Pooled Polymerase Chain Reaction to Detect Tritrichomonas Foetus in Beef Bulls

Abstract

Preputial scraping samples from 305 mixed breed beef bulls were examined for the detection of Tritrichomonas foetus infection. All samples were collected by veterinarians and transported in commercial media to an accredited lab. Upon arrival samples underwent microscopic examination for the presence of Tritrichomonas foetus and were then incubated until 5 days postcollection before final microscopic examination. Culture detected 14 samples with Trichomonad spp.; all were confirmed to be Tritrichomonas foetus by polymerase chain reaction (PCR). After final examination samples were randomly placed in groups of 5 samples; technicians were blinded as to culture results of the individual samples constituting each pool. From each sample within a group, a portion of the fluid sediment was removed and pooled with the other samples of the group to form 61 pools. From each of the formed pools an aliquot was removed for PCR. PCR detected 16 positive pools; an additional 2 positive samples were then identified on individual PCR on samples previously diagnosed as culture negative. Relative to culture, the 95% confidence intervals for sensitivity and specificity of PCR pools to detect Tritrichomonas foetus were 76.8% to 100% (mean value: 100%) and 85.5 to 99.5% (mean value: 93.4%), respectively.

Tritrichomonas foetus is a flagellated protozoa associated with infertility, early embryonic death, and abortion in cattle. 1,6,13,14 The organism is transmitted to susceptible females during coitus by persistently infected asymptomatic bulls and results in vaginitis, cervicitis, endometritis, and abortion. 1,5,6,13,14 Most often infected females recover and after a period of convalescence and sexual inactivity return to normal reproductive performance. 1,6,13,14 Infected cows serve as reservoirs for bulls that become infected and further spread the organism to susceptible cows during breeding. Infected bulls frequently present with no apparent signs of infection. In the United States, some states have perceived T foetus as having such a significant impact that mandatory regulatory programs have been adopted requiring presale and import testing. This is especially important in states where management practices include multiple source cows grazing on communal pastures. Control and eradication programs predominantly rely on single or multiple cultures of preputial scrapings (PPS) of bulls prior to the breeding season. A considerable variation of reported culture sensitivity (Se) exists ranging from as low as 70% to as high as 92%. 5,8,12 However, a potential pitfall of culture alone is the lack of specificity (Sp), and false positive results may lead to the sacrifice of valuable breeding animals at a considerable expense to the producer. 5 In order to avoid undue sacrifice, follow-up polymerase chain reaction (PCR) testing on culture-positive samples has been recommended. PCR utilizing the 5.8S-ITS coding region results in the ability to differentiate T foetus from other tritrichomonads. 2,3,6,10 PCR for T foetus has a reported Se of 98.38% and a Sp of 93.75%, relative to culture. 9,12 Sensitivity for culture alone has been reported as low as 70% for a single culture and as high as 97.3% for 3 successive cultures completed 1 week apart. 9,12 Because of its high reported Se some state T foetus control programs have forgone culture and moved directly to PCR, either as a single stand-alone test or using repeated tests on samples collected within 7 to 10 days of each other. 8,11 Tritrichomonas foetus PCR is potentially the test of choice for the detection of T foetus but is limited by the cost of individual laboratory tests, ranging from $20.00 to $35.00 plus veterinarian charges for sample collection. To determine the presence of T foetus infection within a cattle population by individual T foetus, PCR may be cost prohibitive. The present study examined the potential for pooling samples and as a cost-effective approach to estimate a true prevalence for T foetus in a population of bulls.

Veterinarians certified by the Colorado Department of Agriculture in proper trichomonad collection techniques collected 305 samples from nonvirgin bulls as part of a Colorado T foetus control program. After collection samples were transported via commercial shippers and US Postal Service to an American Association of Veterinary Laboratory Diagnosticians (AAVLD)-accredited laboratory in commercially available transport/culture pouches. a Upon receipt samples were accessioned into the diagnostic laboratory system with special attention given the date of collection of samples and the condition of the samples in regard to shipping. Samples collected and shipped without properly controlling for temperature or with extended in-transit times were removed from the study, and a new submission was requested. Samples received greater than 2 days after collection but fewer than 5 were microscopically examined for viable trichomonads at low power and low light, with special attention at the area near the seams of the pouches, and then placed vertically into racks and incubated at 35°C until day 5 postcollection, when a final microscopic evaluation was accomplished. Samples received that had been collected fewer than 2 days prior to arrival at the laboratory were cultured as described above and read on the second day postcollection, then returned to the incubator until the fifth day postcollection, when the final microscopic evaluation was performed.

A preliminary study was conducted to estimate the number of trichomonads per ml necessary for detection by PCR by doing serial dilutions of a laboratory positive control sample. A count of the laboratory stock solution indicated a concentration of 3,000 organisms per ml were present. Serial dilutions using fluid from unused pouches as the diluent were made from the stock solution until a dilution of 10⁻⁴ was reached. The resulting 5 available culture solutions contained 30 organisms, 3 organisms, 0.3 organisms, 0.03 organisms, and 0.0003 organisms per 10 μl. At each dilution level 14 identical aliquots were made and placed at a −70°C, ensuring no further organism replication occurred. Individual gel-based PCR was accomplished on each diluted sample at each log dilution with the following results. PCR results were interpreted by comparing the test band to the T. foetus positive control band found at 347 bp. Where matches were found between the positive control and the test band, samples were classified as T. foetus detected, while if a match was not found the sample was classified as T. foetus not detected. At 10°, 10⁻¹, and 10⁻² the organism was detected in 100% (42/42) of the samples, while at 10⁻³ 86% (12/14), and at 10⁻⁴ 50% (7/14) yielded positive PCR detection. Based on this preliminary study and an estimate of the number of bulls composing a herd bull battery, a pool size of 5 samples was selected for this study.

All 305 pouches were placed into groups of 5 by a second technician that ensured that only 1 positive would be contained within any of the 61 subgroups. A third technician blinded as to the previous test results was assigned the task of pooling and completing the PCR. All of the pouches were stored vertically to allow any sediment to fall to the bottom of the pouch. After approximately 1 hour a sterile disposable pipette b was used to remove a 1-ml aliquot from each pouch, ensuring to include within the aliquot the majority of the sediment from each pouch. The aliquot was then placed in a Cryogenic Vial c labeled with the sample identification number. From each of the 5 1-ml samples a 250-μl aliquot was removed using a sterile disposable pipette and placed in a 2-ml tube d designated with the pool number. This was repeated with all 305 samples until 61 pools of 5 samples each were formed. Pipettes were discarded after each aliquot was removed to avoid cross-contamination resulting in false-positive results. The remaining portion of each individual sample was retained at −70°C for further testing in case of a PCR-positive result in a pool. From each pool 200 μl of fluid was withdrawn for DNA extraction and PCR. Extraction was accomplished using a commercial kit e following the protocol outlined by the manufacturer. PCR was accomplished following a previously published protocol that amplifies the 5.8S-ITS coding region to detect the presence of T. foetus. 2,3,6

The PCR yielded a positive result for T. foetus in 16 out of the 61 pools, identifying 2 pools containing samples that had previously been considered negative by culture. Individual PCR tests were accomplished on the constituents of the 2 pools previously negative by culture, and a single positive sample was found in each pool. Applying a statistical analysis to the results when pools versus culture were compared, a mean kappa of 0.91 was obtained (with upper and lower confidence intervals (CIs) of 0.79 to 1 at the 95% level); when comparing PCR-positive pools to the PCR of individuals composing the positive pools total agreement, that is, a PCR positive individual was found in each positive pool. To demonstrate repeatability between laboratories, the 16 known positive pools along with 16 known negative pools were forwarded to a state diagnostic laboratory for confirmation of results. The referral laboratory was blinded as to previous test results. Statistical analysis of test results between the submitting lab and referral lab resulted in a kappa statistic of 0.75, indicating substantial agreement between the 2 laboratories. 7 The referral lab failed to detect 4 positive pools but was in total agreement with the submitting laboratory on negative results. The failure to detect the 4 positive samples may be attributable to multiple freeze-thaw cycles of the samples and the use of a tissue DNA extraction procedure versus a blood DNA extraction procedure.

Although positive pools would require further testing to identify positive individuals, pooled diagnostic testing would offer the following advantages 4,15 : 1) cost advantages related to laboratory fees; 2) improved accuracy over repeating a test each time, thus compounding the inherent inaccuracies of the test, which is an important consideration for diseases with low prevalence; 3) more precise estimates of prevalence when Se and Sp are less than 1; and 4) less bias when assumed Se and Sp are not equal to true values. Conversely, the disadvantages of pooling include 4,15 : 1) the potential loss of Se due to dilution; 2) the costs to pool samples; 3) retesting costs on positive pools; and 4) failing to evaluate adequate numbers of pools, before predicting prevalence yielding incorrect results.

Based upon data generated in the present study, pooling of 5 samples did not negatively impact the Se nor Sp of individual PCR tests and exceeded the Se of a single culture. The Se and Sp of pooled testing was 100% when compared to individual PCR tests. The application of Se and Sp of 100% to a pooled prevalence calculator (available at http://203.12.171.42/pprev/), would estimate the point prevalence in this subset of bulls to be 5.9% with the upper and lower CI at the 95% level of 3.7% and 8.85%, respectively. 15 This would compare to a point prevalence value of 4.6% (14 of 305) if prevalence were determined by calculating culture positive divided by total samples cultured. If point prevalence was calculated by dividing the individual PCR positives by the total samples tested by PCR (16 of 305), a point prevalence of 5.2% would be obtained

The use of pooled PCR to detect the prevalence of T. foetus has the potential to provide a valuable tool to aid in governmental decisions concerning implementation of mandatory T. foetus regulations. Another potential application of pooled testing is monitoring herd status after herds or states have been declared free or at low risk of T. foetus infection.

The utilization of pooling and PCR could minimize the number of collections per bull, as well as provide Se that would approach or surpass multiple cultures. The use of PCR relies on DNA, not viable organisms, and is capable of detecting the presence of the organism when improper transport or collection technique has resulted in the loss of viable organisms.