Of Terms in Biology: Phage–Plasmid (1|2)

by Christoph
 

Once you think again about horizontal gene transfer (HGT) and how genomic DNA can pass from one cell to another, you may soon feel quite medieval and ruminate «mille viae ducunt hominem per saecula Romam» (Alain de Lille (1128–1203) in his Liber Parabolarum). One of the main routes of chromosomal and extra‑chromosomal DNA (to Rome) is via transduction by bacteriophages, another via conjugation by plasmids, but there are other main roads too (see the recent post by Mechas and Roberto).
 

In the ancient Roman road network, there are secondary roads and the main roads sometimes intersect, of course, or branch out (for example, the Via Popilia to Messina branched of from the Via Appia from Rome to Brindisi in Capua). When it comes to HGT, Phage–Plasmids (P–Ps) are such an intersection: they transfer horizontally between cells as viruses and vertically within cellular lineages as plasmids (note that "Gene Transfer Agents" (GTAs) are at a different intersection, as Mechas explained here in STC). And P–Ps are "shape shifters." Pfeifer & Rocha (2024) state in the abstract of their paper: "For example, gene loss turns P1-like phage–plasmids into integrative prophages or into plasmids (that are no longer phages). Remarkably, some of the latter have acquired conjugation-related functions to became mobilisable by conjugation."
 

(click to enlarge)
Figure 1a. Comparisons between selected replicons of the well-defined P1 (n=70) and N15 (n=44) groups plotted using genoplotR. Similarity between co-oriented bi-directional best hits (BBH) is shown in red and between anti-oriented ones in blue. Colour intensity reflects the degree of gene similarity. The values of wGRR are shown between the pairs of elements. Source. Frontispiece: Detail from the Tabula Peutin­geriana 1–4th century CE. Facsimile Ed. by Conradi Millieri, 1887/1888. Source

The authors can state this in general terms with some certainty, as they had extracted from various databases a total of ~2,500 phages, ~12,000 plasmids and 780 phage–plasmids (P–Ps) across diverse bacterial phyla in a previous study. Using a sophisticated strategy (too com­plicated to give even a brief outline here), they were able to sort these data sets into several larger groups, and these into subgroups, based on common sequence characteristics. I will now pick out just two groups, the phage P1 group and the phage N15 group, to introduce them in some more detail, because P1 and N15 have long been known to be phage–plasmids.
 

Phage P1  Even before horizontal gene transfer (HGT) became a generally accepted term, E. coli geneticists made ample use of it in the lab: they would never had Barbara Bachman's Linkage map of Escherichia coli K-12, edition 7 (1983) without the many Hfr crossing and P1 transduction experiments performed to map several hundred of the ~4,300 genes of this bacterium. "Hfr crossing" is, in principle, HGT via a conjugative plasmid, and "P1 transduction" is HGT via a bacteriophage. Two main roads (to Rome) were taken! We used P1vir phages, which cannot establish as prophages in hosts (thus forming lysogens) and during their lytic develop­ment and, being a bit sloppy, pack a tiny percentage of chromosome fragments of ~95 kb size into phage heads instead of phage DNA (93.6 kb). In his more recent post Transient Thoughts on Transduction, Roberto went into details and some fascinating intricacies of this jewel of E. coli genetics.
 

Not P1vir, but wild-type P1 phages can establish themselves as prophages in host cells. The P1 prophage replicates as a circular plasmid from one of its two replication origins, oriR, which has the typical architecture of plasmid origins with the repA gene flanked on either side by 19-base-pair repeats, 5 on one side and 9 on the other). The second origin of replication, oriL, within the repL gene that is activated during the lytic cycle. Replication begins by a regular bidirectional θ replication from oriL but later in the lytic phase it switches to σ replication using the host recombination machinery to produce packageable genome concatemers.
 

In their collected data, Pfeifer et al. (2021) found 70 sequences that, divided into two subgroups, make up the P1 group. The exemplary align­ments of a few selected group members in Figure 1a (top part) shows that the members of this group have phage-typical, related genes for virion structure and assembly, partition genes as they are the rule for low-copy number plasmids, and plasmid-like replication genes, but most are quite jumbled up in genomic position and orientation in the individual genomes (except for the more closely related genomes).
 

(click to enlarge)
Figure 1b. Pangenome graph of the N15 group. The node colours indicate the phage–plasmid quotient (PPQ) scores (red for phage- and green for plasmid-association) that are computed from the average number of matches of the gene family with phage and plasmid genomes. The edges indicate contiguity between two genes in the P–P and their thickness indicates the frequency of this contiguity. For clarity, we removed the edges when the neighborhood was rare: for N15 < 25%. Source

Phage N15  Only our older readers will remember, with all due respect, the once prevailing view that eukaryotes are distinguished from prokaryotes by linear chromosomes, while the latter have circular chromosomes and plasmids. In her 2008 post Some Like It Linear, Merry looked at the numerous examples that had overturned this view. And of course she mention­ed the temperate bacteriophage N15 in this post, whose prophage does not integrate into the host genome like λ (technically: lysogenization) but replicates in the cytoplasm as a linear plasmid with covalently closed ends. Recently, I explained in some detail how phage N15 elegantly solves the 'end replication problem' of a linear genome in Borrelia and it's not-so-loose ends (1|2) (7th paragraph).
 

Apart from the fact that its prophage is a linear plasmid, phage N15 resembles phage λ in many respects. Ravin (2015) found that: "genes 1 through 21 there is a one to one correlation with the phage λ genes A through J. ... There is up to 90% sequence identity between the N15 and the λ head gene products. Some regions are more closely related to other lambdoid phages, and starting from gene 17 to gene 23, N15 better matches phages HK97 and HK022." Late during infection of their E. coli host, in both N15 and λ, concatemeric genome oligomers are cut at cos sites and (linear) monomers packaged into phage heads. Linear phage genomes are circularized via their cos sites as soon as they enter a host in a new infection round and replicate by θ replication in the lytic pathway. It is not known in detail, neither for N15 nor for λ, exactly how the switch from θ replication to σ replication is regulated. This switch is required for the formation of genome concatemers for packaging and has been confirmed to operate (Figure 2).
 

Figure 2. Model of the N15 plasmid prophage and lytic replication. A indicates replication of the N15 plasmid pro­phage. B shows lytic replication initiated in the lysogen. Note that the circular head-to-head dimer is supposed to be processed by protelomerase into two circular monomers (only one is shown). The known or suggested participation of the N15 proteins TelN and RepA at the individual steps is shown. Source

The lambdoid phages ϕKO2, pY54, and Myoviridae phages ΦHAP-1, VHML, VP882, Vp58.5, vB_VpaM_MAR, are known members of the N15 group, which has now grown to 44 members thanks to the work of Pfeifer et al. (2021). Hallmarks of their genomes are the presence of a telN gene (protelomerase), sop genes for stable plasmid inheritance (important for low-copy number plasmids), and related plasmid replication genes as, for example, the repA initiator. The "group portrait" shown in Figure 1b is a so-called pangenome graph that represents a collection of aligned graphs as a network of walks through an underlying merged sequence graph. It takes some practice to understand such a pangenome graph – which I lack – but it is easier to "read" than the exemplary alignments of a few selected group members in Figure 1a (bottom part). Such "classical" alignments allow you to recognize genomic rearrangements such as inversions of entire blocks of genes at a glance, but it is hardly possible to look into the details. And they are completely unsuitable for alignments of more than 4–5 examples, let alone 44.
 

Horizontal gene transfer  HGT not only blurs the boundaries between classes of mobile genetic elements (MGEs) but also makes it impossible to reconstruct their genealogy in most cases due to their high mutation rates. It is often not even possible to determine the direction in which a par­ticular HGT occurred, and exchange in both directions is probably the rule rather than the exception. What is accessible to quantitative analysis is a cataloging of HGT events that took place with all probability. Pfeifer & Rocha (2024) turned to the genetic exchanges between MGEs without taking into account the functionality of the exchanged genes and say: "To assess quantitatively the gene flow between different phages, plasmids and P–Ps, we first clustered all recombining genes (RGs) into families by sequence similarity. We used a strict threshold of sequence similarity (≥80% sequence identity and ≥80% coverage based on amino acid sequences) and then used a single-linkage algorithm (each gene in a family has at least one other highly similar member) to produce the clusters. We then analyzed each RG family in terms of the elements showing evidence of genetic exchanges. For each family there are potentially six different types of exchanges: within a given type of MGE and between types. We counted the types of exchange present in each family. For instance, if an RG family is in four plasmids and two P–Ps, we mark the presence of exchanges between plasmids, between P–Ps and between P–Ps and plasmids."
 

Figure 3. The numbers on the arrows represent the number of RG (recombining genes) families with at least one such type of exchange (e.g., at least one exchange between plasmids in orange or between phages and plasmids on the orange to blue arrow). Source

Gene flow is high within a type of MGE, highest among plasmids. Genetic exchanges between phages and plasmids seem to take place at low frequencies. In contrast, exchanges of these two types of elements with P–Ps are much more frequent and the exchange between plasmids and P–Ps even higher than within the P–Ps. P–Ps may thus play a key role in driving gene flow between the other elements (Figure 3).
 

This finding is not just interesting for theory but has practical consequences. When epidemio­logists study the spread of antibiotic resistance in human pathogenic bacteria, they assume that transmission between individual lineages of a species and between different species is most efficient via conjugation, that is, mediated by conjugative plasmids (the famous Shigella 'R-factors' mediating multiple antibiotic resistancies were detected in Japan in the 1950 and are such conjugative plasmids). Conjugation requires spatial proximity of donor and recipient in the lower micrometer range. If there is a non-negligible number of plasmids that can package themselves, including acquired antibiotic resistance genes, and distribute as virions this results in spooky actions at a distance, so to speak (no quantum entanglement implicated though).
 

Phage–Plasmids, finally  It is a relief that the multitude of phage–plasmids (P–Ps) that Eugen Pfeifer, Eduardo Rocha, and their coworkers fished from databases need not lead to even greater confusion in classifications of extrachromosomal elements. They say: "It is usual to class plasmids according to the replication incompatibility (Inc types) and phages to their virion structure (although the genomic relatedness is becoming the new standard). P–Ps can be classed relatively to the phage taxonomy and plasmid incompatibility, because they encode virions and plasmid replicases." But a proper name, phage–plasmids, seems warranted, because the work of Pfeifer & Rocha (2024) makes it clear that we are dealing here with a whole class of genetic elements and not just a few oddities like phages P1, N15, SSU5, and satellite phage P4. It's also really helpful that they suggest "P–P" as an abbreviation because it avoids confusion with one of the many synonyms of PP, not least pyrophosphate (PPi).
 

In the second part, I will introduce some lesser-known examples of phage–plasmids.

 

 

Do you want to comment on this post? We would be happy about it! Please comment on Mastodon, Bluesky, or on 𝕏 (formerly Twitter).