EXPEIUMENTALPARASITOLOGY75,19-30

(1992)

Theileria parva: CD4+ Helper and Cytotoxic T-Cell Clones React with a Schizont-Derived Antigen Associated with the Surface of Theileria parva-Infected Lymphocytes CYNTHIA L. BALDWIN,' The International

KEITH P. IAMS, WENDY C. BROWN,~ AND DENNIS J. GRAB

Laboratory

for Research

on Animal Diseases, P.O. Box 30709, Nairobi,

Kenya

BALDWIN, C. L., IAMS, K. P., BROWN, W. C., AND GRAB, D. J. Theileria parva: CD4+ helper and cytotoxic T-cell clones react with a schizont-derived antigen associated with the surface of Theileria parva-infected lymphocytes. Experimental Parasitology 75, 19-30. Theileria parva is a protozoan parasite which infects and transforms bovine lymphocytes, resulting in a fatal lymphoproliierative disease. There is evidence that immunity to the intralymphocytic schizont stage is mediated by T cells. We have previously reported derivation of CD4+ T-cell clones which recognize parasite-derived antigens presented on the surface of infected cells in conjunction with MHC molecules and partial characterization of the antigens. The present study further evaluated one of these antigens, demonstrating that it could be derived from cells infected with different parasite stocks as well as from purified theilerial schizonts and that it was recognized by primed, but not unprimed, bovine lymphocytes including cytolytic CD4+ T cells. Using a cloned CD4+ cytolytic cell line, lysis of schizont-infected cells was shown to be MHC-restricted but not parasite-strain restricted. In addition we demonstrated that T cells which respond to the HSS antigenic preparation were generated in cattle immunized with parasites from any of the three subspecies of T. parva. The antigenic material was fractionated by sequential subjection to anion-exchange chromatography, hydroxylapatite chromatography, and gel filtration using HPLC, which resulted in recovery of approximately 20% of the antigenic material with more than 106-fold purification in selected fractions. To assess the molecular size of the proteins in the highly purified antigenic fractions, the T. parva-infected lymphocytes were metabolically labeled before fractionation with ‘H-amino acids and the material was analyzed by SDSpolyacrylamide gel electrophoresis and liquid scintillation counting of gel slices. The major protein in these fractions had a molecular mass of 9-10 kDa. Q KJW Academic press, IIIC. INDEX DESCRIPTORSAND ABBREVIATIONS: T-cell clone; Theileria parva; Parasite antigen; CPM, counts per minute (CPM); E:T, effector to target ratio (E:T); high-speed supematant antigenic preparation (HSS); International Laboratory for Research on Animal Diseases (ILRAD); ‘?UDR, ‘2510dodeoxyuridine (‘*‘IUDR); stock (St).

elude the bovine $8 T cell population as evidence by expression of CD3 (Baldwin et al. 1989). Parasitized lymphocytes proliferate in an uncontrolled manner, dependent upon the continued presence of the schizont (Pinder et al. f981), while retaining antigen specificity and function, as measured by cellular cytotoxicity (Baldwin and Teale, 1987). The disease is characterized by an acute lymphoproliferation of parasitized and nonparasitized cells, resulting in death of nonimmune hosts within 3 weeks after infection. Theileriosis, therefore, has a great impact on the cattle population in

INTRODUCTION Theileria parva, the causative agent of East Coast fever or theileriosis, is a protozoan parasite which infects bovine lymphocytes, including CD4+ and CD8+ T cells, B cells, and cells previously called null cells (Baldwin et al. 1988) but now known to in’ To whom correspondence should be addressed at Department of Microbiology, College of Biological Sciences, 484 West 12th Ave., The Ohio State University, Columbus, OH 43210. Tel. 614-292-3819. ’ Present address: Department of Veterinary Microbiology and Parasitology, Texas A & M University, College Station, TX 77843. 19

0014-4894/92$5.00 Copyright 0 1992 by Academic Ress, Inc. All rights of reproduction in any form reserved.

20

BALDWIN ETAL.

East and Central Africa and therefore on the food supply. While cattle can be immunized by infection with the parasite and treatment with antitheilerial drugs (Dolan et al. 1984), they are not necessarily immune to challenge with heterologous parasite stocks (Irvin et al. 1983;Young et al. 1973, 1975;Radley et al. 1975) and carrier status can result, creating a reservoir of parasites. A subunit vaccine which provides broad-based protection is a desirable alternative to currently available methods of immunization. There is evidence that the immune response to the schizont (intralymphocytic) stage of the parasite is predominately T-cell mediated (Emery 1981). Both CD4+ MHC class II-restricted (Baldwin et al. 1987)and CD8+ MHC class I-restricted (Goddeeris et al. 1986)T-cell clones which react with T. parva-infected lymphocytes have been propagated in vitro, but it is not known if both subpopulations are required for immunity to T. parva, nor have the antigens with which they react been purified. Recently, progress has been made toward identifying the antigens using T-cell lines which react with schizont-infected lymphocytes and cloned CD4+ cells derived from such lines (Brown et al. 1989a,b, 1990). It has been possible to demonstrate the existence of at least two different parasite-derived antigens which are associated with the surface of infected lymphocytes. The work described here extends these studies by further characterizing one of these antigen preparations and the immune T cells reactive with it. MATERIALSANDMETHODS animals. Boran cattle (Bos indicus) D768, C447, and C801 were immunized with stocks of T. parva (Muguga) by infection and treatment (Dolan et al. 1984). Animal C800, which was unimmunized, was a monozygotic twin of C801. All experimental cattle were housed at the International Laboratory for Research on Animal Diseases under tick-free conditions. MHC class I phenotypes were determined as described (Teale et al. 1983). Cattle, D503, D504, and Experimental

D487, described elsewhere (Kariuki et al. 1990), were derived by embryo transfer from the same parents and matched for MHC class I specificities. They were immunized with T. parva parva (Uganda), T. parva lawrencei (6998) or T. parva bovis (Boleni), respectively, as described (Kariuki et al. 19!3@). Generation of Theileria parva-infected cell lines. Parasite-infected lymphoblastoid cell lines were established as described (Brown 1979) by in vitro infection of PBMC with sporozoites derived from ticks. The ILRAD stabilates of parasite stocks used were T. parva (Muguga) (st 3087), T. parva (Mariakani) (st 3029), and T. parva (Marikebuni) (st 3014). T-cell clones. CD4+ clones 2B2, 2C3, and 3H12 have been described previously (Brown et al. 1990) and were established essentially as described by Baldwin and co-workers (Baldwin et al. 1987). The cytotoxic CD4+ T-cell clones were established similarly (Baldwin et al. 1987), except PBMC were stimulated three times in vitro with whole irradiated T. parva (Muguga)-infected cells and then cloned by limiting dilution at 3, 1, or 0.3 cells/well using the T. pnrvasoluble antigen preparation, previously designated the high-speed supematant (HSS) antigen (Brown et al. 1990), in the presence of 2 X lo4 y-irradiated autologous PBMC (antigen-presenting cells and tiller cells) and 10% T-cell growth factor (24 hr supematant from Cot&stimulated PBMC). Complete culture medium consisted of RPM1 1640, 10% heat-inactivated fetal bovine serum, 50 pg gentamycin/ml, 5 x 10 -5 M 2-mercaptoethanol and 2 mM t.-glutamine. Cells were cloned in %-well round-bottomed microtiter plates with a total volume of 200 J/well. Wells with cell growth were expanded by seeding the contents of a single well into multiple microwells and stimulating as above. Clones were screened for cytotoxic activity (Goddeeris et al. 1986)and selected clones expanded further in 24well culture plates using the same ratio of ingredients as for microwells. Cell surface phenotypic analysis of the clones with monoclonal antibodies IL-A12, reactive with the bovine homologue of CD4, (Baldwin et al. 1986)and ILA17, reactive with the bovine homologue of CD8, (Ellis et al. 1986)was performed as described (Baldwin et al. 1987). Proliferation assays. Proliferation assays were performed with PBMC, isolated over Ficoll-Paque (Pharmacia, Uppsala, Sweden) gradients as described (Lalor et al. 1986) by incubating 5 X lo5 PBMC/well in 96-well flat-bottomed microtiter plates in a total volume of 200 pl of complete culture medium for 5 days with varying amounts of the materials being tested for antigenicity. Proliferation assays with T-cell clones were performed exactly as described (Brown et al. 1990). Briefly, 3 X lo4 cloned T cells, 1 X lo5 irradiated (5000 rad) autologous adherent PBMC (added as a source of antigen-presenting cells) and antigen were

Theileria parva-SPECIFICCD4+ T CELL CLONES incubated in %-well microtiter plates in a total volume of 200 pi/well. Cultures were incubated at 39°C in a humidified atmosphere of 5% CO, in air for 48 to 96 hr. Some assays were conducted in half-area %-well flatbottom plates (Costar, Cambridge, MA) using one-half the volume and cell numbers indicated above. Proliferation was assessed by pulsing with 0.5 @i/well of ‘*‘I-iododeoxyuridine (IUDR) (Amersham, Amersham, UK) for 4 to 24 hr. (Baldwin et al. 1987). Cytotoxicity assays. Cytotoxicity assays using “‘Indium-oxine (Amersham) were performed as described (Goddeeris et al. 1986). Isolation of schizonts. Schizonts were isolated by homogenizing cells infected with T.p. parva (Muguga) suspended in Hanks’ Balanced Salt Solution (HBSS) without Ca’+/Mg’+ (GIBCO, Paisley, Scotland) at 4°C with 30 strokes of a Teflon motor-driven homogenizer (Arthur H. Thomas, Rochester, NY). The suspension, which comprised broken infected cells and intact schizonts, was diluted threefold with 4°C HBSS without Ca’+/Mg’+ containing 5 mM EDTA (HBSS/ EDTA solution). To remove unbroken T. parvainfected cells, the homogenate was centrifuged at 15Og for 5 min at 4”C, the supematant carefully collected, and again centrifuged at 15Ogfor 5 min. The supematant collected after the second centrifugation was then centrifuged at 800g for 15 mitt, to pellet the schizonts. The pelleted schizonts were resuspended in a suspension of 20% Percoll (Pharmacia), diluted from Percoll stock with the HBSSlEDTA solution, and centrifuged over Ficoll-Hypaque for 15 min at 8OOg, 4°C. Schizonts at the Ficoll-Paque/Percoll interface were collected and washed in HBSS/EDTA solution by centrifugation in a microfuge at 1600g for 5 min. Schizont preparations were assessed for purity, i.e., absence of unbroken infected cells and bovine host cell constituents, by light microscopic evaluation of Giemsa-stained smears and by determining the purity of T. parva RNA isolated from the schizont preparations. Bovine and T. parva RNA can be distinguished based on the sizes of their ribosomal RNA (rRNA) (Gerhards et al. 1989). The sizes of T. parva rRNA are 1.8, 1.9, and 3.8 kb, while the sizes of bovine rRNA are 1.8 and 4.3 kb. RNA was prepared as described previously (Chomczynski and Sacchi 1987) and analyzed by electrophoresis in formaldehyde-containing agarose gels (Sambrook et al. 1989)stained with ethidium bromide. RNA was also prepared from bovine lymphocytes and mixed with schizont RNA, for use as a standard to compare with the migration of the schizont rRNA as were RNA standards (0.24-9.5 kb RNA ladder; BRL, Grand Island, NY). Preparation of the T. parva HSS-soluble antigen. T. parva-infected cells were propagated and the T. parvasoluble HSS antigen was prepared essentially as described (Brown et al. 1989a). T. parva-infected cells were washed and suspended to a concentration of 1 to

21

2 x lo* cells/ml in PBS containing protease inhibitors (E-64, leupeptin, antipain, and cystatin). The cells were disrupted by passage through a French Pressure cell (1500 PSI). The broken cells were centrifuged at 125,OOOg for 60 min and the HSS was filtered through a 0.45~pm filter. HSS was diluted IO-fold with dH20 before passage over a column of DEAE-cellulose (Whatman DE53, Whatman Intl., Ltd., Maidstone, England) equilibrated with PBS diluted lo-fold with dH,O (0.015 M NaCl). The antigen which eluted in the flowthrough fraction was immediately applied onto a 1.7 x 2.6-cm hydroxylapatite (Bio-Gel HPT, Bio-Rad, Richmond, CA) column equilibrated in 10% v/v PBS in dH,O. The column was washed with 25 vol of 50 and 200 nuV NaPO, buffer (pH 7.5), before elution of the antigenic material with 400 n&f NaPO, buffer (pH 7.5). The eluted material was concentrated over a YMlO Amicon membrane (Amicon, Danvers, MA) and applied onto two HPLC columns (200 x 7.5 mm Bio-Rad BIO-SIL TSK-250 and TSK-125) connected in tandem, equilibrated in PBS, pH 7.2, as described (Brown et al. 1990). Elution was performed at room temperature with a flow rate of 0.75 ml/mitt, and protein was monitored at 235 nm. The following protein standards (Bio-Rad) were used for the HPLC analyses: thyroglobulin (670,000 kD), IgG (158,000 kD), ovalbumin (44,000 kD), myoglobin (17,000 kD), and vitamin B,* (1350 kD). Analysis of metabolically radiolabeled HSS proteins. Proteins of T. parva-infected cells were radiola-

beled by propagating the cells in the presence of 5 mCi each of [3H]leucine, [3H]lysine, [‘Hlproline and [3H]tyrosine, in a total volume of 2.5 liters of culture medium, for the final 48 hr of culture (Amersham). Between 1 x lo9 and 2 X lo9 T. parva (Muguga)infected cells were harvested from these cultures at the end of the incubation period. The radiolabeled cells were then used to prepare HSS which was immediately subjected to sequential purification by anion exchange chromatography (DE53), hydroxylapatite chromatography, and HPLC fractionation, as described above. Selected HPLC fractions were pooled and concentrated over a YCOS Amicon membrane (Amicon) according to manufacturer’s instructions and the entire pool was applied to a single lane of an SDS-PAGE gel. All counts per minute (CPM) of labeled material were recovered after concentration. SDS-PAGE was performed (Neville 1971) using slab gels (l-mm thick by 200-mm long) containing linear l&20% polyacrylamide gradients. Samples were heated in electrophoresis sample buffer (0.08 M TrisHCl, pH 6.7,7% glycerol, 3% 2-mercaptoethanol, and 2% SDS). A mixture of colored proteins (21.5-218 kD) and colored/“‘C-labeled proteins [ovalbumin (46 kD), carbonic anhydrase (30 kD), trypsin inhibitor (21.5 kD), lysozyme (14.3 kD), aprotinin (6.5 kD), insulin B-chain (3.4 kD), and insulin a-chain (2.35 kD)] (Rain-

22

BALDWIN ETAL.

bow Markers, Amersham) were used as molecular size markers. The size markers were electrophoresed in lanes on either side of those containing the experimental proteins, with one empty lane in between. After electrophoresis, the gels were sliced widthwise into 0.5cm pieces and lengthwise along the lane boundaries. Gel slices were solubilized in 3% protosol (New England Nuclear, Cambridge, MA) in scintillation vials containing scintillation grade toluene with 0.02% PGPOP and 0.4% PPO at 37°C overnight with shaking. Control slices of gel lanes without sample added were included, to determine the background reading throughout the gradient gel, as well as slices of lanes containing the radiolabeled molecular size markers. Each sample was counted in a scintillation counter for 10 min, with control for quenching.

RESULTS

Evaluation of mitogenicity of HSS. Previously we showed that HSS did not contain IL-2 activity (Brown et al. 1990). In order to establish that the HSS was not mitogenic, PBMC from cattle which were monozygotic twins, one of which was immunized with T. parva (Muguga) and one of which was naive, were compared for their proliferative response to HSS. While there was a low level of proliferation of PBMC from the T. parva-nonimmune twin with the HSS preparation (Table I, CSOl), the proliferation of PBMC from the immune sibling was considerably higher, indicating a response by T cells sensitized to T. parva antigens (Table I, CSOO).

toxic activity for T. parva-infected cells.

HSS was evaluated for its ability to induce CD4+ T cells which were cytolytic for T. parva-infected lymphocytes. To do this, immune PBMC were stimulated in vitro with whole irradiated T. parva-infected cells in the absenceof additional presenting cells and then cloned by limiting dilution culture in the presence of the HSS antigen and autologous presenting cells. Of the 47 clones tested, 10.6% were able to kill T. parva-infected cells, and of those, 3 mediated levels of killing exceeding 50% on autologous T. parva-infected target cells. Clones mediating the highest levels of cytotoxicity were expanded and assessedfor phenotype. None of the cells expanded from the original wells expressed the differentiation antigen CD8 (BoT8 in cattle), while the majority of the cells expanded from individual wells were highly CD4 +. Clone C447.32 was also tested for parasite strain restriction and MHC specificity on a panel of targets (Fig. 1). The levels of killing were less than 50% on all the target cell lines, a characteristic of bovine class IIrestricted cytotoxic cells (Teale et al. 1986). The cytolytic T cell clone was able to men

ET 10/r

E.T 5/l

Generation of CLM+ clones with cytoTABLE I Proliferative Response of PBMC from Theileria purvu-Immunized and -Unimmunized Twin Cattle to HSS-Soluble Antigenic Preparation Source of PBMC

C447lM

E49,A

E-%9/M

TARGET

Dilution of the HSS

C801 Unimmunized

C800 Immunized

No antigen l/5 II25 l/l25 l/525

1,154” (701) 3,498 (575) 3,250 (879) 2,677 (667) 890 (562)

3,104 (597) 41,396 (1,197) 39,711 (2,004) 19,084 (4,832) 7,272 (494)

n Mean (SD in parentheses) of CPM of incorporated ‘2sIUDR is indicated.

I

C 196/M

CELL

120.4/M

D409lA

LINES

FIG. 1. Cytotoxic activity of clone (X47.32 after three stimulations with the HSS antigen. Target cell lines are indicated by the animal from which the cells were obtained and the parasite stock was used to infect them. M, Theiferia parvu parvn (Muguga), A, T.p. parva (Marikebuni). The MHC class I phenotype of the target cells: C447, E49, and D409 were W7/WlO; Cl96 was W7/w6; and 120.4 was MHC mismatched with C447. Effector to target ratios (ET) tested were 10/l (open bars) and 511(hatched bars).

23

Theileria parva-SPECIFICCD4+ T CELL CLONES

diate cytotoxicity against cells infected with different parasite stocks, i.e., T. parva (Muguga) as well as those infected with T. parva (Marikebuni) (Fig. 1, compare C447/M with E49/A). The cytolytic activity was restricted, however, since not all cells infected with T. parva (Muguga) were killed (Fig. 1, see 120.4/M). The MHC class I phenotypes of the target cells were as follows: C447, E49, and D409 were W7/WlO, Cl96 was W7lW6, and 120.4 was MHC mismatched with C447. Our results indicate that restriction of killing did not correlate with expression of MHC class I antigens on the target cells, since some (see E49/A cells), but not all (see D409/A cells), target cell lines which were matched with the T cell clone for both class I haplotypes (W7/ WlO) were killed. We cannot confirm that restriction occurred by MHC class II antigens, since we are unable to distinguish among bovine class II MHC polymorphisms. Generation of HSS-responsive T cells in cattle immunized with different stocks ofT.

parva. Preferred vaccine candidates are antigens which are both common among strains of T. parva parasites and induce responsive T cells following in vivo infection of cattle with different strains of the parasite. We evaluated the induction of HSSresponsive T cells in cattle immunized with stocks representative of the three subspe-

ties of T. parva, i.e., T. parva parva, T. parva lawrencei, and T. parva bovis. PBMC from these cattle were tested in proliferation assays with HSS. The immunized cattle had the sameparents, derived by embryo transfer, and were matched for both age and MHC class phenotype. The results (Table II) indicated that immunization with any of the three subspecies of T. parva induced HSS-responsive lymphocytes in vivo. Proliferative responsesmeasured may include those by both CD4+ cytolytic and CD4+ noncytolytic T cells. Preparation of the HSS antigen from purified schizonts. HSS was prepared from

purified schizonts and fractionated by HPLC to determine if the HSS antigen was derived directly from the T. parva schizonts and, if so, which of the several antigenie peaks obtained by fractionation of whole T. parva-infected cells, described previously (Brown et al. 1990), were derived directly from the schizont rather than from outside the schizont, e.g., in the cytoplasm or organeles of the host cell or associated with the membrane of the infected cell. To evaluate the purity of the schizont preparations prior to fractionation, samples were used to prepare Giemsa-stained smears and RNA. By light microscopy, contamination of the schizonts with whole infected cells was estimated to be at a ratio of 10d7. By RNA analysis, the schizont

TABLEII Proliferative Responses of PBMC from MHC-Matched Cattle Immunized with Different Subspecies of Theileria parva

Parasite stock used for immunization of cattle from which PBMC were derived Dilution of T. parva HSS used to stimulate PBMC l/5 l/z 11125

T.p. lawrencei

T.p. parva

T.p. bovis

w98)

(Uganda)

(Boleni)

7.3” 7.4 1.9

5.5 3.9 1.7

11.1 10.6 1.1

(1Stimulation indices (ratio of the mean CPM of incorporated 1251UDRfrom triplicate cultures stimulated with antigen, divided by the mean CPM of replicate cultures without antigen) of the PBMC proliferation in response to stimulation with T. parva (Muguga) are indicated. All stimulation indices above 3 are significantly different from the controls.

24

BALDWIN

ET AL.

preparations appeared to be highly depleted of bovine RNA contamination (Fig. 2). 2110 Based on schizont counts made by light microscopy in the presence of trypan blue @- 1738 dye, we estimate that 20 to 50% of the $1366 schizonts were recovered in a viable state Yj 994 using this purification procedure. I When HSS produced from purified “0 622 schizonts was fractionated by HPLC gel fil2-x tration, several peaks of schizont-derived 5 10 15 20 25 30 35 antigenic material were eluted which had HPLC FRACTION NO. apparent molecular masses(M,) of approxFIG. 3. Fractionation of HSS produced from puriimately 43, 12, and less than 1 kD (Fig. 3). fied schizonts by gel filtration using HPLC. Protein When a noncytolytic T cell clone was used was monitored at 235 nm (solid line, right ordinate). to assessproliferation, we found that two of Elution of protein standards is indicated at the top of the figure. HPLC fractions were tested for their ability the major schizont HSS antigenic activity to stimulate proliferation of clone 2C3. Proliferation peaks (43 and 12 kD) were distinct from the was measured by CPM of incorporated “‘IUDR vast majority of schizont HSS material and (dashed line, left ordinate). occurred in regions similar to antigenic activity peaks of material derived from whole proliferative responses than did T. parva infected cells (Brown et al. 1990). (Muguga)-infected cells (Brown et al. Analysis of the HSS antigenic material. 1990), HSS prepared from T. parva (MariIn order to more precisely identify the com- akani)-infected cells induced similar levels ponents of the antigenic HSS preparations, of proliferation of clone 2C3 as did HSS further purification and analysis of the an- from the same number of T. parva tigenic material was conducted. For this, (Muguga)-infected cells (21,091 + 1107and whole lymphocytes infected with T. parva 23,085 + 1660 CPM, respectively, with a (Muguga) for two reasons. First, while we background proliferation of 445 ? 79 had previously found that cells infected CPM), indicating that the relative amounts with T. parva (Mariakani) induced stronger of antigenic material in the cell lysates were A

7.54.42.4+

ab bovine rRNA schizont rRNA

1.4-

FIG. 2. (A) Agarose gel electrophoresis of RNA size standards (lane a); a mixture of RNA from bovine lymphocytes and I’heileria parva schizonts (lane b). (B) Agarose gel electrophoresis of RNA size standards (lane a); RNA from two preparations of purified T. parva schizonts (lanes b and e are different loads of the same schizont RNA preparation, as are lanes c and f); a mixture of RNA from bovine lymphocytes and T. parva schizonts (lane d), included as a reference for migration characteristics of the bovine and schizont rRNA.

Theileria pUrVU-SPECIFIC

equivalent. Secondly, as indicated above, purification of schizonts was inefficient since as few as 20% of the schizonts were recovered from the whole infected cells. Preliminary experiments indicated that at neutral pH the majority of HSS antigen did not bind to either DEAE or CM cellulose gel in low salt (0.015 M) buffer (data not shown). DE53 cellulose anion exchange chromatography, therefore, was used for the initial purification step to reduce the extraneous components in freshly prepared HSS supernatant. Material which eluted off DEAE-cellulose in PBS diluted in dH,O (1: 10 v/v) was subjected immediately to hydroxylapatite column chromatography. The antigen which eluted in the 200-400 mM sodium phosphate fraction was concentrated and size fractionated by HPLC. We found it imperative to perform the entire fractionation procedure and test samples in proliferation assays with the T-cell clones in a single day, since storage of the material at any temperature substantially reduced its antigenic activity. The antigenic activity contained within the HPLC fractions following this purification scheme is shown (Fig. 4). Using noncytolytic T cell clones to evaluate antigenicity, antigenic activity peaks and troughs were found to be reproduced among runs and between HSSresponsive T-cell clones 2C3 and 3H12, relative to the elution time and absorbance profiles of the material at 235 nm. The majority of the antigenic activity of the HSS eluted after the majority of the protein eluted. We found that antigenic material continued to elute from the column for up to 18 min after the last protein standard (1.35 kD), as assessed by the ability of the material in the HPLC fractions to stimulate the T-cell clones (Fig. 4). When calculating antigenic content as the ability to induce proliferation of T-cell clones, approximately 20% of the antigenic material was recovered in the HPLC fractions. The material in the HPLC fractions could not be visualized on SDS-PAGE by silver

CD4’

= 8

T CELL

3

5

10

25

CLONES

15

t-PLC

20

25

30

FRACTION

35

40

No.

45

50

55

?i

2

FIG. 4. The HSS material purified as described under Materials and Methods was fractionated by HPLC as the final step. The material was eluted over 50 min and then was evaluated for protein content by measuring absorbance at 235 nm and CPM of incorporated ‘H-amino acids. These two measurements resulted in identical profiles. The solid line indicates absorbance at 235 nm and CPM of incorporated 3H-amino acids; the scale indicated on the outside right ordinate is for CPM of 3H-proteins; the inside right ordinate indicates OD 235 nm, starting at 0 units of absorbance up to 0.1 units indicated by the arrow). HPLC fractions 7-54 were evaluated for their abilities to induce proliferation of clone 2C3 (assessed by CPM of incorporated “%JDR; broken line, left ordinate). Elution of protein size standards are indicated by the arrows at the top of the graph.

staining, due to insufficient quantities of protein, therefore it was necessary to radiolabel the proteins before fractionation and to pool fractions for analysis. The T. parvainfected cell proteins were labeled by culturing the cells with 3H-amino acids. Assessing the HPLC fraction by CPM of incorporated 3H-amino acids resulted in an elution profile identical to that evaluated by absorbance at 235 nm (Fig. 4). The HPLC fractions were assessed for their ability to stimulate the T-cell clones and then pooled, concentrated, and electrophoresed on SDS-polyacrylamide gradient gels, and gel slices were assessed for counts of radioactivity by liquid scintillation. The initial experiments indicated that several protein peaks (see Fig. 5a) were found in the pools of the HPLC fractions 25-55 illustrated in Fig. 4. Since HPLC fractions 37-54 (Fig. 4) had the greatest relative antigenic activity to protein content (using the CPM of incor-

26

BALDWIN

1 2 3 4 5 6 7 8 9 101112131415161718~920

GEL

SLICE

NO.

ET AL.

1

5

10

15

20

GEL

21

SLICE

30

35

40

45

No.

FIG. 5. HSS was prepared from 2 x I@’ Theileriu parva-infected cells which had been metabolically labeled with 3H-amino acids and sequentially purified as indicated under Materials and Methods. (a) Following the purification, HPLC fractions 21-55 were pooled and electrophoresed by SDS-PAGE on a lO-20% gradient gel and the gel was sliced. Labeled proteins were eluted from gel slices and the CPM was measured by liquid scintillation counting. Results from a representative pool are indicated. The ordinate begins at the level of the mean background CPM of control lanes (no protein). The molecular size standards are indicated based on combined information from the colored and radiolabeled protein standards run in lanes in the same gel. (b) SDS-PAGE of pooled and concentrated HPLC fractions 38-54, prepared as described in (a) except that a 620% gradient gel was used.

porated 3H-amino acids as a relative measure of protein, we estimated that the antigen activity relative to protein content was increased more than IO’-fold in the “very late” HPLC fractions 37-54), further experiments were conducted to identify the content of those fractions after pooling. When those fractions were pooled and analyzed by SDS-PAGE, the 9-10 kDa protein was the dominant protein (Fig. 5b).

gen epitopes. The antigenic epitope, which is recognized by the T-cell clones, in the HSS antigenic preparation occurs in cells infected with at least five geographically distinct parasite stocks. This was shown by the ability of cells infected with different parasite stocks to stimulate the individual T-cell clones (Brown et al. 1990). While we have shown that immune T cells recognize parasite-specific antigens on the surface of the infected host cells (BaldDISCUSSION win et al. 1987; Goddeeris et al. 1986; Our interest is in protective immunity to Brown et al. 1990), no parasite-derived anthe schizont stage of T. parva infections. tigens associated with the surface of inSince this stage of the infection is intralym- fected host cells have been identified by anphocytic, immunity is almost certainly tibodies (Creemers 1982), except in one T-cell mediated. It has been shown that T instance (Newson et al. 1986). Studies concells can recognize strain-specific antigens ducted in recent years with virus-infected (Baldwin et al. 1987; Goddeeris et al. 1986) cells indicate that viral proteins inserted in as well as antigens common to cells in- the membrane of host cells are not those fected with any of the three subspecies of T. recognized by immune T cells, rather, impurvu (Kariuki et al. 1990). Recent molec- mune T cells recognize peptides of viral ular analyses of DNA from a large number proteins which are expressed on the host of T. pm-vu isolates indicate that the num- cell surface in conjunction with MHC molber of strains of this parasite may be vast ecules, and, in some cases, are otherwise (Conrad et al. 1987, 1989), emphasizing the only found inside the infected host cell need for vaccines based on common anti- (Townsend et al. 1986). This is also likely to

Theileria parva-SPECIFIC CD~+T CELL CLONES

be the case for the T. parva antigens which stimulate immune T cells. It is, therefore, necessary to employ T cells directly as probes to identify the stimulatory antigens. Feasible approaches to this include biochemical purification, as employed here and previously (Brown et al. 1989b; 1990), and production of recombinant proteins from schizont cDNA libraries (C. L. Baldwin and K. P. Iams, unpublished data). Exogenous or killed antigens such as biochemically fractionated antigens are presented following processing by antigenpresenting cells in conjunction with class II MHC molecules but not with class I MHC molecules (for reviews see Braciale et al. 1987; Long and Jacobson 1989). As a result, it was necessary in this study to focus on identification of antigens which stimulated CD4+ T cells. CD4* T cells have been shown to play a principal role in protective immunity to other intracellular pathogens including Plasmodium chabaudi (Suss et al. 1988; Brake et al. 1988), Mycobacterium spp, (Muller et al. 1987; Pedrazzini et al. 1987; Leveton et al. 1989) and Leishmania spp (reviewed in Liew 1989) and are likely to contribute to immunity to T. parva by their production of growth factors (Baldwin et al. 1986), which are necessary for proliferation of bovine CDS+ lymphocytes (Baldwin et al. 1986), and by lysis of parasite-infected cells, as demonstrated during the course of this study. Since y-interferon and tumor necrosis factor, cytokines important in limiting the intracellular growth of other parasites or transformed cells, are not effective in limiting the growth of schizont-infected lymphocytes (DeMartini and Baldwin 1991), cytotoxicity of infected cells is the most likely mechanism of cellmediated immunity. CD4+ cytotoxic cells specific for host cells infected with viruses (Bourgault et al. 1989) and bacteria (Kaufmann et al. 1987) have also been reported although little is known concerning their role in immunity. Their role is ambiguous since these cells

21

have the potential to kill class II-bearing antigen-presenting cells which are crucial for induction and maintenance of immune responses (Braakman et al. 1987). With reference to theileriosis, however, all schizont-infected host cells (CD4+ and CD8+ T cells, y/6 T cells and B cells) express class II MHC antigens (Black et al. 1981; DeMartini et al. submitted for publication), thereby implicating class II-restricted CD4+ cytotoxic cells in protective immunity. Since there are MHC constraints on recognition of class I-associated T. parva antigens by CD8+ cytotoxic cells (Goddeeris et al. 1990), lysis of infected cells by class II-restricted T. par-vu-specific CD4+ cells may provide an alternative method for clearance of parasitized cells. Identification of parasite molecules which contain antigen epitopes for CD4+ T cells may also reveal antigens containing epitopes which associate with class I MHC molecules to stimulate the potent cytotoxic CD8+ T cells (Pearson et al. 1982; Eugui and Emery 1981), as has been shown for proteins of influenza virus (Braciale et al. 1987). We analyzed HPLC fractions for their ability to stimulate proliferation of the noncytolytic T-cell clones. Our studies did not evaluate the ability of these fractions to sensitize targets for lysis by the HSSresponsive CD4+ cytolytic clones. Since the HSS antigenic preparation is a complex mixture of proteins, the CD4+ cytolytic clones are not necessarily recognizing the same antigenic peptide as the noncytolytic clones. Our results indicate that the antigenie activity which stimulated the noncytolytic clones appeared to be associated with a wide size range of material, when elution time relative to that of the standard proteins was used to predict their size. This occurred whether purified schizonts or whole infected lymphocytes were used to prepare the HSS antigen. Analysis of HPLC fractions containing radiolabeled proteins by SDS-PAGE indicated that proteins of vastly different sizes occurred

28

BALDWINETAL.

within fraction pools and proteins of particular molecular sizes eluted over a large time range, thus appearing in several different fraction pools (C. Baldwin, unpublished data; Grab et al., submitted). The proteins contained in the antigenic material may be interacting with one another, thereby forming aggregates which are disrupted when prepared under reducing conditions for SDS-PAGE analysis. The proteins are apparently interacting with the column matrix, since their actual retention time is not consistent with their predicted retention time based on their molecular size. The antigen epitope recognized by the noncytolytic T-cell clones may be contained in proteins of several sizes, resulting from glycosylation of proteins, formation of multimers, or degradation of the parent protein by the host cell after it leaves the schizont. Although the fate of nonvacuoleassociated intracytoplasmic antigen is not clear, some modifications do occur, as evidenced by the processing of large antigens into peptides for association with MHC class I molecules following injection of the protein antigens into the cytoplasm (Morrison et al. 1986; Moore er al. 1988). The lo-kDa protein identified here may, therefore, be derived from larger proteins. Recent evidence indicates that when HSS proteins, produced as described here, are iodinated after the hydroxyl apaptite fractionation and analyzed by SDS-PAGE, the major protein bands were similar to those predicted here, i.e. 48, 43, 32, 22.5 24, 17, and 10 kDa (Grab et al. 1991; Grab et al., submitted), with the unique proteins in the antigenic pools being the 48, 24, and 10 kDa proteins. The 48- and 24-kDa proteins may be precursors of the lo-kDa peptide. Also the small peptides of ~3.5 kDa may contribute to the antigenicity in the HPLC fractions. Recent technological advances have made it possible to sequence picamole quantities of proteins electophoretically transferred to membranesfrom SDS-PAGE gels (Matsudaira 1987). Cur-

rent efforts are aimed at obtaining a peptide sequence of the IO&Da peptide for further molecular characterization of the protein and its antigenic epitope. ACKNOWLEDGMENTS We would like to thank Dr. S. Kemp for MHC typing of cattle, Drs. D. Whitelaw, 0. Ole-MoiYoi, and T. Asonganyi for assistance with the HPLC, Dr. S. J. Black for helpful discussions, and F. Mbwika, K. Logan and Y. Vejee for technical assistance. This work was supported by the International Laboratory for Research on Animal Diseases.

REFERENCES BALDWIN, C. L., TEALE, A. J., NAESSENS, J. G., GODDEERIS, B. M., MACHUGH, N. D., AND MORRISON, W. I. 1986. Characterization of a subset of

bovine T lymphocytes that express BoT4 by monoclonal antibodies and function: Similarity to lymphocytes defined by human T4 and murine L3T4. Journal of Immunology 136,43854391. BALDWIN, C. L., AND TEALE, A. J. 1987. Alloreactive T cell clones transformed by Theileria parva retain cytolytic activity and antigen specificity. European Journal of Immunology 17, 1859-1862. BALDWIN, C. L., GODDEERIS, B. M., AND MortalSON, W. I. 1987. Bovine helper T-cell clones specific for lymphocytes infected with Theileria parva (Muguga). Parasite Immunology 9, 499-513. BALDWIN, C. L., BLACK, S. J., BROWN, W. C., CONRAD, P. A., GODDEERIS, B. M., KINUTHIA, S. W., LALOR, P. A., MACHUGH, N. D., MORRISON, W. I., MORZARIA, S. P., NAESSENS, J., AND NEWSON, J. 1988. Bovine T cells, B cells and null cells are transformed by the protozoan parasite Z’heileria parva. Infection and Immunity 56, 462-467. BALDWIN, C. L., CLEVERS, H., AND IAMS, K. P.

1989. Cloning and sequencing of cDNA coding for the bovine homologues of CD3 gamma, delta and epsilon. In “Proceedings, Second International Veterinary Immunology Symposium, Hannover, Germany.” [Abstract] BLACK, S. J., JACK, R., LALOR, P., AND NEWSON, J.

1981. Analyses of Theileria-infected cell surface antigens with monoclonal antibodies. In “Advances in the Control of Theileriosis” (A. D. Irvin, M. P. Cunningham, and A. S. Young, Eds.) pp. 327-339. Martinus Nijhoff Publishers, The Hague, Netherlands. BOURGAULT, I., GOMEZ, A., GOMRAD, E., PICARD, F., AND LEVY J. P. 1989. A Virus-S@fiC Cm+

cell-mediated cytolytic activity revealed by CD8 + cell elimination regularly develops in uncloned human antiviral cell lines. Journal of Immunology 142, 252-256.

Theileria

parva-SPECIFIC

BRAAKMAN, E., ROTTEVEEL, F. T. M., VANBLEEK, G., VANSEVENTER, G. A., AND LUCAS, C. J. 1987.

Are MHC class II-restricted cytotoxic T lymphocytes important? Immunology Today 8, 265-267. BRACIALE, T. J., MORRISON, L. A., SWEETSER, M. T., SAMBROOK, J., GETHING, M., AND BRACIALE, V. L. 1987. Antigen presentation pathways to class I and class IX MHC-restricted T lymphocytes. Immunological Reviews 98, 95-l 14. BRAKE, D. A., LONG, C. A., AND WEIDANZ, W. P. 1988. Adoptive protection against Plasmodium chabaudi adami malaria in athymic nude mice by a cloned T cell line. Journal of Immunology 140, 1989-1993. BROWN, C. G. D. 1979. Propagation of Theileria.

In

“Practical Tissue Culture Applications (K. Maramorosch and H. Hirumi, Eds.), pp. 223-254. Academic Press, New York. BROWN, W. C., SUGIMOTO, C., CONRAD, P. A., AND GRAB, D. J. 1989a. Differential response of bovine T cell lines to schizont-derived membrane and soluble antigens of Theileria pawa. Parasite Immunology 11, 567-583. BROWN, W. C., SUGIMOTO, C., AND GRAB, D. J. 1989b. Theileria parva: Bovine helper T cell clones

specific for both infected lymphocytes and schizont membrane antigens. Experimental Parasitology 69, 234-248. BROWN, W. C., LONSDALE-ECCLES, J. D., DEMARTINI, J. C., AND GRAB, D. J. 1990. Recognition of soluble Theileria parva antigen by bovine helper T

cell clones: Characterization and partial purification of the antigen. Journal of Immunology 144,271-277. CHOMCZYNSKI, P., AND SACCHI, N. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Biochemisfry 9, 1x-159. CONRAD, P. A., IAMS, K., BROWN, W. C., SOHANPAL, B., AND OLE-MOIYOI, 0. K. 1987. DNA probes detect genomic diversity in Theileria parva stocks. Molecular and Biochemical Parasitology 25, 213-226. CONRAD, P. A., OLE-MOIYOI, 0. K., BALDWIN, C. L., DOLAN, T. T., O’CALLAGHAN, C. J., NJAMUNGGEH, R. E. G., GROOTENHUIS, J. G., STAGG, D. A., LEITCH, B. L., AND YOUNG, A. S. 1989.

Characterization of buffalo-derived theilerial parasites with monoclonal antibodies and DNA probes. Parasitology 98, 179-187. CREEMERS, P. 1982. Lack of reactivity of sera from Theileria parva-infected and recovered cattle against cell membrane antigens of Theileria parva transformed cell lines. Veterinary Immunology and Immunopathology 3, 427-427. DEMARTINI, J. D., MACHUGH, N. D., NAESSENS, J. G., AND TEALE, A. J. In vivo and in vitro expres-

sion of MHC class II antigens on bovine lympho-

CD4’

T CELL

29

CLONES

cytes infected by Theileria parva, submitted for publication. DEMARTINI, J. D., AND BALDWIN, C. L. 1991. Production of IFN by Theileria parva-infected bovine lymphoblasts and effects of IFN-gamma, TNFalpha and IL-2 on infection and proliferation of parasitized cells. Infection and Immunity, in press. DOLAN, T. T., LINYONI, S. K., MBOGO, S. K., AND YOUNG, A. S. 1984. Comparison of long acting oxytetracycline and parvaquone in immunisation against East Coast fever by infection and treatment. Research in Veterinary ELLIS, J. A., BALDWIN, BENSAID, A., TEALE, AND MORRISON, W. I.

Science 31, 175-176. C. L., MACHUGH, N. D., A. J., GODDEERIS, B. M.,

1986. Characterization by a monoclonal antibody and functional analysis of a subset of bovine T lymphocytes that express BoT8, a molecule analogous to human CDS. Immunology 58, 351-358. EMERY, D. L. 1981. Adoptive transfer of immunity to infection with Theileria parva (East Coast fever) between cattle twins. Research in Veterinary Science

30, 364-367. EUGUI, E. M., AND EMERY, D. L. 1981. Genetically

restricted cell mediated cytotoxicity in cattle immune to Theileria parva. Nature 290, 251-254. GERHARDS, J., GILL, A. C., EHRFELD, A. Y. B., DOBBELAERE,

D. A. E., AND WILLIAMS,

R. 0.

1989. Isolation and characterization of RNA from the intracellular parasite Theileria parva. Molecular and Biochemical Parasitology 34, 15-24. GODDEERIS, B. M., MORRISON, W. I., TEALE, A. J., BENSAID, A., AND BALDWIN,

C. L. 1986. Bovine

cytotoxic T-cell clones specific for cells infected with the protozoan parasite Theileria parva: Parasite strain specificity and class I major histocompatibility complex restriction. Proceedings of the National Academy of Science, USA 83, 5238-5242. GODDEERIS, B. M., MORRISON, W. I., TOYE, P. G., AND BISHOP, R. 1990. Strain specificity of bovine Theileria parva-specific cytotoxic T cells is deter-

mined by the phenotype of the restricting class I MHC. Immunology 69, 39-44. GRAB, D. J., BALDWIN, C., INNIS, L., AND BROWN,

W. C. 1991. Tentative identification of an antigen associated with the surface of Theileria parvainfected lymphocytes. Federation of the American Societies of Experimental

Biology

Journal

5, A570.

[Abstract] IRVIN, A. D., DOBBELAERE, D. A. E., MWAMACHI, E. M., MINAMI, T., SPOONER, P. R., AND OCAMA,

3. G. R. 1983. Immunisation against East Coast fever: correlation between monoclonal antibody profiles of Theileria parva stocks and cross immunity in vivo. Research in Veterinary Science 35, 341-346. KARIUKI, T. M., GROOTENHUIS, J. G., DOLAN, T. T., BISHOP, R. P., AND BALDWIN, C. L. 1990. Immu-

BALDWIN

30

nization with Theileriu parva parasites from buffaloes results in generation of cytotoxic T cells which recognize antigens common among cells infected with stocks of T. parva parva, T. parva bovis and T. purvu luwrencei.

Infection

and Immunity

58, 3574-

3581. KAUFMANN, S. H. E., HUG, E., VATH, U., AND DELIBERO, G. 1987. Specific lysis of Listeria monocytogenes-infected macrophages by class II-restricted L3T4+ T cells. European Journal of Immunology 17, 237-246. LALOR, P. A., MORRISON,W. I., GODDEERIS,B. M., JACK, R. M., AND BLACK, S. J. 1986. Monoclonal antibodies identify phenotypically and functionally distinct cell types in the bovine lymphoid system. Veterinury

Immunology

and Immunoputhology

13,

121-140. LEVETON, C., BA~~ASS, S., CHAMPION, B., LUCAS, S., DESOUZA, B., NICOL, M., BANERJEE,D., AND ROOK, G. 1989. T-cell-mediated protection of mice against virulent Mycobucterium tuberculosis. Infection and Immunity

57, 390-395.

LIEW, F. Y. 1989. Functional heterogeneity of CD4+ T cells in leishmaniasis. Immunology Today 10, 4045. LONG, E. O., AND JACOBSON,S. 1989. Pathways of viral antigen processing and presentation to CTL: defined by the mode of virus entry? Immunology Today 10, 45-48. MATSIJDAIRA,P. 1987. Sequence from picomole quantities of proteins elect&lotted onto polyvinylidene diiuoride membranes. Journal of Biological Chemistry 262, 10,035-10,038. MOORE,M. W., CARBONE,F. R., AND BEVAN, M. J. 1988. Introduction of soluble protein into the class I pathway of antigen processing and presentation. Cell 54, 777-785.

MORRISON, L. A., LUKACHER, A. E., BRACIALE, V. L., FAN, D. P., AND BRACIALE, T. J. 1986. Differences in antigen presentation to MHC class I-and class II-restricted influenza virus-specific cytolytic T cell clones. Journal of Experimental Medicine 163, 903-921. MULLER, I., COBBOLD,S. P., WALDMANN, H., AND KAUFMANN, S. H. E. 1987. Impaired resistance to Mycobacterium tuberculosis infection after selective in vivo depletion of L3T4 + and Lyt-2 + T cells. Infection und Immunity 55, 2037-2041. NEVILLE, D. M., JR. 1971. Molecular weight determination of protein-dodecyl sulfate complexes by gel electrophoresis in a discontinuous buffer system. Journal

of Biological

Chemistry

246, 6328.

NEWSON, J., NAESSENS, J., STAGG, D. A., AND BLACK, S. J. 1986. A cell surface antigen associated with Theileriu parva lawrencei-infected bovine lymphoid cells. Parasite Immunology 8, 149-158. PEARSON,R. W., HEWETT, R. S., ROELANTS,G. E.,

ET AL. STAGG,D. A., AND DOLAN, T. T. 1982. Studies on the induction and specificity of cytotoxicity to Theileriu-transformed cell lines. Journal of Immunology 128, 2509-2513. PEDRAZZINI, T., HUG, K., AND LOUIS, J. A. 1987. Importance of L3T4+ and Lyt-2+ cells in the immunologic control of infection with Mycobucterium bovis strain bacillus calmetteguerin in mice: Assessment by elimination of T cell subsets in vivo. Journal of Immunology

139, 2032-2037.

PINDER, M., KAR, S., WITHEY, K. S., LUNDIN, L. B., AND ROELANTS, G. E. 1981. Proliferation and lymphocyte stimulatory capacity of Theileriuinfected lymphoblastoid cells before and after the elimination of intracellular parasites. Zmmunology 44,51-60. RADLEY, D. E., BROWN,C. G. D., BURRIDGE,M. J., CUNNINGHAM, M. P., KIRIMI, I. M., PURNELL, R. E., AND YOUNG, A. S. 1975.East Coast fever. 1. Chemoprophylactic immunization of cattle against Theileriu purvu (Muguga) and five theilerial strains. Veterinary Parasitology 1, 35-41. SAMBROOK,J., FRITSCH, E. F., AND MANIATIS, T. 1989. “Molecular Cloning: A Laboratory Manual.” 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Suss, G., EICHMANN, K., KURY, E., LINKE, A., AND LANGHORNE, J. 1988. Roles of CD4- and CD8bearing T lymphocytes in the immune response to the erythrocytic stages of Plusmodium chubaudi. Infection

and Immunity

56, 3081-3088.

TEALE, A. J., KEMP, S. J., YOUNG, F., AND SPOONER,R. L. 1983. Selection by major histocompatibility type (BoLA), of lymphoid cells derived from a bovine chimaera and transformed by Theileriu parasites. Parasite Immunology 5, 329-536. TOWNSEND,A. R. M., ROTHBARD,J., GOTCH,F. M., BAHADUR, G., WRAITH, D., AND MCMICHAEL, A. J. 1986. The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic peptides. Cell 44, 959968. YOUNG, A. S., BROWN, C. G. D., BURRIDGE,M. J., CUNNINGHAM, M. P., KIRIMI, I. M., AND IRVIN, A. D. 1973.Observations on the cross-immunity between Theileriu luwrencei (Serengeti) and Theileriu purvu (Muguga) in cattle. International Journal of Parasitology

3, 723-728.

YOUNG, A. S., BROWN, C. G. D., BURIUDGE,M. J., CUNNINGHAM, M. P., MUSISI, T. L., AND PURNELL, R. E. 1975. East Coast fever. 2. Cross immunity trials with a Kenya Strain of Theileriu purvu Lawrencei. Veterinury Purusitology 1, 43-50. Received 8 November 1991; accepted with revision 1 April 1992