The Gene Ail for the Attachment–Invasion Locus Protein of Yer-Sinia Enterocolitica Biotype 1A Strains Is Located on the Genomes of Novel Prophages

Abstract

The attachment–invasion locus protein Ail of pathogenic Yersinia strains is an important virulence factor, both for invasion of eucaryotic cells and for serum resistance. In other Yersinia strains, e.g., those belonging to biotype (BT) 1A of Yersinia enterocolitica, ail has only occasionally been described. Sequence analysis of 370 BT 1A isolates in our laboratory revealed 41 (11.1%) which were ail-positive. Most of these isolates were recovered from minced meat and tonsils of wild boars, and belonged to 17 MLST allele profiles. A closer look at DNA sequences surrounding ail disclosed that the gene in most isolates is embedded in DNA regions encoding phage proteins. The genomes of four prophages belonging to four different phylogenetic clusters were determined and analyzed by in silico studies. These have sizes of 34.9 and 50.7 kb, and are closely related to each other but not to known phages. Unlike other regions of the prophages, the integrases and attachment sites of some of them diverge, leading to different integration sites in the isolates. In a fifth cluster, ail is relocated at a position on the Y. enterocolitica chromosome that is several hundred kilobases apart from those of the other clusters, but surrounded by prophage-related sequences. In addition, highly pathogenic 1B/O:8 strains contain a DNA segment which includes ail and is 65 to 94% identical to the prophage sequences determined in this study.

1. Introduction

Yersiniosis is an infectious disease of the gastrointestinal tract caused by Yersinia (Y.) enterocolitica and, to a lesser extent, by Y. pseudotuberculosis [1,2]. It is the third-most common bacterial enteritis in Europe. Infections are mainly caused by the consumption of raw or undercooked pork [3,4,5,6,7,8,9]. Y. enterocolitica comprises six biotypes (BT) and up to 70 serotypes. While the BTs 1B, 2, 3, 4 and 5 possess the 70 kb virulence plasmid pYV encoding a type III secretion system and effector proteins (Yersinia outer proteins = YOPs) with partly toxic activity for eucaryotic cells, BT 1A strains are generally devoid of pYV [10]. Moreover, some important chromosomally encoded virulence factors of pathogenic biotypes exist only rarely or not at all in BT 1A [11,12]. Two of those are the enterotoxin YstA and the attachment–invasion locus protein (Ail), the latter being involved in both invasion of eucaryotic cells and serum resistance [13,14,15]. To date, the ail gene has only been detected in some BT 1A strains [16,17]. Because of the lack of these virulence factors, BT 1A strains have for a long time been considered to be non-pathogenic. However, in recent years, there have been an increasing number of reports describing BT 1A isolates from clinical cases [18,19,20,21,22,23,24,25,26]. This biotype obviously comprises at least two phylogenetic lineages, each with different virulence factors, some of which are toxins [17,27]. The question arises as to whether horizontal gene transfer may be involved in the heterogeneity of this biotype. In this study, we sequenced 370 BT 1A isolates from different sources. Among these isolates, 41 contained the virulence gene ail. A closer analysis of this gene revealed that, unlike in the strictly pathogenic biotypes, ail in most BT 1A isolates is located on the genomes of prophages. These prophages were then characterized by in silico analyses.

2. Results

2.1. Out of 370 BT 1A Genomes, 41 Contain a Prophage-Associated Ail Gene

Sequence analyses of 370 Y. enterocolitica BT 1A isolates revealed that 41 of them harbor the gene ail. Isolates were obtained between 2013 and 2025 from different sources (mainly minced meat and wild boars). These isolates belonged to 17 different MLST allele profiles (Figure 1A). All ail-positive isolates (n = 41) possessed the enterotoxin gene ystB, while thirty, ten and three isolates additionally possessed the virulence-associated genes hreP, myfA and tccC, respectively (Table S1). For 39 isolates, plasmid-borne sequences of approximately 3 to 106 kb were predicted.

A closer look at the ail gene in these isolates showed that it is up to 99% identical to the ail in pathogenic biotypes, e.g., the bio/serotype 1B/O:8 strain 8081. In addition, the upstream sequence containing the promoter and ribosome-binding sites of ail is similarly related to its counterpart in other biotypes, suggesting that ail may be active in BT 1A.

However, the regions encompassing ail diverge significantly in the various biotypes. In most BT 1A isolates, such as the 24-YE00064 studied here, the gene is surrounded by partitioning genes and genes for cell lysis (lysin and holin), as well as DNA packaging (small and large terminase) typically associated with phage (Table S2). Moreover, genes for phage assembly (capsid and tail) and for genetic switch, as well as integrase and excisionase, were also identified in 37 chromosomes, suggesting that ail is part of a prophage (Table S2).

2.2. Comparison of the Prophages Indicates Relationships Between Them

The 41 prophage sequences identified in the investigated isolates by comparison with the vB_Yen-24-YE00064 prophage form five major clusters (Figure 1A). Among these, the prophages in the clusters C2 to C5 harbor the ail gene, whereas prophages in cluster C1 are devoid of ail because here the gene is located at a different position on the chromosome (see below). An alignment disclosed that most regions of the prophages, particularly those encoding structural proteins, are closely related (up to 85%), whereas the integrase genes show major differences (Figure 1B). Short-read sequencing allowed the prediction of four whole-prophage genomes (vB_Yen_16-YE00051, vB_Yen_20-YE00187, vB_Yen_24-YE00064 and vB_Yen_25-YE00027) belonging to four clusters (Figure 1A). These have genome sizes between 34,918 and 50,744 bp, and are composed of between 55 and 78 Open Reading Frames (ORFs, Table S2). The overall genome organization of the prophages is similar (Figure 2A). As with other temperate phages, ORFs for repressor proteins, cell lysis, DNA packaging, and capsid and tail assembly are clustered, even though some ORFs, particularly those encoding structural proteins, are missing in vB_Yen_16-YE00051. The prophages showed no identities to other phages, and only some relatedness to two Y. enterocolitica BT 1A chromosomes (Y201, CP124238.1; and Y115, CP124259.1).

Three prophages (vB_Yen_16-YE00051, vB_Yen_20-YE00187 and vB_Yen_25-YE00027) have an identical attachment site, att, of 40 bp (Figure 2B). The integrases of these prophages are 100% identical (Figure 2C). By contrast, the prophage vB_Yen_24-YE00064 has an att site of only 27 bp (Figure 2B). The integrase of this prophage is only approximately 35% identical to those of the other ones (Figure 2C). Thus, it is not surprising that the two groups have different integration sites on the bacterial chromosome. While the prophages vB_Yen_16-YE00051, vB_Yen_20-YE00187 and vB_Yen_25-YE00027 are integrated between two genes for hypothetical proteins, vB_Yen_24-YE00064 is integrated between a gene for an integrase and a YebY family protein.

2.3. Relocation of Ail in Cluster C1 and Analysis of 1B/O:8 Strains

Sequence analysis of cluster C1 revealed similar prophage sequences as in isolate 24-YE00064. However, at the position of ail in vB_Yen_24-YE00064, there is a gap in the C1 prophage genome (Figure 3A). In this cluster, ail is located approximately 600 kb apart from the vB_Yen_24-YE00064-related prophage sequences. Interestingly enough, the gene and its adjacent sequences are surrounded by DNA segments which are similarly present in the corresponding prophages, as shown for the prophage vB_Yen_23-YE00044.2 (Figure 3B). The homologous upstream and downstream sequences of ail have lengths of 144 bp and 925 bp, respectively, but are not related to each other. Therefore, it remains open how ail was relocated in isolates belonging to cluster C1. Nevertheless, it is noteworthy that even in the highly pathogenic Y. enterocolitica 1B/O:8 strains 8081 (AM286415.1), WA (CP009367.1) and Billups-1803-68 (CP173224.1), ail is associated with phage genes. Indeed, a stretch of approximately 20 kb of strain 8081 containing ail is similar to vB_Yen_24-YE00064 (Figure 3C). This stretch essentially corresponds to the ØYE200 region identified in 8081 [28]; this, however, has been determined to be a smaller prophage (15.5 kb) without ail. Besides ail, the 20 kb prophage of strain 8081 comprises genes for an integrase, lysis proteins (holin and lysin) and the terminase large subunit. Moreover, the fact that this DNA segment also contains the 27 bp attachment site of vB_Yen_24-YE00064 upstream of the integrase gene, and that this site is linked to tRNA genes, suggests that ail was once associated with a similar prophage. It is conspicuous that the DNA segment in strain 8081 harbors several transposase genes which are lacking in vB_Yen_24-YE00064 and which might have been involved in genetic reassortments.

3. Discussion

The attachment–invasion locus protein Ail of Y. enterocolitica is an important virulence factor which is produced by all pathogenic biotypes of this species, as well as by Y. pseudotuberculosis and Y. pestis [29]. To date, it has only rarely been described in BT 1A strains of Y. enterocolitica and in Y. enterocolitica-like species, e.g., Y. kristensenii, some of which have been reported to contain an additional ail-related gene which may be associated with plasmids or phages [11,30,31]. For this reason, ail is routinely used as a target in detection of pathogenic Y. enterocolitica and Y. pseudotuberculosis by RT-PCR (ISO TS 18867:2015). However, sequencing of 370 BT 1A isolates showed that ail is present more commonly than expected in this biotype. We identified the gene in 41 (11.1%) out of 370 isolates recovered from food and wild boars in the last 12 years. The ail-positive isolates represent a broad range of MLST allele profiles, even though some types (ST304, ST428, ST832) were prevalent. The question arises as to how BT 1A strains may acquire ail. This study suggests that it may occur by lysogenic conversion via temperate phages. Indeed, the analysis of ail-positive BT 1A isolates showed that in most of them the gene is located on a prophage. The prophages are related to each other, but form five different clusters. Up to now, four prophage genomes containing ail could be analyzed in detail. The analysis suggests that three of them (vB_Yen_20-YE00187, vB_Yen_24-YE00064 and vB_Yen_25-YE00027) may be active, because all elements required for the formation of a phage particle are obviously present, whereas the prophage vB_Yen_16-YE00051 is presumably defective because some essential ORFs for capsid and tail proteins are lacking. Indeed, initial induction experiments with an isolate (18-YE00061) belonging to cluster C5 demonstrated that, like other prophages of this isolate, the ail prophage is inducible and forms plaques on some ail-negative BT 1A strains, though only at 37 °C (to be published elsewhere). The fact that genes for structural proteins of the complete prophages are very similar indicates that the corresponding phage particles may have the same morphology. A striking difference, however, pertains to their integration site on the Y. enterocolitica chromosome. Unlike some other parts of their genomes, genes for the integrase and attachment sites are in part highly diverse. As a consequence, the prophages are not integrated at the same position in BT 1A strains. Whether these sites also exist in other Y. enterocolitica biotypes or even other Yersinia species, and whether those strains may also acquire ail by phage-mediated transfer, has still to be studied. It has to be taken into account that to achieve lysogenic conversion the host range of a temperate phage is of major importance. Yersinia enterocolitica BT 1A strains belong to various serotypes, which may determine the host specificity of a phage. It appears, however, that ail prophages are subjected to genetic recombination. This can be clearly seen in cluster C1, comprising related prophages, in which ail including adjacent sequences has been relocated. Similarly, highly pathogenic 1B strains like 8081 contain remnants of ail prophages, suggesting recombination events or even a horizontal gene transfer in the past. Thus, it is conceivable that a related temperate phage of Y. enterocolitica 1B/O:8, which has already been isolated from wild boars in Europe [32], once infected, say, a BT 1A/O:8 strain, followed by recombination with an indigenous prophage. Whether such an event has a selective advantage for BT 1A is not clear. However, it is noteworthy that most isolates of this study were recovered from tonsils of wild boars, where the presence of an ail gene might be beneficial for survival. Because wild boars may come into contact with other animals (e.g., pigs) and can contaminate vegetables and salad plants growing in fields, it is possible that ail-positive Y. enterocolitica isolates obtain access to the food chain. Indeed, the fact that some of the ail-prophage-containing isolates were recovered from minced meat indicates that they were lysogenized with such a phage. This means that the spread of ail in pig farms cannot be ruled out. However, it has still to be demonstrated that Ail is really produced by BT 1A, even though this does seem likely due to the similarity of its ail and upstream sequences to those of pathogenic biotypes. Regrettably, information on the prevalence of temperate Y. enterocolitica phages and their potential to exchange genes is still scarce [33,34]. Popp et al. (2000) isolated eight temperate phages from 170 Yersinia strains by induction with mitomycin C [34]. Seven phages were isolated from Y. enterocolitica, most of which belonged to BT 1A. Some phages were able to infect pathogenic and non-pathogenic Y. enterocolitica strains. It was shown that one of those phages (PY20) isolated from a B4/O:3 strain was able to transduce small Yersinia plasmids, but not the 70 kb virulence plasmid pYV [35]. In another study, ten temperate phages were isolated from 102 pathogenic bio/serotypes of Y. enterocolitica, some of which additionally lysed BT 1A [36]. They formed plaques exclusively at room temperature. However, data on the potential of these phages to transduce DNA have not been published. We will therefore extend our study on temperate phages to investigate whether lysogenic conversion is involved in the spread of the ail gene.

4. Materials and Methods

4.1. Source and Typing of Y. enterocolitica Strains

The 370 isolates analyzed in this study are part of our Yersinia strain collection. This collection comprises both isolates from food sent by local laboratories to the BfR (German Federal Institute for Risk Assessment) for typing and isolates from prevalence studies on Yersinia in wild boars carried out between 2019 and 2025. Here, tonsils of wild boars, shot in the north-east of Germany, were examined for the presence of Yersinia using the EN ISO method 10273:2017 or an improved version of it [37]. The isolates were cultivated on Columbia agar supplemented with 5% sheep blood (bioMérieux Deutschland GmbH, Nürtingen, Germany) at 28 °C for 16–20 h for whole-cell matrix-assisted laser-desorption/ionization time-of-flight mass spectrometric identification (MALDI-TOF MS) using the direct transfer method with HCCA matrix on a Biotyper (Bruker Daltonics GmbH and Co. KG, Bremen, Germany). In addition, physiological and biochemical tests using classical tube and plate procedures for species confirmation and biochemical differentiation were conducted as previously described [38]. Unless otherwise indicated, YP cultivation was conducted at aerobic conditions at 28 °C for 18–24 h using lysogeny broth (LB)-based media. For solid media preparation, LB medium was supplemented with 1.8% Bacto Agar No. 1 (Oxoid Deutschland GmbH, Wesel, Germany) [39,40].

4.2. Genome Sequencing and Bioinformatics Analysis

Whole-genome sequencing (WGS) of Y. enterocolitica BT 1A isolates was performed by short-read, paired-end sequencing (2 × 150 cycles) on a NextSeq500 benchtop device (Illumina Inc., San Diego, CA, USA). Bacterial genomic DNA was extracted from liquid cultures grown at 28 °C for 20–24 h using the PureLink Genomic DNA Mini Kit (Invitrogen, Ebersberg, Germany) according to the recommendation of the manufacturers. DNA sequencing libraries were prepared using the Nextera XT DNA Sample Preparation Kit (Illumina Inc.) [33,36,41,42]. Raw sequencing data were subjected to the Aquamis (v1.4.1) pipeline [43] for quality evaluation, demultiplexing and trimming, while general in silico typing tasks were conducted using BakCharak (https://gitlab.com/bfr_bioinformatics/bakcharak; accessed on 15 July 2025). The BakCharak pipeline (v3.1.6) includes tools for prediction of MLST sequences (pubMLST) and for detection of antimicrobial resistances (AMRfinder [v3.12.8], NCBI resistance gene database [v2024-01-31.1]), plasmids (Abricate [v1.0.1], plasmidfinder [v2.0.1] database [v2022-07-13]), virulence factors (Abricate [v1.0.1], VFDB database [v2024-06-25]), and close reference genomes ((mash [v2.3]), NCBI refseq, plasmid reference finder (mash [v2.3]), NCBI plasmid database [v2.14.1]). In addition, BakCharak was used for plasmid comparison (Blastn [v2.14.1], NCBI plasmid database (accessed on 3 September 2025)) and fastANI species prediction [v1.33] (NCBI refseq database [v2021-11-19]) [43]. Prophage detection was conducted using the Phastest tool (v3.0) for initial screening, while manual curation using Accelrys DS Gene (v2.5; Accelrys Inc., San Diego, CA, USA) was performed to determine the complete prophage sequence from attachment sites of bacteria (attB) and prophages (attP). Genome annotation of the prophage genomes was conducted using the Bacterial and Viral Bioinformatics Resource Center (BV-BRC, v3.54.6). Initial functional prediction of open reading frames (ORFs) was manually curated according to predicted functions of closely related protein sequences derived from blastp searches at NCBI (National Center for Biotechnology Information) [33,36,41,42]. Phylogenetic relationship of vB_Yen_24-YE00064 (reference) to Y. enterocolitica genome datasets encoding Ail were conducted using CSI phylogeny (v1.4; default settings; https://cge.food.dtu.dk/services/CSIPhylogeny/; accessed on 20 September 2025). The software identifies, filter, and validate SNPs from the de novo assembled contigs of Y. enterocolitica BV 1A genomes to construct a phylogenetic tree based on concatenated SNP profiles using default settings [44]. Unless stated otherwise, all software tools are used with default settings. Figures were generated using CorelDraw (v13).

4.3. Genome Accession Numbers

Deposition of prophage genomes at NCBI Genbank was conducted using BankIt for the prophages vB_Yen_24-YE00064 (PV779719), vB_Yen_16-YE00051 (PX531504), vB_Yen_20-YE00187 (PX109664) and vB_Yen_25-YE00027 (PX109665). The genomes of all Y. enterocolitica BV 1A mentioned in Figure 1A and Table S1 are accessible under Bioproject PRJNA1347376.