Possible use of ail and foxA polymorphisms for detecting pathogenic Yersinia enterocolitica.

BACKGROUND: Yersinia enterocolitica is an enteric pathogen that invades the intestinal mucosa and proliferates within the lymphoid follicles (Peyer's patches). The attachment invasion locus (ail) mediates invasion by Y. enterocolitica and confers an invasive phenotype upon non-invasive E. coli; ail is the primary virulence factor of Y. enterocolitica. The ferrioxamine receptor (foxA) located on the Y. enterocolitica chromosome, together with its transport protein, transports a siderophore specific for ferric ion. Currently, ail is the primary target gene for nucleic acid detection of pathogenic Y. enterocolitica.
RESULTS: The genes ail and foxA in 271 pathogenic and 27 non-pathogenic Y. enterocolitica strains isolated from China and 10 reference strains were sequenced, aligned, compared to the ail and foxA sequences of Yersinia enterocolitica subsp. enterocolitica 8081 (Genbank: NC_008800), and analyzed for sequence polymorphism. The ail from the 282 strains showed 3 sequence patterns: 277 strains of serotypes O:3, O:9 and O:5, 27 with identical nucleic acid sequences formed pattern A1; 4 strains of serotype 1B/O:8 with identical nucleic acid sequences formed pattern A2; and one Chinese isolate 2/O:9 formed pattern A3. In the primary coding region of the foxA ORF (Genebank: X60447 nt 433-1866; nt 28 to 1,461 in the ORF), the sequences formed 3 groups and were further divided into 8 sequence patterns.
CONCLUSION: The ail and foxA loci of pathogenic Y. enterocolitica have been analyzed. The ail sequence was highly conserved among the same serotype strains from different sources; and foxA was highly conserved among the pathogenic strains, although there was some sequence diversity. Fewer strains were used from outside China, which is a limitation of the study.

Background

Yersinia enterocolitica is a food-borne pathogen [1] that causes a broad spectrum of clinical syndromes. A primary characteristic of the microorganism is that it penetrates the intestinal epithelial cells and replicates in lymphatic nodules, causing a wide variety of clinical and immunological manifestations [2-4]. All pathogenic Y. enterocolitica strains harbor ail, which is different from the inv sequence (which encodes a protein of similar function), and renders Y. enterocolitica capable of invading the intestinal epithelium. In addition, the Ail protein confers a serum resistance phenotype on Y. enterocolitica [5]. In contrast to inv, which exists in non-pathogenic as well as pathogenic strains of Y. enterocolitica, ail only exists in Y. enterocolitica strains epidemiologically related to human disease [6], and is therefore an important virulence marker. Environmental isolates not associated with disease have a non-functional inv and no ail [7].

Ferric ion uptake is essential for bacterial growth and survival. The supply of iron and production of the siderophore transport system is a central factor in infections with Yesinia pestis and Y. enterocolitica. Pathogenic Y. enterocolitica can be divided into 2 groups, those producing the siderophore, such as biotype 1B/O:8, and those producing no siderophore, as in serotypes O:3 and O:9 [8]; the latter take up ferric ion using ectogenic siderophores, such as ferrioxamin B and ferrioxamin E [9]. The 2 groups have different ferric ion uptake abilities, which may explain the differences in virulence among serotypes [10]. A 77 kDa receptor on the Y. enterocolitica outer membrane [11] combines with ferrioxamin to take up ferric ion rapidly [12]. This process is energy-dependent and requires the action of the TonB protein, part of a complex known as the Ton system. This complex undergoes a conformational change driven by the proton motive force (PMF), which interacts with the outer membrane receptors and activates transport [13]. The FoxA receptor of Y. enterocolitica, the ferrochrome receptor and the TonB-dependent receptor share high amino acid homology [14,15]. The foxA was chosen for study because it exists in all Y. enterocolitica strains.

Using polymorphic gene analysis, we show that combined detection of ail and foxA confirms the identity of pathogenic Y. enterocolitica.

Methods

Bacterial strains and identification of biotype and serotype

We chose 271 pathogenic and 27 non-pathogenic Y. enterocolitica strains isolated from diarrhea patients, animals, food and the environment in China. They included 205 strains of serotype O:9, 72 of serotype O:3, 10 of serotype O:8, 5 of serotype O:5, 3 of serotype O:6, 30 and 3 of undetermined serotype (Table 1), together with 11 reference strains from Europe, the United States and Japan (Table 2). The serotypes, biotypes and pathogenesis of these strains were determined as previously described [16-18].

Table 1

Bio-serotypes of the 298 Y. enterocolitica isolates from China

Host	Pathogenic Strains						Non-pathogenic Strains
	2/O:9	3/O:9	2/O:3	3/O:3	4/O:3	Total	1A/O:3	1A/O:9	1A/O:8	1A/O:5	1A/O:6,30	1A/Un*	Total
Patient	6	2	1	4	1	14				5	3		8
Swine	83	3		46		132	1	2	4				7
Dog	3			13		16						1	1
Fowl	1	1		1		3	1	1					2
Goat				2		2			2				2
Cattle		1				1		1	2				3
Fish				1		1							0
Rat	96	1		1		98						1	1
Rabbit	1					1							0
Food	1	1				2			2			1	3
Environment	1					1							0
Total	192	9	1	68	1	271	2	4	10	5	3	3	27

*: undetermined serotype

Table 2

Sources of the 11 reference strains

Strains No.	Bioserotype	Location	Source
52203	4/O:3	The Pasteur Institute, France	Purchased from the Pasteur Institute by the Institute of Chinese Biomedicine.
52212	4/O:9
52211	1B/O:8
Pa40134	4/O:3	Japan	Provided by Dr. H. Fukushima (Public Health Institute of Shimane Prefecture, Matsue, Japan).
ye3vp-/03	3/O:3
ye3vp5/03	3/O:3
ye4/03	4/O:3
D92	2/O:5,27
Pa12986	1B/O:8
Ye92010	1BO:8
8081	1B/O:8	Complete genome sequence of the highly pathogenic Yersinia enterocolitica subsp. enterocolitica 8081 (Genbank: NC_008800).

Primer nucleotide sequences

The primers for ail and foxA were designed in our laboratory, referencing sequences from GenBank (ail: M29945, foxA: X60447), and synthesized by Shanghai Sangon Biological Engineering & Technology and Service Co., Ltd, China. The primers for ail amplify the entire ORF, while those for foxA amplify the ORF coding region from nt 28 to nt 1,461 (Table 3).

Table 3

Primer sequences and annealing temperatures for ail and foxA.

Target gene and primer direction		Primer Sequences (5'→ 3')	GenBank no.	Location (nt)	Amplicon length	Annealing temp.
ail	Forward	GGT TAT TGT ATT AGT ATT GTT	M29945	446-466	585 bp	57°C
	Reverse	CAG GTG GGT TTT CAC TAT CTG		1031-1051
foxA	Forward	CTC TGC GGA AGA TAA CTA TG	X60447	389-408	1532 bp	58°C
	Reverse	ATC CGG GAA TAA ACT TGG CGT A		1899-1920

PCR, DNA sequencing and sequence analysis

Bacteria were cultured as previously described [18]. The bacterial DNA was extracted using a Blood & Tissue Kit (QIAGEN, USA). PCR was performed in a 200 μl volume containing 10 ng DNA template, 5U Taq DNA polymerase (TaKaRa, China), 0.2 mM of each dNTP, 1 μM of each forward and reverse primer, 1.5 mM MgCl2, 50 mM KCl, and 10 mM Tris-HCl (pH 8.3). Thermal cycling was done in a MJ PTC200 (Bio-Rad, USA) and the conditions were: one cycle of denaturation at 94°C for 5 min, followed by 25 cycles of melting at 94°C for 15 s, annealing for 30 s at various temperatures depending on the primers used (Table 3), elongation at 72°C for 30 s, and a final extension at 72°C for 10 min. Five microliters of PCR product was electrophoresed on a 1.5% agarose gel. The gel image was captured using a Gel Documentation 2000 (Bio-Rad, USA).

The specific PCR products were purified using a Gel Extraction Kit (QIAGEN, USA) and sequenced using an ABI PRISM® BigDyeTM Terminator cycle sequencing Ready Reaction Kit with AmpliTag DNA Polymerase, following the manufacturer's instructions, and an ABI PRISM® 377XL DNA Sequencer (Applied Biosystems, Foster City, CA, USA) at TaKaRa Biotechnology (Dalian) Co., Ltd. The sequences were aligned with the reference sequences. Nucleotide sequence alignments and cluster tree construction were performed using Clustal X (Version 1.8) and MEGA (Version 4).

Results

General features of ail and foxA

ail is located on the Y. enterocolitica chromosome where the ORF encodes a peptide of 178 amino acids, MW: 19,548 Da [19]. There is a typical prokaryotic signal sequence at the N-terminus of the peptide [20] with a cleavage site between residues 23 and 24, where the first 23 amino acids act as a signal sequence [19].

foxA has an ORF of 2,129 bp encoding a protein of 710 amino acids, MW: 78,565 Da. The first 26 amino acids are a signal sequence, and a mature protein of 684 aa, MW: 75,768 Da, is formed after cleavage [14]. There is a sequence ahead of foxA with homology to the putative ferric ion uptake regulator (Fur) of Yersinia [21]. The expression of foxA may be regulated by iron via the Fur protein, as in other known siderophore receptors [14]. Fur may be a transcription inhibition protein acting on the ferric regulation promoter using Fe2+-dependent DNA binding activity homologous to that in E. coli [22-25].

Analysis of ail

The entire ail ORF for 271 pathogenic Y. enterocolitica strains isolated from China and 10 reference strains were analyzed and compared to strain 8081. The data showed that all the strains can be divided into 3 sequence patterns. The Chinese isolates, 270 strains (70 of serotype O:3 and 200 of serotype O:9) and 7 reference strains (5 of O:3, one of O:9 and one of O:5,27), were sequentially identical and formed pattern A1. Four highly pathogenic strains of serotype 1B/O:8 showed identical sequences and formed pattern A2. Compared to pattern A1, pattern A2 showed 21 base mutations among which 9 were sense and 12 were nonsense mutations. In addition, one pathogenic Chinese isolate O:9 serotype (isolated from the tongue of a rat in Ningxia, 1997) showed 3 base mutations compared to the entire ail of pattern A1, one sense and 2 nonsense; it formed pattern A3 (Fig. 1). This new ail genotype was submitted to Genbank and given the accession number GU722202.

Figure 1

Sequence polymorphism in . Each number on the scale indicates the site number in the ORF; red letters indicate the mutated bases; the yellow regions are missense mutations; and the other mutations are nonsense.

Analysis of foxA

Analysis of the primary coding region of foxA from nt 28 to nt 1,461 in 271 pathogenic Y. enterocolitica strains isolated from China and 11 reference strains showed that all the strains can be divided into 3 groups including 8 sequence patterns (Fig. 2). Group I comprised patterns F1, F2 and F3 and included 201 serotype O:9 strains isolated from China and 2 reference strains (one strain O:9 and one O:5,27). Among these, 199 strains of O:9 and 2 reference strains showing identical sequences formed pattern F1, and 2 other strains of O:9 with 2 and 3 base mutations formed patterns F2 and F3. Group II comprised patterns F4 and F5, and included 70 Chinese isolates and 5 reference strains of serotype O:3. Sixty-nine serotype O:3 strains (67 Chinese isolates and2 reference strains) showing identical sequences formed pattern F4; and 6 other strains of O:3 had one base mutation and formed pattern F5. Group III comprised five reference strains including patterns F6, F7 and F8. Pattern F6 (2 Japanese strains) had 2 base mutations compared to pattern F7 (52211). Compared to pattern F7, pattern F8 (8081) had 5 base mutations (Fig. 3).

Figure 2

Phylogenetic tree of . Among the 309 isolates studied, 282 were pathogenic and the others were nonpathogenic. [No.]: the number of the strains of the same serotype in the pattern.

Figure 3

Sequence polymorphism in . The numbers on the scale indicates the site numbers in the ORF; red letters indicate mutated bases; the yellow regions are missense mutations; and the other mutations are nonsense.

To analyze foxA polymorphism in Y. enterocolitica overall, we chose 27 strains of non-pathogenic Y. enterocolitica as controls (Table 1). The results showed 13 sequence patterns for the 27 strains with 10's to 100's more polymorphic sites and no apparent regularity. This indicated that foxA was less polymorphic and more conserved in pathogenic strains than in non-pathogenic strains.

Discussion

Only pathogenic Y. enterocolitica contains ail, which confers a bacterial invasion and serum resistance phenotype, that is an important virulence marker on the chromosome [6,19]. The entire ORF of ail was sequenced and analyzed from strains from different sources and biotypes and serotypes. The data showed that the 282 pathogenic Y. enterocolitica formed 3 sequence patterns (Fig. 1); the strains were pathogenic O:3 and O:9 isolated from various hosts in China and the reference strains. Only one Chinese isolate formed pattern A3, a new ail genotype submitted to Genbank and given the GenBank accession number GU722202. When it was compared to the sequence of pattern A1, three base mutations were found, one sense and two nonsense. We presume that pathogenic Y. enterocolitica had 2 original ail patterns, A1 represented in serotypes O:3 and O:9 and A2 represented in bio-serotype 1B/O:8; pattern A3 may be a mutation of A1. Pathogenic Y. enterocolitica can be divided into a high-pathogenicity group (Y. enterocolitica biogroup 1B) and a low-pathogenicity group (Y. enterocolitica biogroups 2 to 5) on the basis of the lethal infectious dose in the mouse model [26]. The typing of ail in this study is consistent with this grouping of pathogenic strains. Therefore, we presumed that lethality in mice (50% lethal dose <1,000 microorganisms) not only depends on the presence of the yersiniabactin (ybt) locus, which carries the genes for biosynthesis, transport and regulation of the siderophore yersiniabactin [8,27-29], but is also closely related to the ail sequence. We conjecture that synergism with ail is necessary for Y. enterocolitica pathogenesis.

ail is not only an important virulence gene for pathogenic Y. enterocolitica, but also harbors highly conserved sequences, mutation of which may change the virulence of the bacterium. For instance, in the 1B/O:8 strain, which is highly lethal to mice, the ail belongs to pattern A2, while ail in other pathogenic bioserotype strains belongs to pattern A1. So we believe that a change in ail is closely related to the pathogenesis of the strain. A pathogenic O:9 strain isolated from Cricetulus triton in Ningxia contains ail pattern A3, the sequence of which has 3 site mutations, only one being a sense mutation. Further study is needed to see whether amino acid change alters the function of Ail protein or bacterial virulence.

Analysis of the 1,434 base pairs of the foxA primary coding region showed that the foxA sequence correlated with the biotype and serotype of pathogenic Y. enterocolitica. Comparing the primary sequences of groups I and II, 13 base mutations at fixed positions were found; 5 were sense and 8 were nonsense mutations, indicating that the primary difference in the pathogenic Y. enterocolitica foxA was located in these 13 sites. Strain 8081 showed 26 base mutations compared to F1 and 31 compared to F4. From these findings we presume that pathogenic O:3 and O:9 have similar foxA sequences (Fig. 2) and mutation sites additional to strain 8081 bio-serotype 1B/O:8 (Fig. 3). Thus, there is a correlation between pathogenesis and the different bio-serotypes of Y. enterocolitica. More mutation sites and no obvious regulation were found in non-pathogenic Y. enterocolitica foxA, although some strains showed an identical foxA sequence type (Fig. 2). The identical sequence patterns of the pathogenic Y. enterocolitica strains isolated from different areas, at different times and from different host sources show the foxA sequence pattern to be closely correlated to pathogenesis. Unfortunately, fewer strains from outside China were used, which is a limitation of the study and needs adding strains for future study.

ail is a primary marker for pathogenic Y. enterocolitica and is an important tool for detecting it, making it a very important gene to analyze. Some scholars have established a real-time PCR assay to detect Y. enterocolitica using ail or ystA as the target gene [30-33]. According to the current identification standards, strains having no ail and harboring ystB isolated from diarrhea patients are classified as non-pathogenic. However, other researchers believe that strains harboring ystB are pathogenic and cause the diarrhea, as inferred from epidemiology and the etiology of disease outbreaks [34,35]. Therefore, it is necessary to detect other conserved genes as a complement to these markers because reliance on only ail and ystA to detect pathogenesis remains questionable. We chose to detect foxA, which is found in both pathogenic and non-pathogenic Y. enterocolitica. The results showed that both ail and foxA were conserved together in pathogenic strains and can therefore be used to confirm the detection of pathogenic Y. enterocolitica. Currently, we are attempting to extract bacterial DNA from clinical specimens to detect foxA in order to identify Y. enterocolitica directly from humans and other animals; and we have some preliminary data (unpublished).

Almost all Y. enterocolitica carry foxA while pathogenic strains carry ail. It is very important for real-time PCR detection of Y. enterocolitica to study sequence polymorphism in ail and foxA. It will be helpful to design specific primers and probes in the conserved region in order to develop real-time or traditional PCR methods. We are trying to establish a duplex real-time PCR to detect Y. enterocolitica from clinical samples and to confirm its pathogenicity. Designing specific primers for foxA and ail in a combined detection system is valuable for increasing sensitivity and specificity in the detection of pathogenic Y. enterocolitica.

Conclusion

Analysis of polymorphisms in ail and foxA of pathogenic Y. enterocolitica strains from different times and regions showed ail to be an important virulence gene for pathogenic Y. enterocolitica, and that it has a highly conserved sequence. The gene encoding the ferrioxamine receptor, foxA, is also conserved in pathogenic strains, where 2 primary sequence patterns were found. More strains from outside China are needed for further study.

Authors' contributions

YH did most of the PCR work and DNA sequencing. XW analyzed the sequences. ZC did the data clustering and construction of phylogenetic trees. YY and YX identified the biotypes and serotypes of strains. LT wrote the paper. BK and XJ participated in discussion of the study. HJ designed and coordinated the study and drafted the manuscript. All the authors read and approved the final manuscript.