Comparison of antibody- and antigen-detection enzyme immunoassays for the diagnosis of Trypanosoma evansi infections in camels

Recent Progress in the Detection of Surra, a Neglected Disease Caused by Trypanosoma evansi with a One Health Impact in Large Parts of the Tropic and Sub-Tropic World

Surra is a wasting disease triggered by infection with Trypanosoma evansi, a protozoan blood parasite that causes mortality and morbidity in a broad spectrum of wild and domestic animals and occasionally humans. Trypanosoma evansi has the widest geographical spread among all pathogenic trypanosomes, inflicting significant worldwide economic problems due to its adverse effects on meat and milk production. For diagnosis, most endemic countries continue to rely on traditional parasitological and serological techniques, such as the analysis of blood smears by microscopy and the Card Agglutination Test for T. evansi (CATT/T. evansi). Although these techniques suffer from a limited positive predictive value (PPV), resource constraints in endemic countries often hinder the adoption of more advanced diagnostic tools such as PCR. This paper addresses diverse diagnostic approaches for identifying T. evansi and assesses their viability in field settings. Moreover, it underscores the urgency of transitioning towards molecular diagnostic techniques such as Loop-Mediated Isothermal Amplification (LAMP) and Recombinase Polymerase Amplification (RPA) for dependable high-PPV point-of-care (POC) diagnostics. Finally, this review delves into strategies to enhance and refine next-generation diagnostics for Surra as part of a One Health approach.

An Unbiased Immunization Strategy Results in the Identification of Enolase as a Potential Marker for Nanobody-Based Detection of Trypanosoma evansi

Trypanosoma evansi is a widely spread parasite that causes the debilitating disease “surra” in several types of ungulates. This severely challenges livestock rearing and heavily weighs on the socio-economic development in the affected areas, which include countries on five continents. Active case finding requires a sensitive and specific diagnostic test. In this paper, we describe the application of an unbiased immunization strategy to identify potential biomarkers for Nanobody (Nb)-based detection of T. evansi infections. Alpaca immunization with soluble lysates from different T. evansi strains followed by panning against T. evansi secretome resulted in the selection of a single Nb (Nb11). By combining Nb11-mediated immuno-capturing with mass spectrometry, the T. evansi target antigen was identified as the glycolytic enzyme enolase. Four additional anti-enolase binders were subsequently generated by immunizing another alpaca with the recombinant target enzyme. Together with Nb11, these binders were evaluated for their potential use in a heterologous sandwich detection format. Three Nb pairs were identified as candidates for the further development of an antigen-based assay for Nb-mediated diagnosis of T. evansi infection.

Systematic review and meta-analysis on the global distribution, host range, and prevalence of Trypanosoma evansi

BACKGROUND

Surra is an animal trypanosomosis, caused by infection with Trypanosoma evansi and leading to severe economic loss due to mortality and morbidity. Compared to tsetse-transmitted animal trypanosomoses, little attention is given to the epidemiology and control of surra. Understanding its epidemiology is a first step in local and global efforts to control the disease. We conducted a systematic review and meta-analysis of published studies on distribution, host ranges and prevalence of T. evansi infection.

METHODS

Four electronic databases were searched for publications on T. evansi that met our inclusion criteria for the systematic review. Subsets of publications were subjected to meta-analysis for the pooled prevalence of T. evansi in various hosts as determined by multiple detection methods.

RESULTS

A total of 272 references published between 1906-2017 were included. Trypanosoma evansi was reported from 48 countries; largely confined to Africa and Asia with publications on natural T. evansi infections from 77% (n = 48) of countries, contrasting with seven countries in South America, and four in Europe where T. evansi is not endemic but was imported with infected animals. Although surra is a notifiable disease, many countries do not report surra cases to OIE. Trypanosoma evansi was mainly reported from dromedary camels in Africa and the Middle East, water buffaloes, cattle, dogs and horses in East and Southeast Asia. In South America, the acute form of the disease was reported in horses and dogs. Surra was also reported in a wide range of wild animals. Some rare human cases occurred in India and Vietnam. Meta-analysis on a subset of 165 publications indicated pooled prevalence of T. evansi in domestic animals ranging from 14-31%, 6-28% and 2-9% using respectively antibody detection, molecular and parasitological tests, with camels as the most affected, followed by buffalo and cattle.

CONCLUSIONS

This study illustrates that T. evansi affects a wide range of domestic and wild animals in Africa, Asia and South America with highest prevalence observed in dromedary camels. For successful control of T. evansi, both locally and globally, the role of wild animals in the epidemiology of surra needs further investigation.

A study on prevalence and molecular characterization of trypanosomal species infecting equines in Lahore region, Pakistan

Trypanosomiasis is an important protozoal disease with a diverse range of susceptible host including human. In the current study, molecular characterization of prevalent species was done through a pan-trypanosome polymerase chain reaction (PCR) followed by restriction fragment length polymorphism (RFLP). A total of three hundred (n = 300) equines including horses, donkeys and mules (100 each) were randomly selected and the equine blood samples were subjected to screening for trypanosomes through microhaematocrit centrifuge technique (MHCT), conventional PCR, semi-nested PCR and RFLP. Overall prevalence of trypanosomal species was 8% (24/300) as revealed by MHCT and species wise prevalence in horses, donkeys and mules was 4.33% (13/300), 1.33% (4/300) and 2.33% (7/300), respectively. Conventional and semi-nested PCR depicted an overall prevalence of 21% (63/300) and species wise prevalence in horses, donkeys and mules was 12% (36/300), 3.67% (11/300) and 5.33% (16/300), respectively. RFLP analysis of the semi-nested products, using Msp1 and Eco571 enzymes, negated the presence of T. congolense, T. brucei, T. vivax, T. theileri, and T. vivax in the positive samples and revealed that the animals might be suffering from T. evansi infection as the enzymes used were not able to detect this species. This hypothesis was further confirmed by using T. evansi specific primers which depicted all of the 63 samples were positive for T. evansi. It is inferred that T. evansi is the major trypanosome species prevalent in equines. Furthermore, PCR is more sensitive as compared to microscopic examination and the pan-trypanosome PCR-RFLP assay is suitable for carrying out laboratory diagnosis of field samples and epidemiological studies. Further studies on the possibilities of use of other restriction enzymes may help to improve the species specificity of the assay.

Generation of a nanobody targeting the paraflagellar rod protein of trypanosomes

Trypanosomes are protozoan parasites that cause diseases in humans and livestock for which no vaccines are available. Disease eradication requires sensitive diagnostic tools and efficient treatment strategies. Immunodiagnostics based on antigen detection are preferable to antibody detection because the latter cannot differentiate between active infection and cure. Classical monoclonal antibodies are inaccessible to cryptic epitopes (based on their size-150 kDa), costly to produce and require cold chain maintenance, a condition that is difficult to achieve in trypanosomiasis endemic regions, which are mostly rural. Nanobodies are recombinant, heat-stable, small-sized (15 kDa), antigen-specific, single-domain, variable fragments derived from heavy chain-only antibodies in camelids. Because of numerous advantages over classical antibodies, we investigated the use of nanobodies for the targeting of trypanosome-specific antigens and diagnostic potential. An alpaca was immunized using lysates of Trypanosoma evansi. Using phage display and bio-panning techniques, a cross-reactive nanobody (Nb392) targeting all trypanosome species and isolates tested was selected. Imunoblotting, immunofluorescence microscopy, immunoprecipitation and mass spectrometry assays were combined to identify the target recognized. Nb392 targets paraflagellar rod protein (PFR1) of T. evansi, T. brucei, T. congolense and T. vivax. Two different RNAi mutants with defective PFR assembly (PFR2^RNAi and KIF9B^RNAi) were used to confirm its specificity. In conclusion, using a complex protein mixture for alpaca immunization, we generated a highly specific nanobody (Nb392) that targets a conserved trypanosome protein, i.e., PFR1 in the flagella of trypanosomes. Nb392 is an excellent marker for the PFR and can be useful in the diagnosis of trypanosomiasis. In addition, as demonstrated, Nb392 can be a useful research or PFR protein isolation tool.

Trypanosoma evansi: a comparative study of four diagnostic techniques for trypanosomosis using rabbit as an experimental model

The goal of this study was to compare two parasitological diagnostic techniques, such as by Micro-Haematocrit Centrifugation Technique (MHCT) and Direct Microscopic Examination (DME) with a serological method (iELISA), and a molecular procedure PCR, in rabbits experimentally infected with Trypanosoma evansi, in order to determine their sensitivity throughout the course of disease. The parasitological methods were not able of detecting the presence of the parasite during the phases of low parasitemia, the prepatency period and the chronic phase. In contrast, PCR detected T. evansi in the prepatency and chronic phase, when increase the amount of DNA from 100 to 300ng. 100% detection was observed with iELISA only in the chronic stage of the disease. In the acute phase, all samples were positively diagnosed using either MHCT or PCR, whereas only few samples were diagnosed by DME. Samples obtained from day 15 post infection were also detected by iELISA. The highest diagnostic register during the course of infection was achieved by the PCR technique (93.8%), followed by iELISA (71.1%), MHCT (59%) and DME (13.6%). Therefore, we recommend the use of PCR in epidemiological studies in order to implement sanitary control plans for the improvement of livestock productivity in the country.

Development of a TaqMan PCR assay for the detection of Trypanosoma evansi, the agent of surra

A TaqMan PCR assay was developed for the detection of Trypanosoma evansi. The assay targets the internal transcribed spacer 1 (ITS-1) region of rRNA. The ITS-1 region of eleven strains of T. evansi from widely separated geographical regions were sequenced and alignments compared. Primers and probe for the test were designed from these sequence data. The assay was tested using blood from infected rats and was found to be sensitive, detecting less than one genomic equivalent of T. evansi. The assay has been tested against 10 different species of trypanosomes found in native animals in Australia and did not detect any of these trypanosome species. Time course experiments using rats infected with T. evansi were performed to compare the TaqMan assay with the Haematocrit centrifugation test (HCT) and the mouse inoculation (MI) assay. The assay was more sensitive than the HCT but not as sensitive as the MI. The TaqMan assay has the ability to rapidly detect T. evansi and determine the number of organisms present in a blood sample from an infected animal. This is the first time a TaqMan assay has been developed for the detection of T. evansi.

A comparative evaluation of parasitological, serological and DNA amplification methods for diagnosis of natural Trypanosoma evansi infection in camels

A representative number of 217 camels (Camelus dromedarius) from different areas of western Rajasthan State, India, were examined from July 2002 to May 2003 for Trypanosoma evansi infection. The tests used were parasitological (wet blood film, WBF; stained thin blood smear, TBS), immunodiagnostic (double antibody sandwich enzyme linked immunosorbent assay for antigen detection, Ag-ELISA), and DNA amplification by polymerase chain reaction (PCR). These techniques were compared and the best efficiency was found for the last named (PCR). A prevalence of T. evansi infection was detected in 17.05, 9.67, 4.60 and 4.14% by PCR, Ag-ELISA, TBS and WBF with a sensitivity of 100, 56.75, 27.02 and 24.32%, respectively. PCR revealed a specific 227bp band in positive samples. The intensity of PCR bands was variable in different test samples depending upon the level of infection in the test samples. The history of intermittent fever, emaciation, oedema, poor body condition significantly correlated with positive serological status in ELISA as well as trypanosome DNA detection by PCR.

Detection of Trypanosoma evansi in camels using PCR and CATT/T. evansi tests in Kenya

Camel trypanosomosis (Surra) causes high morbidity and is an impediment to the camel husbandry in Kenya. The lack of a sensitive diagnostic test has hindered the collection of accurate epidemiological data and institution of control programmes. A cross-sectional study was conducted in three districts of Kenya to estimate the prevalence of Trypanosoma evansi (T. evansi) and to compare four diagnostic tests: polymerase chain reaction (PCR), card agglutination test (CATT/T. evansi), microhaematocrit centrifugation technique (MHCT) and mouse inoculation (MI). A total of 549 camels were randomly sampled. The overall prevalence of Surra was 5.3% using MHCT, 26.6% using PCR and 45.9% using CATT/T.evansi. There was a significant difference (P < 0.001) between PCR and CATT/T.evansi test, MHCT and MI in detection of T. evansi. The prevalence of T. evansi was 39.8% in Samburu, 24.7% in Nanyuki and 14.4% in Isiolo districts using PCR. A male camel was 2.6 times more likely to be infected with T. evansi compared to a female camel (OR = 3.0% CI: 1.6, 4.1), while an adult camel was 2.2 times more likely to be infected compared to non-adults (OR = 2.2; 95% CI: 1.2, 5.0). There was a poor association between the presence of the published clinical signs and seropositivity (kappa = 0.12), PCR (kappa = 0.11) and MHCT (kappa = 0.05). However, there was a higher agreement between farmers' classification of disease with the PCR test (kappa = 0.5, n = 61). The mean PCV varied with age, presence of infection, locality and gender, with the lowest mean PCV being recorded in MHCT-positive animals (20.97 +/- 0.5) and from infected calves (19.5 +/- 1.2). This study shows that PCR was more sensitive in detecting T. evansi than other tests used. Further, the prevalence of T. evansi in the camel herds sampled is higher than that previously reported in Kenya, and that the judgment by camel keepers may be a reliable "pen-side" diagnostic test for Surra. Considering the low sensitivity of parasitological techniques in detection of chronic T. evansi infection and high cost of PCR, development of a sensitive pen side diagnostic test, with a low cost is still a priority.

Evaluation of antigen and antibody rapid detection tests for Trypanosoma evansi infection in camels in Kenya

The card agglutination test for Trypanosoma evansi (CATT/T. evansi) for the detection of antibodies, and Suratex for the detection of circulating antigens were compared in a cross-sectional study involving camels in eastern and central parts of Kenya. Of the 2227 camels screened, 2038 were owned by nomadic pastoralists in T. evansi endemic areas in eastern Kenya. A herd of 86 camels were from a ranch in Mugwoni. In Athi River area, 35 camels belonged to Kenya Trypanosomiasis Research Institute, and 68 were slaughter animals. Diagnostic sensitivity estimates were obtained by testing sera from 51 camels that had been found to be parasitologically positive by the haematocrit centrifugation technique, buffy-coat technique and mouse inoculation. Diagnostic specificity was estimated by testing sera from 35 camels known to be trypanosome-free. Positive and negative predictive values (NPVs) were calculated using a range of prevalence values. The sensitivity of CATT/T. evansi (68.6%) was higher than that of Suratex (58.8%), but not significantly. Both tests had equally high specificity (100%). The overall prevalence was 2.3% (51 out of 2227) by parasite detection, 32.2% (327 out of 1017) by CATT/T. evansi and 19.6% (188 out of 961) by Suratex. Overall, there was a positive association between CATT/T. evansi and Suratex though the strength of association was low (McNemar's test=46.12, P=0.001; kappa=0.26, CI: 0.20-0.33). Parasite prevalence ranged from 0% in several herds to 27.8% in a herd in Isiolo. Prevalence was highest in Isiolo with 2.5% (51 out of 2030) by parasitological detection, 38.8% (321 out of 828) by CATT/T. evansi and 21.9% (169 out of 772) by Suratex. In Mugwoni prevalence was 7 and 18% by CATT/T. evansi and Suratex, respectively, and no parasites were detected. In Athi River Suratex detected 2.9% (3 out of 103) positive while CATT/T. evansi and parasitological methods gave negative results. At prevalence values between 10 and 100%, CATT/T. evansi as well as Suratex had infinitely high positive predictive values, whereas Suratex had a lower NPV than CATT/T. evansi. In conclusion, results of this study showed that CATT/T. evansi and Suratex were able to detect aparasitaemic infections rapidly and were more sensitive than parasitological methods in revealing the true extent of trypanosomosis in a herd. The tests effectively complemented parasitological methods in the detection of T. evansi infections in camels.

Genetic relatedness among Trypanosoma evansi stocks by random amplification of polymorphic DNA and evaluation of a synapomorphic DNA fragment for species-specific diagnosis

In this study we employed randomly amplified polymorphic DNA patterns to assess the genetic relatedness among 14 Brazilian Trypanosoma evansi stocks from domestic and wild hosts, which are known to differ in biological characteristics. These akinetoplastic stocks were compared with one another, to three Old World (Ethiopia, China and Philippines) dyskinetoplastic stocks of T. evansi, and also with Trypanosoma equiperdum, Trypanosoma brucei brucei, Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense. Randomly amplified polymorphic DNA analysis showed limited heterogeneity in T. evansi stocks from different hosts and geographical regions of the world, or in other species of the subgenus Trypanozoon. However, minor variations generated random amplification of polymorphic DNA analysis disclosed a pattern consisting of a unique synapomorphic DNA fragment (termed Te664) for the T. evansi cluster that was not detected in any other trypanosome species investigated. Pulsed field gel electrophoresis analysis demonstrated that the Te664 fragment is a repetitive sequence, dispersed in intermediate and minichromosomes of T. evansi. Based on this sequence, we developed a conventional PCR assay for the detection of T. evansi using crude preparations of blood collected either on glass slides or on filter paper as template DNA. Our results showed that this assay may be useful as a diagnostic tool for field-epidemiological studies of T. evansi.

Preliminary studies by ELISA on the antigen and antibody dynamics in the early stages of experimental infections with Trypanosoma evansi in cattle

Six 6-month-old bulls were experimentally infected with five different isolates of Trypanosoma evansi; two received the same isolate and the other four received different isolates. The parasitaemias and serum antigen levels were monitored regularly by the haematocrit centrifuge technique (HCT) and antigen-detection ELISA (Ag-ELISA), respectively. Trypanosomal antigen was demonstrated by the Ag-ELISA by 10-14 days post inoculation in four cattle, while parasitaemias were first found to be positive in individual cattle over a longer period of time post inoculation (6-28 days). In two cattle, the Ag-ELISA values were also positive when the animals were found to harbour trypanosomes by the HCT and only turned negative 3 days after treatment, while the ELISA values fluctuated during the experiment in another two bulls. The remaining two cattle never produced positive ELISA results despite positive parasitological results. The antibody titres in all six cattle started to rise around 10 days post inoculation and then stayed high throughout the experiment. It was concluded that the Ag-ELISA would produce some false negative results in the early stages of T. evansi infection owing to variations in the balance of parasitaemia and antibody levels in the circulation, and in the pathogenicity of parasite strains.

A preliminary comparative study of a dipstick colloidal dye immunoassay and two antigen-detection ELISAs for diagnosis of Trypanosoma evansi infection in cattle

A dipstick colloidal dye immunoassay (DIA) was developed for the field diagnosis of Trypanosoma evansi infection using affinity-purified polyclonal antibodies (PcAbs) and the monoclonal antibody (McAb) 8B9. PcAbs were adsorbed onto Palanil Red dye particles and used as dye reagents. Dipsticks were dotted with four different antibodies; normal rabbit and mouse IgGs as negative controls, and anti-T. evansi PcAb and McAb 8B9, which capture trypanosome antigens in the tested samples. Since the dye reagent bound to the captured antigens, the presence of coloured dots on the dipstick identified trypanosome infections. The sensitivity of the DIA was compared with two antigen detection ELISAs (Ag-ELISA); one was PcAb-based and the other was based on a combination of the same Mc- and PcAbs as were employed for the DIA. With a positive serum, the DIA detected trypanosomal antigen up to a dilution of 1:500 for both the PcAb and McAb dots, at which dilution the PcAb- and combination-based Ag-ELISA gave positive OD readings of 0.13 and 0.36, respectively. When 124 field sera were tested, circulating antigens were detected in 51 (41%) samples by the DIA, and 76 (61%) and 49 (40%) samples by the PcAb- and combination-based Ag-ELISAs respectively, of which 48 (63%) and 34 (69%) were also positive by the DIA.

Incidence, biochemical and haematological alterations induced by natural trypanosomosis in racing dromedary camels

In this study 10.67% incidence of trypanosomosis was observed in 150 racing camels by Suratex (latex agglutination). A significant decrease (P<0.05) in RBC, Hb, PCV and lymphocyte was observed, while a significant (P<0.05) increase in WBC and neutrophils was noted in trypanosomosis positive samples. The blood chemistry parameters indicated that there was significant (P<0.05) decrease in iron and albumin, but no significant alteration was observed in alkaline phosphatase (ALP), alanine aminotransferase (ALT), asparate aminotransferase (AST), gamma glutamyltransferase (GGT), lactic dehydrogenase (LDH), urea, total protein, calcium, creatinine, phosphorous and magnesium.

Seroprevalence of Trypanosoma evansi in dromedaries (Camelus dromedarius) from the Canary Islands (Spain) using an antibody Ab-ELISA

After the description in Grand Canary Island of a case of dromedary trypanosomosis caused by Trypanosoma evansi in 1998, an indirect enzyme immunoassay for the detection of specific anti-T. evansi IgG (Ab-ELISA) was used to assess the seroprevalence of this disease on the Canary Islands. Seroprevalence was 9.0% in the four studied islands (Gran Canaria, Tenerife, Lanzarote and Fuerteventura), varying from 10.0 to 7.5% by island (not significantly different). Prevalence using Ab-ELISA was higher than that observed when a parasitological method (microscopic observation of blood smears) was used (1.3%).

Haemoparasite infections in newly introduced dairy cattle in Loei Province, Thailand: Trypanosoma evansi antigen levels by ELISA referring to abortion

An antigen-detection enzyme-linked immunosorbent assay (Ag-ELISA) based on affinity-purified polyclonal antibody was utilised as an immunodiagnosis for Trypanosoma evansi infections in cattle. Five hundred pregnant heifers were introduced in Loei province, Thailand and a total of 271 samples were collected at 37 farms during four visits over a year commencing a month after the introduction. Each time the haematocrit centrifuge technique (HCT) was carried out as a field diagnosis for T. evansi, and sera were examined for trypanosomal antigen levels by Ag-ELISA. At the first sampling over 80% of the cattle were positive for trypanosome antigens by Ag-ELISA although the titres were low. Soon after, aborted cases at the late stage of pregnancy were reported and at the second sampling in the rainy season, 25.5% of the cattle sampled were found to harbour T. evansi by HCT. This time the infection rate by Ag-ELISA was 52.9% with high antigen levels. Between the first and second sampling nine cattle out of 51 aborted, which was suspected to be due to T. evansi. As soon as treatment with a trypanocidal drug was started, abortion cases decreased. However, the infection rate remained high during the rainy season when Tabanus flies were active. As the climate became cool and dry, the antigen levels in the area lowered and the positive rate by Ag-ELISA dropped to 32.3%.

Enzyme-linked immunosorbent assay (ELISA) for detection of anti-Trypanosoma evansi equine antibodies

The standardization of ELISA for the detection of anti-Trypanosoma evansi antibodies in naturally and experimentally infected horses is described. Bayesian analysis was used to establish the cutoff between positive and negative sera. In order to determine the assessment of the ELISA test, the results obtained were compared with those from an IFA. A relative sensibility of 98.39%, a specificity of 95.12% and a predictive value of 96.83% were determined. The standardized technique was used to evaluate the antibody production against trypanosome in an experimentally infected equine, in which the sera converted 15 days after infection. The test was also used for a study of sera prevalence in a non-random sample from two different populations. A prevalence of 81.7% in workhorse and 57.14% in stable horses was found.

Effect of anti-immunoglobulin antibodies produced in cattle infected with Trypanosoma evansi on antigen detection ELISA

The possibility of interference with the antigen-detection ELISA for trypanosomosis by anti-rodent IgG antibodies produced in cattle infected with Trypanosoma evansi was investigated. Two different ELISA for detection of trypanosome antigen and three different systems for anti-rodent IgG antibody detection were established. The former two were respectively polyclonal antibody-based and a combination of monoclonal and polyclonal antibody-based assays. The latter three were also adapted for detection of anti-mouse IgG, anti-rabbit IgG and anti-IgG antibodies cross-reactive with both rabbit and mouse IgGs. A total of 170 samples were collected from a dairy cattle farm where an outbreak of T. evansi infection was reported. One hundred and two cattle (59%) were found to be positive for trypanosome antigens by the polyclonal antibody-based assay and 86 (51%) were positive by the combination-based system. On the other hand, 51 (30%) and 10 (6%) of cattle had anti-rabbit and anti-mouse IgG antibodies respectively but none had antibodies cross-reactive with both IgGs. Of the 102 cattle positive for trypanosome antigens in the polyclonal antibody-based ELISA, 48 (47%) were also anti-rabbit IgG antibody positive. It is concluded that antigen detection ELISA based on a single-species immunoglobulin for capture and detection might misdiagnose T. evansi infection. Results indicate that this bias will be avoided if reagents for capture and detection are derived from different species.

The use of enzyme linked immunosorbent assays to investigate the prevalence of Trypanosoma equiperdum in Ethiopian horses

A field study involving 309 horses was undertaken in the provinces of Arsi and Bale in the Ethiopian highlands to investigate the prevalence of Trypanosoma equiperdum infections using enzyme linked immunosorbent assays (ELISAs) for the detection of both trypanosomal antigen and antibody. Adult horses of both sexes were examined for clinical signs of T. equiperdum infection and serum samples were collected for the assays. One hundred and one horses showed the presence of trypanosomal antibodies in their serum and 70 animals showed typical clinical signs of dourine. Nineteen horses showed the presence of trypanosomal antigen. Eight horses were positive for both T. equiperdum antibody and antigen. Blood and genital washes from seven antigenaemic horses were inoculated into mice and rabbits in an attempt to isolate trypanosomes but none became infected. Statistical analysis of the results of antibody assays indicated that there were significant differences in the distribution of serologically positive horses in the different clinical groupings, with seropositivity increasing with the severity of the observed clinical signs (P < 0.001). There was also a positive correlation between the presence of circulating trypanosomal antigen and clinical evidence of infection. Although it was not possible to obtain direct parasitological evidence of infection, the results of the serological assays, together with the clinical signs of disease observed in many of the horses, provide strong circumstantial evidence that T. equiperdum occurs in Arsi and Bale provinces of Ethiopia. Furthermore, in view of the large number of horses in Ethiopia and the unrestricted movement of animals throughout the country it is likely that dourine may be more widespread in Ethiopia than is currently realised. The assays used show potential for diagnosis of dourine, but to be widely applied in field situations for the diagnosis and control of dourine in Africa they require validation of their specificity and sensitivity.