INTRODUCTION
Bovine trichomonosis caused by Tritrichomonas foetus can have a significant impact in the reproductive performance of cattle (Ondrak, Reference Ondrak2016). The protozoan parasite T. foetus is sexually transmitted in cattle resulting in early embryonic deaths, abortion and infertility (BonDurant, Reference BonDurant1997; Ondrak, Reference Ondrak2016). The parasite is also naturally present as a commensal in pig caecum, stomach and nose (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012; Doi et al. Reference Doi, Abe and Oku2013; Mueller et al. Reference Mueller, Morin-Adeline, Gilchrist, Brown and Šlapeta2015). In domestic cats T. foetus causes feline enteric trichomonosis (Levy et al. Reference Levy, Gookin, Poore, Birkenheuer, Dykstra and Litaker2003). Comparative analysis of feline, porcine and bovine isolates of T. foetus led to the recognition of ‘bovine’ and ‘feline’ genotypes for isolates from cattle and pigs, and cats, respectively. The bovine isolates from Argentina, Czech Republic and Australia all shared identical sequences across ten DNA loci (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). No isolate from African cattle has been compared, despite wide distribution of T. foetus in African cattle (Pefanis et al. Reference Pefanis, Herr, Venter, Kruger, Queiroga and Amaral1988; Madoroba et al. Reference Madoroba, Gelaw, Hlokwe and Mnisi2011).
A recent retrospective study from six Southern African countries demonstrated wide prevalence of major reproductive infection in Southern African cattle (Madoroba et al. Reference Madoroba, Gelaw, Hlokwe and Mnisi2011). Tritrichomonas foetus was detected in 142 (4·1%, n = 3458) bull sheath washings and scrapings, and Campylobacter fetus was present in 1·9% (60/3161) samples (Madoroba et al. Reference Madoroba, Gelaw, Hlokwe and Mnisi2011). The study demonstrated T. foetus prevalence of 4·5, 3·8 and 3·3% in bulls from South Africa (90/11999), Namibia (45/1201) and Botswana (7/210), respectively (Madoroba et al. Reference Madoroba, Gelaw, Hlokwe and Mnisi2011). Tritrichomonas foetus was not detected in Swaziland (n = 7) or Zambia (n = 41), most likely due to the low number of samples investigated. Few studies exist about the prevalence of T. foetus in Southern Africa. Cattle farming remains important in Southern Africa with many regions where cattle is predominantly serviced naturally providing opportunities for venereal infections such as trichomonosis. Vaccine TrichGuard® (Zoetis, South Africa) and TrichGuard® V5L (Zoetis, South Africa) is marketed in Southern Africa for aid in prevention of T. foetus or in combination with C. fetus and three species of Leptospira, respectively. TrichGuard vaccine is a killed vaccine containing axenic trophozoites in adjuvant against T. foetus for cattle (Baltzell et al. Reference Baltzell, Newton and O'Connor2013).
The aim of this study was to culture and identify T. foetus from cattle in Namibia and compare their molecular identity to known genotypes of T. foetus. To do so, bull samples were collected and cultured. Established cultures were characterized using molecular diagnostic techniques and the genotype confirmed by multilocus genotyping.
MATERIALS AND METHODS
Sample collection
Between 2014 and 2015 about 4000 bulls present in herds from central regions of Namibia (Khomas, Hardap, Erongo, Omaheke and Otjozondjupa) were sampled once for T. foetus by collecting preputial samples using sheath washing technique (Irons et al. Reference Irons, Henton and Bertschinger2002). The same samples were inoculated into a Steve's transport medium (Steve's TM, Vrede Veterinary Laboratory, South Africa) and delivered to the Central Veterinary Laboratory of Windhoek (CVL) within 72 h after the sampling.
Culture of Tritrichomonas foetus
Samples in transport medium were examined directly under a standard light microscope (Olympus BX53 equipped with Olympus SC100 high-resolution digital colour camera) using a magnification of 1000× (100× objective and 10× eye piece) for the presence of motile protozoa with three flagella. Giemsa-stained smears were used to confirm trichomonad morphology. Prepared enriched ‘Trichomonas medium’ and Thioglycollate medium U.S.P. (Oxoid LTD, Basingstoke, Hampshire, England) were dispensed aseptically into sterile McCartney bottles in 15 mL aliquots and used as culture medium for all the samples collected for isolation of T. foetus with microscopic examination of the medium at intervals from day 1 to day 7 after inoculation and incubation at 37 °C. The results were recorded as positive when trichomonad organisms displaying unique morphological characteristics were present, or negative if there was no growth of trichomonads.
DNA isolation
DNA from 36 T. foetus positive cultures was extracted using the Maxwell®16 Cell LEV DNA Purification Kit (Promega, Madison, WI, USA) according to the manufacturer's instructions with an elution volume of 50 µL. DNA was shipped to the Faculty of Veterinary Sciences, The University of Sydney and stored at −20 °C prior molecular characterisation.
Molecular diagnostics for Tritrichomonas foetus
Diagnostic T. foetus real-time polymerase chain reaction (PCR) was based on primers TFR3 [S0001] (5′-CGG GTC TTC CTA TAT GAG ACA GAA CC-3′) and TFR4 [S0002] (5′-CCT GCC GTT GGA TCA GTT TCG TTA A-3′) amplify region of the internal transcribed spacer (ITS) rDNA (Felleisen et al. Reference Felleisen, Lambelet, Bachmann, Nicolet, Muller and Gottstein1998; Mueller et al. Reference Mueller, Morin-Adeline, Gilchrist, Brown and Šlapeta2015). The real-time PCR reactions used KAPA SYBR® FAST qPCR (2×) Master Mix (Kapa Biosystems, Inc., MA, USA) on CFX96 Touch™ Real-Time PCR Detection System with the corresponding CFX Manager 3·1 software (BioRad, Australia). The volumes of the real-time PCR reactions were made up to 20 µL, including 2 µL of template DNA (all samples were used at its original concentration or diluted 1:10). The PCR mix included primers at a final concentration of 200 nm. PCR conditions were 3 min at 95 °C, followed by 40 cycles of 3 s at 95 °C and 20 s at 63 °C as previously described by Mueller et al. (Reference Mueller, Morin-Adeline, Gilchrist, Brown and Šlapeta2015). ddH2O served as a negative control. Positive PCRs were directly sequenced by Macrogen Ltd. (Seoul, Korea; http://dna.macrogen.com/). Oligonucleotides were synthesised by Macrogen Inc. (Seoul, Korea).
Trichomonas foetus DNA Test Kit (VetMAX™-Gold Trich Detection Kit, 4483869, Life Technologies, Thermo Fisher, Australia), a test based on TaqMan probe qPCR, was used according to the manufacturer's instruction on CFX96. The amplification on CFX96 and data were analysed with the corresponding software (BioRad, Australia). The real-time PCR threshold was arbitrarily set to a single threshold at 100 rfu. Data (C t -values) were interpreted according to the manufacturer's instructions (Life Technologies, Thermo Fisher, Australia).
Conventional trichomonad PCR amplifying ITS rDNA region based on primers TFR1 [S0062] (5′-TGC TTC AGT TCA GCG GGT CTT CC-3′) and TFR2 [S0063] (5′-CGG TAG GTG AAC CTG CCG TTG G-3′) as previously described using MyTaq™ Red Mix (BioLine, Australia) (Felleisen, Reference Felleisen1997; Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012).
Multilocus genotying of Tritrichomonas foetus
Protein coding genes, cysteine proteases (CP1, 2, 4–9) and cytosolic malate dehydrogenase 1 (MDH1) were amplified from T. foetus DNA as previously described (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012; Sun et al. Reference Sun, Stack and Šlapeta2012). Oligonucleotides were synthesised by Macrogen Inc. (Seoul, Korea). All PCR amplifications were performed with MyTaq™ Red Mix (BioLine, Australia) in a total volume of 30 µL. Primers were added at a concentration of 0·25 µ m each. The PCR was run using the following cycling conditions: 95 °C for 15 s, 55 °C for 15 s and 72 °C for 20 s for 35 cycles. All reactions were initiated at 95 °C for 2 min and concluded at 72 °C for 7 min. PCRs were amplified in the Verity PCR cycler (Thermo Fisher Scientific, Australia). Each PCR mix contained 2 µL of the sample DNA. All PCRs were run with negative controls (ddH2O). All PCRs that yielded unambiguous single bands of the expected size were directly and bidirectionally sequenced using amplification primers at Macrogen Ltd. (Seoul, Korea).
DNA sequences were verified, assembled and multiple sequence alignments with T. foetus and Tritrichomonas mobilensis were produced in CLC Main Workbench 6·8·1 (CLC bio, a QIAGEN Company, Denmark; http://www.clcbio.com/). Individual gene alignments were concatenated in CLC Main Workbench. Split networks from distances (K2P) were calculated using NeighborNet and cluster network visualized in SplitsTrees 4 (v 4·14·3; http://www.splitstree.org/) (Huson and Bryant, Reference Huson and Bryant2006).
Sequence data and data accessibility
Sequences were assembled, aligned with related sequences and analysed using CLC Main Workbench 6·8·1 (CLC bio, Denmark) and deposited in GenBank (National Center for Biotechnology Information, NCBI) under the Accession Numbers: KX425856-KX425919.
RESULTS
Thirty-six cultures contained trichomonads morphologically consistent with T. foetus. All but two (94%, 34/36) tested positive with an average C t -value of 21·31 (min. 14·53, max. 29·83) using T. foetus specific qPCR based on TFR3/4 primers targeting ITS rDNA (Table 1). The C t -value was not returned for NAM-1 and NAM-19. A diluted (1:10) DNA template confirmed presence of T. foetus DNA in 97% (35/36) samples (C t -value mean 25·02, min. 18·11, max. 34·85), including NAM-1 sample (C t -value = 30·28). Sample NAM-19 did not return C t -value in both runs and was considered T. foetus negative. Furthermore, NAM-19 remained negative using trichomonad universal primer set TFR1/2.
Summary of molecular diagnostics for Tritrichomonas foetus isolates
Duplicate USDA-licensed diagnostic T. foetus specific qPCR confirmed presence of T. foetus DNA in 35 samples (C t -values: average 25·41, min. 20·10, max. 30·96) using DNA diluted 1:10 (Table 1). Sample NAM-19 did not return C t -values.
PCR products from qPCR based on TFR3/4 primers amplifying ITS rDNA were subjected to DNA sequencing to further identify the genotype of T. foetus (Table 1). Previously, ITS rDNA single polymorphism has been shown diagnostic between the ‘bovine’ and ‘feline’ genotype of T. foetus (Šlapeta et al. Reference Šlapeta, Craig, McDonell and Emery2010, Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). All 38 DNA sequences of ITS rDNA revealed a novel genotype with an AA insertion (adenosines) in ITS1 rDNA compared with the ‘bovine’ or ‘feline’ genotype of T. foetus (Fig. 1). The ITS2 rDNA for all 35 DNA sequences matched the ‘bovine’ genotype of T. foetus with C (cytosine) at the diagnostic residue (Šlapeta et al. Reference Šlapeta, Craig, McDonell and Emery2010) (Fig. 1).
Comparison between Tritrichomonas foetus genotypes and Tritrichomonas mobilensis. (A) Multiple sequence alignment of internal transcribed spacer region of ribosomal RNA gene of T. foetus and T. mobilensis. Identical residues presented as dots. Differences highlighted. (B) Multiple sequence alignment at cysteine protease 2 (CP2) of T. foetus and T. mobilensis. (C) Pairwise comparison of CP2 nucleotide sequences (upper) and amino acid sequences (lower). The top right represents pairwise similarity and bottom left presents number of differences. (D) Split cluster network analysis (K2P model) calculated using NeighborNet using all 10 loci and 4554 alignment positions in the final dataset.
The presence of the novel T. foetus genotype prompted the characterization of cysteine proteinase 2 (CP2) locus of T. foetus previously recognized as the most divergent (22 nt differences) between the known ‘feline’ and ‘bovine’ genotypes (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012; Suzuki et al. Reference Suzuki, Kobayashi, Osuka, Kawahata, Oishi, Sekiguchi, Hamada and Iwata2016). Amplification of CP2 was successful for 36% (13/35) DNA samples that were considered positive for T. foetus using the qPCR assays. NAM-2, NAM-3, 5, 30, 34, 35, 13, 20, 21, 26, 27, 28 and 32 samples amplified CP2 and were those with lower C t values in diagnostic PCR (Table 1). All obtained 669 nt long CP2 sequences were identical to each other and distinct by 15 nt from the ‘bovine’ T. foetus genotype coding for 9 amino acid differences (Fig. 1; Table 2). The remaining DNA samples produced insufficient PCR product for direct sequencing.
Multilocus comparison of Tritrichomonas foetus ‘Southern African’ genotype with T. foetus and T. mobilensis
CP, Tritrichomonas sp. cysteine protease; nt, nucleotide; MDH, malate dehydrogenase; ITS, internal transcribed spacer.
a Polymerase chain reaction (PCR) amplified and sequenced DNA without PCR primers.
Multilocus genotyping based on protein coding genes confirmed unique status of the studied isolates (Table 2). Eight protein coding genes (CP1, CP2, CP4, CP5, CP6, CP7, CP8 and MDH1) and ITS locus have been amplified and sequenced for two isolates NAM-27 and NAM-30. Comparison with T. foetus ‘bovine’ genotype and T. foetus ‘feline’ genotype, the loci were equally (97·8–99·8 and 97·8–99·8%, respectively) identical at the nucleotide sequence with the studied isolates. The least identical (97·8%) locus was CP2 and CP6 between T. foetus ‘bovine’ genotype and studied isolates (Table 2). Amplification of CP8 from NAM-27 yielded an insufficient PCR product for direct sequencing, therefore primers F1/R2 (Sun et al. Reference Sun, Stack and Šlapeta2012) producing a shorter product were successfully applied.
The closest sister species to T. foetus is T. mobilensis (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). Six nucleotide sequences for protein coding genes and ITS are available for T. mobilensis spaning over 2472 residues (Table 2). The novel T. foetus genotype has the least variable sites (1%, 24/2472) with T. mobilensis compared with 25 and 33 with T. foetus ‘feline’ and ‘bovine’ genotype, respectively. Inferring the phylogenetic relationship between the T. foetus genotypes and T. mobilensis using split networks revealed splits leading to T. foetus (Fig. 1).
DISCUSSION
This study revealed the presence of a novel genotype of T. foetus in Southern African cattle. All previously characterized T. foetus from cattle (Europe, America, Australia) were identical at the ITS marker and/or using multilocus approach (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). The Southern African T. foetus isolates form a new genetic lineage, the T. foetus ‘Southern African’ genotype. The T. foetus ‘Southern African’ genotype is detectable using either the USDA-licensed diagnostic qPCR or commonly used PCR/qPCR based on the TFR3/4 primer pair (Felleisen et al. Reference Felleisen, Lambelet, Bachmann, Nicolet, Muller and Gottstein1998).
The T. foetus ‘Southern African’ genotype is closely related to T. foetus ‘bovine’ genotype from cattle and pigs, T. foetus ‘feline’ genotype from cats, and T. mobilensis. Experimental infections with the non-homologous genotype (cattle with the T. foetus ‘feline’ genotype and vice versa) are possible under laboratory conditions, but they do not result in the same disease outcome (Stockdale et al. Reference Stockdale, Rodning, Givens, Carpenter, Lenz, Spencer, Dykstra, Lindsay and Blagburn2007, Reference Stockdale, Dillon, Newton, Bird, Bondurant, Deinnocentes, Barney, Bulter, Land, Spencer, Lindsay and Blagburn2008). Across six protein coding gene loci (2199 nt), the T. foetus ‘Southern African’ genotype is most closely related to T. mobilensis with only 20 nt (<1%) single nucleotide polymorphisms. These changes only code for nine amino acid changes in CP2 and a single amino acid change in MDH1. The genetic distance is comparable with the distance of the T. foetus ‘feline’ genotype to T. mobilensis (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). It is therefore worthwhile to consider if the pathobiology of the T. foetus ‘Southern African’ genotype is distinct to that of those studied previously (Stockdale et al. Reference Stockdale, Rodning, Givens, Carpenter, Lenz, Spencer, Dykstra, Lindsay and Blagburn2007, Reference Stockdale, Dillon, Newton, Bird, Bondurant, Deinnocentes, Barney, Bulter, Land, Spencer, Lindsay and Blagburn2008). In vitro culture studies demonstrate functional differences between isolates and could be used to compare the T. foetus ‘bovine’ genotype with the ‘Southern African’ genotype (Tolbert et al. Reference Tolbert, Stauffer and Gookin2013, Reference Tolbert, Stauffer, Brand and Gookin2014).
It was previously thought that the T. foetus ‘bovine’ and ‘feline’ genotypes strictly isolated from cattle and cats, respectively, was a strong indicator that they possess distinct epidemiology and evolutionary trajectories (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). This also implied a single origin of the bovine T. foetus that is now challenged with the discovery of the T. foetus ‘Southern African’ genotype. Currently, it is impossible to unambiguously resolve the origin of bovine isolates, despite using multiple genes and assuming T. mobilensis as the sister to all T. foetus. Recent studies of transcriptomes of feline, bovine and porcine T. foetus using RNA-seq have demonstrated the opportunities of next-generation sequencing to better understand the biology and identity of closely relating microorganism (Morin-Adeline et al. Reference Morin-Adeline, Lomas, O'Meally, Stack, Conesa and Šlapeta2014, Reference Morin-Adeline, Mueller, Conesa and Šlapeta2015).
Over the past 6 years T. foetus was characterized from several countries using molecular techniques that unambiguously identify the strain and genotype (Šlapeta et al. Reference Šlapeta, Craig, McDonell and Emery2010; Doi et al. Reference Doi, Abe and Oku2013; Mueller et al. Reference Mueller, Morin-Adeline, Gilchrist, Brown and Šlapeta2015; Arranz-Solís et al. Reference Arranz-Solís, Pedraza-Díaz, Miró, Rojo-Montejo, Hernández, Ortega-Mora and Collantes-Fernández2016; Suzuki et al. Reference Suzuki, Kobayashi, Osuka, Kawahata, Oishi, Sekiguchi, Hamada and Iwata2016). To date, only two main genotypes and close allies have been identified (Šlapeta et al. Reference Šlapeta, Müller, Stack, Walker, Lew-Tabor, Tachezy and Frey2012). Identification of this new genotype of T. foetus demonstrates the need for wider global sampling to fully understand the diversity and origin of T. foetus that causes disease in cattle or cats. The genetic diversity of T. foetus has potential implication for efficacy of the killed, whole-cell T. foetus vaccine (TrichGuard®) if a different T. foetus genotype is used in the vaccine than the one circulating in cattle to be vaccinated. Recent systematic review of the efficacy of the T. foetus vaccine identified potential over-estimate of efficacy and limited or no evidence that vaccination decreases infections or open risk in heifers (Baltzell et al. Reference Baltzell, Newton and O'Connor2013).
The outcome of this study is the discovery of a yet unknown ‘Southern African’ genotype of T. foetus in Namibian cattle, distinct from the T. foetus ‘bovine’ genotype from Europe, South and North America and Australia. Identification of the new genotype of T. foetus demonstrates the need for wider global sampling to fully understand the diversity and origin of T. foetus causing disease in cattle or cats.
FINANCIAL SUPPORT
This study was funded in part through the Faculty of Veterinary Science, University of Sydney diagnostic laboratory (to A. C. and J. Š.) and the Directorate of Veterinary Services, Ministry of Agriculture, Water and Forestry of Namibia (to U. M., K. K. and S. K.).
