- Email: [email protected]
Monensin use against Neospora caninum challenge in dairy cattle
Veterinary Parasitology 175 (2011) 372–376
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Short communication
Monensin use against Neospora caninum challenge in dairy cattle J.A. VanLeeuwen a,b,∗ , S. Greenwood c , F. Clark c , A. Acorn c , F. Markham c , J. McCarron a,b , R. O’Handley d a
Centre for Veterinary Epidemiological Research, University of Prince Edward Island, Charlottetown, PEI, Canada C1A 4P3 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, 550, University Ave, Charlottetown, PEI, Canada C1A 4P3 c Department of Pathology/Microbiology, Atlantic Veterinary College, University of Prince Edward Island, 550, University Ave, Charlottetown, PEI, Canada C1A 4P3 d School of Animal and Veterinary Sciences, University of Adelaide, Roseworthy Campus, South Australia 5371, Australia b
a r t i c l e
i n f o
Article history: Received 26 July 2010 Received in revised form 30 September 2010 Accepted 8 October 2010 Keywords: Neospora caninum Randomized controlled challenge trial Monensin Cattle
a b s t r a c t Using a randomized controlled trial design, a randomly allocated intervention group of 15 cows received a slow-release bolus that delivered 100 days of monensin. The negative control group of 15 cows received a placebo bolus that was identical to the monensin bolus, except without the monensin. Two weeks after bolus administration, all cows were challenged with a 2 ml subcutaneous injection of a live tachyzoite suspension. Whole blood and serum samples were collected from each cow every week for the first month postchallenge, and then every 2 weeks for the next 2 months. The extracted DNA from whole blood was tested for the Nc-5 gene fragment of Neospora caninum using a quantitative real time polymerase chain reaction. Serum was tested for antibodies to N. caninum using the IDEXX ELISA. Cows treated with monensin boluses had a significantly lower humoral immune response than cows treated with placebo boluses at one time point post-challenge (week 4 post-challenge). However, when adjusting for repeated measures within cows, the P value for this humoral difference was 0.098. No DNA for N. caninum was detected in either group, likely due to study design features. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Neospora caninum is a costly intracellular protozoan that is a major cause of epidemic and endemic abortions in cattle world-wide. Cattle remain infected for life and can undergo repeat abortions. The abortions can lead to reduced milk production (Tiwari et al., 2007), lower reproductive efficiency (VanLeeuwen et al., 2010b), and premature culling (Tiwari et al., 2005). Total direct production costs associated with neosporosis in Canadian dairy cattle were determined to be $2305/year/infected herd of
∗ Corresponding author at: 550, University Ave, Charlottetown, PEI, Canada C1A 4P3. Tel.: +1 902 566 0457; fax: +1 902 620 5053. E-mail address: [email protected] (J.A. VanLeeuwen). 0304-4017/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.vetpar.2010.10.016
50 cows (Chi et al., 2002). Similar cost estimates have been determined elsewhere (Trees et al., 1999; Bartels et al., 2006; Dubey et al., 2007), demonstrating that N. caninum is an economically important disease of the dairy industry. Cattle (intermediate hosts) become infected by transplacental (vertical) transmission of the fast-growing form (tachyzoite) of the protozoan from dam to calf, or by ingestion (horizontal) of oocysts from exposure to feces from wild or domestic dogs (definitive hosts) that are infected and actively shedding the parasite (McAllister et al., 1998; Gondim et al., 2004). After cattle become infected, either vertically or horizontally, their immune system tries to eliminate the infection, at which time, the circulating protozoa transform into hardy slowly dividing forms (bradyzoites) and encyst in nerve cells where they
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
can remain dormant for months or years (Dubey et al., 2007). To date, there has been no effective treatment demonstrated against N. caninum. However, we conducted a risk factor analysis of 240 dairy farms across Canada and showed that cows receiving monensin (a polyether ionophore produced by Elanco, based in Greenfield, Indiana) as a feed additive were 1.5 times less likely (P < 0.05) to be infected with N. caninum than cows not receiving monensin (VanLeeuwen et al., 2010a). A similar trend was seen in supplementary analyses (Duffield T., Ontario Veterinary College, Guelph, ON, Canada, personal communication) of abortions due to N. caninum in herds in Ontario (Hobson et al., 2005). In laboratory experiments, monensin was shown to reduce the population of N. caninum tachyzoites in cell culture by at least 95% at concentrations as low as 0.001 ng/ml (Lindsay et al., 1994). Serum concentrations in cows given monensin boluses achieve these levels (Bagg R., Elanco Animal Health Ltd., Guelph, ON, Canada, personal communication). No studies have looked at the effect of monensin on the hardy bradyzoites in the lab or in the field. Clinical trials are needed to determine the effect of monensin on N. caninum in cows. The study objective was to determine if monensin had an effect on N. caninum tachyzoite infection in cattle. The study was undertaken after approval of the Animal Care Committee at the University of Prince Edward Island. 2. Materials and methods The study group started out with all of the 30 nonmilking, non-pregnant cows in the teaching barn at the Atlantic Veterinary College (AVC) because commercial dairy farmers would not want their animals to be intentionally infected with N. caninum. Because the AVC teaching cows were purchased cull cows used for teaching (and research), limited information on the cows was available. Breed and weight were available and recorded. There were 6 Jersey cows, with the remainder being Holstein. The average weight was 702 kg, with a range of 532–820 kg. The cows were all non-pregnant and not bred during the study, so N. caninum infection could not cause abortions. The study was a randomized controlled trial with two groups of cows. On January 2nd, 2009, half of the 30 cows were randomly allocated to the intervention group (using computer-generated random numbers), and they received a slow-release bolus that delivered 100 days of monensin. The remaining cows in the negative control group received a placebo bolus that was identical to the monensin bolus, except without the monensin. The bolus serial numbers were recorded but AVC barn staff reported no regurgitated boluses. Just prior to the random allocation and administration of boluses, serum samples from all cows were obtained and tested for antibodies against N. caninum using the Herdchek anti-Neospora IDEXX ELISA (Westbrook, ME, USA), according to the manufacturer’s instructions. Only one cow was found to be positive for N. caninum on serology, and it was in the placebo group so it was removed from the final dataset, along with one randomly selected cow from the intervention group.
373
Three weeks after bolus administration, all cows were challenged with a 2 ml subcutaneous injection of a live tachyzoite suspension (NC-1, 2.5 × 106 tachyzoites/ml within 2 h of harvest), as described (Liddell et al., 1999) and successfully utilized in a previous vaccination efficacy trial using N. caninum challenge (O’Handley et al., 2003). Over the next 3 months, whole blood and serum samples were collected from each cow every week for the first month post-challenge, and then every 2 weeks. Whole blood samples (1 ml) underwent DNA extraction (Qiagen DNeasy Blood and Tissue kit according to the manufacturer’s instructions) within 8 h of sampling, and the DNA samples were frozen at −20 ◦ C until all samples were collected. The extracted samples were split into 3 aliquots and tested for the presence and estimate of numbers of tachyzoites (outcome variables of interest), using a quantitative real-time polymerase chain reaction (qPCR) for the Nc-5 gene fragment of N. caninum that was validated and optimized at the AVC, based on a recently published protocol (Okeoma et al., 2005). Standard curves were constructed for both N. caninum (Nc-5) and bovine (28S rRNA) genes to ensure comparable PCR efficiency. The PCR reactions for Nc5 and 28S rRNA genes were run in triplicate and duplicate, respectively, on all samples. Positive controls for N. caninum were detected at Ct 19–20, and samples were only considered positive if a Ct value (20–40) was detected in at least two of the triplicate blood samples. For serum testing, 1 ml of serum from each cow was frozen at −20 ◦ C until all samples were collected, and then tested for antibody levels to N. caninum using the Herdcheck Anti-Neospora IDEXX ELISA (Westbrook, ME, USA). Sample-to-positive ratios (S/P ratio) on the ELISA were recorded. A cow was classified as infected with N. caninum if the sample-to-positive ratio (S/P ratio) on the ELISA was ≥0.5, as recommended by the manufacturer of the test kit. At this cut-off, the test has a sensitivity of 93% and specificity of 94% (Wapenaar et al., 2007). All researchers, laboratory staff, and animal care-givers were blinded to the group status of the cows, except one administrative technician. The continuous outcome variables of interest (S/P ratios for ELISA results, and number of tachyzoites for PCR results) were tested for normality using the sktest command (i.e. skewness and kurtosis) in STATA Version 11 (College Station, TX, USA) and if the variables were not normally distributed, they were transformed using a natural logarithm transformation (ln). Because S/P ratios ranged from 0.01 to 1.78, a value of 1 was added to every S/P ratio, prior to transformation, so that the ln S/P ratio would not be a negative number. The means and standard deviations of the continuous outcome variables (including ln transformed outcomes, where applicable – which were converted back to the regular scale for presentation) were calculated, by treatment group and time point. Significant differences in the outcome variables were compared (P < 0.05) between the two groups at each time point using an unpaired t-test. Using the dichotomous outcome variables of interest (seropositivity and PCR positivity), chi-square tests were used to determine differences in the proportion of positive cows between the two groups at each time point.
374
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
Table 1 Post-challenge serological results of N. caninum antibody testing for cows receiving either a monensin (n = 14) or placebo (n = 13) bolus, by week of study. Placebo group Range S/P ratio Week 0 Week 1 Week 2 Week 3 Week 4 Week 6 Week 8 Week 10
0.03–0.18 0.01–0.19 0.08–1.70 0.17–1.78 0.12–1.51 0.08–1.29 0.05–0.97 0.05–0.98
Monensin group Average ln S/P ratioa
S.D. ln S/P ratioa
% Positive
0.086 0.065 0.672 0.810† 0.834* (0.05) 0.703 0.564 0.490
0.039 0.043 0.341 0.335 0.288 0.239 0.221 0.205
0% 0% 62% 77%† 77%† 69%† 69%† 54%
Range S/P ratio 0.02–0.24 0.01–0.14 0.09–1.76 0.26–1.38 0.28–1.19 0.23–1.22 0.12–0.85 0.05–0.82
Average ln S/P ratioa
S.D. ln S/P ratioa
% Positive
0.095 0.058 0.485 0.563 0.554 0.541 0.407 0.344
0.053 0.038 0.291 0.205 0.169 0.212 0.208 0.195
0% 0% 43% 43% 43% 36% 36% 29%
a ln S/P ratio = Sample-to-positive ratios were transformed using a natural logarithm due to non-normal distribution, then averages and standard deviations were determined, and then these calculations were converted back to the regular scale. * P < 0.05 with specific P-value in parentheses. † P < 0.10.
In order to control for clustering of repeated test measures within cows, XTGEE models were used to determine significant associations for the continuous outcome variables, using an AR correlation structure, in SAS Version 10 (Cary, NC, USA). For dichotomous outcomes, XTGEE was again used, using the binomial family, logit link, and AR correlation structure. The following sample size calculation was performed to determine the number of cows needed. Using an alpha (type 1 error) of 5% (level of confidence of 95%) and a beta (type 2 error) of 20% (power = 80%), the sample size needed in each treatment group was 16 in order to detect a 35% difference in presence of tachyzoites or positive titres between the two groups. Therefore, our sample size was close to this desired number. 3. Results One cow in the placebo group was lost to follow-up for reasons unrelated to the research project (became very fractious and therefore unsuitable for retention). Therefore, the final dataset included 13 cows in the placebo group and 14 cows in the monensin group. The final sampling of blood was conducted 10 weeks after challenge. With the random allocation, there were 4 and 2 Jerseys in the placebo and monensin groups, respectively. The overall average weight of the cows was 702 kg, with the placebo and monensin groups weighing, on average, 699 and 706 kg, respectively. There were no statistically significant differences in weight or breed between the groups (unpaired t-test and Chi-square test, respectively), suggesting that the random allocation process succeeded in equally balancing the effects of confounding variables, such as breed and weight, among the two groups. The S/P ratios were not normally distributed and therefore were ln transformed. The cows in the placebo group had significantly higher ln S/P ratios than the cows in the monensin group at week 4, with similar trends (P values between 0.05 and 0.15) from week 2 through to week 10 (Table 1). There were no significant differences between the proportions of positive serological results for the placebo (10 of 13 were positive for weeks 3 and 4) and monensin (6 of 14 were positive for weeks 3 and 4) groups, although
these differences approached statistical significance (P values between 0.05 and 0.1) for weeks 3 through 8. None of the cows in either group had positive PCR results at any of the time points of the study. All positive controls for N. caninum were detected in each qPCR assay, and the bovine 28S rRNA gene was detected in all samples, confirming that the PCR assay was functioning properly. In the final XTGEE model of ln S/P ratio, controlling for clustering of measurements at various time points within cows, breed, weight and treatment group did not remain significant (P > 0.05). However, there was a trend (P = 0.098) toward cows in the placebo group having a higher average ln S/P ratio compared to cows in the monensin group (results not shown due to not statistically significant at P < 0.05). Similarly, the XTGEE model of seropositivity determined that there was only a trend (P = 0.08) toward more cows with a positive N. caninum titre in the placebo group than in the monensin group (results not shown again due to not statistically significant at P < 0.05). No cows had any adverse events related to the administration of the boluses or injections of the N. caninum challenge. 4. Discussion This study demonstrated a significant effect of monensin on reducing the humoral response to challenge infection of cows with N. caninum at one time point. However, the statistical analyses that controlled for repeated measures within cows did not show a statistically significant difference in N. caninum titres between the placebo and monensin groups over the length of the study, for various reasons. First, substantially elevated titres only occurred for part of the study timeframe. It was expected that weeks 1 and 2 would be less likely to show significant differences between groups because of the time required for the cows to mount a detectable immune response post-challenge. The average S/P ratios peaked in week 4, waning thereafter, with some cows even converting from positive to negative titres. Antibody titres were seen to peak at approximately a month post-vaccination and postchallenge in another clinical trial on N. caninum (O’Handley et al., 2003), with conversion from positive to negative titres being reported elsewhere as well (Sager et al., 2001).
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
Second, the limited sample size made it more difficult to find a significant difference in S/P ratios and seropositivity, although we did use all cows available at our institution. Alternatively, the lack of significance could have been due to random variation and no real difference between groups. More research is needed to confirm or refute our results. Clearly, the monensin group did end up having some cows that appeared to become infected with N. caninum, based on substantial rises in titres (2 monensin-treated cows in particular rose from 0.03 and 0.08 to 1.12 and 1.68, respectively), suggesting that, if monensin does have some effect on N. caninum infection, other factors must be involved in determining whether monensin boluses can potentially prevent infection. Perhaps one factor could be the level of monensin in each animal, which was not measured in this study. The boluses are manufactured to release 335 mg/d, but the levels circulating in the body of any given cow would depend on the consistency of release of monensin from the bolus, the weight of the cow, and the rate of metabolism of the monensin in each cow. The 2 monensin cows with high S/P ratios, mentioned above, had weights of 790 and 780 kg, respectively, higher than average in the group. Perhaps one monensin bolus does not lead to a sufficient level of monensin circulating in the body of a heavy cow to reduce the likelihood of N. caninum infection. Further research could help to elucidate these factors. The reasons for none of the cows in either group having positive PCR results at any of the time points of the study are likely related to the methodology and dynamics of the parasite after experimental challenge. Although PCR methods generally have the benefit of being sensitive and specific, they still have limits of detection. The current test was able to detect from −1 to 10−4 (77.6–0.077 pg) copies of N. caninum tachyzoite per mg of total DNA. Therefore, if the number of circulating tachyzoites in infected cows was below this limit, a false negative result would be obtained, leading to reduced test sensitivity. Several scenarios could arise in which this could happen: (1) if the inoculation dose was too low and the number of tachyzoites/ml was too dilute for detection (5 × 106 tachyzoites/cow was chosen from a previous study that was able to detect parasitemia by PCR in sheep (O’Handley et al., 2003)); (2) if the tachyzoites were viable at injection but subsequently became non-viable and therefore were unable to reach general circulation from the subcutaneous injection site (tachyzoites were injected within 2 h of preparation and we have been able to maintain viability at 4 ◦ C for between 5 and 6 h); (3) if viable tachyzoites were cleared rapidly from circulation; and (4) if parasitemia was transient and did not coincide with sampling (Macaldowie et al., 2004). The sporadic nature of N. caninum DNA detection by PCR has been observed previously with no clear consensus. Numerous reasons surrounding cattle immunocompetence, sex, and pregnancy status have been considered. However, temporal studies in mice have revealed that tachyzoites were quickly cleared from circulation during the first week post-inoculation, and only occasionally reached PCR detection limits over time (CollantesFernandez et al., 2006). Before recommendations can be made on monensin use for N. caninum, further research is needed in other larger
375
populations of cattle, preferably pregnant, with a modified sampling timeframe for detecting tachyzoite DNA. Furthermore, trials on monensin’s effectiveness on more natural modes of N. caninum transmission (e.g. bradyzoite recrudescence leading to trans-placental transmission, or oocyst ingestion followed by sporozoite and then tachyzoite infection) would also be needed before recommendations could be made. Conflict of interest statement The authors have no financial or personal relationships with other people or organizations that could inappropriately influence (bias) their work. Acknowledgements The authors acknowledge the technical support of T. Andrews, R. Milton, and L. Dalziel (Department of Health Management, Atlantic Veterinary College). This work was carried out with the financial support of the Atlantic Veterinary College and Elanco Animal Health. The funding sources agreed with the proposed study design but did not influence, in any way, the study design, collection, analysis or interpretation of the data, or the writing of the manuscript and its submission for publication. References Bartels, C.J., van, S.G., Veldhuisen, J.P., van den Borne, B.H., Wouda, W., Dijkstra, T., 2006. Effect of Neospora caninum-serostatus on culling, reproductive performance and milk production in Dutch dairy herds with and without a history of Neospora caninum-associated abortion epidemics. Prev. Vet. Med. 77, 186–198. Chi, J., VanLeeuwen, J.A., Weersink, A., Keefe, G.P., 2002. Direct production losses and treatment costs from bovine viral diarrhoea virus, bovine leukosis virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum. Prev. Vet. Med. 55, 137–153. Collantes-Fernandez, E., Lopez-Perez, I., Alvarez-Garcia, G., Ortega-Mora, L.M., 2006. Temporal distribution and parasite load kinetics in blood and tissues during Neospora caninum infection in mice. Infect. Immun. 74, 2491–2494. Dubey, J.P., Schares, G., Ortega-Mora, L.M., 2007. Epidemiology and control of neosporosis and Neospora caninum. Clin. Microbiol. Rev. 20, 323–367. Gondim, L.F., McAllister, M.M., Pitt, W.C., Zemlicka, D.E., 2004. Coyotes (Canis latrans) are definitive hosts of Neospora caninum. Int. J. Parasitol. 34, 159–161. Hobson, J.C., Duffield, T.F., Kelton, D., Lissemore, K., Hietala, S.K., Leslie, K.E., McEwen, B., Peregrine, A.S., 2005. Risk factors associated with Neospora caninum abortion in Ontario Holstein dairy herds. Vet. Parasitol. 127, 177–188. Liddell, S., Jenkins, M.C., Dubey, J.P., 1999. A competitive PCR assay for quantitative detection of Neospora caninum. Int. J. Parasitol. 29, 1583–1587. Lindsay, D.S., Rippey, N.S., Cole, R.A., Parsons, L.C., Dubey, J.P., Tidwell, R.R., Blagburn, B.L., 1994. Examination of the activities of 43 chemotherapeutic agents against Neospora caninum tachyzoites in cultured cells. Am. J. Vet. Res. 55, 976–981. Macaldowie, C., Maley, S.W., Wright, S., Bartley, P., Esteban-Redondo, I., Buxton, D., Innes, E.A., 2004. Placental pathology associated with fetal death in cattle inoculated with Neospora caninum by two different routes in early pregnancy. J. Comp. Pathol. 131, 142–156. McAllister, M.M., Dubey, J.P., Lindsay, D.S., Jolley, W.R., Wills, R.A., McGuire, A.M., 1998. Dogs are definitive hosts of Neospora caninum. Int. J. Parasitol. 28, 1473–1478. O’Handley, R.M., Morgan, S.A., Parker, C., Jenkins, M.C., Dubey, J.P., 2003. Vaccination of ewes for prevention of vertical transmission of Neospora caninum. Am. J. Vet. Res. 64, 449–452. Okeoma, C.M., Stowell, K.M., Williamson, N.B., Pomroy, W.E., 2005. Neospora caninum: quantification of DNA in the blood of naturally
376
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
infected aborted and pregnant cows using real-time PCR. Exp. Parasitol. 110, 48–55. Sager, H., Fischer, I., Furrer, K., Strasser, M., Waldvogel, A., Boerlin, P., Audige, L., Gottstein, B., 2001. A Swiss case–control study to assess Neospora caninum-associated bovine abortions by PCR, histopathology and serology. Vet. Parasitol. 102, 1–15. Tiwari, A., VanLeeuwen, J.A., Dohoo, I.R., Keefe, G.P., Haddad, J.P., Tremblay, R., Scott, H.M., Whiting, T., 2007. Production effects of pathogens causing bovine leukosis, bovine viral diarrhea, paratuberculosis, and neosporosis. J. Dairy Sci. 90, 659–669. Tiwari, A., VanLeeuwen, J.A., Dohoo, I.R., Stryhn, H., Keefe, G.P., Haddad, J.P., 2005. Effects of seropositivity for bovine leukemia virus, bovine viral diarrhoea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum on culling in dairy cattle in four Canadian provinces. Vet. Microbiol. 109, 147–158. Trees, A.J., Davison, H.C., Innes, E.A., Wastling, J.M., Tenter, A.M., Shirley, M.W., 1999. Towards evaluating the economic impact of bovine
neosporosis. Special issue: COST820: Vaccines against animal coccidioses. Papers from the Annual Workshop, Toledo, Spain. Int. J. Parasitol. 29, 1195–1200. VanLeeuwen, J.A., Haddad, J.P., Dohoo, I.R., Keefe, G.P., Tiwari, A., Scott, H.M., 2010a. Risk factors associated with Neospora caninum seropositivity in randomly sampled Canadian dairy cows and herds. Prev. Vet. Med. 93, 129–138. VanLeeuwen, J.A., Haddad, J.P., Dohoo, I.R., Tiwari, A., Keefe, G.P., Stryhn, H., Tremblay, R., 2010b. Associations between reproductive performance and seropositivity for bovine leukemia virus, bovine viral-diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in Canadian dairy cows. Prev. Vet. Med. 94, 54–64. Wapenaar, W., Barkema, H.W., VanLeeuwen, J.A., McClure, J.T., O’Handley, R.M., Kwok, O.C., Thulliez, P., Dubey, J.P., Jenkins, M.C., 2007. Comparison of serological methods for the diagnosis of Neospora caninum infection in cattle. Vet. Parasitol. 143, 166–173.
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Short communication
Monensin use against Neospora caninum challenge in dairy cattle J.A. VanLeeuwen a,b,∗ , S. Greenwood c , F. Clark c , A. Acorn c , F. Markham c , J. McCarron a,b , R. O’Handley d a
Centre for Veterinary Epidemiological Research, University of Prince Edward Island, Charlottetown, PEI, Canada C1A 4P3 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, 550, University Ave, Charlottetown, PEI, Canada C1A 4P3 c Department of Pathology/Microbiology, Atlantic Veterinary College, University of Prince Edward Island, 550, University Ave, Charlottetown, PEI, Canada C1A 4P3 d School of Animal and Veterinary Sciences, University of Adelaide, Roseworthy Campus, South Australia 5371, Australia b
a r t i c l e
i n f o
Article history: Received 26 July 2010 Received in revised form 30 September 2010 Accepted 8 October 2010 Keywords: Neospora caninum Randomized controlled challenge trial Monensin Cattle
a b s t r a c t Using a randomized controlled trial design, a randomly allocated intervention group of 15 cows received a slow-release bolus that delivered 100 days of monensin. The negative control group of 15 cows received a placebo bolus that was identical to the monensin bolus, except without the monensin. Two weeks after bolus administration, all cows were challenged with a 2 ml subcutaneous injection of a live tachyzoite suspension. Whole blood and serum samples were collected from each cow every week for the first month postchallenge, and then every 2 weeks for the next 2 months. The extracted DNA from whole blood was tested for the Nc-5 gene fragment of Neospora caninum using a quantitative real time polymerase chain reaction. Serum was tested for antibodies to N. caninum using the IDEXX ELISA. Cows treated with monensin boluses had a significantly lower humoral immune response than cows treated with placebo boluses at one time point post-challenge (week 4 post-challenge). However, when adjusting for repeated measures within cows, the P value for this humoral difference was 0.098. No DNA for N. caninum was detected in either group, likely due to study design features. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Neospora caninum is a costly intracellular protozoan that is a major cause of epidemic and endemic abortions in cattle world-wide. Cattle remain infected for life and can undergo repeat abortions. The abortions can lead to reduced milk production (Tiwari et al., 2007), lower reproductive efficiency (VanLeeuwen et al., 2010b), and premature culling (Tiwari et al., 2005). Total direct production costs associated with neosporosis in Canadian dairy cattle were determined to be $2305/year/infected herd of
∗ Corresponding author at: 550, University Ave, Charlottetown, PEI, Canada C1A 4P3. Tel.: +1 902 566 0457; fax: +1 902 620 5053. E-mail address: [email protected] (J.A. VanLeeuwen). 0304-4017/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.vetpar.2010.10.016
50 cows (Chi et al., 2002). Similar cost estimates have been determined elsewhere (Trees et al., 1999; Bartels et al., 2006; Dubey et al., 2007), demonstrating that N. caninum is an economically important disease of the dairy industry. Cattle (intermediate hosts) become infected by transplacental (vertical) transmission of the fast-growing form (tachyzoite) of the protozoan from dam to calf, or by ingestion (horizontal) of oocysts from exposure to feces from wild or domestic dogs (definitive hosts) that are infected and actively shedding the parasite (McAllister et al., 1998; Gondim et al., 2004). After cattle become infected, either vertically or horizontally, their immune system tries to eliminate the infection, at which time, the circulating protozoa transform into hardy slowly dividing forms (bradyzoites) and encyst in nerve cells where they
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
can remain dormant for months or years (Dubey et al., 2007). To date, there has been no effective treatment demonstrated against N. caninum. However, we conducted a risk factor analysis of 240 dairy farms across Canada and showed that cows receiving monensin (a polyether ionophore produced by Elanco, based in Greenfield, Indiana) as a feed additive were 1.5 times less likely (P < 0.05) to be infected with N. caninum than cows not receiving monensin (VanLeeuwen et al., 2010a). A similar trend was seen in supplementary analyses (Duffield T., Ontario Veterinary College, Guelph, ON, Canada, personal communication) of abortions due to N. caninum in herds in Ontario (Hobson et al., 2005). In laboratory experiments, monensin was shown to reduce the population of N. caninum tachyzoites in cell culture by at least 95% at concentrations as low as 0.001 ng/ml (Lindsay et al., 1994). Serum concentrations in cows given monensin boluses achieve these levels (Bagg R., Elanco Animal Health Ltd., Guelph, ON, Canada, personal communication). No studies have looked at the effect of monensin on the hardy bradyzoites in the lab or in the field. Clinical trials are needed to determine the effect of monensin on N. caninum in cows. The study objective was to determine if monensin had an effect on N. caninum tachyzoite infection in cattle. The study was undertaken after approval of the Animal Care Committee at the University of Prince Edward Island. 2. Materials and methods The study group started out with all of the 30 nonmilking, non-pregnant cows in the teaching barn at the Atlantic Veterinary College (AVC) because commercial dairy farmers would not want their animals to be intentionally infected with N. caninum. Because the AVC teaching cows were purchased cull cows used for teaching (and research), limited information on the cows was available. Breed and weight were available and recorded. There were 6 Jersey cows, with the remainder being Holstein. The average weight was 702 kg, with a range of 532–820 kg. The cows were all non-pregnant and not bred during the study, so N. caninum infection could not cause abortions. The study was a randomized controlled trial with two groups of cows. On January 2nd, 2009, half of the 30 cows were randomly allocated to the intervention group (using computer-generated random numbers), and they received a slow-release bolus that delivered 100 days of monensin. The remaining cows in the negative control group received a placebo bolus that was identical to the monensin bolus, except without the monensin. The bolus serial numbers were recorded but AVC barn staff reported no regurgitated boluses. Just prior to the random allocation and administration of boluses, serum samples from all cows were obtained and tested for antibodies against N. caninum using the Herdchek anti-Neospora IDEXX ELISA (Westbrook, ME, USA), according to the manufacturer’s instructions. Only one cow was found to be positive for N. caninum on serology, and it was in the placebo group so it was removed from the final dataset, along with one randomly selected cow from the intervention group.
373
Three weeks after bolus administration, all cows were challenged with a 2 ml subcutaneous injection of a live tachyzoite suspension (NC-1, 2.5 × 106 tachyzoites/ml within 2 h of harvest), as described (Liddell et al., 1999) and successfully utilized in a previous vaccination efficacy trial using N. caninum challenge (O’Handley et al., 2003). Over the next 3 months, whole blood and serum samples were collected from each cow every week for the first month post-challenge, and then every 2 weeks. Whole blood samples (1 ml) underwent DNA extraction (Qiagen DNeasy Blood and Tissue kit according to the manufacturer’s instructions) within 8 h of sampling, and the DNA samples were frozen at −20 ◦ C until all samples were collected. The extracted samples were split into 3 aliquots and tested for the presence and estimate of numbers of tachyzoites (outcome variables of interest), using a quantitative real-time polymerase chain reaction (qPCR) for the Nc-5 gene fragment of N. caninum that was validated and optimized at the AVC, based on a recently published protocol (Okeoma et al., 2005). Standard curves were constructed for both N. caninum (Nc-5) and bovine (28S rRNA) genes to ensure comparable PCR efficiency. The PCR reactions for Nc5 and 28S rRNA genes were run in triplicate and duplicate, respectively, on all samples. Positive controls for N. caninum were detected at Ct 19–20, and samples were only considered positive if a Ct value (20–40) was detected in at least two of the triplicate blood samples. For serum testing, 1 ml of serum from each cow was frozen at −20 ◦ C until all samples were collected, and then tested for antibody levels to N. caninum using the Herdcheck Anti-Neospora IDEXX ELISA (Westbrook, ME, USA). Sample-to-positive ratios (S/P ratio) on the ELISA were recorded. A cow was classified as infected with N. caninum if the sample-to-positive ratio (S/P ratio) on the ELISA was ≥0.5, as recommended by the manufacturer of the test kit. At this cut-off, the test has a sensitivity of 93% and specificity of 94% (Wapenaar et al., 2007). All researchers, laboratory staff, and animal care-givers were blinded to the group status of the cows, except one administrative technician. The continuous outcome variables of interest (S/P ratios for ELISA results, and number of tachyzoites for PCR results) were tested for normality using the sktest command (i.e. skewness and kurtosis) in STATA Version 11 (College Station, TX, USA) and if the variables were not normally distributed, they were transformed using a natural logarithm transformation (ln). Because S/P ratios ranged from 0.01 to 1.78, a value of 1 was added to every S/P ratio, prior to transformation, so that the ln S/P ratio would not be a negative number. The means and standard deviations of the continuous outcome variables (including ln transformed outcomes, where applicable – which were converted back to the regular scale for presentation) were calculated, by treatment group and time point. Significant differences in the outcome variables were compared (P < 0.05) between the two groups at each time point using an unpaired t-test. Using the dichotomous outcome variables of interest (seropositivity and PCR positivity), chi-square tests were used to determine differences in the proportion of positive cows between the two groups at each time point.
374
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
Table 1 Post-challenge serological results of N. caninum antibody testing for cows receiving either a monensin (n = 14) or placebo (n = 13) bolus, by week of study. Placebo group Range S/P ratio Week 0 Week 1 Week 2 Week 3 Week 4 Week 6 Week 8 Week 10
0.03–0.18 0.01–0.19 0.08–1.70 0.17–1.78 0.12–1.51 0.08–1.29 0.05–0.97 0.05–0.98
Monensin group Average ln S/P ratioa
S.D. ln S/P ratioa
% Positive
0.086 0.065 0.672 0.810† 0.834* (0.05) 0.703 0.564 0.490
0.039 0.043 0.341 0.335 0.288 0.239 0.221 0.205
0% 0% 62% 77%† 77%† 69%† 69%† 54%
Range S/P ratio 0.02–0.24 0.01–0.14 0.09–1.76 0.26–1.38 0.28–1.19 0.23–1.22 0.12–0.85 0.05–0.82
Average ln S/P ratioa
S.D. ln S/P ratioa
% Positive
0.095 0.058 0.485 0.563 0.554 0.541 0.407 0.344
0.053 0.038 0.291 0.205 0.169 0.212 0.208 0.195
0% 0% 43% 43% 43% 36% 36% 29%
a ln S/P ratio = Sample-to-positive ratios were transformed using a natural logarithm due to non-normal distribution, then averages and standard deviations were determined, and then these calculations were converted back to the regular scale. * P < 0.05 with specific P-value in parentheses. † P < 0.10.
In order to control for clustering of repeated test measures within cows, XTGEE models were used to determine significant associations for the continuous outcome variables, using an AR correlation structure, in SAS Version 10 (Cary, NC, USA). For dichotomous outcomes, XTGEE was again used, using the binomial family, logit link, and AR correlation structure. The following sample size calculation was performed to determine the number of cows needed. Using an alpha (type 1 error) of 5% (level of confidence of 95%) and a beta (type 2 error) of 20% (power = 80%), the sample size needed in each treatment group was 16 in order to detect a 35% difference in presence of tachyzoites or positive titres between the two groups. Therefore, our sample size was close to this desired number. 3. Results One cow in the placebo group was lost to follow-up for reasons unrelated to the research project (became very fractious and therefore unsuitable for retention). Therefore, the final dataset included 13 cows in the placebo group and 14 cows in the monensin group. The final sampling of blood was conducted 10 weeks after challenge. With the random allocation, there were 4 and 2 Jerseys in the placebo and monensin groups, respectively. The overall average weight of the cows was 702 kg, with the placebo and monensin groups weighing, on average, 699 and 706 kg, respectively. There were no statistically significant differences in weight or breed between the groups (unpaired t-test and Chi-square test, respectively), suggesting that the random allocation process succeeded in equally balancing the effects of confounding variables, such as breed and weight, among the two groups. The S/P ratios were not normally distributed and therefore were ln transformed. The cows in the placebo group had significantly higher ln S/P ratios than the cows in the monensin group at week 4, with similar trends (P values between 0.05 and 0.15) from week 2 through to week 10 (Table 1). There were no significant differences between the proportions of positive serological results for the placebo (10 of 13 were positive for weeks 3 and 4) and monensin (6 of 14 were positive for weeks 3 and 4) groups, although
these differences approached statistical significance (P values between 0.05 and 0.1) for weeks 3 through 8. None of the cows in either group had positive PCR results at any of the time points of the study. All positive controls for N. caninum were detected in each qPCR assay, and the bovine 28S rRNA gene was detected in all samples, confirming that the PCR assay was functioning properly. In the final XTGEE model of ln S/P ratio, controlling for clustering of measurements at various time points within cows, breed, weight and treatment group did not remain significant (P > 0.05). However, there was a trend (P = 0.098) toward cows in the placebo group having a higher average ln S/P ratio compared to cows in the monensin group (results not shown due to not statistically significant at P < 0.05). Similarly, the XTGEE model of seropositivity determined that there was only a trend (P = 0.08) toward more cows with a positive N. caninum titre in the placebo group than in the monensin group (results not shown again due to not statistically significant at P < 0.05). No cows had any adverse events related to the administration of the boluses or injections of the N. caninum challenge. 4. Discussion This study demonstrated a significant effect of monensin on reducing the humoral response to challenge infection of cows with N. caninum at one time point. However, the statistical analyses that controlled for repeated measures within cows did not show a statistically significant difference in N. caninum titres between the placebo and monensin groups over the length of the study, for various reasons. First, substantially elevated titres only occurred for part of the study timeframe. It was expected that weeks 1 and 2 would be less likely to show significant differences between groups because of the time required for the cows to mount a detectable immune response post-challenge. The average S/P ratios peaked in week 4, waning thereafter, with some cows even converting from positive to negative titres. Antibody titres were seen to peak at approximately a month post-vaccination and postchallenge in another clinical trial on N. caninum (O’Handley et al., 2003), with conversion from positive to negative titres being reported elsewhere as well (Sager et al., 2001).
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
Second, the limited sample size made it more difficult to find a significant difference in S/P ratios and seropositivity, although we did use all cows available at our institution. Alternatively, the lack of significance could have been due to random variation and no real difference between groups. More research is needed to confirm or refute our results. Clearly, the monensin group did end up having some cows that appeared to become infected with N. caninum, based on substantial rises in titres (2 monensin-treated cows in particular rose from 0.03 and 0.08 to 1.12 and 1.68, respectively), suggesting that, if monensin does have some effect on N. caninum infection, other factors must be involved in determining whether monensin boluses can potentially prevent infection. Perhaps one factor could be the level of monensin in each animal, which was not measured in this study. The boluses are manufactured to release 335 mg/d, but the levels circulating in the body of any given cow would depend on the consistency of release of monensin from the bolus, the weight of the cow, and the rate of metabolism of the monensin in each cow. The 2 monensin cows with high S/P ratios, mentioned above, had weights of 790 and 780 kg, respectively, higher than average in the group. Perhaps one monensin bolus does not lead to a sufficient level of monensin circulating in the body of a heavy cow to reduce the likelihood of N. caninum infection. Further research could help to elucidate these factors. The reasons for none of the cows in either group having positive PCR results at any of the time points of the study are likely related to the methodology and dynamics of the parasite after experimental challenge. Although PCR methods generally have the benefit of being sensitive and specific, they still have limits of detection. The current test was able to detect from −1 to 10−4 (77.6–0.077 pg) copies of N. caninum tachyzoite per mg of total DNA. Therefore, if the number of circulating tachyzoites in infected cows was below this limit, a false negative result would be obtained, leading to reduced test sensitivity. Several scenarios could arise in which this could happen: (1) if the inoculation dose was too low and the number of tachyzoites/ml was too dilute for detection (5 × 106 tachyzoites/cow was chosen from a previous study that was able to detect parasitemia by PCR in sheep (O’Handley et al., 2003)); (2) if the tachyzoites were viable at injection but subsequently became non-viable and therefore were unable to reach general circulation from the subcutaneous injection site (tachyzoites were injected within 2 h of preparation and we have been able to maintain viability at 4 ◦ C for between 5 and 6 h); (3) if viable tachyzoites were cleared rapidly from circulation; and (4) if parasitemia was transient and did not coincide with sampling (Macaldowie et al., 2004). The sporadic nature of N. caninum DNA detection by PCR has been observed previously with no clear consensus. Numerous reasons surrounding cattle immunocompetence, sex, and pregnancy status have been considered. However, temporal studies in mice have revealed that tachyzoites were quickly cleared from circulation during the first week post-inoculation, and only occasionally reached PCR detection limits over time (CollantesFernandez et al., 2006). Before recommendations can be made on monensin use for N. caninum, further research is needed in other larger
375
populations of cattle, preferably pregnant, with a modified sampling timeframe for detecting tachyzoite DNA. Furthermore, trials on monensin’s effectiveness on more natural modes of N. caninum transmission (e.g. bradyzoite recrudescence leading to trans-placental transmission, or oocyst ingestion followed by sporozoite and then tachyzoite infection) would also be needed before recommendations could be made. Conflict of interest statement The authors have no financial or personal relationships with other people or organizations that could inappropriately influence (bias) their work. Acknowledgements The authors acknowledge the technical support of T. Andrews, R. Milton, and L. Dalziel (Department of Health Management, Atlantic Veterinary College). This work was carried out with the financial support of the Atlantic Veterinary College and Elanco Animal Health. The funding sources agreed with the proposed study design but did not influence, in any way, the study design, collection, analysis or interpretation of the data, or the writing of the manuscript and its submission for publication. References Bartels, C.J., van, S.G., Veldhuisen, J.P., van den Borne, B.H., Wouda, W., Dijkstra, T., 2006. Effect of Neospora caninum-serostatus on culling, reproductive performance and milk production in Dutch dairy herds with and without a history of Neospora caninum-associated abortion epidemics. Prev. Vet. Med. 77, 186–198. Chi, J., VanLeeuwen, J.A., Weersink, A., Keefe, G.P., 2002. Direct production losses and treatment costs from bovine viral diarrhoea virus, bovine leukosis virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum. Prev. Vet. Med. 55, 137–153. Collantes-Fernandez, E., Lopez-Perez, I., Alvarez-Garcia, G., Ortega-Mora, L.M., 2006. Temporal distribution and parasite load kinetics in blood and tissues during Neospora caninum infection in mice. Infect. Immun. 74, 2491–2494. Dubey, J.P., Schares, G., Ortega-Mora, L.M., 2007. Epidemiology and control of neosporosis and Neospora caninum. Clin. Microbiol. Rev. 20, 323–367. Gondim, L.F., McAllister, M.M., Pitt, W.C., Zemlicka, D.E., 2004. Coyotes (Canis latrans) are definitive hosts of Neospora caninum. Int. J. Parasitol. 34, 159–161. Hobson, J.C., Duffield, T.F., Kelton, D., Lissemore, K., Hietala, S.K., Leslie, K.E., McEwen, B., Peregrine, A.S., 2005. Risk factors associated with Neospora caninum abortion in Ontario Holstein dairy herds. Vet. Parasitol. 127, 177–188. Liddell, S., Jenkins, M.C., Dubey, J.P., 1999. A competitive PCR assay for quantitative detection of Neospora caninum. Int. J. Parasitol. 29, 1583–1587. Lindsay, D.S., Rippey, N.S., Cole, R.A., Parsons, L.C., Dubey, J.P., Tidwell, R.R., Blagburn, B.L., 1994. Examination of the activities of 43 chemotherapeutic agents against Neospora caninum tachyzoites in cultured cells. Am. J. Vet. Res. 55, 976–981. Macaldowie, C., Maley, S.W., Wright, S., Bartley, P., Esteban-Redondo, I., Buxton, D., Innes, E.A., 2004. Placental pathology associated with fetal death in cattle inoculated with Neospora caninum by two different routes in early pregnancy. J. Comp. Pathol. 131, 142–156. McAllister, M.M., Dubey, J.P., Lindsay, D.S., Jolley, W.R., Wills, R.A., McGuire, A.M., 1998. Dogs are definitive hosts of Neospora caninum. Int. J. Parasitol. 28, 1473–1478. O’Handley, R.M., Morgan, S.A., Parker, C., Jenkins, M.C., Dubey, J.P., 2003. Vaccination of ewes for prevention of vertical transmission of Neospora caninum. Am. J. Vet. Res. 64, 449–452. Okeoma, C.M., Stowell, K.M., Williamson, N.B., Pomroy, W.E., 2005. Neospora caninum: quantification of DNA in the blood of naturally
376
J.A. VanLeeuwen et al. / Veterinary Parasitology 175 (2011) 372–376
infected aborted and pregnant cows using real-time PCR. Exp. Parasitol. 110, 48–55. Sager, H., Fischer, I., Furrer, K., Strasser, M., Waldvogel, A., Boerlin, P., Audige, L., Gottstein, B., 2001. A Swiss case–control study to assess Neospora caninum-associated bovine abortions by PCR, histopathology and serology. Vet. Parasitol. 102, 1–15. Tiwari, A., VanLeeuwen, J.A., Dohoo, I.R., Keefe, G.P., Haddad, J.P., Tremblay, R., Scott, H.M., Whiting, T., 2007. Production effects of pathogens causing bovine leukosis, bovine viral diarrhea, paratuberculosis, and neosporosis. J. Dairy Sci. 90, 659–669. Tiwari, A., VanLeeuwen, J.A., Dohoo, I.R., Stryhn, H., Keefe, G.P., Haddad, J.P., 2005. Effects of seropositivity for bovine leukemia virus, bovine viral diarrhoea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum on culling in dairy cattle in four Canadian provinces. Vet. Microbiol. 109, 147–158. Trees, A.J., Davison, H.C., Innes, E.A., Wastling, J.M., Tenter, A.M., Shirley, M.W., 1999. Towards evaluating the economic impact of bovine
neosporosis. Special issue: COST820: Vaccines against animal coccidioses. Papers from the Annual Workshop, Toledo, Spain. Int. J. Parasitol. 29, 1195–1200. VanLeeuwen, J.A., Haddad, J.P., Dohoo, I.R., Keefe, G.P., Tiwari, A., Scott, H.M., 2010a. Risk factors associated with Neospora caninum seropositivity in randomly sampled Canadian dairy cows and herds. Prev. Vet. Med. 93, 129–138. VanLeeuwen, J.A., Haddad, J.P., Dohoo, I.R., Tiwari, A., Keefe, G.P., Stryhn, H., Tremblay, R., 2010b. Associations between reproductive performance and seropositivity for bovine leukemia virus, bovine viral-diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in Canadian dairy cows. Prev. Vet. Med. 94, 54–64. Wapenaar, W., Barkema, H.W., VanLeeuwen, J.A., McClure, J.T., O’Handley, R.M., Kwok, O.C., Thulliez, P., Dubey, J.P., Jenkins, M.C., 2007. Comparison of serological methods for the diagnosis of Neospora caninum infection in cattle. Vet. Parasitol. 143, 166–173.