RESEARCH ARTICLE

Investigation of virulence genes in clinical isolates of Yersinia enterocolitica Haoxuan Zheng, Yong Sun, Zhengguo Mao & Bo Jiang Institute of Digestive Disease, Nanfang Hospital, Southern Medical University, Guangzhou, China

Correspondence: Bo Jiang, Institute of Digestive Disease, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China. Tel./fax: 186 20 6164 1541; e-mail: [email protected] Received 16 March 2008; revised 16 April 2008; accepted 8 May 2008. First published online 16 June 2008. DOI:10.1111/j.1574-695X.2008.00436.x Editor: Patrick Brennan Keywords investigation; virulence genes; Yersinia enterocolitica .

Abstract In this study, we aimed to investigate the distribution of virulence genes in clinical isolates of pathogenic Yersinia enterocolitica. Two thousand six hundred stool samples were collected from 2600 patients with diarrhea, and were tested using the culture method and real-time PCR. Then, all isolates of pathogenic Y. enterocolitica cultured from the culture method were examined for virulence genes (inv, ail, ystA, ystB, ystC, yadA, virF) by PCR and for the presence of plasmid by four phenotypic tests. As a result, 160 pathogenic strains were successfully detected by the culture method, including bio/serotype 1A/unknown (4), 1B/unknown (8), 2/O:9 (39), 2/unknown (7), 3/O:3 (22), 3/unknown (6), 4/O:3 (55), 4/unknown (10) and 5/unknown (9). The positive rate of virulence genes tested in 160 isolates was inv (100%), ail (94%), ystA (93%), ystB (7.5%), ystC (5%), yadA (89%) and virF (82%) while the phenotypic test included autoagglutination (87%), binding of crystal violet (89%), calcium-dependent growth (74%) and Congo red absorption (78%), respectively. Finally, we found that not all pathogenic Y. enterocolitica necessarily carry all traditional virulence genes in both chromosomes and plasmids to cause illness. Perhaps, some of them, lacking some traditional virulence genes, contain other unknown virulence markers that interact with each other and play an important role in the diverse pathogenesis of pathogenic Y. enterocolitica.

Introduction Yersinia enterocolitica is a food-borne pathogen causing symptoms such as fever, diarrhea, nausea, vomiting and abdominal pain (Feng & Weagant, 1994; Bottone, 1997), and causes diseases that range from self-limiting gastroenteritis to fatal septicemia (Feng & Weagant, 1994). Extraintestinal complications caused by virulent Y. enterocolitica such as liver and spleen abscesses (Rabson et al., 1975), pneumonia (Portnoy & Martinez, 1979), septic arthritis (Taylor et al., 1977), meningitis (Sonnenwirth, 1970), empyema (Clarridge et al., 1983) and endocarditis (Urbano-Marquez et al., 1983) have been widely reported. Human yersiniosis is attributed to contaminated pork, milk, water and tofu consumption as well as blood transfusion (Feng & Weagant, 1994). Generally, Y. enterocolitica can be classified into biotype 1A, regarded as nonpathogenic (Bottone, 1997), and the pathogenic biotypes 1B, 2, 3, 4 and 5. Both species can also be divided into different serotypes with predictive values for pathogenicity. Serological and biochemical classifications, 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

however, are time-consuming and laborious, and are not usually available in routine laboratories. Alternative phenotypical tests (Yang & Fang, 2003), such as calciumdependent growth at 37 1C, Congo red binding, pyrazinamidase testing and autoagglutination testing, all have limited predictive value for the pathogenicity of Y. enterocolitica. Moreover, the above tests are frequently ambiguous to read and their outcome may be unreliable, because they depend on the presence and expression of (plasmid-borne) virulence genes and the virulence plasmid pYV can easily be lost depending on the culture conditions. Therefore, differentiation of pathogenic strains should not rely solely on the expression or the detection of the virulence plasmid but also on the detection of chromosomal virulence factors (Thoerner et al., 2003). In recent years, PCR assays and other molecular methods have been developed as efficient tools for identifying pathogenic Y. enterocolitica, targeting chromosomal genes such as ail (attachment invasion locus, which mediates cell invasion) (Miller et al., 1989, 1990), inv (invasive gene, which FEMS Immunol Med Microbiol 53 (2008) 368–374

369

Investigation of virulence genes in Y. enterocolitica

mediates cell invasion) (Rasmussen et al., 1994), ystA (which is responsible for the production of a heat-stable enterotoxin in virulent Y. enterocolitica) (Delor et al., 1990), ystB and ystC (which have been observed to encode an enterotoxin present mainly in avirulent Y. enterocolitica) (Ramamurthy et al., 1997) and rfbC (which can be used to identify pathogenic Y. enterocolitica O:3 strains) (Weynants et al., 1996), as well as plasmid genes such as yadA (whose product is involved in autoagglutination, serum resistance and adhesion) (Skurnik & Wolf-Watz, 1989), and virF or lcrF (for Y. enterocolitica and Yersinia pseudotuberculosis, respectively, which encodes transcriptional activators of the yop regulon) (Cornelis et al., 1989). In addition, molecular methods have also been used for typing and in epidemiological studies of Y. enterocolitica, including restriction analysis of both plasmids and chromosomes, randomly amplified polymorphic DNA analysis, ribotyping and pulsed-field gel electrophoresis (Falca˜o et al., 2006). In the past, some researchers (Thoerner et al., 2003; Thisted Lambertz & Danielsson-Tham, 2005; Falca˜o et al., 2006) have conducted the analysis of distribution of several virulence genes in Y. enterocolitica with limited clinical isolates, but they did not specify that whether or not it was necessary for pathogenic Y. enterocolitica to carry all traditional virulence genes in both chromosomes and plasmids to cause illness. In this study, considering the above question, we investigated the distribution of virulence markers (inv, ail, ystA, ystB, ystC, yadA, virF) in 160 pathogenic strains of Y. enterocolitica cultured from patients with diarrhea, using common PCR and four phenotypic tests.

Materials and methods Samples Two thousand six hundred diarrheal stool samples (from 2600 patients, males : females = 1540 : 1060, children : adults = 1446 : 1154, spring : summer : autumn : winter = 469 : 500 : 775 : 856) were randomly and continuously collected from May 2005 to January 2008 from in- and outpatients in Nanfang Hospital (1600 samples), Zhujiang Hospital (500 samples) and Guangdong Provincial People’s Hospital (500 samples), who defecated 4 3 times per day accompanied by changes of stool character, e.g. wateriness. All samples were used for clinical investigation with permission. Stool samples (about 1 g each) were kept in Cary–Blair solution (5 mL) and immediately sent to the laboratory.

was added to the above liquid when cooled down to 50 1C. The solution pH was adjusted to 8.4. Finally, the solution was autoclaved (15 min at 121 1C) and then aseptically transferred to a sterile test tube.

Yersinia -selective enrichment and Cefsulodin--Irgasan--Novobiocin (CIN) plate (YSE--CIN) Briefly, tryptone (Oxide) 10 g, MOPS (Sigma) 2.75 g, Tris (Sigma) 2.75 g, D-sorbital (Sigma) 20 g and no. 3 bile salt (Sigma) 1.5 g were dissolved in 1 L of double distilled water. The solution was autoclaved (15 min at 121 1C) and then cooled down to room temperature. Nine hundred and eighty-three milliliters of solution was aseptically transferred to a flask of suitable size and the following ingredients were added: two vials of Yersinia-selective supplement (CIN, Oxide Cat. No. SR0190E) and 5 mL 0.8 M sterilized KClO3 solution. The solution pH was adjusted to 8.4  0.1 using sterile 1 M KOH (Sigma) solution (c. 10 mL). The content of YSE was modified slightly compared with Knutsson et al., 2002. A CIN plate (Oxide, Cat. No. CM 0653B) was prepared according to the manufacturer’s instructions. The YSE–CIN method was more efficient than the other three traditional culture methods by the McNemer test in a previous study (Zheng et al., 2006).

Culture method Stool (0.5 g) was added to normal saline (4.5 mL). After thorough homogenization, 500 mL was placed into YSE for incubation (48 h, 16 1C) and then 5 mL was streaked on a CIN plate (24 h, at 28 1C). Yersinia enterocolitica colonies on the CIN plate presented as a 1–2 mm cardinal red center surrounded by a diaphanous toroidal ring with a limpid rim. After examination by three experienced researchers, all suspected Y. enterocolitica colonies were subjected to 15 biochemical tests (supplied by Me´ rieux Foundation) including the growth test 37 1C ( ) and 25 1C (1), V–P 37 1C ( ) and 25 1C(1), oxidase ( ), phenylalanine ( ), mannitol (1), sorbitol (1), cellobiose (1), raffinose ( ), melibiose ( ), ornithine decarboxylase (1), lysine decarboxylase ( ), simmons citrate ( ) and koser citrate (1). The strain, consistent with the above as well as positive for both esculin (25 1C) and salicin (37 1C), was regarded as avirulent Y. enterocolitica, while negative ones were considered as virulent Y. enterocolitica (Yang & Fang, 2003). All tests were performed three times under the same conditions.

Cary--Blair solution Generally, sodium thioglycollate (Sigma) 1.5 g, dibasic sodium phosphate (Sigma) 1.1 g and sodium chloride (Sigma) 5 g were heated and dissolved in 991 mL of double distilled water and then 1% calcium chloride solution 9 mL FEMS Immunol Med Microbiol 53 (2008) 368–374

Real-time PCR: specific primer and probe, and stool DNA extraction The primers and probe sequences (targeting ystA, which exists only in virulent Y. enterocolitica) reported by (Ibrahim 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

370

et al., 1997) were modified to suit the requirements of the assay and synthesized as in previous studies (Zheng et al., 2006, 2007). All stools were selectively enriched in YSE at 16 1C for 18 h. Compound (200 mL) was pipetted into a microtube and DNA extraction was performed as before. Finally, a 5-mL aliquot of extraction served as the template for each realtime (RT)-PCR amplification. The RT-PCR components and conditions were the same as before. Each set of reactions included triplicate wells that were no-template controls, negative controls and positive controls and were performed three times. The reaction was performed on an M-J Opticon-2 DNA Engine using the equation DRQ = RQ1 RQ (Bassler et al., 1995). A positive interpretation for pathogenic Y. enterocolitica was based on a threshold of four times the average DRQ value of negative controls (stool without virulent Y. enterocolitica) (Sanchez-Vizcaino & Cambro-Alvarez, 1987), which enabled us to eliminate several other bacteria that fell in the positive interpretation range.

Classification of biogroup in Y. enterocolitica Classification of six biogroups (1A, 1B, 2, 3, 4, 5) was performed on all clinical isolates with biochemical tests (supplied by Me´ rieux Foundation) according to the standard made by Wauters (Wauters et al., 1987; Bottone, 1997).

Detection of virulence genes by PCR All clinical isolates from the culture method (Table 1) and reference strains (Table 2) were investigated for the presence of virulence genes three times with empty and negative controls, such as inv, ail, ystA, ystB, ystC, yadA, virF and rfbC. PCR agents and conditions for each gene were referred to the following papers: inv (Rasmussen et al., 1994), ail (Nakajima et al., 1992), ystA (Ibrahim et al., 1997), ystB (Ramamurthy et al., 1997), ystC (Ramamurthy et al., 1997), yadA (Feng & Weagant, 1994), virF (Thoerner et al., 2003) and rfbC (Weynants et al., 1996).

Reference strains and diagnostic serum All strains (Table 2) from American Type Culture Collection (ATCC), Chinese Medical Culture Center (CMCC) and Centers for Disease Control and Prevention (CDC) were used as reference strains. Diagnostic serum (Y. enterocolitica O:3 and O:9) was purchased from Statens Serum Institut as before. The serodiagnostic test was performed according to the manufacturer’s instructions. 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

H. Zheng et al.

Phenotypic tests The phenotypic tests, including temperature-dependent autoagglutination, calcium-dependent growth, Congo red absorption, pyrazinamidase production and binding of crystal violet, were performed three times on all clinical isolates and some reference strains (virulent and avirulent Y. enterocolitica) as described (Yang & Fang, 2003). Pyrazinamidase production should be negative for virulent Y. enterocolitica while the other four tests should be positive for the presence of plasmid in virulent Y. enterocolitica (Yang & Fang, 2003).

Results and discussion All 2600 diarrheal stools were detected by RT-PCR and the culture method (YSE–CIN). The total number of positive samples was 188. Among these samples, 178 positive samples were confirmed by RT-PCR while 160 were identified by the culture method. All 160 clinical isolates from the culture method, such as four strains of biotype 1A (traditionally regarded as avirulent) and some strains of unknown serotypes (despite lack of traditionally important virulence markers), were fully confirmed to be virulent by biochemical tests. As summarized in Table 3, 10 positive samples (one strain of bio/serotype 1A/unknown, three strains of 2/ unknown, two strains of 4/O:3, one strain of 4/unknown, three strains of 5/unknown) confirmed by the culture method could not be identified by RT-PCR while 28 samples positive in RT-PCR were negative in the culture method. Generally, PCR inhibitors found in tissues and feces may have significant adverse effects on the efficiency and sensitivity of PCR-based assays (Chen et al., 1997). Therefore, these 10 negative samples in RT-PCR might contain inhibitors that may prevent a molecular reaction. However, subsequent common PCR performed on these 10 isolates found none of them with ystA but ystB1 or ystC1, which showed that these 10 virulent strains did not carry the ystA gene so that it was impossible for RT-PCR (targeting ystA) to detect them. On the other hand, 28 negative samples in the culture method statistically showed its lower sensitivity than RT-PCR. As a whole, all 160 isolates from the culture method were examined for virulence genes in chromosomes and plasmids by common PCR. The distribution of bio/serotype in 160 pathogenic strains (Table 1) was 1A/unknown (4), 1B/ unknown (8), 2/O:9 (39), 2/unknown (7), 3/O:3 (22), 3/ unknown (6), 4/O:3 (55), 4/unknown (10) and 5/unknown (9) while the positive rate of virulence genes tested (Table 1) was inv (100%), ail (94%), ystA (93%), ystB (7.5%), ystC (5%), yadA (89%), virF (82%) and rfbC (49%), respectively. Theoretically, pathogenic strains should contain all virulence genes in chromosomes (inv, ail, ystA) and plasmids (yadA, virF), all of which may interact with each other to FEMS Immunol Med Microbiol 53 (2008) 368–374

371

Investigation of virulence genes in Y. enterocolitica

Table 1. Distribution of virulence genes in clinical isolates of Yersinia enterocolitica Bio/serotype (number) 1A/unknown (4) 2 strains 1 strain 1 strain 1B/unknown (8) 4 strains 3 strain 1 strain 2/O:9 (39) 35 strains 4 strains 2/unknown (7) 4 strains 3 strains 3/O:3 (22) 18 strains 3 strain 1 strain 3/unknown (6) 4 strains 2 strains 4/O:3 (55) 45 strains 5 strains 3 strains 2 strains 4/unknown (10) 4 strains 3 strains 2 strains 1 strain 5/unknown (9) 6 strains 2 strains 1 strains Total (160)

inv

ail

4 1 1 1 8 1 1 1 39 1 1 7 1 1 22 1 1 1

1

ystA 3 1

6 1 1 55 1 1 1 1 10 1 1 1 1 9 1 1 1 160

1 8 1 1 1 39 1 1 7 1 1 22 1 1 1 4 1 55 1 1 1 1 8 1 1 1 6 1

1 8 1 1 1 39 1 1 4 1 22 1 1 1 6 1 1 53 1 1 1 7 1 1 1

ystB

ystC

4 1 1 1 1

1

yadA 2 1

0

4

3

1 0

1 0

1

0

1 0

0

0

2

1

3

rfbC 0

1 1

1

1 0

virF

7

4

0

1

39 1 1 7 1 1 21 1 1

35 1

0

7 1 1 18 1

0

6 1 1 45 1

4 1

7

22 1 1 1 0

55 1 1 1 1 0

1 1

Binding of crystal violet

3 1 1

1 1

6 1 1 50 1 1

Autoagglutination

Calciumdependent growth

3

2

1 1 7

Congo red absorption

1

7

Pyrazinamidase production

3

0

1 1 4

0

1 1

1 1

1

4 1

39 1 1 3

35 1

35 1

0

4

0

1 21 1 1

39 1 1 7 1 1 21 1 1

18 1

0

6 1 1 50 1 1

6 1 1 50 1 1

4

3

7 1 1 18 1

1

4

0

1

4 1

45 1

50 1 1

0

7

3

3

1 1

1 1

1

1

1 6

2

1 1

1

6

6

1

1

142

131

1

6

0

0

1

1

6 1

139

142

118

125

1 4

1 150

148

12

1 8

1 78

The positive rate of virulence genes tested in 160 isolates was inv (100%), ail (94%), ystA (93%), ystB (7.5%), ystC (5%), yadA (89%), virF (82%) and rfbC (49%) while the phenotypic test included autoagglutination (87%), binding of crystal violet (89%), calcium-dependent growth (74%), Congo red absorption (78%) pyrazinamidase production (2.5%), respectively. PCR results were consistent three times.

cause illness in humans. However, in this study, we found that only inv existed in all 160 clinical isolates whereas ail, ystA, yadA and virF were present in some of them, more or less. Most of the 160 strains were positive for inv, ail, ystA, yadA and virF, but others were positive for only some of them, which could still be pathogenic to humans, indicating that perhaps not all virulence genes were required to exist and be expressed to cause illness, or the presence of specific virulence genes might closely correlate to the bio/serotypes of Y. enterocolitica. Moreover, pathogenic Y. enterocolitica may contain other unknown important virulence genes located in the chromosome and plasmid, which played a key role in pathogenesis in association with traditional genes FEMS Immunol Med Microbiol 53 (2008) 368–374

confirmed by PCR. On the other hand, one question arises as to whether the virulence gene in the chromosome would be lost just as in the plasmid due to the conditions in the culture method. Although loss of plasmid in Y. enterocolitica is universal, loss of virulence genes in the chromosome is rarely reported. If it really exists, it could be because of various factors including media and temperature. Altogether, loss or lack of some traditional virulence markers in the chromosome is needed to carry out further research. The distribution of virulence markers in biotype 1A (Table 1, confirmed to be virulent by biochemical tests, regardless of different virulence genes and serotypes) was two strains (inv1, ystA1, ystB1, yadA1, virF1), one strain 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

372

H. Zheng et al.

Table 2. Distribution of virulence gene in reference strains

Bio/serotype (number)

inv

ail

ystA

ystB

ystC

yadA

virF

rfbC

Virulent Y.e (4) 4/O:3 2/O:9 2/O:5,27 1B/O:8 Avirulent Y.e (2) 1A/O:6,30 1A/O:7,8 Other species of yersinia (3) Y. intermediaic Y. kristenseniic Y. pseudotuberculosis Other enteric bacteria (4) E. coli O:157 Campylobacter jejuni Pseudomonas aeruginosa

4 1 1 1 1 0

4 1 1 1 1 0

4 1 1 1 1 0

0

0

4 1 1 1 1 0

1 1

0

4 1 1 1 1 0

0

0

0

0

1

1

0

0

0

0

2 1 1 0

0

0

1 0

1 0

0

Autoagglutination

Binding of crystal violet

4 1 1 1 1 0

3 1 1 1 0

ND ND ND ND ND ND

Calciumdependent growth

Congo red absorption

Pyrazinamidase production 0

1 1 0

4 1 1 1 1 0

ND ND ND

ND ND ND

ND ND ND

ND ND ND

ND ND ND

ND ND ND

ND ND ND

ND ND ND

3 1

2 1 1

0

ND, not done. Results from all reference strains were normal.

Table 3. Detection rate of two methods by the McNemer test Culture method RT-PCR

1

1

150 10 160

Total

Total 28 2412 2440

178 2422 2600

The detection rate of RT-PCR was significantly higher than the culture method by the McNemer test (w2 = 7.61, P 4 0.05).

(inv1, ystB1, ystC1, virF1) and one strain (inv1, ail1, ystA1, ystB1). YstB1 has been reported in nonpathogenic biotype 1A whereas inv1, ail1, ystA1, yadA1 and virF1 have seldom been described by other researchers. In this study, we found four strains of biotype 1A with different virulence markers besides ystB1, similar to (Thoerner et al., 2003), which indicated that some strains of biotype 1A were virulent and indeed could cause illness in humans. Interestingly, virF was positive in one strain while yadA was negative; both of these were located in the plasmid. Furthermore, no plasmid was detected in one strain whereas both yadA and virF were positive in another strain. The reason for this was unknown but it may be related to different pathogenesis in biotype 1A. In the past, biotype 1A strains, possibly lacking virulence markers (Pierson & Falkow, 1990; Delor & Cornelis, 1992), were thus regarded as avirulent. Nevertheless, biotype 1A strains have been reported to be associated with nosocomial and food-borne outbreaks of diarrhea (Greenwood & Hooper, 1990; Butt et al., 1991) and were once again 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

confirmed to be virulent in this study. Previous research has also indicated that gastroenteritis caused by biotype 1A strains was indistinguishable from that caused by primary pathogenic biotypes (Morris et al., 1991; Burnens et al., 1996). In summary, according to the above data, we speculate that a small part or some of biotype 1A strains, carrying virulence genes similar to strains of biotype 1B and 2 –5 as well as ystB1 or other unknown virulence markers, may perhaps be pathogenic to humans, such as in this study, while further research is needed with respect to pathogenesis of biotype 1A that may vary from other biotypes. Of the other 156 strains of biotype 1B and 2–5 (Table 1, confirmed to be virulent by biochemical tests, regardless of different virulence genes and serotypes), the predominant genotype (120 strains) was inv1, ail1, ystA1, yadA1 and virF1, with traditional virulence markers in both the chromosome and the plasmid, whereas the other 36 clinical isolates were too diverse to draw a conclusion. YstB1 or ystC1 sporadically occurred in one strain of 1B/unknown, four strains of 2/O:9, three strains of 2/unknown, one strain of 3/O:3, three strains of 4/unknown and three strains of 5/ unknown, which showed that ystB1 or ystC1 may not be refined to biotype 1A (Thoerner et al., 2003; Singh & Virdi, 2004). Surprisingly, RfbC was found in one strain of 5/ unknown besides all O:3 strains but the serodiagnostic test indicated it as being negative. Perhaps rfbC existed in this strain but no antigen of O:3 occurred. Moreover, positive for yadA (89%, Table 1) or virF (82%, Table 1) also presented FEMS Immunol Med Microbiol 53 (2008) 368–374

373

Investigation of virulence genes in Y. enterocolitica

(Thoerner et al., 2003), as in biotype 1A, in three strains of 1B/unknown, four strains of 2/O:9, three strains of 3/O:3, four strains of 4/O:3 and three strains of 4/unknown while no plasmid was found in 13 strains, including one strain of 1B/unknown, one strains of 3/O:3, 5 strains of 4/O:3, three strains of 4/unknown and three strains of 5/unknown. The presence of plasmid is a prerequisite for full virulence of Y. enterocolitica, which codes for an array of proteins and, in a virtual stepwise fashion, guides the invading yersinial pathogen past numerous host defense mechanisms to establish itself within its ecologic niche extracellularly or within macrophages (Bottone, 1997). Thereby, the plasmid in virulent strains is easy to understand whereas the presence of either yadA or virF in the plasmid (Thoerner et al., 2003) and absence of a plasmid in virulent strains were rare. As is known, the plasmid may be lost when the temperature in the culture method is above 30 1C, but the temperature in this study was set o 30 1C so that there is no possibility of loss of plasmid caused by temperature. However, whether the plasmid may be lost below 30 1C is unknown and worth further research. Perhaps, other unknown virulence genes in plasmids with a function to similar yadA or virF may exist and may play a role in the illness instead of the possibly absent yadA or virF while the finding of 13 pathogenic strains without yadA and virF may probably contribute to the loss of plasmid due to some unknown reasons. Also, it is possible that virulent strains without a plasmid may mediate the invasion in humans only by traditional and other unknown virulence genes in the chromosome, which is a hypothesis and should be validated in the future. Anyway, due to different distributions of virulence markers, especially the conflicting occurrence of yadA and virF in the plasmid (Thoerner et al., 2003), it is certain that pathogenesis in virulent Y. enterocolitica is not single but diverse. The positive rate of phenotypic tests performed on 160 isolates was autoagglutination (87%), binding of crystal violet (89%), calcium-dependent growth (74%), Congo red absorption (78%) and pyrazinamidase production (2.5%). For a long time, phenotypic tests have been regarded as time-consuming and laborious and their results were unreliable due to subjective and objective factors. Hence, three experienced technicians were employed to perform all the biochemical tests above and phenotypic tests here. Among 160 isolates, 146 were positive for the presence of plasmid by PCR (either yadA1 or virF1), and hence the positive rate of phenotypic tests can be revised as follows: autogglutination (95%), binding of crystal violet (97%), calcium-dependent growth (81%) and Congo red absorption (86%). In addition, 14 isolates without a plasmid by PCR (neither yadA1 nor virF1) were negative in phenotypic tests. Interestingly, one strain of 1A/unknown without yadA was positive in autoagglutination while four strains of 2/unknown with yadA were negative, which could not be explained by this FEMS Immunol Med Microbiol 53 (2008) 368–374

study but indicates that autoagglutination may relate to other unknown genes except yadA. The above results were slightly different from other studies (with a lower positive rate) (Falca˜o et al., 2006). Perhaps, this may be correlated to the strains from different sources and areas as well as level of personal operation. Generally, these results showed a wellestablished correlation between virulence markers and all phenotypic tests while the inconsistency between them may be that these genes were silenced by mutation or that the conditions used in phenotypic tests were not appropriate to induce expression of virulence markers in some strains or that some unknown virulence genes participated in phenotypic tests when the traditional ones were absent or not functional. In this study, we conducted an investigation of the distribution of virulence genes (inv, ail, ystA, ystB, ystC, yadA, virF, rfbC) in 160 pathogenic strains of Y. enterocolitica from patients with diarrhea, using common PCR and four phenotypic tests, in order to find out whether or not pathogenic Y. enterocolitica would carry all traditional virulence genes in both chromosomes and plasmids that cause illness. In summary, from the above results, we found that not all pathogenic Y. enterocolitica necessarily carry all traditional virulence genes in both chromosomes and plasmids to cause illness whereas some of them, lacking some traditional virulence genes, may contain other unknown virulence markers that interact with each other and play an important role in the diverse pathogenesis of pathogenic Y. enterocolitica.

References Bassler HA, Flood SA, Livak KJ, Mawmaro J, Knorr R & Batt CA (1995) Use of a fluorogenic probe in a PCR-based assay for the detection of Listeria monocytogenes. Appl Environ Microbiol 6: 3724–3728. Bottone E (1997) Yersinia enterocolitica: the charisma continues. Clin Microbiol Rev 10: 257–276. Burnens AP, Frey A & Nicolet J (1996) Association between clinical presentation, biogroups and virulence attributes of Yersinia enterocolitica strains in human diarrheal disease. Epidemiol Infect 116: 27–34. Butt HL, Gordon DL & Lee-Archer T (1991) Relationship between clinical and milk isolates of Yersinia enterocolitica. Pathology 23: 153–157. Chen S, Yee A, Griffiths M, Larkin C, Yamashiro CT, Behari R, PaszkoKolva C, Rahn K & Grandis de SA (1997) The evaluation of a fluorogenic polymerase chain reaction assay for the detection of Salmonella species in food commodities. Int J Food Microbiol 35: 239–250. Clarridge J, Roberts C, Peters J & Musher D (1983) Sepsis and empyema caused by Yersinia enterocolitica. J Clin Microbiol 17: 936–938. Cornelis G, Sluiters C & de Rouvroit CL (1989) Homology between VirF, the transcriptional activator of the Yersinia

2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

374

virulence regulon, and AraC, the Escherichia coli arabinose operon regulator. J Bacteriol 171: 254–262. Delor I & Cornelis GR (1992) Role of Yersinia enterocolitica Yst toxin in experimental infection of young rabbits. Infect Immun 60: 4269–4277. Delor I, Kaeckenbeeck A & Wauters G (1990) Nucleotide sequence of yst, the Yersinia enterocolitica gene encoding the heat-stable enterotoxin, and prevalence of the gene among pathogenic and nonpathogenic yersiniae. Infect Immun 58: 2983–2988. Falca˜o JP, Falca˜o DP & Pitondo-Silva A (2006) Molecular typing and virulence markers of Yersinia enterocolitica strains from human, animal and food origins isolated between 1968 and 2000 in Brazil. J Med Microbiol 55: 1539–1548. Feng P & Weagant SD (1994) Yersinia. Food-Borne Diseases Handbook: Diseases Caused by Bacteria (Hui YH, Gorham JR, Murrell KD & Cliver DO, eds), pp. 427–460. Dekker, New York. Greenwood MH & Hooper WL (1990) Excretion of Yersinia spp. associated with consumption of pasteurized milk. Epidemiol Infect 104: 345–350. Ibrahim A, Liesack W, Griffiths MW & Robins-Browne RM (1997) Development of a highly specific assay for rapid identification of pathogenic strains of Yersinia enterocolitica based on PCR amplification of the Yersinia heat-stable enterotoxin gene (yst). J Clin Microbiol 35: 1636–1638. Knutsson R, Fontanesi M & Grage H (2002) Development of a PCR-compatible enrichment medium for Yersinia enterocolitica: amplification precision and dynamic detection range during cultivation. Int J Food Microbiol 72: 185–201. Miller VL, Farmer III JJ & Hill WE (1989) The ail locus is found uniquely in Yersinia enterocolitica serotypes commonly associated with disease. Infect Immun 57: 121–131. Miller VL, Bliska JB & Falkow S (1990) Nucleotide sequence of the Yersinia enterocolitica ail gene and characterization of the Ail protein product. J Bacteriol 172: 1062–1069. Morris JG, Prado V & Ferreccio C (1991) Yersinia enterocolitica isolated from two cohorts of young children in Santiago, Chile: incidence of and lack of correlation between illness and proposed virulence factors. J Clin Microbiol 29: 2784–2788. Nakajima H, Inoue M & Mori T (1992) Detection and identification of Yersinia pseudotuberculosis and pathogenic Yersinia enterocolitica by an improved polymerase chain reaction method. J Clin Microbiol 30: 2484–2486. Pierson DE & Falkow S (1990) Nonpathogenic isolates of Yersinia enterocolitica do not contain functional inv-homologous sequences. Infect Immun 58: 1059–1064. Portnoy D & Martinez LA (1979) Yersinia enterocolitica septicemia with pneumonia. Can Med Assoc J 120: 61–62. Rabson AR, Hallett AF & Koornhof HJ (1975) Generalized Yersinia enterocolitica infection. J Infect Dis 131: 447–451. Ramamurthy T, Yoshino KI & Huang X (1997) The novel heat-stable enterotoxin subtype gene (ystB) of Yersinia

2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

H. Zheng et al.

enterocolitica: nucleotide sequence and distribution of the yst genes. Microb Pathog 23: 189–200. Rasmussen HN, Rasmussen OF & Andersen JK (1994) Specific detection of pathogenic Yersinia enterocolitica by twostep PCR using hot-start and DMSO. Mol Cell Probes 8: 99–108. Sanchez-Vizcaino JM & Cambro-Alvarez M (1987) Execution of the ELISA technique. Enzyme Immunoassay Techniques in Animal and Plant Disease, 2nd edn. pp. 23–24. Office International des Epizooties, Paris. Singh I & Virdi JS (2004) Production of Yersinia stable toxin (YST) and distribution of yst genes in biotype 1A strains of Yersinia enterocolitica. J Med Microbiol 53: 1065–1068. Skurnik M & Wolf-Watz H (1989) Analysis of the yopA gene encoding the Yop1 virulence determinants of Yersinia spp. Mol Microbiol 3: 517–529. Sonnenwirth AC (1970) Bacteremia with and without meningitis due to Yersinia enterocolitica, Edwardsiella tarda, Comamonas terrigena, and Pseudomonas maltophilia. Ann NY Acad Sci 174: 488–502. Taylor BG, Zafarzai MZ, Humphreys DW & Manfredi F (1977) Nodular pulmonary infiltrates and septic arthritis associated with Yersinia enterocolitica bacteremia. Am Rev Respir Dis 116: 525–529. Thisted Lambertz S & Danielsson-Tham ML (2005) Identification and characterization of pathogenic Yersinia enterocolitica isolates by PCR and pulsed-field gel electrophoresis. Appl Envir Microbiol 71: 3674–3681. Thoerner P, Bin Kingombe CI & B¨ogli-Stuber K (2003) PCR detection of virulence genes in Yersinia enterocolitica and Yersinia pseudotuberculosis and investigation of virulence gene distribution. Appl Envir Microbiol 69: 1810–1816. Urbano-Marquez A, Estruch R & Agusti A (1983) Infectious endocarditis due to Yersinia enterocolitica. J Infect Dis 148: 940. Wauters G, Kandolo K & Janssens M (1987) Revised biogrouping scheme of Yersinia enterocolitica. Contrib Microbiol Immunol 9: 14–21. Weynants V, Jadot V & Denoel PA (1996) Detection of Yersinia enterocolitica serogroup O:3 by a PCR method. J Clin Microbiol 34: 1224–1227. Yang SH & Fang H (2003) Yersinia enterocolitica; in Pathogen and Bacteriology for Animal and People (in Chinese), pp. 558–592. Hebei Scientific and Technological Publishing House, Shijiazhuang. Zheng HX, Zhang MJ, Sun Y & Jiang B (2006) Detection of Yersinia enterocolitica in diarrheal stool by real-time PCR (in Chinese). Zhonghua Yi Xue Za Zhi 86: 2281–2284. [Zheng HX, Wang JD, Zhang MJ, Sun Y, Jiang B (2006) Yersinia enterocolitica identification in stool samples using real-time PCR (in English). Diagn Microbiol Infect Dis (available on PubMed)]. Zheng H, Wang J, Sun Y & Jiang B (2007) Clinical isolation and characterization of Yersinia enterocolitica in China using realtime PCR and culture method. Digestion 75: 199–204.

FEMS Immunol Med Microbiol 53 (2008) 368–374