Article — Artikel

Evaluation of PCR assays for the detection of fetus in bovine preputial scrapings and the identification of subspecies in South African field isolates

T Schmidta*, E H Venterb and J A Picardb

upon the collection and maintenance of ABSTRACT the fragile C. fetus bacterium, which has As a result of the high lability and slow growth of Campylobacter fetus subspecies, the laboratory limited viability outside the host8. Most of diagnosis of bovine genital has always been difficult. This is especially the cattle-breeding farms in South Africa true under South African conditions, where farms are far apart, laboratories are only present tend to be extensive in nature and in major centres and there are high ambient temperatures. In order to overcome the short- located long distances from diagnostic comings associated with traditional diagnostic methods, the implementation of a molecular laboratories. In many cases the consulting assay was sought. This work describes how a previously published PCR assay (MG3F/ veterinarian may need to travel long MG4R primers) was adapted, optimised and applied in the diagnostic laboratory to test distances to sample animals and the task preputial samples directly for the presence of Campylobacter fetus. Field evaluation of the of ensuring that samples reach the labora- assay revealed an analytical sensitivity and specificity of 85.7 % and 99 %, respectively. tory timeously becomes a logistical Subsequent genotyping and phenotyping of a diverse collection of South African field dilemma. In order to address these prob- isolates revealed that South Africa has an unexpected and previously unreported high incidence of Campylobacter fetus subsp. venerealis biovar intermedius strains. These strains lems, alternative diagnostic methods were not identified correctly by the subspecies-specific primer set evaluated. Until such have been sought. time that cost- effective genotyping methods are available to diagnostic laboratories in Polymerase chain reaction (PCR) assays South Africa, and other countries with these atypical Campylobacter fetus subsp. venerealis present a number of advantages over strains, the need for bacterial culture will persist. Identification to subspecies level of iso- more traditional methods and the tech- lates at present remains dependent upon a single phenotypic criterion, namely tolerance to nology is becoming increasingly more 1 % glycine. accessible to diagnostic laboratories. Keywords: bovine genital campylobacteriosis, Campylobacter fetus fetus, Campylobacter fetus Apart from exhibiting high specificities venerealis biovar intermedius, culture, polymerase chain reaction. and sensitivities, the ability of these types of assays to detect non-viable organisms Schmidt T, Venter E H, Picard J A Evaluation of PCR assays for the detection of could also potentially resolve problems Campylobacter fetus in bovine preputial scrapings and the identification of subspecies in encountered in the field with the preser- South African field isolates. Journal of the South African Veterinary Association (2010) 81(2): vation of organism viability for subsequent 87–92 (En.). Allerton Provincial Veterinary Laboratory,Private Bag X2, Cascades, 3202 South 22 Africa. bacteriological culture . A number of C. fetus-species- and sub- species-specific PCR assays have been de- scribed. Of these the multiplex PCR assay INTRODUCTION economic importance to the cattle industry initially evaluated by Hum and colleagues5 Genital campylobacteriosis is a trans- worldwide1 and is considered to be one of has been described extensively. This assay, missible genital disease of cattle caused by the most important infectious causes of which is designed to identify and differ- the bacterium Campylobacter fetus subsp. poor calving rates in southern Africa. entiate the 2 C. fetus subspecies, has been venerealis (Cfv). The disease is characterised Regional veterinary laboratories in South evaluated by various researchers and by temporary infertility of female cattle, Africa and Namibia have reported be- been shown to be both sensitive and spe- early embryonic mortality, irregular tween 0 and 12 % prevalences based on cific13,20. Good correlation of subspecies oestrus cycles, delayed conception and bacterial isolations made from bull sheath identification results, albeit not 100 %, infrequent abortions6. Most cows recover wash submissions. One particular study was achieved when the PCR method was from infection and oestrus cycles will conducted on bulls which were commu- compared with traditional biochemical recommence after a few months but the nally grazed, reported a substantially and genotyping methods, including herd experiences a decreased pregnancy higher prevalence of almost 29 %14.Itis random amplification polymorphic DNA rate and a prolonged calving season8. believed that the availability of (RAPD-PCR); pulsed field gel electropho- Infected bulls usually show no clinical and practices have resis (PFGE) and amplified fragment signs of infection but become carriers and decreased the prevalence of the disease15 length polymorphism (AFLP)10,13. infect females at service2. but many researchers still consider the The aim of this investigation was to Campylobacteriosis is of considerable prevalence of bovine venereal campylo- assess the suitability of the C. fetus-specific bacteriosis to be underestimated as a primer set, initially evaluated by Hum aAllerton Provincial Veterinary Laboratory, Private Bag consequence of the lack of sensitive and and colleagues5, to detect C. fetus directly X2, Cascades, 3202 South Africa. 4 bDepartment of Veterinary Tropical Diseases, Faculty of reliable techniques for diagnosis . in bovine preputial specimens. The multi- Veterinary Science, University of Pretoria, Private Bag Currently diagnosis of genital campylo- plex assay was used in conjunction with X04, Onderstepoort, 0110 South Africa. bacteriosis relies on the cultivation and traditional phenotyping tests to establish *Author for correspondence. E-mail: [email protected] identification of the causative organ- the suitability of the assay to assign South Received: February 2010. Accepted: April 2010. ism1. The success hereof is dependent African C. fetus field isolates to subspecies.

0038-2809 Jl S.Afr.vet.Ass. (2010) 81(2): 87–92 87 Table 1: PCR results obtained using the species-specific primer set (MG3F and MG4R). (VenSF (5’-CTTAGCAGTTTGCGATAT TGCCATT-3’) and VenSR (5’-GCTTTTG Microorganism Source of isolate PCR result* AGATAACAATAAGAGCTT-3’) primer (MG3F and MG4R species-specific set). The test was performed as described 5 primers) previously with no modifications . PCR products were analysed by electropho- Campylobacter fetus subsp. fetus (Cff) NCTC 10842 (ATCC 27354) + resis using 1.5 % agarose gels in 1 X Cff 5515 (ATCC 33247) + Tris-Borate-EDTA (TBE) buffer at 100 V. Campylobacter fetus subsp. venerealis (Cfv) NCTC 10354 (ATCC 19438) + Following staining in ethidium bromide Cfv LMG 6570 + (0.5 µg/m ), the DNA bands were visual- Campylobacter fetus subsp. venerealis Unknown + ised using a gel documentation system bv intermedius (Cfv-i) and sized against a 100-bp DNA ladder Cfv-i Unknown + C. fetus South African field isolates (n = 40)† + (Fermentas). C. jejuni Avian intestine – C. sputorum subsp. bubulus Bovine preputial specimen – Evaluation of the specificity of the PCR C. sputorum subsp. bubulus Bovine preputial specimen – The specificity of the C. fetus-specific C. jejuni Unknown – uniplex PCR was determined by testing C. coli ATCC 43478 – C. fetus reference and South African field C. jejuni ATCC 29428 – isolates (n = 46) in addition to a collection C. sputorum subsp. bubulus Bovine preputial specimen – of other Campylobacter species (n = 10) and C. hyointestinalis Bovine faeces – organisms which may be encountered in C. sputorum subsp. bubulus Bovine preputial specimen – the genital tract of cattle (n = 18). Further- C. hyointestinalis Unknown – more, amplicons from one of the Cfv Brevundimonas vesicularis Bovine preputial specimen – reference strains and a field isolate were Brevundimonas vesicularis Bovine preputial specimen – sent for sequencing. Stenotrophomonas maltophilia Bovine preputial specimen – Arcobacter skirrowii Bovine intestine – Evaluation of the sensitivity of the PCR Corynebacterium striatum Bovine preputial specimen – and the influence of potential inhibitors Histophilus somni Bovine vaginal aspirate – Histophilus somni Bovine preputial specimen – Dilutions of a Cfv reference strain Ochrobacter sp. Bovine faeces – (ATCC 19438), prepared in modified 7 Histophilus somni ATCC 70025 – Weybridge transport medium ,were Staphylococcus aureus ATCC 25923 – subjected to the PCR. Aliquots of each Escherichia coli ATCC 25922 – dilution were also plated out on blood Pseudomonas aeruginosa ATCC 27853 – agar and incubated at 37 °C in a micro- Staphylococcus epidermidis ATCC 12228 – aerophilic atmosphere for 72 h to deter- Bovine preputial specimen – mine bacterial counts. The detection limit Tritrichomonas foetus Bovine preputial specimen – of the PCR assay was taken as the lowest Arcanobacterium pyogenes ATCC 19411 – bacterial inoculum (cfu/m ) that yielded a Proteus sp. Unknown – positive PCR result. Once the detection Enterococcus faecalis ATCC 29212 – limit was established, the 2 dilutions immediately above the limit were spiked *+ = Positive; – = Negative †Details of origin given in Table 4. with varying concentrations (1 %, 2 %, 5 %, 10 %, 20 % and 50 % (v/v)) of blood, MATERIALS AND METHODS and tissue kit). Extraction was carried out urine, faeces and semen. Aliquots of each according to the manufacturer’s instruc- spiked sample were subjected to PCR. Bacterial and protozoan cultures tions except when processing diagnostic Forty-six Campylobacter fetus isolates and specimens where 500 µ of sample was sub- Evaluation of the sensitivity of PCR and 27 bacterial and protozoan isolates, jected to extraction and final elution of culture assays using spiked bovine genomically similar to C. fetus or occupying DNA was done using only 100 µ of the preputial specimens a similar biological niche, were obtained elution buffer. Preputial specimens were collected from reference collections or were isolated from bulls younger than 2 years of age from diagnostic material (Table 1). All bac- PCR protocol and which had previously been tested terial isolates were grown on Columbia The C. fetus-specific uniplex PCR using and found to be negative for C. fetus. agar (Oxoid) supplemented with 5 % the MG3F(5’-GGTAGCCGCAGCTGC Between 4 and 8 animals were sampled sheep blood, at 37 °C under microaerophilic TAA GAT-3’) and MG4R (5’-TAGCTA and the preputial material inoculated into atmospheric conditions. Microaerophilic CAATAACGACAACT-3’) primers was car- Weybridge medium. Samples were atmospheric conditions were generated ried out as described by Schulze and col- pooled in the laboratory before being in anaerobic jars using Campylobacter gas leagues16. For the processing of diagnostic divided into aliquots and inoculated with generating kits (Oxoid). The Tritrichomonas specimens, the volume of DNA template dilutions of the Cfv reference strain foetus isolates were maintained in Tricho- added to the reaction mixture was 0.2 µ (ATCC 19438) prepared in phosphate monas medium (Oxoid) at 37 °C. and the volume of Taq polymerase incor- buffered saline containing 0.02 % porated into the reaction mixture was Tween 80. Inoculated aliquots were kept PCR template preparation increased to 0.6 µ (Roche 1 IU/µ ). The at room temperature for 72 h. During this DNA was isolated from diagnostic spec- multiplex PCR was used to identify bacte- time, material was removed from each imens and bacterial suspensions using a rial isolates to both species (MG3F and aliquot at 0, 24, 48 and 72 h post-inocula- commercial kit (Qiagen® DNeasy blood MG4R primer set) and subspecies level tion for PCR and culture. For bacterio-

88 0038-2809 Tydskr.S.Afr.vet.Ver. (2010) 81(2): 87–92 Table 2: Detection limits of Campylobacter fetus subsp. venerealis (cfu/m ) in the culture and PCR assays obtained over a period of 72 hours using spiked preputial material.

Culture PCR 0 h 24 h 48 h 72 h 0 h 24 h 48 h 72 h

First run 77 000† * ND * 77 77 ND 0 Second run 0.6 6 60 60 6 60 6 6 Third run 1 1000 100 1000 10 10 10 10

†Isolated only after filtration onto enriched BCA, not by direct plating onto Skirrow’s agar. *Owing to problems with Pseudomonas contamination, Cfv could not be recovered. ND: not tested. logical culture, 100 µ of each aliquot was of approximately 750-bp when tested tion limit of both the culture and PCR spread plated onto Skirrow’s agar against all C. fetus isolates. No amplicons assays were monitored over a period of (Oxoid) and approximately 300 µ filtered were observed when any of the other 72 hours and recorded as the lowest Cfv (0.65 µm cellulose acetate filter) onto Campylobacter strains or the collection of inoculum (Cfv/m ) that yielded a positive blood agar supplemented with FBP genital-associated microbes which were result. Results are summarised in Table 2. Campylobacter enrichment supplement tested. A summary of the results is illus- The 1st trial run was hampered by the (Oxoid). Following incubation at 37 °C trated in Table 1. presence of Pseudomonas aeruginosa in the under microaerophilic conditions, all The specificity of the primer pair was samples. All Skirrow’s plates were over- plates were examined for colonies typical further demonstrated by DNA sequenc- grown with Ps. aeruginosa preventing the of C. fetus, Gram-stained and counted11. ing and alignment of the PCR amplifica- detection and recovery of any C. fetus tion products from 1 of the Cfv reference colonies. Field evaluation strains (LMG 6570) and a field isolate All bovine preputial scrapes (n = 212) (Inqaba Biotec). A BLAST search, using Field evaluation submitted in Weybridge transport me- the assembled sequencing data, revealed During the investigation period 212 bull dium to Allerton Provincial Veterinary 100 % similarity between the 2 sequences preputial scrapes were received and Laboratory, KwaZulu-Natal, South Africa obtained and C. fetus sequences loaded processed using PCR and bacteriological between June 2007 and March 2008 were in GenBank (Accession Numbers: CP000- culture. The results obtained using both tested using the uniplex PCR assay as well 487.1; AY158814.1 and AY 158813.1). methods are presented in Table 3. as the traditional bacteriological culture The determination of sensitivity of the From this data the analytical sensitivity method. It was decided to only test PCR revealed a detection limit of 615 and specificity of the PCR assay was preputial scrapes in modified Weybridge cells/m Weybridge medium, or 6.15 cells calculated using guidelines12. A sensitivity transport medium as this has been per PCR reaction. The 2 Cfv dilutions of 85.7 % and a specificity of 99 % were proven to give the highest sensitivity above the detection limit were spiked obtained. when culturing is used8,17. All samples with varying concentrations of faeces, were processed within 36 hours of collec- blood, urine and semen. Significant inhi- Identification of subspecies tion. Following incubation, all plates were bition of the PCR reaction was noted for The multiplex PCR was evaluated using examined for the presence of small, samples spiked with faeces; a 10-fold 6 reference C. fetus isolates representing smooth, shiny colonies having a slightly decrease in sensitivity was observed with both sub species as well as C. fetus subsp. grey to pink appearance. Suspect colo- aliquots spiked with 1 % faeces and com- venerealis biovar intermedius (Cfv-i). A nies were Gram-stained and identified as plete inhibition observed with samples single PCR amplicon of approximately C. fetus if they were oxidase positive, grew spiked with 2 % faecal material. Con- 750-bp was obtained with all 6 cultures. A at 25 °C, were resistant to nalidixic acid versely, blood, urine and semen seem- 2nd amplicon of approximately 180-bp and sensitive to cephalothin and did not ingly had no impact on the PCR results at was obtained with both of the Cfv isolates produce hydrogen sulphide in triple the concentrations tested even when the but not the Cff or the Cfv-i reference iso- sugar iron agar11. test aliquots were spiked with as much as lates. 50 % of the potential inhibitor. Phenotyping and PCR subspecies iden- Identification of subspecies tification results for the 6 C. fetus reference Campylobacter fetus reference and field Evaluation of the sensitivity of PCR cultures and 40 field isolates are summa- isolates were identified to subspecies and culture assays using spiked rised in Table 4. Based on the pivotal level using the multiplex PCR and 2 bovine preputial specimens glycine tolerance test, 5 of the 40 field phenotyping tests namely: tolerance to The evaluation of sensitivity of the PCR isolates were identified as Cff. All of the 1 % glycine (BDH) and hydrogen sul- on spiked bovine specimens were carried remaining isolates were unable to grow phide production using lead acetate pa- out in triplicate using preputial material on blood agar supplemented with 1 % per11. Additionally the ability to reduce sel- collected from different herds. The detec- glycine; by virtue of the fact that all of enite and susceptibility to metronidazole and cefoperazone was determined11,16. Table 3: Comparison of the bacterial culture and PCR results obtained for the screening of Campylobacter fetus in clinical specimens. RESULTS Culture-positive Culture-negative Total PCR specificity and sensitivity PCR positive 7 2 9 The MG3F and MG4R C. fetus-specific PCR negative 1 202 203 primers were shown to be highly specific Total 8 204 212 yielding a single, distinct PCR amplicon

0038-2809 Jl S.Afr.vet.Ass. (2010) 81(2): 87–92 89 these isolates produced hydrogen with the uniplex PCR (Table 1). None of Pseudomonas aeruginosa are known to sulphide in L-cysteine supplemented the other isolates tested produced ampli- produce the enzyme DNase3. The activity medium, they were classified as Cfv-i. cons, indicating that the test has an of this enzyme could have resulted in Using the multiplex PCR assay all iso- analytical specificity of 100 %. It is inter- the degradation of DNA present in the lates yielded the same results: spe- esting to note that initial publications5,13,20 sample. Problems resulting from the con- cies-specific amplicons were observed reported the generation of a PCR ampli- tamination of preputial cultures by Pseu- but no subspecies-specific amplicons. The con of about 960-bp in size. The amplicon domonas spp. have been reported8.Itis VenSF and VenSR subspecies-specific has, however, been more recently reported interesting to note, however, that none of primer set used in the assay targets a seg- and sequenced as 750-bp9,12,23. The reason the diagnostic samples tested during field ment of the Cfv genome. The generation for the discrepancy of almost 200-bp in evaluation exhibited such extensive Pseu- of an amplicon of approximately 182-bp is size has not been explained. domonas contamination as was noted with indicative that the test organism is Cfv. During sample collection preputial ma- this test run, suggesting that in practice Campylobacter fetus subsp. fetus is identi- terial can quite easily become contami- this type of contamination may be a rare fied by exclusion. Based on the PCR test nated with urine, semen, faeces and/or occurrence. results obtained, all the field isolates blood. All 4 of these components have Owing to geographical and other logis- should be classified as Cff. been documented as containing elements tical reasons, private practitioners in which may interfere with PCR tests24.It South Africa typically face difficulties in DISCUSSION was consequently deemed necessary to getting specimens to the laboratory In veterinary diagnostic laboratories the evaluate the influence of each of the within the 6-hour viability period of the culture and isolation of C. fetus is the tradi- components on the sensitivity of the PCR bacterium. The use of a transport enrich- tional method for the diagnosis of bovine assay. The detection limit of the assay was ment medium (TEM) is a prerequisite as it campylobacteriosis. The inherent limita- seemingly unaffected when samples permits a more lenient window period tions of this methodology prompted this spiked with up to 50 % of urine, semen or between sample collection and laboratory investigation to identify and evaluate a blood were tested. Faecal contamination, testing to be observed. As shown by the more sensitive means of detecting C. fetus. however, was shown to have a major test results obtained here, the detection Molecular techniques, particularly PCR, inhibitory effect on the assay even at limit of the culture assay decreased with are becoming more practical and afford- concentration levels as low as 1 %. The time, even when use was made of a TEM. able for diagnostic laboratories to incorpo- use of an alternative DNA extraction Nonetheless, Cfv was still recoverable rate into their test portfolio. The specificity protocol could be investigated as a means from samples up to 72 hours post-inocula- and sensitivity of assays, combined with of overcoming this potential problem but tion. Comparatively, the sensitivity of the the speed at which samples can be pro- the inclusion of an internal control in the PCR did not appear to be affected by time, cessed and results obtained, are appeal- PCR assay would probably be the most with the detection limit remaining un- ing attributes22. feasible way of ensuring that false nega- changed between 0 and 72 hours. The The multiplex PCR initially evaluated tive results are not reported as a result of greater sensitivity achieved with PCR at by Hum and colleagues5 is the most exten- PCR failure due to the presence of inhibi- 48 and 72 hours post-inoculation of sam- sively described assay for the identifica- tors in the samples. ples (Table 2) is of particular interest as tion of C. fetus at species and subspecies The spiking of preputial material was these are the time intervals more likely to level. In the multiplex format many inves- carried out in triplicate using material be experienced in the field. tigators have used the PCR for the direct collected by different practitioners and Field evaluation of the PCR assay was identification of subspecies of bacterial from animals in different herds. Sampling carried out over a 10-month period using isolates. No publications to date have criteria were instituted to ensure as far as diagnostic specimens submitted to Allerton reported on the use of this PCR for the possible that only animals negative for Provincial Veterinary Laboratory. All direct screening of clinical specimens. The C. fetus were sampled. The 1st run of bovine sheath scrapes submitted in Wey- ability to screen clinical specimens directly spiked preputial material clearly illus- bridge transport medium were tested could potentially improve the overall trated the advantage of PCR over bacte- using both the culture method and the sensitivity and accuracy of diagnostic rial culture as a means of detecting C. PCR assay. Of the 212 samples received results as many of the shortcomings fetus. The presence of Pseudomonas in the and tested, 4.2 % were found to be posi- conventionally encountered with the preputial material dramatically impeded tive using the PCR assay while 3.8 % were culture method would be eliminated. As the recovery of Cfv. Pseudomonas colonies, found to be positive using the culture a possible screening assay only the resistant to the in the Weybridge method. A high degree of agreement was species-specific primer set was evaluated. transport and Skirrow’s media, readily observed between the PCR and the culture It was reasoned that as a result of the grew on the culture plates, obscuring any results (Table 3). Discrepancies between reportedly low prevalence of C. fetus in Cfv colonies which were plated out. results were only noted on 2 occasions. In South Africa6 and the improved test sensi- Using the PCR assay however, Cfv was the 1st instance 1 of the samples was tivity achieved with uniplex test formats16 detected at 77 Cfv/m (Table 2) at both 0 h found to be positive by PCR but negative it would be more feasible to screen and 24 h. Surprisingly, no amplicons were on culture. Both tests were repeated with samples using only the C. fetus-specific obtained when the 72 h spiked aliquots the same results being obtained. This primer set. was tested. In view of the fact that PCR discrepancy is possibly a demonstration The specificity of the species-specific amplicons were visualised when the 72 h of the higher sensitivity of the PCR assay; uniplex PCR was demonstrated by testing spiked aliquots from both subsequent test illustrating the advantages of the PCR in a collection of reference and field C. fetus runs were analysed, it is hypothesised detecting C. fetus in low numbers where isolates as well as taxonomically that the failure to detect Cfv is a result of the culture could possibly fail, especially related to C. fetus or known to exist in the presence of Pseudomonas and possibly in the presence of large numbers of con- similar microbiological niches. All C. fetus other contaminants that had increased in taminant bacteria or where non-viable isolates yielded a single PCR amplicon of number in the pooled sample. bacteria are present. The converse situation approximately 750-bp in size when tested Certain strains of Pseudomonas such as was observed with the 2nd anomalous

90 0038-2809 Tydskr.S.Afr.vet.Ver. (2010) 81(2): 87–92 Table 4: Subspecies identification of Campylobacter fetus isolates using phenotyping the PCR and culture results was surpris- characterisation tests and a PCR assay (shaded area = the reference strains). ing, particularly since the shortcomings associated with the culture method are Source of isolate PCR results* PCR Phenotyping* Phenotyping well documented8. The success achieved subspecies ID tests subspecies ID results results with culturing is attributed in part to the fact that the laboratory imposed rigorous sampling guidelines; only sheath scrape samples submitted in Weybridge trans- port medium were accepted. Sheath

primer pair scrapes are preferable to sheath washes, a technique still used by many local practi- tioners, since the scraping technique has C. fetus (MG3F/MG4R) Subspecies primer pair (VenSF/VenSR) Hydrogen sulphide production Growth on 1% glycine medium been shown to recover larger numbers of NCTC 10842 + – Cff + + Cff Cfv and provide ‘cleaner’ samples with ATCC 33247 + – Cff + + Cff less bacterial contamination17. In addition ATCC 19438 + + Cfv – – Cfv to sampling criteria, the culture proce- LMG 6570 + + Cfv – – Cfv dure in the laboratory was enhanced by 10 + – Cff + – Cfv-i 136 + – Cff + – Cfv-i ensuring that all samples were plated out KwaZulu-Natal 1991 + – Cff + + Cff onto 2 different media; 1 selective and KwaZulu-Natal 1996 + – Cff + – Cfv-i 1 FBP-enriched medium. This practice is KwaZulu-Natal 2006 + – Cff + – Cfv-i often compromised, particularly in some Western Cape + – Cff + + Cff of the smaller regional veterinary labora- Limpopo 2006 + – Cff + – Cfv-i tories where budget constraints are prev- Mpumalanga 2006 + – Cff + – Cfv-i alent. Mpumalanga 2006 + – Cff + – Cfv-i In order to evaluate the described sub- KwaZulu-Natal 2006 + – Cff + – Cfv-I species-specific primer set, VenSF and KwaZulu-Natal 1990 + – Cff + + Cff VenSR, a collection of South African C. fe- KwaZulu-Natal 2007 + – Cff + – Cfv-i Limpopo 2007 + – Cff + – Cfv-i tus field isolates were identified to sub- Unknown + – Cff + – Cfv-i species using the multiplex PCR as well as Uknown + – Cff + – Cfv-i traditional phenotyping techniques. Cur- Unknown + – Cff + – Cfv-i rently, tolerance to 1 % glycine is the only Unknown + – Cff + – Cfv-i internationally accepted phenotypic test KwaZulu-Natal 2007 + – Cff + + Cff prescribed for the differentiation of the KwaZulu-Natal 2007 + – Cff + + Cff 2 C. fetus subspecies11. The reproducibility Mpumalanga 2007 + – Cff + – Cfv-i of this assay is, however, poor and the test Mpumalanga 2007 + – Cff + – Cfv-i can give ambiguous results19. Other bio- Mpumalanga 2007 + – Cff + – Cfv-i chemical tests have been described for the Mpumalanga 2007 + – Cff + – Cfv-i KwaZulu-Natal 2007 + – Cff + + Cff purpose of determining subspecies but KwaZulu-Natal 2007 + – Cff + – Cfv-i contradictory results have been obtained KwaZulu-Natal 2007 + – Cff + – Cfv-i both in this study (results not shown) as 16 Western Cape 2007 + – Cff + – Cfv-i well as by Schulze and colleagues . North West Province 2007 + – Cff + – Cfv-i The C. fetus isolates tested represented a North West Province 2007 + – Cff + – Cfv-i geographically diverse collection which North West Province 2007 + – Cff + – Cfv-i was sourced from different veterinary Mpumalanga 2007 + – Cff + – Cfv-i laboratories across the country. Further- Western Cape 2007 + – Cff + – Cfv-i more, some of the field isolates were Mpumalanga 2007 + – Cff + – Cfv-i recovered from culture collections and North West province 2007 + – Cff + – Cfv-i North West Province 2007 + – Cff + – Cfv-i represent a genetic gene pool that existed KwaZulu-Natal 2007 + – Cff + – Cfv-i up to 2 decades ago. With the exception Mpumalanga 2008 + – Cff + – Cfv-i of 5 of the field isolates, which were iden- KwaZulu-Natal 2008 + – Cff + – Cfv-i tified as Cff, all the isolates were identified Mpumalanga 2008 + – Cff + – Cfv-i as Cfv-i using phenotyping methods. The KwaZulu-Natal 2008 + – Cff + – Cfv-i PCR results, however, indicated that all KwaZulu-Natal 2008 + – Cff + – Cfv-i field isolates were Cff. The lack of corre- KwaZulu-Natal 2008 + – Cff + – Cfv-i spondence obtained was somewhat startling. Previous reports have only inti- *+ = Positive; – = negative. mated minor incongruencies when com- paring these 2 approaches to identifying result where a negative PCR result was possible DNA units in the PCR reaction subspecies5,9 although a recent investiga- obtained but the sample was positive on mix could have resulted in the diluting tion23 alluded to greater problems when a culture. This result is harder to explain. out of the target DNA. In spite of the 2 collection of United Kingdom isolates The presence of trace amounts of PCR anomalous results obtained, it is impor- were tested. These investigators found inhibitors cannot be ruled out. Neither tant to note that neither of the tests under that only 3 isolates, out of a group of 19, can the possibility that the small volume evaluation failed to detect a positive herd which were initially phenotyped as Cfv, (500 µ ) of the aliquot used for for DNA when repeat testing was carried out. gave the same results for subspecies iden- extraction and subsequent dilution of The high degree of agreement between tification using this specific multi-

0038-2809 Jl S.Afr.vet.Ass. (2010) 81(2): 87–92 91 plex PCR assay. After further analysis the the pivotal glycine tolerance test is carried Manual of diagnostic tests and vaccines for researchers proposed that the discrepan- out in triplicate for all isolates being tested terrestrial animals (mammals, birds and bees) (6th edn). Office International des Épi- cies were the result of an unusual Cfv clone in order to ensure the accuracy of the test zooties, Paris: 46–55 circulating within the cattle population in results. 13.OnSLW,Harrington C S 2001 Evaluation the United Kingdom (UK) and that the of numerical analysis of PFGE-DNA multiplex PCR assay was consequently ACKNOWLEDGEMENTS profiles for differentiating Campylobacter unsuitable for the subspecific identifica- We wish to sincerely express our grati- fetus subspecies by comparison with tion of UK isolates. tude to all the colleagues who submitted phenotypic, PCR and 16S rDNA sequenc- samples and made this study possible. ing methods. Journal of Applied Microbiology Unfortunately the initial evaluation of 90: 285–293 the multiplex PCR did not take Cfv-i REFERENCES 14. Pefanis S M, Herr S, Venter C G, Kruger L P, strains into consideration5. In fact, many Queiroga C C, Amaral L 1988 Trichomo- 1. Brooks B W,Devenish J, Lutze-Wallace C L, of the phenotyping and genotyping in- niasis and campylobacteriosis in bulls in Milnes D, Robertson, R H, Berlie-Surujballi the Republic of Transkei. Journal of the South vestigations which have been carried out G 2004 Evaluation of a monoclonal anti- African Veterinary Association 59: 139–140 in this field have seemingly ignored the body-based enzyme-linked immunosor- 15. Schmidt-Dumont A M A 1984 The incidence existence of this particular group of bacte- bent assay for detection of Campylobacter of Campylobacter fetus and Tritrichomonas ria. This is to some extent understandable fetus in bovine preputial washing and vagi- foetus in cattle in South West Africa/Namibia. nal mucus samples. Veterinary Microbiology since the initial description of this group Proceedings of the TenthInternational Congress 103: 77–84 on Animal Reproduction and Artificial Insemi- of isolates was very vague and reference 2. Garcia M M, Eaglesome M D, Rigby C 1983 18 nation, Urbana Champaign, Illinois, June isolates were not readily available .Re- important in veterinary 1984: 10–14 medicine. Veterinary Bulletin 53: 793–811 cently, however, with the availability of 16. Schulze F,Bagon A, Muller W,Hotzel H 2006 3. Gilardi G L 1985 Pseudomonas. In Lenette E high resolution genotyping tools such as Identification of Campylobacter fetus subspe- H, Balows A, Hausler, W J, Shadomy, H J cies by phenotypic differentiation and AFLP and multilocus sequence typing (eds) Manual of clinical microbiology (4th PCR. Journal of Clinical Microbiology 44: (MLST), isolates that have been pheno- edn). American Society of Microbiology, 2019–2024 Washington: 350–372 typed as Cfv biovar intermedius, have 17. Tedesco L F, Errico F, Del Baglivi L P 1977 4. Hum S, Quinn K, Kennedy D 1994 Diagno- been more specifically characterised. These Comparison of three sampling methods for sis of bovine venereal campylobacteriosis techniques have supported phenotyping the diagnosis of genital vibriosis in the bull. by ELISA. Australian Veterinary Journal 71: Australian Veterinary Journal 53: 470–472 results and have in fact provided evi- 140–143 dence for the reclassification of Cfv-i as a 5. Hum S, Quinn K, Brunner J, OnSLW1997 18. Van BergenMAP,Linnane S, van Putten J P, 19 Wagenaar J A 2005 Global detection and separate subspecies . Evaluation of a PCR assay for identification identification of Campylobacter fetus subsp. The apparent absence of classical Cfv and differentiation of Campylobacter fetus subspecies. Australian Veterinary Journal 75: venerealis. Revue Scientifique et Technique, field strains and the widespread distribu- 827–831 Office International des Épizooties 24: tion of Cfv-i in South Africa render the 6. Irons P C, Schutte A P, van der Walt M L, 1017–1026 multiplex PCR unsuitable for local use. In Bishop G C 2004 Genital campylobacte- 19. Van BergenMAP,Simons G, van der Graaf-van Bloois L, van Putten J P M, order to resolve the subspecies identifica- riosis in cattle. In CoetzerJAW,Tustin R C (eds) Infectious diseases of livestock with special Rombout J, Wesley I, Wagenaar J A 2005 tion dilemma, efforts need to be redirected reference to southern Africa Vol. 3 (2nd edn), Amplified fragment length polymorphism towards the sequence analysis of the Cfv-i Oxford University Press, Cape Town: 1459– based identification of genetic markers and genome. The identification of unique 1468 novel PCR assay for differentiation of gene sequences, or lack of specific se- 7. Lander K P 1990 The development of a Campylobacter fetus subspecies. Journal of Medical Microbiology 54: 1217–1234 quences, may allow for the development transport and enrichment medium for Campylobacter fetus. British Veterinary Journal 20. Vargas A C, Costa M M, Vainstein M H, of a primer set which can be used to differ- 146: 327–333 Kreutz L C, Neves J P 2003 Phenotypic and entiate these strains from classical Cfv 8. Monke H J, Love B C, Wittum T E, Monke molecular characterization of bovine and Cff isolates. Until such time, the D R, Byrum B A 2002 Effect of transport Campylobacter fetus strains isolated in Brazil. reliance on bacteriological culture for enrichment medium, transport time, and Veterinary Microbiology 93: 121–132 21. Wagenaar J A, van BergenMAP,Newell isolation, identification and subspecies growth medium on the detection of Campylobacter fetus subsp. venerealis. Journal D G, Grogono-Thomas R, Duim B 2001 differentiation of C. fetus will persist. of Veterinary Diagnostic Investigation 14: Comparative study using amplified frag- Despite the practical drawbacks typi- 35–39 ment length polymorphism fingerprinting, cally associated with bacteriological cul- 9. Muller W,Hotzel H, Schulze F 2003 [Identi- PCR genotyping, and phenotyping to differentiate Campylobacter fetus strains ture the relatively high success rate fication and differentiation of Campylobacter fetus subspecies by PCR] Deutsche Tierärz- isolated from animals. Journal of Clinical achieved during this study has been par- liche Wochenschrift 110: 55–59 (In German) Microbiology 39: 2283–2286 ticularly encouraging. In order to ensure 10. Newell D G, Duim B, van Bergen M A P, 22. Willoughby K 2003 The ABC of PCR. In optimal recovery it is advocated that Grogono-Thomas R, Wagenaar J A 2000 Practice 25: 140–145 preputial scrapes are collected and sub- Speciation, subspeciation and subtyping of 23. Willoughby K, Nettleton P F, Quire M, mitted for laboratory examination in an Campylobacter spp. associated with bovine Maley M A, Foster G, Toszeghy M, Newell infertility and . Cattle Practice 8: D G 2005 A multiplex polymerase chain appropriate transport medium such as 421–425 reaction to detect and differentiate Cam- Weybridge. Samples should be kept cool 11. OIE 2008 Bovine genital campylobac- pylobacter fetus subspecies fetus and Cam- and reach the laboratory within 36 hours teriosis. In Manual of diagnostic tests and pylobacter fetus – species venerealis: use on of collection. Culturing onto at least 2 dif- vaccines for terrestrial animals (mammals, birds UK isolates of C. fetus and other Campylo- ferent culture media (one selective and and bees) (6th edn). Office International des bacter spp. Journal of Applied Microbiology 99: Épizooties, Paris: 661–670 758–766 1 enriched) is advised. As far as identifica- 12. OIE 2008 Validation and quality control of 24. Wilson I G 1997 Inhibition and facilitation of tion to subspecies of isolates is concerned, polymerase chain reaction methods used nucleic acid amplification. Applied and Envi- diagnosticians are advised to ensure that for the diagnosis of infectious diseases. In: ronmental Microbiology 63: 3741–3751

92 0038-2809 Tydskr.S.Afr.vet.Ver. (2010) 81(2): 87–92