- Email: [email protected]
Susceptibility of African buffalo and Boran cattle to Trypanosoma congolense transmitted by Glossina morsitans centralis
Veterinary Parasitology, 35 (1990) 219-231 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
219
S u s c e p t i b i l i t y of A f r i c a n B u f f a l o and B o r a n Cattle to Trypanosoma congolense t r a n s m i t t e d by
Glossina morsitans centralis J.G. GROOTENHUIS 1, R.H. DWINGER 2'*, R.B. DOLAN 3, S.K. MOLO02 and MAX MURRAY T M
1National Veterinary Research Centre (NVRC), Kenya Agricultural Research Institute (KARD, PO Kabete (Kenya) 2International Laboratory [or Research on Animal Diseases (ILRAD), PO Box 30709, Nairobi (Kenya) 3PO Box 24437, Nairobi (Kenya) (Accepted for publication 13 September 1989)
ABSTRACT Grootenhuis, J.G., Dwinger, R.H., Dolan, R.B., Moloo, S.K. and Murray, M. 1990. Susceptibility of African buffalo and Boran cattle to Trypanosoma congolense transmitted by Glossina morsitans centralis. Vet. Parasitol., 35: 219-231. Four African buffalo (Syncerus caller) and four Boran cattle (Bos indicus) were each exposed to the bites of 10 tsetse flies infected with Trypanosoma congolense. Although both groups of animals became infected, the buffalo showed no clinical signs of trypanosomiasis while the cattle suffered from the disease characterized by pronounced skin reactions, high parasitaemia and severe anaemia. The prepatent periods in the buffalo varied from 18 to 27 days in comparison with 11 to 14 days in the cattle. In the buffalo, skin reactions were only detectable by histological examination of skin biopsies, the peak of parasitaemia was at least a hundredfold below that in cattle and after 54 days parasites were no longer detected. In contrast, the cattle had a continuous high parasitaemia until they were treated with a trypanocidal drug 60 days after infection. Neutralizing antibody to metacyclic trypanosomes appeared in the buffalo during the prepatent period, 15-20 days after infection, whereas in cattle neutralizing antibody was not detected until 10 days after the first peak of the parasitaemia, 25-30 days after infection.
INTRODUCTION
African wildlife has survived in an environment where tsetse flies and trypanosomes have operated as effective agents of natural selection over many millions of years. Highly efficient mechanisms of co-existing with these parasites have thus evolved in wild animal populations. Domestic cattle, however, *Present address: ITC, PMB 14, Banjul, The Gambia. **Present address: Department of Veterinary Medicine, University of Glasgow, Scotland.
0304-4017/90/$03.50
© 1990 Elsevier Science Publishers B.V.
220
J.G. GROOTENHU1S ET AL.
are a relatively recent introduction to Africa (Epstein, 1971) and although trypanotolerance is recognised in certain breeds, it does not appear to be of the degree observed in African wild Bovidae. Elucidation of the mechanisms of trypanotolerance is important both in understanding the various aspects of the response to the parasite and in increasing trypanotolerance through selection; thus trypanotolerant cattle breeds are receiving increasing attention (Murray et al., 1984). However, the marked resistance to trypanosomiasis exhibited by wild African bovids makes them an important host system for comparative studies of trypanotolerance. The African buffalo and domestic cattle both belong to the subfamily Bovinae and are equally attractive to tsetse (Grootenhuis et al., 1982 ); however, buffalo survive in areas infested with tsetse which preclude cattle. In this paper the susceptibility of buffalo and cattle are compared following experimental infection with Trypanosoma congolense transmitted by Glossina morsitans centralis. MATERIALSAND METHODS
Experimental animals Four African buffalo and four Boran cattle had been reared in an environment free of trypanosomiasis at the National Veterinary Research Centre, Kabete. The buffalo group consisted of three two-year-old males and one twoyear-old female. The cattle were four 2-year-old steers. Six months before the experiment, three of the buffalo and three of the cattle were experimentally infected with Trypanosoma vivax and treated with the trypanocidal drug diminazene aceturate (Berenil, Hoechst AG, Frankfurt, W. Germany) at a dose of 10 mg kg -1 body weight (Dwinger et al., 1986). One buffalo and one steer remained unexposed to trypanosomes. None of the animals had serum antibodies against T. congolense metacyclic or blood stream forms as determined by indirect fluorescent antibody and serum neutralization tests. During the experiment the animals were kept in fly proof isolation units. For routine sampling the buffalo were manually restrained as were the experimental cattle. A/J mice were obtained from the ILRAD breeding colony and used for serum neutralization tests and for sub-inoculation of blood and skin extracts.
Trypanosomes T. congolense (IL 1180) was derived after one passage in mice from a clone IL 968 (Nantulya et al., 1984), a derivative of STIB 212, isolated from a lion in Tanzania (Geigy and Kauffmann, 1973).
SUSCEPTIBILITY TO T. CONGOLENSE
221
Tsetse flies Teneral G. m. centralis from the ILRAD-bred colony were allowed to feed for 21 days on the clipped flanks of a goat infected with T. congolense. The tsetse were not fed for two days, then allowed to probe singly on slides preheated to 37 °C and the salivary probes were examined for the metatrypanosomes. Those which had mature infections were fed on the shaved flanks of the buffalo or cattle and the bite sites were marked with a felt-tipped pen.
Experimental design The four buffalo and four Boran cattle were each bitten on the flank by 10 tsetse infected with T. congolense. One animal of each species received an additional 10 infected tsetse bites on the other flank. Observations were made in all animals on changes in skin thickness, parasite levels in the blood, packed red cell volume (PCV) percentage and immune response. The skin of the two animals which received 10 additional tsetse bites was biopsied and examined for the presence of trypanosomes and for histological changes.
Sampling techniques The buffalo were weighed before and after the experiment. Blood was collected from the jugular vein in evacuated tubes containing Na-ethylene-diamine-tetra-acetate ( E D T A ) . P C V percentage was measured and the buffy coat was examined for trypanosomes using phase contrast microscopy (Murray et al., 1977). The number of parasites was estimated by the semi-quantitative scoring method according to Paris et al., (1982). The skin thickness was measured daily for a period of 18 days at each marked bite site using Vernier callipers. The buffalo had a mean skin thickness of 13.5 mm and the cattle of 12.5 mm at the sites on their flanks to be exposed to the bites of tsetse flies. Skin biopsies were taken on Days 0, 5, 8, 10, 12, 14 and 18 after infection from those animals which had received additional tsetse bites. The animals were sedated using 2 % xylazine ( R o m p u n R, Bayer, Leverkusen, W. Germany) at a dose rate of 1 ml 100 k g - 1 body weight. The skin was shaved and disinfected and a piece measuring 2 × 0.5 cm was excised, including the epidermis, the dermis and the panniculus carnosus. The skin biopsy material was divided into two pieces, one for histological preparation and one for sub-inoculation into mice. The incision was closed using silk mattress sutures. The skin sample for subinoculation was cut under aseptical conditions into small pieces using scissors and a scalpel blade and suspended in Eagle's modified minimum essential medium ( p H 7.2, containing Earle's salts, NaHCO3 and 20 m M buffer) with 15% foetal calf serum. The material was centrifuged
222
J.G. GROOTENHUIS ET AL.
for 10 min at 180 g and the supernate was collected. Two A / J mice were each inoculated intraperitoneally with 0.5 ml of the supernate. A second pair of mice was inoculated with 0.5 ml of blood collected in E D T A on the same day as the skin biopsy. Subinoculations of the skin extract were continued at days on which skin biopsies were taken until parasites were detected in the blood of the donor animal or until the mice subinoculated with blood from the buffalo and the cattle became parasitaemic.
Histology Skin biopsies were fixed in mercuric chloride-formalin for 12-24 h, routinely processed, sections 3-5 # m in thickness were cut and stained with Mayer's haematoxylin and eosin, Giemsa and Toluidine blue (Drury and Wallington, 1980).
Serology Metacyclic neutralization infectivity tests ( M N I T ) were carried out using metacyclic trypanosomes originating directly from tsetse. Amounts of 20/11 of test sera from the buffalo and the cattle were placed into the wells of leucocyte migration plates (Sterilin, Middlesex, England) which were preheated to 37 ° C. Infected tsetse were allowed to probe singly through the netting of their tube into the serum in the wells of the plate. Four tsetse were allowed to probe into each serum sample. More of the same test serum was added to the 20/~l serum sample containing the metacyclic trypanosomes to obtain a volume of 600/tl. This mixture was collected into a syringe and left for 30 min at room temperature before intraperitoneal inoculation into three mice, each receiving 200 #1. The M N I T was repeated using metacyclic T. congolense IL 1180 cultured in vitro. Six mice were used per serum sample. Each of the mice was inoculated with 200 #l of a 1 : 10 serum dilution in phosphate buffered saline (pH 7.2) containing 200 metacyclic trypanosomes. In both tests the mice were checked twice a week for parasitaemia by microscopic examination of tail blood and neutralization was judged to have occurred if the mice failed to become infected, provided that the control mice became infected. RESULTS All the buffalo and cattle became infected. No differences were found between the animals previously exposed to T. vivax and the two non-exposed animals. Therefore, in further analysis individuals belonging to one species were treated as one group. The buffalo showed no clinical signs of trypanosomiasis. There was no febrile response, no pallor of the mucous membranes and
SUSCEPTIBILITY TO T. CONGOLENSE
223
a mean body weight increase of 9.5% was observed during the experimental period. In contrast, the cattle suffered from the disease, lost condition and had to be treated with diminazene aceturate 53 days after infection because of their low PCV values (20%).
Skin reactions The data for all tsetse feeding sites on the four buffalo and the four cattle are shown in Table 1 and plotted in Fig. 1. No clear pattern of skin thickness changes were observed in the buffalo, while the cattle data showed a distinct peak reaction on Day 12, with an average skin thickness increase > 5 mm. The differences between the groups were significant on an analysis of variance
(P<0.01). Skin thickness changes measured on sites which did not show a pattern of increased thickness over time could vary as much as 3 mm. Therefore an increased skin thickness > 4 mm, was considered to be a sign of a reaction. Skin reactions > 5 mm were macroscopically detectable. Skin reactions were therefore divided into three categories: < 4 ram; between 4 and 5 mm; a n d > 5 mm (Table 1). Seven out of the 40 sites on buffalo showed detectable skin reactions, but none of these reacting sites produced skin thickness increases > 5 ram. In contrast, 35 out of the 40 bite sites on the cattle had skin reactions > 5 mm. TABLE 1
Pattern of skin thickness increases following measurement of 10 single tsetse bite sites on four African buffalo and four Boran cattle. Each tsetse fly was shown to be infected with trypanosomes (T. congolense) before biting the skin Experimental animals
Mean initial skin thickness (ram)
No. of bite sites with skin thickness increases of: < 4 mm 4-5 mm > 5 mm
12 13 16 13
8 10 8 7
2 0 2 3
0 0 0 0
16 10 12 12
0 1 1 1
0 0 2 0
10 9 7 9
Buffalo 5641 6017 6020 6666
Cattle 279 283 288 931
224
J.G. G R O O T E N H U I S E T A1,.
/_/\
6J
©
E = 5 4
o
2 ,_
~ jo~o.~ ~
o\
,,,..o
/O~o_._._o...o o
.,~- o~.~:.-.,~.,~i~o
"-
"
0
...................................................................
.~, ......... p
\./
...........
l
I
I
I
I
I
l
I
I
I
2
4
fi
8
10
12
14
16
18
20
Days
after
infection
Fig. 1. M e a n increase in skin thickness in four African buffalo (closed circles) a n d four Boran cattle (open circles). Each p o i n t represents the m e a n of daily changes of 40 tsetse bite sites on each of the species. T h e tsetse flies were infected with T. congolense.
Histology
The histological architecture of the buffalo skin before infection was generally similar to that of cattle skin. However, it differed in having fewer hair follicles and therefore fewer associated glandular elements per unit area. There was no obvious erector pili muscle. A microscopic grid was used to determine the comparative density of sweat glands, sebaceous glands and hair follicles. A mean number of 3.0 sweat gland vesicles, 3.0 sebaceous glands and 3.8 hair follicles were found per mm 2 in histological sections of normal buffalo skin. In the uninfected Boran, a mean number of 9.1 sweat glands, 5.3 sebaceous glands and 7.6 hair follicles were found per mm 2 of skin. The buffalo had a sparse resident population of small lymphocytes and mast cells in the connective tissue surrounding vascular trunks and the above mentioned skin elements (Fig. 2A). In the histological sections taken from the Boran steer before infection an obvious but scanty resident population of cells, including mast cells, lymphocytes and occasional eosinophils, was observed (Fig. 3B). In the buffalo a minor inflammatory skin reaction was observed in contrast to the major changes observed in the Boran steer (Figs. 2B and 3B). The buffalo first showed a response to the bite of an infected tsetse eight days after infection. At this time, there was an increase in cellularity of the papillary and reticular dermis. The cellular infiltrate consisted of lymphocytes and mast cells. There was some oedema and thickening of the sweat gland walls. Ten days after infection the numbers of eosinophils increased in the infiltrating cell pop-
SUSCEPTIBILITYTO T. CONGOLENSE
225
Fig. 2. Skin of an African buffalo before (Fig. 2A) and 12 days after the bite of one G. m. centralis infected with T. c o n g o l e n s e (Fig. 2B ). Fig. 2A, in the papillary and reticular dermis sweat glands, hair follicles and a pilosebaceous unit can be seen; the dermis contains a scanty population of small lymphocytes and few mastcells. Fig. 2B, there is moderate infiltration (arrows) of small lymphocytes and eosinophils around vascular trunks and to a lesser extent around pilosebaceous units. There is also pronounced thickening of the sweat gland epithelium (arrow) (H.E. 110 × ).
ulations around pilo-sebaceous units and vascular trunks throughout the dermis. Certain sweat glands were obliterated as shown by disrupted vesicular walls and the lumina being replaced by inflammatory cells. There was also some proliferation of fibroblasts. This inflammatory response intensified on Day 12, at which time there was diffuse cellular infiltration, further degenerative changes of the sweat glands and intense perivascular cuffing of the vas-
226
J.G. GROOTENHUISET AL
Fig. 3. Skin of a Boran steer before (Fig. 3A) and 10 days after the bite of one G. m. centrali,s infected with T. congolense. Fig. 3A, larger numbers of cells consisting of small lymphocytes, some eosinopbils and few mastcells, are present than in the skin section taken at the same time from the buffalo; numerous sweat glands and hair follicles should be noted. Fig. 3B, there is massive cellular infiltration throughout the dermis {H.E. 110 X ).
culature in the muscular dermis. The predominant cell types were lymphocytes, eosinophils, and fibroblasts with very few neutrophils present (Fig. 2B ). Fourteen days after infection there was a reduction in cellular infiltration. Destruction of some of the sweat glands was seen as a result of proliferative changes followed by cellular infiltration. Infiltrates around the vascular trunks showed many pyknotic cells. Eosinophilic granulocytes were numerous in the cellular infiltrates. Cellular infiltration was further reduced 18 days after infection. By
SUSCEPTIBILITY TO T. CONGOLENSE
227
this time the number of sweat glands had increased and there were clear signs of regression of the inflammatory response. No trypanosomes were seen in the histological sections of the buffalo skin at any time. In the Boran, an inflammatory response was noted five days after infection, characterized by marked cellular infiltration around the vascular trunks in the papillary and reticular dermis. The cell types consisted mainly of lymphocytes, with some eosinophils and mast cells. Eight days after infection the number of cells had increased and the infiltration extended diffusely throughout the layers of the skin. At this time, trypanosomes were detected and most readily seen in the hypodermis. By Day 10 there were clear signs of necrosis indicated by the destruction of collagen, the appearance of pyknotic cells, karyorhexis and degenerative changes of the walls of the sweat glands {Fig. 3B ). No clear signs of regression of the inflammatory response were noticeable in the histological sections from biopsies taken on Days 12, 14 and 18 after infection.
A
5x106 6 5
o
o.o
10`4 4
o/
o
\
o.o
3 2
.f ..._..z',,...v....
5x102 1
1o
I
40!
I
[
I
a
I
/_.Xo.
35 | \
I°'oG / ~,*-e.~ o.411%.o,/
o~O
\'-'/
e
o~o
u
25
°~o.~. ° \o.~.
20
°""o"~'°~.o 0
10
0 Days
0 after
40
50
infection
Fig. 4. Mean parasitaemia (Fig. 4A) and PCV (Fig. 4B ) in four buffalo (closed circles) and four Boran cattle (open circles) following infection with T. co ngolense transmitted by G. m. centralis.
228
,I.G. GROOTENHUIS ET AL.
Parasitaemia and anaemia
By sub-inoculation of skin material from buffalo into mice, trypanosomes were first detected 10 days after infection. Eight days later parasites were shown to be present in the blood of one of the buffalo by mouse sub-inoculation and buffy coat examination. In the cattle, trypanosomes were detected in the skin by Day 5 and in the blood by Day 14. In the buffalo, the prepatent period varied between 18, 19, 20 and 27 days approximately twice as long as in the cattle (11, 12 and 14 days). The variation in the pattern ofparasitaemia and PCV changes between individual animals within species, was very slight; only the mean values for the two species are therefore shown in Fig. 4. Parasitaemia in the buffalo was only intermittently detectable while the cattle had continuously detectable parasitaemia. The parasitaemia levels in the buffalo were always 1-2 loglo below the parasitaemia levels of the cattle in which peaks of 106 trypanosomes ml- 1 of blood were attained. The mean PCV in the buffalo fluctuated around pre-infection levels for the first 18 days following infection, after which an increase in PCV was observed. In the cattle there was a continuous decline in PCV to mean levels of 20% at the end of the observation period, 53 days after infection. TABLE
2
Detection of metacyclic neutralizing antibody in sera from buffalo and cattle following infection with T. congolense Buffalo
Days after infection 11
15
20
26
33
53
(T+V)
(V)
(T)
(V)
(T)
(V)
(T)
5641
-
-
--
+
+
+
+
6017
-
-
+
+
+
+
+
6020
--
--
--
+
+
+
+
6666
-
--
+
+
+
+
+
0
Cattle
Days a~erin%ction 0
22
25
30
40
70
(T+V)
(V)
(T)
(V)
(V)
(T)
279
-
-
+
+
+
+
283
-
-
+
+
+
+
288
-
-
+
+
+
+
931
-
-
-
+
+
+
trypanosomes, obtained by letting tsetse probe into undiluted test serum, were inoculated into mice. V - 2 0 0 metacyclic trypanosomes cultured in vitro were mixed with test serum to a final serum dilution o f 1 / 1 0 and inoculated into each o f 6 mice.
T =metacyclic
SUSCEPTIBILITY TO T. CONGOLENSE
229
Serology Both buffalo and cattle produced neutralizing antibodies to metacyclic trypanosomes. In two out of four buffalo neutralizing antibodies to the metacyclic forms of T. congolense (IL 1180) were detected on Day 15 (Table 2). In the other two buffalo, the neutralizing antibodies were detected on Day 20. In three of the cattle neutralizing antibodies were detected on Day 25, and in the fourth animal 30 days after infection. Antibodies persisted throughout the observation period in both buffalo and cattle. DISCUSSION Although the buffalo in this experiment were susceptible to infection with
T. congolense, they were apparently unaffected by trypanosomiasis. The buffalo showed no clinical signs of the disease and they increased in weight during the time of the experiment. Their trypanotolerance appeared to be associated with the nature of the local skin reaction, the relatively low number of trypanosomes in the blood, the early antibody response and the absence of anaemia. In the buffalo, no skin thickness increase could be observed macroscopically, whereas in cattle 35 out of 40 bites developed into conspicuous chancre reactions with increases in skin thickness of > 5 mm. Proliferation of parasites in the buffalo skin appeared to be slower than in cattle as indicated by the fact that parasites in the skin of the buffalo were detected approximately one week later than in the skin of the cattle. The inflammatory response in the skin of the buffalo was mild in comparison to the severe reaction observed in the cattle. Similar differences have been observed by Akol et al. (1986), comparing trypanotolerant taurine West African cattle breeds with the more susceptible West African zebu. The type of cellular changes seen in the buffalo and cattle appeared to be largely similar and comparable to earlier reports (Akol and Murray, 1982; Akol et al., 1986), except for the absence of infiltration by neutrophils early in the course of infection and the striking prominence of eosinophilic granulocytes during the peak of the reaction in the buffalo. The prepatent period in the buffalo was almost twice as long as in the cattle. The levels of parasitaemia were usually more than a hundred fold below the parasitaemia levels in the cattle. These differences could be due to inhibition of trypanosome growth by non-immune parasite growth regulating factors in the skin or by limitation of trypanosome growth owing to an early immune response. Serological investigations showed that metacyclic neutralizing antibodies appeared in three of the buffalo before detection of parasitaemia. In the cattle, antibodies were not found until 13 to 18 days after the first detection of parasites in the blood. In experimental T. brucei infections in mice, differentiation from slender to stumpy senescent trypanosomes was required to ac-
230
,I.G. GROOTENHUIS ET AL.
tivate the protective immune response in the resistant mouse strain. The early accumulation of senescent trypanosomes correlated with an earlier priming of the immune system which appeared to be the basis for the relative resistance of the trypanotolerant mice (Sendashonga and Black, 1982; Black et al., 1985 ). Similarly, destruction of T. congolense metacyclic forms in the buffalo skin might have caused neutralizing antibodies to occur before parasites could be detected. The buffalo developed no anaemia and showed no signs of disease. There was in fact an increase in PCV from Day 20 onwards (Fig. 4B) in each of the four buffalo. However, this increase was not attributable to an increase in reticulocytes. Further investigations are required to investigate the biological significance of this increase. The buffalo were able to self-cure and gain weight during the course of infection. The trypanotolerance of the buffalo was manifest in a variety of ways; there was limitation of trypanosome growth at the level of the skin as well as in the circulation and the appearance of neutralizing antibody prior to detection of trypanosomes in the blood. Except for the early appearance of neutralizing antibody, the resistance to T. congolense infection of buffalo resembles the response of trypanotolerant cattle as reported by Akol et al. (1986). However, the reactions of the buffalo were much more pronounced; there was an absence of macroscopically detectable skin reactions, no decrease in PCV and no first peak of parasitaemia similar to levels observed in susceptible breeds of cattle. Thus, the buffalo provides an excellent model for examination of the trypanosome growth control mechanisms produced by the trypanotolerant host. Further examination of these mechanisms may lead to improved methods of trypanosomiasis control in cattle, either by identifying markers for selection of trypanotolerant animals or by providing methods to enhance factors which impede parasite growth in the skin or in the blood. The present data do not confirm whether trypanosome growth control is primarily a result of early events taking place at skin level or a systemic effect. This question will be addressed in future experiments by intravenous inoculation of the same trypanosomes in another group of buffalo. ACKNOWLEDGEMENTS The authors wish to thank Drs. H. and K. Hirumi for providing metacyclic forms of T. congolense IL 1180, cultured in vitro. Invaluable technical assistance was given by Mr. John Ejwau and Mr. G. Lamb. The care and attention given by A.A. Dallmeijer in hand-rearing one of the buffalo is greatly appreciated. This paper is published with the permission of the Director of Veterinary Services of Kenya.
SUSCEPTIBILITY TO T. CONGOLENSE
231
REFERENCES Akol, G.W.O. and Murray, M., 1982. Early events following challenge of cattle with tsetse infected with Trypanosoma congolense: development of the local skin reaction. Vet. Rec., 110: 295-302. Akol, G.W.O., Authie, E., Pinder, M., Moloo, S.K., Roelants, G.E. and Murray, M., 1986. Susceptibility and immune responses of Zebu and taurine cattle of West Africa to infection with Trypanosoma congolense transmitted by Glossina morsitans centralis. Vet. Immunol. Immunopathol., 11: 361-373. Black, S.J., Sendashonga, C.N., O'Brien, C., Borowy, N.K., Naessens, J., Webster, P. and Murray, M., 1985. Regulation of parasitaemia in mice infected with Trypanosoma brucei. Curr. Top. Microbiol. Immunol., 117: 93-118. Drury, R.A.B. and Wallington, E.A., 1980. Carleton's Histological Techniques. 5th ed., Oxford University Press, Oxford, 520 pp. Dwinger, R.H., Grootenhuis, J.G., Murray, M, Moloo, S.K. and Gettinby, G., 1986. Susceptibility of buffaloes, cattle and goats to tsetse-transmitted infection with different stocks of Trypanosoma vivax. Res. Vet. Sci., 41: 307-315. Epstein, H., 1971. The Origin of Domestic Animals of Africa. Africana Publishing Corporation, New York, Vol I, 573 pp. Geigy, R. and Kauffmann, M., 1973. Sleeping sickness survey in the Serengeti area (Tanzania) 1971. Part 1. Examination of large mammals for trypanosomes. Acta Trop., 30: 12-23. Grootenhuis, J.G., Varma, Y., Black, S., Moloo, S.K., Akol, G.W.O., Emery, D.L. and Murray, M., 1982. Host response of some African wild Bovidae to experimental trypanosome infection. In: E. Karbe and E.K. Freitas, (Editors), Workshop on Trypanotolerance and Animal Production, GTZ Publication No. 116, Eschborn, pp. 337-342. Murray, M., Murray, P.K. and McIntyre, W.I.M., 1977. An improved parasitological technique for the diagnosis of African trypanosomiasis. Trans. R. Soc. Trop. Med. Hyg. 71: 325-326. Murray, M., Trail, J.C.M. and Grootenhuis, J.G., 1984. Trypanotolerant livestock: potential and future exploitation. Outlook on Agriculture, 13: 1,433-511. Nantulya, V.M., Musoke, A.J,, Rarangirwa, F.R. and Moloo, S.K., 1984. Resistance of cattle to tsetse-transmitted challenge with Trypanosoma brucei or Trypanosoma congolense after spontaneous recovery from syringe passaged infections. Infect. Immun. 43: 735-738. Paris, J., Murray, M. and McOdimba, F., 1982. A comparative evaluation of the parasitological techniques currently available for the diagnosis of African trypanosomiasis in cattle. Acta Trop. 37: 307-316. Sendashonga, C.N. and Black, S.J., 1982. Humoral responses against Trypanosoma brucei variable surface antigen are induced by degenerating parasites. Parasit. Immunol. 4: 245-257.
219
S u s c e p t i b i l i t y of A f r i c a n B u f f a l o and B o r a n Cattle to Trypanosoma congolense t r a n s m i t t e d by
Glossina morsitans centralis J.G. GROOTENHUIS 1, R.H. DWINGER 2'*, R.B. DOLAN 3, S.K. MOLO02 and MAX MURRAY T M
1National Veterinary Research Centre (NVRC), Kenya Agricultural Research Institute (KARD, PO Kabete (Kenya) 2International Laboratory [or Research on Animal Diseases (ILRAD), PO Box 30709, Nairobi (Kenya) 3PO Box 24437, Nairobi (Kenya) (Accepted for publication 13 September 1989)
ABSTRACT Grootenhuis, J.G., Dwinger, R.H., Dolan, R.B., Moloo, S.K. and Murray, M. 1990. Susceptibility of African buffalo and Boran cattle to Trypanosoma congolense transmitted by Glossina morsitans centralis. Vet. Parasitol., 35: 219-231. Four African buffalo (Syncerus caller) and four Boran cattle (Bos indicus) were each exposed to the bites of 10 tsetse flies infected with Trypanosoma congolense. Although both groups of animals became infected, the buffalo showed no clinical signs of trypanosomiasis while the cattle suffered from the disease characterized by pronounced skin reactions, high parasitaemia and severe anaemia. The prepatent periods in the buffalo varied from 18 to 27 days in comparison with 11 to 14 days in the cattle. In the buffalo, skin reactions were only detectable by histological examination of skin biopsies, the peak of parasitaemia was at least a hundredfold below that in cattle and after 54 days parasites were no longer detected. In contrast, the cattle had a continuous high parasitaemia until they were treated with a trypanocidal drug 60 days after infection. Neutralizing antibody to metacyclic trypanosomes appeared in the buffalo during the prepatent period, 15-20 days after infection, whereas in cattle neutralizing antibody was not detected until 10 days after the first peak of the parasitaemia, 25-30 days after infection.
INTRODUCTION
African wildlife has survived in an environment where tsetse flies and trypanosomes have operated as effective agents of natural selection over many millions of years. Highly efficient mechanisms of co-existing with these parasites have thus evolved in wild animal populations. Domestic cattle, however, *Present address: ITC, PMB 14, Banjul, The Gambia. **Present address: Department of Veterinary Medicine, University of Glasgow, Scotland.
0304-4017/90/$03.50
© 1990 Elsevier Science Publishers B.V.
220
J.G. GROOTENHU1S ET AL.
are a relatively recent introduction to Africa (Epstein, 1971) and although trypanotolerance is recognised in certain breeds, it does not appear to be of the degree observed in African wild Bovidae. Elucidation of the mechanisms of trypanotolerance is important both in understanding the various aspects of the response to the parasite and in increasing trypanotolerance through selection; thus trypanotolerant cattle breeds are receiving increasing attention (Murray et al., 1984). However, the marked resistance to trypanosomiasis exhibited by wild African bovids makes them an important host system for comparative studies of trypanotolerance. The African buffalo and domestic cattle both belong to the subfamily Bovinae and are equally attractive to tsetse (Grootenhuis et al., 1982 ); however, buffalo survive in areas infested with tsetse which preclude cattle. In this paper the susceptibility of buffalo and cattle are compared following experimental infection with Trypanosoma congolense transmitted by Glossina morsitans centralis. MATERIALSAND METHODS
Experimental animals Four African buffalo and four Boran cattle had been reared in an environment free of trypanosomiasis at the National Veterinary Research Centre, Kabete. The buffalo group consisted of three two-year-old males and one twoyear-old female. The cattle were four 2-year-old steers. Six months before the experiment, three of the buffalo and three of the cattle were experimentally infected with Trypanosoma vivax and treated with the trypanocidal drug diminazene aceturate (Berenil, Hoechst AG, Frankfurt, W. Germany) at a dose of 10 mg kg -1 body weight (Dwinger et al., 1986). One buffalo and one steer remained unexposed to trypanosomes. None of the animals had serum antibodies against T. congolense metacyclic or blood stream forms as determined by indirect fluorescent antibody and serum neutralization tests. During the experiment the animals were kept in fly proof isolation units. For routine sampling the buffalo were manually restrained as were the experimental cattle. A/J mice were obtained from the ILRAD breeding colony and used for serum neutralization tests and for sub-inoculation of blood and skin extracts.
Trypanosomes T. congolense (IL 1180) was derived after one passage in mice from a clone IL 968 (Nantulya et al., 1984), a derivative of STIB 212, isolated from a lion in Tanzania (Geigy and Kauffmann, 1973).
SUSCEPTIBILITY TO T. CONGOLENSE
221
Tsetse flies Teneral G. m. centralis from the ILRAD-bred colony were allowed to feed for 21 days on the clipped flanks of a goat infected with T. congolense. The tsetse were not fed for two days, then allowed to probe singly on slides preheated to 37 °C and the salivary probes were examined for the metatrypanosomes. Those which had mature infections were fed on the shaved flanks of the buffalo or cattle and the bite sites were marked with a felt-tipped pen.
Experimental design The four buffalo and four Boran cattle were each bitten on the flank by 10 tsetse infected with T. congolense. One animal of each species received an additional 10 infected tsetse bites on the other flank. Observations were made in all animals on changes in skin thickness, parasite levels in the blood, packed red cell volume (PCV) percentage and immune response. The skin of the two animals which received 10 additional tsetse bites was biopsied and examined for the presence of trypanosomes and for histological changes.
Sampling techniques The buffalo were weighed before and after the experiment. Blood was collected from the jugular vein in evacuated tubes containing Na-ethylene-diamine-tetra-acetate ( E D T A ) . P C V percentage was measured and the buffy coat was examined for trypanosomes using phase contrast microscopy (Murray et al., 1977). The number of parasites was estimated by the semi-quantitative scoring method according to Paris et al., (1982). The skin thickness was measured daily for a period of 18 days at each marked bite site using Vernier callipers. The buffalo had a mean skin thickness of 13.5 mm and the cattle of 12.5 mm at the sites on their flanks to be exposed to the bites of tsetse flies. Skin biopsies were taken on Days 0, 5, 8, 10, 12, 14 and 18 after infection from those animals which had received additional tsetse bites. The animals were sedated using 2 % xylazine ( R o m p u n R, Bayer, Leverkusen, W. Germany) at a dose rate of 1 ml 100 k g - 1 body weight. The skin was shaved and disinfected and a piece measuring 2 × 0.5 cm was excised, including the epidermis, the dermis and the panniculus carnosus. The skin biopsy material was divided into two pieces, one for histological preparation and one for sub-inoculation into mice. The incision was closed using silk mattress sutures. The skin sample for subinoculation was cut under aseptical conditions into small pieces using scissors and a scalpel blade and suspended in Eagle's modified minimum essential medium ( p H 7.2, containing Earle's salts, NaHCO3 and 20 m M buffer) with 15% foetal calf serum. The material was centrifuged
222
J.G. GROOTENHUIS ET AL.
for 10 min at 180 g and the supernate was collected. Two A / J mice were each inoculated intraperitoneally with 0.5 ml of the supernate. A second pair of mice was inoculated with 0.5 ml of blood collected in E D T A on the same day as the skin biopsy. Subinoculations of the skin extract were continued at days on which skin biopsies were taken until parasites were detected in the blood of the donor animal or until the mice subinoculated with blood from the buffalo and the cattle became parasitaemic.
Histology Skin biopsies were fixed in mercuric chloride-formalin for 12-24 h, routinely processed, sections 3-5 # m in thickness were cut and stained with Mayer's haematoxylin and eosin, Giemsa and Toluidine blue (Drury and Wallington, 1980).
Serology Metacyclic neutralization infectivity tests ( M N I T ) were carried out using metacyclic trypanosomes originating directly from tsetse. Amounts of 20/11 of test sera from the buffalo and the cattle were placed into the wells of leucocyte migration plates (Sterilin, Middlesex, England) which were preheated to 37 ° C. Infected tsetse were allowed to probe singly through the netting of their tube into the serum in the wells of the plate. Four tsetse were allowed to probe into each serum sample. More of the same test serum was added to the 20/~l serum sample containing the metacyclic trypanosomes to obtain a volume of 600/tl. This mixture was collected into a syringe and left for 30 min at room temperature before intraperitoneal inoculation into three mice, each receiving 200 #1. The M N I T was repeated using metacyclic T. congolense IL 1180 cultured in vitro. Six mice were used per serum sample. Each of the mice was inoculated with 200 #l of a 1 : 10 serum dilution in phosphate buffered saline (pH 7.2) containing 200 metacyclic trypanosomes. In both tests the mice were checked twice a week for parasitaemia by microscopic examination of tail blood and neutralization was judged to have occurred if the mice failed to become infected, provided that the control mice became infected. RESULTS All the buffalo and cattle became infected. No differences were found between the animals previously exposed to T. vivax and the two non-exposed animals. Therefore, in further analysis individuals belonging to one species were treated as one group. The buffalo showed no clinical signs of trypanosomiasis. There was no febrile response, no pallor of the mucous membranes and
SUSCEPTIBILITY TO T. CONGOLENSE
223
a mean body weight increase of 9.5% was observed during the experimental period. In contrast, the cattle suffered from the disease, lost condition and had to be treated with diminazene aceturate 53 days after infection because of their low PCV values (20%).
Skin reactions The data for all tsetse feeding sites on the four buffalo and the four cattle are shown in Table 1 and plotted in Fig. 1. No clear pattern of skin thickness changes were observed in the buffalo, while the cattle data showed a distinct peak reaction on Day 12, with an average skin thickness increase > 5 mm. The differences between the groups were significant on an analysis of variance
(P<0.01). Skin thickness changes measured on sites which did not show a pattern of increased thickness over time could vary as much as 3 mm. Therefore an increased skin thickness > 4 mm, was considered to be a sign of a reaction. Skin reactions > 5 mm were macroscopically detectable. Skin reactions were therefore divided into three categories: < 4 ram; between 4 and 5 mm; a n d > 5 mm (Table 1). Seven out of the 40 sites on buffalo showed detectable skin reactions, but none of these reacting sites produced skin thickness increases > 5 ram. In contrast, 35 out of the 40 bite sites on the cattle had skin reactions > 5 mm. TABLE 1
Pattern of skin thickness increases following measurement of 10 single tsetse bite sites on four African buffalo and four Boran cattle. Each tsetse fly was shown to be infected with trypanosomes (T. congolense) before biting the skin Experimental animals
Mean initial skin thickness (ram)
No. of bite sites with skin thickness increases of: < 4 mm 4-5 mm > 5 mm
12 13 16 13
8 10 8 7
2 0 2 3
0 0 0 0
16 10 12 12
0 1 1 1
0 0 2 0
10 9 7 9
Buffalo 5641 6017 6020 6666
Cattle 279 283 288 931
224
J.G. G R O O T E N H U I S E T A1,.
/_/\
6J
©
E = 5 4
o
2 ,_
~ jo~o.~ ~
o\
,,,..o
/O~o_._._o...o o
.,~- o~.~:.-.,~.,~i~o
"-
"
0
...................................................................
.~, ......... p
\./
...........
l
I
I
I
I
I
l
I
I
I
2
4
fi
8
10
12
14
16
18
20
Days
after
infection
Fig. 1. M e a n increase in skin thickness in four African buffalo (closed circles) a n d four Boran cattle (open circles). Each p o i n t represents the m e a n of daily changes of 40 tsetse bite sites on each of the species. T h e tsetse flies were infected with T. congolense.
Histology
The histological architecture of the buffalo skin before infection was generally similar to that of cattle skin. However, it differed in having fewer hair follicles and therefore fewer associated glandular elements per unit area. There was no obvious erector pili muscle. A microscopic grid was used to determine the comparative density of sweat glands, sebaceous glands and hair follicles. A mean number of 3.0 sweat gland vesicles, 3.0 sebaceous glands and 3.8 hair follicles were found per mm 2 in histological sections of normal buffalo skin. In the uninfected Boran, a mean number of 9.1 sweat glands, 5.3 sebaceous glands and 7.6 hair follicles were found per mm 2 of skin. The buffalo had a sparse resident population of small lymphocytes and mast cells in the connective tissue surrounding vascular trunks and the above mentioned skin elements (Fig. 2A). In the histological sections taken from the Boran steer before infection an obvious but scanty resident population of cells, including mast cells, lymphocytes and occasional eosinophils, was observed (Fig. 3B). In the buffalo a minor inflammatory skin reaction was observed in contrast to the major changes observed in the Boran steer (Figs. 2B and 3B). The buffalo first showed a response to the bite of an infected tsetse eight days after infection. At this time, there was an increase in cellularity of the papillary and reticular dermis. The cellular infiltrate consisted of lymphocytes and mast cells. There was some oedema and thickening of the sweat gland walls. Ten days after infection the numbers of eosinophils increased in the infiltrating cell pop-
SUSCEPTIBILITYTO T. CONGOLENSE
225
Fig. 2. Skin of an African buffalo before (Fig. 2A) and 12 days after the bite of one G. m. centralis infected with T. c o n g o l e n s e (Fig. 2B ). Fig. 2A, in the papillary and reticular dermis sweat glands, hair follicles and a pilosebaceous unit can be seen; the dermis contains a scanty population of small lymphocytes and few mastcells. Fig. 2B, there is moderate infiltration (arrows) of small lymphocytes and eosinophils around vascular trunks and to a lesser extent around pilosebaceous units. There is also pronounced thickening of the sweat gland epithelium (arrow) (H.E. 110 × ).
ulations around pilo-sebaceous units and vascular trunks throughout the dermis. Certain sweat glands were obliterated as shown by disrupted vesicular walls and the lumina being replaced by inflammatory cells. There was also some proliferation of fibroblasts. This inflammatory response intensified on Day 12, at which time there was diffuse cellular infiltration, further degenerative changes of the sweat glands and intense perivascular cuffing of the vas-
226
J.G. GROOTENHUISET AL
Fig. 3. Skin of a Boran steer before (Fig. 3A) and 10 days after the bite of one G. m. centrali,s infected with T. congolense. Fig. 3A, larger numbers of cells consisting of small lymphocytes, some eosinopbils and few mastcells, are present than in the skin section taken at the same time from the buffalo; numerous sweat glands and hair follicles should be noted. Fig. 3B, there is massive cellular infiltration throughout the dermis {H.E. 110 X ).
culature in the muscular dermis. The predominant cell types were lymphocytes, eosinophils, and fibroblasts with very few neutrophils present (Fig. 2B ). Fourteen days after infection there was a reduction in cellular infiltration. Destruction of some of the sweat glands was seen as a result of proliferative changes followed by cellular infiltration. Infiltrates around the vascular trunks showed many pyknotic cells. Eosinophilic granulocytes were numerous in the cellular infiltrates. Cellular infiltration was further reduced 18 days after infection. By
SUSCEPTIBILITY TO T. CONGOLENSE
227
this time the number of sweat glands had increased and there were clear signs of regression of the inflammatory response. No trypanosomes were seen in the histological sections of the buffalo skin at any time. In the Boran, an inflammatory response was noted five days after infection, characterized by marked cellular infiltration around the vascular trunks in the papillary and reticular dermis. The cell types consisted mainly of lymphocytes, with some eosinophils and mast cells. Eight days after infection the number of cells had increased and the infiltration extended diffusely throughout the layers of the skin. At this time, trypanosomes were detected and most readily seen in the hypodermis. By Day 10 there were clear signs of necrosis indicated by the destruction of collagen, the appearance of pyknotic cells, karyorhexis and degenerative changes of the walls of the sweat glands {Fig. 3B ). No clear signs of regression of the inflammatory response were noticeable in the histological sections from biopsies taken on Days 12, 14 and 18 after infection.
A
5x106 6 5
o
o.o
10`4 4
o/
o
\
o.o
3 2
.f ..._..z',,...v....
5x102 1
1o
I
40!
I
[
I
a
I
/_.Xo.
35 | \
I°'oG / ~,*-e.~ o.411%.o,/
o~O
\'-'/
e
o~o
u
25
°~o.~. ° \o.~.
20
°""o"~'°~.o 0
10
0 Days
0 after
40
50
infection
Fig. 4. Mean parasitaemia (Fig. 4A) and PCV (Fig. 4B ) in four buffalo (closed circles) and four Boran cattle (open circles) following infection with T. co ngolense transmitted by G. m. centralis.
228
,I.G. GROOTENHUIS ET AL.
Parasitaemia and anaemia
By sub-inoculation of skin material from buffalo into mice, trypanosomes were first detected 10 days after infection. Eight days later parasites were shown to be present in the blood of one of the buffalo by mouse sub-inoculation and buffy coat examination. In the cattle, trypanosomes were detected in the skin by Day 5 and in the blood by Day 14. In the buffalo, the prepatent period varied between 18, 19, 20 and 27 days approximately twice as long as in the cattle (11, 12 and 14 days). The variation in the pattern ofparasitaemia and PCV changes between individual animals within species, was very slight; only the mean values for the two species are therefore shown in Fig. 4. Parasitaemia in the buffalo was only intermittently detectable while the cattle had continuously detectable parasitaemia. The parasitaemia levels in the buffalo were always 1-2 loglo below the parasitaemia levels of the cattle in which peaks of 106 trypanosomes ml- 1 of blood were attained. The mean PCV in the buffalo fluctuated around pre-infection levels for the first 18 days following infection, after which an increase in PCV was observed. In the cattle there was a continuous decline in PCV to mean levels of 20% at the end of the observation period, 53 days after infection. TABLE
2
Detection of metacyclic neutralizing antibody in sera from buffalo and cattle following infection with T. congolense Buffalo
Days after infection 11
15
20
26
33
53
(T+V)
(V)
(T)
(V)
(T)
(V)
(T)
5641
-
-
--
+
+
+
+
6017
-
-
+
+
+
+
+
6020
--
--
--
+
+
+
+
6666
-
--
+
+
+
+
+
0
Cattle
Days a~erin%ction 0
22
25
30
40
70
(T+V)
(V)
(T)
(V)
(V)
(T)
279
-
-
+
+
+
+
283
-
-
+
+
+
+
288
-
-
+
+
+
+
931
-
-
-
+
+
+
trypanosomes, obtained by letting tsetse probe into undiluted test serum, were inoculated into mice. V - 2 0 0 metacyclic trypanosomes cultured in vitro were mixed with test serum to a final serum dilution o f 1 / 1 0 and inoculated into each o f 6 mice.
T =metacyclic
SUSCEPTIBILITY TO T. CONGOLENSE
229
Serology Both buffalo and cattle produced neutralizing antibodies to metacyclic trypanosomes. In two out of four buffalo neutralizing antibodies to the metacyclic forms of T. congolense (IL 1180) were detected on Day 15 (Table 2). In the other two buffalo, the neutralizing antibodies were detected on Day 20. In three of the cattle neutralizing antibodies were detected on Day 25, and in the fourth animal 30 days after infection. Antibodies persisted throughout the observation period in both buffalo and cattle. DISCUSSION Although the buffalo in this experiment were susceptible to infection with
T. congolense, they were apparently unaffected by trypanosomiasis. The buffalo showed no clinical signs of the disease and they increased in weight during the time of the experiment. Their trypanotolerance appeared to be associated with the nature of the local skin reaction, the relatively low number of trypanosomes in the blood, the early antibody response and the absence of anaemia. In the buffalo, no skin thickness increase could be observed macroscopically, whereas in cattle 35 out of 40 bites developed into conspicuous chancre reactions with increases in skin thickness of > 5 mm. Proliferation of parasites in the buffalo skin appeared to be slower than in cattle as indicated by the fact that parasites in the skin of the buffalo were detected approximately one week later than in the skin of the cattle. The inflammatory response in the skin of the buffalo was mild in comparison to the severe reaction observed in the cattle. Similar differences have been observed by Akol et al. (1986), comparing trypanotolerant taurine West African cattle breeds with the more susceptible West African zebu. The type of cellular changes seen in the buffalo and cattle appeared to be largely similar and comparable to earlier reports (Akol and Murray, 1982; Akol et al., 1986), except for the absence of infiltration by neutrophils early in the course of infection and the striking prominence of eosinophilic granulocytes during the peak of the reaction in the buffalo. The prepatent period in the buffalo was almost twice as long as in the cattle. The levels of parasitaemia were usually more than a hundred fold below the parasitaemia levels in the cattle. These differences could be due to inhibition of trypanosome growth by non-immune parasite growth regulating factors in the skin or by limitation of trypanosome growth owing to an early immune response. Serological investigations showed that metacyclic neutralizing antibodies appeared in three of the buffalo before detection of parasitaemia. In the cattle, antibodies were not found until 13 to 18 days after the first detection of parasites in the blood. In experimental T. brucei infections in mice, differentiation from slender to stumpy senescent trypanosomes was required to ac-
230
,I.G. GROOTENHUIS ET AL.
tivate the protective immune response in the resistant mouse strain. The early accumulation of senescent trypanosomes correlated with an earlier priming of the immune system which appeared to be the basis for the relative resistance of the trypanotolerant mice (Sendashonga and Black, 1982; Black et al., 1985 ). Similarly, destruction of T. congolense metacyclic forms in the buffalo skin might have caused neutralizing antibodies to occur before parasites could be detected. The buffalo developed no anaemia and showed no signs of disease. There was in fact an increase in PCV from Day 20 onwards (Fig. 4B) in each of the four buffalo. However, this increase was not attributable to an increase in reticulocytes. Further investigations are required to investigate the biological significance of this increase. The buffalo were able to self-cure and gain weight during the course of infection. The trypanotolerance of the buffalo was manifest in a variety of ways; there was limitation of trypanosome growth at the level of the skin as well as in the circulation and the appearance of neutralizing antibody prior to detection of trypanosomes in the blood. Except for the early appearance of neutralizing antibody, the resistance to T. congolense infection of buffalo resembles the response of trypanotolerant cattle as reported by Akol et al. (1986). However, the reactions of the buffalo were much more pronounced; there was an absence of macroscopically detectable skin reactions, no decrease in PCV and no first peak of parasitaemia similar to levels observed in susceptible breeds of cattle. Thus, the buffalo provides an excellent model for examination of the trypanosome growth control mechanisms produced by the trypanotolerant host. Further examination of these mechanisms may lead to improved methods of trypanosomiasis control in cattle, either by identifying markers for selection of trypanotolerant animals or by providing methods to enhance factors which impede parasite growth in the skin or in the blood. The present data do not confirm whether trypanosome growth control is primarily a result of early events taking place at skin level or a systemic effect. This question will be addressed in future experiments by intravenous inoculation of the same trypanosomes in another group of buffalo. ACKNOWLEDGEMENTS The authors wish to thank Drs. H. and K. Hirumi for providing metacyclic forms of T. congolense IL 1180, cultured in vitro. Invaluable technical assistance was given by Mr. John Ejwau and Mr. G. Lamb. The care and attention given by A.A. Dallmeijer in hand-rearing one of the buffalo is greatly appreciated. This paper is published with the permission of the Director of Veterinary Services of Kenya.
SUSCEPTIBILITY TO T. CONGOLENSE
231
REFERENCES Akol, G.W.O. and Murray, M., 1982. Early events following challenge of cattle with tsetse infected with Trypanosoma congolense: development of the local skin reaction. Vet. Rec., 110: 295-302. Akol, G.W.O., Authie, E., Pinder, M., Moloo, S.K., Roelants, G.E. and Murray, M., 1986. Susceptibility and immune responses of Zebu and taurine cattle of West Africa to infection with Trypanosoma congolense transmitted by Glossina morsitans centralis. Vet. Immunol. Immunopathol., 11: 361-373. Black, S.J., Sendashonga, C.N., O'Brien, C., Borowy, N.K., Naessens, J., Webster, P. and Murray, M., 1985. Regulation of parasitaemia in mice infected with Trypanosoma brucei. Curr. Top. Microbiol. Immunol., 117: 93-118. Drury, R.A.B. and Wallington, E.A., 1980. Carleton's Histological Techniques. 5th ed., Oxford University Press, Oxford, 520 pp. Dwinger, R.H., Grootenhuis, J.G., Murray, M, Moloo, S.K. and Gettinby, G., 1986. Susceptibility of buffaloes, cattle and goats to tsetse-transmitted infection with different stocks of Trypanosoma vivax. Res. Vet. Sci., 41: 307-315. Epstein, H., 1971. The Origin of Domestic Animals of Africa. Africana Publishing Corporation, New York, Vol I, 573 pp. Geigy, R. and Kauffmann, M., 1973. Sleeping sickness survey in the Serengeti area (Tanzania) 1971. Part 1. Examination of large mammals for trypanosomes. Acta Trop., 30: 12-23. Grootenhuis, J.G., Varma, Y., Black, S., Moloo, S.K., Akol, G.W.O., Emery, D.L. and Murray, M., 1982. Host response of some African wild Bovidae to experimental trypanosome infection. In: E. Karbe and E.K. Freitas, (Editors), Workshop on Trypanotolerance and Animal Production, GTZ Publication No. 116, Eschborn, pp. 337-342. Murray, M., Murray, P.K. and McIntyre, W.I.M., 1977. An improved parasitological technique for the diagnosis of African trypanosomiasis. Trans. R. Soc. Trop. Med. Hyg. 71: 325-326. Murray, M., Trail, J.C.M. and Grootenhuis, J.G., 1984. Trypanotolerant livestock: potential and future exploitation. Outlook on Agriculture, 13: 1,433-511. Nantulya, V.M., Musoke, A.J,, Rarangirwa, F.R. and Moloo, S.K., 1984. Resistance of cattle to tsetse-transmitted challenge with Trypanosoma brucei or Trypanosoma congolense after spontaneous recovery from syringe passaged infections. Infect. Immun. 43: 735-738. Paris, J., Murray, M. and McOdimba, F., 1982. A comparative evaluation of the parasitological techniques currently available for the diagnosis of African trypanosomiasis in cattle. Acta Trop. 37: 307-316. Sendashonga, C.N. and Black, S.J., 1982. Humoral responses against Trypanosoma brucei variable surface antigen are induced by degenerating parasites. Parasit. Immunol. 4: 245-257.