaques were obtained using a sterile pipette tip, transferred to 2 ml SM buffer (5.8 g NaCl₂, 2 g MgSO₄•7H₂O, 7.5 g Tris–HCl (Trizma, pH 7.5;Sigma-Aldrich, CA, USA), and 0.1 g Gelatin (Sigma-Aldrich, CA, USA) in 1000 ml MilliQ water), and vortexed. The phage solution was then centrifuged (10 000 × g, 10 min) and 0.22 μm filtered (Qmax syringe filter, Frisenette, Denmark). The isolated plaques were subjected to three rounds of purification, plating a dilution series of the phage suspension using the double agar overlay method described as above. For each round a single plaque was purified and the final purified phage was named Pseudomonas phage KT1.

Susceptibility of all 96 environmental isolates to phage KT1 was tested using the small drop spot assay [34], where drops of purified phage were spotted onto lawns of the individual bacteria embedded in soft agar. In the case of growth inhibition in the spotted zone, the environmental isolate was considered susceptible to the phage. Prior to the isolation procedures, the sterility of the 0.02 μm filtered water used for the slurries was tested for bacterial and phage isolation. As no bacteria or phages were isolated from the seawater used for dilution, it was concluded that the isolates were derived from the sediment samples. The isolated bacteria (all Pseudomonas sp.; see below) were grown in MB or on MB agar, unless otherwise specified.

Phage and host characteristics

Phage buoyant density, morphology, and life cycle characteristics

Phage buoyant density was determined using density gradient centrifugation of the phage stock in a CsCl gradient (100 000 × g, 60 min, Beckman Optima Ultracentrifuge, CA, USA) and subsequent collection and quantification of infective phages in 45 density fractions to identify the density layer containing the most phages.

Pseudomonas phage KT1 morphology was observed using a JEM-1010 transmission electron microscope (Jeol, Tokyo, Japan) by mounting them on negatively stained copper grids as previously described [35].

Phage kinetic parameters were determined by one-step growth experiments to measure the latency time, burst size, and adsorption rate of the phage during infection [36]. Phage adsorption rate (K) was calculated from the decrease in unabsorbed phages over time according to the following equation: K = 2/3/B*log(P0/Pt), where B is the concentration of bacteria (cells per milliliter), P0 = plaque-forming unit (PFU) at time zero, and Pt = PFU in the supernatant (i.e. phages not adsorbed) at time t (min). The adsorption rate (K) is the velocity constant (ml/min) [33]. For latency times and burst sizes, samples for PFU were collected every 10 min for 70 min and quantified using the spot assay [34].

DNA extraction and sequencing of pseudomonas phage_KT1 phage and KT_2_4 host and environmental isolates

Six environmental bacterial isolates, including the WT host, were selected for full genome sequencing. Single colonies were transferred to MB and incubated overnight at room temperature before DNA extraction was performed using the Wizard Genomic DNA Purification Kit (Promega, WI, USA) following the manufacturer’s instructions. DNA was sequenced on an Illumina DNBseq platform by BGI (Shenzhen, China). Whole-genome preprocessing and assembly were performed using Geneious Prime (2023.1).

For whole genome sequencing of phage KT1, the phage was proliferated in liquid cultures using the wild-type of KT-2-4. In triplicates, 1 ml phage stock was added to 25 ml bacterial cultures at an OD₆₀₀ of 0.1 and incubated overnight at room temperature. The lysed cultures were centrifuged (10 000 × g, 20 min), and the supernatant were filtered (0.2 μm Sterivex, Millipore, MA, USA). The phage stocks were then concentrated using polyethylene glycol 8000 (PEG-8000, Sigma-Aldridge, CA, USA) as described in Castillo et al. [37]. The phage pellet was resuspended in 1 ml SM buffer without gelatin before DNA was extracted and sequenced as described above for the bacterial isolates. Genome annotations were done using DRAM (v0.1.2) [38] and when possible, checked for consensus with Prokka (v1.14.5) [39].

Phage and host phylogeny

Phylogenetic trees were constructed for the isolated bacterium Pseudomonas sp. and phage_KT1. The reference database for the bacterium was made from 113 Pseudomonas genomes downloaded from NCBI (MD, USA, October, 2023) and 10 Pseudomonas genomes from deep-sea or hadal environments [40–44]. Cellvibrio Japonicum Ueda107 was used as the outgroup. The genes 16S rRNA, rpoD, rpoB, and gyrB were extracted from each genome and the phylogenetic tree was constructed as described Gomila et al. [45]. In brief, the phylogenetic tree was constructed on the concatenated, individual gene alignments using mafft —auto (v7.520) [46], in the order: 16S rRNA, gyrB, rpoD, and rpoB. A maximum-likelihood tree was generated using IQ-TREE (v2.2.2.6) [45]and visualized using FigTree (v1.4.3). For the phage phylogeny, a full genome blast search was attempted with no results. Instead, a reference set of phage genomes were compiled from all Pseudomonas phages from marine sources (NCBI, MD, USA, October 2023) together with the top 20 best hits and hits originating from environmental sources from a BLASTp search of the terminase large subunit gene from phage_KT1. The phage sequences where downloaded, reannotated using DRAM [38] and the terminase large subunit gene was extracted from each genome or contig. Alignment and tree building were done as described above.

Distribution pf phage and host in environmental samples

Metagenomic sequencing and screening for Pseudomonas KT_2_4 sequences in environmental samples

To investigate the presence and relative abundance of the host organism, KT_2_4 in situ, a per-sample coverage was estimated using metagenomic samples generated from the original sediment samples. DNA extraction, library construction sequencing and raw read quality filtration is described in detail in Trouche et al. [47], (2023). The raw, quality checked reads were mapped onto the genome of KT_2_4 using Bowtie2 [48] and Samtools [49]. MetaBAT2 [50] was used to generate the per-sample coverage table (Table S2).

Quantification of phage spatial distribution in the sediment

To determine the in situ density of phages infective against the Pseudomonas sp. host in the environmental samples, a PFU assay was done using the original glycerol-amended sediment slurries diluted 10x and 100x. One hundred microliters of either dilution was plated as described above using the wild-type KT_2_4 and the double agar overlay method. PFUs were enumerated from the dilutions and used as a proxy for total abundance of infective phages in the sediment samples.

Pressure effects on host growth, phage viability, and phage-host interactions

To quantify the effects of hydrostatic pressure on the growth of the Pseudomonas sp. KT_2_4 isolate, bacterial culture was inoculated in 10% MB medium at a starting density of ~10⁶ cells ml⁻¹ before filling 6 ml exetainers in 10 replicates. Five replicates were incubated at atmospheric pressure (controls) or 1000 bar, respectively. At 1000 bar, the pressure corresponded to the hydrostatic pressure at 10000 m water depth. In parallel, the effects of pressure on phage_KT1 infectivity were determined at atmospheric pressure (control), 500 bar, and 1000 bar in five replicates of 3 ml exetainers containing (~10⁶ PFU ml⁻¹ in SM buffer).

Pressure incubations were conducted in rotating cylindrical stainless-steel tanks (100 mm in diameter and 400 mm in length; Dustec GmbH, Wismar, Germany) ([51]. The experiment was run for 72 h at 4°C, and samples for CFU and PFU counts were collected immediately before pressurization (T_start) and at the end (T_end) of the experiment. Control incubations were performed at atmospheric pressure in acrylic cylinders in the dark and rotated at 5 rpm on custom-made roller tables [51].

Infection experiments and quantification of lysogeny and host resistance

To examine phage infection characteristics and development of resistance, the wild-type KT_2_4 host was inoculated in triplicate in 100 ml MB medium at OD = 0.1 with 3 × 10⁸ PFU ml⁻¹ of phage_KT1 with control cultures without phages. Samples were collected every 2 h for 12 h to measure OD₆₀₀ and PFU. At the last time point, samples were spread on MB agar, and 100 single colonies were obtained and tested for susceptibility to Pseudomonas phage_KT1 by spot assay (Table S3).

Analysis of lysogenization by pseudomonas phage KT1 in experimental and environmental pseudomonas sp. isolates.

A subset of 83 of the isolates obtained from the infection experiment above (Table S3) was further screened for lysogenization with phage_KT1 phage analyzing for the presence of three phage_KT1 genes: A polymerase (ORF 7), a hypothetical protein (ORF 33), and a structural gene (ORF 2) (Fig. S1). DNA was extracted from all the individual isolates, and primers were designed for PCR amplification of the three phage genes (Table S4). If all three genes were present in the genomic DNA of an isolate, the bacterium was considered to be a lysogen.

This analysis was repeated on 96 isolates from the environmental samples to examine the presence and distribution of Pseudomonas phage_KT1 as a prophage in the environmental populations.

From these analyses, three different infection phenotypes of the host were identified: (i) a phage susceptible type, which did not contain the prophage (e.g. the wild-type strain KT_2_4 (WT)), (ii) a lysogenic phenotype (R+) resistant to phage infection because it carried the phage_KT1 as a prophage, and (iii) a resistant phenotype (R-) with no prophage, which had acquired resistance in some other way. From here on, the three infection phenotypes are referred to as WT, R+, and R-, respectively.

Genome annotation of Pseudomonas phage_KT1 revealed an AMG encoding a putative chitinase, which was annotated as a different gene than the chitinase gene found in the host genome. Therefore, a series of experiments was designed to test and quantify the extent to which this chitinase gene encoded a functional protein, its metabolic implications in the lysogens, as well as the general implications of phage resistance. For these experiments, we selected representative clones for each of the three infection phenotypes of the Pseudomonas sp. KT_2_4 (i.e. WT, R+ and R-). Activity, growth rate, and biofilm formation of these three strains were quantified under different substrate conditions, with and without chitin.

Effects of lysogeny on host cell performance

Chitinase activity

Chitinase activity of the three strains was quantified using both a fluorogenic assay in a 96-well plate with a chitin derivative as the substrate [52] and a plate assay with chitin-enriched agar. For the fluorogenic assay, the bacterial strains were grown to exponential phase (OD₆₀₀ = 0.5–0.55) in 1% MB medium enriched with 0.01% chitin (Sigma-Aldrich, CA, USA), and then incubated with 5 μM (final concentration) of the substrate methylumbelliferyl(MUF)-N-acetyl-β-D-glucosaminide (Sigma-Aldrich, CA, USA) in triplicate for 1 h using growth medium as a blank. Chitinase activity was calculated from the increase in fluorescence caused by the enzymatic degradation of the non-fluorescent MUF-N-acetyl-β-D-glucosaminide to the fluorescent MUF product during incubation, and quantified on a plate reader (FLOUstar OPTIMA, BMG Labtech GMbH, Ortenberg, Germany) at Ex_355nm and Em_460nm. For the plate assay, MB plates amended with 2% final concentration of chitin and spots of 10 μl of overnight cultures (OD₆₀₀ = 1.1) of the bacterial strains were added and incubated for 48 h at room temperature. As the embedded chitin was degraded by the chitinase secreted by the bacteria growing in the center for the plate, a clearing zone around the bacterial colony developed. Chitinase activity was measured as the diameter of the clearing zone.

Growth rate and net cell production

Growth rate and net cell production of the three strains were measured in the presence of chitin. Overnight cultures of the strains in 10% MB medium were inoculated into 50 ml 1% MB medium amended with 0.01% chitin in triplicate in 250 ml square bottles (Nunc, ThermoFisher Scientific, MD, USA) at a final concentration of ~1 × 10⁶ cells ml⁻¹. The cultures were incubated for 60 h at 16°C, every 6–10 h, 1 ml was removed for quantification of cell density, fixed with 2% final concentration of glutaraldehyde, and stored at −80°C. For quantification of cell density on a FACS Canto II flow cytometer (Becton-Dickinson, NJ, USA), samples were diluted in TE buffer, and the cells were stained with the fluorescent SYBR-Green stain for 5 min [31]. The 60 h experiments covered the entire exponential growth curve of the strains that entered the stationary phase after ~50 h of incubation. The growth rate was calculated from the exponential increase in cell density during the initial 36 h, and net cell production was defined as the maximum cell density reached after ~55–60 h of incubation.

Chitinase expression levels

Expression levels of the prophage-encoded chitinase gene in Pseudomonas sp. KT_2_4 (R+) was evaluated by RT–qPCR. Overnight cultures and Pseudomonas sp. KT_2_4 (R+) in 10% MB was inoculated into 60 ml 1% MB in 250 ml square bottles (Nunc) with and without 0.01% chitin supplementation. Pseudomonas sp. KT_2_4 (WT) was analyzed in parallel as a negative control. The experiment was done in triplicates, resulting in 12 bottles in total. The cultures were incubated for 5 h, allowing the bacteria to enter exponential stage of growth. Samples of 20 ml were taken from each replicate, followed by centrifugation at 2500 g for 10 min. The supernatant was discarded, and the bacterial pellet was used for total RNA extraction, using PureLink RNA Mini Kit (Invitrogen, MA, USA) following the manufacturer’s instructions. Before eluting RNA from the column, DNae treatment (Invitrogen, MA, USA) was performed twice. The final RNA was measured by Qubit™ RNA High Sensitivity (HS) method, according to manufacturer’s protocol (Invitrogen, MA, USA).

Reverse transcription of the RNA was performed using PrimeScript RT Master Mix (Takara Bio Inc., Tokyo, Japan) after adjusting the amount of total RNA to the same level in each reaction, following manufacturer’s protocol. The obtained cDNA was used to evaluate the expression levels of chitinase in the presence and absence of chitin for the two strains by quantitative PCR, using TB Green Premix Ex Taq II *(Takara Bio Inc, Tokyo, Japan). Two sets of primers targeting the chitinase gene of Pseudomonas phage_KT1 were designed: chi_01F (5’-TTACCGGACGTGCGAACTAT-3′) and chi_01R (5’-CCGTTGATGCGCTTAGTGAT-3′), chi_02F (5’-GTCGATCACTGAGCAGCAAC-3′), and chi_02R (5’-CACGCACATAACGCAATTGG-3′). Two housekeeping (HK) genes (recA and anr) that have been used in a previous study to normalize mRNA from Pseudomonas sp. strains in qRT-PCR were used [53]. Specific primers targeting the HK genes of Pseudomonas sp. KT_2_4 were designed as follows: recA_01F (5’-GAGGTTGTCGGTAGCGAAAC-3′) and recA_01R (5’-GATCTTGCTGCCGTTGTAGG-3′), anr_01F (5’-AGCAAATCACCGGTTTCCAC-3′), and anr_01R (5’-GGTTTTCTTCGACAGCAGCA-3′), respectively. The PCR mixture for each reaction was prepared as follows: TB Green Premix Ex Taq II 12.5 μl, Forward Primer (10 μM) 1 μl, Reverse Primer (10 μM) 1 μl, nuclease-free purified water 8 μl, RT cDNA mix (sample) 2.5 μl. The qPCR amplifications were performed in hard-shelled PCR low-profile, semi-skirted plates (BioRad, CA, USA) covered with an adhesive plate sealing film (BioRad, CA, USA) using a Bio-Rad CFX Connect (BioRad, CA, USA), following a 2-step PCR protocol including melting curve.

Biofilm formation

Biofilm-forming properties of the strains were quantified using the crystal violet method [54] and modified to 96-well plates. Twenty microliters of overnight cultures in MB were inoculated into 200 μl of fresh MB in eight replicates, and the plates were incubated for 24 h at room temperature. After staining with crystal violet, washing, and detachment of the stain using 96% ethanol, the absorbance was measured at Abs_595nm in the plate reader. A parallel experiment in which the WT strain was incubated with phage_KT1 for 24 h prior to biofilm measurement was performed to examine the role of phage infection in biofilm formation.

Effect of substrate composition on prophage induction

To examine the effect of substrate composition on prophage induction dynamics, changes in the induction of prophages from the lysogenic strain R(+) were quantified in response to medium strength and chitin availability in the medium. The strain was cultured in a concentration series of MB covering 0.1%, 1%, 10%, and 100% medium strength, with and without 0.01% chitin. Growth was monitored by CFU counts, and the induction of prophages was quantified by PFU as a proxy for the concentration of viable phage_KT1 obtained as the phage density leveled out when the culture reached the stationary phase.

Results

Characterization of the phage-host system

The phage isolate had a siphoviridae-like morphology with a buoyant density value of 1.44 g ml⁻¹ and a size of 140 nm (Fig. S2). Life cycle characteristics showed an adsorption rate of 9.712 x10⁻¹⁰ ± 1.28 *10⁻¹⁰ ml min⁻¹, a latent period of 40 min and a burst size of 4.5 ± 2.0 phages cell⁻¹ in MB medium at 20°C (Fig. S2).

The five environmental bacterial isolates and phage host KT_2_4 were identified as belonging to the genus Pseudomonas based on the 16S rRNA gene. To confirm their position within the Pseudomonas group, an additional multisequence alignment analysis was done. The five environmental isolates appeared to be the same strain of Pseudomonas sp. and were closely related, but not identical, to the host KT_2_4. All six Pseudomonas isolates from the present study were placed within a subgroup under Pseudomonas fluorescens [55]. The isolates in this subgroup were primarily associated with the soil and the rhizosphere. Furthermore, none of the Pseudomonas genome sequences in this study clustered closely with previously reported Pseudomonas strains found in deep-sea environments (Fig. S3). The isolated phage formed a sub-cluster with three other phages. The closest relative was an undescribed prophage found in P. fluorescens strain Ps 77 (GenBank: LCYB01000031). A temperate Marinobacter phage infecting algae-associated bacteria [56] and a metagenome-assembled viral bin identified as a Caudoviricetes (GenBank: BK029073) were the other two members in the subcluster (Fig. S4).

Ninety-six Pseudomonas isolates obtained from the environmental samples were screened for lysogenization with phage_KT1 phage based on the presence of three phage_KT1 genes (Table S1). Of these, the majority (95%) were PCR-positive for all three phage genes, and thus contained phage_KT_2_4 as a prophage and were resistant to re-infection. Of the remaining 5% (4 isolates) that were non-lysogenic, only two were susceptible to infection by Pseudomonas phage_KT1.

Genome sequencing of the original bacterial isolate KT_2_4, used as a host strain for isolation of the phage showed a 6 103 812 bp genome bacterium belonging to the Pseudomonas genus. The genome had 60% GC content and contained five putative prophage genomes and seven questionable CRISPR arrays.

Whole genome sequencing of phage_KT1 revealed a 40 631 bp genome with 52 open reading frames and a GC content of 57.7% (Fig. S1, Table S5).

The genomes of the host KT_2_4 and phage_KT1 each encoded a putative chitinase gene. The bacterial chitinase gene is 1065 bases long and translates to 354 amino acids. The chitinase gene was annotated as glycosyl hydrolase in the GH18 family, which is a group of chitinases catalyzing random endo-hydrolysis of the (1- > 4)-beta-linkages in chitin and chitodextin compounds. In the bacterial genome, the chitinase is preceded by lytic polysaccharide monooxygenase (LPMO), a group of proteins that oxidizes the cleavage of glycosidic bonds. Downstream of the chitinase a gene encoding a potential repressor similar to the GntR transcriptional regulators. Besides this, most of the surrounding genes of the chitinase are hypothetical proteins with unknown function. The putative chitinase gene sequence of phage_KT1 differs from that of the bacterial chitinase gene and is much smaller than its bacterial counterpart, with only 534 bases and 177 amino acids. The phage gene was annotated as a chitinase belonging to the GH19 family.

Distribution of the Pseudomonas KT_2_4 and infective phages in Kermadec Trench sediments

The in situ abundance of the phage infective units (Fig. 1A) showed the presence of the phage both inside the hadal trench (Sites 4 and 5) and in sediment from the abyssal plain (Site 7). Furthermore, viable phages were found at all the sampled depths (0–33 cm). At all sites, there was a gradual increase in viral density with sediment depth, from 163 ± 117 to 294 ± 1135 PFU ml sediment⁻¹ in surface sediment to a maximum of 1380 ± 543 PFU ml sediment⁻¹ and 1495 ± 470 PFU ml⁻¹ at 25–35 cm depth at hadal Sites 4 and 5, respectively. In sediment from the abyssal plain, phage abundance reached 2430 ± 1352 PFU ml sediment⁻¹ at a depth of 15–20 cm.

Figure 1

(A) Distribution of infective phages (PFU) in the sediments at the hadal Sites 4 and 5 and the abyssal Site 7. (B) Distribution of the relative abundance of Pseudomonas sp. KT_2_4 genomes in sediments obtained from the metagenomes of the samples from the hadal Site 6 and abyssal Site 7 horizontal lines indicate the oxygen penetration depths at the three sites (from Glud et al 2021 [22]).

In the metagenomic data, the relative coverage of Pseudomonas KT_2_4 suggested that it was most abundant in the surface sediment and decreased gradually with depth to 5–10 cm where some variation was observed between hadal and abyssal sediment at 15–30 cm depth (Fig. 1B).

Effects of pressure on phage and host stability and interactions

To investigate the effects of hydrostatic pressure, we exposed the phage and the bacterial host to different pressures, both individually and in combination. In the absence of the host, the infective phage density remained relatively constant for the duration of the experiment at different pressures (Fig. S5), varying from 1.2 × 10⁶ PFU ml⁻¹ at T = 0 to 2.1 × 10⁶ ± 6.2 × 10⁵ PFU ml⁻¹ at atmospheric pressure, 1.7 × 10⁶ ± 5.3 × 10⁵ PFU ml⁻¹ at 500 bar and 1.8 x10⁶ ± 4.1 × 10⁵ PFU ml⁻¹ at 1000 bar with no significant effects of pressure on phage infectivity after 68 h of incubation (P > .05, Fig. S5).

The bacterial incubations were performed at atmospheric pressure and 1000 bar pressure, and here there was also no significant (t-test, P > .05) effect of pressure on cell growth during the incubation as the cell density increased from 1.0 × 10⁶ ± 5.9 × 10⁵ CFU ml⁻¹ at T = 0 to ~1 × 10⁷ CFU ml⁻¹ at both atmospheric pressure and 1000 bar at the end of the incubation (Fig. S5).

The combined phage-host experiment was also conducted at atmospheric pressure and 1000 bar pressure, and showed similar results, with no significant effect (t-test, P > .05) of pressure on phage production. During the incubation, phage production increased from ~3 × 10³ phages ml⁻¹ at t = 0 to 1.0 and 0.8 × 10⁴ phages ml⁻¹ in the atmospheric and 1000 bar incubations, respectively, after 72 h (Fig. S6). Similarly, there was no significant effect on bacterial density, which increased from 2.4 × 10⁶ cells ml⁻¹ at t = 0 to 2.4 × 10⁷ ± 1.1 × 10⁷ bacteria ml⁻¹ and 2.1 × 10⁷ ± 1.2 × 10⁶ bacteria ml⁻¹ at atmospheric and 1000 bar pressure, respectively after 72 h incubation (Fig. S6). Overall, these experiments suggest that both host and phage are piezotolerant.

Phage-host interactions and selection for lysogenization

When the phage-sensitive wild type strain KT_2_4 was inoculated at OD = 0.1 with 3 × 10⁸ PFU ml⁻¹ of Pseudomonas phage_KT1, there was a strong controlling effect of the phage on bacterial growth relative to control cultures without phage during 12 h incubation. The control cultures grew exponentially and reached a stationary phase at an OD of 1.5, whereas the phage-exposed culture remained at an OD of ~0.35 until 8 h, and then increased to 0.63 after 12 h. The production of 1.4 × 10⁹ phages ml⁻¹ thus reduced the bacterial cell density to 43% of the control density (Fig. 2).

Figure 2

Growth of Pseudomonas sp. KT_2_4 WT host in the presence of Pseudomonas phage_KT1 and control cultures without the phage, along with phage production.

Of the 100 clones isolated from the phage-exposed cultures after 12 h, 61 clones were resistant to phage infection and the rest showed some variation in sensitivity to infection (Table S3). A subset of 83 clones was further tested for lysogenization, 55% of which were lysogenized, whereas 45% did not contain the prophage. Lysogenization provided total resistance against re-infection by Pseudomonas phage_KT1, whereas only one of the non-lysogenic isolates, KT_2_4 (R −) (Table S3), was resistant to phage infection. Overall, phage exposure was selected for a mixed population of primarily phage-resistant lysogens and phage-sensitive non-lys ogens.

Effects of lysogenization and resistance on host performance

Based on the different genotypes and phenotypes of isolates obtained from environmental samples as well as from culture experiments, we used three representative isolates for subsequent analysis of the effects of lysogeny and resistance on the performance of Pseudomonas sp. (Table S6), with emphasis on the potential role of the prophage-encoded chitinase gene.

Quantification of chitinase activity by two independent methods demonstrated significant increases in chitinase activity in Pseudomonas sp. KT_2_4 (R+) relative to the non-lysogenic WT control (Fig. 3). Using the fluorogenic approach with MUF-N-acetyl-β-D-glucosaminide assay, the lysogen showed a 52% increase in chitinase activity (P < .001), whereas chitinase activity quantified from the enzyme-driven clearing zone in a chitin-embedded agar, showed a 64% increase relative to the WT strain (P < .002) (Fig. 3A). In the chitin-containing growth medium, the increased chitinase activity coincided with a significantly higher growth rate (0.106 ± 0.003 h⁻¹) of the lysogenic strain than obtained for wild-type and phage-resistant strains (0.0744 ± 0.002 h⁻¹ and 0.071 ± 0.002 h⁻¹, respectively) (Fig. 3B; P < .002).

Figure 3

(A) Chitinase activity of the WT, R+, and R- strains measured using fluorogenic and plate assays. (B) Growth rate of the same strains during exponential growth in chitin-enriched medium in batch culture.

A comparison of net cell production during a 60-h experiment at 16°C for the three selected clones in media with and without 0.01% chitin supported the observed increase in the performance of the lysogenic strain KT_2_4 (R+) compared with the non-lysogenic strains (Fig. S7). In the presence of chitin, strain KT_2_4 (R+) produced 12% and 18% more cells than the non-lysogen KT_2_4 WT KT_2_4 (R-), respectively, confirming the benefit of the prophage-encoded chitinase in supporting host growth. The addition of chitin to the 1% MB medium showed that chitin significantly promoted net growth (P < .05) in all three strains, emphasizing that the chitinase detected in the core host genome also supported net cell growth in these incubations (Fig. S7). In the wild-type strain KT_2_4 WT, this corresponded to an 8% increase in net growth, whereas the 17% increase in growth of the lysogen in chitin-enriched medium again emphasized the importance of the prophage-encoded chitinase for chitin degradation in this Pseudomonas sp. (Fig. S7). The fact that the net production of the WT and lysogenic strain did not differ significantly in the absence of chitin suggests that there was no apparent fitness cost of having the prophage during growth in 1% MB medium. The lowest net production was observed in the non-lysogenic phage-resistant strain KT_2_4 (R-) in the absence of chitin, suggesting that resistance may impose a fitness cost in this medium compared with the phage-sensitive wild-type strain (Fig. S7) [57].

Prophage-encoded chitinase expression

Quantitative RT-PCR showed a 2.3 to 3.3-fold increase in prophage-encoded chitinase gene expression in lysogenic strain KT_2_4 (R+) in 1% MB supplemented with 0.01% chitin compared to the control condition of 1% MB without chitin supplementation (Fig. S8). The WT strain KT_2_4 did not show significant amplification of this gene. Specifically, the first set of primers for the chitinase gene generated a 2.7- and 3.3-fold increase in expression compared to the control condition without chitin when calibrated with recA and anr HK genes, respectively. The second set of primers confirmed the upregulation of chitinase in the presence of chitin by 2.3- and 2.7-fold, respectively, when normalized to recA and anr (Fig. S8).

Effects of phage and prophage on host biofilm formation

The influence of phage-host interactions on host biofilm formation was examined by quantifying the biofilm-forming properties of each individual strain as well as of the wild-type strain after exposure to the phage. The lysogenic strain KT_2_4 (R+) produced significantly (P < .05) more biofilm than the wild-type strain KT_2_4 (Fig. S9), and also exposure of the sensitive wild-type strain to the free phage promoted biofilm formation. The resistant non-lysogenic strain also produced less biofilm than the lysogenic strain, but the difference was not statistically significant (t-test, P > .05).

Phage-host dynamics and the spontaneous induction of prophages

Spontaneous induction of Pseudomonas Phage_KT1 prophages correlated with substrate concentration and composition. During growth in 1% MB, the net release of free phages from the lysogenic Pseudomonas sp. KT_2_4 (R+) was <1200 PFU ml⁻¹ during 12 h incubation in the presence of chitin, whereas the net release of free phages by induction reached 9000 PFU ml in the absence of chitin (Fig. 4A). Comparing the prophage induction across different media strengths showed a 10-fold decrease in the concentration of free phages from ~3.5 × 10⁵ PFU ml⁻¹ in 0.1% MB to ~3.5 × 10⁴ PFU ml⁻¹ in 100% MB medium after 6 h of incubation. This corresponded to a decrease in cell-specific prophage induction of ~4.0 x10⁴ PFU million cells⁻¹ in 0.1% MB to ~10 PFU million cells⁻¹ in 100% MB, with a gradual decrease in cell-specific induction with an increase in medium strength (Fig. 4B). Combining the effects of medium strength and chitin amendment further confirmed that the medium composition affected prophage induction, as indicated by the 44-fold decrease in prophage induction upon the addition of chitin in 0.1% MB (Fig. 4B). This effect of chitin amendment on prophage induction was reduced with increasing medium strength, and at high MB concentrations (>10%), chitin addition had no effect on spontaneous induction rates (Fig. 4B).

Figure 4

Prophage induction in Pseudomonas sp. KT_2_4 (R+) at various substrate conditions. (A) Release of free Pseudomonas phage_KT1 particles during 12 h growth of the lysogenic host in 1% MB medium in the presence and absence of 0.01% chitin, (B) relative prophage induction expressed as the release of free Pseudomonas phage_KT1 particles per million cells during exponential growth of lysogen at different medium strengths in the presence (black bars) and absence (white bars) of chitin.

Discussion

This study describes a widespread, piezotolerant Pseudomonas sp. phage-host system isolated from one of the deepest benthic sites in the global ocean and demonstrates a substrate-controlled interaction between the Pseudomonas host KT_2_4 and its Pseudomonas phage_KT1, which increases the fitness of both, suggesting a mutualistic co-existence of phage and host as an adaptation to the environment they occupy.

Characterization and distribution of pseudomonas KT_2_4 and pseudomonas phage_KT1

Pseudomonads are commonly found among culturable bacteria isolated on rich growth media from marine environments, reflecting their wide distribution and ability to respond quickly to nutrient input. Early reports of bacterial isolation from hadal trenches described a new species, Pseudomonas bathycetes, obtained from both Philippine Trench and Mariana Trench sediments at >9000 m depth [58]. More recently, several other species of Pseudomonas have been isolated from both sediments and deep waters of the Mariana Trench [40–43]. In addition, a temperate bacteriophage (vB_PstS-1) was identified upon induction by a Pseudomonas stutzeri 1–1-1b host isolated from a water sample collected at 7000 m depth in the Japan Trench [28]. However, the current isolate from the Kermadec Trench has little similarity to other deep-sea Pseudomonas isolates, which are also distributed broadly across the phylogenetic tree. Similarly, phage KT1 did not cluster with other deep-sea phages, suggesting a wide distribution of Pseudomonas phages across different habitats and spatial scales. The current isolation of Pseudomonas sp. and phage from Kermadec trench sediments is therefore in line with previous reports on the dominance of pseudomonads in culturable communities from deep-sea samples. The presence of both Pseudomonas KT_2_4 host and Pseudomonas phage_KT1 at all three sites, covering a horizontal distance of ~100 km, a difference in water depth from 6000 m to 10 000 m, and a vertical profile in the sediment from to 0–33 cm, suggests that this phage host system is well adapted to the different environmental conditions that prevail in these habitats [22, 57, 59]. This is also supported by the small genomic variability among the sequenced isolates of both phage and host, which suggests efficient dispersal of these phylotypes and/or limited selection for diversification, leading to a stable genomic composition of these microorganisms on a kilometer scale. Similarly, the large majority (98%) of lysogens among the environmental isolates suggested a high dispersal efficiency of the phage and co-existence with a relatively small sub-population of phage-sensitive strains that ensures re-infection and production of new phages.

The increase in infective phages with sediment depth at all stations suggested either increased prophage induction or perhaps a slower decay of infective phages in the deeper layers. The oxygen conditions at the three sites varied considerably, with oxygen penetration depths ranging from 4 cm at Site 4 to 21 cm at Site 7 [22], emphasizing that the Pseudomonas host and phage are produced and viable under both oxic and anoxic conditions. The majority of Pseudomonas species are facultative anaerobes and capable of using nitrite or nitrate as terminal electron acceptor [60], supporting the ubiquitous distribution of this phage-host system in hadal sediments and its tolerance to various environmental conditions. The observed densities ranging from 200 to 2600 PFU ml sediment⁻¹ demonstrate that Pseudomonas phage_KT1 only represents a minor fraction of the total viral community of 10⁸–10⁹ viruses ml⁻¹ [31], which is in line with the general finding of very high viral diversity in these environments [25].

Analysis of the metagenomic datasets from Sites 4 and 5 confirmed the presence of the Pseudomonas KT_2_4 host strain in all the samples down to >30 cm in the sediment. However, in contrast to the vertical distribution of infective phages, the relative coverage of the Pseudomonas host in the metagenomes suggested that the population was most abundant in the surface sediment and decreased gradually with depth to 5–10 cm. This apparent inverse relationship between the distribution of phage_KT1 and its host suggested a phage-driven reduction in host densities in the surface sediments, where the growth rates of Pseudomonas are expected to be highest. However, other factors are likely involved in shaping the vertical distribution of phage and host.

The wide distribution of viable phages and hosts in these deep-sea environmental conditions was also supported by the tolerance to pressurization by both Pseudomonas sp. and phage, as cultivability and infectivity, respectively, were unaffected by pressures of up to 1000 bar. Piezotolerant properties were also detected in P. stutzeri 1–1-1b from the Japan trench, which were able to grow at hydrostatic pressures ranging from 0.1–70 MPa [28].

The wide distribution of Pseudomonas phage_KT1 in and adjacent to the Kermadec Trench is in line with recent observations of the Halomonas phage vB_HmeY_H4907, isolated from surface sediment at 8900 m depth in the Mariana trench [29]. The mining of global viral metagenomic datasets for the presence of sequence homology to phage vB_HmeY_H4907 led to the discovery of a global distribution of the Halomonas phage covering bathypelagic, temperate, and tropical epipelagic and mesopelagic sampling sites. Our observations therefore support previous results and add to the increasing evidence that culturable Gram-negative Gammaproteobacteria, such as halobacteria and pseudomonads, are generalists that can grow and interact with their phages across the large temperature and pressure ranges from the surface ocean to the deep sea and from the aerobic water column to anoxic sediments.

The observed tolerance of both phage and host to exposure to the extreme hydrostatic pressure of 1000 bars demonstrated that this phage-host system can thrive under all pressure conditions in the ocean and emphasizes the versatile nature of pseudomonads. There are only a few studies on the pressure effects of phages and phage-host interactions, and the current results are, to our knowledge, the first to demonstrate phage infection and the production of progeny phages under deep-sea pressure conditions. Previous work [61] on surface water cyano- and roseophages exposed to long-term hydrostatic pressure (2 y, 30 MPa) showed variable effects on phage stability and infectivity, likely related to capsid size, volume, and shape [61]. Phages with larger icosahedral capsids were completely disrupted by pressure, whereas Siphoviridae phages with smaller icosahedral capsids (<60 nm) were relatively stable and maintained intact structures and infectivity [61]. Therefore, phages might require specific structural properties to maintain infectivity at hadal depths.

Effects of the prophage on growth and chitinase activity of Pseudomonas KT_2_4 (R+)

When the non-lysogenic wild-type strain KT_2_4 was exposed to phage infection, the majority of the phage-susceptible population was killed, driving a strong selection for resistant clones, most of which were lysogenized, and only a minority had become resistant by other mechanisms, probably mutational changes. Representatives of these three genotypes (KT_2_4 WT: sensitive non-lysogen, KT_2_4 (R+): resistant lysogen, and KT_2_4 (R-): resistant non-lysogen) showed significant differences in growth characteristics.

The experimental comparison of growth and chitinase activity of the non-lysogen strain KT_2_4 WT and the lysogenized derivative KT_2_4 (R+) showed systematic positive effects of harboring the prophage in the lysogens when grown on a chitin-amended medium. This was not only directly reflected in the 1.5- fold higher cell-specific chitinase activity in KT_2_4 (R+) relative to the non-lysogen KT_2_4 WT, but also translated into 1.4- and 1.1-fold higher growth rates and net cell accumulation, respectively, in batch cultures. A 2.2- to 3.3-fold increase in prophage-encoded chitinase expression during exponential growth in the presence of chitin relative to conditions without chitin confirmed that the prophage chitinase gene contributed to host chitinase production. The fact that the induction of prophage induction was almost eliminated in the presence of chitin during exponential growth, further confirmed that the observed upregulation of the prophage-encoded chitinase gene in chitin-supplemented MB was associated with its function as an auxiliary metabolic gene inside the host genome.

The presence of the prophage-encoded chitinase thus improved the host’s ability to degrade chitin and significantly boosted growth performance. Chitin is a major component of crustacean exoskeletons, one of the most abundant polysaccharides in the ocean, and an important source of carbon and nitrogen [62].Therefore, the capacity to utilize chitin is a major advantage for lysogens in chitin-rich environments. In contrast, the observed decrease in cell-specific chitinase activity with medium strength suggested that the expression of chitinase is regulated by the availability of other substrates and indicated that Pseudomonas KT_2_4 downregulated chitinase excretion if other more labile substrates are available. The data confirmed that these chitinases are indeed functionally hydrolytic enzymes used to acquire substrate for growth in a process that is strongly regulated by the presence of chitin and general substrate availability.

As expected, the phage-resistant non-lysogenic strain KT_2_4 (R-), derived from the sensitive wild type upon phage exposure, did not show any difference in chitinase activity or overall growth rate. However, the reduced net cell production relative to wild-type KT_2_4 suggested a fitness cost related to phage-resistance mutations, as previously observed in connection with phage resistance in marine bacteria [63].

In addition to increased chitinase activity, the lysogenic strain KT_2_4 (R+) also formed more biofilms than the non-lysogens KT_2_4 WT and KT_2_4 (R-). The positive effects of prophages on biofilm formation have been observed in several lysogens [13, 21, 64, 65] and are most likely related to the release of extracellular DNA and other lysate products released upon prophage induction, which contributes to biofilm stability and structure. The observed stimulation of biofilm formation by exposing the phage susceptible strain KT_2_4 WT to the phage has also been found in marine phage-Vibrio interactions and has been suggested as a phage defense mechanism, as the biofilm matrix offers protection against phage infection for bacteria embedded in the biofilm [13, 66].

Regulation of prophage induction

A recent demonstration that the lysis-lysogeny switch of some phages is regulated by host quorum sensing (QS) signaling (e.g. [21]) and a phage-encoded QS-like signal system [67] illustrated that prophage induction is modulated by a variety of environmental factors related to cell- and phage-density-dependent signal molecules.

A 4000-fold decrease in the prophage induction rate with increasing nutrient concentration from 0.1% to 100% MB medium supported the previous demonstration that high cell density favors lysogeny [21]. However, the observed influence of chitin amendment on prophage induction strongly indicated that substrate composition works as an additional environmental factor contributing to the regulation of the lysis-lysogeny switch by repressing prophage induction. Together, these observations imply that the interaction between the host and the prophage is regulated to optimize the utilization of the prophage-encoded chitinase under chitin-enriched conditions, whereas prophage release is favored under low cell density and low chitin conditions. The results also indicated that, at high nutrient concentrations and a decrease in the relative importance of chitin for supporting cell growth, the regulatory effect of chitin on prophage induction disappeared.

Overall, we suggest that prophage induction is regulated by various environmental conditions that select for either repression or activation of induction, allowing the coexistence of phages and hosts. For the host, the trade-offs between the benefit of prophage-driven chitinase activity and phage resistance on the one hand, and the cost of mortality by prophage induction (lysogen) and lytic infection (non-lysogen) on the other hand, thus seem to promote the co-existence of the lytic and lysogenic genotypes. From the perspective of temperate phages, the prophage-encoded enhancement of host fitness, protection of the phage genome inside the host genome, and occasional lytic infection securing renewal and dispersal of the phage population are also factors that contribute to explaining the phage-host co-existence and widespread occurrence in these deep-sea sediments.

Ecological implications

The accumulation of in silico-based evidence that phage-encoded AMGs play important roles as modulators of host metabolic functions [68, 69] emphasizes the importance of using model systems for to quantify phage-encoded contributions to bacterial biogeochemical cycling in marine environments [70]. We propose to distinguish between AMGs as being either collaborative or selfish AMGs. Selfish AMGs are primarily encoded by lytic phages, where the AMG confers a direct advantage to the phage during proliferation by supporting key cellular functions of the host to maximize viral production. This has been observed for the psbA and psbD genes that are activated during phage infection of the cyanobacteria Synechococcus sp. WH7803 [71]. In contrast, collaborative AMGs can be found in lysogenic phages and provide an advantage to both the host and phage. In this case, we demonstrated a prophage-encoded chitinase utilized by the host during cellular growth, which provided a metabolic advantage to the host and prophage, and potentially also improved cellular conditions for the prophage when entering the lytic phase. Increased biofilm formation in the lysogen (R+) has been observed in association with prophage induction in other Pseudomonas and Vibrio species [65], and demonstrates that the prophage also confers other traits to its host, which are beneficial during surface colonization. Furthermore, the inhibition of phage induction in the presence of chitin suggests that the host can manipulate the lifecycle of the prophage, adding to the collaborative properties of this AMG. The collaborative nature of the interaction likely allows the host to adapt to environmental conditions, and we propose that in chitin-enriched environments, such as amphipod carcasses and exoskeletons, there is a selection for lysogens that promote biofilm formation and chitin degradation. At low chitin concentrations, the induction rate increases, allowing lytic infection of the smaller WT subpopulation (~2% of the total population) (Fig. 5), which secures the recruitment of new phage particles and lysogenization of new host cells.

Figure 5

Conceptual model of the potential interaction between chitinase-encoding Pseudomonas phageKT1 and its Pseudomonas sp. host in deep-sea sediments in the presence of a chitin source such as an amphipod carcass. Prophage-containing lysogens (dark cells) comprise the majority of the Pseudomonas population and are favored upon colonization of the chitin skeleton. Consequently, lysogens dominate the colonizing community. Prophage induction further stabilizes bacterial biofilms and increases the chitin degradation efficiency. As the chitin source is being used and bacteria are released from the biofilm, the prophage induction rate increases, and free phages are released. A small subpopulation of phage-susceptible non-lysogens (light cells) is infected and lysed, thus maintaining the production of new phages, which ensures lysogenization of new hosts and dissemination of the AMG within the Pseudomonas population (made in BioRender).

Overall, the present description of a collaborative interaction between a prophage and its host, combined with previous indications of a dominance of temperate phages and a large representation of AMGs extracted from metagenomes of deep-sea benthic environments [25], suggests that prophage-encoded metabolic functional genes play an important role in host biogeochemical cycling. Recent findings of high genomic diversity in hadal sediments and indications of carbohydrate degradation capacity as an important driver in structuring the microbial communities [5], further emphasize the potential importance of phage-based dissemination of these traits. Our study provides experimental evidence that phage-bacteria interactions can affect bacterial metabolism through the regulation of prophage induction and integration, and we suggest that the variable organic matter composition in marine sediments selects for specialized and mutualistic phage-bacteria interactions and prophage-driven adaptations to variable environmental conditions. As this is a single-model system from a complex habitat, more studies are required to link the observed potential of prophage-encoded AMGs from metagenomics with an understanding of their quantitative role and regulation in benthic biogeochemical cycling.

Acknowledgements

We would like to thank the captain, crew, and scientific personnel of the RV Tangaroa (TAN1711). Anni Glud is acknowledged for her help with on board sample processing and work with pressure experiments.

Author contributions

MM, SJT, DC, and PGK conceived the idea, designed and performed the experiments and analyzed the data. MM and SJT wrote the paper with contributions from all authors. All authors contributed to the interpretation and discussion of the data, and all have approved the final version of the manuscript.

Conflicts of interest

The authors declare no competing interests.

Correspondence and requests for materials should be addressed to Mathias Middelboe.

Funding

Ship time partly funded by Coasts & Oceans Centre of New Zealand’s National Institute of Water & Atmospheric Research [NIWA], awarded to Ashley A. Rowden and R.N.G). The study was also financially supported by the HADES-ERC Advanced grant 669947 and the Danish National Research Foundation grant DNRF145 to establish the Danish Center for Hadal Research both awarded to R.N.G, and by the Independent Research Fund Denmark grant 10.46540/4283-00101B to M.M and S.J.T.

Data availability

The genome sequence raw data are deposited at DDBJ/ENA/GenBank with accession numbers PQ316073 (Bacteriophage_KT1), CP166962 (Pseudomonas KT_2_4 WT), JBJGCQ000000000 (Pseudomonas sp. KT_Env-2), JBJGCR000000000 (Pseudomonas sp. KT_Env-10), JBJGCS000000000 (Pseudomonas sp. KT_Env-17), and JBJGCT000000000(Pseudomonas sp. KT_Env-32).