crAssphage genomes identified in fecal samples of an adult and infants with evidence of positive genomic selective pressure within tail protein genes

Abstract

crAssphages are a broad group of diverse bacteriophages in the order Caudovirales that have been found to be highly abundant in the human gastrointestinal tract. Despite their high prevalence, we have an incomplete understanding of how crAssphages shape and respond to ecological and evolutionary dynamics in the gut. Here, we report genomes of crAssphages from feces of one South African woman and three infants. Across the complete genome sequences of the South African crAssphages described here, we identify particularly elevated positive selection in RNA polymerase and phage tail protein encoding genes, contrasted against purifying selection, genome-wide. We further validate these findings against a crAssphage genome from previous studies. Together, our results suggest hotspots of selection within crAssphage RNA polymerase and phage tail protein encoding genes are potentially mediated by interactions between crAssphages and their bacterial partners.

The human gut virome is dominated by bacteriophages (Dutilh et al., 2014; Minot et al., 2011; Reyes et al., 2010). In the last decade, a ~97kb circular DNA phage genome, commonly referred to as crAssphage (from cross assembly of 12 fecal metagenomes), was identified and found to be the most abundant virus in the human gut (Dutilh et al., 2014). crAssphage-like contigs and genomes have been de novo assembled from fecal samples from various regions of the world and across health and disease states, emphasizing their high prevalence among human enteric microbiota (Edwards et al., 2019). Morphologically, crAssphages appear similar to members of the family Podoviridae (Shkoporov et al., 2018; Yutin et al., 2018), yet sequence-based approaches suggest a novel family level grouping within the order Caudovirales (Yutin et al., 2018). Comparative genomic analyses of human-associated crAssphage taxa have partitioned them into four candidate subfamilies composed of ten putative genera (Guerin et al., 2018).

Despite the abundance of crAssphage-like contigs and genomes detected in human fecal microbiota, little is known about the lifestyle or selective pressures acting on crAssphage genomes. Based on analysis of CRISPR spacers and other genomic features, it was speculated that crAssphages prey upon members of Bacteroides (Dutilh et al., 2014). Intuitively, this fits well with the Bacteroidetes-dominated community profile of human gut microbiota. Recently, certain members of the crAssphage group (ΦcrAss001) have been shown to stably infect Bacteroides intestinalis and are able to maintain long-term persistence in vivo, though the mechanisms underlying this relationship are unknown (Shkoporov et al., 2018). Furthermore, after long-term (23 days) interaction experiments between Bacteroides phage, ΦcrAss001 (accession # MH675552), and B. intestinalis, approximately half of isolated B. intestinalis colonies demonstrated complete resistance to the phage, indicating rapid evolution of bacterial interaction factors (Shkoporov et al., 2018). However, it remains unclear how this relationship impacts selective pressures acting across the crAssphage genome.

Our knowledge of bacterial-phage coevolution in the human gut is largely unexplored (Minot et al., 2011; Scanlan, 2017). Antagonistic interactions between bacteria and phages are crucial determinants of genomic evolution for both partners, and several studies have described these trends across diverse systems and environments (Brockhurst et al., 2007; Gomez and Buckling, 2011; Martiny et al., 2014; Schwartz and Lindell, 2017). In the gut environment, Minot et al. (2013) described rapid nucleotide substitution rates (>10⁻⁵ per nucleotide per day) in the lytic phage family Microviridae. With the establishment of a bacterial host for crAssphage taxa, evolutionary studies of selection in these host cells are likely to reveal some of the genomic consequences of bacterial-phage relationships in the gut.

Stool (sample ID: M186D4) was collected from a 24-year-old woman living with HIV enrolled in the InFANT study (Tchakoute et al., 2018) at a periurban clinic in Cape Town, South Africa. The participant was six days postpartum, was living with HIV with a CD4 lymphocyte count of 265 cells/mm³, and had initiated antiretroviral therapy during pregnancy with tenofovir, emtricitabine and efavirenz. In addition to the adult sample, four samples were also collected from infants without HIV (samples IDs C0521BD4, C0521BW15, C0526BW15 and C0531BW4). One infant (C0521B) samples were obtained at two time points; 4 days and 15 weeks of age. The samples were transported on ice and stored at −80°C until nucleic acid extraction. After defrosting, approximately 0.5g of fecal samples were homogenized in 20ml SM buffer as previously described (Kraberger et al., 2018) and centrifuged at 10,000 × g for 10 min. The resulting supernatant was filtered sequentially through 0.45μm and 0.2μm Minisart (Sartorius AG, Germany) syringe filters. Filtered supernatants were incubated with lysozyme and Turbo DNAse (Thermo Fisher Scientific, USA) at 37°C for 1 hour to degrade nucleic acids not enclosed in virus particles.

Viral nucleic acid was extracted from 200μl of the filtrate using the High Pure viral nucleic acid kit (Roche Diagnostics, USA) following the standard protocol. Circular viral DNA was amplified using rolling circle amplification (RCA) with Illustra TempliPhi 100 amplification kit (GE Healthcare, USA). The RCA products were used to prepare a 250 bp insert DNA library for each sample following the manufacturer’s standard protocol. Shotgun sequencing was performed on an Illumina HiSeq 2500 platform using 150bp paired end chemistry at Novogene (Hong Kong), or on an Illumina Novaseq6000 platform at North West Genomics Center (Seattle, WA USA).

Raw sequencing reads were trimmed using Trimmomatic v0.36 (Bolger et al., 2014) and then de novo assembled using SPAdes v 3.11.1 (Nurk et al., 2017). All de novo assembled contigs of >500 nts were aligned and annotated using a BLASTx search (Altschul and Lipman, 1990) against NCBI’s viral RefSeq protein database. Annotations were only applied if the E-value of the top hit was ≤ 1e⁻⁵. Open reading frames (ORFs) were identified using Glimmer (Delcher et al., 1999; Kearse et al., 2012) and annotated using an in-house reference protein database compiled from NCBI’s GenBank resource.

Sequence similarity searches of de novo assembled contigs resulted in the identification of various prokaryote and eukaryote-infecting viruses. In each of the fecal samples, we identified complete crAssphage genomes (Figure 1; 97,757 – 99,121 nts) that are most closely related to uncultured crAssphage (GenBank accession #BK010471), CrAssphage apr34_000142F (MK415399), CrAssphage ES_ALL_000190F (MK415403), CrAssphage FA1–2_000172F (MK415404), CrAssphage GF1–2_000079F (MK415408), CrAssphage LMMB (MT006214), CrAssphage sp. O-152 (MK069403) and CrAssphage YS1–2_2437 (MK415410). The five South African crAssphage genomes (referred by the patient fecal sample ID M186D4, C0521BD4, C0521BW15, C0526BW15 and C0531BW4) were fully closed with 6–420-fold median coverage (Figure 1A). We identified and annotated >80 ORFs (Figure 1C) and protein coding sequences. Of the 10 candidate genera proposed by Guerin et al (2018), the crAssphages in Figure 1 belonged in candidate genus I, with a genome-wide GC content of 29%. The two crAssphages from the same infant at day 4 (crAssphage C0521BD4, MW067000) and week 15 (crAssphage C0521BW15, MW067001) are 100% identical likely suggesting a dominant variant or temporal stability of the crAssphage. In general, the five South African crAssphages (Figure 1) all are more closely related to each other, sharing >93% genome-wide pairwise identity, versus 83–91% identity to the other eight most closely related crAsspages from Japan and Mexico, as well as the composite assembly of uncultured crAssphage (BK010471). Thus, it seems that the 13 crAssphages listed in Figure 1 form a distinct group of closely related phages.

Figure 1. Summary of crAssphage genomes identified in this study (in bold) as well as those that are most closely related.

A. Details of the read coverage and mapping for the five South African crAssphages identified in study and a pairwise identity comparison of the most closely related crAssphage genomes. B. An unrooted Maximum Likelihood phylogenetic tree inferred using PHYML 3.0 (Criscuolo, 2011) with substitution model GTR+G+I. Bootstrapped (1000 iterations) branch support is provided at each branch. C. Illustration of the crAssphages genome organization. Genes /coding regions discusses further in the manuscript are color codes. Blastn based similarity between genomes was determined using EasyFig (Sullivan et al., 2011) are shown.

For variant analysis, trimmed paired end reads were mapped to the genomes of the five South Africa crAssphages using the BWA v0.7.15 with default parameters (Li and Durbin, 2009). Samtools (v0.1.19) mpileup utility (Li et al., 2009) was used to calculate the per-nucleotide read coverage and variant frequency. Variants were detected and filtered using VarScan (Koboldt et al., 2009). Variants were only investigated further if there were, at minimum, five high quality reads supporting the variant, which was the lowest threshold for significance in our dataset utilizing an α value of 0.05. P values were calculated using a Fisher’s Exact test on read counts supporting the reference and variant alleles using VarScan (Table S1). Functional effects of identified variants were predicted using SnpEff (Cingolani et al., 2012) against a custom annotation database generated as part of this study and annotated as nonsynonymous (N_d) or synonymous (S_d) substitutions. The total number of nonsynonymous (N) and synonymous (S) sites for each protein coding sequence was calculated using SnpGenie (Nelson et al., 2015). The numbers of nonsynonymous (d_N) and synonymous (d_S) substitutions per site were estimated using the Jukes-Cantor formula:

$$dN=\frac{3}{4}\ln \left(1-\frac{4p_N}{3}\right)dS=\frac{3}{4}\ln \left(1-\frac{4p_S}{3}\right)$$

where p_N and p_S indicate the proportions of nonsynonymous and synonymous substitutions, respectively, and can be estimated by $p_N=\frac{N_d}{N}$ and $p_S=\frac{S_d}{S}$.

Across the genome assembled from the adult woman, crAssphage M186D4 (MK238400), we detected 40 significant single nucleotide polymorphisms (SNP) and one insertion. After filtering, 40 SNPs were retained, of which 37 fell within coding regions (Table 1), leading to a variation rate of 1 variant per 2,443 nucleotides. The nucleotide substitution profile was composed of 33 transitions and 7 transversions, for a transition to transversion ratio of 4.71. Of the 37 variants that occurred within coding regions, 30 (78.9%) resulted in nonsynonymous substitutions (N_d), seven (18.4%) resulted in synonymous (S_d) mutations, and only one (2.6%) resulted in a nonsense mutation. The only insertion fell within a coding sequence for a tail tubular protein P22 (Table 1).

Table 1.

Functional annotation of variants in crAssphages M186D4, C0521BW15, and C0526BW15.

Genome	Gene	Start	Stop	Length	Direction	Nd	Nonsense mutations	Sd	Insertion	N	S
M186D4	putative ssb single stranded DNA-binding protein	388	1215	827	forward	3	0	1	0	631.5	193.5
M186D4	hypothetical protein	1381	1725	345	forward	2	0	3	0	267.8333333	74.16666667
M186D4	putative SWI2/SNF2 ATPase 252C non-canonical Walker A motif	19772	20590	819	forward	1	0	0	0	645	171
M186D4	putative deoxynucleoside monophosphate kinase	20936	21811	876	forward	1	0	0	0	684	189
M186D4	coil containing protein	22045	22827	783	forward	0	0	1	0	619.6666667	160.3333333
M186D4	putative RNAP catalytic subunit fused to unknown RNAP subunit	29489	41764	12276	reverse	5	0	2	0	8090.833333	2295.166667
M186D4	putative RNAP associated protein fused to zincinprotease	41832	47720	5889	reverse	4	0	0	0	3900.833333	1085.166667
M186D4	putative tail tubular protein P22 gp4	58024	59040	1016	reverse	0	0	1	1	683.6666667	174.3333333
M186D4	putative phage tail fiber protein (DUF3751)	61309	62769	1461	reverse	7	0	0	0	984.3	260.67
M186D4	putative tail protein	65848	66618	770	reverse	2	0	0	0	498.3333333	137.6666667
M186D4	putative Bacon (Bacteroidetes-Associated Carbohydrate-binding) domain containing protein	69585	73469	3885	reverse	2	0	0	0	2600.5	780.5
M186D4	putative tail protein	73473	74294	821	reverse	1	0	0	0	528.6666667	155.3333333
M186D4	putative tail fiber protein	74320	76413	2093	reverse	2	0	0	0	1367.833333	402.1666667
M186D4	putative plasmid replication initiation protein RepL	96778	97632	855	forward	0	1	0	0	678.1666667	173.8333333
C0521BW15	putative phage tail fiber protein (DUF3751)	62256	63716	1460	reverse	7	0	0	0	1081.666667	322.3333333
C0526BW15	hypothetical protein	84080	85237	1157	reverse	1	0	0	0	662.6666667	201.3333333
C0526BW15	putative phage tail fiber protein (DUF3751)	63302	64762	1460	reverse	12	0	0	0	905.1666667	246.8333333
C0526BW15	putative tail protein	75466	76260	794	reverse	1	0	0	0	492.8333333	128.1666667
C0526BW15	putative tail protein	78409	81015	2606	reverse	1	0	0	0	1599.666667	461.3333333

N_d, Nonsynonymous mutations; S_d, synonymous mutations; N, nonsynonymous sites; S, synonymous sites

The distribution of SNPs across the genome of crAssphage M186D4 (MK238400) was nonuniform. An accumulation of mutations was apparent in two distinct genomic regions (Figure 2). These occurred in RNA polymerase genes that lie within genomic regions from ~30,000–48,000 nts and phage tail proteins in genomic regions from ~58,000–76,000 nts. Averaged across the genome, the d_N/d_S ratio was 0.54. When considering only RNA polymerase genes, the d_N/d_S ratio was 1.89. In phage tail proteins, d_N/d_S ratio was even further elevated, at 2.66. The gene with the highest SNP count was a putative phage tail-collar protein (DUF3751), which had a total of seven SNPs (Table 1). The variant with the highest frequency across the dataset was a T → C transition occurring within a putative Bacteroides-associated carbohydrate-binding often N-terminal (BACON) domain containing protein. When removing genes encoding RNA polymerase subunits and phage tail proteins from consideration, the genome-wide d_N/d_S ratio fell to 0.34.

Figure 2. Selective pressures are variable across the genome of several South African crAssphages.

A. The distribution of variants per kilobase across the genomes of three South African crAssphages. The region encoding RNA polymerase genes is colored blue. The region encoding phage tail proteins is colored red. B. The fold coverage of quality filtered reads across the genomes of each crAssphage taxon. CrAssphages M186D4 (MK238400) in grey lines, C0521BW15 (MW067001) in orange lines, C0526BW15 (MW067002) in blue lines.

To assess whether selective pressures were elevated in these genes across distinct crAssphage taxa, we performed variant annotation and analysis of the original crAssphage genome sequence (referenced here as uncultured crAssphage (BK010471); Figure S1, Table S2), as well as the two additional novel genomes generated as part of this study with sufficient high-quality reads of >50x median coverage. Many publicly available complete crAssphage genome sequences have been cross assembled from multiple individuals or represent pooled samples (Dutilh et al., 2014; Reyes et al., 2010; Shkoporov et al., 2018), which may introduce significant sequence variation and confound variant analysis. However, the individual datasets from which uncultured crAssphage (BK010471) was cross assembled were primarily composed of related mother-child pairs with low interpersonal viral diversity, thus representing an attractive option for cross validation (Reyes et al., 2010). Raw nucleotide sequences were downloaded from the NCBI Sequence Read Archive (SRA) from Study SRP002523. The raw reads from SRA sample SRR073438 were selected for variant analysis because that sample was collected exclusively from a related mother-twin triad. Genes and annotations were transferred from the associated published genome (GenBank accession #BK010471) using Prokka (Seemann, 2014). As observed for crAssphage M186D4 (MK238400), we detected a nonuniform distribution of SNPs across the genome of uncultured crAssphage (BK010471), despite even sequencing depth (Figure S1). Consistent with observations from crAssphage M186D4, SNPs persistently occurred in genes annotated as RNA polymerase and tail proteins, often resulting in nonsynonymous substitutions (Table S2). For uncultured crAssphage (BK010471), the genes with the highest SNP counts were a putative tail fiber protein and putative tail protein UGP073, both with three nonsynonymous substitutions (N_d) and zero synonymous substitutions (S_d, Table S2). Because all synonymous substitutions (S_d) occurring at significant levels were distributed across protein coding genes other than RNA polymerase and tail protein genes, we were not able to calculate d_S values for those genes. However, average d_N values were similar to those observed in tail protein genes in crAssphage M186D4 (MK238400) (uncultured crAssphage (BK010471): 0.0017, crAssphage M186D4 (MK238400): 0.0029). Similarly, for RNA polymerase genes, d_N values were comparable between both genomes (uncultured crAssphage (BK010471): 0.0002, crAssphage M186D4 (MK238400): 0.0008). We performed variant analysis on two of the additional novel taxa identified in this study which both closed with >50x coverage. Uncultured crAssphages C0521BW15 (MW067001) and C0526BW15 (MW067002) displayed comparable d_N values in tail protein genes (Figure 2; Table S2).

As an ecological niche, the gut environment imposes a dynamic range of selective pressures on the genomes of gut microbiota. Here, we assembled and analyzed five novel circularized full genomes of human crAssphages with >90% sequence similarity to that of the first described crAssphage genome (Dutilh et al., 2014). We identified differential selective pressures along these genomes, with hotspots of selection targeting groups of genes. We present evidence of substantial genetic heterogeneity and dynamic selective pressure across the genome of crAssphages M186D4 (MK238400), C0521BW15 (MW067001), C0526BW15 (MW067002), and uncultured crAssphage (BK010471). Many nucleotide polymorphisms occurred in protein coding regions and yielded a high ratio of substitution rates at nonsynonymous to synonymous sites (d_N/d_S) differentially across the genome, specifically implying strong positive selection at RNA polymerase and phage tail protein genes, and purifying selection genome wide.

The genomic selective pressures that we detected were primarily focused on two regions undergoing relatively elevated positive selection. These regions, from 30–48kb and 58–76kb, are comprised of genes from various clusters of orthologous groups (COGs), but the variation was localized into genes annotated as RNA polymerase subunits and phage tail proteins, respectively. Phage tail proteins are essential in mediating bacterial cell attachment and genome delivery. The putative protein with the highest SNP count and most variants is the putative phage tail fiber protein (DUF3751). This protein, while functionally uncharacterized, has been shown to mediate antagonistic interactions with bacterial taxa and is a conserved element in phage-derived bacterial tailocin complexes (Ghequire et al., 2015). Tailocins are bacterial protein complexes co-opted from bacteriophage that are morphologically and functionally similar to phage tail proteins and are critical to eukaryotic and bacterial cell binding (Ghequire and De Mot, 2015; Hockett et al., 2015). Yutin et al. predicted that crAssphages likely possessed a Podoviridae-like morphology based on protein sequence analysis (Yutin et al., 2018), and Shkoporov et al. confirmed this for the crass-like phage, ΦcrAss001, using imaging approaches (Shkoporov et al., 2018). However, most of the genomic variability between ΦcrAss001 and other crAssphages exists in the region encoding for tail proteins (Shkoporov et al., 2018), making direct comparison difficult. As such, it remains unclear if the putative DUF3751 tail fiber protein discussed here is involved in structural or host receptor recognition, attachment or adsorption functions. Additionally, we detected a high frequency variant in the BACON domain containing protein. de Jonge et al. recently showed that in crAssphage taxa where the BACON domain gene is proximal to tail protein genes, there is elevated evolutionary expansion of the crAss-BACON ORF, suggesting a potential role in Bacteroides cell surface binding (Jonge et al., 2019). Though it is unclear how this domain physically interacts with bacterial cell receptors, positive selection along this gene and tail proteins may represent a Red Queen scenario where the phage tail protein is adapting to constantly evolving target cell surface proteins. It is well known that coevolution with bacterial hosts directs phage genomic evolution and selective pressures. Adaptive evolution of phage tail proteins has been well documented in marine bacteriophage communities (Angly et al., 2009) and has been shown to facilitate an expanded host range. Our analysis of geographically distinct crAssphage variants, here, as well as recent work by Siranosian et al. suggest this phenomenon may also occur in the human gut (Siranosian et al., 2020). Additionally, long-term (23 day) phage/B. intestinalis co-cultivation experiments have shown that founder strains of crAssphage have severely limited ability to infect bacterial strains passaged for the length of the experiment (Shkoporov et al., 2018) indicating rapid, likely antagonistic, coevolution between crAssphage and bacterial host strains. This reduction in infection rate likely reflects nonsynonymous changes in bacterial cell surface receptors that are mediated by interaction with crAssphage tail proteins.

In summary, our results suggest genome-wide purifying selection in South African crAssphages, with episodic shifts of strong positive selection within RNA polymerase and phage tail protein genes. Elevated selective pressure of phage tail proteins may be due to antagonistic coevolution between crAssphage and bacterial targets, though further work is needed to demonstrate this conclusively.

Highlights:

• Five complete crAssphage genomes from feces of South African woman and unrelated infants.

• Elevated positive selection in RNA polymerase and phage tail protein genes.

• Hotspots of selection may be mediated by selection pressure from bacterial partners.

Data availability:

The annotated genome sequences for crAssphages M186D4, C0521BD4, C0521BW15, C0526BW15, C0531BW4, are available under GenBank accession numbers MK238400, MW067000, MW067001, MW067002, and MW067003, respectively. Raw sequencing reads have been deposited in the NCBI SRA under BioProject PRJNA526942. The data, functions, and R script are available at https://github.com/itsmisterbrown/crAssphage_M186D4_analyses.