Comparison of the performance of three PCR assays for the detection and differentiation of Theileria orientalis genotypes

Background Oriental theileriosis is a tick-borne disease of bovines caused by the members of the Theileria orientalis complex. Recently, we developed a multiplexed tandem (MT) PCR to detect, differentiate and quantitate four genotypes (i.e., buffeli, chitose, ikeda and type 5) of T. orientalis. In this study, we used MT PCR to assess the prevalence and infection intensity of four T. orientalis genotypes in selected cattle herds that experienced oriental theileriosis outbreaks in New Zealand, and compared the sensitivities and specificities of MT PCR, PCR-high resolution melting (PCR-HRM) and a TaqMan® qPCR. Methods MT PCR, PCR-HRM analysis for T. orientalis and a TaqMan® qPCR assay for ikeda genotype were employed to test 154 and 88 cattle blood samples from North (where oriental theileriosis outbreaks had occurred; designated as Group 1) and South (where no outbreaks had been reported; Group 2) Islands of New Zealand, respectively. Quantitative data from MT PCR assay were analyzed using generalized linear model and paired-sample t-test. The diagnostic specificity and sensitivity of the assays were estimated using a Bayesian latent class modeling approach. Results In Group 1, 99.4% (153/154) of cattle were test-positive for T. orientalis in both the MT PCR and PCR-HRM assays. The apparent prevalences of genotype ikeda in Group 1 were 87.6% (134/153) and 87.7% (135/154) using the MT PCR and Ikeda TaqMan® qPCR assays, respectively. Using the MT PCR test, all four genotypes of T. orientalis were detected. The infection intensity estimated for genotype ikeda was significantly higher (P = 0.009) in severely anaemic cattle than in those without anaemia, and this intensity was significantly higher than that of buffeli (P < 0.001) in the former cattle. Bayesian latent class analysis showed that the diagnostic sensitivities (97.1-98.9%) and specificities (96.5-98.9%) of the three PCR assays were very comparable. Conclusion The present findings show the advantages of using the MT PCR assay as a useful tool for in-depth epidemiological and transmission studies of T. orientalis worldwide.


Background
Theileria species (Apicomplexa: Piroplasmida; Theileriidae) are tick-transmitted intracellular protists that infect various domestic and wild ruminants worldwide [1]. Depending on their pathogenicity, Theileria species can be divided into two groups: (i) those that transform host cells (including T. annulata and T. parva) and (ii) those that do not (e.g., Theileria orientalis complex) [2]. Theileria annulata and T. parva cause the most severe forms of bovine theileriosis, whereas members of T. orientalis complex were recognized to usually cause a less severe form (oriental theileriosis) in cattle. However, in recent years, T. orientalis has been associated with significant outbreaks of oriental theileriosis in Australia [3,4] and New Zealand [5,6], causing pyrexia, haemolytic anaemia, productivity losses, abortions and/or mortality in dairy and beef cattle [3][4][5][6][7][8].
Recently, McFadden et al. [5] reported that genotype chitose was linked to an outbreak of oriental theileriosis in apparently naïve cattle being transported to the northern part of the North Island of New Zealand. Since August 2012, the number of theileriosis outbreaks has increased considerably in this geographical region, and in some herds, there have been significant morbidity, mortality and productivity losses [6]. In recent studies, the identification of genotypes of T. orientalis was carried out on samples collected from herds with theileriosis outbreaks, and genotype ikeda was identified [6,13].
Clinical signs, haematology, serology and/or molecular tools have been used for the diagnosis of theileriosis in New Zealand [5,13]; however, owing to increasing numbers of oriental theileriosis outbreaks in this country, there has been a need for a rapid and accurate method of identification of pathogenic genotypes of T. orientalis. Therefore, Pulford et al. (D. J. Pulford, and A. M. J. McFadden, unpublished observations) developed PCRhigh resolution melting (PCR-HRM) and TaqMan W qPCR methods employing the MPSP gene as a marker for the detection of genotype ikeda of T. orientalis in blood samples from cattle in New Zealand. Although the combined use of the PCR-HRM and TaqMan W qPCR allows the identification of the ikeda genotype, these assays do not specifically identify or differentiate other genotypes of T. orientalis (i.e., buffeli, chitose and type 5), or quantify the amount of DNA of these genotypes (D. J. Pulford, and A. M. J. McFadden, unpublished observations).
To overcome the limitations associated with conventional molecular tools (such as low diagnostic sensitivity and/or specificity, quantitation and time required for testing), Perera et al. [14] recently established and validated a multiplexed tandem PCR (MT PCR) assay for the simultaneous detection, differentiation and quantitation of four of the commonest genotypes (i.e., buffeli, chitose, ikeda and type 5) of T. orientalis in Australasia. Subsequently, Perera et al. [15] used MT PCR to estimate the prevalence and intensity of these four genotypes in cattle in 15 dairy herds in the State of Victoria, Australia, and demonstrated the utility, high performance, throughput and convenience of the assay for diagnostic and epidemiological applications. Using MT PCR, these authors were able to determine the prevalence of four common T. orientalis genotypes (i.e., buffeli, chitose, ikeda and type 5) and quantify the DNA copy number for each of these genotypes in individual cattle.
As three novel molecular-diagnostic assays have been developed independently in Australia (MT PCR) and New Zealand (PCR-HRM and TaqMan W qPCR), the aims of the present study were to: (i) employ MT PCR to estimate the prevalence and infection intensity of T. orientalis genotypes buffeli, chitose, ikeda and type 5 in selected cattle herds that had experienced oriental theileriosis outbreaks in New Zealand, and (ii) compare the sensitivities and specificities of all three assays.

Methods
Farms, demographic characteristics of cattle, and blood collection Blood samples (n = 154) were collected by registered veterinarians from 103 beef (Angus, Hereford or crosses) or dairy (Friesian or crosses) herds experiencing oriental theileriosis (Group 1) in four regions (i.e., Auckland, East Cape, Northland and Waikato) from the North Island, New Zealand, between July and December 2013 ( Figure 1; Table 1). The impact of this disease was recorded as 'moderate' (1-2 cattle died) on 61 farms and 'severe' (≥5 cattle died and > 30% of the herd were anaemic) on one farm ( Figure 1). Clinical signs recorded were acute death, pale vulval mucous membranes, anorexia, lethargy, tachycardia and/or tachypnoea. At necropsy, affected cattle were often jaundiced, with watery un-clotted blood in major blood vessels. Information on age and breed of cattle in each herd (where possible) was obtained. In addition to samples from cattle in herds suffering from outbreaks, blood samples (n = 88) from nine unaffected herds were randomly collected from asymptomatic beef cattle (Group 2) in the Southland region of the South Island, New Zealand, with no history of theileriosis (Table 1). Although ticks are abundant in North Island, there was no evidence of ticks present in the herds representing Group 2.

Isolation of genomic DNA
Individual blood samples (i.e., 40 μl of blood from each individual) were diluted 1:5 in sterile Elix water, and genomic DNA was then extracted from 200 μl of each sample using the QIAxtractor system and the DX universal liquid sample DNA extraction kit (Cat. No. 950107; Qiagen) using the manufacturer's protocol (eluting into 200 μl).  Disease groups were categorized according to clinical signs and haematological parameters. 'NR' , 'ANT' , 'MO' and 'SO' denote no disease reported, anaemia negative Theileria (i.e., anaemia not observed but Theileria piroplasms were observed on blood smears), moderate outbreak (i.e., more than one animal with anaemia) and severe outbreak (i.e., with multiple deaths), respectively.

PCR-HRM
To amplify all genotypes of T. orientalis, this assay was per-

Statistical analyses
For MT PCR, the DNA copy numbers of individual genotypes of T. orientalis recorded in each blood sample was log-transformed. A generalized linear model was then used to compare the relative intensity of infection for each genotype in individual cattle among three categories of anaemia (i.e., severe anaemia, mild anaemia and no anaemia). A paired-sample t-test was used to compare the intensity of infection of the genotypes shown to be dominant, based on prevalence within severely, mildly or non-anaemic cattle (Group 1), using the genotype with the lowest mean DNA copy number as the reference category. The software package SPSS Statistics 22 (IBM) was used for statistical analyses; a P-value of < 0.05 was considered statistically significant. The diagnostic specificity and sensitivity of the MT PCR assay were estimated using a Bayesian latent class modeling approach [17,18] as described previously [14]. Briefly, two sets of conditionally dependent tests (i.e., Set 1; MT PCR and TaqMan W qPCR for genotype ikeda only, Set 2; MT PCR and PCR-HRM for T. orientalis) were carried out on two distinct cattle populations (i.e., Groups 1 and 2) in the absence of a 'gold standard' (i.e., reference samples of known disease status). The apparent prevalence was assumed to be distinct for each population and the diagnostic specificity and sensitivity of each test was assumed to be constant across the two populations. To allow for zero-infection prevalences, the prevalence in each population was modelled as a mixture of point mass at zero and a continuous beta distribution, as described by Branscum et al. [17]. Diagnostic specificity and sensitivity of each test was assumed to be constant across the two populations. The three molecular diagnostic tests were assumed to be dependent (conditional on infection status), because they had the same biological basis, i.e., the detection of DNA of T. orientalis genotypes. Information about the diagnostic specificity and sensitivity of the MT PCR [14], and Ikeda TaqMan W qPCR and PCR-HRM were modelled using independent and informative beta-distributions elicited from two technical experts [R. B. Gasser and A. M. J. McFadden, respectively] with knowledge of the study populations and diagnostic test performance. The technical experts were not involved in the sample collection and/or testing. Agreement statistics (prevalence-adjusted bias-adjusted Kappa, PABAK) [19] were directly calculated as model outputs. Final inferences were presented as the 50%, 2.5% and 97.5% quantiles of the marginal posterior distributions for each of the parameters, corresponding to a posterior median point estimate and a 95% probability interval (95% PI), respectively.

Apparent prevalence of T. orientalis genotypes and estimation of infection intensity by MT PCR
In Group 1, 99.4% (153/154) of cattle sampled from North Island were test-positive for T. orientalis by MT PCR, and 100% apparent prevalence was recorded for cattle from the Auckland, East Cape and Northland regions, and 98.9% (86/87) in Waikato region (Table 1). In Group 2 cattle, there was a low prevalence (2.3%; 2/88) of genotypes buffeli, chitose and type 5 of T. orientalis; genotype ikeda was absent (Table 1). All four genotypes (i.e., buffeli, chitose, ikeda and type 5) were detected in all four geographical regions of the North Island, and buffeli had the highest prevalence (98.0%; 150/153), followed by ikeda (87.6%; 134/153), chitose (52.9%; 81/ 153) and type 5 (19.0%; 29/153). Herds in Auckland, East Cape and Waikato had the highest prevalence of genotype buffeli while those in Northland had both genotypes ikeda and buffeli in higher prevalences. In individual cattle, multiple genotypes of T. orientalis were more commonly detected than single genotypes (Figure 2). Mixed infections with genotypes buffeli and ikeda showed the highest prevalence (46.4%) followed by a combination of genotypes buffeli, chitose and ikeda (25.5%) (Figure 2).
Although genotype buffeli was prevalent in almost all farms in North Island, the relative average intensity of infection with genotype ikeda was dominant (with an average of 349,271 DNA copies), followed by buffeli (170,786), chitose (70,684) and type 5 (11,252) ( Table 1). With the exception of East Cape, outbreaks in all other regions of North Island showed higher infection intensity with genotype ikeda than with genotype buffeli (Table 1).
Based on HCT values (130 blood samples), cattle in Group 1 were categorized as severely anaemic (HCT < 0.15; n = 49), mildly anaemic (HCT 0.15-0.24; n = 31) or non-anaemic/ normal (HCT > 0.24; n = 50) ( Table 2). Haematocrit values were not available for the cattle in Group 2, but were assumed to be normal given that blood had been collected from clinically normal cattle and there was no indication of ikeda or its vector being present in the region. The apparent prevalences of genotypes buffeli and ikeda were high (100%) in severely anaemic cattle. However, in both mildly anaemic and non-anaemic cattle, the apparent prevalence of genotype buffeli was higher (96.8% and 92%, respectively). In nonanaemic cattle, the prevalence of genotype ikeda was lower (66%) than that of genotypes buffeli or chitose (Table 2).
Overall the intensity of infection with genotype ikeda was higher in infected cattle in severely anaemic, mildly anaemic and non-anaemic categories ( Table 2). Intensity of infection with genotype ikeda was significantly higher (P = 0.009) in severely anaemic cattle than that in nonanaemic cattle (Table 3). In mildly anaemic cattle, the intensity of infection with genotype type 5 was significantly higher (P = 0.006) than in non-anaemic cattle ( Table 3). Paired-sample t-test of data for severely anaemic cattle showed that intensity of infection with genotype ikeda was significantly higher than that with buffeli (P < 0.001) ( Table 4).

SSCP-coupled sequencing of MT PCR amplicons
SSCP analysis of 50 of all 154 amplicons representing the genotypes buffeli (n = 15), chitose (n = 10), ikeda (n = 15) and type 5 (n = 10) showed four main profiles. Minor profile variation was detected within genotype buffeli (data not shown), which was reflected in a minute peak melting temperature difference (0.8-0.9°C). DNA sequencing of amplicons revealed that a nucleotide variation of 1.4% was linked to the target region (data not shown).

Apparent prevalence of T. orientalis genotypes by T. orientalis PCR-HRM analysis and Ikeda TaqMan W qPCR
In Group 1, the prevalence of infection with all genotypes of T. orientalis by PCR-HRM was estimated at

Comparison of performance of the three molecular diagnostic methods
All samples that were test-positive by PCR-HRM were also test-positive by MT PCR assay for one or more genotypes of T. orientalis. All samples test-negative by PCR-HRM were also test-negative for T. orientalis genotypes buffeli, chitose, ikeda or type 5 by MT PCR. Of the 135 samples that were test-positive by Ikeda TaqMan

Discussion
This is the first application of the newly developed semiquantitative MT PCR assay [14] for the simultaneous detection and differentiation of four genotypes of T.  orientalis in New Zealand. The MT PCR assay uses three independent markers (i.e., genotype-specific primers) in the p23 gene, MPSP gene and ITS-1 of nuclear ribosomal DNA regions for the simultaneous detection of genotypes buffeli, chitose and type 5, and ikeda, respectively. In contrast, the TaqMan W qPCR and PCR-HRM assays (D. J. Pulford, and A. M. J. McFadden, unpublished observations) both employ primers to the MPSP gene to detect genotypes ikeda and T. orientalis, respectively. Comparison of the MT PCR with the two other PCR techniques using Bayesian latent class modeling showed that the diagnostic sensitivities (≥ 97.1%) and specificities (≥ 96.5%) of the three molecular tools were comparable.
Using both MT PCR and PCR-HRM, we observed that 99.4% (153/154) and 2.3% (2/88) of cattle were infected with T. orientalis using samples from the North (Group 1) and South (Group 2) Islands of New Zealand, respectively. Using the MT PCR assay, we identified all four genotypes of T. orientalis (i.e., buffeli, chitose, ikeda and type 5) in all four regions (i.e., Auckland, East Cape, Northland and Waikato) studied on the North Island, with buffeli having the highest apparent prevalence (98.0%). Using this assay, we were also able to detect the presence of genotypes chitose and type 5 on the South Island, not detectable by the other two tests, and to confirm the absence of genotype ikeda, in accordance with results for both MT PCR and Ikeda TaqMan W qPCR.
In 1982, James et al. [13] reported T. orientalis cases in New Zealand from both dairy and beef cattle, mostly from Wellsford (the northern-most town in the Auckland region). The detection of T. orientalis in Southland cattle in the present study suggests that there may have been  some recent tick activity in the region, although the information on the distribution of the proposed tickvector (Haemaphysalis longicornis) of T. orientalis in New Zealand is lacking for the South Island [20]. A possible reason for the detection of T. orientalis in the Southland might have been the transportation of cattle from the North to the South Island, although this explanation seems unlikely, as beef cattle from Southland tested here were less than two years of age, and because the movement of young beef cattle from the North to the southern South Island for fattening is uncommon in New Zealand. This proposal and the detection of T. orientalis in the South Island [21] prompt a future investigation into the current distribution of H. longicornis on the South Island of New Zealand. This study reports the first detection of genotype ikeda in New Zealand using ITS-1. Previously, McFadden et al. [5] reported genotype chitose in New Zealand using primers targeting the mitochondrial gene cox3 and a region in the nuclear ribosomal small subunit (18S). Although previous studies have identified all four genotypes (i.e., buffeli, chitose, ikeda and type 5) using conventional PCR, employing primers to the p23 and MPSP genes (D. J. Pulford, and A. M. J. McFadden, unpublished observations), the apparent prevalence of T. orientalis determined here is inferred to be considerably more accurate than conventional PCR, due to a 1,000 times higher analytical sensitivity of the MT PCR [14]. In addition, the MT PCR assay is able to estimate the infection intensity for each genotype based on the DNA copy number detected. These results reveal a relationship between a high intensity of genotype ikeda and severe anaemia in cattle affected by oriental theileriosis based on HCT values (< 0.15), suggesting that genotype ikeda is the pathogenic genotype in New Zealand, as is the case for regions in Australia [14,15]. The apparent absence of genotype ikeda from cattle from Southland, where no outbreaks have been reported, also provide support for this proposal. We also found that the intensity of infection with genotype type 5 was significantly higher (P = 0.006) in mildly anaemic cattle than in non-anaemic cattle (Table 3); however, this result should be interpreted with caution due to small sample size in each category of cattle tested here.
It appears that genotype ikeda is the main pathogenic genotype involved in oriental theileriosis in cattle in New Zealand. Although both genotypes chitose and/or ikeda were suggested to be linked to symptomatic oriental theileriosis in cattle in Australasia, based on the presence and/or prevalence of genotypes by conventional molecular diagnostic tools [3][4][5][6][7][10][11][12], the application of MT PCR in Australia to detect and estimate the infection intensities of individual T. orientalis genotypes provided evidence that genotype ikeda was the main pathogenic genotype present in south-eastern Australia [15]. The present study shows a similar finding for New Zealand. However, based on conventional molecular tools, McFadden et al. [5] showed that genotype chitose was associated with a varying severity of anaemia in dairy cattle infected with T. orientalis. The present study showed that MT PCR and PCR-HRM had very similar diagnostic specificities and sensitivities for the detection of genotypes buffeli, chitose, ikeda and type 5 of T. orientalis (Tables 5 and 6). In addition, MT and Ikeda TaqMan W qPCRs had similar diagnostic specificity and sensitivity for detecting genotype ikeda, and the results from MT PCR were comparable to the combined results obtained by both PCR-HRM and Ikeda TaqMan W qPCR (which are the molecular tools currently used in New Zealand; D. J. Pulford, and A. M. The diagnostic sensitivity of the MT PCR was defined based on a cut-off of ≥ 1 DNA copies. a PI, probability interval. b ρ+ is the conditional correlation between MT PCR and PCR-HRM outcomes for infected animals, and ρis the conditional correlation for uninfected animals. PABAK = prevalence-adjusted bias-adjusted Cohen's kappa (ĸ) statistic. pD = effective dimension (model complexity). DIC = Deviance information criterion, a generalisation of Akaike's Information Criterion (AIC).
J. McFadden, unpublished observations). The estimated cost per sample (including the cost for DNA extraction) by MT PCR is AUD 22, which is approximately twice that of PCR-HRM and Ikeda TaqMan W qPCR (i.e., AUD 10). MT PCR assay used here can test 12 samples per hour, whereas the time taken to test 12 samples by PCR-HRM is 1.5 h and an additional 1.5 h for Ikeda TaqMan W qPCR (i.e., a total of 3 hours). Importantly, MT PCR assay is performed in a robotic platform reducing the labour, contamination and PCR inhibition involved with many conventional and qPCR assays. In addition, the analytical MT PCR assay allows a qualitative (detection and differentiation) and semi-quantitative evaluation of four distinct genotypes of T. orientalis at once, which cannot be achieved using the other two PCRs.

Conclusion
In conclusion, MT PCR assay used here has some important advantages over other diagnostic techniques currently used within and outside of Australasia for the detection and differentiation of T. orientalis. Findings from the present study provide further support for the proposal that MT PCR assay could be used to predict the risk of clinical disease developing in cattle, based on DNA copy number thresholds of pathogenic and apathogenic genotypes. Although PCR-HRM and Ikeda TaqMan W qPCR have proven to be useful diagnostic tools in a recent epidemic of oriental theileriosis in New Zealand, MT PCR allows the simultaneous detection of four common genotypes (buffeli, chitose, ikeda and type 5) of T. orientalis complex and an estimation of the intensity of infection with these genotypes, thereby making it a useful tool for in-depth epidemiological and transmission studies.