OUP user menu

Isolation of Arcobacter species from animal feces

Ellen Van Driessche, Kurt Houf, Jan Van Hoof, Lieven De Zutter, Peter Vandamme
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00840-1 243-248 First published online: 1 December 2003

Abstract

A previously developed Arcobacter isolation protocol for poultry skin and meat was validated for the isolation of Arcobacter from feces of livestock animals. Good repeatability, in-lab reproducibility and sensitivity were achieved and the specificity was improved by additional incorporation of cycloheximide and increase of the novobiocin concentration in the selective supplement. The limit of detection of quantitative and qualitative analysis was 102 and 100 cfu g−1 feces, respectively. From fecal samples collected at slaughterhouse. Arcobacter was isolated from 43.9% of porcine, 39.2% of bovine, 16.1% of ovine and 15.4% of equine samples. All three animal-associated Arcobacter species were isolated and levels up to 103 cfu g−1 feces were determined.

Keywords
  • Arcobacter
  • Isolation
  • Feces

1 Introduction

The genus Arcobacter was described in 1991 [1] to accomodate bacteria formerly known as aerotolerant campylobacters. Arcobacters are Gram-negative non-spore-forming rods with a single polar flagellum. They differ from the closely related campylobacters in their ability to grow aerobically from 15 up to 42°C. Within the genus Arcobacter, four species are presently recognized: Arcobacter nitrofigilis, a nitrogen-fixing plant associated species [2] and the animal and human related species A. butzleri, A. cryaerophilus and A. skirrowii [3,4].

A. butzleri and A. cryaerophilus have been isolated from humans with abdominal illness and septicemia [5,6] and non-human primates with chronic diarrhea [7]. They are associated with reproductive problems in various farm animals [811]. Arcobacters have been detected in river water samples, ground water sources, sewage sludges and water distribution pipe surfaces [1215]. They have been cultured from raw milk from mastitis outbreaks in cattle [16] and from stomachs of pigs with gastric ulcers [17]. Arcobacters have been found in feces from clinically healthy and ill livestock animals [1821] and on food of animal origin such as beef [22], pork [22,23], turkey [24], duck [25] and chicken [26,27]. This contamination of food from animal origin is assumed to occur by fecal contamination during slaughter [28].

Arcobacters were first isolated from bovine and porcine fetuses in 1977, using the semi-solid Leptospira isolation medium Ellinghausen, McCullough, Johnson, and Harris-polysorbate 80 (EMJH-P80) supplemented with 5-fluorouracil and rabbit serum [29]. Different studies report Arcobacter occurrence in feces of livestock animals detected by different isolation methods and molecular techniques [20,21,30,31]. Those techniques, usually developed for the recovery of arcobacters from foods, were never validated for the Arcobacter isolation from feces. Only Kabeya et al. [21] was successful in isolating the three animal-associated Arcobacter species from porcine and bovine fecal samples using the method of de Boer et al. [22].

Recently, an Arcobacter enrichment broth and Arcobacter selective agar were developed for the quantitative and qualitative isolation of all Arcobacter species from poultry skin and meat with maximal suppression of the accompanying flora [32]. The developed selective supplement contained 5-fluorouracil, amphotericin, cefoperazone, novobiocin and trimethoprim. The aims of the present study were to validate the method developed by Houf et al. [32] for the Arcobacter isolation from animal fecal samples and to determine the prevalence and contamination level of Arcobacter in feces from Belgian clinically healthy animals at slaughterhouse.

2 Materials and methods

2.1 Bacterial cultures

Six Arcobacter reference strains (Table 1) were obtained from the BCCM/LMG Bacteria Collection, Ghent University (Ghent, Belgium). The strains were grown on blood agar plates (Mueller Hinton, CM 337, Oxoid, Basingstoke, UK and 50 ml l−1 lysed defibrinated horse blood [E&O Laboratories Ltd., Bonnybridge, Scotland]) and incubated for 48 h at 28°C under microaerobic conditions by evacuating 80% of the normal atmosphere and introducing a gas mixture of 8% CO2, 8% H2 and 84% N2 into the jar. This incubation atmosphere was used for all further Arcobacter isolations. From each strain, a bacterial suspension was prepared in 10 ml of sterile Arcobacter enrichment broth (containing 24 g l−1Arcobacter broth [CM 965-Oxoid, Basingstoke, UK], and the previously developed selective supplement (100 mg l−1 5-fluorouracil, 10 mg l−1 amphotericine B, 16 mg l−1 cefoperazone, 32 mg l−1 novobiocine and 64 mg l−1 trimethoprim) [32] with an optical density of 0.9 (measured at 660 nm) which corresponded with a bacterial concentration of approximately 109 cfu ml−1. Serial 10-fold dilutions in sterile Arcobacter enrichment broth were prepared to obtain 104 and 103 cfu ml−1 for use in the validation of the direct isolation method and 103 to 100 cfu ml−1 for use in the validation of the enrichment medium. Counts of the expected 103 and 104 cfu ml−1 bacterial suspensions were performed in 10-fold by inoculating 100 µl by the spiral plating method onto Arcobacter selective agar plates [32] (containing 24 g l−1Arcobacter broth, 12 g l−1 Agar Technical No.3 [L13-Oxoid] and the selective supplement described above). Following incubation at 28°C for 24–72 h, the average and the standard deviation of the logarithm of the colony counts were calculated.

View this table:
Table 1

Arcobacter collection strains used in this study

OrganismsNo.Strain numberSourceCountry
A. butzleri1LMG 10240Horse, fecesCanada
2LMG 10828Human, fecesUnited states
A. cryaerophilus3LMG 9867Equine fetus, spleenNorthern Ireland
4LMG 9908Porcine fetusNorthern Ireland
A. skirrowii5LMG 9912Bovine fetus, thoracalNorthern Ireland
6LMG 14985Bull, preputial washUnited states
  • T: Type strain

2.2 Validation of the isolation procedure

Only fecal samples, negative for Arcobacter as determined by the validated method, were used to determine the sensitivity of the direct and enrichment methods. Fecal samples were taken rectally of livestock animals, including cows, sheep, pigs and horses. For the examination of the natural Arcobacter contamination, one g feces per sample was transferred into a sterile stomacher bag. Next, nine ml Arcobacter enrichment broth supplemented with 50 ml l−1 lysed defibrinated horse blood was added. The mixtures were homogenized for 1 min with a stomacher blender at normal speed. After homogenization, 100 µl of each homogenate was inoculated onto two Arcobacter selective agar plates by spiral plating. Plates were incubated for 24–72 h at 28°C. The remaining homogenates were incubated for 48 h at 28°C. Following incubation, 50 µl of the homogenates was streaked onto Arcobacter selective agar plates. Plates were incubated for 24–72 h at 28°C and checked every 24 h for bacterial growth. All colonies were counted and subcultured onto blood agar plates. For identification at species level, the previously developed m-PCR [27] using the primers ARCO, BUTZ, CRY1, CRY2 and SKIR was used. Samples tested negative in the m-PCR, were retested by an Arcobacter-genus-specific PCR assay using the primers Arco I and Arco II [33].

For the direct isolation procedure, repeatability, in-lab reproducibility, sensitivity and specificity were evaluated. For each of the six Arcobacter reference strain tested, one g feces was taken in duplicate from each of the different animal species and transferred into sterile stomacher bags. Each sample was spiked respectively with one ml of the Arcobacter reference strain dilutions expected to be 104 and 103 cfu ml−1. Then, eight ml Arcobacter enrichment broth was added. After homogenization, 100 µl was transferred onto Arcobacter selective agar plates by means of the spiral plater. A total of 20 samples were taken for each suspension prepared for each of the six. The remaining inoculated homogenates were stored for 5 days at 4°C. After chill-storage, 100 µl was transferred 20 times onto Arcobacter selective agar plates. All agar plates were incubated for up to 5 days at 28°C and checked every 24 h for the presence of bacterial growth. The colonies were counted and identified by m-PCR. The average and standard deviation of the logarithm of the 20 colony counts were calculated.

For the enrichment method, the detection limit was determined. Therefore, five times 1 g feces was taken and transferred into sterile stomacher bags. Eight ml Arcobacter enrichment broth was added. Four samples were respectively spiked with 1 ml of the bacterial suspensions of the Arcobacter strain tested, ranging from 103 to 100 cfu ml−1. The fifth sample was spiked with 1 ml sterile Arcobacter enrichment broth and acted as blank. After homogenization, all samples were incubated at 28°C. After 24 h, 48 h and 72 h of incubation, 50 µl of the homogenates was streaked onto Arcobacter selective agar plates. The plates were incubated and checked every 24 h for bacterial growth up to 3 days. This was repeated for each animal species.

Results were analyzed using univariate analysis of variance (P<0.05).

2.3 Prevalence of Arcobacter in feces of livestock animals

To test the effectiveness of the method in vivo, fecal samples from livestock animals were collected during slaughter from October 2002 to May 2003. Isolation, enumeration and identification of Arcobacter from feces was performed as described above, except for the selective supplement, that was adapted according to observations during validation.

3 Results

3.1 Validation of the isolation procedure

Arcobacter was not detected in fecal samples used in the validation studies. For the direct isolation method, the results of the average colony counts and standard deviations of the spiked fecal homogenates are shown in Table 2. For A. butzleri, after 24 h of incubation, added at 103 and 104 cfu ml−1, colony counts ranged from 102 to 103 and 103 to 104 cfu ml−1, respectively. For A. cryaerophilus, an incubation time of 48 h was required to obtain the same results. For A. skirrowii however, these colony counts were obtained only after 72 h of incubation.

View this table:
Table 2

Average colony counts and standard deviations (SD) of the fecal homogenates spiked with the bacterial suspensions 104 (dilution −5) and 103 (dilution −6) cfu ml−1

Strain no. in ExpColony count
Original bacterial dilution −5Spiked horse feces with dilution −5Spiked sheep feces with dilution −5Spiked cow feces with dilution −5Spiked pig feces with dilution −5Original bacterial dilution −6Spiked horse feces with dilution −6Spiked sheep feces with dilution −6Spiked cow feces with dilution −6Spiked pig feces with dilution −6
AverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSD
14.540.053.620.043.650.053.640.043.680.043.570.042.670.052.690.062.720.062.730.14
24.660.063.800.043.800.083.810.043.730.053.750.062.830.062.780.062.830.062.720.05
34.640.053.750.053.860.103.740.053.720.053.740.022.850.042.810.072.820.042.820.03
44.540.053.660.063.670.053.710.063.750.083.470.072.670.062.720.072.690.072.830.04
53.990.103.820.063.820.053.760.07NDND2.990.132.720.062.730.072.770.06NDND
64.790.483.560.063.910.063.950.073.660.073.410.612.720.062.890.072.920.072.760.05
  • ND=Not done.

  • Strain designation as in Table 1.

  • Exp=Experiment.

  • Average and SD of 20 replicates prepared from a single suspension.

To determine the in-lab reproducibility, the spiked homogenates of strains 1, 4 and 6 were stored for 5 days at 4°C. The average colony counts and standard deviations are presented in Table 3. Storage of these homogenates resulted in colony counts that had decreased significantly (P<0.05) in comparison with the colony counts of the original spiked homogenates.

View this table:
Table 3

In-lab reproducibility: average colony counts and standard deviations (SD) of spiked homogenates after storage

Strain no.Colony count
Spiked horse feces with dilution −5Spiked sheep feces with dilution −5Spiked cow feces with dilution −5Spiked pig feces with dilution −5Spiked horse feces with dilution −6Spiked sheep feces with dilution −6Spiked cow feces with dilution −6Spiked pig feces with dilution −6
AverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSDAverageSD
13.340.033.230.033.290.032.340.082.230.091.930.142.250.071.410.16
43.160.053.210.043.280.023.350.052.170.092.220.092.280.092.460.08
63.050.053.360.033.420.063.570.042.090.142.640.092.580.072.570.10
  • Strain designation as in Table 1.

For the enrichment method, the detection limit was set on 100 cfu g−1. At this concentration, A. butzleri could be isolated after 24 h of enrichment, followed by 24 h of plate incubation. A. cryaerophilus and A. skirrowii required an enrichment of 48 h followed by 48 h of plate incubation. An enrichment of 72 h was not required for any of the Arcobacter species tested.

During validation two phenomena were observed. First, after incubation for more than 72 h, small whitish colonies could be observed in samples of all animal species tested. These colonies tested negative in both Arcobacter species and genus PCR. An increase of the novobiocin concentration to 64 mg l−1 improved the specificity of the method. Secondly, after 48 h of incubation fungal growth was observed on a number of Arcobacter selective agar plates, especially plates inoculated with porcine or bovine feces. This fungal growth could be delayed to more than three days by the additional incorporation of 100 mg l−1 of cycloheximide into the selective supplement.

3.2 Prevalence of Arcobacter in feces of livestock animals

The prevalence of Arcobacter species in feces from Belgian livestock animals was determined, using the validated method for direct isolation and enrichment.

For cattle, using direct isolation, five of 51 (10%) bovine fecal samples tested positive for the presence of Arcobacter. A. butzleri was isolated from two samples and A. cryaerophilus from one. Two animals shed two Arcobacter species simultaneously: A. butzleri and A. skirrowii were isolated from one sample and A. cryaerophilus and A. skirrowii were found in another bovine fecal sample. After enrichment, the overall number of positive samples increased to 20 (39%), but co-colonization was not longer detected. A. butzleri was isolated from 13 samples, A. cryaerophilus from five and A. skirrowii from two.

For pigs, in 23 of the 82 (28%) fecal samples, arcobacters were isolated by direct isolation. A. butzleri was identified in 18 samples and A. cryaerophilus in three. Two pigs shed A. butzleri and A. cryaerophilus simultaneously. After enrichment, arcobacters were isolated from 36 (44%) porcine fecal samples. A. butzleri was present in 29 and A. cryaerophilus in seven samples. After enrichment, co-colonization was not detected.

Of the 62 sheep samples examined, three (5%) were positive after direct isolation, and 10 (16%) after enrichment. A. cryaerophilus was the only species found by direct isolation. After enrichment, A. butzleri was isolated from two samples, A. cryaerophilus from eight.

After enrichment, A. butzleri was isolated from two out of 13 (15%) equine samples examined. No arcobacters were isolated by direct isolation.

4 Discussion

In the present study, a direct isolation and enrichment method for Arcobacter species from poultry skin and meat [32] was validated for the isolation of arcobacters from animal fecal samples. The method allowed the isolation of A. butzleri, A. cryaerophilus and A. skirrowii, both during the validation of the method and the examination of slaughterhouse fecal samples. With this method, the Arcobacter bacterial load in feces could be quantified, with a good suppression of the accompanying fecal flora.

Good repeatability, in-lab reproducibility and sensitivity were achieved, and moreover, a small adaptation of the selective supplement resulted in an improvement of the specificity of the method. We propose the name Arcobacter selective isolation broth (ASIB) and Arcobacter selective isolation agar (ASIA) for the Arcobacter enrichment broth and Arcobacter selective agar containing the modified selective supplement for isolation of Arcobacter from feces.

Using the validated enrichment method, 43.9% of porcine, 39.2% of bovine, 16.1% of ovine and 15.4% of equine samples were positive for the presence of Arcobacter at slaughterhouse level. In the present study, 62 sheep were examined using the validated method. To our knowledge, this is the first report of Arcobacter isolation from feces of healthy sheep. The isolation of arcobacters from animal fecal samples has been reported using other methods, but the prevalence was lower than in the present study. Differences in Arcobacter prevalence in feces from livestock animals reported in the literature can be ascribed to several causes. First of all, the isolation medium has a considerable impact on the number of Arcobacter-positive samples. In a study from Golla et al. [20], Arcobacter was detected in 4.5% of the 200 fecal swabs from cattle using the method by Johnson and Murano [34] and in 2.5% of the samples using the Collins method [23]. Using the method of de Boer et al. [22], arcobacters were isolated from 10% porcine and 3.6% bovine fecal samples collected at the abattoir [21]. However, A. cryaerophilus and A. skirrowii are susceptible to lower piperacillin concentrations than that incorporated in Arcobacter selective broth [35], which can lead to a lower detection rate of those species. Secondly, as previously described for Campylobacter [36], the sampling method, swab versus rectally collected fecal samples, may influence the number of positive samples. Also the season may have an influence on the Arcobacter prevalence [30]. Although as yet there are no reports on the seasonal variation of arcobacters in feces, there are studies that mention a higher isolation rate in spring and summer on poultry meat [24] and in water [14]. Moreover, the animals’ age should be taken into account when results are compared. Hume et al. [31] found only four of 200 piglets positive for the presence of Arcobacter, whereas 23 of 79 sows examined shed arcobacters. Finally, the origin of the samples, such as country, farm or slaughterhouse, may affect the results.

In conclusion, the validated and optimized isolation protocol in the present study is a reliable method for the isolation of Arcobacter from feces of livestock animals and can be recommended for further epidemiological research. The Arcobacter species associated with human and animal illness are effectively isolated with a good inhibition of the accompanying fecal flora. The Arcobacter prevalences in feces from livestock animals obtained with this protocol are higher than those reported in other studies. The use of the direct plating provides an opportunity to obtain information on the bacterial load in fecal samples and excludes the possibility of (growth) selection occurring during enrichment.

Acknowledgements

This work was supported by the Research Fund of the Ghent University, Belgium, code nr. BOF 2002/DRMAN/063. We thank the BCCM/LMG Bacteria collection, Ghent University (Ghent, Belgium) for providing the Arcobacter strains used in this study. The technical assistance of Bieke Verbeke and Jasmien Taildeman was greatly appreciated. We thank Dr. J. Van Hende for the statistical analysis.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
View Abstract