Veterinary Microbiology 93 (2003) 261273

Detection and characterization of feline Bartonella henselae in the Czech Republic

kb, R.S. Weyantc, J. Janec ekb, A. Nemeca, O. Meltera,*, K. Herc e, P. Brannyb J. Mecerad, L. Gonzorova
a

roba rova 48, 10042 Prague, Czech Republic National Institute of Public Health, S b den ska 1083, 14220 Prague, Czech Republic Institute of Microbiology, V c Centers for Disease Control and Prevention, Atlanta, GA 30333, USA d ka 902, 19800 Prague, Czech Republic Veterinary Care Center, Gen. Janous e dku 8, 12000 Prague, Czech Republic Veterinary Clinic, Na Hra

Received 27 May 2002; received in revised form 10 December 2002; accepted 23 January 2003

Abstract The aims of the study were to characterize isolates of Bartonella henselae and to determine the prevalence of bacteremic domestic cats in urban and suburban parts of Prague, Czech Republic. Five (18%) gram-negative fastidious bacterial single-cat isolates were recovered from 27 hemocultures incubated without previous freezing. Four of these isolates originated from ea infested stray cats (n 6) and one from a shelter cat without any ectoparasites (n 21). None of the 34 previously frozen specimens from ea free pet cats yielded any bacteria. All ve isolates were catalase and oxidase negative. Their enzymatic activity, RFLP prole of citrate synthetase gene (gltA) and DNA DNA hybridization results were typical of B. henselae. According to their PvuII and BglI ribotypes the isolates could be allocated to two homogeneous groups. Ribotype HindIII and RFLP of 16S23S rRNA spacer region analysis gave unique proles different from those of Bartonella quintana, Bartonella elizabethae and Bartonella clarridgeiae. The 16S rRNA type-specic amplication revealed an identical prole typical of B. henselae genotype II for all the cat isolates studied. Pulsedeld gel electrophoresis (PFGE) assigned a different prole to each of the isolates studied. Determination of the enzymatic activity, RFLP of gltA gene, RFLP of 16S23S rRNA spacer region, and HindIII ribotype could be efcient tools for identication of B. henselae isolates. Ribotyping (PvuII, BglI), 16S rRNA typing and PFGE may be useful methods to prospect ecology and epidemiology of the agent. # 2003 Elsevier Science B.V. All rights reserved.
Keywords: Bartonella henselae; Cat; Molecular typing; Czech Republic

* Corresponding author. Tel.: 420-2-67082266; fax: 420-2-72730428. E-mail address: melter@szu.cz (O. Melter).

0378-1135/03/$ see front matter # 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0378-1135(03)00032-4

262

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

1. Introduction At present, the genus Bartonella includes 16 species, six of which (Bartonella bacilliformis, Bartonella henselae, Bartonella quintana, Bartonella clarridgeiae, Bartonella elizabethae, Bartonella vinsonii subspp. berkhofi and arupensis) may cause human infections (Birtles and Raoult, 1998; Droz et al., 1999; Welch et al., 1999). The most relevant medical impact is ascribed to B. henselae, responsible for cat-scratch disease (CSD), vasoproliferative illnessbacillary angiomatosis, fever and other clinical symptoms depending on the organs affected (Fournier and Raoult, 1998). Epidemiological studies support the idea that cats, and particularly young kittens, are reservoirs for B. henselae infection (Zangwill et al., 1993; Child et al., 1994; Groves and Harrington, 1994; Chomel et al., 1995). The suspected vector is the cat ea (Ctenocephalides felis) (Koehler et al., 1994; Chomel et al., 1996). The prevalence of B. henselae bacteremic cats varies widely from 141% with the regions and categories of the cats studied (Koehler et al., 1994; Chomel et al., 1995; Branley et al., 1996; Bergmans et al., 1997; Heller et al., 1997; Sander et al., 1997; Arvand et al., 2001). B. quintana is the causative agent of trench fever, nowadays also described as urban sickness of homeless and alcoholic individuals and transmitted by their ectoparasites (Pediculus humanus). Human infections caused by other Bartonella spp. are very rare (Birtles and Raoult, 1998). One of the aims of this study was to determine the prevalence of B. henselae isolates in the blood of domestic cats since the agent had never been cultured in the Czech Republic before the present study although co-operating infectionists had announced sporadically suspected CSD cases. A polyphasic analysis was used to identify the isolates, to analyze their properties and to compare them with those of selected strains of Bartonella species. The cat venous blood specimens were cultured and the suspected isolates were identied and subtyped by the methods as described previously: enzymatic detection (Welch et al., 1993), RFLP of gltA gene (Regnery et al., 1991), RFLP of 16S23S rRNA spacer intergenic region (Matar et al., 1993), 16S rRNAS type specic amplication (Bergmans et al., 1996), PFGE (Roux and Raoult, 1995; Sander et al., 1998) and DNADNA hybridization (Brenner et al., 1993). Moreover, digestion of DNA of the isolates by HindIII, PvuII and BglI, blotting of the DNA fragments onto nylon membrane and hybridization by digoxigenin probe specic for 16S23S rRNA (Gerner-Smidt, 1992) were performed.

2. Materials and methods 2.1. Bacterial strains Houston 1 (ATCC 49882T) was used as the B. henselae type strain. Other Bartonella strains were obtained from the American type culture collection: B. elizabethae ATCC 49927T, B. clarridgeiae ATCC 51734T, and from the Collection de lInstitut Pasteur: B. quintana CIP 103739. An additional B. henselae strain G9866, originally cultured from the groin lymph node of a human in Michigan, USA, was used.

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

263

2.2. Cat specimens and isolates Sixty-one specimens (C1C61) were collected from cats living in urban and suburban parts of Prague in the Czech Republic located at least 5 km from each other. The stray cats studied were trapped in one urban and two suburban localities. Pound cats lived in two urban shelters and pet cats lived in one urban and two suburban parts of the city. Age of the cats varied from 3 months to 13 years and mean age was 35 months. Specimens C1, C2, C25C27 were taken from ea infested stray cats whose age ranged from 3 months to 12 years and mean age was 27 months; specimens C4C24 originated from shelter cats of the same age range and mean age of 32 months and specimens C28C61 were obtained from pet cats occasionally staying outdoors showing an age range from 1 to 13 years and mean age of 36 months (Table 1). The shelter and pet cats were not infested by ectoparasites while investigated. In 1995, 2 ml of venous blood (specimens C1C27) were collected into tubes containing home-made transport medium [(2 ml Brain Heart Infusion Broth (Oxoid Limited, UK), K2EDTA (2 mg/ml) and 0.2% of saponin (SigmaAldrich, Germany)] to be processed immediately after transport to the laboratory. The tube was gently agitated for 5 min. Lysed blood (0.2 ml) from each specimen was poured on one Columbia agar (Oxoid) plate with 5% of sheep blood and one chocolate agar plate prepared with MuellerHinton agar (Oxoid) enriched with 5% of sheep blood. In the year 2000, specimens C28C61 were collected into evacuated tubes (Vacutainer, Becton Dickinson, USA) and stored at 20 8C for up to 45 days until all of them were gathered to be referred to the laboratory and inoculated onto the media described above.

Table 1 Characteristics of cats Cat C1 C2 C3 C4 C5C24 C25 C26b C27 C28C36 C37C61 Category CFU Stray Stray Stray Shelter Shelter Stray Stray Stray Pets Pets 1320 1140 1460 1350 1280 Age 8m 8m 8m 2.5 y 3 m5 y 10 y 3m 1y 18 y 113 y Gender M M M M-N M, F, N F F F M, F, N M, F, N Fleas Yes Yes Yes No No Yes Yes Yes No No Processing Directly Directly Directly Directly Directly Directly Directly Directly Stored at 20 8Cc Stored at 20 8C
a

Prague Suburban (Roztoky) Suburban (Roztoky) Suburban (Roztoky) Urban Strana) (Mala Suburban cholupy) (H. Me Urban (Vinohrady) Urban (Albertov) novice, Suburban (Kla . Most) C

Year 1995 1995 1995 1995 1995 1995 1995 1995 2000 2000

Stray cats: age range, 3 months12 years, mean age 27 months; shelter cats: age range, 3 months12 years, mean age 32 months; pet cats: age range, 112 years, mean age 36 months; CFU, colony forming units of B. henselae per ml of cat venous blood; Month: m; year: y; M: male; F: female; N: neutered cat; M-N: male neutered cat. a Processed directly after collection into home-made transport medium. b Kitten with cachexia. c Collected into Vacutainer tubes and stored at 20 8C until the collection was completed (up to 45 days) and then processed by plating on agar.

264

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

The plates were incubated at 35 8C in humid microaerobic atmosphere with 5% CO2 for 45 days. All colonies that had appeared on the media were examined by gram-staining, catalase (5% H2O2) and oxidase tests (Gordon and McLeods reaction, oxidase identication sticks, Oxoid). Each gram-negative coccobacillary isolate showing neither catalase nor oxidase activity was labeled with the same code number as its source blood specimen and was analyzed by the methods described later. CFUs per ml were counted from the average colony numbers growing on one plate of Columbia agar with 5% of sheep blood and one plate of chocolate agar with 5% of sheep blood. 2.3. Detection of enzymatic activity The isolates were tested by the Rapid Anaerobe Panel and Rapid Yeast Identication Panel (Baxter Diagnostics Inc., USA) for presumptive identication of B. henselae isolates as described by Welch et al. (1992, 1993). 2.4. Preparation of whole-cell DNA for PCR and ribotyping DNA was isolated by the phenolchloroform method (Sambrook et al., 1989) from 20 mg of cells grown for 7 days on Columbia agar with 5% sheep blood in a microaerobic atmosphere. Puried DNA was dissolved in 50 ml of 10 mM TrisHCl, pH 8, 1 mM EDTA. 2.5. Ribotyping Restriction enzymes HindIII, BglI and PvuII (MBI, Fermentas, Lithuania) were used. Digestion, separation of fragments, Southern blotting, 16S and 23S Escherichia coli rRNA probe preparation and hybridization were carried out using the Gerner-Smidt (1992) procedure slightly modied by Melter et al. (1999). RFLP proles and ribotypes, which differed from each other in position of one or more bands, were considered as different proles. 2.6. Amplification of a portion of citrate synthase gene (gltA) and digestion The method originally described by Regnery et al. (1991) was used with minor modications. The reaction mixture consisted of deoxynucleoside triphosphates (200 mM each), primers (25 mM each, Invitrogen, UK), 0.5 U Taq polymerase (Promega, USA). PCR amplications were performed in 100 ml reactions in a thermal cycler PTC100 (MJ Research Inc., USA). Approximately 100 ng of puried genomic DNA were added to the mixture. PCR amplications were performed with 40 cycles of denaturation (94 8C, 30 s), annealing (48 8C, 30 s) and extension (72 8C, 1 min) followed by a single nal extension (72 8C, 5 min). Amplied products were digested by TaqI restrictase (MBI, Fermentas) as recommended by the manufacturer. Ten microliter aliquots of each product were separated by electrophoresis in 3% Metaphor agarose (FMC, USA) at 150 V for 1 h. DNA restriction fragments of the isolates and those of B. henselae (Houston 1 and G9866), B. quintana, B. elizabethae and B. clarridgeiae were compared.

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

265

2.7. Polymorphism of 16S23S rRNA intergenic spacer region and 16S rRNA type The method originally described by Matar et al. (1993) was applied. The reaction mixture consisted of deoxynucleoside triphosphates (200 mM each), primers (100 pM each, Invitrogen, UK), Taq polymerase (Top-Bio, 1U, Czech Republic) and chromosomal DNA from B. henselae (100 ng) in 50 ml of Taq buffer (Top-Bio). PCR amplications were performed in a thermal cycler Mastercycler (Eppendorf, Germany). PCR amplication of the gene was performed with initial denaturation (95 8C, 2 min), followed by 30 cycles of denaturation (95 8C, 30 s), annealing (42 8C, 1 min) and extension (72 8C, 2 min), with a single nal extension (72 8C, 5 min). Amplicons were digested by AluI restriction endonuclease and separated by electrophoresis at 100 V for 1 h in 2% gel and the resulting proles were compared with those of other Bartonella strains. The method originally described by Bergmans et al. (1997) for the 16S rRNA typespecic amplications was applied with minor modications. Concentrations of all ingredients in the reaction mixture remained the same as for amplication of citrate synthase gene. PCR amplications were performed with initial denaturation (94 8C, 2 min) and 30 cycles of denaturation (94 8C, 30 s), annealing (55 8C, 30 s) and extension (72 8C, 1 min) followed by a single nal extension (72 8C, 5 min). 2.8. DNADNA hybridization The hydroxyapatite method was used as described by Brenner et al. (1993). The source of labeled DNA was isolate C27. DNADNA hybridizations were performed at 55 8C. 2.9. Pulsed field gel electrophoresis Bacteria cultured in a 5% CO2 atmosphere on Columbia agar with 5% sheep blood were harvested after 10 days. Due to strong autoagglutination of the cells, 20 mg aliquots were homogenised in Eppendorf tubes containing 10 mM TrisHCl (pH 8.0), 1 M NaCl with a strip of sterile nylon monol to be swirled with a glass stick. The re-suspended bacteria were pipetted while pressing the monol strip against the bottom of the tube. Further steps of DNA preparation and restriction digestion with enzyme SmaI (20U, New England Biolabs, USA) were performed as described by de Lencastre et al. (1994). The PFGE parameters were 21 h, 1.515 s, 6 V, 120 8C and 14 8C. The proles were evaluated visually and those differing in the positions of more than three bands were considered as different.

3. Results Five single cat isolates (C1, C2, C4, C26 and C27) (overall prevalence 8%) were obtained from 61 cat hemocultures. All of these isolates were obtained from directly plated stray or shelter cat specimens without previous freezing. No cultures were obtained from frozen samples C29C61 which were stored at 20 8C before plating until the collection was completed (145 days). Source of all these specimens were ea free pet cats.

266

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

Prevalence rates of bacteremia were 67% in stray cats (isolates C1, C2, C26 and C27) and 5% in shelter cats (isolate C4). Three male and two female cats suffered from bacteremia. Mean age of the bacteremic animals was 12 months, with an age range from 3 to 30 months. Four out of the ve bacteremic cats (80%) were 1-year-old or younger (Table 1). Isolates C1 and C2 were obtained from stray siblings living in the same locality. The other isolates were cultured from cats living in different parts of the city. All isolates originated from clinically healthy cats, except C26, cultured from an afebrile cachectic kitten whose mother was not infected with Bartonella. All isolates but C4 were from cats infested by Ctenocephalides felis. Shelter cats and pet cats did not harbour this parasite while investigated in contrast to all stray cats that were infested by eas (n 6) (Table 1). All isolates of B. henselae were detected in initial cultures after 7 days of incubation on both the culture media, except isolate C2, which appeared in initial culture on chocolate agar only. Interestingly, in all positive specimens, the concentrations of bacterial cells were estimated to be approximately 103 per ml of venous blood (Table 1). The bacteria adhered to the media surface, isolate C4 was even embedded in the medium. All isolates grew as a typical mixture of rough colonies ranging from 0.5 to 2 mm after 2 weeks of culture. The bacteria gave smooth and homogeneously sized colonies in following passages. All isolates shared an identical enzymatic activity prole (10077240) on the Rapid Anaerobe Panel. They were positive for bis-p-nitrophenyl-phosphate, L-leucine-bnaphthylamide, DL-methionine-b-naphthylamide, L-lysine-b-naphthylamide [alkaline], L-lysine-b-naphthylamide [acid], glycylglycine-b-naphthylamide, glycine-b-naphthylamide, L-arginine-b-naphthylamide and L-tryptophan-b-naphthylamide. Moreover, the isolates expressed enzymatic activity with glycyl-L-arginine-4-methoxy-b-naphthylamide, L-arginyl-L-arginine-b-naphthylamide, L-lysyl-L-alanine-4-methoxy-b-naphthylamide and L-alanine-4-methoxy-b-naphthylamide on the Rapid Yeast Identication Panel (prole 035700000). The HindIII ribotype was identical for both of the B. henselae strains (Houston-1 and G9866) and isolates and differed from those of B. quintana, B. elizabethae and B. clarridgeiae. Ribotypes BglI and PvuII allocated the isolates into two groups (Fig. 1, Table 2). The proles of gltA genes of the isolates were identical and specic for B. henselae. The RFLP proles of gltA genes of B. quintana, B. elizabethae and B. clarridgeiae differed from that inherent to B. henselae (Table 2). Amplied gltA gene of B. clarridgeiae gave a 380 bp amplicon which was not cut during enzymatic digestion with TaqI restrictase as described by Gureld et al. (1997). In agreement with these authors, a ne but irreproducible band (nearly 125 bp) was detected. Analysis of the 16S23S rRNA spacer region of all isolates studied (C1, C4, C26, C27) and B. henselae strain G9866 revealed a unique prole B and type strains ATCC 49882 revealed prole A as reported by Matar et al. (1993). Type strains of B. elizabethae, B. clarridgeiae, B. henselae and strain B. quintana CIP 103739 showed distinct proles (Table 2). All isolates studied belong to 16S rRNA type II as described by Bergmans et al. (1996). B. henselae type strain (Houston-1) belonged to type I and strain G9866 showed the characteristics of type II.

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

267

Fig. 1. (a) HindIII, PvuII and BglI ribotypes of cat isolates and Bartonella spp. strains molecular size marker, l/ HindIIIStyI; HindIII profiles: lanes 1 and 2, B. henselae strains Houston 1 and G 9866; lanes 37, cat isolates C1, C2, C4, C26 and C27; lane 8, B. elizabethae ATCC 49927; lane 9, B. clarridgeiae ATCC 51734; lane 10, B. quintana CIP 103739. PvuII profiles: lanes 11 and 12, B. henselae strains Houston 1 and G 9866; lanes 1316, cat isolates C1, C4, C26 and C27; lane 17: B. elizabethae ATCC 49927; lane 18, B. clarridgeiae ATCC 51734; lane 19, B. quintana CIP 103739; BglI profiles: lanes 20 and 21, B. henselae strains Houston 1 and G 9866; lane 2225, cat isolates C1, C4, C26 and C27; lane 26, B. elizabethae ATCC 49927; lane 27, B. clarridgeiae ATCC 51734; lane 28, B. quintana CIP 103739. (b) Dendrogram of the HindIII, PvuII and BglI ribotypes. The dendrogram was constructed by the unweighted pairgroup method using percent disagreement distance measure (software STATISTICA 5.5, StatSoft Inc.). Houston 1, G9866: B. henselae strains; C1, C4, C26, C27, cat isolates; BQ, B. quintana CIP 103739; BE, B. elizabethae ATCC 49927; BC, B. clarridgeiae ATCC 51734.

DNADNA hybridization relatedness values between isolate C27 and strain Houston 1, C4 isolate and C1 isolate was 99, 99 and 95%, respectively. Strains of the other species gave lower percentages of similarity: B. quintana (67%), B. elizabethae (59%) and B. clarridgeiae (34%) (Table 2).

268

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

Table 2 Genotype characteristics of cat isolates and Bartonella spp. strains Strain/isolate RFLP gltA TaqI A A A A A A A B C D RFLP 16S23S rRNA AluI A B B ND B B B Cb Db Eb 16S rRNA type Type Type Type ND Type Type Type ND ND ND I II II II II II Ribotypes HindIII PvuII A A A A A A A B C D A A B ND A B B C D E BglI A A B ND A B B C D E PFGE SmaI A B C ND D E F ND ND ND DNADNA hybridization (55 8C) 99 ND 95 ND 99 ND 100 59 67c 34

Bh Houston 1 Bh G9866 Isolate C1 Isolate C2a Isolate C4 Isolate C26 Isolate C27 B. elizabethae B. quintana B. clarridgeiae

Bh B. henselae, B. elisabethae ATCC 49927, B. quintana CIP 103739, B. clarridgeiae ATCC 51734, ND not determined. a Isolate C2: lost before the study was completed. b C, D, E profiles different from that described for B. henselae by other authors. c B. quintana strain VR 358 was used for DNADNA analysis.

PFGE showed a distinct prole for each of the isolates studied and B. henselae strains (Table 2, Fig. 2).

4. Discussion Although national data on B. henselae infection in humans were not available, the cooperating infectionists announced sporadically suspected cases of cat-scratch disease (CSD), usually manifested as lymphadenitis. We took this as a challenge to our experimental work. Our rst steps were to culture the agent and to determine its prevalence in natural reservoirs. Venous blood specimens of 61 cats of different categories were cultured and all gramnegative, oxidase and catalase negative, slow growing isolates were studied by polyphasic analysis. We were unable to obtain B. henselae from the specimens collected into Vacutainer tubes with EDTA, frozen and stored at 20 8C for up to 45 days (C29C61). Interestingly, other authors reported even enhanced culture sensitivity after previous freezing of cat venous blood specimens for 26 days (Brenner et al., 1997). These authors used a lower temperature to freeze (65 8C) the specimens collected into commercial tubes with EDTA. The source of all our frozen specimens were ea free pet cats (n 34) and their mean age, 36 months, was higher compared to that of Czech stray and shelter cats (27 and 32 months, respectively). We suspect that the storage of the specimens in the freezer had a negative effect on the number of viable cells. Nevertheless, potential effect of other variables is to be taken into account, such as higher mean age and the presence of eas which are risk factors for acquiring B. henselae infection by cats (Koehler et al., 1994; Chomel et al., 1995; Sander et al., 1997).

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

269

Fig. 2. (a) Schematic graph of PFGE profiles of SmaI digested DNA of B. henselae strains and cat isolates; lane 1, B. henselae Houston 1; lane 2, B. henselae G9866; lanes 36: cat isolates C1, C4, C26 and C27; lane 7, low range PFG marker (New England Biolabs, USA). (b) Dendrogram of the PFGE profiles. The dendrogram was constructed by the unweighted pairgroup method using percent disagreement distance measure (software STATISTICA 5.5, StatSoft Inc.). Houston 1, G9866, B. henselae strains; C1, C4, C26, C27, cat isolates.

270

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

The overall prevalence of bacteremia in cats was 8% while the specimens processed without previous freezing (C1C27) showed a prevalence rate of 18%. The latter blood specimens (C1C27) originated from six stray and 22 shelter cats. Each of the B. henselae isolates was cultured from a single cat. None of all 61 cats studied was infected or co-infected with B. clarridgeiae as described by Arvand et al. (2001) for German stray cats, by Bergmans et al. (1996) for Dutch shelter cats, by Chomel et al. (2002) for Danish cats, by Gureld et al. (1997, 2001) and Heller et al. (1997) for French cats. Neither co-infection with two different 16S rRNA genotypes of B. henselae, as described by Chomel et al. (2002) in American cats and by Gureld et al. (1997, 2001) in French cats, has been detected in our study. Proportions of bacteremic cats correlated with their category (four stray cats and one shelter cat) and age (four cats were 1-year-old or younger) but not with gender (three males and two females) as shown in Table 1. Sander et al. (1997) also conrmed more frequent bacteremia in younger domestic cats. Among 100 cats investigated, 18% of the cats younger than 2 years (i.e. 11 out of 60) and 5% of the cats aged above 2 years (2 out of 40) were bacteremic. In contrast to our results, female cats were infected more frequently (21%, 10 out of 47) compared to male cats (6%, 3 out of 53). Four out of the ve Czech bacteremic cats were infested with ectoparasites which seems to be an important risk factor for bacteremia as described by Koehler et al. (1994) and Chomel et al. (1996). The bacteremia prevalence rates were 67% for stray cats (four out of six cats) and 5% for shelter cats (one out of 22 cats). The bacteremia prevalence rate calculated for age group 1 year or younger was 80% (four from ve cats were bacteremic). Other authors reported different prevalence rates of B. henselae bacteremia in cats in Europe. In Germany, Sander et al. (1997) indicated 13% in Freiburg and Arvand et al. (2001) 1% for pet cats and 19% for stray cats in Berlin. In both, the Netherlands (Bergmans et al., 1997) and Denmark (Chomel, 2002), the B. henselae bacteremia in cat hemocultures was reported to be concordantly 22%. In France, 37% of stray cat hemocultures tested positive for this agent (Heller et al., 1997). Higher percentages of B. henselae bacteremic cats were reported in the USA: 41% (Koehler et al., 1994) and 40% (Chomel et al., 1996). In Australia this rate reached 33% (Branley et al., 1996). The enzymatic activities of the Czech isolates were characteristic of B. henselae but none of these isolates expressed activity to L-prolin-b-naphthylamide as reported by Gureld et al. (1997) and by Welch et al. (1993) for 96% of B. henselae isolates. RFLP of gltA gene proved useful in identication of B. henselae isolates as documented by others (Regnery et al., 1991; Joblet et al., 1995; Arvand et al., 1998). We could recommend this method for identication of Bartonella spp. isolates because of its sensitivity, simplicity and rapidity. RFLP analysis of 16S23S rRNA spacer region gave an identical prole B and analysis of 16S RNA conrmed type II for all cat isolates (C1, C4, C26, C27) and strain G8866 while the type strain showed prole A and type I (Table 2). All the isolates analyzed were of type II, i.e. the one described by European authors to be the most frequent among the cat isolates. Interestingly, Matar et al. (1993) could allocated eleven B. henselae strains to six proles. Bergmans et al. (1996) could differentiate two proles (A and B) in CSD patient samples and cat isolates. Their prole A isolates belonged to 16S rRNA type I and prole B isolates belonged to type II. Prole A prevailed in CSD patient specimens (75%), being

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

271

less frequent in Dutch cat specimens (28%). In contrast, prole B was identied in 25% of the patients and 56% of the cats (Bergmans et al., 1996, 1997). Chomel et al. (2002) reported that type II was inherent to 20 out of 21 cats B. henselae isolates (95%). Heller et al. (1997) detected close numbers of B. henselae strains of type I (17 strains) and type II (18 strains) in 94 stray cats. Gureld et al. (2001) identied 50 and 15% of B. henselae isolates from 72 French domestic cats as being of types II and I, respectively. Arvand et al. (2001) proved type I in 95% (n 19) and Sander et al. (1998) in 6% of the German cats studied. The variation in frequency of 16S rRNA types I and II reported by different authors could be explained probably by different numbers of isolates analysed. Maruyama et al. (2000) classied all 13 Japanese B. henselae isolates from four cats as belonging to type I. PFGE was the most discriminatory method to be able to assign a different prole to each of the B. henselae strains studied. This conclusion is in agreement with Roux and Raoult (1995) who analyzed four B. henselae strains only. Maruyama et al. (2000) did not detect any identical proles among American, French and Japanese isolates of B. henselae. Some Japanese isolates from the same cat shared the same PFGE prole. He also reported identical PFGE proles for some USA isolates of cat and human origins. All of the French isolates were of different PFGE proles. Sander et al. (1997) found 10 and Arvand et al. (2001) 11 proles in their sets of 18 and 24 strains, respectively. To our knowledge, ribotyping was used for the rst time in the present study for analysis of B. henselae isolates. Ribotypes PvuII, BglI distinguished two types among the isolates analyzed and HindIII identied them to the species level. Although time-consuming, analysis of a larger number of B. henselae strains could prove reliability of HindIII ribotyping as an identication tool for B. henselae. All the methods described above were capable either of identifying or of typing the B. henselae isolates and could be useful in ecology of the agent in cats or in nding possible epidemiological links between feline and human infections.

Acknowledgements The study was supported by Grants No. 310/98/0417 awarded to P. Branny and No. 204/ 02/D121 awarded to K. Herc k by the Grant Agency of the Czech Republic. We are extremely grateful to Dr. J. Schindler, Third Medical Faculty of Medicine, Charles University, Prague, and Dr. D.J. Brenner, Centers for Disease Control and Prevention, Atlanta for their kind support and advice. We also thank E. Kodytkova for revision of the manuscript and Dr. V. Naxera, Veterinary Ambulance, Prague, for valuable help in contacting cat breeders.

References
Arvand, M., Wendt, C., Regnath, T., Ullrich, R., Hahn, H., 1998. Characterization of Bartonella henselae isolated from bacillary angiomatosis lesions in a human immunodeficiency virus-infected patient in Germany. Clin. Infect. Dis. 26, 12961299.

272

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

Arvand, M., Klose, A.J., Schwartz-Porsche, D., Hahn, H., Wendt, C., 2001. Genetic variability and prevalence of Bartonella henselae in cats in Berlin, Germany, and analysis of its genetic relatedness to a strain from Berlin that is pathogenic for humans. J. Clin. Microbiol. 39, 743746. Bergmans, A.M.C., Schellekens, J.F.P., Van Embden, J.D.A., Schouls, L.M., 1996. Predominance of two Bartonella henselae variants among cat-scratch disease patients in the Netherlands. J. Clin. Microbiol. 34, 254260. Bergmans, A.M.C., Jong, C.M.A., Amerongen, G., Schot, C.S., Schouls, L.M., 1997. Prevalence of Bartonella species in domestic cats in the Netherlands. J. Clin. Microbiol. 35, 22562261. Birtles, R., Raoult, D., 1998. The genus Afipia nad Bartonella. In: Schmidt, A. (Ed.), Bartonella and Afipia Species Emphasizing Bartonella henselae. Karger, Basel, pp. 131. Branley, J., Wolfson, C., Waters, P., Gotlieb, P., Bradbury, R., 1996. The prevalence of Bartonella henselae in an Australian cat population. Pathology 28, 262265. Brenner, D.J., OConnor, S.P., Winkler, H.H., Steigerwalt, A.G., 1993. Proposals to unify the genera Bartonella and Rochalimaea, with description of Bartonella quintana comb. nov., Bartonella vinsonii comb. nov., and Bartonella elizabethae comb. nov., and to remove the family Bartonellaceae from the order Rickettsiales. Int. J. Syst. Bacteriol. 43, 777786. Brenner, S.A., Rooney, J.A., Manzewitsch, P., Regnery, R.L., 1997. Isolation of Bartonella (Rochalimaeae) henselae: effects of methods of blood collection and handling. J. Clin. Microbiol. 35, 544547. Child, J.E., Rooney, J.A., Cooper, J.L., Olson, J.G., Regnery, R.L., 1994. Epidemiologic observation of infection with Rochalimaea species among cats living in Baltimore, MD. JAVMA 204 (11), 17751778. Chomel, B.B., Abbott, R.C., Kasten, R.W., Floyd-Hawkins, K., Kass, P.H., Glasser, C.A., Pedersen, N.C., Koehler, J.E., 1995. Bartonella henselae prevalence in domestic cats in California: risk factors and association between bacteremia and antibody titers. J. Clin. Microbiol. 33, 24452450. Chomel, B.B., Kasten, R.W., Floyd-Hawkins, K., Chi, B., Yamamoto, K., Robert-Wilson, J., Gurfield, A.N., Abbott, R.C., Pedersen, N.C., Koehler, J.E., 1996. Experimental transmission of Bartonella henselae by cat flea. J. Clin. Microbiol. 34, 19521956. Chomel, B.B., Bolous, H.J., Petersen, H., Kasten, R.W., Yamamoto, K., Chang, C.C., Gandoin, C., Boullin, C., Hew, C.M., 2002. Prevalence of Bartonella infection in domestic cats in Denmark. Vet. Res. 2, 205 213. de Lencastre, H., Couto, I., Santos, I., Melo-Cristino, J., Torres-Pereira, A., Tomasz, A., 1994. Methicillinresistant Staphylococcus aureus clonal disease in a Portuguese hospital: characterization of clonal types by combination of DNA typing methods. Eur. J. Clin. Microbiol. Infect. Dis. 13, 6473. Droz, S., Chi, B., Horn, E., Steigerwald, A.G., Whitney, A.N., Brenner, D.J., 1999. Bartonella koehlerae sp. nov., isolated from cats. J. Clin. Microbiol. 37, 11171122. Fournier, P.E., Raoult, D., 1998. Cat-scratch disease and an overview of other Bartonella henselae-related infections. In: Schmidt, A. (Ed.), Bartonella and Afipia Species Emphasizing Bartonella henselae. Karger, Basel, pp. 3262. Gerner-Smidt, P., 1992. Ribotyping of the Acinetobacter calcoaceticusAcinetobacter baumannii complex. J. Clin. Microbiol. 30, 26802685. Groves, M.G., Harrington, K.S., 1994. Rochalimaea henselae infections: newly recognized zoonoses transmitted by domestic cats. JAVMA 204 (2), 267271. Gurfield, A.N., Boulouis, H., Chomel, B.B., Heller, R., Kasten, R.W., Yamamoto, K., Piemont, Y., 1997. Coinfection with Bartonella clarridgeiae and Bartonella henselae and with different Bartonella henselae strains in domestic cats. J. Clin. Microbiol. 35, 21202123. Gurfield, A.N., Boulouis, H., Chomel, B.B., Kasten, R.W., Heller, R., Bouillin, C., Gandoin, C., Thibault, D., Chang, C.C., Barrat, F., Piemont, Y., 2001. Epidemiology of Bartonella infection in domestic cats in France. Vet. Microbiol. 80, 185198. Heller, R., Artois, M., Xemar, V., De Briel, D., Gehin, H., Jaulhac, B., Monteil, H., Piemont, Y., 1997. Prevalence of Bartonella henselae and Bartonella clarridgeiae in stray cats. J. Clin. Microbiol. 35, 1327 1331. Joblet, C., Roux, V., Drancourt, M., Gouvernet, J., Raoult, D., 1995. Identification of Bartonella (Rochalimaea) species among fastidious gram-negative bacteria on the basis of the partial sequence of the citrate synthase gene. J. Clin. Microbiol. 33, 18791883. Koehler, J.E., Glaser, C.A., Tappero, J.W., 1994. Rochalimaea henselae infection. JAMA 271, 531535.

O. Melter et al. / Veterinary Microbiology 93 (2003) 261273

273

Maruyama, S., Kasten, R.W., Boulouis, H.J., Gurfield, N.A., Katsube, Y., Chomel, B.B., 2000. Genomoc diversity of Bartonella henselae isolates from domestic cats from Japan, the USA and France by pulsed-field gel electrophoresis. Vet. Microbiol. 79, 337349. Matar, G.M., Swaminathan, B., Hunter, S.B., Slater, L.N., Welch, D.F., 1993. Polymerase chain reaction-based restriction fragment length polymorphism analysis of a fragment of the ribosomal operon from Rochalimaea species for subtyping. J. Clin. Microbiol. 31, 17301734. Melter, O., Santos Sanches, I., Schindler, J., Aires de Sousa, M., Mato, R., Kovarova, V., Zemlickova, H., Lencastre, H., 1999. Methicillin-resistant Staphylococcus aureus clonal types in the Czech Republic. J. Clin. Microbiol. 37, 27982803. Regnery, R.L., Spruill, C.L., Plikaytis, B.D., 1991. Genotypic identification of Rickettsiae and estimation of intraspecies sequence divergence for portions of two rickettsial genes. J. Bacteriol. 173 (5), 15761589. Roux, V., Raoult, D., 1995. Inter and intraspecies identification of Bartonella (Rochalimaea) species. J. Clin. Microbiol. 33, 15731579. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning, vol. 3, 2nd ed. Cold Spring Harbor Laboratory Press, New York, p. F3. hler, K., Pelz, K., Gramm, E., Brendt, W., 1997. Detection and identification of two Bartonella Sander, A., Bu henselae variants in domestic cats in Germany. J. Clin. Microbiol. 35, 584587. Sander, A., Ruess, M., Bereswill, S., Schuppler, M., Steinbrueckner, B., 1998. Comparison of different DNA fingerprinting techniques for molecular typing of Bartonella henselae isolates. J. Clin. Microbiol. 36, 2973 2981. Welch, D.F., Pickett, D.A., Slater, L.N., Steigerwalt, A.G., Brenner, D.J., 1992. Rochalimaea henselae sp. nov., a cause of septicemia, bacillary angiomatosis, and parenchymal bacillary peliosis. J. Clin. Microbiol. 30, 275 280. Welch, D.F., Hensel, D.M., Pickett, D.A., San Joaquin, V.H., Robinson, A., Slater, L.N., 1993. Bacteremia due to Rochalimaea henselae in a child: practical identification of isolates in the clinical laboratory. J. Clin. Microbiol. 31, 23812386. Welch, D.F., Carroll, K.C., Hofmeister, E.K., Persing, D.H., Robison, D.A., Steigerwalt, A.G., Brenner, D.J., 1999. Isolation of a new subspecies, Bartonella vinsonii subsp. Arupensis, from a cattle rancher: identity with isolates found in conjunction with Borrelia burgdorferi and Babesia microti among naturally infected mice. J. Clin. Microbiol. 37, 25982601. Zangwill, K.M., Hamilton, D.H., Perkins, B.A., Regnery, R.L., Plikaytis, B.D., Hadler, J.L., Cartter, M.L., Wenger, J.D., 1993. Cat-scratch disease in Connecticut. New Engl. J. Med. 239 (1), 813.