HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Yoh, M. Articles by Honda, T. Search for Related Content PubMed PubMed Citation Articles by Yoh, M. Articles by Honda, T. Agricola Articles by Yoh, M. Articles by Honda, T.

Importance of Providencia species as a major cause of travellers’ diarrhoea

^1,2Center for Emerging Infectious Diseases¹ and Department of Bacterial Infections², Research Institute for Microbial Diseases, Osaka University, 3-1 Yamadaoka, Suita, Osaka 565-0871, Japan ³Kansai Airport Quarantine Station, 1 Senshu Kuko Naka, Tajiri-cho, Sennan-gun 549-0011, Japan

Correspondence Myonsun Yoh myon{at}biken.osaka-u.ac.jp

Received July 30, 2004
Accepted July 24, 2005

In this study the importance of Providencia species as a cause of travellers’ diarrhoea was examined using a selective medium developed by the authors. Providencia species could easily be distinguished from other enteric pathogens by the colour of the colonies obtained. Nine strains of Providencia alcalifaciens, nine of Providencia rettgeri and five of Providencia stuartii were isolated from 130 specimens, representing a surprisingly high incidence of infection compared with other pathogens isolated on SS agar and TCBS agar. Patients infected with P. rettgeri complained of abdominal pain, as for other Providencia species, but also of vomiting, which is rather characteristic of P. rettgeri infection. To analyse the pathogenicity of these isolates, their invasiveness was examined using Caco-2 cells. Most of the P. rettgeri strains invaded Caco-2 cells. Random amplified polymorphic DNA (RAPD) fingerprinting showed the same profile for two P. rettgeri isolates from individuals travelling in the same tour group. The results show that Providencia species, especially P. rettgeri, might cause diarrhoea, and that these species are important pathogens.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Providencia alcalifaciens has been described as a possible diarrhoea-causing pathogen in travellers and children in developing countries (Albert et al., 1998; Guth & Perrella, 1996; Haynes & Hawkey, 1989; Janda et al., 1998). In 1996, a large outbreak of food-borne infection, involving 290 patients, occurred in Japan. We investigated the aetiology of the outbreak and concluded that the cause was P. alcalifaciens (Murata et al., 2001). To show the role of Providencia species in bacterial gastroenteritis, Janda et al. (1998) evaluated their ability to invade eukaryotic cell lines. They reported that some P. alcalifaciens strains recovered from people with diarrhoea were invasive in HEp-2 and other eukaryotic cell lines but Providencia stuartii and Providencia rettgeri were not. As the enteropathogenicity of the other species of Providencia has not been established, Providencia species other than P. alcalifaciens are considered to be commensal flora in the gastrointestinal tract, and their detection as pathogens in patients with diarrhoea is not usually attempted. Furthermore, selective media to isolate Providencia species have not been developed.

To evaluate not only P. alcalifaciens but also other Providencia species for their ability to cause diarrhoea, we established an effective selective medium (polymyxin-mannitol-xylitol medium for Providencia, PMXMP). A medium for the detection and recognition of P. alcalifaciens in faeces [P. alcalifaciens medium (PAM)] was previously reported (Senior, 1997). We modified PAM to detect Providencia species with ease and simplicity.

In the present study, we found a high incidence of Providencia species, especially P. rettgeri, as a cause of travellers’ diarrhoea, and examined the invasiveness of these pathogens in Caco-2 cells. Albert et al. (1998) reported vomiting in diarrhoeal children infected with P. alcalifaciens, but to our knowledge there have been no reports so far on adult patients infected with other Providencia species. The results of this study provide evidence that P. rettgeri strains can cause diarrhoea. This is the first report on the importance of P. rettgeri as a cause of gastroenteritis.

	METHODS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Bacterial strains.

P. alcalifaciens, P. rettgeri, and P. stuartii strains were isolated at the Kansai airport quarantine station in 2002 (Table 1). P. alcalifaciens RIMD 1656012, P. stuartii RIMD 1660001, P. rettgeri RIMD 1658001, Providencia rustigianii RIMD 1659002, Escherichia coli RIMD 0509770 (enterotoxigenic E. coli, ETEC), E. coli RIMD 0509952 (enterohemorrhagic E. coli, EHEC), Klebsiella pneumoniae RIMD 1102001, Salmonella enteritidis RIMD 1933006, Shigella boydii RIMD 1998008, Proteus vulgaris RIMD 1643003, Proteus mirabilis RIMD 1641002, Morganella morganii RIMD 1642003, Pseudomonas aeruginosa RIMD 1603014 and Serratia liquefaciens RIMD 3126001 were used for evaluation of the medium (Table 2). P. alcalifaciens RIMD 1656001, Salmonella typhimurium RIMD 1985089 and E. coli RIMD 05091207 (enteroinvasive E. coli, EIEC) were used as controls in the test of invasiveness.

Preparation of PMXMP medium.

We modified P. alcalifaciens medium (PAM) (Senior, 1997), containing 10 g Lab Lemco powder (Oxoid), 5 g sodium deoxycholate, 0.8 g anhydrous disodium hydrogenphosphate, 80 mg phenol red dye and 12 g agar in 1 l distilled water, by adding sugars (maltose and xylose) separately sterilized at 100 °C for 10 min, and polymyxin B sulfate (Pfizer) (250 000 units in 5 ml distilled water sterilized by filtration; pore size, 0.2 µm). After the base medium with these sugars was cooled to 50 °C, polymyxin B was added.

Evaluation of PMXMP.

Seven strains of P. alcalifaciens, three strains of P. stuartii, two strains of P. rettgeri and one strain of other species, including the strains listed in Table 2, were inoculated onto PMXMP to observe their growth. After 48 h incubation, colony formation and the colour of colonies were observed. Strains listed in Table 2 were inoculated into 1 ml Luria–Bertani (LB) broth and incubated overnight at 37 °C with rotation. The dilution of cultures with PBS was done serially and 0.1 ml of diluent was plated on PMXMP and tryptic soy agar (TSA, Difco) as non-selective medium. The numbers of colony-forming units were counted on TSA after 24 h and on PMXMP after 48 h incubation.

Identification of Providencia species.

The suspicious colonies coloured red to pink on PMXMP were mostly Providencia species or M. morganii, and occasionally S. liquefaciens and Pseudomonas species. The ornithine decarboxylase test distinguished Providencia species from M. morganii and Serratia liquefaciens, and the oxidase test distinguished Providencia species from Pseudomonas species. To determine the species among the genus of Providencia, the API 20E biochemical strip for the family Enterobacteriaceae (API System, BioMérieux) was used. The adonitol utility test helped to distinguish P. alcalifaciens and P. rustigianii. API 20E was also used to confirm M. morganii, Serratia liquefaciens and Pseudomonas species.

Isolation of diarrhoeal pathogens from faeces at Kansai airport quarantine station.

In the Kansai airport quarantine station, faeces from patients with diarrhoea were usually inoculated onto Salmonella–Shigella agar (SS agar) (Nissui) and into alkaline peptone water. Cultures enriched for 6 h in alkaline peptone water were inoculated onto thiosulfate citrate bile saccharose agar (TCBS) (Nissui) to isolate Vibrio cholerae and other Vibrio species. From SS agar and TCBS, Shigella species, Salmonella species, Plesiomonas species, enteropathogenic E. coli, Aeromonas species, Vibrio parahaemolyticus, V. cholerae, Vibrio fluvialis and Vibrio mimicus could be isolated. For this study, faeces were also inoculated on PMXMP by the direct method. Enrichment of the culture (in alkaline peptone water) before the inoculation on PMXMP did not affect the detection of Providencia species. After 40 h at 37 °C, red to pink colonies on PMXMP were collected. Suspicious colonies without ornithine decarboxylase activity were identified using the API 20E system (bioMérieux) and adonitol fermenting test. Surveillance was done for a period of 1 month (from 15 June 2002 to 14 July 2002) at the Kansai airport quarantine station. The number of travellers with diarrhoea during this period was 130 (all except one were adults). The majority of travellers who complained of diarrhoea were returning from South-East Asia.

Healthy controls.

As it is hard to collect the stools of healthy returning travellers, we examined the stools of food handlers, who are obligated to regularly give stool samples to confirm that they are not infected by enteropathogens. During one month, 177 specimens were analysed by the PMXMP method.

Invasion assay.

The invasion assay was conducted essentially as described elsewhere (Akeda et al., 1997). In brief, we grew Caco-2, a human colon carcinoma cell line, in Dulbecco's modified Eagle medium (DMEM) supplemented with 10 % fetal calf serum (FCS) in 5 % CO₂ at 37 °C. Assays were performed in 48-well tissue culture plates (Corning). For the invasion assay, cultures were grown to stationary phase at 37 °C in LB broth with rotation. The bacterial cultures were diluted in PBS to give a m.o.i. of 1. Cell monolayers were challenged with bacterial suspensions for 3 h at 37 °C in 5 % CO₂ and then washed three times in PBS to remove non-invasive bacteria. DMEM supplemented with 10 % FCS containing gentamicin (100 µg ml^–1) was then added to each well to kill any extracellular bacteria not removed by the PBS washes. After an additional 2 h incubation, the monolayers were lysed with 0.1 % Triton X-100 in PBS for 20 min at 37 °C with reciprocal shaking (150 r.p.m.). To determine the numbers of intracellular bacteria, we plated serial dilutions of the lysed monolayers in PBS on TSA (Merck) and incubated the samples overnight at 37 °C. Assays were repeated three times, each time in duplicate. Results are shown as the percentage of bacteria internalized, which was calculated from the mean number of colony-forming units of the bacterial cells internalized in the Caco-2 cells and the number of bacteria in the original inoculum.

Plasmid profiles.

Using the method of Birnboim & Doly (1979), we extracted plasmid DNA from overnight cultures of P. alcalifaciens strains and P. rettgeri strains, and examined them by electrophoresis in 0.5 % agarose gels.

Random amplified polymorphic DNA (RAPD) fingerprinting.

Genomic DNA was prepared from overnight cultures (1 ml) in LB broth at 37 °C with a DNA purification kit (QIA amp DNA mini kit; QIAGEN KK) as recommended by the manufacturer. RAPD reactions with the genomic DNA preparations were performed using beads (Ready-To-Go RAPD Analysis Beads, Amersham Biosciences) according to the manufacturer's directions. For the RAPD analysis, primers 3 (5'-d[GTAGACCCGT]-3') and 5 (5'-d[AACGCGCAAC]-3') were used. Amplification products were separated by electrophoresis in 2 % agarose (Invitrogen) and visualized by ethidium bromide staining.

	RESULTS AND DISCUSSION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Among Enterobacteriaceae, including Providencia species, M. morganii is the only species that does not ferment maltose-xylose like Providencia species. As shown in Table 2, E. coli and Salmonella species could grow on PMXMP but their colonies were yellow, distinct from the red to pink colonies formed by Providencia species. The rate of growth inhibition (colony formation) on PMXMP for E. coli or Salmonella species was extremely high, about 10⁵–10⁷ times that for Providencia species. Since M. morganii formed colonies with a red to pink colour like Providencia species, we distinguished them by examining ornithine decarboxylase activity using Mller ornithine medium (Eiken kagaku): the former is positive and the latter negative for the activity.

As shown in Table 1, 23 strains of Providencia species were isolated on PMXMP from 20 patients (15.4 %) among 130. In addition, other species were isolated by routine methods (SS agar and TCBS), including Plesiomonas shigelloides from 13 patients (10 %) and Vibrio parahaemolyticus from nine patients (6.9 %) (Table 3). P. rettgeri was isolated in nine cases, P. alcalifaciens in nine cases and P. stuartii in five cases. Four of 20 subjects with Providencia species were also infected with some other species of enteric bacteria (patients 3, 6, 8 and 17 in Tables 1 and 3). Two different species of Providencia were isolated simultaneously in three cases (patients 5, 12 and 13 in Table 1). In the remaining Providencia cases, one Providencia species only was isolated.

The symptoms of patients from whom Providencia species were isolated are summarized in Table 1. Diarrhoea was seen in all subjects except one (patient 18), who complained of vomiting on the day the faeces were obtained. Seven of the 20 patients complained of abdominal pain and four patients had a fever. Vomiting was seen in five patients from whom P. rettgeri was isolated.

Among 177 specimens from healthy controls, six specimens (seven strains) were Providencia-positive (3.4 %). Three strains of P. rettgeri, one strain of P. stuartii, one strain of P. alcalifaciens and two strains of P. rustigianii were isolated. One specimen contained both P. rustigianii and P. alcalifaciens.

Although nine P. rettgeri strains were isolated from the diarrhoeal patients in our case, P. rettgeri has not to our knowledge been reported previously as an enteropathogenic bacterium. To analyse the pathogenicity of these strains, we performed an invasion test using Caco-2 cells (Fig. 1). Four of the nine P. rettgeri strains and two of the nine P. alcalifaciens strains were highly invasive. Four P. rettgeri strains, three P. alcalifaciens strains and one P. stuartii strain showed weak invasive activity. As a positive reference, we used a P. alcalifaciens strain isolated previously in Japan (Murata et al., 2001), EIEC and Salmonella typhimurium; all were positive as expected (Fig. 1). None of seven strains from healthy controls showed invasive activity.

View larger version (33K): [in this window] [in a new window]   Fig. 1. Invasion of Caco-2 cells by Providencia species strains. PA, P. alcalifaciens; PR, P. rettgeri; PS, P. stuartii; EI, enteroinvasive E. coli; ST, Salmonella typhimurium. Assays were repeated five times. Results are shown as mean ± SD percentages of bacteria internalized in the Caco-2 cells, which was calculated as the mean number of colony-forming units of bacterial cells internalized/number of bacteria in original inoculum.

In some Enterobacteriaceae, a large plasmid, the virulence plasmid, has been implicated in the ability of the organism to invade eukaryotic cells (Hale, 1991). We analysed the plasmid profile of the P. alcalifaciens and P. rettgeri strains, and found that all strains had a large (>100 kbp) plasmid (data not shown). We could not find, however, any relationship between invasiveness in Caco-2 cells and the possession of a common plasmid in the P. rettgeri strains.

Nine strains of P. rettgeri isolated from travellers with diarrhoea returning from various countries of South-East Asia were analysed further for genetic relatedness. Among the travellers from whom P. rettgeri was isolated, patients 5 (RIMD 1658019) and 7 (RIMD 1658023) (Table 1) belonged to the same tour group. These two strains showed the same genotype on RAPD fingerprinting (Fig. 2). The results strongly suggest that patients 5 and 7 were infected with P. rettgeri abroad and developed diarrhoea, and that the strains were not commensals in the gastrointestinal tract.

View larger version (72K): [in this window] [in a new window]   Fig. 2. RAPD profiles obtained from genomic DNA of P. rettgeri strains. Lane 1, RIMD 1658024 from patient 3; lane 2, RIMD 1658016 from patient 4; lane 3, RIMD 1658019 from patient 5; lane 4, RIMD 1658022 from patient 6; lane 5, RIMD 1658023 from patient 7; lane 6, RIMD 1658011 from patient 8; lane 7, RIMD 1658014 from patient 10; lane 8, RIMD 1658015 from patient 11; lane 9, RIMD 1658006 from patient 18; lane M, DNA molecular marker, X174/HaeIII (Nippon Gene).

Our results showed that PMXMP medium is an excellent selective medium for the isolation of Providencia species. One can not distinguish Providencia species from other non-lactose-fermenters such as Salmonella species or Shigella species using DHL, MacConkey agar or SS agar as a selective medium for the isolation of bacteria from faeces. Testing at the Kansai airport quarantine station using PMXMP as a selective medium for Providencia species showed a high incidence of Providencia species in diarrhoeal faeces. Since faeces from diarrhoeal patients are usually inoculated into alkaline peptone water to enrich V. cholerae at the quarantine station, we also used alkaline peptone water in this study. We obtained the same results with the direct method and the enrichment method using alkaline peptone. Of 20 patients from whom Providencia species were isolated, other enteropathogens were also isolated in four, but only Providencia species were isolated in the remaining 16. Although P. alcalifaciens seems to be an enteropathogen (Albert et al., 1998; Guth & Perrella, 1996; Haynes & Hawkey, 1989; Murata et al., 2001), P. rettgeri and P. stuartii have not been reported to cause human diarrhoea until now. Furthermore, most of the P. rettgeri strains showed invasive activity in Caco-2 cells but P. stuartii did not. These results suggest that P. rettgeri is a diarrhoea-causing pathogen.

Providencia species are likely to be pathogens for diarrhoea among Japanese travellers abroad. This is in agreement with the report of Haynes & Hawkey (1989) in Britain, who analysed the relationship between P. alcalifaciens and travellers’ diarrhoea. We isolated not only P. alcalifaciens but also P. stuartii and P. rettgeri from travellers with diarrhoea. Some of the patients infected with Providencia species complained of abdominal pain, fever and vomiting. As detailed information on clinical symptoms is not available, the relation between invasive activity and clinical symptoms is not clear. Vomiting was a rather characteristic complaint among patients with diarrhoea caused by P. rettgeri. Although the cause of the vomiting is not clear yet, another virulence factor like emetic toxin might be produced by P. rettgeri strains.

The virulence plasmid possessed by Shigella species and EIEC has been implicated in the ability of the organism to invade eukaryotic cells (Hale, 1991). Guth & Perrella (1996) reported that the invasive ability of P. alcalifaciens strains in HeLa cells was not related to the possession of a plasmid, but our previous results suggested that a large plasmid might in some way be related to the invasiveness of P. alcalifaciens strains in Caco-2 cells (Murata et al., 2001). Here we could not find any relationship between invasiveness of P. rettgeri in Caco-2 cells and the possession of a (large) virulent plasmid. P. rettgeri might carry invasion-associated genes chromosomally like other enteric pathogens.

The epidemiological typing of P. rettgeri isolates by RAPD fingerprinting was also performed. Two strains from patients who travelled together showed the same genotype on RAPD fingerprinting, while all the other strains differed. As the two patients were not related and were not even living in the same household, it is not possible that they were exposed to the same pathogen in their daily life. The finding suggests that these two individuals became infected with P. rettgeri while abroad. As we cannot rule out that parasitic or viral agents were associated with the travellers’ diarrhoea, further analysis is required.

This is the first report that suggests that P. rettgeri is an enteropathogen in adults. We believe that P. rettgeri is a cause of diarrhoea among travellers returning to Japan. The mechanism behind the diarrhoea caused by P. rettgeri may be invasiveness but further study is necessary.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This work was supported by a grant-in-aid for Scientific Research on Priority Areas (C) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. We appreciate the assistance of members of the Bacterial Clinical Laboratory, the Research Foundation for Microbial Diseases of Osaka University, with the analysis of healthy controls.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES