Cryptic Prophages Contribution for Campylobacter jejuni and Campylobacter coli Introgression

Campylobacter coli and C. jejuni, the causing agents of campylobacteriosis, are described to be undergoing introgression events, i.e., the transference of genetic material between different species, with some isolates sharing almost a quarter of its genome. The participation of phages in introgression events and consequent impact on host ecology and evolution remain elusive. Three distinct prophages, named C. jejuni integrated elements 1, 2, and 4 (CJIE1, CJIE2, and CJIE4), are described in C. jejuni. Here, we identified two unreported prophages, Campylobacter coli integrated elements 1 and 2 (CCIE1 and CCIE2 prophages), which are C. coli homologues of CJIE1 and CJIE2, respectively. No induction was achieved for both prophages. Conversely, induction assays on CJIE1 and CJIE2 point towards the inducibility of these prophages. CCIE2-, CJIE1-, and CJIE4-like prophages were identified in a Campylobacter spp. population of 840 genomes, and phylogenetic analysis revealed clustering in three major groups: CJIE1-CCIE1, CJIE2-CCIE2, and CJIE4, clearly segregating prophages from C. jejuni and C. coli, but not from human- and nonhuman-derived isolates, corroborating the flowing between animals and humans in the agricultural context. Punctual bacteriophage host-jumps were observed in the context of C. jejuni and C. coli, and although random chance cannot be fully discarded, these observations seem to implicate prophages in evolutionary introgression events that are modulating the hybridization of C. jejuni and C. coli species.


Introduction
Campylobacter species related to campylobacteriosis cases shows that 83.9% of the identified bacteria were C. jejuni, 10.3% were C. coli, while the remaining are other Campylobacter species [1]. Thus, the thermotolerant C. jejuni and C. coli are the major sources of human campylobacteriosis [2][3][4], mostly attributed to poultry meat handling and consumption [5,6]. These two species are thought to have diverged over 6500 years ago during the Neolithic revolution, coinciding with animal domestication and changes in agriculture practices, while C. coli population diverged into three distinct clades (clades 1, 2, and 3) by about 1700-1000 years ago [4,7]. The appearance of clonal complexes in the population occurred way after the species divergence [4].

Identification of Prophages
Within the 22 C. jejuni and 82 C. coli newly sequenced genomes, a total of 402 prophage regions were predicted (123 PHASTER-identified and 279 PHT-identified, mean length of 18,340.47 ± 9773.85 bp, ranging from 4543 to 50,845 bp-data not shown). Considering only predicted prophages without homology with plasmids, larger than 25,000 bp, presenting structural proteins, and classified as intact (by PHASTER) or active (by PHT) reduces the list to nine by PHASTER (2 in C. jejuni and 7 in C. coli) and 29 by PHT (4 in C. jejuni and 25 in C. coli) ( Table 1). PHASTER and PHT clearly identified three groups of prophages with homology with CJIE1, CJIE2, and CJIE4. The nine PHASTER-identified prophages evidenced homology with CJIE1 (average percent identity of 81.1% for C. coli and 93.1% for C. jejuni predicted prophages). Among the PHT-predicted prophages, 28 had homology with CJIE2 (average percent identity of 54.5% for C. coli and 52.8% for C. jejuni predicted prophages), and one with CJIE4 (percent identity of 91.7% for a C. coli prophage). These observations together with the phylogenetic analysis (see below) pinpoint the existence of two novel prophages in C. coli that for their similarity with C. jejuni prophages were named Campylobacter coli integrated element 1 (CCIE1) and Campylobacter coli integrated element 2 (CCIE2). Although CJIE1-like prophages have been considerably reported in C. coli genomes [42], the majority (if not all) of them may be, in fact, CCIE1 prophages. Interestingly, CJIE2 prophages were only reported in C. jejuni isolates so far [12,[42][43][44][45][46][47], whereas, in contrast, CCIE2 prophages were found in both C. coli and C. jejuni genomes. CCIE1 and CJIE1 are very similar to each other (≈80% sequence identity), while CCIE2 and CJIE2 do not show such similarity (≈50% sequence identity). Significant deletions and rearrangements in CJIE2-like prophages were described [44,45], potentially explaining the low coverage of homologous regions when comparing CJIE2 and CCIE2 prophages. Notably, CJIE1-CCIE1 and CJIE2-CCIE2 display several genes shared in block with same organization (data not shown). Both CCIE1 and CCIE2 showed genome length of 38,556 bp and 36,356 bp, respectively, consistent with CJIE1 (34,403 bp) and CJIE2 (40,268 bp).

Characterization of the New CCIE1 and CCIE2 Prophages
A more detailed analysis was performed on the CCIE1 region within C. coli Cc63-H-18 genome, and on the CCIE2 region within C. coli Cco1598-H-13 genome (Figure 1, Tables 2 and S3). Most predicted CDS in both CCIE1 e CCIE2 were matched either by sequence or structure with phage genes (Table S3). Several hypothetical proteins were annotated as a result of the lack of knowledge surrounding Campylobacter spp. phages. A bacterial putative NADH-ubiquinone oxidoreductase, located at the 3 edge of CCIE2 genome, was identified, potentially representing a watermark from a past phage transduction event. More specifically, the peripheral location of the putative bacterial gene in the prophage genome points to specialized transduction, in which the bacteriophage packages its genome with flanking bacterial DNA taken during chromosomal excision [48]. genome points to specialized transduction, in which the bacteriophage packages its genome with flanking bacterial DNA taken during chromosomal excision [48].
Regarding structural analysis, CCIE1 seems a near-complete prophage lacking only one essential structural protein, possibly explaining the general failure of induction attempts of its close homologue CJIE1 prophage [15,49], while not objecting a case of apparent CJIE1 induction success [12]. However, it should not be ruled out that induction failure may be due to inefficient experimental conditions. On the other hand, CCIE2 was predicted to have several structural proteins missing, similar to its homologue CJIE2 [12]. The lack of structural proteins is a common feature of cryptic prophages [50], which suggests that CCIE2 is possibly incomplete. Hypothetical proteins for which no annotation update was possible are represented in yellow. Regulation proteins and genome processing proteins are depicted, respectively, in blue and green, while lysis-related proteins and structural proteins are highlighted in red and grey, respectively. Partial dns nuclease found in CCIE1 is shown in lilac. For space simplification, the linear prophage region is represented as a circular genome, with start (first residue) and end (last residue) highlighted at 0′ position of each representation. The figure was obtained using Geneious Prime 2020.1.1. Further description of the annotated regions is available on Table S3. Hypothetical proteins for which no annotation update was possible are represented in yellow. Regulation proteins and genome processing proteins are depicted, respectively, in blue and green, while lysis-related proteins and structural proteins are highlighted in red and grey, respectively. Partial dns nuclease found in CCIE1 is shown in lilac. For space simplification, the linear prophage region is represented as a circular genome, with start (first residue) and end (last residue) highlighted at 0 position of each representation. The figure was obtained using Geneious Prime 2020.1.1. Further description of the annotated regions is available on Table S3.
Regarding structural analysis, CCIE1 seems a near-complete prophage lacking only one essential structural protein, possibly explaining the general failure of induction attempts of its close homologue CJIE1 prophage [15,49], while not objecting a case of apparent CJIE1 induction success [12]. However, it should not be ruled out that induction failure may be due to inefficient experimental conditions. On the other hand, CCIE2 was predicted to have several structural proteins missing, similar to its homologue CJIE2 [12]. The lack of structural proteins is a common feature of cryptic prophages [50], which suggests that CCIE2 is possibly incomplete.  . Although several tail-related proteins were annotated in the genome, no tail-like structures were observed on CJIE1 particles (c), and tail-resembling structures were inconsistently observed for CJIE2 (white arrow head on (d)).

Prophage Induction Assays
Despite being reported as mitomycin C-inducible [12], to our knowledge CJIE1 induction was not described in detail and there are a couple reports of induction failure [15,16]. CJIE2, on the other hand, was reported as incomplete; thus, likely not inducible [12]. For the CJIE1-like prophage, PCR after exonuclease treatments rendered phagic gene amplification ( Figure 2a, lanes 6, 12, and 18). Similar results were obtained for the CJIE2-like prophage (data not shown), suggesting that both C. jejuni-harbored prophages are inducible. The detection of circular phagic genomes of not-induced bacteria (noninduced control), points towards a basal spontaneous release of phage particles. For the CCIE1-like prophage, PCR after exonuclease treatments rendered no phagic gene amplification (Figure 2b, lanes 6, 12, and 18). Similar results were obtained for the CCIE2-like prophage (data not shown), suggesting that CCIE1 and CCIE2 were not inducible, thus rendering no PCR amplification. Despite these observations, no phage-induced lysis was observed upon application of PCR positive phage precipitates on the potentially indicator strains (data not shown).
Negative staining transmission electron microscopy was performed on PCR-positive phage precipitates, and phage-like structures were observed (arrows on Figure 2c,d).
Although scarce, such structures were observed in all the PCR-positive phage precipitates, showing roughly icosahedral heads with 49 ± 4 nm (CJIE1- Figure 2c) and 50 ± 3 nm (CJIE2- Figure 2d). Even though several tail-related proteins were annotated in the genome and predicted by VIRFAM, no tail-like structures were observed on CJIE1 particles ( Figure 2c) and tail-resembling structures were inconsistently observed for CJIE2 (Figure 2d, white arrow head). This feature may be due to the lack of tail-related proteins rendering tailless phages or possibly due to some limitations on negative staining which may not reveal phage tails because of their limited density [28]. The observed bacteriophage-like structures conjugated with the PCR detection of circular phage DNA on culture supernatants upon induction points to the production of CJIE1 and CJIE2 bacteriophage particles, thus not supporting the previous reports of CJIE1 and CJIE2 as incomplete, at least in the assayed Campylobacter spp. strains.

Identification of CJIE1-, CJIE4-, and CCIE2-like Prophages within Campylobacter spp.
To understand the phage dynamics on Campylobacter spp. population, the reference CJIE1-1 and CJIE4-1 prophages, as well as the newly identified CCIE2 prophage, were used as template model prophages, since the prophages identified using PHASTER and PHT fell essentially in these three groups (Figure 3a). Integrated elements CJIE3 and CJIE5 were not included as models for this screening as neither PHASTER-nor PHT-predictions evidenced any of these prophages. To double-check this, BLASTn querying CJIE3 and CJIE5 were conducted, identifying no CJIE3-like regions in the newly sequenced C. coli and C. jejuni genomes, but verifying the presence of CJIE5 in all genomes. Indeed, CJIE5, which is likely not a prophage, is also present in C. jejuni RM1221 (where all were initially identified) and in C. jejuni NCTC11168 (described as a prophageless strain). The total number of prophages found was 177 (Table 3 and Table S1). In detail, the screening on the 104 genomes of this study rendered a total of 39 Campylobacter spp. isolates carrying CJIE1like prophages, 53 carrying CCIE2-like prophages, and 4 carrying CJIE4-like prophages. Among these genomes, 12.5% (13/104) harbored two prophages, either CJIE1 and CJIE2 or CJIE2 and CJIE4 (consult Tables S1 and S4 for details). Screening on 736 publicly available genomes rendered a total of 48 Campylobacter spp. isolates carrying CJIE1-like prophages, 24 carrying CCIE2-like prophages, and 9 carrying CJIE4-like prophages ( Table 3). Among these genomes, 1.1% (8/736) harbored two prophages, CJIE1 and CJIE2 (Table S5). The prophage presence ratio on Portuguese isolates was much higher than in the publicly available genomes of Campylobacter spp. (Table 3), and it does not appear to be related with ST or CC type (Tables S1, S4 and S5). Different Campylobacter spp. populations, either in terms of location, host of isolation, or phenotypic characteristics, seem to differentially carry CJIE elements (Table S6), supporting a potential role of these prophages in the modulation of carriers genomic and phenotypic features [14,16,17,52], and in the evolution and ecological adaptation of the isolates (as reviewed by Harrison and Brockhurst (2017) [53]). However, our approach of considering only prophages with template coverage over 50% may have led to the discard of prophages, namely remnant, truncated, or mosaic prophages, which may have been analyzed in the previously mentioned studies. Nonetheless, the reverse reasoning is also valid: it is possible that we have considered distinct phages to be similar to the reference phages due to sequence coverage above the 50% mark. For databases-retrieved isolates, we should remind that only isolates with information regarding MLST, country of isolation, and host organism were selected, which may have led to the non-analysis of prophages eventually carried by the remaining genomes available in the databases.

Insertion Sites of the Prophages Identified
The insertion site of CJIE1-like prophages was the least conserved, with over a dozen insertion sites identified (Table S4A). This volatility of insertion sites of CJIE1-CCIE1 prophages is in agreement with other previous descriptions of varied insertion sites [12,42,49,54,55]. Even with diverse insertion sites in the population, CJIE1-like prophages of isolates of the same ST largely shared the same integration sites, especially in the context of C. coli. Interestingly, the original insertion site of CJIE1 in C. jejuni RM1221 was not between the determined insertion sites [12].
Among CCIE2-like prophages, the insertion sites were highly conserved, pointing to vertical prophage transmission, with prophages within C. coli genomes normally inserting between a tRNA-Leu gene (locus tag: N149_0910) and an adjacent hypothetical protein coding sequence, identified as a putative NTPase or cell division-related protein (locus tag: N149_01865) (Table S4B). C. jejuni CCIE2-like prophage genomes were usually inserted between a tRNA-Arg gene and an adjacent hypothetical protein coding sequence, which could not be identified either by sequence or structure homology (locus tag: Cj0494) (Table S4B). Consistent with previous observations, this insertion site is the same used in C. jejuni RM1221 by CJIE2 [12], which was also described to be majorly conserved [44,45,47].
Similarly, most CJIE4-like prophages shared a common insertion site in C. jejuni isolates between a tRNA-Met (upstream of locus tag: Cj1282) and a tRNA-Phe genes (downstream of locus tag: Cj1283) (Table S4C). This insertion site transversal to C. jejuni isolates is coincident with the original insertion site of CJIE4 in C. jejuni RM1221 where it was first identified [12], and with previous reports [14,54], reinforcing the conservation of this prophage insertion site. The insertion site for CJIE4-like prophages harbored by C. coli isolates was hard to determine precisely, but generally seems to be between a tRNA-Met gene (locus tag: N149_0550, described in the reference genome as a tRNA-Ile gene) and an adjacent hypothetical protein-coding sequence, identified as a putative conjugative transfer protein (upstream of locus tag: N149_0550) (Table S4C).

Phylogenetic Analysis of the Prophages Identified
The maximum-likelihood phylogenetic tree (Figure 3a) clearly separated the three groups of prophages identified, CJIE1-, CJIE4-, and CCIE2-like prophages. Inside each group, the separation between prophages harbored by human-derived isolates (unfilled markers) and nonhuman-derived isolates (filled markers) was weak, as expected by the typical epidemiology and transmission mechanisms of Campylobacter spp. with isolates flowing between animals and humans in the agricultural context suggested by the clonal complexes of isolates [4]. CJIE3 did not cluster with either of these groups, highlighting its genomic divergence. In the CJIE1-and CCIE1-like prophages cluster, there was a clear separation between C. coli and C. jejuni harbored prophages. The cluster of CJIE4-like prophages (blue markers) was more well mixed, with three prophages harbored by C. coli isolates (Cc60-H-18, Cc65-H-18, and Cc72-H-18, arrowed in Figure 3b) being displayed among the C. jejuni-carried prophages and not closer to the remaining prophages harbored by the C. coli isolates, especially the early-diverging CcoCVM41970-H-11 prophage (Figure 3b). Although the isolate Cc72-H-18 has a new reported MLST profile with no clonal complex assigned, the isolates Cc60-H-18 and Cc65-H-18 belong to ST-828 complex, within C. coli clade 1, for which introgression events were described [9]. These three prophages can point to potential host-jumps and introgression events, but a larger set of CJIE4-like prophages would be necessary for strong conclusions. To test introgression, the ABBA-BABA test was performed [35,36], and when introgression was detected, the admixture proportion was determined [37,38]. Comparing the genomes of prophages reveals that CJIE4-like prophage of C. coli genomes Cc72, Cc60_1, and Cc65_1 (arrow in Figure 3b) and CJIE4-like prophage from C. jejuni genomes share more derived alleles than expected by chance. The resulting positive D-statistic of 0.6426 and Z-score of 7.583 (Table S7A) suggests introgression between these prophages, with an estimated admixture of 88.31% (Table S7A).
For CCIE2-like prophages, a sharp separation between C. coli-harbored and C. jejuni-harbored prophages was observed (Figure 3c), segregating CJIE2-and CCIE2-like prophages. The CJIE2 prophage integrates the cluster comprising only prophages harbored by C. jejuni isolates, while CCIE2 prophage was in the cluster comprising majorly CCIE2like prophages harbored by C. coli isolates. In this last cluster, two noticeable exceptions were found: the CCIE2-like prophages carried by C. jejuni Cj5-H-17 and CjeCJ017CCUA-H-01 (arrows on Figure 3c). These punctual observations suggest that some Campylobacter spp. bacteriophages could have the ability to infect both C. coli and C. jejuni isolates in interspecies infections, as both contributor and consequence phenomena of introgression events. The ABBA-BABA test verified the existence of prophage introgression for CCIE2like prophages in C. jejuni prophages. Comparing the genomes of prophages reveals that CCIE2-like prophage of C. jejuni genomes CjeCJ017CCUA and Cj5-H-17 (arrow in Figure 3c) and CCIE2-like prophage from C. coli genomes share more derived alleles than expected by chance. Again, the resulting positive D-statistic of 0.3696 and Z-score of 4.958 (Table S7B) suggests introgression between these prophages, with an estimated admixture of 39.32% (Table S7B). Additionally, the positive D-statistic of 0.3144 with a Z-score of 5.572 suggests introgression of CJIE2 in CCIE2 present in C. jejuni genomes, with an estimated admixture of 39.46% (Table S7B).  Prophages harbored by C. coli are represented as circles while prophages harbored by C. jejuni are represented as triangles. Filled markers are displayed for Campylobacter spp. isolates obtained from nonhuman hosts, while isolates from human hosts display unfilled markers. On CCIE2-like cluster, two internal clusters can be distinguished: one composed of C. jejuni-harbored prophages (light green) that includes CJIE2 prophage (square); and the other composed majorly of C. coli-harbored prophages (dark green) that include the new identified CCIE2 prophage (square), potentially representing a C. coli homologue of CJIE2 not yet described. Similarly, on the CJIE1-like cluster, two internal clusters can be distinguished: one composed of C. jejuni-harbored prophages (light red) that includes all CJIE1 series prophages (squares); and the other composed of C. coli-harbored prophages (dark red) including CCIE1, a not-yet-described C. coli homologue of CJIE2. Clusters depicting potential host jumps are highlighted (dashed rectangles) and are presented in (b,c). Cluster (b) refers to CJIE4-like prophages (blue). Cluster (c) refers to CCIE2-like prophages (green). No potential host-jumps were detected in CJIE1-like prophages (red cluster on (a)). Potential introgression events are emphasized (arrows). Tree scales and key are presented. Prophages used to construct the tree are detailed on Table S4.

Newly Identified MLST Allelic Variations and MLST Profiles
Among the analyzed C. coli genomes, two new allelic variations and eight new MLST profiles were identified and submitted to PubMLST database [56]. The new profiles were assigned a new non-existing ST with eight being assigned to pre-existing clonal complexes and three having no clonal complex assigned (Table S1). As expected, the majority of our new-profile isolates was assigned to ST-1150 and ST-828 clonal complexes, which dominate clinical and farm isolates [4].

Nucleases Encoded by Campylobacter Prophages
For both CJIE1 and CJIE4, the endonuclease genes dns and nucA, respectively, were described to prevent carrier isolates' natural transformation [18,19]. CJIE2 also carries a copy of nucA identical to the CJIE4 one [19]. As these endonucleases can have a major role on shaping Campylobacter spp. genome evolution and population structure by potentially modeling introgression and contributing to the maintenance of stable lineages, nucleases carried by the identified prophages were analyzed (Table 4). Interestingly, all CJIE1 and CCIE1-like prophages carried a version of dns gene either in a complete form or in a partial form. All partial forms were found in C. coli-harbored prophages (CCIE1-like prophages) while only two C. coli-harbored prophages carried the integral dns gene, contrarily to C. jejuni (Table 4), thus possibly facilitating the C. jejuni-to-C. coli introgression process, as it is suggested to occur [4], rather than the opposite direction. * Genes with coverage >90% were considered as complete, while genes with coverage 50-90% were considered as partial; n.d., lower coverages were reported as not detected.

Discussion
Prophages are commonly found in C. jejuni and C. coli genomes, although their prevalence may be unbalanced worldwide, since the prevalence of detected prophages diverge between studies [42,43,54]. A manual curation of prophages predicted using specific software [21,22] led to the identification of three main prophage types found in C. jejuni and C. coli genomes, the CCJIE1 and CCIE1 group, the CJIE2 and CCIE2 group, and the CJIE4 group. Among these, we have introduced two new prophages, CCIE1 and CCIE2, found in C. coli genomes, whose close homologues are the C. jejuni prophages CJIE1 and CJIE2, respectively. The new prophage found was CCIE1, a 38,556 bp region, probably incomplete and homologous to CJIE1. The other new prophage, CCIE2, is a 36,356 bp region that, although harboring a wide range of bacteriophage-like proteins, is probably an incomplete prophage, as supported by the inability to induce this prophage. This region has several common features with CJIE2 prophage identified in C. jejuni RM1221 and was found to represent a close homologue of CJIE2 prophage in C. coli. Although not unequivocally, induction assays and analysis point towards the possibility of CJIE1 and CJIE2 inducibility, even though at poor induction rates. It is possible that both CJIE1-CCIE1 and CJIE2-CCIE2 prophages are currently on the edge of becoming cryptic prophages, with most lacking essential structural proteins leading to failure upon induction assays, while some maintain a minimal manageable set of essential structural proteins, making possible the (poor) induction observed for these prophages, namely, the C. jejuni-harbored homologues.
The prophages found in a large collection of C. jejuni and C. coli genomes clustered into three main groups, as supported by the phylogenetic tree, including CJIE1-CCIE1, CJIE2-CCIE2, and CJIE4. The genetic similarity between the pairs CJIE1-CCIE1 and CJIE2-CCIE2 and the phylogenetic clustering according to species points to prophage acquisition previous to the speciation event of C. jejuni and C. coli, followed by co-evolution between prophages and their bacterial hosts [29,[57][58][59]. Furthermore, when there is conservation of the prophage insertion site, prophage vertical transmission may be in place [60,61].
The global scenario suggests that all the three analyzed prophage groups (CJIE1-CCIE1, CJIE2-CCIE2, and CJIE4) have different homologues running in either Campylobacter species, even for CJIE4-like prophages, leaving open that a C. coli homologue of CJIE4 may also exist. However, we may not discard the hypothesis that CJIE4-like prophages switch between host species (C. jejuni and C. coli), not denoting a canonical prophage sequence for each species. CJIE4-like prophage kept the same insertion site in both C. jejuni and C. coli and shared the same phylogenetic cluster, suggesting cross-species transmissions, in line with what was observed for streptococcal prophages [62]. In these cases, prophages may be evolving independently from the host bacteria, driven by ecological relatedness rather than evolutionary relatedness of the host bacteria [62].
Phages are known to be highly species-specific, infecting only one species and sometimes only some strains of a certain species [63]. Remarkably, we found prophages from C. coli in C. jejuni genomes (CCIE2-like prophage) and vice versa (CJIE4-like prophage). Crossspecies transmission of prophages have also been described for streptococcal prophages [62], suggesting that some prophages are capable of species spillover that may erode species differences over time. Furthermore, conserved endonuclease genes such as dns and nucA, which inhibit natural transformation capability of Campylobacter spp. [18,19], were identified among the analyzed prophages, suggesting that prophage presence may be associated with more stable lineages after prophage acquisition.
Indeed, we have found punctual introgression events for CJIE4-like and CCIE2-like prophages. The introgression events found in C. coli and C. jejuni agree with previous observations. It should be noted that Campylobacter spp. have a very plastic and dynamic genome evidencing transference of genetic material between C. jejuni and C. coli. Consequently, there is sharing of several molecular mechanisms [4]. In fact, ST-828 and ST-1150 clonal complexes were originated in clade 1 of C. coli by the accumulation of C. jejuni DNA [9]. Despite that C. jejuni Cj5-H-17 and CjeCJ017CCUA-H-01 are not assigned to a clonal complex, the harbored prophages are phylogenetically close to CCIE2 prophages harbored by C. coli isolates belonging to ST-828 complex (clade 1). Introgression events are described to be more common in C. jejuni-to-C. coli direction [9], but for CCIE2-like prophage the spillover appears to occur from C. coli-to-C. jejuni. For CJIE4-like prophages, introgression appears to occur in the classic C. jejuni-to-C. coli direction and involving ST-828 complex (C. coli Cc60-H-18 and Cc65-H-18 belong to ST-828 complex, while C. coli Cc72-H-18 does not fit in any pre-existing clonal complex). Introgression can impact clinical and ecological traits of Campylobacter spp. isolates and suggests the maintenance of cell mechanisms between these two species, potentially facilitating bacteriophage host-jumps. Although present in a smaller number of strains, the introgression events detected point to introgression driven by prophages. Conjugative plasmids have also been associated with introgression events [64], supporting interspecies introgression through horizontal gene transfer. This spillover of genes from one species into the gene pool of another species shakes species boundaries, which are semipermeable especially at specific genome regions [65]. The mechanism behind bacteriophage host-jumps and introgression phenomena may be mediated by the capacity of the phage to infect both species, or by uptake of DNA since Campylobacter species are naturally competent [66]. The absence of induction evidence suggests that CCIE1 and CCIE2 prophages are cryptoprophages-incomplete and dysfunctional versions of the prophages that were once able to complete lytic cycles but were domesticated by the host and condemned to perpetual lysogeny [67]. Although defective, it was previously shown that cryptophages can contribute to host fitness [67], potentially explaining the degree of conservation observed. This observation is consistent with the reported difficulties for obtaining infectious particles of the CCIE1 C. jejuni-harbored homologue, CJIE1 [15,16]. The same was here observed for CCIE2 phage, suggested as incomplete by VIRFAM analyses and by its similarities with the incomplete CJIE2. Although not assayed in this work, mainly due to lack of harboring-strain availability, CJIE4-like prophages were also reported to be difficult to induce, thus possibly being incomplete [14]. Interestingly, and contrary to the aforementioned [12,15,16], our induction assays suggest CJIE1 and CJIE2 as inducible. Bacteriophages may play a role in introgression, even in the absence of infectious particles production. It would also be interesting to understand whether the punctual host-jumps mentioned can represent an opportunistic transduction. In other words, they can represent the carriage of genomes from putatively incomplete prophages within infectious particles of complete bacteriophages co-infecting the same host. This phenomenon is called molecular piracy and was already described in Caudovirales bacteriophages, the order of all Campylobacter phages described so far [68][69][70]. Prophage introgression by random chance due to natural competence in Campylobacter spp. cannot be fully discarded, and additional work is required to better clarify the role of prophages for population structure. Further studying Campylobacter spp. prophage genomics and phylogenetics is of major interest and may give insight not only into bacteriophage biology, but also on Campylobacter population structure and introgression phenomena.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10030516/s1, Table S1: General characterization of the newly sequenced Campylobacter spp. isolates analyzed in the present study; Table S2: Primer pairs used for phage circular DNA detection; Table S3: Predicted annotated regions within CCIE1 and CCIE2 prophage genomes; Table S4: General characterization of the prophages identified in the present study within Campylobacter spp. population; Table S5: General characterization of database isolates harboring prophages; Table S6: Presence of C. coliand C. jejuni-harbored prophages in different studies; and Table S7: ABBA-BABA test considering the following tree sample organization (((P1, P2) P3) P4)). Funding: F.F.V. is funded by Fundação para a Ciência e a Tecnologia (FCT) through an Assistant Researcher grant CEECIND/03023/2017, and a project grant (PTDC/BTM-SAL/28978/2017) that supported this work. The work is partially supported by Na-tional funds from FCT, projects UIDB/04138/2020 and UIDP/04138/2020. Campylobacter strains were sequenced under the GenomePT project (POCI-01-0145-FEDER-022184), supported by COMPETE 2020-Operational Programme for Competitiveness and Internationalization (POCI), Lisboa Portugal Regional Operational Programme (Lisboa2020), Algarve Portugal Regional Oper-ational Programme (CRESC Algarve2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF), and by FCT. This work was also supported by Fundos FEDER through the Programa Operacional Factores de Competitivida-de-COMPETE and by Fundos Nacionais through the FCT within the scope of the project UID/BIM/00009/2019 (Centre for Toxicogenomics and Human Health-ToxOmics).

Informed Consent Statement: Not applicable.
Data Availability Statement: Raw sequence reads of bacterial genomes, which are representative of the prophages diversity, are deposited in the European Nucleotide Archive (ENA), under the study accession numbers PRJEB46733 and PRJEB46750. Campylobacter prophage sequences are available under the GenBank accession numbers MZ667634-MZ667639.