Human-Gut Phages Harbor Sporulation Genes

ABSTRACT Spore-forming bacteria are prevalent in mammalian guts and have implications for host health and nutrition. The production of dormant spores is thought to play an important role in the colonization, persistence, and transmission of these bacteria. Spore formation also modifies interactions among microorganisms such as infection by phages. Recent studies suggest that phages may counter dormancy-mediated defense through the expression of phage-carried sporulation genes during infection, which can alter the transitions between active and inactive states. By mining genomes and gut-derived metagenomes, we identified sporulation genes that are preferentially carried by phages that infect spore-forming bacteria. These included genes involved in chromosome partitioning, DNA damage repair, and cell wall-associated functions. In addition, phages contained homologs of sporulation-specific transcription factors, notably spo0A, the master regulator of sporulation, which could allow phages to control the complex genetic network responsible for spore development. Our findings suggest that phages could influence the formation of bacterial spores with implications for the health of the human gut microbiome, as well as bacterial communities in other environments.

Viruses of microbes, such as bacteriophages, play an important role in shaping gut microbiomes (1). Phage fitness is thought to be enhanced through the encoding of bacterial-like auxiliary metabolic genes (AMGs) that can reprogram and sustain host metabolism during infection (5). The acquisition of other, nonmetabolic genes may allow phages to alter other aspects of bacterial physiology (6). One of the most important determinants of phage fitness is the metabolic activity of the host cell (7,8). Rates of bacterial metabolism span orders of magnitude, ranging from exponential growth to being nearly inert when cells engage in certain types of dormancy, such as sporulation (9,10). By entering a state of reduced metabolic activity, microorganisms can defend themselves against phage attack (11,12), altering selection in ways that could modify coevolutionary dynamics.
Previous work has demonstrated that some phage genomes possess sporulation genes (13)(14)(15). In one example, homologs of sporulation-specific sigma factors (sigG and sigF) were identified in both lytic and lysogenic phages (13). These sigma factors are essential for the developmental transition of a vegetative cell to an endospore (16). When expressed in a host (Bacillus subtilis), the phage-carried sigma factors activate sporulation transcriptional pathways and depress spore yield by up to 99% (13). To date, there has not been any systematic analysis of the prevalence and distribution of sporulation genes in phages. Thus, it remains unknown whether modification of host sporulation is a common phage strategy. In this study, we search for homologs of sporulation genes in genomic and metagenomic data to determine whether phages employ this strategy in human gut microbiomes.
Identifying sporulation homologs in viral genomes and metagenomes. We identified sporulation genes in viral genomes and uncultivated viral genomes (UViGs) using DRAM-v (17) (see Fig. S1 and Text S1 in the supplemental material). Specifically, we targeted homologs of well-characterized sporulation genes found in B. subtilis and Clostridioides difficile. We reasoned that phage-carried genes can affect sporulation only if they are in phages that infect a spore-forming host. We therefore designed an enrichment test to identify homologs of sporulation genes that were preferentially found in phages that infect spore-forming hosts. We first evaluated our search strategy by looking for sporulation genes in genomes of phage isolates for which the host was known. Next, we applied the same approach to UViGs assembled from human-gut environments, and for which host predictions had been made in the original studies (18,19). To minimize the potential for contamination by bacterial sequences, we inspected the annotations of 6,542 gut-derived UViGs in which sporulation genes were detected (Fig. S2), with an average of 117 UViGs inspected per enriched sporulation gene (Fig. S3).
Phages possess nonregulatory genes required for sporulation. Our search identified 31 phage-carried homologs of sporulation genes (Table 1). These sporulation genes were enriched in phages that infect spore-forming hosts (Fig. 1). Many of the phage-carried homologs were nonregulatory genes involved in an assortment of sporulation-related processes such as chromosome partitioning, DNA damage repair, and cell wall-associated functions ( Table 1). The acquisition of these genes might allow phages to promote or impede specific steps of spore development or its eventual germination. For example, phages may use chromosome segregation genes to increase the probability of entrapment (and survival) of the phage genome in the spore during the asymmetric division separating the developing spore from the mother cell (20). Alternatively, it is possible that some of these genes are used by phages for functions other than sporulation. Chromosome segregation genes are known to be used by phages that establish extrachromosomal, plasmid-like lysogeny (21). Likewise, cell wall hydrolases used by the host to restructure the cell during sporulation (cwlJ, sleB, spoIID) could be repurposed as endolysins to burst the host cell at the completion of the phage lytic cycle (22). Further experimental investigation will be required to establish the phage functions of the sporulation gene homologs that we have catalogued in this work.
Phages also possess genes that regulate sporulation. Nearly one-third of the sporulation homologs (n = 9) identified in phage genomes and metagenomes are transcriptional regulators (Table 1). This finding is different than most examples of AMGs, where the phages control a metabolic process by phage-carried enzymes or by expression of modulators of host enzyme activity (5). It may be that phages manipulate host sporulation by interfering with the tightly regulated transcriptional program that is essential for this complex developmental process (2). Such findings are consistent with recent experimental findings regarding sporulation-specific sigma factors, where the ectopic expression of phage-carried sigG and sigF homologs altered the transcriptional program of B. subtilis, resulting in reduced spore yield (13). SigE Most notable among phage-carried regulators are homologs of spo0A, the master regulator of sporulation initiation that is conserved among all spore-forming bacteria (4). Interestingly, the homologs found in phages are truncated versions of spo0A that contain the DNA-binding effector domain, but not the receiver domain (Fig. 2). The latter is responsible for modifying the DNA-binding activity in response to environmental and physiological signals received via the phosphorelay signal-transduction system (23). The truncation suggests that phage-carried spo0A may not require the normal host signals to activate or repress the initiation of host sporulation (24). In addition to transcriptional regulators, phage genes included other potential posttranscriptional regulators (RNA binding spoVG, and translation-related genes cca and smpB). Taken together, the bioinformatic findings here and laboratory results (13) suggest some phages may overcome dormancy defenses by targeting the regulation of sporulation. Compared to the use of nonregulatory genes, this is likely to be a more efficient strategy for altering the course of a complex cellular program.
Phage-carried sporulation genes occur in diverse environments. The recovery of phage-carried sporulation genes is not restricted to the human gut. We identified sporulation genes in UViGs originating from diverse environments (Table S1). Of the 30 sporulation genes identified in gut-derived UViGs, we found 23 that also occur in phages from terrestrial and aquatic environments (Fig. S4). Thus, phage manipulation of sporulation may be a common phenomenon in environments where spore-forming bacteria are found.
Implications and future directions. Sporulation is an ancient, complex, and important trait that contributes to the persistence and transmission of beneficial and pathogenic members of the mammalian gut microbiome. While sporulation can reduce virus infection, our analysis supports the view that phages may use host-like genes to overcome this host defense mechanism (25). Specifically, our study provides genomic and metagenomic evidence that phages carry homologs of multiple sporulation genes, which may influence the transition of bacteria between active and dormant states in host-associated and environmental ecosystems. The evolutionary drivers and ecological consequences of phage-carried sporulation genes remain to be investigated (15). Our  (18,19) using DRAM-v. However, genes enriched among phages infecting spore-forming hosts (purple points) were mostly sporulation-related genes. The number of homologs detected for each gene is plotted on the x axis. For each of the genes, a hypergeometric enrichment test evaluated if it was found in phages that infect spore-forming hosts more than the random expectation given the number of phages infecting spore-forming hosts among all phages surveyed. The enrichment result [-log 10 (hypergeometric P value)] is plotted on the y axis. Purple points represent enriched genes with an adjusted P value of ,10 26 (horizontal dashed line) and a sample size of .30 (vertical dashed line). Representative names of B. subtilis sporulation genes are provided for genes that were enriched and of viral origin.
Human-Gut Phages Harbor Sporulation Genes mBio work demonstrates how partitioning phages by a specific host trait (e.g., sporulation) can be used to identify genes used by phages to influence the same host trait. Data availability. The data products of this study, including sequence data for sporulation genes identified in UViGs, are available at https://doi.org/10.5281/zenodo .7742410. All code used in this study is available at https://github.com/LennonLab/ spore_amg and in the Zenodo repository linked above.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. TEXT S1, DOCX file, 0.03 MB.  The alpha helices of the effector domain are indicated above the sequences. Sequence colors correspond to physicochemical properties of amino acids, using the Zappo coloring scheme. Information on functional and structural domains is from reference 26.