Journal of Infection (1994) 29, 3o5-3IO

SHORT COMMUNICATION

Bio typ ing o f C a m p y l o b a c t e r j e j u n i a n d C a m p y l o b a c t e r coli in fec t ions in Spain

Aria Jim6nez,* Jorge B. VeDizquez,~" Joaquln Rodrlguez,* Benilda Chom6n~ and Tom,is G. Villa~

* Laboratory of Microbiology, Residencia Sanitaria Ntra. Sra. del Cristal, Orense, "~ Laboratory of Food Technology, Faculty of Food Science and Technology, University of Santiago de Cornpostela at Lugo, 27071 Lugo, Laboratory of Clinical Analysis, Hospital Mddico-Quirfzrgico de Conxo, Santiago de Cornpostela and ~ Department of Microbiology, Faculty of Pharmacy, University of Santiago de Cornpostela, I57O6 Santiago de

Cornpostela, Spain

Accepted for publication IO June I994

Summary

Biochemical profiles were studied in IO2 Campylobacter isolates from patients affected by enteric disease in Orense (Spain) over a I-year period. The isolates were identified as hippurate-positive Campylobacter jejuni (n = 9o), Campylobacter coli (n = m) and hippurate-negative C. jejuni (n = 2). Seventy-seven of the hippurate-positive and both hippurate-negative C. jejuni isolates were biotyped as C. jejuni subsp, jejuni biotype I, nine as C. jejuni subsp, jejuni biotype 2 and four as C. jejuni subsp, doylei. Hippurate- hydrolysis was correlated to a simple scheme based on L-arginine arylamidase production, propionate assimilation and malate assimilation which yielded sensitivity and specificity values of o'9o and I.OO, respectively. Effective grouping of nalidixic acid-resistant C. jejuni and C. coli isolates (38"5 % of the total) was also achieved.

Introduction

Microorganisms belonging to the genus Carnpylobacter have been established as major enteric pathogens in humans throughout the world. 1-~ The prevalence of Carnpylobacter infection is quite high in Orense (northwestern Spain) and is the main cause of human enteric disease in this region together with Salmonella.

Biotyping of clinical Carnpylobacter strains is sometimes complicated by the isolation of hippurate-negative C. jejuni or both nalidixic acid-resistant (NAR) C. jejuni and C. coil < 7 Indeed, the prevalence of NAR Carnpylobacter is quite high in patients affected by enteric disease in our region, 5 causing problems in the grouping of clinical isolates. The high prevalence of NAR hippurate- positive Carnpylobacter strains in our region and other regions of Europe might suggest the need for a taxonomic reorganisation in this genus. In this report we describe the effective biotyping of Campylobacter involved in human infections in Orense, northwestern Spain, during a I-year-period.

Address correspondence to : Prof. Tomfis G. Villa, Department of Microbiology, Faculty of Pharmacy, University of Santiago de Compostela, 15706 Santiago de Compostela, Spain.

ox63-4453/94/o6o3o5 +06 $08.00/o

14

© I994 The British Society for the Study of Infection

5IN 29

306 A. JIMI~NEZ ET AL.

Materials and methods

Isolation o f microorganisms and media

Stool samples were collected from patients affected by enteric disease in the Residencia Sanitaria Ntra. Sra. del Cristal (Orense, northwestern Spain) between June I992 and May I993 and processed for Salmonella and Shigella species, enterotoxigenic Escherichia coli (ETEC), Vibrio species, rotavirus and parasites by standard methods.* The isolation of Campylobacter was carried out on Campylosel medium (BioM6rieux, Marcy l'Etoile, France) at 49. °C under microaerophiUic conditions (gas-generating kit, Oxoid Ltd., London, England). Identification of strains as belonging to the genus Campylobacter was carried out using the following tests: Gram's stain, production of oxidase (Mast Diagnostics Ltd., Merseyside, England), production of catalase and agglutination test (Meridian Diagnostics, Cincinnati, OH, U.S.A.).

Biotyping o f Campylobacter strains

Hippurate-hydrolysis was determined as described elsewhere2 The pheno- typical behaviour of the Campylobacter strains was studied using a standardised API CAMPY system (BioM6rieux) according to the manufacturer 's instruc- tions. The following tests were considered: assimilation of carbon sources (glucose, succinate, acetate, propionate, malate and citrate); production of enzymes (urease, nitrate reductase, esterase, hippurate hydrolase, gamma glutamyl transferase, triphenyl tetrazolium chloride reductase, pyrrolidonyl arylamidase, L-arginine arylamidase, L-aspartate arylamidase and alkaline phosphatase); susceptibility to antimicrobial agents (nalidixic acid, cefazoline and erythromycin); and production of hydrogen sulphur. After 24 hours of incubation at 37 °C, the reactions were read and biotyping carried out according to the API CAMPY Analytical Profile Index. A laboratory strain of C. jejuni was used as reference.

Results

Biotyping of Campylobacter strains

Stool samples were collected from 1561 patients affected by enteric disease in the Residencia Sanitaria Ntra. Sra. del Cristal (Orense, northwestern Spain) between June 1999 . and May 1993. From this, 361 samples (23"2 %) carried pathogenic flora which in IO2 (6"5%) cases proved to be Campylobacter. Ninety of the lO2 Campylobacter isolates were characterised as hippurate- positive C. jejuni, ten as C. coli and two as hippurate-negative C. jejuni. Seventy-seven of the 9o hippurate-positive C. jejuni strains were biotyped as C. jejuni subsp, jejuni biotype I, nine as C. jejuni subsp, jejuni biotype 2 and four as C. jejuni subsp, doylei. Both hippurate-negative C. jejuni strains were characterised as C. jejuni subsp, jejuni biotype I.

Table I summarises the results obtained in the biochemical characterisation of the ioz Campylobacter isolates. None of the io9 strains produced urease, L- aspartate arylamidase or sulphur dioxide nor assimilated glucose, but all of them were catalase producers. The nine strains of C. jejuni subsp, jejuni biotype 2 produced gamma glutamyl transferase (GGT), alkaline phosphatase

Table I

Biotyping of Campylobacter

Biotyping of the IO2 isolates of C. jejuni and C. coli

307

C. jejuni C. jejuni subsp, jejuni subsp, doylei C. coli

Biotype i Biotype 2 Tes t (n = 79) a (n = 9) (n = 4) (n = IO)

Hippura te hydrolysis 77 (97"4) 9 (Ioo) 4 (I00) 0

Product ion tests g r e a s e o o o o Catalase 79 (IOO) 9 (IOO) 4 (IOO) IO (IOO) Esterase 68 (86'I) 7 (77"7) 4 (IOO) Io (Ioo) G G T o 9 (I00) 0 0 Pyrrol idonyl arylamidase 27 (34"2) 2 (22'2) o o L-arg in ine arylamidase I (I '3) o o 5 (50) L-aspar ta te arylamidase o o o o Alkaline phosphatase 73 (92'4) 9 ( too) 4 (IOO) 9 (90) H2S product ion o o o o

Reduc t ion tests Nitrates 73 (92"4) 7 (77'7) o IO (IOO) T T C 53 (67"I) 4 (44"4) I (25) 9 (9 o)

Assimilat ion tests Glucose o o o o Succinate 73 (92"4) 9 ( ioo) 3 (75) IO ( ioo) Acetate 5 (6"3) I ( I I ' I ) O 0 Propionate o o o 5 (50) Malate 69 (87"3) 9 (Ioo) 3 (75) 3 (3o) Citrate 25 (31"6) 3 (33"3) 2 (5o) I (IO)

Susceptibi l i ty tests b Cefazoline 72 (9 I ' I ) 9 (IOO) I (25) IO (IO0) Nalidixic acid 3I (39"2) 3 (33'3) I (25) 3 (3 O) Ery th romyc in 0 0 0 2 (20)

a Results are expressed as the number and percentage (in brackets) of posit ive strains. b Resistant strains are considered as positive.

and assimilated both succinate and L-malate, while none produced L-arginine arylamidase or assimilated propionate (Table I). The most significant biochemical difference between the biotypes I and z of C. jejuni subsp, jejuni was the non-production/production of G G T , respectively. Strains belonging to the C. jejuni subsp, doylei mainly differed from the C. jejuni isolates in the absence of nitrate reduction and the non-production of G G T or any of the three arylamidases tested (Table I). With respect to the isolates of C. coli, all were nitrate reducers, esterase producers and succinate assimilators. By contrast, none of the ten isolates of C. coli produced G G T , pirrolyidonyl arylamidase of L-aspartate arylamidase nor assimilated acetate (Table I).

The next step in this work was the selection of a small number of biochemical tests that could be correlated to hippurate-hydrolysis in our clinical isolates of Campylobacter. With this aim, all the IO2 isolates and two C. jejuni reference strains were carefully studied with respect to their

14-2

308 A. J IMI~NEZ E T A L .

Table II (a) Correlation between three biochemical tests and the hippurate hydrolysis test; (b) Rapid biotyping scheme for C. jejuni

(a) L-arginine arylamidase-, propionate , malate+/hippurate+: Sensitivity o'9o Specificity I 'oo Positive predictive value I.OO Negative predictive value o'53

(b) Isolates of C. jejuni: GGT-, Nitrate + : C. jejuni subsp, jejuni biotype I. GGT +, Nitrate + : C. jejuni subsp, jejuni biotype 2. GGT +, Nitrate-: C. jejuni subsp, doylei. GGT-, Nitrate-: C. jejuni subsp, doylei.

biochemical profiles. The following tests were selected: L-arginine aryl- amidase and both propionate and malate assimilation. As can be observed in Table II, 90 % of the hippurate-positive strains of C. jejuni were L-arginine arylamidase negative, propionate negative and malate positive. None of the ten hippurate-negative strains of C. coli showed this profile, so no false positive results were obtained. The NAR group showed similar behaviour: 90 % of the NAR hippurate-positive strains were grouped as C. jejuni following the above- described identification scheme. It should be noted that both hippurate- negative strains of C. jejuni considered in this work followed the cited biochemical pattern. Thus, although Table II shows the correlation of this proposed scheme with hippurate-hydrolysis, this scheme could be helpful in the elucidation of hippurate-negative strains likely to be C. jejuni. This rapid scheme could also be of use in distinguishing between nalidixic acid-resistant mutants of C. jejuni, an increasing clinical problem, and natural NAR species such as C. fetus and C. laridis. Table II also shows a simple grouping of C. jejuni in subspecies and biotypes based on two biochemical tests: nitrate reduction and the production of GGT.

D i s c u s s i o n

We have described a simple biochemical method that helped us in the identification of Campylobacter clinical strains. This method was based in the following tests included in the API CAMPY identification system : production of L-arginine arylamidase, assimilation of propionate and assimilation of malate. L-arginine arylamidase had been previously included in a grouping scheme for Campylobacter. TM Ninety percent of the 90 hippurate-positive C. jejuni isolates were L-arginine arylamidase negative, propionate negative and malate positive, while none of the ten C. coli isolates showed this biochemical pattern. This implied that both the specificity and positive predictive values were I'O, due to the absence of false positives. Other authors have described hippurate-negative isolates of Campylobacter genetically related to C. jejuni: hippurate hydrolysis in some of these strains cannot be observed even with the use of several chromatographic techniques. 7 The two hippurate-negative isolates of C. jejuni proved to be L-arginine arylamidase negative, propionate

Biotyping of Campylobac te r 3o9

negative and malate positive, suggesting that the a fore-ment ioned rapid scheme of identification could be helpful to dist inguish be tween C. coli and hippurate-negat ive C. jejuni. On the other hand, resistance to nalidixic acid was found in 38"2 % of our isolates of Campylobacter 87"2 % of them being C. jejuni and 12"8 % C. coli. In recent years, a high prevalence (in some cases higher than 3o %) of N A R mutants of Campylobacter has been repor ted in clinical laboratories. 5'11'12 Al though the identification system (API C A M P Y Analytical Profile Index) considered the resistance to nalidixic acid as a discordance in the identification of bo th C. jejuni and C. coli, the basic biochemical profile cor responded with high probabi l i ty to these two species rather than to other species of Campylobacter. Other authors have descr ibed problems in the identification of N A R mutants of Campylobacter isolated f rom patients affected by enteritis, tradit ionally caused by C. jejuni and C. coli. 6' 11 T h e rapid identification scheme descr ibed in this paper could also be useful in the differentiation be tween nalidixic acid-resistant (NAR) mutants of C. jejuni (these showing increasing prevalence in the clinical laboratory) and wild strains of C. fetus and C. laridis. Additionally, rapid grouping of our isolates of C. jejuni into subspecies and bio types was easily achieved by considering nitrate reduct ion and the produc t ion of G G T , as descr ibed in Tab le II . T h e cause of such a high prevalence of N A R mutants could be the extensive use of quinolones in veterinary medicine in the large meat and milk industries of the area surrounding Orense, as suggested for other areas of Europe by other authors. 15 T h e f requency of muta t ion to a N A R pheno type was much higher in our Campylobacter isolates than in other genera: this would indicate that the source of N A R Campylobacter mutants might be the poul t ry (exposed to high levels of quinolones) and that these pathogens would infect humans through the food chain.

(The authors wish to thank all the people at the Laboratory of Microbiology of the Residencia del Cristal (Orense) for their excellent technical assistance and Nick Skinner for editing the English manuscript.)

R e f e r e n c e s

I. Butzler JP, Skirrow MB. Campylobacter enteritis. Clin Gastroenterol 1978; 8: 737-765. 2. Blaser MJ, Taylor DN, Feldman RA. Epidemiology of Campylobacter jejuni infections.

Epidemiol Rev 1983; g : I 5 7 - I 7 6 . 3- Karmali MA, Penner JL, Fleming PC, Williams A, Hennessy JN. The serotype and

biotype distribution of clinical isolates of Campylobacterjejuni and Campylobacter coli over a three-year period. J Infect Dis I983 ; 147: 243-245.

4. Penner JL. The genus Campylobacter: a decade of progress. Clin Microbiol Rev 1988; x: I57-I72.

5. Jim~nez A, Velfizquez JB, Rodriguez J, Tinajas A, Villa TG. Prevalence offluoroquinolone resistance in clinical strains of Campylobacter jejuni isolated in Spain. J Antimicrob Chemother 1994; 33: I88--I90-

6. Altwegg M, Burnens A, Zollinger-Iten J, Penner JL. Problems in identification of Campylobacter jejuni associated with acquisition of resistance to nalidixic acid. J Clin Microbiol 1987; 25: I807-I808.

7. Torten PA, Patton CM, Tenover FC et al. Prevalence and characterization of hippurate- negative Campylobacter jejuni in King County, Washington. J Clin Microbiol 1987; 25: I747-I752.

3zo A. JIMENEZ E T A L .

8. Murray RGE, Brenner D J, Bryant MP, Holt JG, Krieg NR, Moulder JW, Pfennig N, Sneath PHA, Staley JT, Eds. Bergey's Manual of Systematic Bacteriology. Baltimore: The William & Wilkins Co., I984-I989.

9. Harvey SM. Hippurate hydrolysis by Campylobacter fetus. J Clin Microbiol I98O; IX: 435-437.

IO. Elharrif Z, M6graud F. Characterization of thermophilic Campylobacter. II. Enzymatic profiles. Curt Microbiol I986; I3: 317-32z.

zI. Endtz HPh, Ruijs GJ, van Klingeren B, Jansen WH, van der Reyden T, Mouton RP. Quinolone resistance in Campylobacter isolated from man and poultry following the introduction of fluoroquinolones in veterinary medicine. J Antimicrob Chemother I99I ; z7: I99--208.

I2. Segreti J, Gootz TD, Goodman LJ, Parkhurst GW, Quinn JP, Martin BA, Trenholme GM. High-level quinolone resistance in clinical isolates of Campylobacter jejuni. J Infect Dis I992; I65: 667-670.