Molecular Genetic and Functional Analyses of Surface Molecules of Theileria annulata Sporozoites. By Frank Katzer For the submission as D. Phil. Thesis University of York July 1995 1 Abstract: The molecular genetics and potential function of two sporozoite surface antigens (SPAG-1 and SPAG-2) of Theileria annulata have been investigated using a variety of molecular and immunochemical techniques. Both of these antigens are candidates for inclusion in a subunit vaccine. In this thesis I will concentrate on the role of SPAG-1, since SPAG-2 only became available during the later stages of this research. I have been able to demonstrate that SPAG-1 is encoded by a single copy gene. Further, I have identified four SPAG-1 alleles using a PCR-based approach. The sequences of two different full-length SPAG-1 alleles were compared. The comparison of these revealed that the C- and N-termini are highly conserved and that the central region of the antigen gene is highly polymorphic. The implications for vaccine development and functional importance are discussed. I mapped the 5' end of the SPAG-1 mRNA in an attempt to identify the promoter region involved in the stage specific regulation of the gene. The comparison of the upstream regions of SPAG-1 and p67, the T. parva homologue, resulted in the identification of two palindromic, 6 bp long sequence motifs, which are conserved in the 5' region of both genes. They are situated in a highly conserved stretch near the predicted mRNA initiation site and one of these palindromic sequences is repeated. The function of these motifs is unknown. I provide evidence for the existence of a cryptic, 30 bp long, intron in the SPAG-1 gene. I have been able to show that recombinant SPAG-1 and SPAG-2 bind to a subset of bovine peripheral blood mononuclear cells. Some of these cells are targets for sporozoite invasion. The importance of SPAG-1, SPAG-2 and the elastin receptor for sporozoite invasion is discussed. My results show that SPAG-1 also binds to BL3 cells which are also targets for sporozoite invasion. The SPAG-1 receptor number on BL3 cells is presented. 2 Contents Title Abstract. Contents. Table Index. Figure Index Abbreviations. Acknowledgements. Declaration. Chapter 1. Introduction. I. I. The Parasite. 1.1.2. The Phylum. 1.1.3. The Distribution of the Theileria genus and their vectors. 1.1.4. The Life cycle. 1.2. Tropical Theileriosis. 1.2. The disease. 1.2.2. The immune response of the bovine host. Pages 1 2 3 6 7 9 10 11 12 12 12 15 16 21 21 23 1.3. Treatment and Control. 27 1.3.1. Treatment of tropical theileriosis. 27 1.3.2. Vector control. 28 1.3.3. Vaccination with attenuated macroschizont-infected cell lines. 29 1.3.4. The infection and treatment method. 30 1.3.5. Recombinant sub-unit vaccines. 31 1.4. Sporozoite Antigens. 32 1.4.1. The sporozoite surface antigen SPAG-1.32 1.4.2. The sporozoite surface antigen SPAG-2.36 1.5. Antigenic Polymorphism. 36 1.5.1. Antigenic polymorphism in Plasmodium with relevance to vaccine development. 37 1.5.2. The detection of polymorphism in Theileria. 38 1.5.3. The study of polymorphism for treatment and vaccine purposes. 39 1.6. Stage-specific gene regulation. 40 1.6.1. Stage-specific gene regulation in trypanosomes. 40 1.6.2. Stage-specific gene regulation in Plasmodium. 41 1.6.3. Stage-specific gene regulation in Theileria. 46 1.7. Host cell recognition and invasion. 49 1.7.1. Host cell recognition and invasion by malaria parasites. 49 1.7.2. Host cell attachment and invasion by Toxoplasma tachyzoites. 54 1.7.2. The process of host cell recognition and invasion by Theileria sporozoites. 1.7.3. Identification of target cells for invasion by Theileria 56 sporozoites. 59 1.8. Objectives of my work. 62 Chapter 2. Materials and Methods. 64 2.1 Materials. 64 2.1.1. Buffers. 64 2.1.2. Bacterial culture media. 65 3 2.1.3. 2.1.4. 2.1.5. 2.1.6. 2.1.7. 2.1.8. 2.1.9. 2.1.10. 2.1.11. 2.1.12. 2.1.13. 2.1.14. 2.1.15. 2.2. Methods. 2.2.1. 2.2.2. 2.2.3. 2.2.4. 2.2.5. 2.2.6. 2.2.7. 2.2.8. 2.2.9. 2.2.10. 2.2.11. 2.2.12. 2.2.13. Solutions for extracting plasmid DNA. Solutions for restriction digests. Solutions for agarose gel electrophoresis. Solutions for preparing GST fusion proteins. Solutions for SDS polyacrylamide gel electrophoresis. Solutions for Western blotting. Solutions for biotinylation and flow cytometry. Solutions for protein iodinations and binding assay. Tissue culture media. Solutions for extracting eukaryotic DNA. Solutions for Southern blotting. Solutions for RNA extractions and primer extensions Solutions for working with bacteriophage X. Tissue culture. Preparation of DNA. Electrophoresis and Southern blotting. Molecular cloning techniques. DNA sequencing. Preparation of recombinant proteins. Protein analysis. Flow cytometry. Iodination and 1251 labelled protein binding assay. Sporozoite RNA extraction. Sl intron mapping. Primer extension. %gtl 1 library screening. Chapter 3 Sequence comparison of SPAG-1 alleles: practical and functional importance. 66 67 68 68 69 71 71 72 73 73 74 74 76 78 78 79 81 83 86 86 88 89 90 91 92 92 93 95 3.1. Introduction. 95 3.1.1. The sequence analysis of SPAG-1 with respect to functional importance. 3.1.2. The sequence analysis of SPAG-1 with respect to vaccine 95 development. 96 3.2. Results. 96 3.2.1. Allelic RFLP analysis demonstrates that SPAG-1 is a single copy gene. 3.2.2. Sequence analysis of a genomic copy (allele gH3.4) of SPAG-1. 3.2.3. The comparison of the cDNA (cH) and genomic (gH3.4) sequence of SPAG-1. 3.2.4. Further analysis of variable and constant regions of SPAG-1 alleles. 3.2.4.1. Sequence analysis of the most polymorphic region of SPAG-1. 3.2.4.2. Sequence analysis of the C terminal constant region of SPAG-1. 3.2.5. Comparison of SPAG-1 with the p67 antigen of Theileria parva: implication for the composition of the 1A7 epitope. 3.2.6. The Eco RI RFLP pattern is based on the loss of an internal 96 99 104 107 107 112 115 Eco RI site. 115 3.3. Discussion. 118 3.3.1. SPAG-1 is a single copy gene. 118 3.3.2. Implications of the cDNA and genomic SPAG"1 sequence comparison. 118 4 3.3.3. Implications of the sequence comparison of the four SPAG-1 alleles. 119 3.3.4. Implications for vaccine development. 119 3.3.5. Implications for the functional importance of SPAG-1.120 Chapter 4 Regulation of the expression of SPAG-1. 122 4.1 Introduction 122 4.2. Results. 122 4.2.1. Sequencing of the 5' untranslated region of SPAG-1.122 4.2.2. Sequence comparison of p67 and gH3.4.124 4.2.3. Sequence comparison of the 5' untranslated region of 4 alleles of SPAG-1.127 4.2.4. Mapping the beginning of the SPAG-1 mRNA. 127 4.2.5. Putative promoter binding sites in the 5' untranslated region of SPAG-1.130 4.2.6. Evidence for an intron in SPAG-1.132 4.2.7. Confirmation of intron by S1 mapping. 134 4.2.8. Comparison of the SPAG-1 intron to other Theileria introns. 136 4.3. Discussion. 137 4.3.1. Stage-specific regulation of SPAG-1.137 4.3.2. The SPAG-1 intron. 139 Chapter 5 Binding studies with recombinant sporozoite antigens. 141 5.1. Introduction. 141 5.1.1. Are SPAG-1 and SPAG-2 ligands for host cell recognition/ invasion? 141 5.2. Results 143 5.2.1. Cloning and sequence data for SPAG-1 and the SPAG-2 constructs. 143 5.2.2. Expression of SPAG-1 and the SPAG-2 constructs. 148 5.2.3. Cleavage of SPAG-1 and the SPAG-2 constructs. 152 5.2.4. Do SPAG-1 and SPAG-2 bind to host cells? 154 5.2.5. Which are the target cells for SPAG-1 and SPAG-2 binding? 162 5.2.6. SPAG-1 binds to BL3 cells but SPAG-2 does not. 166 5.2.7. BL3 cells do not express the elastin receptor. 168 5.2.8. How many SPAG-1 receptors are there on BL3 cells? 168 5.3. Discussion. 173 5.3.1. Confirmation of the validity of the protein constructs. 173 5.3.2. Are SPAG-1 and SPAG-2 involved in host cell recognition and invasion? 174 5.3.3. Is the elastin receptor involved in the process of host cell recognition? 175 5.3.4. Future work and unanswered questions. 176 Chapter 6 General Discussion. 178 6.1. SPAG-1 and sub-unit vaccine development. 178 6.2. Stage-specific regulation of the SPAG-1 gene. 180 6.3. Functional importance of the SPAG molecules. 181 References. Appendix Publications 184 227 5 Table Legends Page Table 1: Theileria species, their tick vectors and their distribution. 13 Table 2: Plasmodium genes and their stage specific gene regulation. 42 Table 3: Transcription factors and the DNA sequence motif they bind to. 45 Table 4: Stage specifically regulated Theileria genes which are under transcriptional control. 48 Table 5: Comparison of target cells for sporozoite invasion by T. annulata and T. parva. 61 Table 6: Origins of cloned macroschizont infected cell lines. 79 Table 7: The oligonucleotide primers used for sequencing and cloning. 84 Table 8: Cloned macroschizont infected cell lines, the SPAG-1 alleles and the Eco RI RFLPs they contain. 99 Table 9: Monoclonal antibodies used in the two-colour flow cytometry. 162 6 Figure Index Page Figure 1: The distribution of Theileria annulata. 15 Figure 2: The life cycle of Theileria annulata. 17 Figure 3: cDNA sequence of T. annulata surface antigen (SPAG-1). 34 Figure 4: Mimicry of bovine elastin by T. annulata surface antigen SPAG-1.35 Figure 5: Diagrammatical representation of the invasion of a leucocyte by a Theileria sporozoite. 58 Figure 6: Southern blot analysis of the SPAG-1 associated RFLPs. 97 Figure 7: A Diagrammatic representation of the sub-cloning of gH3.4 into plasmid pSPAG1.4 and pSPAG3.4.101 Figure 8: Diagrammatic representation of the SPAG-1 gH3.4 sequence obtained for each DNA orientation. 102 Figure 9: DNA sequence of the gH3.4 allele of SPAG-1.103 Figure 10: Comparison of SPAG-1 protein sequences, derived from the DNA sequences of alleles gH3.4 and cH. 105 Figure 11: Schematic representation of the PCR based cloning and sequencing strategy. 108 Figure 12: Agarose gel with PCR products, stained with. ethidium bromide. 109 Figure 13: Agarose gel with DNA from pGEM-T clones, stained with ethidium bromide. 109 Figure 14: Comparison of the SPAG-1 alleles over the most polymorphic region (bases 850-1107, amino acids 284-369). 110 Figure 15: Comparison of the SPAG-1 alleles over the most constant region (bases 2419-2706, amino acids 806-902). 113 Figure 16: Schematic representation of the comparison of the cH allele of SPAG-1 and p67 of T. parva. 116 Figure 17: Comparison of the SPAG-1 alleles across the internal Eco RI site. 117 Figure 18: Diagrammatic representation of the sequencing information for the 5' untranslated region of SPAG-1.123 Figure 19: The sequence of 1141 bp of the 5' untranslated region of SPAG-1 and the first 297 bp of the SPAG-1 gene. 125 Figure 20: Sequence comparison of the 5' untranslated sequence of SPAG-1 of T. annulata and p67 of T. parva. 126 Figure 21: Sequence comparison of the first 350 bp of the 5' untranslated sequence of SPAG-1 from 4 different alleles. 128 7 Figure 22: Autoradiograph of the primer extension analysis. 129 Figure 23: Sequence motifs in the 5' untranslated region of SPAG-1.131 Figure 24: The SPAG-1 intron. 133 Figure 25: Autoradiograph of the SI nuclease intron mapping experiment. 135 Figure 26: Sequence comparison of introns found in Theileria. 136 Figure 27: Diagrammatic representation of the SPAG-2 constructs. 145 Figure 28: Diagrammatic representation of the SPAG-2 sequence information. 146 Figure 29: DNA sequence of SPAG-2 (KP8). 147 Figure 30: Coomassie stained SDS PAGE gel of the GST-SPAG proteins. 149 Figure 31: Western blot of GST-SPAG-1.151 Figure 32: Sequence data of the junctions of the SPAG-2 clones in the pGEX vector system. 152 Figure 33: Coomassie stained SDS PAGE gels of the cleaved SPAG-2 during different stages of purification. 153 Figure 34: Coomassie stained SDS PAGE gel of the cleaved SPAG proteins. 155 Figure 35: Western blot of biotinylated protein detected with Extravidine. 156 Figure 36: Single colour flow cytometry using a dilution series of biotinylated proteins and PBM cells. 158 Figure 37: Two colour flow cytometry of PBM cells using biotinylated cleaved SPAG-1 and a panel of monoclonal antibodies. 163 Figure 38: Two colour flow cytometry of PBM cells using biotinylated cleaved KP8 and a panel of monoclonal antibodies. 164 Figure 39: Two colour flow cytometry of PBM cells using the biotinylated cleaved C350 and a panel of monoclonal antibodies. 165 Figure 40: One colour flow cytometry of BL3 cells using biotinylated cleaved SPAG-1.167 Figure 41: One colour flow cytometry of BL3 cells using the anti-elastin receptor antibody BCZ. 169 Figure 42: Autoradiograph of 1251 labelled SPAG-1 and KP8 on a SDS PAGE gel. 171 Figure 43: Binding curves of SPAG-1 to BL3 cells. 172 8 Abbreviations APS BCIP BoLA bp BSA cDNA cpm CTVM DEPC DMSO DNA MT EDTA GST IAA IPTG kb kDa LB mAb MHC mRNA NBT NMS CD p67 PBM PBS PCR PIPES PMSF RFLP RNA RPMI RT SDS SDS-PAGE SPAG-1 SPAG-2 SSC TaA TaH TBL 1E THE X-Gal Ammonium persulphate 5-bromo-4-chloro-3-indolyl phosphate Bovine leucocyte antigen Base pairs Bovine serum albumin Complementary deoxy-ribonucleic acid Counts per minute Centre for Tropical Veterinary Medicine (Edinburgh) Diethyl pyrocarbonate Dimethyl sulfoxide Deoxy-ribonucleic acid Dithiothreitol Ethyl enediaminetetra-acetic acid Glutathione-S-transferase Isoamyl alcohol Isopropylthio-ß-D-galactoside Kilobases Kilodaltons Luria-Bertani medium Monoclonal antibody Major histocompatibility complex Messenger ribo-nucleic acid Nitro blue tetrazolium Normal mouse serum Optical density T. parva 67 kDa sporozoite antigen Peripheral blood mononuclear cells Phosphate buffered saline Polymerase chain reaction Piperzaine-NN'-bis-2-ethane sulphonic acid Phenylmethylsulphonyl fluoride Restriction fragment length polymorphism Ribo-nucleic acid Roswell Park Memorial Institute medium Room temperature Sodium dodecyl sulphate Sodium dodecyl sulphate polyacrylamide gel electrophoresis T. annulata sporozoite surface antigen 1 T. annulata sporozoite surface antigen 2 Sodium citrate Theileria annulata Ankara Theileria annulata Hisar Theileria annulata infected bovine lymphocytes Tris-EDTA Tris-sodium chloride-EDTA 5-bromo-4-chloro-3-indolyl-ß-D-galactoside 9 Acknowledgements I have to thank Dr. R. Hall for his helpful discussions, supervision and proof reading. I am indebted to the Department of Biology at the University of York for providing my fees, the Institute for Animal Health in Compton for the CASE Studentship and my parents for providing my maintenance money. I also have to thank Dr. W. I. Morrison for his helpful discussions, funding and use of materials and equipment during my visits to Compton. I also have to thank Dr C. J. Howard, P. Sopp and K. Griffith for their help and advice during the flow cytometry studies. I would like to thank Duncan Brown, Susanna Williamson, Lesley Bell-Sakyi and Gwen Wilkie from the Centre for Tropical Veterinary Medicine, Edinburgh, for providing sporozoite material, the cloned macroschizont cell lines and the 41311 and I A7 monoclonal antibodies. I also have to thank Dr. P. Knight for providing the SPAG-2 clones and her sequencing data. I am also indebted to the Nuffield Foundation for providing a travel grant for me to visit the laboratory of Prof. R. Mecham at the Washington University Medical Centre, St. Louis. I would like to thank Prof. R. Mecham for his kind hospitality, teaching me the iodination and binding procedures and for the provision of the two monoclonal antibodies BCZ and BA4. I also have to thank Meg Stark and Pete Crosby for taking the pictures displayed in this thesis. In Dr. R. Hall's laboratory I would like to thank Daniel Lawson for helping me to get started using the computers, Nicky Boulter for discussing my thesis and making life in the laboratory more enjoyable. I also wish to thank Rob Somerville for being Rob and for keeping me amused, Andy Harrison for helpful discussions and William Starritt for looking after me. Finally but not least I thank Dr. Phil Hunt for providing the ground for my work, his helpful comments and discussion. 10 Declaration: The studies reported in this thesis are the work of the author with the help of those people listed in the acknowledgements. This thesis has not been submitted previously for the award of a degree to any university. The following publications include work described in this thesis: Boulter, N., Knight, P. A., Hunt, P. D., Hennessey, E. S., Katzer, F., Tait, A., Williamson, S., Brown, D., Baylis, H. A. and Hall, R. (1994) Theileria annulata sporozoite surface antigen (SPAG-1) contains neutralizing determinants in the C terminus. Parasite Immunology 16,97-104. Katzer, F., Carrington, M., Knight, P., Williamson, S., Tait, A., Morrison, I. W. and Hall, R. (1994) Polymorphism of SPAG-1, a candidate antigen for the inclusion in a sub-unit vaccine against Theileria annulata. Molecular and Biochemical Parasitology 67,1-10. Katzer, F., Knight, P., Williamson, S., Morrison, I. W. and Hall, R. (1994) Sporozoite surface molecules of Theileria annulata : Variation and Function. pp 105-109 In: Proceedings of the Third EU Workshop on Tropical Theileriosis. (Antalya, Turkey, 1994). Knight, P., Musoke, A. J., Gachanja, J. N., Nene, V., Katzer, F., Boulter, N., Hall, R., Brown, C. G. D., Williamson, S., Kirvar, E., Bell-Sakyi, L., Hussain, K. and Tait, A. (submitted) Conservation of neutralising determinants between the sporozoite surface antigens of Theileria annulata and T. parva. Experimental Parasitology. 11 Chapter 1 Introduction and literature review. 1.1. The Parasite. The protozoan parasite Theileria annulata was first described by Dschunkowsky and Luhs (1904) and is the causative agent of tropical theileriosis or Mediterranean Coast fever. World-wide this parasite threatens an estimated 250 million cattle and is of major economical importance due to constraints in livestock production in endemic areas due to its disease. The eradication of this parasite is hindered by its ability to infect cells of the hosts' immune system and to infect several host species. 1.1.2. The Phylum. Theileria annulata belongs to a group of tick-borne apicomplexan parasites which infect a variety of wild and domestic animals throughout the world. The genus Theileria contains a number of species, some of which infect cattle as shown in Table 1. The different species and their characteristics have been reviewed by Uilenberg (1981), Dolan (1989) and Morzaria and Nene (1990). Other apicomplexan parasites which are of major importance are Plasmodium, Eimeria, Babesia, Sarcocystis and Toxoplasma. AII apicomplexan parasites are characterised by possession of an apical complex during at least one of their life cycle stages and by exhibiting at least one intracellular stage within the mammalian host. For example, they invade either erythrocytes, leucocytes or hepatocytes. The closest related parasites to Theileria are those of the genus Babesia. These two genera can be distinguished by their morphology and their target cells since Babesia species only invade erythrocytes while Theileria species 12 1 tý 'fl w ,,,; " wO u ý .! C p E aS iG ... x '0 ~ .r H yýý 3 oý H V.:.: ' 3ý cc y en ý .+ý uý oý ýo 10 % .. tu -0 " -: " ý ... u Ü c ,rN H ý Z O\ -. .! L b. xO^ ~ :. u mý. '" C9N [ a0 ýý 'C üu ý ü 6° == ýý ý ý ý'v 0: ees 's et M hbN ýý g 00 "_ uýooeý :_ .CN u I, 3r., "_ = aa. 1e Rý w Vý mOa ýüM ,_ 00 cýxz ý°: ý Fý°=°= .ý = en CL w ý " w~ o w; ="IE wý aw ä y+~ü 0 ý a U ý ýý E Ew c .ý ýý ^ ý¢ ý a ü b äö ý, w " 42 uu^ýý ed ä E 1u ý tu .a ' r, Co : In ýäý"v y %1 zi 'C7 . ýa 6. ýýý a == F+ r, = H 1"n :; ýüa h 4"' """ "In _u I. O aoO"- pO zwvýý ýa ww ¢o OOu výzO vý -*S a. «s ¢U ýw < a. O. c 2U w ¢ e,.., a u a o .: u ý > O U ._1.. 1 > Uu ._ ý>Ö a u . . "F y < ¢ y o _ uwm "º- , co a ._ . H o H cý ýo - ... , y 0 In te .. vu "° O CvI.. 00 ý O. _ °° p U 'C ;r - O °° _ 'ü v .ü ' ü o Ca >ÜE- E-1 :2 r_ = ý `° o wU ýc OE"1 " cn ý ö Q ß g q h Q, y y y ý h h 4) . r3 ý I. .ý 'C h tl iz r- B: r- E E Q6 %. ) ö ,v ý tl p C >, ý ý ý= 4 "4 g -O y O ;r b h d `: ý 4z I. .O p ý b b b ýv "O Q° U ` ý V m f :x ýV ý tl 2 2 . .. y ý :: ý ýý ý °' I .b %) ". N d ä ý" O ý O, r '" v u b ÖO tl ý oi r y tl y hý ý CZ cQ cz t3 h yr4 ho0 f3Qp Wh u u h b "ý "ý ' i d .. r b tl d O d ý F4 Er ý F4 N: ý ý E4 . -: ý 00 oý ý v c O N "Ir Ir O ý ý a ý c .ý ý ý, E 0 4. n v 4) ý b I. V .Z e ý C L4 Vf I- C r.. U ý > . ýe 1 :: V w w h v ýu v a h r3 "r \ " 13 also invade leukocytes. Molecular biological techniques may soon play a more important role in the classification and the development of phylogenetic trees for apicomplexan parasites. The phylogenetic relationship between Theileria and other apicomplexan parasites has been established and is based on the sequence comparison of the small sub-unit ribosomal RNA molecules (Barta et al., 1991; Gajadhar et al., 1991). The current classification of the Theileria genus is as follows: Subkingdom Phylum Class Subclass Order Family Genus Protozoa Apicomplexa Sporozoea Piroplasmia Piroplasmida Theileriidae Theileria Classification according to Irvin (1987) and Irvin and Morrison (1987). 1.1.3. The distribution of the Theileria genus and their vectors. The genus Theile ri a is found across all continents but is less common in the Americas (Uilenberg, 1981). The economically more important species which infect cattle can be distinguished by the morphology of their piroplasms and schizonts, by serological differences detected by indirect fluorescent antibody tests (Kimber et al., 1973), and differences in their geographical distribution, pathogenicity and vector species. There are five economically important Th eile ri a species and due to differences in treatment and prevention of disease it is important to be able to distinguish between the species. T. parva is found in Eastern and Central Africa where it is transmitted by Rhipicephalus species ticks and is the cause of East Coast Fever, Corridor Disease and January disease. T. mutans is also found in Africa, is transmitted by ticks of the Amblyomma genus and causes Benign African theileriosis I. In Africa one can also find T. taurotragi, which 14 Figure 1: The distribution of Theileria annulata. The parasite is endemic in the shaded areas of the map of the world. This figure was adapted from Neitz (1957). 1c causes Benign African theileriosis II, and like T. parva, is transmitted by ticks of the Rhipicephalus genus. Diagnosis is made more difficult by another species, T. velifera, which is non-pathogenic but also occurs in Africa and is transmitted by ticks of the genus Amblyomma. The mammalian host for all these parasites is cattle, although some species also infect a variety of native mammals. In general, the symptoms of the disease are more profound in exotic and cross-bred cattle than in native breeds. This is a major constraint on the improvement of milk production in these countries. T. annulata the causative agent of tropical theileriosis has a much wider distribution; it is found in Southern Europe, Northern Africa, Egypt to the Sudan, the Middle East, India, parts of the former Soviet Union and southern China. T. annulata is transmitted by ticks of the genus Hyalomma, and depending on the geographical location, different species are of importance. For example the most important tick vector in India is Hyalomma anatolicum anatolicum, while in the former Soviet Union it is H. scupense and in North Africa it is H. detritum (Singh et al., 1986). A diagrammatic representation of the world wide distribution of T. annulata is shown in Figure 1. Another Theileria species which infects cattle and is of economical importance is T. sergenti. The disease associated with this parasite is called Oriental theileriosis, and is found in the eastern part of the former Soviet Union, Japan, South Korea and eastern China. It is transmitted by ticks of the genus Haemaphysalis. Other Theileria species which are also of economical importance but do not infect cattle are T. hirci and T. camelensis. The former is transmitted by ticks of the genus Hyalomma and causes disease in sheep and goats while the latter infects camels but its arthropod host is unknown. These Theileria species, their tick vectors, distribution and the disease they cause are listed in Table 1 together with the relevant references. 1.1.4. The life cycle. T. annulata has a complicated two-host life cycle which is typical of the whole Theileria genus. A simplified diagrammatic form is shown in Figure 2. It involves three main phases of multiplication. These are sporogony, after sexual reproduction in the invertebrate host, followed by two asexual phases of multiplication in the vertebrate host: 16 \ 16 15 %4 ýb ý Tick Salivary Gland Tick Gut 2 3,, 6 Cr 'f `c; in't. "i 1 tiiýq-C tº/ eK 6 ý 6 O Figure 2: The life cycle of Theileria annulata. 1. Sporozoite in saliva of a feeding tick. 2. Sporozoite invading a leukocyte. 3. Macroschizont. 4. Macroschizonts divide in synchrony with host cells. S. Merogony, release of merozoites. 6. Merozoite. 7. Merozoite invading an erythrocyte. 8. Piroplasms. 9. Release of ovoid stages from blood masses in tick gut. 10. Microgamete. 11. Macrogamete. 12. Zygote. 13. Infected gut epithelial cell. 14. Kinete. 15. Kinete invades salivary gland cell. 16. Asexual reproductionof kinete. 17. Sporoblast. 18. Release of sporozoites. (This figure is adapted from: Mehlhorn and Schein, 1984; Dolan, 1989; Tait and Hall, 1990). Bovine Host 17 3 schizogony and merogony. The life cycle of T. annulata has been studied by light and electron microscopy, and the following account is based on observations and descriptions by Uilenberg (1981), Fawcett et al. (1982), Mehlhorn and Schein (1984), Tait and Hall (1990) and Morzaria and Nene (1990). T. annulata is transmitted from an infected tick of the Hya lomma genus to the vertebrate host (cattle Bos taurus or Bos indicus or domestic buffalo, Bubalis bubalis), when the tick is feeding. Most vertebrates can only be infected by ticks or by artificial inoculation since no transovarial transmission has been detected. During a blood meal an infected adult tick will inject saliva containing Theileria sporozoites into the blood stream of the bovine host. Once in the blood, the sporozoites rapidly invade their host cells. Generally, the target cells for T. annulata are leukocytes, such as B cells and macrophages (Glass et al., 1989). The invasion process and the target cells are discussed in more detail in sections 1.7.2 and 1.7.3 respectively. The invasion process is very rapid, taking around three minutes in vitro (Jura, 1984). During the invasion process the sporozoite is enclosed by the host cell membrane. Subsequent to invasion this membrane is broken down, probably by secretions of the sporozoite, and finally the sporozoite comes to lie in the cytoplasm where it develops into the trophozoite. This process is distinct from other apicomplexan parasites such as Eimeria and Plasmodium, where the sporozoite is retained in a parasitophorous vacuole. The process from beginning of invasion until development of the trophozoite is completed within 30 minutes in vitro (Jura et al., 1983). The trophozoite enlarges by ingesting the cytoplasm of the host cell, and undergoes repeated nuclear division which results in the formation of the macroschizont. The macroschizont then transforms and immortalises its host cell. The macroschizont can first be detected approximately three days after infection (Mehlhorn and Schein, 1984). A typical macroschizont contains six to eight nuclei, but in some cases many more have been observed (Kurtti et al., 1981). The process by which the macroschizont immortalises its host cell is only poorly understood, but elevated levels of some transcription factors (Ap-1, Jun, 18 NF1 and NF-icB) have been found in infected cells (Baylis et al., 1995; Ivanov and Williams, 1991). Changes in phosphoprotein and protein kinase activities have also been detected (Dyer et al., 1992). Possible mechanisms for transformation of the host cell are discussed by Dyer and Tait (1987). The process of synchronous division of host cell and macroschizont is called schizogony, which is the first phase of asexual multiplication in the bovine host. The next phase of asexual multiplication for T. annulata in the bovine host is called merogony, and this phase follows schizogony. After 8-10 days merozoites start to form from the nuclei at the periphery of the schizont. They appear "rosette-like" and are termed microschizonts; they can be seen under a light microscope. The merozoites develop rhoptries and an apical polar ring and they bud from the surface of the schizont in a synchronous manner (Shaw and Tilney, 1992). Free merozoites in the bloodstream of the bovine host contain a rhoptry complex, a mitochondrion, ribosomes and a nucleus. They are 1-2 µm long and their outer surface consists of a cell membrane and two closely apposed inner membranes (Mehlhorn and Schein, 1984). Once in the bloodstream, the -merozoites rapidly invade erythrocytes. In an infected cow up to 90% of all erythrocytes can be infected. The process of invasion of erythrocytes by merozoites is only poorly understood, but is thought to be mediated by ligand-receptor interactions as demonstrated in malaria (Kawamoto et al., 1990). After the invasion the erythrocyte membrane enveloping the merozoite is disintegrated, probably by secretions from the rhoptries, and the parasite differentiates into the piroplasm form. Two different forms of piroplasms have been observed: a) slender comma shaped forms or b) ovoid or spherical forms (Mehlhorn and Schein, 1984). The frequency with which these two different forms are observed varies for different Th eileria species, e. g. for T. annulata both forms occur in approximately equal numbers whereas in T. parva 80% of all piroplasms are comma shaped. It has been proposed that another cycle of division is linked to the comma-shaped piroplasms. These are thought to divide by binary fission and this nuclear division is associated with cellular division. As a result ,n no multi-nucleate schizont-like stages can be observed. This process has been described in vitro by Conrad et al. (1985), and similar forms were also seen in infected cattle. The merozoites released from erythrocytes are identical to those released from schizont-infected lymphocytes and can re-infect erythrocytes. The relative importance of the two forms for the maintenance of infection is unknown. When a tick feeds on an infected host, piroplasms are taken up during the blood meal and so transmission between vertebrate and invertebrate hosts occurs. In the gut of the tick, the spherical piroplasms develop into microgametes. These can be observed 2-4 days after the tick stops feeding (Mehlhorn and Schein, 1984). Larger gametes can also be found in the gut of the tick and these gametes are called macrogametes. These two types of gametes fuse to form the zygote, which develops into a kinete; this is a club-like, uninucleate motile stage which in turn invades the gut epithelial cells of the tick. Shortly before the molting of the skin of the tick, kinetes of T. annulata can be observed to migrate in the haemolymph on their way to the salivary gland. Once a kinete reaches the salivary gland it infects salivary gland cells and is thought to lie dormant until the tick starts feeding (Mehlhorn and Schein, 1984). The kinetes are found in the cytoplasm of the gland cells where they undergo a series of differentiation and multiplication steps, after the tick starts to feed. Three to five days after feeding is initiated the sporozoites are formed which, when transmitted to the bovine host, are infective and complete the life cycle of T. annulata. In total, about 50,000 sporozoites can be produced from a single infected salivary gland cell. 20 1.2. Tropical Theileriosis. 1.2.1 The disease. Worldwide there are 250 million cattle at risk of tropical theileriosis. In endemic areas, indigenous cattle are usually infected as calves and tend to develop only mild symptoms and recover readily. After this recovery they exhibit protective immunity to further infections but are persistent carriers (Brown, 1990a). In recent years, attempts to increase the milk yield in Third World countries have resulted in the introduction of cross-bred and exotic cattle. These were found to be highly susceptible to tropical theileriosis and exhibit a mortality of 40-60 %. Tropical theileriosis is therefore of great economical importance due to the constraints it places on the increase of milk production in endemic areas. The epidemiology of tropical theileriosis is described by Uilenberg (1981). In subtropical areas the disease has a seasonal character, and most cases are found during the summer months when the ticks by which the disease is transmitted are most active. As would be expected the seasonality of infection is less pronounced in the tropics. An important characteristic of the disease, which could make its eradication virtually impossible, is that exotic and cross-bred cattle, once recovered, remain healthy carriers. These animals thus act as a reservoir for future infections. Another reservoir host is the Asian water buffalo, which only experiences mild disease symptoms and also remains a carrier. The description below of the pathology of the disease is based on papers by Neitz (1957), Laiblin (1978), Uilenberg (1981), Mehlhorn and Schein (1984) and Tait and Hall (1990). The severity of the disease varies, depending on the susceptibility of the animal, the virulence of the parasite strain and the number of the sporozoites with which the animal is infected. In general, an infected animal develops a fever about 2 weeks after the tick starts to feed. The onset of fever in mechanically- infected animals is usually delayed. The symptoms of a typical acute infection are as follows. Two days before the onset of the fever, (which is usually above 410C), schizonts 21 can be detected in the bloodstream of the infected animal. The fever is accompanied by swelling of the superficial lymph glands, closely followed by swelling of the regional lymph nodes draining the sites of infection. Other symptoms are the cessation of rumination, drooling from the mouth, swelling of the eyelids, accelerated pulse and breathing, general weakness, a reduction in milk production and diarrhoea. After prolonged infection, blood and mucus can be observed in the faeces. The red blood cell count also drops from 7 to 3 million cells per mm3. The animal becomes markedly emaciated. If the animal continues to feed and the erythrocyte count recovers then the animal has a good chance of recovering completely. If, however, the erythrocyte count does not recover this will lead to severe anaemia (an erythrocyte count of less than 1 million per mm3) and the animal will die, usually 8-15 days after the onset of the disease. It is not clear how disease progression is related to parasite development. However, the main pathogenic effects occur during the intra-lymphocytic stage i. e. schizogony, when the infected lymphocytes multiply. It has been suggested that these displace uninfected lymphocytes in the lymph node tissue and thereby induce symptoms similar to leucosis (leucocyte depletion). Haemolytic anaemia, which can be observed during the later stages of infection, occurs during merogony and it has been shown that up to 90 % of all erythrocytes can be infected. The severe anaemia is more likely to be due to phagocytosis of infected erythrocytes than to direct lysis of the infected cells by the parasite. Parasite induced auto-immune responses may be another cause of anaemia, as suggested by Uilenberg (1981). The symptoms for other Theileria species vary. For example in T. mutans, the pathogenesis is predominantly due to the erythrocytic stage, while in T. parva piroplasm replication is not found and haemolytic anaemia is uncommon (Morzaria and Nene, 1990). The disease is usually diagnosed by evaluation of the clinical symptoms. The diagnosis can be substantiated by detection of macroschizonts and/or piroplasms in Giemsa stained blood or tissue smears. Other techniques are being developed, especially for locations where more than one Theileria species is endemic. Such techniques could involve indirect immunofluorescent antibody (IFA) tests and 22 enzyme-linked-immuno-sorbent-assays (ELISAs) against macroschizonts and piroplasms, or techniques based on Southern blotting or PCR (Allsopp et al., 1989; Katende et al., 1990; Ben Miled et al., 1994) 1.2.2. The immune response of the bovine host. The immune responses of the bovine host to T. annulata infection have been reviewed by Hall (1988), Tait and Hall (1990) and Brown (1990). Protective immune responses against tropical theileriosis have been observed after an animal has overcome the infection following either natural or mechanical infection or immunisation with attenuated macroschizont-infected cell lines (described in section 1.3.4. ). The immunity, if not challenged subsequently through natural infection, lasts up to three years. The host's immune responses to the individual life stages of the parasite are discussed below. The sporozoite. Immunity against sporozoites is not essential for a protective immune response, since cattle infected with attenuated macroschizont- infected cell-lines develop a protective immune response which does not recognise the sporozoite stage (Brown et al., 1990). However, humoral immune responses against the sporozoite stage have been detected (Gray and Brown, 1981). They show that serum from immune cattle is capable of neutralising sporozoite infectivity of lymphocytes in vitro. This was confirmed by Preston and Brown (1985). Subsequently, monoclonal antibodies were raised against surface molecules from sporozoites and some of these were found to prevent sporozoite infection of lymphocytes in vitro (Williamson, 1988; Williamson et al., 1989). Two of these antibodies, 1A7 and 4B11, and the antigens they bind to will be discussed in section 1.4. These findings imply that if sufficiently high antibody titres were present in cattle they might prevent infection. Data to support this theory were obtained from an immunisation trial using a part of a recombinant sporozoite antigen expressed in the el loop of Hepatitis B core antigen (Boulter et al., 1995). 23 Similarly, immune responses against T. parva have also been found to be directed against the sporozoite stage. So far only humoral factors capable of neutralising sporozoite infectivity in vitro have been identified (Musoke et al., 1982; Dobbelaere et al., 1984; Musoke et al., 1984). These immune responses in T. parva target p67 (Nene et al., 1992) and the 104 kDa microneme-rhoptry protein (lams et al., 1990). The genes encoding both these proteins have been cloned and sequenced. Immunisation trials using recombinant p67 resulted in the protection of the majority of immunised cattle (Musoke et al., 1992; Musoke et al., 1993), indicating humoral immune responses against the sporozoite could be protective in vivo, although the neutralisation titre does not correlate with protection. The macroschizont. The macroschizont stage seems to be the most important target for protective immune responses, as immunisations of naive cattle with attenuated macroschizont-infected cell lines induce protective immunity (Hall, 1988; Brown et al., 1990). This immunity seems to be predominantly a cellular response employing a variety of lymphocyte sub-populations, while the humoral response, though observed, was not found to be of importance (Pipano, 1977). It was shown that cytotoxic cells directed against the macroschizont can be found in animals which recover, or have recovered, from tropical theileriosis whereas none were found in animals which died of the disease (Preston et al., 1983). Both cytotoxic T cells as well as NK cells were postulated to be involved in this immune response. Subsequently it has been shown that the immunological memory contains cytotoxic T cells which are directed against surface antigens of the macroschizont stage (Preston and Brown, 1988). They also found macrophage-mediated cytostasis to be of importance and this response could be observed consistently after both immunisation and challenge. Three origins of the antigens recognised by the immune system have been postulated (Hall, 1988): a) they could be parasite antigens which are expressed on the surface of the parasitised cell; b) they could be surface antigens of the host which have been altered by the parasite; 24 or c) they could be host antigens which are not normally expressed. Attempts have been made to isolate antigens from the surface of macroschizont-infected cell lines. The method chosen was to raise monoclonal antibodies against surface molecules of macroschizont infected cell lines (Shiels et al., 1986). One of these monoclonal antibodies, 4H5, binds to a 95-120 kDa surface molecule of macroschizont infected lymphocytes (Shiels et al., 1989). 4H5 lyses macroschizont infected cells in a complement-dependent manner and suppresses infected cell proliferation (Preston et al., 1986). Unfortunately no data is available to indicate whether this antigen induces cell-mediated immune responses to the macroschizont stage, but it is worth studying this antigen further with regard to inclusion in a recombinant vaccine. In T. parva, it has been shown that the most important immune response against East Coast Fever is cellular (Morrison et al., 1989). Some of the antigens recognised by the bovine immune system have been isolated and they are reviewed by Morrison et al. (1989). Both cytotoxic and T helper clones, which react with macroschizont-infected lymphocytes, have been derived from T. parva-infected animals (Brown et al., 1990). When these cytotoxic T cell clones were analysed they were found to be either parasite strain specific or cross-reactive. This indicates that the T cell epitopes seen by the cytotoxic T cell clones can be conserved but can also be polymorphic (Brown et al., 1990) The merozoile/piroplasm. Serum from immune cattle does not appear to react with infected erythrocytes but a humoral response to merozoites and piroplasms was found (Shiels et al., 1989). The antigens recognised by the immune serum can be immuno-precipitated and seem to originate from the merozoite stage. The response of the bovine immune system to these antigens has not yet been elucidated. If the immune system can decrease the number of merozoites it might be able to reduce the anaemia associated with the disease and also reduce the piroplasm transmission to the tick. Until recently it was difficult to isolate sufficient numbers of merozoites to either study their invasion of erythrocytes or the effect of UNIVERSITY OF YORK LI8RARY 25 the bovine immune system on their survival. However, merozoites have now been produced in vitro by culturing macroschizont-infected lymphocytes at 41°C (Glascodine et al., 1990). Subsequently, a surface antigen of the merozoite and piroplasm stage was isolated (Dickson and Shiels, 1993). The role of this antigen, and its interaction with the bovine immune system, however, still remain to be investigated. The tick. Little is known about the bovine immune response to the tick vector of T. annulata. However, it may be important to investigate the immune responses to the tick; generally, ticks remain attached to their host for at least a week, which is sufficient time to allow the immune system to raise an adequate response. Innate and acquired immune responses to ixodid ticks, to which the Hyalomma genus belongs, have been described (Wakelin, 1984). It has been shown that tick bites induce an inflammatory reaction and in immune hosts elicit a rapid immune response which might prevent the tick from feeding and may even result in its death. The immune response of guinea pigs to the ixodid tick Amblyomma americanum was studied by Brown and Askenase (1983). They found an inflammatory response at the site of penetration of the mouth parts of the tick. Initially the bite site was infiltrated predominantly by neutrophils and after 3-5 days by basophils and eosinophils. The recruitment of basophils and eosinophils was induced by sensitised T cells and IgGI antibodies produced in response to A. americanum antigens. This immune response seemed to be typical for a number of ixodid ticks. If an immune response to ticks of the Hyalomma genus could be induced in cattle, which would prevent ticks from feeding, it could stop the transmission of T. annulata. 26 1.3. Treatment and Control. There are three main types of control for tropical theileriosis and East Coast fever: chemotherapy, vector control and vaccination. These control measures are reviewed by Uilenberg (1981), Irvin and Morrison (1987), Dolan (1989), Brown (1990), Morzaria and Nene (1990), Musisi (1990) and Tait and Hall (1990). The following account of present control measures for tropical theileriosis is based on these reviews. 1.3.1. Treatment of Tropical Theileriosis. Chemotherapy has been used on a large scale to treat T. parva infection. It has also been used as a component in the immunisation of cattle using the "Infection and Treatment" method (see section 1.3.4. ). In T. annulata, on the other hand, chemotherapy has hardly been used, although it is very effective. This is due to high costs and the availability of attenuated cell line vaccines (see section 1.3.3. ) The earliest drug used was chlorotetracycline. During the last 15 years many new drugs have been tested, some of which were found to be promising. These drugs were tested by screening for anti-theilerial activity in in vitro cultures (Brown, 1989). These tests were followed by in vivo trials. Drugs which were found to be successful are: the anticoccidial drug halofuginone, a napthaquinone menoctone and the menoctone analogues parvaquone and buparvaquone. Menoctone and buparvaquone were found to be active against T. annulata and T. parva macroschizonts in culture (McHardy, 1978; McHardy et al., 1985). Subsequently, field trials have shown that buparvaquone is a highly effective therapeutic agent against both T. annulata and T. parva infection in cattle (McHardy, 1991). Interestingly, drugs which are effective in the treatment of Babesia, Eimeria and Plasmodium were ineffective against Theileria (McHardy, 1978). It was postulated that these drugs, proguanil, diaveridine and chloroquine, fail to penetrate the infected lymphocyte. 27 1.3.2. Vector control. One of the methods used for controlling tropical theileriosis and East Coast fever is by minimising tick infestation. The current method for reducing the tick burden is by either spraying or dipping cattle in acaricides such as butocarb or amitraz. The advantage of this method is that it disposes of all ticks regardless of which parasitic infection they carry. Therefore, one can reduce the incidence of Theileria infection as well as other tick-borne diseases, e. g. heartwater and babesiosis by dipping cattle with acaricides. Dipping or spraying of cattle is usually done once a week since the acaricide residues are sufficient for about four days and sporozoite transmission only occurs three days after tick attachment (Urquhart et al., 1987). This method is more commonly used for the prevention of East Coast fever rather than tropical theileriosis. There are some disadvantages associated with this method. For example one will not be able to eradicate Th eile ri a by spraying or dipping cattle in acaricides since there are other reservoir hosts. Therefore this method will only permit the disease to be kept under control. Further, it is a relatively expensive method for Third World countries if it is used on a large scale. For this technique to be effective, farms need to be highly organised and rigorous as any lapse in treatment could lead directly to outbreaks of infection. Another problem is the development of resistance of ticks to the acaricide over prolonged use. Other methods of controlling infection are the separation of infected cattle from uninfected susceptible cattle, and prevention of infected cattle moving into areas where the disease is otherwise not found. Cattle have been kept in zero grazing environments and thereby the cattle were isolated from infected stock, but this is a very expensive method and has been used only in very rare circumstances. In Morocco, it was found that some ticks behave as barn ticks and these hibernate in cracks in the walls of sheds in which the animals are kept. In these circumstances, spraying the sheds with acaricides proved effective. 28 1.3.3. Vaccination with attenuated macroschizont- infected cell lines. The use of attenuated macroschizont infected cell lines as vaccines is reviewed by Pipano (1981), Hall (1988) and Brown (1990). This method is very efficient and is widely used in the prevention of tropical theileriosis. The development of this method was possible after continuous cultures of macroschizont-infected cell lines had been established in vitro (Hulliger et al., 1964). These infected lymphocyte cell lines became attenuated during their prolonged culture in vitro. This means that if these cells are injected into cattle they are less virulent, producing milder clinical symptoms and a lower parasitaemia after increasing time in culture. A cell-line is fully attenuated when it no longer produces any piroplasms when injected into cattle. The attenuation of T. annulata infected leucocytes is achieved in between 20-300 passages (2 months to over 2 years) in culture depending on the T. annulata isolate. The mechanisms involved in attenuation are not yet understood but generally, when attenuated, the parasites lose their ability to differentiate into piroplasms (Pipano, 1989) and in some cases changes in the expression of proteases occur (Baylis et al., 1992a). Once attenuation has been achieved, 106 - 107 infected cells are usually used per inoculation (Hall, 1988). Normally one immunisation is adequate to induce cross-protective immunity since the immunity is reinforced by subsequent tick challenges in the field. However, Friesian cattle usually require a second immunisation from a heterologous schizont stock of a lower passage culture to provide adequate protection (Pipano, 1981) The attenuated macroschizont vaccine was first developed in Israel over 20 years ago (Pipano, 1981) and has drastically reduced tropical theileriosis there (Brown, 1990). Subsequently, this vaccine has become established in India where it has also proved very effective (Singh, 1990). It is currently on trial in Iran, Russia, Morocco and Turkey (Hall, 1988). So far, it has not been possible to develop an attenuated vaccine for T. parva. It has been postulated that the reason for this is failure of transfer of schizonts into lymphocytes of the host itself (Dolan, 1989). Musisi (1990) suggested that the BoLA mis-match inhibits the development of immunity in T. parva but not in T. annulata. 29 Unfortunately, this method is not without its disadvantages. For example, a "cold chain" is required. This means that the macroschizont- infected cells need to be kept at 40C or frozen between the laboratory, where they were generated, and the site of the immunisation. This might pose a problem for immunisations in areas which are not easily accessible (Tait and Hall, 1990). Another potential problem might be the infection of the immunised cattle with other pathogens. As the macroschizont-infected cell lines were established from other animals, these could have been infected with other pathogens, or subsequently the cell cultures could be contaminated, i. e. in the laboratory. Another risk of vaccination with attenuated macroschizont-infected cell lines is that the parasite could become virulent again. This has not yet been detected but if it did happen it could result in the infection of susceptible cattle. Another drawback is that the immunisation of cattle with attenuated cell lines does not protect the cattle from becoming healthy carriers following subsequent tick challenges in the field (Tait and Hall, 1990). This technique will therefore never eradicate tropical theileriosis. 1.3.4. The infection and treatment method. The infection and treatment method is reviewed by Brown (1990) and Morzaria and Nene (1990). In this immunisation method, cattle are infected using either virulent sporozoites, from a cryo-preserved stabilate, or infected ticks and subsequently are treated with chemotherapeutic agents during the latent period. The drugs used tend to be tetracyclines and more recently buparvaquone. Simultaneous administration of drugs and parasites have also been tested. This resulted in reduced symptoms and cross-protective immunity was usually still found. This method is used on a large scale with T. parva in Burundi, Kenya, Malawi, Rwanda and Zambia since no attenuated macroschizont vaccines are available. This method is rarely used in T. annulata since it tends to be more expensive than attenuated cell vaccines. The main disadvantage of this method is the high cost of the chemotherapeutic drugs for Third World countries. There are other practical disadvantages, as this method needs to be administered by a qualified veterinarian and the cattle become carriers. There is also a 30 danger that the parasite may be passed on to uninfected cattle (Musisi, 1990; Mozaria and Nene, 1990). This method also shares two disadvantages with the attenuated macroschizont vaccine, i. e. the "cold chain" and infection with other pathogens. Another problem is that the infection and treatment method fails to induce cross protective immunity. Therefore a "cocktail" of several parasite strains are included but as a result one might introduce parasite strains to regions where they were not originally found (Musisi, 1990). It is also expensive to titre a dose of stabilate as it has to be tested on cattle and there is a large error range attached to dose evaluation (Musisi, 1990). 1.3.5. Recombinant sub-unit vaccines. There is no sub-unit vaccine available for the prevention of either tropical theileriosis or East Coast fever, but a large effort has been made to develop such vaccines. The target for such a vaccine is to induce cross-protective immunity, to be inexpensive so that people in the Third World can afford it, to be easy to administer (ideally in a single dose) and to have a long shelf life. It would have advantages over the other preventative measures, since a cold chain would not be necessary, the product would not be contaminated with other pathogens and it could not revert to virulence. So far a number of antigens of both T. annulata and T. parva have been isolated from several life cycle stages but small vaccination trials have only been conducted with three sporozoite antigens. p67, a sporozoite surface antigen of T. parva, was the first candidate to give some positive results. Most cattle immunised with p67 were protected against homologous challenge (Musoke et al., 1992; Musoke 1993). Similarly, positive results were obtained by Boulter et al. (1995) using a part of the cloned gene encoding the surface antigen SPAG-1 of T. annulata (this antigen is discussed in more detail in section 1.4.1. ). Another trial using part of the SPAG-2 antigen (a different surface antigen of T. annulata sporozoite is discussed in section 1.4.2. ) did not result in any detectable protection (Knight personal communication). Some of these trials indicate that sporozoite antigens might induce protective immunity, but it seems that a sub-unit vaccine containing 31 antigens from all parasite life-cycle stages might be more effective. If the vaccine is solely based on the sporozoite stage, some sporozoites might escape the immune responses due to their rapid invasion of host cells. If a sporozoite manages to escape the immune response it has the potential to develop into a macroschizont and subsequently into thousands of merozoites. The inclusion of antigens from different life cycle stages in the multi-stage vaccine might help to overcome the parasites at later stages of the life cycle. This might also reduce the probability of selecting polymorphic parasite strains since variation must arise in more than one of the vaccine components. 1.4. Sporozoite Antigens. So far only one sporozoite antigen of T. annulata has been cloned and sequenced completely, while another has been only partly cloned and sequenced. These two antigens have been studied in this thesis. The relevant background of both antigens is described below. 1.4.1. The sporozoite surface antigen SPAG-1. A panel of mouse monoclonal antibodies was raised against the sporozoite stage of T. annulata (Williamson, 1988). One of these antibodies, 1A7, was found to block invasion of sporozoites into purified bovine peripheral blood mononuclear cells in vitro by 66% (Williamson, 1988 and Williamson et al., 1989). IA7 was shown to react with the surface of sporozoites using an indirect fluorescent antibody test, and binds to a protein (called SPAG-1) which is only expressed at the sporozoite stage. This protein is completely absent during subsequent life cycle stages of the parasite (Williamson et al., 1989). The protein recognized by 1A7 has a molecular weight of 104 kDa and is processed into forms on the mature sporozoite with molecular masses of 85 kDa, 70 kDa, 63 kDa and 54 kDa, which are recognised by 1A7 on Western blots (Williamson et al., 1989). 1A7 was used to screen a ? lgtl1 expression library containing genomic T. annulata Hisar DNA. A clone, called Xgtll-SR1, containing a 330 bp insert was isolated and sequenced (Williamson, 1989). SRI was used to 32 screen a cDNA library containing T. annulata Hisar DNA and a clone of 2.8 kb was isolated and sequenced. This clone contained the full length SPAG-1 gene (Hall et al., 1992) whose sequence is shown in Figure 3. The SRI region maps to the C-terminus of SPAG-1. The 1A7 epitope has been mapped to 16 amino acids within the first 45 amino acids of SRI (Boulter et al., 1994). SPAG-1 has been shown to be expressed on the surface of the sporozoite via immuno-gold electron microscopy (Knight 1993). Sequence comparison of the predicted SPAG-1 amino acid sequence to p67, a sporozoite surface antigen of T. parva, revealed an overall identity of 47% (Nene et al., 1992 and Katzer et al., 1994). Another interesting finding was that the predicted SPAG-1 amino acid sequence contains two blocks of repeats as shown diagrammatically in Figure 4a. SPAG-1 contains a total of 17 repeats of the PGVGV motif, which is identical to the repeats found in bovine elastin (Hall et al., 1992 and Hall, 1994). Bovine elastin has 11 tandem repeats of this motif. An alignment of the first block of the Theileria repeats and the bovine sequence is shown in Figure 4b. It has been speculated that the importance of these repeats lies in the evasion of the hosts immune system. The blocking data of the 1A7 monoclonal antibody gave the first indication that SPAG-1 is a candidate antigen for the inclusion is a sub- unit vaccine. Further support of the importance of this antigen as a sub- unit vaccine candidate was provided by data from a vaccination trial using the SRI region of SPAG-1, expressed in the el loop of the hepatitis B core antigen (Boulter et at., 1995). This trial indicated that SPAG-1 induces protective immune responses against the parasite. 33 1 CTTTTMAAAGCACTMCATTUGAATTTMATTTCATTTTCCMCACTCAAýY: ATCAATATTA7'ACACTTfCTGTTCACCATTCCGGCTA 90 MNI1 11 FLLTIPAI 91 TT7'TTCTATCTCGAGCGGACMCATC: CC7CCGGGACAMGTTCTAC: AACCTCTAAACCCAGTCCCCTACTMCCCTAGAATCGCCCCTAA 180 FVSGADKMPACESSRTSKPS 1' 1. VTLESAVT 181 CACAACCTTCAAAGGACCCATTCAAGACAATTACTGCCTfGTCAAAAGCAACAAAACTAI'GGAACTCAGCCGTATCAGTATCAGCTGACT 270 QPSKDPFKTISALSY. ATKVWY. SAVSVSGDS 271 CTAAGACTGTACCTACTCCACTTTCGGAACCAATCATCACTCCATCTTTTCAAGAACCACTA7'C"fCMGMCTTGMTTCCAATCAGATA 360 KTVPTPVSEPMITRSFQEPVS0ELEFQSDT 361 C'ICAAATTAATCACTCACCATCCGGTTCAGATCAGGATGAGCATGACCATGACGATGAGGAGGAACMGMCACCATAAATCTACCTCAT 450 EIN 6E Sc SCSDEDEDDDDDEEEEEDDKSTSS 451 CTAAAAACGGAAAAGGCAGCCCAAAAGCTCAGCCTGGACTATCTTCAAGCACTACATCCTCACCAAGTCCAACATCTCCAACTACAACAT 540 KNGKGSPKAQPGVSSSSTSSASPTSPTTTL 541 TATCACAAACTGGAT'lGCGACCAAGTGGTTCCCACGCTCAACAACATCCCCCTCTAGG7GTTCCAGGACTTCGTCTTCCACCACTACCTG 630 SQTGLGPSCSHAQQDPGVGVPGVGVPGVGV 631 TTCCAGGACTAGGTGTTCCAGGAGTAGGTGTTCCACG7GTAGGTCT000AGGTGTAGGAGGTGTTCCCGGAGTTGGCGTTGCACCAGGGG 720 PGVGVPGVGVPGVCVPGVCGVPGVGVAPGV 721 TAGGTGTTCCAGGACTTGGTG CACCAGGTG AGGTGTTGGAGCTGATA TAG TTGCCTGGAAGTGGTGGTC AGCA., GAG 810 GVPGVCVAPGVCVG JA DSSCLPGSGCLCAGA 811 CAAAGGC7C', GGAAAGGTCAAGGATCTOGTCTACAGGGACCACGAGGTGTTGGACTAGTACC7GCTGTACG7GTAGCAGCTTCTTCTTCTT 900 KACKCQCSCLQCPCGVCV VIP CVG VIA ASSSS 901 CACCAGGAAAACCTCCAGGAGTAGGAGCAGGACTTAT000TGGAGTTCt-'TGTACCAGCACAAGGAGGAGTAATAATTGCTCCGCCAGCAG 990 G -VP GVGVAPGVG -V7-1 PGKPPGVGAGVM IP GVGVRAQGGVIIGAYGv 991 TAGCAGGTGTGCCAGGAGGAAACCCAGCACAACCAGTATCTCAAGAACTTGAACTGAAATCAGACACTGAAATTAATGAGTCAGGTTCCA 1080 AGVPGGKPGQPVSQELELKSDTEI Nb ESGSS 1081 GTTCAGAAGGGGAAGACGATGACCATGAAGAAGAGGAAGAAGAAAATAAATCTACCTCATCTAAAGCAGCAGGAGGAAAGGCTGCAAAAG 1170 SEGEDDDDEEEEEENKSTSSKGAGGY. AGKG 1171 GTCAAGGATCTGTATCACCAGGAGGAGGATCCTCAGCAACTCAAACATCTCCAACTACAACACCACAATCTGGCTTGGCATCAAGTGCTT 1260 QCSVSPCGGSSASQTSPTTTPQSCLASSCS 1261 CTCATGCTCAACAAAGTCCTCAACAAGATCCAGCGCCTAGTAAACCTAGTGCAGGAGGTCTCCCAGCAGTTGGAGTTCC7GGTCTTGGCG 1350 HAQQSPQQDPAPSKPSGGGV -PG VGVPGVGV 1351 TTCCCGCTGTTGGAGTACCAGGACTAGGAGTTGCGCCGGGACTTGGTGTTGTACCTGGAGTAGGGC. GTGCAACAACTTCTTCATCATCAA 1440 IPGVGVPGVGVAPGVGVVPGVGG IA TTSSSST 1441 CAACTTCAACTTCAACTTCAACTACTACTACTACTACAACTTCATCACGAAAACCTTCAGACCAAGGAACCCATGCfACTTCTCCAAGAA 1530 TSTSTSTTTTTTTSSGKPSDQGSHGTSPRN 1531 ATGCACTAACCAGACAAACTGACTCAATATCAGGA000ATACCATCACCAGGAGATCCAAGAGCAATTACTCGACAAATCGGTGAAGCAG 1620 AVTRQTDSISGPIPSPGDPRAITGQMGEGE 1621 AAAGGTTTGCTGTACACTTCCTGGGAGATTTTAAACCAAAACCAAGGAGATATGAAGGACAAGGAACAGATGCAGTAAAACTAAAACAAT 1710 RFAVQFLGDFKPKPRRYECQGTDAVKLKQF 1711 TCATTTTCGAAGAGGTCAAATCGCTGGTGCAAACCTTAATAAACCTTAAATTAGCAATTGCAAACGACTTTGTTGAAATCACTGAAAACT 1800 IFEEVKSLVQTLINLKLAIANDFVEISEKL 1801 TGAAAAAGAAAAATCAAAATTACGTACCGAAATTAAAGTTCTTAAAAGGAGAACAATTTGACACCAAACAGAAGGTACCCAACGTACTAA 1890 KKKNQNYVPKLKLLKGEQFDTKQKVANVLK 1891 AACCGTTCAATTCTCTGTACTTCGTATTTTTTATCAACCTTAACCTAGCGAAACAAGTTAACAAACCGCAAGAATTCCCAGAATTTCTTT 1980 CFNSLYFVFFMNLNLAKEVNKPEELAEFLN 1981 GGAAACTAAATACAATCCCAGATAAAGTAGGAAGAGAATT'IGAGTTACCAATAGAAAAAACTAAAGGTTCAGAGAAAAAGAAGGAATTAG 2070 KLNTIPDKVGREFELAIEKTKGSEKKKELE 2071 AAGAAGCATTTAATTCAATAGGGTTAGGTTTCAAAATAGCACACTACGCAACAAATCACATCCTCTCAAGTATAACAAATTCAGTCTACT 2160 EAFNSIGLGFKIAQYATNDILSSITNSVYS 2161 CCCTGATAAAACTAAAGAATTTTGGAGATCATTTTGTTACCGAACTAAGAAAGTCACTGCAAATGGTTCCACACCAAAACAACCTAAACG 2250 LIKLKNFGDDFVTEVRKSLQMVPHQKNL Nh G 2251 GATCAGCATTTATAGTCAAAATCTCAGAAATAATCAACAAAAAAGGAACAGAAGATCAGGATCAAACATCAGGAAGIGGGTCAAAAGGAA 2340 SAFIVKISEIINKKCTEDQDQTSGSGSKCT 2341 CAGAAGGAGGATCACTAAGGCCGCAAGATTTGACAGAAGAAGAAGTTTTGAAACTTCTGCATGAACTAGTGAAGGATCTAACCGAAGAAC 2430 EGCSLRGQDLTEEEVLKVLDELVKDVSEEH 2431 ATGTTGGAATAGGACATTTAAGTGACCCAACTAGCACAACACCAAATGCAAAACCAGCCGAACTTGGACCTTCACTAGTGATACAAAATG 2520 VCICDLSDPSSRTPNKPAELCPSLVIONV 2521 TACCGTCAGACCCCTCAAAAGTGACACCAACACAGCCTTCAAATT'IGCCACAAGTACCAACAACAGGCCCGGCGAICCGGACGGATGGAA 2610 PSDPSKVTPTQPSNLPQVPTTCPCNGTDCT 2611 CAACAACAGGACCAGGTGGAAACGGGGAAGGAGGCAAAGATTTGAACCAAGGACAAAACAAAGAACCATTATTTCAAAAGATCAAAAACA 2700 GGN PGVGVPGVGVPGVGV c 2701 AACTCTTGGGCTCAGGATTCGAAGTCGCAAGTATTATTATACCAATGACAACAATCATATTCAGCATAGTCCACTAAAACTAAAAACACA 2790 LLGSGFEVASIIIPMTTIIFSIVH 2791 ACTAACCACACTAATTTATAATATACAAAAAAAAA 2825 Figure 3: cDNA sequence of T. annulata surface antigen (SPAG-1). Reproduced from Hall et al. (1992). The predicted amino acid sequence is shown below. The PGVGV repeats showing homology to elastin are boxed. The C-terminal region encoded by the insert of Xgtll-SRI, used to isolate this clone, is underlined. Putative N-linked glycosylation sites are marked with arrowheads. --------- ------ ------------ ---- ---- ------- PGVGVPGVGVAPGVGVVPGVGG PGVGVPGVGV 34 a) [A7 b) : [PGVGV]n : amino acids 179-236 and 424-456 : SR1 : amino acids 778-891 S 179 PGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGGVPGVGVAPGVGVPGVGVAPGVGV E 335 PGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVG. VPGVGV. PGVGVPGVGV. PGVGV Figure 4: Mimicry of bovine elastin by T. annulata surface antigen SPAG -1. a) Schematic representation of the structure of the SPAG-1 polypeptide emphasizing the location of the elastin homologous regions and the SRI region. b) Sequence comparison between the most N terminal block of the SPAG-1 PGVGV repeats and bovine elastin. The VGVAPG regions are underlined. S denotes the SPAG-1 sequence and E denotes the bovine elastin sequence. 35 1.4.2. The sporozoite surface antigen SPAG-2. 41311, another monoclonal antibody raised against T. annulata sporozoites, blocks the invasion of sporozoites into host cells by about 100% (Williamson, 1988; Knight, 1993). The antigen, to which 4B 11 binds has a molecular weight of 17 kDa, as detected by Western blot analysis of mature sporozoite material. The antigen identified by 4B 11 was named SPAG-2 and like SPAG-1 it is also processed. The processed forms of SPAG-2 have molecular weights of 150,67 and 17-20 kDa as seen on Western blots (Knight, 1993). The 41311 antibody was used to screen a ), gill expression library containing genomic T. annulata DNA and a 2, gtll clone containing a 980 bp insert was isolated. This clone was called Xgtll-KP8. KP8 was partially sequenced and cloned into pGEX-IXT and called pGEX-KP8 (Knight, 1993). 41311 reacts with the GST-KP8 fusion protein (expressed by pGEX-KP8) on Western blots (Knight, 1993) confirming that the sequence cloned codes for the parasite protein recognised by 4B11. Unlike SPAG-1, SPAG-2 is not exclusively expressed during the sporozoite stage, but also during the macroschizont stage as shown by Northern blotting (Knight, 1993). No sequence homology between KP8 and any known gene could be detected when the gene banks were searched (Knight, personal communication). The remaining part of the SPAG-2 gene is currently being cloned and sequenced (Knight, personal communication). 1.5. Antigenic Polymorphism. Parasites use four main techniques to evade the immune system of their vertebrate host (Borst, 1991): (1) invading cells or hiding in sites in the body where the immune system is less effective, (2) mimicry of host proteins, (3) suppressing the immune system of the host and (4) antigenic polymorphism. In Theileria all these mechanisms have been observed during some of the life-cycle stages in the vertebrate host and the importance of antigenic polymorphism is discussed in more detail below. When a molecule is considered for inclusion in a sub-unit vaccine it is essential to understand the degree and nature of variation in its structure. Ideally the aim is to identify an immunologically important 36 region which is not polymorphic and will induce immune responses cross-protective against different strains of the parasite. 1.5.1. Antigenic polymorphism in Plasmodium with relevance to vaccine development. People who overcome malaria infection are usually protected against homologous challenge but not against different parasite isolates. This observation has subsequently been linked to the existence of antigenic polymorphism among malaria parasites in natural populations (Langsley and Roth, 1987; Kemp et al., 1990). Antigenic polymorphism has been well documented, for example in two of the major malarial vaccine candidates, merozoite surface protein-1 (MSP-1) and cirumsporozoite protein (CSP) (Miller et al., 1993). Attempts to map the host's immune responses to locations on known antigens have shown that the immune system tends to interact with polymorphic regions of these antigens (McCutchan et al., 1992; Mendis et al., 1991). Conserved and cross-protective epitopes prove difficult to find in malaria parasites probably due to immunological selection. For example, B-cell epitopes have been located in variant parts of the CSP antigen of Plasmodium vivax (Rosenberg et al., 1989), the MSP-1 (Peterson et al., 1990; Tanabe et al., 1987) and MSA2 of P. falciparum (Peterson et al., 1990; Fenton et al., 1990), the erythrocyte surface antigens of both P. falciparum (Marsh and Howard, 1986; Mendis et al., 1983) and P. vivax (Mendis et al., 1988) and the sexual stage antigens of P. vivax (Premawansa et al., 1990) and, to a lesser extent, of P. falciparum (Petterson et al., 1987). Variation has also been observed in regions where T-cell epitopes are located in some of these antigens (Good et al., 1988a and 1988b; Lockyer et al., 1989; Guttinger et al., 1988; Suss et al., 1993). As a result it is very difficult, if not impossible, to find non-polymorphic T-cell or B-cell epitopes. Therefore, it will be necessary to devise a sub-unit vaccine which includes a cocktail of variants of polymorphic epitopes from as many strains as possible in order to induce protective immune response against a range of parasite strains. 37 1.5.2. The detection of polymorphism in Theileria. Polymorphism in Theileria has been well documented and studied. The first documented evidence for strain variation of T. annulata is based on the results observed during the vaccination of calves by the infection and treatment method. It was shown that calves vaccinated by this method were protected against homologous challenges but were not protected against a heterologous challenge (Gill et al., 1980). This indicates the existence of different parasite strains from distinct geographical areas. In T. parva the variation of geographically distinct stocks was characterised by raising monoclonal antibodies against macroschizonts from different stocks. Some monoclonal antibodies raised this way were specific for certain T. parva strains (Minami et al., 1983). This shows that different T. parva strains exist and these antibodies can be used to characterise parasite stocks found in the field. Subsequently, monoclonal antibodies were raised which demonstrate variation between T. annulata strains thus allowing strain typing (Shiels et al., 1986). Glucose phosphate isomerase polymorphisms have also been demonstrated in both T. annulata and T. parva (Melrose et al., 1984 and Wilkie et al., 1986). This is another method which can be employed in strain typing. More recent advances in molecular biology have allowed further characterisation of the polymorphism at the DNA level. These were first discovered by studying the melting point of DNA from different T. parva strains (Allsopp et al., 1988) and later by studying restriction fragment length polymorphism (RFLP) patterns (Allsopp et al., 1988, Morzaria et al., 1990 and Bishop et al., 1993). RFLPs were found to exist for T. annulata DNA digested with Eco RI and probed with the SRI fragment of SPAG-1 (Williamson, 1988, Williamson et al., 1989, Knight, 1993 and Katzer et al., 1994). PCR, is another technique which has also been used extensively in an attempt to detect and characterise polymorphic strains of T. annulata (Ben Miled et al., 1994). Further indications of polymorphism at the protein level were obtained by SDS-PAGE. The T. parva polymorphic immunodominant macroschizont antigen (PIM) was shown to have a size polymorphism (Toye et al., 1991). Subsequently, similar size polymorphisms were 38 observed in the T. parva immunodominant schizont surface antigen from different stocks using 2 dimensional SDS-PAGE followed by Western blotting (Sugimoto et al., 1992). This antigen is probably be the same as PIM. A protein size polymorphism was also detected for the major merozoite surface molecule in T. annulata. This protein was shown to have a molecular mass of 30-kDa in isolates from Ankara (Turkey) and a molecular mass of 32-kDa in isolates from Gharb (Morocco) (Dickson and Shiels 1993). This size difference on Western blots is probably due to differential glycosylation. The detection of size polymorphism is yet another technique which can be used for parasite strain typing but yields little information with relevance to vaccine development and none about the function of a protein. 1.5.3. The study of polymorphism for treatment and vaccine purposes. So far the main emphasis of studying polymorphism was on the identification of all polymorphic parasite strains in a given geographical location. This would permit the selection of appropriate strains for the infection and treatment method for T. parva and also the selection of cross protective attenuated cell lines for vaccinations against T. annulata. So far no single technique has been able to detect all polymorphic parasite strains known and it is very unlikely that any one technique will ever be able to do so. It might be more successful to develop an infection and treatment method for T. parva which incorporates all parasite strains found in a given geographical location. A better alternative to either the infection and treatment method for T. parva and vaccination with attenuated cell lines for T. annulata is the development of subunit vaccines which consist of cross-protective epitopes from various life cycle stages of all known parasite strains for either T. annulata or T. parva . One might also be able to identify protective T-cell or B-cell epitopes which are identical for all Theileria species. 39 1.6. Stage-specific gene regulation. Stage specific gene regulation has been well documented and studied in protozoan parasites. Some of these stage-regulated genes encode molecules which are involved in cell adhesion and penetration as well as those which are involved in the evasion of the host defence mechanism (Parsons, 1990). Although many stage-regulated parasite genes have been isolated, we are just beginning to understand which mechanisms of gene regulation are of importance for protozoan parasites. The stages at which the expression of genes can be controlled are at the pre-transcriptional, transcriptional and post-transcriptional levels. Transcriptional regulation can either be constitutive, suppressible or inducible. Post-transcriptional regulation can be controlled by mRNA stability, mRNA splicing or by regulating translation. Transcriptional regulation is predicted to be the most economical mode of regulation (Latchman, 1990) and the expression of many genes has been found to involve this process; some examples will be described in the following sections. 1.6.1. Stage specific gene regulation in trypanosomes. The best studied parasites with regard to gene regulation are the trypanosomes. One of the transcriptionally-regulated genes in these organisms encodes PARP (procyclic acidic repetitive protein)(Clayton et al., 1989). It's mRNA is a-amanitin resistant, indicating that it is transcribed by RNA polymerase I (Rudenko et al., 1989). A single stranded DNA binding site has been identified in its predicted promoter region (Brown and van der Ploeg, 1994). The VSG (variant surface glycoprotein) genes of trypanosomes are a well studied gene family and their transcriptional control involves gene rearrangements (Pays et al., 1989; Latchman, 1990). About 1000 VSG genes are known but only one VSG transcription initiation site has been identified. During the promastigote stage, this transcription initiation site is operating constitutively, but only one VSG gene is transcribed at a time and the others lie dormant. The switching of expression from one gene to another happens by duplication of one of the silent genes which 40 replaces the previously transcribed gene at the expression site (Pays et al., 1989; Latchman, 1990). Post-transcriptional gene regulation has also been demonstrated in trypanosomes. For example the expression of the phosphoglycerate kinase isozyme is regulated by mRNA stability (Gibson et al., 1988). The fructose biphosphate aldolase gene is also post-transcriptionally regulated. This was suggested since the enzyme is stage-specifically expressed but the mRNA is not stage-specifically transcribed; the factors involved still need to be identified (Vijayasarathy et al., 1990; Parsons, 1990). In general, the regulation of trypanosome gene expression is well studied, whereas for a lot of other important parasites very little is known. The reason why so much more is known about gene regulation in trypanosomes is linked to the transfection systems which are available for these organisms. Once trypanosomes could be transformed, it became possible to study and isolate promoters, possible enhancers and other factors which might be involved in stage-specific gene regulation, such as regions responsible for mRNA stability. For example, transformation experiments have shed some light on the function of the PARP promoter (Rudenko et al., 1990; Lee and Ploeg, 1990) and the VSG promoter (Lee, 1995). 1.6.2. Stage-specific gene regulation in Plasmodium. Stage-specific gene regulation is also well documented in Plasmodium but very little is known about the mechanisms involved in this process since a stable transformation system has only just been achieved (Wu et al., 1995). A whole array of Plasmodium genes have been cloned and some of them are known to be stage-specifically regulated. The most common stage for gene regulation is at the level of transcription. Table 2 contains ten Plasmodium genes all of which are stage-specifically regulated and involve some level of transcriptional control. More detailed studies might reveal that post-transcriptional levels of control are also effective at later stages after the initial 41 wÖ ý C . ýý ed C id 0 0 p, C C C C _ 0 ý+ Ö C> Ö Ö Ü poCt. C" Q -, O O% C VN N w' C Ö Ö ý aCi äu .w _ _. " .» c> p 4. c> sC C yl C st C YC dÄC NA Q~ Ö ~ ' ý 4- ^ ýw" , ýý ä C R ý, ý 0 C 0 C 0 C O c ä C ä C ý. cL :, h ä p 0 0 O _ ý. ý 0 4) ý1 4 ý. w, v ,I ý. ^ ^ w ~ ' . + ""i y " ý. r1 ý u i 0 +. I +Y C tn " v; 0° ° C C 7 "'", aý C + E 7 ° 7 ý + E ý O O O O C C C C C ' ýý p C 4) ý00 .. ed 7 b 0 0 C> pý p0 ä f 4) wC CAd ý: Q .+° OQ .. N ý ý Nä ý; ýä ý ýý A pý V .. Oý ýýý _" \ ý Oý C> ý. Oý d 6> 00 OG ý O+ C "-+ 1.: A h ,ý "-" --. C 00 Oý 00 dl Oý 00 4.4 l7 N .., N .., OA.. 4) .. ý , h, . ,n , ... 8. . - 6" -". ä-- äa . bý ý 2 41 öý °' CO a CO xý a 3 x 3 3 O A A A A A ^' A A A A A A C C C C C C C C C C Co - ° ,° *O O C CO O O °d O O ÖA r Cý. ' -. 92. ". C. ". a " -. O ,w Cs. pý ý. . ý a . -. CL ý. a -. a ä º- t) ý u ý T: .... ': - '. - Z) .. 1:: d y y y CO y-b y Ci u Ci V u (-"ý ý" A CO A C A C Aý C C Cw C C N ~ ý ~ A ry A M A ý. A ". CO ý. CO ý. .: 4) ý, 2 "CVC _ oýo w ý ü u ý u ^ oýo "ý O ., M C oo N ýýD «S VYp e° le c:, o OO o CO ä Co 0 Co 0 A . a.. 6-. W 00 O, > 0ý CO ý 3 ". ýi ý b ý r- w- , ý -c> .. ý 0 c> N -". 0 c4 VJ . ""+ d 4- "^ --ý y -"r ^Ov. o CO oh O -"n Ch --. -, -ý h ' N -' b. " s. ý -- 3 y . ": ä w: ý. b v C7 .. y- 4) 4) O ~ ý ^ «+ H 4) 4) 4) N . O. \ N Vi ý .Cý .+ 3 .. .. O p N ýh tý p ~ O .. ~.. p 0 Vl N 4) eu 4) ý ^ A tn Ný i.... 00 vC N O O 00 N N N "' " O Op W ON b. 6.. , CL O , 7+ Ö ý ý pQ N ý Ay OO u2 Cs. 91 Cv y 4) N ý ; 'Ci . -"+ ON O C> O fý P 00 -Z ý ON ý. 00 C> C 00 C 00 '. O `° ~ 00 ° u w- ý 00 E u d'ý -6 uý C Oý E. ' . - Oý .. O O% .. ü ý 3 0N o e > ` w O4- ý 4) N -r i x ' "-°-. - h y O ý 4) «0 - w m äa x ä 4) :2 < °° ý .. . - ° x z etý a N A - x Q) Ca c C Q CO .ä c C z Q~ ý o. y a Q Cý C _ß 7 00 d ý. aý C ý b0 tý V a a h aý eo ea .. h i. ýv C eo h ý c ý ý N 41 3 ! Si Ey 42 regulation of transcription. One known case is pfMSA-1, where the gene is transcribed during the erythrocytic stages (Myler, 1989) but the mRNA only accumulates during the late erythrocytic stages, as demonstrated in a comparison of Northern blot and nuclear run-on data (Lanzer et al., 1992). The py230 gene is the Plasmodium yoelii equivalent to the pfMSA-1 gene of Plasmodium falciparum. Interestingly it has been shown that the pfMSA-1 gene is post-transcriptionally as well as transcriptionally regulated (Myler, 1989: Lanzer et al., 1992), while in the case of py230, transcriptional regulation is the only detected mode of regulation so far (Lewis, 1990). This shows that further levels of gene regulation might act at later stages which have not yet been identified. The process of stage-specific gene regulation in Plasmodium is being studied in order to gain a better understanding of parasite genetics. In the future this might lead to a form of treatment which is based entirely on a genetic approach by disrupting parasite specific gene expression. This might act by blocking developmental stages of the parasite by inhibiting unique, parasite-specific transcription factors. The first step towards an understanding of transcriptional regulation in Plasmodium was achieved by mapping the site for transcription initiation of the CS gene (Ruiz i Altaba et al., 1987). This gene is transcribed only during the sporozoite stage. Subsequently the site for transcription initiation has been mapped for another four Plasmodium genes: pfMSA-1 (Myler, 1989), py230 (Lewis, 1990), GBP 130 (Lanzer et al., 1990) and KAHRP (Lanzer et al., 1992). Interestingly, it has been noted that both KAHRP and GBP 130 have unique transcription initiation sites (Lanzer et al., 1990; Lanzer et al., 1992) while CS, HRP II and pfMSA-1 have multiple sites (Ruiz i Altaba et al., 1987; Myler, 1989; Lewis, 1990). For the actin genes, pf-actin I and pf-actin II, of Plasmodium falciparum it has been shown that the sites for mRNA initiation are a maximum of 1100 bp and 450 bp, respectively, 5' from the ATG start codon (Wesseling et al., 1989). The beginning of the mRNA has not yet been mapped to specific sequences since only 250 bp and 331 bp, respectively, of the 5' untranslated region has been cloned and sequenced for these genes. Once the transcription initiation sites for the five genes were mapped and the sequences of the 5' regions of these initiation sites were available, it became possible to search for putative transcription factor 43 binding sites. In eukaryotes it has been shown that promoter binding proteins usually bind within 100 bp of the transcription initiation site (Dynman & Tjian, 1985; Maniatis et al., 1987). In the 5' region of the multiple transcription initiation site of the py230 gene TATA boxes were found. The consensus sequence of a TATA box is TATA A/T A A/T (Corden et al., 1980) and has been shown to be the site responsible for binding by the transcription factor TFIID (Horikoshi et al., 1989). TFIID is responsible for transcription of genes by RNA polymerase II (Buratowski et al., 1989; Pugh & Tjian 1992; Flores et al., 1992) but is also needed for the transcription of genes which do not contain a TATA box in their promoter region (Pugh & Tjian 1991). The TATA box binding protein (TBP) of Plasmodium falciparum has been cloned and sequenced (McAndrew et al., 1993). Although TATA boxes were found in the 5' regions of stage-specifically regulated genes in Plasmodium their role in stage-specific gene regulation is questionable as the genome of these parasites is highly A and T rich and the intergenic regions consists of 90% A's and T's (Hyde and Sims, 1987). The TATA box motif has also been found in abundance in both constitutive and developmentally regulated genes, for example the a-tubulin I, the a-tubulin II and ß-tubulin genes (Holloway et al., 1989; Holloway et al., 1990). When the 5' region of the transcription sites of the pfMSA-1 gene were searched for other motifs, recognized by other transcription factors, two immunoglobulin octamer (Oct-1) sequences were found but no CCAAT box (Lewis, 1990). The consensus sequences for Oct-1 binding and the CCAAT box is shown in Table 3. Since two Oct-1 sites were found in the region 5' of transcription initiation it has been suggested that Oct- 1 might be involved in transcriptional regulation of pfMSA-1 (Lewis, 1990) but more evidence is needed to substantiate this theory. The 5' region of the GBP130 gene was also analysed and it was shown that the transcription initiation site maps to 980 bp upstream of the ATG codon (Lanzer et al., 1990). The GBP 130 gene is developmentally regulated as shown in Table 2 and transcripts of the gene are only found during the trophozoite stage (Lanzer et al., 1990). The region 5' to transcription initiation contains a region which is homologous to the SV40 enhancer sequence (Lanzer et al., 1992 and Weiher 1983). Subsequently it has been shown that nuclear extracts from the 44 Name of Sequence Name of DNA- Binding Protein DNA Sequence Relevant references ANTP TCAATTAAAT Treisman et al.. 1990 API TGAGTCG Lee et al., 1987 AP2 COOCAGGC Ima awa et al. 1987 bicoid TCTAATCCC Hanes and Brent, 1989 CRE CREB t/g a CGTCA Yamamoto et al., 1988 CCAAT BOX CTF, NFl TGTGGC7NNNAGCCAA Santoro et al.. 1988; Mermod et al., 1989 C/EBP ATTGCGCAAT Landschultz et al., 1989 ets CACITCCT Was lk et al., 1990 FOS same sequence as API Neuber et al., 1989 Ftz TCAATTAAATGA Jaynes and O'Farrell. 1988 GRE Promoter Glucocorticoid GGTACANNNTGTWT Hollenberg and Evans, 1988 GRE Su pressor ATYACNNNNTGATCW Godowski et al., 1988 GAL 4 TGTGGATATATG Carey et al., 1990 GCN4 TGA c TCAT Hope and Struhl. 1986 HSE CTNGAATNTTCTAGA Bienz and Pelham, 1986 HSU2 GOGTItCGGGA Gaston and Fried, 1992 H2TF1 TGGGATTCCCCA Sin Rh et al., 1988 JUN same sequence as API Neuber et al., 1989 KROX-20 GCGG cIg GGCG Nardelli et al.. 1991 NF-icB TGGGGATTCCOCA Lenardo and Baltimore, 1989 mb c AAC cG Biedenkapp et al.. 1988 myc CACGTG Murre et al.. 1989 Octl, Oct2 ATGCAAAT or ATTTGCAT Falkner et al., 1986: Latchman, 1990 Octl+VP16 TAATG a/g AT O'Hare and Goding. 1988; Preston et al., 1988 Oestrogen . AGGTCANNNTGACCT Kumar and Chambon, 1988 Pit-1 ATGAATA t/a Ingraham et al.. 1988 rel GTGGAGATGGGGAATCCCCA Gilmore, 1990 Serum Response Element GATGTCCATATTAGGACATC Norman et al., 1988 si -1 GAGGAA Goebl. 1990 Spl Box 0000 000 Courey and Tjian, 1988. Kadonga et al., 1987 TATA Box TFIID TATA t/a A t/a Corden et al., 1980 Thyroid Response Element Thyroid hormone/ Retinoic acid TCAGGTCATGACCTGA Evans, 1988; Umesono and Evans, 1989 Table 3: Transcription factors and the DNA sequence motifs they bind to. erythrocytic stages of the parasite bind to this region (Lanzer et al., 1993), indicating that it is very likely to be involved in the regulation of the gene. The 5' region prior to the transcriptional initiation site of the CS gene also has homology to the SV40 enhancer sequence (Lanzer et al., 1992; Lanzer et al., 1993). A transformation system needs to be operational to determine whether these sequences or any other sequences, which are speculated to be promoter binding sites, are involved in stage specific gene regulation. The stable transformation of Plasmodium blood stages has only just been achieved (Wu et al., 1995) and this system will hopefully provide the required method to study the involvement of these sequence motifs in stage-specific gene regulation in Plasmodium . 1.6.3. Stage-specific gene regulation in Theileria. Little is known about the process of stage specific gene regulation in Theileria. The first indication that some level of control of expression must occur in T. annulata is the existence of life-cycle stage specific antigens. The existence of stage specific antigens was first recognized by raising monoclonal antibodies against surface molecules of different parasite stages for vaccine development. Shiels et al. (1986) raised a panel of monoclonal antibodies against macroschizonts. It was shown that some of these mAbs reacted exclusively with schizonts, some reacted with schizonts, piroplasms and sporozoites, while others just reacted with schizonts and piroplasms or schizonts and sporozoites. It was therefore concluded that Theileria expresses life-cycle stage specific antigens. Subsequently, another four monoclonal antibodies were raised which were specific for macroschizonts and these did not react with piroplasms, kinetes or sporozoites (Shapiro et al., 1987). These immunological experiments showed that *stage specific gene expression occurred, but they yielded no information as to which level of regulation is of importance. More information about gene regulation in Theileria was obtained when the first genes were isolated and molecular biology techniques could be employed to study at which stage their gene regulation was controlled. So far three genes have been isolated which are transcribed 46 and expressed constitutively. Firstly the apocytochrome B gene of T. annulata (Megson et al., 1991) has been characterised. This gene is located on the 6.3 kb extrachromosomal DNA element (Hall et at., 1990). It is thought that this gene product might be a suitable target for the development of a treatment for tropical theileriosis, as its expression is thought to be essential for the parasite (Megson et al., 1991). The second constitutively expressed gene is the cysteine protease of T. annulata (Baylis et al., 1992) and the third is the hsp 70.1 gene of T. annulata (Mason et al., 1989). The latter has been shown to be transcribed during the sporozoite, schizont and piroplasm stages (Mason et al., 1989) but the level of expression is still further inducible. Although these genes are constitutively transcribed during the life-cycle stages in the bovine host very little is known about gene expression during the early life-cycle stages in the tick. If these stages were studied, it might become apparent that some genes which are constitutively expressed in the bovine host might not be transcribed at all during the stages in the tick, or at least not for all stages in the tick. So far, most Theileria genes isolated are stage-specifically expressed and all these genes are regulated to some extent at the transcriptional level. A list of these genes is shown in Table 4. Some of these genes are transcribed only during a single stage of the life cycle of the parasite, for example p67 (Nene et al., 1992). It is only transcribed during the sporozoite stage and ORF-1, an open reading frame which is located 5' to the p67 gene, is only transcribed during the schizont stage and not during the sporozoite nor the piroplasm stages (Nene et al., 1992). The function of this gene is as yet unknown (Nene et al., 1992). 3' to the p67 gene is another open reading frame, ORF-2, which is transcribed during the sporozoite, schizont and piroplasm stages. Its function is unknown but it is another potential constitutively expressed gene (Nene et al., 1990). Interestingly the 117 kDa rhoptry protein of T. annula to is only found in merozoite rhoptries and not in those of the sporozoite, and the gene is transcribed 2 to 3 days prior to rhoptry formation. Afterwards, transcription is down regulated (McDonald, personal communication). These examples indicate that transcriptional control is very important for the development of Theileria but unfortunately very little is known about how these genes are regulated. A ý7 Transcription initiation sites have been mapped for only one Theileria gene, the hsp 70.1 gene (Mason et al., 1989). The beginning of the mRNA was mapped to 215 bp 5' of the ATG start codon using the SI mapping technique. Two sequences which show high homology to the heat-shock element binding site consensus sequence, shown in Table 3, were found in the 5' region upstream of the transcription initiation site (Mason et al., 1989). Furthermore, a putative TATA box was found in the 5' region and polyadenylation signals were found in the 3' untranslated region. Since the hsp 70.1 gene is heat inducible it was speculated that these conserved promoter binding sequences are functional in Theileria. To date, no promoter binding proteins have been isolated in Theileria. Name of gene Stage where the gene product is found. References P protein schizont, piroplasm Baylis et al., 1992 PIM sporozoite, schizont Toe et al., 1991 ORF-1 schizont Nene et al., 1992 SPAG-1 sporozoite Williamson et al., 1989 SPAG-2 sporozoite, schizont Knight. 1993 67 sporozoite Nene et al., 1992 117 kDa rhoptry protein merozoite McDonald, pers. comm. 30-32 kDa merozoite antigen merozoite, piroplasm Glascodine et al., 1990 Dickson and Shiels, 1993 Table 4: A list of Theileria genes which are stage specifically regulated and the level of regulation involves some transcriptional regulation. A0 1.7. Host cell recognition and invasion. In order to evade the immune responses of their host, a very successful strategy has been developed by a number of protozoan parasites; they become intracellular for one or more of their life-cycle stages (Bloom, 1979; Borst, 1991). The parasites have develop adaptive processes that allow host cell penetration and intracellular survival. Ideally, vaccine and treatment development should target the interception or destruction of the parasite before it becomes intracellular and thereby establishes an infection. Therefore, one needs to understand how the parasite interacts with host cells and how defence mechanisms function against them (Vandenberg and Stewart, 1990). 1.7.1. Host cell recognition and invasion by malaria parasites. Plasmodium exhibits three invasive stages: a) the sporozoite invading hepatocytes in the vertebrate host, b) the merozoite whose target cell is the erythrocyte and c) the ookinete which infects the midgut endothelial cell of the mosquito vector (Sinden, 1985). The first two stages are of importance for the development of therapeutic and preventative treatments. The specificity of the recognition event and the speed of the invasion process implies that the invasion process is receptor-ligand mediated (Hadley et al., 1986). Further, it has been suggested that the invasion of Plasmodium is a multi-step process requiring a series of interactions involving several different host and parasite molecules (Mitchell et al., 1986). The current knowledge of the mechanisms of host recognition and invasion during hepatocyte and erythrocyte invasion are described below. Recognition of hepatocytes by Plasmodium sporozoites. When sporozoites are transmitted by their mosquito vector into their mammalian host they circulate through the blood stream until they reach the liver. Once there they have to penetrate or move between either Kupffer cells or endothelial cells, which line the lumen of the 49 hepatic sinusoid, in order to reach their target cells, the hepatocytes (Vanderberg and Stewart, 1990). The sporozoites are well adapted and as many as 95% of all injected sporozoites will reach and invade their target cells (Vanderberg, 1968). The CS protein has been proposed as a candidate ligand for the putative hepatic cell receptor (Nussenzweig and Nussenzweig, 1989). This antigen was proposed for three reasons: a) it is expressed on the surface of the infective (salivary gland) sporozoite (Yoshida et al., 1981) but is only present in small numbers on non- infective sporozoites from oocysts (Aikawa et al., 1981); b) the CS protein contains a region, region II, which bears a striking homology to a cell adhesion domain of thrombospondin (Prater et al., 1991; Tuszynski et al., 1989) and to similar regions of several other proteins (Clarke et al., 1990; Hedstrom et al., 1990; Robson et al., 1988; Goundis and Reid, 1988); and c) sporozoite invasion of hepatocytes can be blocked by monoclonal antibodies which react with the CS protein (Potocnjak et al., 1980). Two other characteristics of the CS protein have been suggested as assisting in host cell invasion. The first is that the CS protein contains a Cat + binding region which interacts with the phospholipid membrane of host cells. It is thereby of critical function during attachment, invasion and the development of the malaria parasites in hepatic cells (Verdini et al., 1991). The second observation is that the CS protein possess both positively and negatively charged sites which are accessible on the surface of the sporozoite, and are thought to be components of surface ligands during hepatocyte invasion (Mathews and Vanderberg, 1994). Conclusive evidence showing that the CS protein is involved in hepatocyte invasion was provided by Cerami and co-workers (1992). They showed that recombinant CSP of P. falciparum binds specifically to the membrane surface of hepatocytes and not to other cells. They also showed that the binding can be inhibited by synthetic peptides representing the evolutionarily-conserved region II of the CS protein. Further, this region has been shown to be responsible for binding to the heperan sulfate proteoglycans on the surface of the hepatocytes (Cerami et al., 1992; Panacake et al., 1992; Frevert et al., 1993; Frevert, 1994; Cerami et al., 1994). The binding domain has been found in CS proteins from all Plasmodium species as well as another sporozoite protein SSP2/TRAP (Rogers et al., 1992a; Rogers et al., 1992b; Wizel et al., 1994; Robson et al., 1988). The SSP2/TRAP protein is located in the micronemes and the surface of the sporozoite stage and has been shown to bind to sulfated cn glycoconjugates on the surface hepatocytes (Miller et al., 1993). It has been proposed that the recognition and initiation of hepatocyte invasion is achieved by a cascade of receptor-ligand interaction. It is therefore likely that the CS protein and SSP2/TRAP act together with other sporozoite proteins during the invasion process (Hedstorm et al., 1990; Moelans, et al., 1991; Fidock et al., 1994a; Fidock et al., 1994b). Interestingly, the immunodominant microneme protein of Elm eria tenella has the same thrombospondin-related region as the CS protein and it has also been speculated that it is involved in host cell invasion (Clarke et al., 1990; Tomley et al., 1991). Another surface molecule of the sporozoite stage, CSP-2 has only just been isolated (Sina et al., 1995). Since monoclonal antibodies directed against this protein block the invasion of hepatocytes in vitro this might indicate that this protein is also involved in the invasion of host cells (Sina et al., 1995). Recognition of erythrocytes by merozoites of Plasmodium. Over 40 years ago, it was suggested that the invasion of erythrocytes by merozoites is receptor-ligand mediated (McGhee, 1953). Although the process of merozoite attachment and invasion has been described in detail at the light microscope level (Dvorak et al., 1975) and by ultrastructural studies with the electron microscope (reviewed by Bannister and Dluzewski, 1990; Aikawa et al., 1978; Perkins, 1989), the molecules involved and their roles are still poorly understood. One of the main candidate ligands on the merozoite surface is the MSP-1 protein. This antigen is processed and the resulting polypeptides are found uniformly distributed across the surface of the merozoite (Holder et al., 1987, McBride and Heidrich, 1987). Perkins and Rocco (1988) have demonstrated that this protein binds to erythrocytes and they speculated that the polypeptide binds to sialic acid on the erythrocyte surface. When the structure of MSP-1 was investigated, two epidermal growth factor (EGF) modules were found in the 19 kDa C-terminal region of the antigen (Blackman et al., 1991a). The 19 kDa polypeptide is linked to a GPI-anchor and remains on the surface of the merozoite during the invasion process (Blackman et al., 1990a; Blackman et al., 1991b). Subsequently it has been shown that this cleavage is intrinsic to successful erythrocyte invasion and that this process is facilitated by a serine protease (Blackman et al., 1993). The proteolytic cleavage process of MSP-1 by a serine protease has also been confirmed in vivo (Odea et al., 1995). Further evidence that this C-terminal region is involved in the invasion of erythrocytes was obtained by the use of antibodies in invasion blocking studies. A monoclonal antibody which binds to the first EGF module was shown to block merozoite binding to erythrocytes (Chappel & Holder, 1993: Su et al., 1993). Other monoclonal antibodies which react with the 19 kDa C-terminal fragment also block invasion (Blackman et al., 1990a; Cooper et al., 1992). Therefore, this C-terminal polypeptide seems important in the invasion process and the role of the EGF modules in this process needs to be evaluated further. Binding studies have revealed that the MSP-1 protein also binds to sugar residues found on the surface of erythrocytes (Qazi et al., 1994). It seems likely that the MSP-1 protein is one of the first receptor- ligand interactions between the merozoite and the erythrocyte. This interaction might induce the next step in the cascade, i. e. the release of EBA-175 from the micronemes. EBA-175 was shown to belong to a family of microneme proteins (Camus and Hadley, 1985; Haynes et at., 1988; Wertheimer and Barnwell, 1989; Adams et al., 1992; Peterson et al., 1995) which can bind to the sialic acid moiety on glycophorin A (Orlandi et at, 1990; Klotz et al 1992; Orlandi et at., 1992; Sim et al., 1994). Sequence analysis of the microneme proteins of the above gene family reveals that these proteins are related and exhibit dimorphism (Klotz et al., 1992; Prickett et at., 1994). Similar to MSP-1, EBA-175 is proteolyticly cleaved during the invasion process into a 65-kDa fragment whose binding to erythrocytes is sialic acid independent (Kain et at., 1993). It is predicted that this cleavage process is involved in the formation of the moving junction between the merozoite and the erythrocyte membrane and thereby permitting membrane fusion with the host cell (Kain et al., 1993). The region involved in the interaction with the host cell membrane has been mapped to a 42 amino acid peptide in the C-terminus of EBA-175 (Sim et at., 1994a; Sim et al., 1994b). It has been proposed that EBA-175 is the most important ligand for binding of merozoites to glycophorin A on erythrocytes (Sim, 1995). However, other evidence exists that P. falciparum can utilize other receptors, such as glycophorin B (Dolan et at., 1994). Therefore it seems likely that parasite has multiple genes with different erythrocyte-receptor specificities which can be switched on and off (Miller et at., 1994). 52 Another merozoite protein has been isolated and is called MCP-1 (merozoite cap protein-1). This protein follows the distribution of the moving junction during the invasion of erythrocytes. Sequence conservation between this protein and bacterial and eukaryotic proteins has indicated that it has oxidereductase activity as well as another domain which may mediate interactions of MCP-1 with the cytoskeleton in erythrocytes (Hudson-Taylor, et al., 1995). During this and later stages of the invasion process, other proteins will be released from the merozoite such as further erythrocyte membrane binding proteins (Elmoudni et al., 1993), serine-proteases and phospholipases (Braunbreton et al., 1994; Florent et al., 1994), protein kinases (Ward et al., 1994) and various rhoptry proteins (Grellier et al., 1994: Samyellowe et al., 1995; Ndengele et al., 1995). A different receptor on the erythrocyte has been implicated in merozoite invasion of P. knowlesi and P. vivax. This is the Duffy blood group antigen (Miller et al., 1975; Miller et al., 1979; Barnwell et at., 1989). One line of evidence for this was provided by monoclonal antibodies raised against the Duffy blood group antigen, as they block merozoite invasion (Barnwell et al., 1989). When the invasion of P. falciparum merozoites was investigated it was shown that there was no correlation of the expression of the Duffy blood group antigen and susceptibility of target erythrocytes (Holder, 1994). It has been shown that the Duffy blood group antigen consists of the IL8 receptor and MGSA (melanoma growth stimulatory activity) receptor (Horuk et al., 1993). Mutations in the MGSA receptors resulted in the loss of infectivity of the mutated cell lines (Hesselgesser et at., 1995). The family of Duffy blood group antigen binding proteins was isolated from P. knowlesi and P. vivax, was shown to contain homology to EBA-175 and their binding domain was mapped to a cystein-rich domain (Chitinis and Miller, 1994). This binding domain is hypervariable but this variability is limited (Tsuboi et at., 1994). Although different Plasmodium species might utilise different ligand-receptor interactions for both the initial and subsequent invasion steps the general principal is the same and the parasite proteins involved seem to be related. Therefore, for invasion to be successful a cascade of parasite and host cell protein-protein G 0) interactions has to take place and as a result the invading parasite manages to hide inside the host cell where it is protected from most of the hosts immune responses. 1.7.2. Host cell attachment and invasion by Toxoplasma tachyzoites. Toxoplasma tachyzoites exhibit a very broad host cell range. They can invade virtually all nucleated cell types from warm-blooded vertebrate hosts (Werk, 1985). The invasion event is not due to phagocytosis as previously thought but is an active process. During this process no host cell membrane ruffling, actin microfilament reorganisation or tyrosin phosphorylation of host proteins takes place (Morisaki et al., 1995). These effects have, however, been seen during phagocytosis (Morisaki et al., 1995). The notion that trachyzoites can invade host cells by phagocytosis was based on the observation that phagocytic cells take up Toxoplasma parasites and that the parasite can survive in these cells. This however, is due to some parasites escaping from the phagosome by a process analogous to invasion (Morisaki et al., 1995). The invasion process of tachyzoites is very similar to the invasion process of Pla sni odium because a cascade of interactions between the parasite and the host cell has to take place. Initially the tachyzoite loosely attaches to the host cell and rearranges itself to bring the apical pole in close apposition to the membrane (Wong and Remington, 1993; Werk, 1989). These processes involve receptor-ligand interactions and are followed by the protrusion of the conoid. Then a moving junction is formed between the parasite and the host cell membrane which is accompanied by microneme exocytosis. Rhoptry exocytosis is the next step and when the parasite is engulfed completely the prasitophorous vacuole is formed and transformed into an environment suitable for the parasite with the help from the dense granules (Joiner and Dubremetz, 1993; Dubremetz and Schwartzman, 1993). The complete process is extremely fast lasting between 25 and 40 seconds (Morisaki et al., 1995). The attachment process of the tachyzoite to the host cell and the initiation of the invasion process involves an array of receptor-ligand interactions which have partly been identified. One of the these interactions is between parasite laminin and the laminin receptor on host cells (Furtado et al., 1992) and it was also shown that the parasite laminin interacts with the 01 integrin receptor and colagen type IV (Joiner et al., 1990). Tachyzoite binding to host cells can be blocked by the addition of monoclonal antibodies against laminin and the a6 chain of the integrin family (Furtado et al., 1992). Two further parasite proteins have been identified which interact with the host cell membrane. These are parasite lectin like molecules (Robert et al., 1991) and major parasite protein SAG-l. SAG-1 was shown to bind glycosylated host cell receptors (Minco et al., 1993; Grimwood& Smith, 1992) and therefore is important for parasite attachment. The importance of SAG-1 was supported by data which shows that antibodies against SAG-1 but not SAG-2 block tachyzoite invasion. (Kasper &Mineo 1994). The first set of proteins which originate from organelles and are of importance during the invasion process are the microneme proteins. Three microneme proteins have been identified in Toxoplasma (MCI1: 60 kDa, MCI2: 120 kDa, MCI3: 90 kDa) but their role during the invasion process still needs to be identified (Achbarou et al., 1991) During the next stage the content of the rhoptries is released. The main components of these are lipids (Foussard et al., 1991) but at least two rhoptry proteins have been cloned and studied. The first is ROP-1 which has one acidic and one basic domain that may confer a large spectrum of binding capabilities (Schwartzman 1986; Ossorio et al., 1992). It has been suggested that this protein may facilitate parasite motility during invasion (Kasper and Mineo, 1994). The second rhoptry protein is ROP-2 which is inserted into the parasitophorous vacuole membrane and is thought to allow interactions of the parasite with the cytoplasm of the host cell (Beckers et al., 1994). It has been shown that when the parasitophorous vacuole membrane is formed it contains either channels or transporters which regulate the trafficking between the parasite vacuole and the cytoplasm (Schwab et al., 1994: Demelo et al., 1992). Finally there is another group of proteins which play an important role during the later stages of the invasion process. These are the proteins released from the dense granules and they are thought to be 55 responsible for the remodelling of the parasitophorous vacuole. Five of these proteins have been cloned and sequenced but the function of all five needs to be further studied. GRA-1 is a 24 kDa soluble calcium binding protein (Cesbron-Delauw et al., 1989; Sibley et at., 1995), while the function of the second protein GRA-2 is not yet known. However, it is soluble and has a molecular mass of 28.5 kDa (Prince et at., 1989; Mercier et al., 1993). The third, GRA-3, is a soluble 30 kDa protein which is thought to interact with the parasitophorous vacuole membrane (Bermudes et at., 1994; Ossorio et at., 1994). Both the function of GRA-4 (40 kDa) and GRA-5 (21 kDa) is not known but both proteins contain a transmembrane domain which might allow interactions of the parasite with the host cell (Mevelec et at., 1992; Lecordier et at., 1993). 1.7.3. The process of host cell recognition and invasion by Theileria sporozoites. Host cell recognition and invasion by Theileria sporozoites is only poorly understood. It became possible to study this process after the infection and transformation of host cells by Theileria parva sporozoites was achieved in vitro (Brown et al., 1971) and the first insight into the invasion process was based on observations by electron microscopy (Fawcett et al., 1982). It was shown that the interaction of the sporozoite with the membrane occurs in two steps (Webster et al., 1985) as illustrated in Figure 5. First, the sporozoite loosely binds in any orientation to the surface of the host cell. This interaction is either terminated and the sporozoite is released or the parasite becomes more closely attached, leading to invasion (Webster et al., 1985). The membrane of the sporozoite comes into very close apposition with the membrane of the host cell. 'Zippering up' of the membranes then occurs and the sporozoite sinks into the host cell until the rim of the invagination pocket closes and fuses over the parasite (Fawcett et al., 1982). This process of internalisation is very fast, taking less than three minutes (Shaw et al., 1991). The invasion process is ligand-receptor mediated and it was thought that the sporozoite enters by passive endocytosis (Fawcett et al., 1984). It has been demonstrated that a surface antigen, p67 of T. parva, which is thought to be a candidate for host cell recognition and invasion is shed from the sporozoite during the 56 invasion process (Dobbelaere et al., 1985). During the entry process of the host cell by the sporozoite the micronemes and rhoptries discharge their contents and the enveloping host cell membrane is lysed (Fawcett et al., 1984 and Shaw et al., 1991). Thus the parasite comes to lie free in the cytoplasm. The initial findings for T. parva attachment and invasion were largely confirmed for T. annulata (Jura et al., 1983). However conflicting results were obtained for the two species regarding the energy requirements of the sporozoite during the invasion process. In contrast to Fawcett's conclusion with T. parva, the process of invasion by T. annulata sporozoites is an active process. It is initiated by the sporozoite and involves either proteins or glycoproteins as receptors on the host cell (Jura, 1984). Although it is theoretically possible that the sporozoites of T. annulata and T. parva use very different methods for host cell invasion, this is thought to be very unlikely. The issue was resolved much later when the entry of T. parva sporozoites into bovine lymphocytes was re-examined. It was shown that whilst the initial interaction with the host cell is a chance event independent of temperature (Shaw et al., 1991), all subsequent stages of invasion are temperature dependent, requiring the participation of live, intact sporozoites and host cells. The process involves some rearrangement of the host cell surface membrane and cytoskeleton (Shaw et al., 1991). As yet none of the receptors involved in the invasion process of T. annulata are known, but in T. parva there is evidence that the MHC class I molecules is essential in the process of sporozoite entry into bovine leucocytes (Shaw et al., 1995). s7 Figure S: Diagrammatical representation of the invasion of a leucocyte by a Theileria sporozoite. Panel a) shows the initial loose interaction of the sporozoite surface coat with the surface of leucocytes (arrow). Panels b) and c) show the entry of the sporozoite into the leucocyte: the parasite and lymphocyte membrane are in close apposition; as the entry progresses, the sporozoite plasmalemma is 'zippering up' with the membrane of the lymphocyte. CQ 1.7.4. Identification of target cells for invasion by Theileria sporozoites. The method of in vitro infection and transformation of host cells by Theileria sporozoites (Brown et al., 1971) has also allowed us to study and characterise the target host cells for sporozoite invasion and transformation. The first indication that T. parva sporozoites infect a specific population of leucocytes was obtained when Stagg et al. (1981) found that the sporozoites only bind to and invade 25% of purified leucocytes. Later it was shown that T cells are the likely target for T. parva sporozoites since the transformed cell lines express T cell markers as defined by specific monoclonal antibodies (Pinder et at., 1981). Similar results to Stagg et at. (1981) were also observed for T. annulata (Jura et al., 1983). Thus T. annulata sporozoites commonly bind to only 10% - 20% (but occasionally up to 40%) of purified leucocytes (Jura et al., 1983). There were indications that there might be different target populations within leucocytes, since sporozoites bind to some cells only at one pole, but evenly across the surface of other cells (Jura et at., 1983). It was also shown that Theileria sporozoites are species-specific in the host cell invasion and transformation process. T. parva infects and transforms only peripheral blood leucocytes of cattle (ie Bos taurus and Bos indicus) and buffalo (Syncerus caffer) (Stagg et al., 1983) while T. annulata infects cells from a wider host range. T. annulata infects peripheral blood leucocytes from the Bovidae family ie. cattle (Bos taurus) and buffalo (both Syncerus caffer and Bubalus bubalis), goats (Capra hircus) and sheep (Ovis aries) (Steuber et al., 1986). However, T. annulata does not transform peripheral blood leucocytes from buffalo in vitro although these cells can be infected by T. annulata sporozoites (Steuber et al., 1986 and Chaudhri and Subramanian 1992). The identification of cells infected by Theileria sporozoites has been made more difficult since it has become apparent that when the parasite transforms a leucocyte it changes the phenotype of the cell (Stagg et al., 1981). The changes include an increase in the amount of cytoplasm and alterations in the surface of the parasitised cell. The composition of the surface molecules changes so that for example, aT cell infected with T. parva, which did not express any MHC class 11 molecule prior to infection, expresses this marker once the cell is 59 transformed (Morrison et al., 1986 and Lalor et al., 1986). Therefore, to identify target cells for Theileria invasion, one cannot rely on the phenotype of the transformed cells as this might give misleading results. The target cells for T. parva invasion were investigated by attempts to infect subpopulations of sorted peripheral blood mononuclear cells in vitro. The identified target cells are T cells, B cells and Null cells while neither monocytes nor neutrophils could be infected (Baldwin et al., 1988). Other targets for T. parva sporozoites include eosinophils and fibroblasts (Morrison et al., 1986). It was also observed that, generally, all transformed B cells lose the expression of their surface immunoglobulin. Further, the transformed cells started to express T cell markers such as CD2 and CD8 (if they did not express them already (Baldwin et al., 1988)). The up-regulation of CD8 expression was also seen after sorted CD4+ CD8- cells were infected in vitro (Emery et al., 1988). Conrad et al. (1989) observed that genotypically distinct T. parva parasites induce different levels of expression of CD6, CD8 and Null cell markers after the transformation of a cloned host cell line. It has also been observed that cells after infection by T. annulata have an altered membrane composition. This includes increased MHC class II expression (Glass and Spooner, 1990) and expression of parasite specific antigens (Shiels et al., 1986 and Shiels et al., 1989). When the target cells for T. annulata sporozoites were compared to those of T. parva it became apparent that both parasites infect and transform different cell types as shown in Table 5 (Spooner et al., 1988, Glass et al., 1989 and Spooner et al., 1989). Thus T. annulata sporozoites preferentially infect MHC class II positive cells, whilst T cells were only infected at a very low frequency. In complete contrast, T. parva preferentially infects T cells and does not infect monocytes at all, although the latter are the main target for T. annulata. Furthermore, B cells are infected much more efficiently by T. annulata than T. parva (Glass et al., 1989 and Spooner et al., 1989). In summary the main targets for T. annulata infection are MHC class II positive cells, macrophages, monocytes and B cells. It is questionable whether T cells are a target for the in vivo infection by T. annulata sporozoites, since they can be infected only at a very low frequency in vitro, and if infected they rapidly lose their T cell markers such as CD4, CD8 (Innes et al., 1989, Spooner 1988). However, some evidence obtained from in vivo derived T. 60 annulata infected cell lines indicates that the sporozoites might infect a wider range of cells in vivo, than in vitro, since one such cell line expressed CD2, CD3 and the 7/8 T cell receptor indicating that this cell line is of T cell origin. But it did not express the other T cell markers CD4 or CD8 (Howard et al., 1993). T. annulata T. parva. MHC class II+ cells +++ + MHC class 11- cells - ++ Monocytes +++ --- B cells ++ + T cells +/- +4-'- Macrophages + +/- Fibroblasts + + Table 5: Comparison of target cells for sporozoite invasion by T. annulata and T. parva. L1 1.8. Objectives of my work. The aim of this thesis is to further the understanding of the functional and practical importance of the sporozoite antigens SPAG-1 and SPAG-2. Unfortunately not the whole SPAG-2 gene is available as discussed in section 1.4.2. and the part of the gene which had been cloned was made available only during the latter stages of my D. Phil. Therefore, most of my work concentrates on SPAG-1. When I started my D. Phil, some preliminary evidence was shown for the existence of polymorphism in the SPAG-1 gene (Dr. R. Hall, personal communication). Therefore one of my objectives was to first confirm that SPAG-1 shows antigenic polymorphism and then to establish the degree of polymorphism. A full length cDNA sequence of SPAG-1 was available and a genomic SPAG-1 clone also existed, but only a very limited amount of sequence data was available for this clone. My primary objective was to sequence the genomic clone of SPAG-1 and to establish the level of polymorphism between the cDNA and genomic copy of SPAG-1. It was predicted that this sequence comparison might reveal important information for future vaccine development. The next objective was to establish the degree of polymorphism of SPAG-1 at the DNA and protein levels between and within isolates. A SPAG-1 polymorphism was detected previously by RFLP analysis of piroplasm DNA digested with Eco RI and probed with the SRI region of SPAG-1 (Williamson, 1988; see section 1.5.2. ). To characterise this polymorphism a PCR based cloning and sequencing approach was chosen. The template DNA investigated was isolated from cloned macroschizont-infected cell lines which originated from two independent parasite isolates: one from Hisar and the other from Ankara. SPAG-1 is a stage specifically regulated gene which is only expressed during sporozoite development (Williamson et al., 1989). A genomic SPAG-1 clone was available which contained the 5' untranslated region of approximately I kb. Therefore, one of my objectives was to map the mRNA initiation site in an attempt to locate the promoter region involved in the stage specific regulation of SPAG-1 transcription. A 1) Further, attempts were made to isolate DNA binding proteins which are involved in this regulation process, as this might reveal novel transcription factors in Theileria as well as novel DNA binding sites. Hall and co-workers (1992) observed three VGVAPG hexa-peptides within the predicted SPAG-1 amino acid sequence. This hexa-peptide is also found in bovine elastin and there it was identified to be the region to which the elastin receptor binds (Mecham et al., 1989). This observation and the fact that 1A7, a monoclonal antibody against SPAG-1, blocks invasion of sporozoites into host cells, led to the theory that SPAG-1 might be a ligand on the sporozoite (Hall et al., 1992). It was further proposed that SPAG-1 binds to the elastin receptor on the host cell surface during the recognition event of host cell invasion (Hall et al., 1992). Therefore, my aims were to show that recombinant SPAG-1 and subsequently SPAG-2 are ligands which are involved in host cell invasion. The approach chosen was to use recombinant parasite proteins in binding assays in an attempt to show that these two antigens bind specifically to putative host cells. Another of my objectives was to investigate the role of the elastin receptor for SPAG-1 binding and its involvement in host cell invasion. 63 Chapter 2 Materials and Methods 2.1. Materials. 2.1.1. Buffers. The abbreviations given for the buffers listed in this section are used throughout the thesis. Unless otherwise stated, all buffers were made up in ultrapure water prepared by reverse osmosis and deionisation using the Purity-Labwater water purifier. 0.5 M EDTA pH 8.0 Disodium ethelenediaminetetra-acetate. 2H20 was made up to 0.5 M and the pH was adjusted to 8.0 with NaOH pellets. 10 x PBS (phosphate buffered saline) was as follows: 1.5 M NaCI 160 mM Na2HPO4.2H20 40 mM NaH2PO4. H20 The pH was adjusted to 7.3 with HCI. lx TE comprises: 10 mM Tris base 1 mM EDTA, diluted from the 0.5 M EDTA pH 8.0 stock, pH was adjusted to 7.5 with HCI. 1.5 M Tris-HC1, pH 6.8 1.5 M Tris base was made up in H2O and the pH was adjusted to 6.8 with HC1. 1M Tris-HCI, pH 7.5 1M Tris base was made up in H2O and the pH was adjusted to 7.5 with HC1. 64 1M Tris-HCI, pH8.0 1M Tris base was made up in H2O and the pH was adjusted to 8.0 with HCI. 1.5 M Tris-HCI, pH 8.8 1.5 M Tris base was made up in H2O and the pH was adjusted to 8.8 with HCI. Competent cell buffer I 30 mM KCOOH 50 mM MnCl2 100 mM KCl 10 mM CaC12.2H20 15 % (v/v) glycerol Competent cell buffer II 10 mM Na-MOPS, pH 7.0 75 mM CaC12.6H20 10 mM KCl 15 % (v/v) glycerol 2.1.2. Bacterial culture media. All bacterial culture media were made up using deionised water and were sterilised by autoclaving for 20 minutes at 15 lb/square inch, 1200C. The following media were used: Luria-Bertani (LB) medium 10 g NaCI 5g Bacto yeast extracts 10 g Bacto-tryptone per litre When growing E. coli for the preparation of recombinant proteins, MgSO4 and glucose was added to 10 mM and 0.2% respectively before use. 65 Ampicillin stock 100 mg ml-1 ampicillin (sodium salt) was made up in H20, filter sterilized and stored at -20°C. Amplicillin was added to culture media at a final concentration of 100 pg ml-1 when E. soli harbouring recombinant plasmids carrying amplicillin resistance genes were grown. LB agar 15 g Difco Agar was added per litre of LB medium prior to autoclaving. TYM medium 2% Bacto tryptone 0.5 % Yeast extracts 0.1 M NaCl 10 mM MgSO4.7H20 X-gal stock 20 mg ml-1 5-bromo-4-chloro-3-indolyl-ß-D-galactoside was made up in dimethylformamide and it was stored at -20°C wrapped in aluminium foil. 1M IPTG stock 238 mg ml-1 isopropylthio-ß-D-galactoside was made up in H20, filter sterilized and stored at -200C. 2.1.3. Solutions for extracting Plasmid DNA. Solution I 50 mM glucose 25 mM Tris-HC1 pH 8.0, diluted from the 1M stock 10 mM EDTA, diluted from the 1M stock sterilized by autoclaving. 66 Solution II 0.2 M NaOH 1 %SDS filter sterilized. Solution III 5M potassium acetate pH adjusted to 4.5 with CH3COOH and sterilized by autoclaving. Chloroform/isoamyl alcohol (IAA) 1 part IAA was equilibrated with 24 parts chloroform. Phenol Distilled aqua phenol (Applegen) was equilibrated with TE buffer and stored a 40C in a dark bottle. Phenol/chloroform 50 % equilibrated phenol and 50 % chloroform/IAA were mixed and stored at 40C. 2.1.4. Solutions for restriction digests. React buffer I 500 mM Tris-HCI, pH 8.0 100 mM MgC12 React buffer II 500 mM Tris-HCI, pH 8.0 100 mM MgC12 500 mM NaCl React buffer III 500 mM Tris-HC1, pH 8.0 100 mM MgC12 IM NaCl 67 React buffer IV 200 mM Tris-HCI, pH 7.4 50 mM MgC12 500 mM KCl 2.1.5. Solutions for agarose gel electrophoresis. Ethidium bromide stock solution This was made up in H2O at 10 mg ml-1 and stored protected from light. Low and high melting point electrophoresis agarose were purchased from BRL. Molecular weight markers 1 kb ladder (BRL), fragment sizes from 75 bp-12.2 kb These were made up according to manufacturer's instructions. 10xTBE 0.9 M Tris base 0.9 M Boric acid 20 mM EDTA (diluted from the 0.5 M EDTA pH 8.0 stock solution) This solution was made up in H20. 2.1.6. Solutions for preparing GST fusion proteins. MTPBS 150 mM NaCI 16 mM Na2HPO4 4 mM NaH2PO4 Sterilized by autoclaving, and stored at 4°C. MTPBS, 1% Triton 5 ml Triton X-100 were added to 500 ml MTPBS. 68 MTPBS, NaCI 3.15 M NaCl 16 mM Na2HP04 4 mM NaH2PO4 Sterilized by autoclaving, and stored at 4°C. 1M1,10-Phenanthroline 1M1,10-phenanthroline monohydrate was made up in ethanol and stored at -200C. 100 mM PMSF 100 mM PMSF was dissolved in DMSO and stored at 4°C. DNase I stock 10 mg ml-1 DNase I was made up in H20, filter sterilized and stored at - 20°C. Elution buffer 50 mM Tris-HC1 pH 8.0, diluted from 1M stock 0.15 % reduced-glutathione Made up immediately before use. Reaction buffer for Factor Xa 20 mM Tris-HCI pH 8.0, diluted from the 1M stock 100 mM NaCI 2 mM CaCl2 Made up immediately before use. 2.1.7. Solutions for SDS polyacrylamide gel electrophoresis. 30 % Acrylamide 30 % (w/v) acrylamide was made up in H20, deionized, filtered and stored in a dark bottle at 4°C. 69 1% Bisacrylamide 1% (w/v) bisacrylamide was made up in H20, deionized, filtered and stored in a dark bottle at 4°C. 25 % APS 25 % (w/v) ammonium persulfate was made in H20, and stored at -20°C. Coomassie blue R250 stain 0.5 mg ml-1 Coomassie blue R250 powder was dissolved in 30 % methanol, 10 % acetic acid in H20. 3x Sample buffer 6% SDS 3% glycerol 0.05 % bromophenol blue 1.5 % 2-mercaptoethanol 187 mM Tris-HCI, pH 6.8, diluted from the 1.5 M stock solution This was made in H2O and stored at RT. 5x Gel running buffer 7.7 M glycine 1M Tris base 2% SDS 40 mM EDTA This was made up in H20. Destaining solution 30 % methanol 10 % acetic acid This was made in H20. Molecular weight markers BDH high molecular weight markers were used, They have a molecular weight of 200 kDa, 116 kDa, 97 kDa, 77 kDa, 55kDa and 43 kDa. The low molecular weight markers were obtained from Sigma and they have a molecular weight of 66 kDa, 45 kDa, 36 kDa, 29 kDa, 24 kDa, 20 kDa and 14 kDa. 70 2.1.8. Solutions for Western blotting. Transfer buffer 192 mM glycine 25 mM Tris base 20 % methanol Made up immediately before use in H20. 10 x TS 20 mM Tris-base 150 mM NaCl 10 x TS/Triton 0.5 % Triton X-100 in 10 x TS. Blocking buffer 5% non-fat skimmed milk powder in 1x TS. NBT BCIP 5% nitro blue tetrazolium in 70 % dimethylformamide. 5% bromochloroindolyl in 100 % dimethylformamide. A-P buffer 100 mM NaCl 5 mM MgC12 100 mM Tris base pH was adjusted to 9.5 with HCI, autoclaved and stored at 4°C. 2.1.9. Solutions for biotinylation and flow cytometry. 10 % NMS 10 % normal mouse serum was made up in 1x Dulbecco's PBS (Gibco). 71 Biotinylation buffer 3 mg ml-1 protein, which will be biotinylated 1 mg ml-1 NHS Biotin 10 % DMSO 0.1 M sodium hydrogen carbonate This was made up in H20. FITC The stock bought from Vector Laboratories was used at a 1/100 dilution in 1x Dulbecco's PBS (Gibco). Streptavidin/Phycoerythrin 10 µg ml-1 R-Phycoerythrin Streptavidin (Vector Laboratories) was made up in 1x Dulbecco's PBS (Gibco). Protein dilutions x µg Protein 0.5% BSA These were made in 1x Dulbecco's PBS (Gibco) and stored at -20°C. 2.1.10. Solutions for protein iodination and binding assays. TBS 20 mM Tris base 500 mM NaCl This was made up in H2O and the pH was adjusted to 7.5 with HC1. lodogen/chloroform 1 mg ml-1 iodogen (Pierce) was dissolved in chloroform. Binding buffer 500 µg ml-1 Bovine serum albumin (BSA) 2 mM MgC12 2 mM CaC12 5 mM glucose Made up in TBS, immediately before use. 72 2.1.13. Solutions for Southern blotting. Depurination solution 0.25 M HCI. Denaturing solution 1.5 M NaCl 0.5 M NaOH Neutralizing solution 1M Tris base 1.5 M NaCl pH adjusted to 8.0 with HC1. Hybridisation buffer 7% SDS in 125 mM phosphate bufffer, pH 7.5 Wash buffer 0.2 % SSC 0.1 % SDS 2.1.14. Solutions for RNA extractions and primer extensions. H20, DEPC treated 0.1 % DEPC (v/v) in H2O This was left overnight and then autoclaved to remove remaining DEPG 4M Guanidium isothiocyanate 4M guanidium isothiocyanate was made up in H2O and stored at -20°C. 0.5 M EDTA, DEPC treated The EDTA was made as described in 2.1.1. but before autoclaving it was treated with DEPC. 74 4M NaCOOH, pH 6.0, DEPC treated 4M NaCOOH was made up in H2O, the pH was adjusted to 6.0 with NaOH pellets, and then DEPC treated and autoclaved. 3M NaCOOH pH 6.2, DEPC treated 3M NaCOOH was made up in H20, the pH was adjusted to 6.2 with NaOH pellets, then was DEPC treated and autoclaved. 10 x One Phor ALL buffer (Pharmacia) 100 mM Tris-acetate pH 7.5 100 mM Magnesium acetate 500 mM Potassium acetate 3 mM ADP 3 mM ADP was made up in DEPC treated H20. 0.5 M PIPES, DEPC treated 0.5 M PIPES in H2O, DEPC treated and autoclaved. 3M NaCl, DEPC treated 3M NaCl in H20, DEPC treated and autoclaved. Hybridisation buffer 40 mM PIPES, pH 6.4, diluted from 0.5 M DEPC treated stock 1 mM EDTA pH 8.0, diluted from 0.5 M DEPC treated stock 0.4 M NaCl, diluted from the 3M DEPC treated stock Made up in formamide, giving a final concentration of 79 % formamide. I 0.1 M DTT 0.1 M DTT was made in DEPC treated H2O and stored at -20°C 5x Super Script buffer (Gibco BRL) 250 mM Tris-HC1, pH 8.3 375 mM KC1 15 mM MgC12 75 10 mM dNTP stock 2.5 mM dATP 2.5 mM dCTP 2.5 mM dGTP 2.5 mM dTTP in DEPC treated H20, stored at -200C. TE, DEPC treated This was prepared as described in 2.1.1 but before autoclaving it was DEPC treated. RNase A (DNase free) 20 mg ml-1 stocks were made in H2O, boiled for 15 minutes to destroy DNases and stored at -20°C. 1x Loading dye (Pharmacia) 0.3 % Bromophenol blue 0.3 % Xylene cyanol FF 10 mM EDTA, pH 7.5 97.5 % deionized formamide 2.1.15. Solutions for working with bacteriophage X. SM phage dilution buffer 0.1 M NaCl 8 mM MgSO4 50 mM Tris-HCI, pH 7.5, diluted from the 1M stock 0.001 % gelatin Sterilized by autoclaving. THE-50 10 mM Tris-HC1 (pH 7.5) 50 mM NaCI 1 mM EDTA 1 mM DTT Made up immediately just before use. 76 SW-block 2.5 % (w/v) dried milk powder 25 mM Hepes (pH 8.0) 1mMDTT 10 % (v/v) glycerol 50 mM NaCI 1 mM EDTA Made up immediately just before use. Top agarose 0.7 % agarose in LB medium, and was sterilized by autoclaving. 77 2.2. Methods. 2.2.1. Tissue culture. Culture of macroschizont infected cell lines from stocks cryopreserved in liquid nitrogen. Vials containing cryopreserved macroschizont cell lines were thawed at 37 °C. The contents were added to 10 ml prewarmed culture medium (see section 2.1.11. ) and the cells were pelleted at RT, 1100x g for 5 minutes. The cells were then resuspended in 5 ml culture medium, checked under a bifocal invertal optical microscope and incubated in a 50 cm3 tissue culture flask in an incubator at 37 °C in 5% CO2. The next day another 5 ml of culture medium were added. The method for culturing macroschizonts was adapted from Brown et al. (1987). Generally the cultures were passaged three times a week by diluting the cultures down to 2x 105 cells ml-1 with TBL medium. Cloning of macroschizont-infected cell lines by limiting dilution. The method for cloning macroschizont infected cell lines was adapted from Wathanga (1984) and Williamson (1988). Cells from an exponentially growing culture were diluted to an estimated 10 cells in 10 ml conditioned culture medium (see section 2.1.11. ). The 10 ml of medium containing the cells was mixed and 100 tl aliquots were plated out in 96 well tissue culture plates. The plates were checked to see whether any of the wells contained more than one cell per well, and any that did were discarded. The cells were allowed to grow up and then the process was repeated. Finally, the resulting cloned macroschizont cultures were expanded up to 10 ml cultures and treated as described above. 78 Cell lines. The cloned cell lines used in this thesis and their origins are listed in Table 6. All cloned cell lines with the exception of TaHBL3b were kindly given to me by Duncan Brown of the CTVM in Edinburgh. Other cell lines used were BL3 as described by Theilen et al. (1968), TaHBL3 (Baylis et al., 1992), TaH BL20 and Tall 46 (Shiels et al., 1986). Table 6 Origins of cloned macroschizont cell lines Cloned macroschizont cell line Origin of cell line TaA 46A Wathang a (1984) TaA 139D4 Wilkie (pers. comm. ) / Williamson (1988) TaA 139D7 Wilkie (Pers. comm. ) TaA 139E3 Wilkie (pers. comm. ) TaA 139E5 Wilkie (pers. comm. ) / Williamson (1988) TaA 46.2 Wathan ga (1984) TaH 46.2 Wathan ga (1984) TaH 46.3 Wathan ga (1984) TaH 46.4 Wathan ga (1984) TaA 139D6 Wilkie (pers. comm. ) TaA 46.3 Wathan ga (1984) TaHBL3b Katzer 2.2.2. Preparation of DNA. Plasmid mini preparation. This method was taken from Sambrook et al. (1989). 5m1 bacterial overnight cultures were spun down for 10 minutes at 1000 g. The resulting cell pellet was resuspended in 100 gl of cold solution I (see section 2.1.3). 200 gl of freshly made solution II (see section 2.1.3) was added and the solution was mixed, before 150 µl of solution III (see section 2.1.3) was added. This was mixed again and spun down (10,000x g 79 for 10 minutes). The DNA was extracted from the supernatant by phenol and phenol/chloroform extraction and was precipitated with ethanol. The DNA pellet was resuspended in TE. Wizard minipreps. The Wizard DNA Miniprep kit was purchased from Promega and the DNA extraction was conducted according to the manufacturer's instructions. For this method a 1.5 ml bacterial overnight culture was used to extract DNA, The cells were lysed using the buffers provided and the DNA was bound to a DNA binding resin, provided by the kit. The resin was purified on Wizard Miniprep columns and the DNA was eluted with TE which had been heated to 55°C. Preparation of single stranded M13. An overnight E. coli culture was diluted (1/100) with fresh 2x TY medium (1.5 ml -5 ml total volume). This culture was inoculated with a single plaque of bacteriophage M13 using a sterile wooden cocktail stick. This culture was grown for 5 hours at 37 °C. After the incubation 1.5 ml of the culture were spun down in an Eppendorf centrifuge (10,000x g for 10 minutes) and the supernatant was recentrifuged in order to pellet all bacterial cells. 200 µ1 of 20 % PEG/2.5 M NaCI was added to the supernatant, mixed and incubated for 15 minutes at RT to precipitate the M13 phage. The phage were spun down by centrifugation. Single stranded DNA was isolated by phenol/chloroform extraction, ethanol precipitated and resuspended in 20 gI TE. Genomic DNA preparations. For DNA preparations, macroschizont cultures were grown up in 100 ml of TBL medium (see 2.1.11. ) to a cell density of 2.0 x 106 cells per ml (2 x 108 cells in total). The cells were spun down for 10 minutes at 500x g, washed twice in PBS and once in 1x SSC. The cells were spun down again and resuspended in 4 ml 1x SSC, 4 ml THE and 1 ml 10 % sarkosyl. The suspension was mixed gently, Proteinase K was added to a final concentration of 100 ng/µl and was incubated for 2 hours at 550C. Following this incubation, the lysate was extracted once with phenol, 80 once with phenol/chloroform and once with chloroform, in order to remove any protein and other contaminants. The aqueous layer was dialysed against several changes of TE buffer at 4°C. Finally, the DNA concentration was estimated by measuring the optical density (OD) at 260 n m. Piroplasm DNA. T. annulata piroplasm DNA extracted from cultures which originated from field isolates from Ankara, Turkey (Schein et al., 1975) and Hisar, Morocco (Gill et al., 1976) was kindly given to me by Dr. R. Hall. 2.2.3. Electrophoresis and Southern blotting. Restriction digests. All restriction digests were carried out using the manufacturer's restriction enzymes (BRL), the appropriate buffer and recommended temperature. Plasmid DNA was digested for an average of two hours while genomic DNA was left to digest overnight. Agarose gel electrophoresis. DNA digested by restriction endonucleases was analysed by agarose gel electrophoresis as described by Sambrook et al. (1989) in TBE buffer (see section 2.1.5). Plasmid DNA was electrophoresed for a period of 1-4 hours, while cut genomic DNA was run out at a low voltage over night. Generally 1 gg plasmid DNA, 8-10 gg macroschizont cell line DNA or 2 tg piroplasm DNA was loaded per track. Molecular weight markers, described in section 2.1.5, were also loaded to allow the estimation of size of the DNA fragments. The DNA samples were diluted in type III gel loading buffer (Sambrook et al., 1989). Ethidium bromide at 1 µg ml-1, was included in the gel so that the DNA could be visualised on a UV transiluminator. 81 Recovery of DNA fragments from low melting point agarose gels. DNA fragments were run out on agarose gels made with low melting point agarose. The required fragment was cut out of the gel and the agarose was melted with 3 volumes of H2O at 65 °C. This was then phenol/chloroform extracted until all the agarose had been removed. Finally the DNA fragment was precipitated with ethanol. DNA recovery from agarose gels via Geneclean. The Geneclean II kit, purchased from BIO 101, was also used, following the manufacturer's instructions, to purify DNA fragments from agarose gels. In this method the band was exised from an agarose gel, 3 volumes of Nal stock solution was added and the agarose was dissolved at 50°C for up to 1 hour. Glassmilk was then added to the suspension and incubated, with occasional mixing, at RT for up to another hour allowing the DNA to bind to the glassmilk. The glassmilk/DNA complex was washed three times with New wash and finally the DNA was eluted from the glassmilk using TE (see section 2.1.1), warmed to 50°C. Random prime-labelling. DNA fragments for making labelled probes were cut out of low melting point agarose gels, melted in 3 volumes of H2O and made single stranded by boiling for 5 minutes. Approximately 25 ng of DNA were labelled with 50 jCi oc32P dCTP using the Random Priming DNA Labelling Kit (Boehringer-Mannheim Pharmaceuticals) according to the manufacturer's instructions. After labelling, the reaction mix was run through a spin column, as described by Sambrook et at. (1989), to separate the labelled probe from unincorporated radioactive nucleotides. Before use, the probe was boiled at 950C for 5 minutes to make the DNA single stranded. 82 Southern blotting. This method was adapted from the standard Southern blotting method (Sambrook et al., 1989 and Southern, 1975). After electrophoresis the gel was incubated for 15 minutes in the depurination solution (see section 2.1.13. ), followed by 20 minutes in denaturing solution (see section 2.1.13. ) and finally 30 minutes in neutralizing solution (see section 2.1.13. ). After a brief wash in H2O the DNA was transferred from the gel onto a nylon membrane (Hybond N, Amersham) overnight as described by Southern (1975). The membrane was washed in 2x SSC and then the DNA was fixed to the membrane by UV cross-linking for 7 minutes. Once the DNA was fixed to the membrane, the membrane was prehybridised in hybridisation buffer (see section 2.1.13. ) for 1 hour at 65°C. The probe prepared as discussed previously, was boiled and then added to the hybridisation buffer. This was followed by overnight incubation at 65°C with constant agitation. The membrane was then washed once with wash buffer (see section 2.1.13. ) at RT, followed by 2-3 further washes at 650C. The excess fluid was removed and the membrane was exposed to Kodak X-OMAT S film for several days in a film cassette containing an intensifying screen at -70° C. Finally the film was developed in a Kodak X-OMAT Developer. 2.2.4. Molecular cloning techniques. Competent cells. 20 ml of TYM broth were inoculated with the E. coli strain tgl recO and the bacteria were grown overnight. The culture was then diluted with 100 ml of warm TYM broth and the cells were grown until an OD of 0.5-0.9 at 600 nm. The culture was diluted again with 500 ml of warm TYM and was grown until an OD of 0.6 at 600 nm. The culture was then rapidly chilled and the bacteria were harvested by centrifugation. The cells were resuspended in 100 ml of cold competent cell buffer I and were left for 5 minutes. The cells were collected again by centrifugation and 83 resuspended in 20 ml of cold competent cell buffer H. Aliquots of 0.5 ml were pipetted into pre-chilled microfuge tubes and frozen in liquid nitrogen. These aliquots were stored for up to 2 month at -700C. Oligonucleotide primers. Table 7 shows all oligonucleotide primers used both for PCR cloning and generation of PCR fragments as well as oligonucleotide primers used for sequencing. Name of primer Sequence of primer 5' to 3' Origin of Sequence Position on Sequence 420 ct ccaattcttcc ttt SPAG-1, cDNA 1919-1900 646 ggagtagacttggcctag SPAG-1, cDNA 344-326 647 ctt ct a atcctcctcc SPAG-1, cDNA 1158-1139 710 ccaa aa ccat tacttc SPAG-1, cDNA 1450-1470 711 aa aa tttt aaa tttt SPAG-1, cDNA 2667-2688 815 tctaca acca a SPAG-1, cDNA 784-804 932 t actat ct aatat att SPAG-1, cDNA 2722-2700 FKI tataatattcatc tt at tt SPAG-1, cDNA -13 to +13 GEX For gcatggcctttgcaggg GEX Vector 855-871 pGEX Rev ct cat t gtcagaggttttcaccg pGEX Vector 1014-988 S p6 gatttaggtgacactatag pGEM-T 143-126 T7 taatac tcactataggg pGEM-T 2987-3 Y2 t tacca a aaa c SPAG-1, cDNA 945-963 Y3 gcggacaagatgcctgcggg SPAG-1, cDNA 53-83 Y11 cca atacaaaaat acc SPAG-1, cDNA 31-33 Y50 aatattatctcaaaacgagtgtg SPAG-1, DNA -371 to -347 Y51 ca ct tataa acc t ttta tc SPAG-1, DNA -241 to -263 Y82 atcctc aa ata c ccaa SPAG-2, cDNA 4-27 Y83 aattctactatcc aa ttccct SPAG-2, cDNA 367-353 Y84 aattccaaatt ctcccctt tt SPAG-2, cDNA 667-651 Table 7: The oligonucleotide primers used for sequencing and cloning. The name of the primers, their sequence, the origin of their sequence and the position they map to are listed. 84 Transformation. In general 1/3 of a ligation mix was used to transform 200 tl of competent E. coli strain tgl recO. The ligation mix and freshly thawed cells were mixed and incubated on ice for 30 minutes. Then 400 µl of LB was added to the suspension, mixed and the cells heat shocked for 2 minutes at 420C. After heat-shocking, the cells were incubated for 1 hour at 37 °C and aliquots of 40 µl, 75 µl, 150 µl and 300 µl were plated onto LB agar plates containing ampicillin. DNA amplification by polymerase chain reaction (PCR). The primers used for PCR are listed in the above table. In all cases the PCR reaction mix was made up to 50 µl containing 50 µmoles of each primer, 100 ng genomic DNA, 10 µM dNTP, 3 units of Taq XL and 1x Taq Buffer (Northumbria Biologicals Ltd). The reaction mix was incubated at 96°C for 5 minutes and this was followed by 30 cycles of 1 minute at 94°C, 1 min at 60°C and 2 minutes at 700C after which the reaction was left at 70°C for 10 minutes. The PCR products were separated by agarose gel electrophoresis, from which fragments were purified using the Geneclean method (see above for method). PCR cloning into pGEM-T. The PCR products were electrophoresed on agarose gels, the required DNA fragments were cut out, purified via Geneclean and cloned into pGEM-T (Promega) following the manufacturer's instructions. PCR cloning into pGEX-2T. PCR fragments were first cloned into pGEM-T (as above) and were sequenced to check for PCR errors. Correct fragments were cut out of pGEM-T using convenient Ban: HI and Eco RI restriction sites. The DNA fragment was purified from an agarose gel using Geneclean (as above). The purified fragment was ligated overnight into aBa ni HI / Eco RI digested pGEX-2T vector with a 3: 1 insert to vector ratio. 1/3 of the ligation mix was used to transform E. coll. 85 2.2.5. DNA sequencing. T7 DNA polymerase sequencing of double-stranded and single- stranded DNA. Sequencing was carried out using the dideoxy chain termination method of Sanger and co-workers (1977). A T7 DNA sequencing kit (Pharmacia) was used for all sequencing reactions. Double stranded DNA was prepared using the Wizard Miniprep method (see section 2.2.2. ), denatured with NaOH and ethanol precipitated. The primer was annealed to the denatured DNA at 37°C. Single stranded DNA was prepared from M13 as discussed in section 2.2.2. and the primer was annealed to the DNA at 650C. The remaining sequencing reactions were conducted according to the manufacturer's instructions using 35S dATP. 2.5 pl of the reaction was loaded onto 6% acrylamide, 0.5 x TBE non-gradient gels (Sambrook et al., 1989). The gels were fixed in 20 % ethanol, 20 % acetic acid for 30 minutes before being dried down onto 3MM paper at 80 °C for 2 hours. Finally the gels were exposed to Kodak X-OMAT S film in a film cassette at RT for 24 hours. 2.2.6. Preparation of recombinant proteins. Expression of GST-fusion proteins. 100 ml overnight cultures of E. coli tglrecO containing pGEX plasmids were grown up in LB medium (see section 2.1.2. ) and diluted by adding 1 litre of prewarmed LB. The cells were allowed to grow for 2 hours before protein expression was induced by adding IPTG to a final concentration of 1µM. The culture was then grown for another 2 hours at 37°C. The cells were then collected by centrifugation, resuspended in MTPBS/Triton (see section 2.1.6. ), containing 10 mM 1,10- phenanthroline, 10 mM PMSF and 10 gg DNase I. The cells were lysed by sonication and insoluble proteins and cell debris were removed by 86 centrifugation. The supernatant was passed through a glutathione 4B Sephadex column (Pharmacia) in order to affinity purify the fusion protein. The column was washed several times with MTPBS/Triton and with MTPBS. Finally the bound fusion protein was eluted with the elution buffer (see section 2.1.6. ). The eluted protein was collected in 1 ml fractions and subsequently visualised on SDS-PAGE (using a method adapted from Laemmli (1970)). The fractions containing protein were pooled and the protein concentration was estimated via the Bradford assay. Bradford assay. The protein concentration was estimated via the Bradford assay (Bradford, 1976), by mixing diluted protein, as well as known bovine serum albumin (BSA) standards, with Bradford reagent. These solutions were kept at RT for 15 minutes and then their OD was measured at 595 nm. The readings of the protein standards were converted into a standard curve which was used to estimate the protein concentration of the sample in question. Thrombin cleavage. After the protein concentration was estimated, the fusion protein was cleaved using thrombin to separate the parasite protein from its GST fusion partner. The protein was dialysed against 1x PBS (see 2.1.1. ) containing 2.5 mM CaC12.2 µg thrombin (Sigma) was added for every mg of fusion protein (Knight, 1993). The reaction mix was incubated for 2 hours at RT. The cleaved protein was then purified by running the reaction mix through a glutathione 4B sepharose column (Pharmacia) to bind the GST and the run-through was collected. The run-through was poured over the same regenerated column twice more, to make sure that all the uncleaved protein, as well as the GST, was removed. The cleaved recombinant protein was then requantified and dialysed against the required buffer for biotinylation or iodination as necessary. 87 Factor Xa cleavage For Factor Xa cleavage the OST fusion protein was dialysed against the appropriate reaction buffer (see section 2.1.6. ). 5-10 mg of fusion protein were incubated with 50 µg activated Factor X for up to 12 hours at RT. The cleaved protein was purified using the same method as described in Thrombin cleavage (see above). 2.2.7. Protein analysis. SDS"polyacrylamide gel electrophorsis. The method for SDS-PAGE was adapted from Laemmli (1970). The mini Protean II gel apparatus from BioRad was used. A resolving gel was poured first consisting generally of 10 % acrylamide and 0.13 % bisacrylamide in a 3.75 M Tris-HC1 pH 8.8 buffer and polymerized with 0.05 % TEMED and 0.05 % APS. The stacking gel consisted of 5% acrylamide, 0.135 % bisacrylamide, 125 mM Tris-HC1 pH 6.8, polymerised with 0.05 %APS and 0.1 % TEMED. Protein samples were diluted in Loading dye (see 2.1.7. ) and before application to the gel, and were heated for 5 minutes at 95°C. Generally, the gels were run at 200V for a period of about 45 minutes. The gels could then either be stained in Coomassie blue stain and destained in destaining solution (see 2.1.7. ), or were used for Western blotting. Western blotting. The method for Western blotting was adapted from Hunt and Hall (1993) and Towbin et al. (1979). Protein bands (0.5-1 µg) from SDS-PAGE gels were transferred onto nitrocellulose membrane (Sartorius) using a BioRad Protean II Western blotting apparatus. The SDS-PAGE gel was placed on the nitrocellulose membrane and placed in the Western blotting tank with transfer buffer (see section 2.1.8. ). The protein was usually transfered for 75 minutes at 150 mA. The membrane was blocked in blocking buffer (see section 2.1.8. ) for 30 minutes and was then incubated for 1 hour in primary antibody diluted in blocking buffer. The 88 antibody dilution chosen was according to Knight (1983). After this incubation the membrane was washed once in 1x TS/Triton and three times in 1x TS. The membrane was then incubated in the secondary antibody diluted in blocking buffer for a further hour. The membrane was then washed as before and bands reacting with the antibodies were detected using in A-P buffer with 0.06 % NBT and 0.03 % BCIP (see section 2.1.8. ). 2.2.8. Flow cytometry. Preparation of PBM cells. Bovine peripheral blood mononuclear cells (PBM) were prepared from fresh heparinised blood from calves. 10 ml of blood was mixed with an equal volume of Dulbecco's PBS and placed on a 10 ml histopaque 1086 (Sigma) bed. This was centrifuged at 1500x g for 30 minutes at RT to separate the PBM cells. After the spin, ' the PBM were pipetted off the histopaque/PBS interphase. They were then washed three times in Dulbecco's PBS and counted using a haemocytometer. Single colour Flow cytometry. PBM cells were prepared as above and 2x 105 cells were used per reaction. The cells were plated out in 96 well microtitre plates and were pelleted by centrifugation for 2 minutes. The cells were then resuspended in a volume of 25 µl containing either a diluted biotinylated antibody or a biotinylated protein. They were incubated at 4°C for 1 hour, washed three times in Dulbecco's PBS, and incubated in 25 gl of 1/100 diluted phycoerythrin streptavidin for 15 minutes. The cells were then washed twice in Dulbecco's PBS and analysed using a Becton and Dickinson FACScan flow cytometer. 89 Two colour Flow cytometry. The method for 2-colour flow cytometry was adapted from Sopp et al. (1991). PBM cells were prepared as above and 2x 105 cells were used per reaction. The cells were plated out in 96 well microtitre plates and were pelleted by centrifugation for 2 minutes. The cells were resuspended and incubated in 25µl of diluted monoclonal antibodies for 30 minutes at RT. The cells were washed 3 times in Dulbecco's PBS and incubated for 30 minutes in 1/100 diluted FITC-conjugated secondary antibody. After this incubation the cells were washed again in Dulbecco's PBS and incubated in 1% normal mouse serum for 15 minutes. This was followed by another wash and the cells were then incubated in either biotinylated proteins or biotinylated antibodies (30 minutes at RT). The cells were washed three times again and incubated in 1/100 diluted phycoerythrin streptavidin for 15 minutes. Then they were washed once more and analysed using a Becton and Dickinson FACScan flow cytometer. 2.2.9. Iodination and 1251 labelled protein binding assay. Iodination of proteins. This method was adapted from Sambrook et al. (1989) and Wrenn et al. (1988). 20 pl of 1 mg/ml iodogen (Pierce) in chloroform was pipetted into an eppendorf microcentrifuge tube which had been washed previously in ethanol. The iodogen was dried in an air stream. In a well ventilated fume hood, 40 pl of protein at a concentration of 1 mg/ml was added to this iodogen covered tube, together with 10 pI0.3 M phosphate buffer, pH 6.8, and 500 pCi of Na-I125. This was incubated for 10 minutes at RT. Then 750 pl of TBS was added to the reaction and the mixture was loaded on an equilibrated PD-10 G-25 (Sephadex) column (Pharmacia). The eppendorf tube was rinsed once more with 750 pl TBS which was also added to the column. The column was allowed to run dry before 5 ml of TBS was added to it. Fractions of 1 ml were collected and checked for radiation using a Geiger counter. The fractions containing most 90 radiation were pooled and their protein content as well as the success of the iodination reaction were tested by running a fraction of the sample onto a SDS PAGE gel. Binding assay. BL3 cells were grown in 100 ml cultures which were kept growing exponentially. These cells were harvested by centrifugation and were washed three times in Dulbecco's PBS. The cell number was estimated by counting a proportion on a haemocytometer. On average 4x 105 cells were used per reaction. These were incubated with a dilution series of iodinated protein. Each reaction with a given protein concentration was performed in triplicate and another identical triplicate was set up with a fifty-fold excess of unlabelled protein as a competitor. The cells were incubated for 1 hour at RT. The cells were then washed three times with TBS and tranferred to a fresh eppendorf tube. The cells were then spun down and a count of emitting gamma radiation was taken using ay radiation counter. 2.2.10. Sporozoite RNA extraction. Extraction of RNA from infected-tick salivary glands. The salivary glands of T. annulata infected ticks, which had been fed for two days on the ears of laboratory rabbits, were dissected by Lesley Bell-Sakyi. These salivary glands were homogenised in a sterile homogenises containing 4 ml of 4M Guanidium isothiocynate (see section 2.1.14. ) and ethanol precipitated. The RNA, DNA and protein was pelleted by centrifugation, and the pellet was dissolved in 5 ml of 50 mM EDTA (DEPC treated). This was followed by a phenol/chloroform extraction including 2 back extractions with 2.5 ml 50 mM EDTA (DEPC treated). 30 ml of 4M NaCOOH pH 6.0 was added and the solution was incubated overnight at 4°C, to precipitate the RNA. The RNA was pelleted by centrifugation and the very small pellet was dried and resuspended in 100 µl H2O (DEPC treated). The RNA was then quantified as described below. 91 RNA Quantification. The RNA was diluted 1/100 in DEPC-treated H2O and OD readings were taken against a blank containing DEPC-treated H2O at 260 and 280 nm. The RNA concentration was calculated knowing that 1 OD unit at 260 nm equals 40 µg ml-1 RNA and the purity of the RNA sample was determined by the 260/280 nm ratio. 2.2.11. Si intron mapping. The B am HI / Acc I fragment of the gH3.4 allele (position 1201- 2179 on the genomic SPAG-1 sequence) was cloned into M13 mp18 and M13 mpl9. The two resulting vectors were grown up (as described in section 2.2.2. ) with the addition of 120 t il inorganic 32phosphate (lOmCi/ml orthophosphoric acid). The resulting radioactively labelled single-stranded M13 DNA was extracted as described in section 2.2.2.50 µg of RNA, extracted from infected-tick salivary glands (see section 2.2.10. ) was added to 2 µl of the single-stranded M13 DNA. The RNA and the DNA were denatured in formamide and resuspended in hybridisation buffer and annealed at 50 °C for 14 hours. The RNA-DNA hybrid was ethanol precipitated and resuspended in S1-buffer. 600 units of Si nuclease were added to the suspension and incubated for 45 minutes at 37 °C to digest any single stranded DNA or RNA. After this incubation the suspension was phenol/chloroform extracted and the resulting RNA-DNA hybrid was ethanol precipitated and resuspended in loading dye. The product was denatured by heating to 95°C for 5 minutes and run out on a6% sequencing gel and visualised by autoradiography. 2.2.12. Primer extension. Endlabelling of primer. This method has been adapted from Sambrook et al. (1989). 0.5 µg of double stranded oligonucleotide were labelled in 1x One-Phor-All Buffer in the presence of 10 units of T4 polynucleotide kinase, 300 4M ADP and 50 µCi y-32P ATP, in a final volume of 10 µl. This reaction mix 92 was incubated for 1 hour at 37°C after which the unincorporated radioactive nucleotides were separated from the labelled primer using sephadex spin columns (Sambrook et al., 1989). Primer extension. This method was also adapted from Sambrook et al. (1989). 30 µg of infected tick salivary gland RNA and 1% of the endlabelled oligonucleotide were precipitated together in the presence of ethanol and sodium acetate. The pelleted RNA and primer were dried briefly and resuspended in hybridisation buffer (see secetion 2.1.14. ). The hybridisation buffer was heated to 85°C for 10 minutes and the primer was annealed to the RNA overnight at 37°C. The annealed primer and RNA were precipitated in ethanol, and then the primer was extended using reverse transcriptase in the appropriate buffer in the presence of RNA Guard, 1 mM dNTP and 10 mM DTT. The remaining RNA was digested with RNase I which was DNase free. The solution was then phenol/ chloroform extracted and the DNA was precipitated in ethanol. The DNA was collected by centrifugation, air dried and resuspended in 10 µl of 1x loading dye. Half of the extended primer was run out on a6% sequencing gel. 2.2.12. %gt11 library screening. Plating out bacteriophage ?,. Bacterial strain Y1090R- was grown up overnight in LB medium containing 0.2 % maltose, 10 mM MgSO4 and 50 tg ml- I ampicillin. The bacteria were harvested by centrifugation and resuspended in 1/4 of the volume of the overnight culture of 10 mM MgSO4. The bacteria could then be stored at 4°C for up to a week. LB plates containing ampicillin (50µg ml-1) were dried and prewarmed to 37°C. Top agarose was melted and cooled to 42°C. Plating bacteria and phage were mixed and incubated for 20 minutes at 37°C. Then the phage and bacteria were mixed with 3 ml of top agar and poured onto the LB plates. The plates were incubated overnight at 42°C. 93 Preparation of filters for library screening. Plating bacteria were prepared as above. 21 cm x 21 cm sterile plates containing 250 ml LB agar containing ampicillin were used for library screening. It was aimed to plate out 5x 104 phage per plate with 1.5 ml of plating cells. These were incubated for 20 minutes at 37°C and were plated out with 30 ml of top agarose. The plates were left at RT for 5 minutes to allow the top agar to set and were then incubated for 3 to 4 hours at 42°C. In the meantime, nitrocellulose membrane (Sartorius), measuring 20 cm x 20 cm, was soaked in 10 mM IPTG for 30 minutes and then dried. Once the phage plaques became visible on the LB plate, the membrane was placed on top of the top agarose. The phage were then left to grow up overnight. The following morning the filter was removed from the plate and was ready to be screened. Filter screening for DNA binding proteins. The method for screening the filters was adapted from Latchman (1993) and Singh et al. (1990). The filter was blocked in 500 ml of SW- block for 1 hour. Then the filter was incubated with 100 ml of THE-50 containing an endlabelled double stranded DNA probe and non-specific competitor DNA. The filter was incubated at RT for 1 hour and was then washed in 3 to 4 washes of THE-50, after which the filter was exposed to X-Ray film. 94 Chapter 3 Sequence comparison of SPAG-1 alleles: practical and functional importance. 3.1 Introduction. It has been proposed that SPAG-1 might function in evasion of the host's immune system and as a ligand for host-cell recognition (Hall et al., 1992). Due to selection pressure, sequences specifying structures involved in invasion of host cells by parasites will tend to be conserved. Structures mediating immune evasion could also be conserved in some situations e. g. where they are involved in molecular mimicry, or at the other extreme selection may produce extreme antigenic polymorphism. When such polymorphism occurs in immunologically relevant antigens, it poses a serious problem for the development of vaccines based on these molecules. This situation is widespread amongst infectious agents (Mendis et al., 1991). With the above considerations in mind I have obtained and compared the sequences of SPAG-1 alleles and identified polymorphic and conserved regions. The implications of the results of this analysis for the function of SPAG-1 and for vaccine development are discussed in this chapter. 3.1.1. The sequence analysis of SPAG-1 with respect to functional importance. The regions of SPAG-1 which may be involved in the invasion process of host cells, and those that are required for maintaining the conformation, and processing of the protein would be expected to be conserved among T. annulata strains. Other regions which are involved in the expression of the molecule on the surface of the sporozoite and its membrane anchor regions are expected to be conserved between species in related proteins, such as p67 of T. parva (Nene et al., 1992). Therefore I decided to study SPAG-1 sequences from different isolates in order to identify conserved regions of this molecule. The sequences of these 95 regions, if homologous to previously described motifs, might help to elucidate their function and that of the SPAG-1 protein. These results might provide further support for the theory that SPAG-1 is involved in the process of host cell recognition and invasion. Nevertheless these results will indicate a) whether more than one SPAG-1 allele exists, b) the extent of the SPAG-1 polymorphism and c) which parts of the protein are most conserved. 3.1.2. Sequence analysis of SPAG-1 with respect to vaccine development. The identification of variable and constant parts of SPAG-1 will be of importance for the future development of sub-unit vaccines against T. annula to because SPAG-1 is a candidate for inclusion in a sub-unit vaccine against tropical theileriosis. Determining the variablility of particular sequences will permit the inclusion of the optimal recombinant SPAG-1 components, so that a future vaccine could consist of constant, cross-protective regions of the protein and possibly a cocktail of all identified polymorphic regions. Although a single variant of one of the polymorphic regions will not induce cross-protective immune responses, it might nonetheless contain essential B cell and/or T cell epitopes for the induction of strain specific immune responses. 3.2 Results 3.2.1. Allelic RFLP analysis demonstrates that SPAG-1 is a single copy gene. Previous data demonstrated that the SRI fragment hybridised to 3 EcoRI restriction fragments (3.4,4.8 and 6.0 kb) in piroplasm DNA (Williamson et al., 1989). These data are confirmed and extended in the Southern blots shown in Figure 6. It can be seen that the piroplasm DNA of Ankara origin displays three bands, one at 3.4-kb, one at 4.8-kb and one at 6.0-kb (Figure 6, lanes 1 and 18). The DNA of Hisar origin shows 96 12345 ý+ - ýw t$ 100.6--w 0 67 w %, No .ý6.0 kb ý 4.8 kb ý 3.4 kb 89 10 11 12 13 14 15 16 17 18 ý " ýýýý -0 , .W ý 6.0 kb ý 4.8 kb ý 3.4 kb - w4 Figure 6: Southern blot analysis of the SPAG-1 associated RFLPs. Genomic DNA was digested with Lrco RI and the blot was probed with the SPAG-1 insert derived from Xgt11-SRI (Williamson et al., 1989). Lane 3 contains 15 µg DNA from uninfected BL3 cells, lanes 1 and 18 contain 2µg and 20 µg piroplasm DNA extracted from Ta Ankara, respectively and lanes 2 and 17 contain an equal amount of Ta Hisar stocks piroplasm DNA. Lanes 4-16 contain 15 µg of DNA extracted from the following cell lines: TaA 139D4 (lane 4), TaH 46.2 (lane 5), TaA 139D6 (lane 6), TaHBL3b (lane 7), TaA 46A (lane 8), TaA 139D7 (lane 9), TaA 139E3 (lane 10), TaA 46.2 (lane 11), TaA 139E5 (lane 12), TaA 46.3 (lane 13), TaHBL3 (lane 14), TaHBL20 (lane 15) and Tal-146 (lane 16). The bars mark the segregating RFLPs and give their sizes in kb. 97 only 2 bands, one at 3.4-kb and one at 6.0-kb (Figure 6, lanes 2 and 17) whereas Williamson et al. (1989) found that the 4.8 kb band was weakly visible in the Hisar piroplasm DNA. The presence of the 4.8-kb fragment appears to fluctuate during passage of the Hisar stock. This observation alone suggests that the multiple bands result from mixtures of distinct parasite stocks. To ascertain whether SPAG-1 is encoded by a single copy gene, an RFLP analysis was performed on DNA extracted from cloned macroschizont-infected cell lines (Table 8). In preliminary experiments I found that none of the available 11 clones contained the 6.0 kb EcoR I RFLP. I therefore isolated a twelfth clone called TaHBL3b from the TaHBL3 line (Table 8, see Materials and Methods for details). The extracted DNA was digested with Eco RI and subjected to Southern blotting with the SRI probe. The result of this analysis is shown in Figure 6. Uninfected BL3, a bovine negative control (lane 3), shows, as expected, no band which indicates that the SRI probe does not hybridise to bovine DNA. Lanes 4 to 13 contain the DNA of different cloned macroschizont-infected cell lines and in each track only a single band is visible. Each band co-migrates with one of the RFLPs found in the piroplasm DNA and each of the RFLP types is represented i. e. 3.4 kb in tracks 4,5,8-10,12,13; 4.8 kb in tracks 6 and 11; and 6.0 kb in track 7. Thus these RFLPs segregate clonally and this provides strong evidence that the SPAG-1 gene exists as a single copy. Each of the three restriction fragments behaves as alternative forms of a gene at a single locus and therefore I will henceforth adopt the term allele. The different alleles are named according to their isolate name and the restriction fragment size. Thus the alleles from the Hisar and Ankara isolates are prefixed by an H and an A respectively e. g. H3.4 is the Hisar allele marked by the 3.4 kb EcoRI fragment. A summary of the 12 cloned lines, their RFLP type and their allele designation is provided in Table 8. 98 Table 8 Cloned macroschizont infected cell lines, SPAG-1 alleles and Eco RI-SRI fragments they contain. Cloned macroschizont cell line Size of Eco RI-SRI fragments Allele TaA 46A 3.4 kb A 3.4 TaA 139D4 3.4 kb A 3.4 TaA 139D7 3.4 kb A 3.4 TaA 139E3 3.4 kb A 3.4 TaA 139E5 3.4 kb A 3.4 TaA 46.2 3.4 kb A 3.4 TaH 46.2 3.4 kb H 3.4 TaH 46.3 3.4 kb H 3.4 TaH 46.4 3.4 kb H 3.4 TaA 139D6 4.8 kb A 4.8 TaA 46.3 4.8 kb A 4.8 TaHBL3b 6.0 kb H 6.0 TaA and TaH denote T. annulata Ankara and Hisar, respectively. 3.2.2. Sequence analysis of a genomic copy (allele gH3.4) of SPAG-1. Since the above data demonstrate that there is polymorphism at the DNA level associated with the SPAG-1 locus, I decided to establish to what extent this was reflected in sequence variation of the structural gene. Therefore two pUC18 plasmids, containing a partially sequenced SPAG-1 genomic clone, were kindly given to me by Dr. R. Hall. The inserts of these plasmids originated from a phage clone which was isolated from a? EMBL3 library containing genomic Ta Hisar piroplasm DNA (Katzer et al., 1994). The phage clone contains a 13 kb insert which hybridises to the SRI probe (Williamson et al., 1989). One plasmid (pSPAG3.4) contains a 3.4-kb EcoRl fragment, which hybridises to the SRI probe on Southern blotting. The other plasmid (pSPAG1.4) contains a 1.4-kb EcoRI fragment, which is located adjacent to the 3.4-kb fragment in the original ?. EMBL3 clone and codes for the 5' region of the SPAG-1 gene (Figure 7). A panel of 182 m 13mp 18 clones containing random fragments (average size 300 bp) generated by sonicating pSPAG3.4 was also made available by Dr. R. Hall. The sequencing of these clones was at a very preliminary stage of analysis when I received them. The limited 99 information that was available indicated that this genomic SPAG-1 clone is not identical to the published SPAG-1 cDNA sequence (Hall et al., 1992). For ease of description these different copies will be referred to hereafter as gH3.4 (g. enomic I isar 3.4 kb EcoRI RFLP) and cH (P-D NA fromflisar). Compilation of my data from the 3.4 kb genomic fragment demonstrated that the first 297 bases of the coding sequence were lacking as deduced by comparison with the cH sequence. I subsequently obtained this information from a 600 bp Hind III-Eco RI fragment of pSPAG1.4 (see Figure 7). This was sub-cloned into ml3mpl8/19 and sequenced. The genomic SPAG-1 sequence data to this point was assembled using the GelAssemble programme of the UWGCG computer package (Devereux, 1989). A diagram of the alignment of 23 of the sequences (thin arrows) obtained from the m13 clones is shown in Figure 8. The thin arrows depicted in this figure represent only those sequences which provide all the essential information that could be derived from the analysis at this stage. For the sake of clarity the other 162 m13 clones which duplicated and consolidated this sequence information are not included in this diagram. Figure 8a shows the sequence information obtained in the 5' to 3' orientation and Figure 8b shows the sequence information obtained in the 3' to 5' orientation. In three locations sequences were only obtained in one direction and these are marked as shaded boxes (Figure 8). These sequence gaps were filled by double stranded DNA sequencing using specially designed oligonucleotides (Y2, Y3 and Y4, ). The sequence obtained with these primers is shown in Figure 8 by the thicker arrows. The final DNA sequence and its predicted amino acid sequence is shown in Figure 9. 100 A) B) ý C) 1.4 kb ý I I , ºdo 3.4 kb ý 3.4 kb Figure 7: A diagrammatic representation of the sub-cloning gH3.4 into plasmid pSPAG1.4 and pSPAG3.4. A) shows a representation of the genomic SPAG-1 fragment from the original EMBL3 clone; B) shows how the SPAG-1 gene was subcloned in two fragments, one was 1.4 kb and the other 3.4 kb in length; C) depicts the final pUC plasmids, pSPAGI. 4 and pSPAG3.4, containing the genomic SPAG-1 allele gH3.4. 101 A) sp19 sp17 sp176 sp69 sp45 sp71 sp39 sp15 sp123b sp143 sppb sp147 Shl º -º II"; i: i::: i;. ý ii: ci:: i:. {i; ü i: PAG-1 S ;::::::: Eco RI Y3 mmOlw- Y2 mmmNP-- B) sp157 sp149 sp117 sp120 sp98 sp152 sp116 sp162 5x8 d18 f- -010 q Eco RI SPAG-1 ýý Y4 Figure 8: Diagrammatic representation of the SPAG"1 gH3.4 sequence obtained for each DNA orientation. The large bar with SPAG-1 written in it represents the SPAG-1 gene. The internal Eco R1 site is marked. The thin arrows represent the sequence information obtained for the ml3mpl8 clones named at the left hand site. The boxed in areas represent region for which sequence information was obtained using one of the two DNA strands. The darker arrows represent the sequence obtained by double stranded DNA sequencing with SPAG-1 specific oligonucleotides. The name of the oligonucleotides used is written next to the beginning of the sequence. A) The sequence information for the 5' to 3' DNA orientation. B) The sequence information for the 3' to 5' DNA orientation. 102 1 CTTATGCTTAGGAGTAACATIGAAATTTAAATTTCATTTTCCAAAACTCAACGATGAACATTTTACACTTTCTGTTGACCATTCCGGTCA 90 MNILHFLLTIPVI 91 TTTTTGTATCTGGAGCGGACAAGATGCCTGCGGGAGAAAGTTCTAGAACCTCTAAACCCAGTCCCCTAGTAACCCTAGAATCAGCGGTAA 180 FVSGADKMPAGESSRTSKPSPLVTLESAVT 181 CACAACCTTCAAAAGACCCATTCAAGACAATTAGTGCCTTGTCAAAAGCAAGºAAAGTATGGAAGICAGCGGTATCAGTATCAGGTGACT 270 QPSKDPFKTISALSKATKVWKSAVSVSGDS 271 CTAAGACTGTTCCTACTCCAGTTTCGGAACCAATTATTACTCGATCITTTCAAGAACCAGTATCTCAAGAACTTGAATTCCAATCAGATA 360 KTVPTPVSEPIITRSFQEPVSQELEFQSDT 361 CTGAAATTAATGAGTCAGGATCCGGTTCAGATGAGGATGACGATGACGATGAGGAGGAGGAAGAAGAAGACGATGGATCCGGCTCATCTA 450 EINESGSGSDEDDDDDEEEEEEDDGSGSSK 451 AAGGTGCAAAAGGAAGCCCAAAAGCTCAGGCTGCAGTA'lCTPCAAGCAGTACATCCACAGCAAGTCCAAGTCTCCAACTACAACATCAT 540 GAKGSPKAQAAVSSSSTSTASPTSPTTTSS 541 CACAACCTGGCTTGGGATCAAGTGGTTCACACGGTCAACAAGTTCCAGGTGTAGGTGTTCCAGGTGTAGGAGTTCCAGGIGTAGGAGTTC 630 QPGLGSSGSHGQQVPGVGVPGVGVPGVGVP 631 CAGGTGTAGGAGTTCCAGGTGTAGGAGTTCCAGGTGTAGGAGTTCCAGGTGTAGGAGTTCCCGGTGTAGGTGTTCCCGGTGTAGGTGTTC 720 GVGVPGVGVPGVGVPGVGVPGVGVPGVGVP 721 CAGGGGTAGGCGTTCCAGGGGTAGGCGTACCAGGGGTTGGAGCTGTACCTGGAGIG('. GGGGGCTTC', GAGATGATAGTAGTGCATTGCCTG 810 GVGVPGVGVPGVGAVPGVGGLGDDSSALPG 811 GAAGTGGTGGTCTTGGAGCAGGAGCAAAGGCtC: GGAAAGGTCCAGGATCTGGTTTACAGGGACCAGGAGGTGTTGGACCAGGAGTACCTG 900 SGGLGAGAKAGKGPGSGLQGPGGVGPGVPG 901 GTGTAGGTGATGCAGC i-i'C-i-i'Ci-itT7"rACCAGGAAAACCTCCAGGTGTAGGAGTTCCTGGAGCAGTAGAACCTGGATTACCAGGAGCAG 990 VGDAASSSLPGKPPGVGVPGAVEPGLPGAA 991 CAGGTGTACCAGGAGGAAAGGCAGGAAAATCGAATAAATCTTCCGACCCTGAATTAGATTTT<'. GAGATGAATCTGATGGGTCAGGATCCG 1080 GVPGGKAGKSNKSSDPELDFGDESDGSGSG 1081 GTTCAGCAGGGGACGACGATGACAATGACGATGAAGAGGAAGAAGACGATAAATCTACCTCATCTAAAGGAGCAGGAGCAAA000TGGGA 1170 SAGDDDDNDDEEEEDDKSTSSKGAGAKAGK 1171 AAGGTCAAGGATCTGTATCACCAGGAGGAGGATCCTCAGCAAGTCAAACATCTTCAACTACAACATCATCACAATCTGGCTTGGCACCAA 1260 GQGSVSPGGGSSASQTSSTTTSSQSGLAPS 1261 GTGGTTCTCACCCTCAACAAGTTCCTCAACAAGATCCAGCGCTTAGTCAACCTAGTGGAGGAGGTGTGCCCGGAGTTGGAGTTCCCGGAG 1350 GSHPQQVPQQDPALSQPSGGGVPGVGVPGV 1351 TTGGAGTTCCIGGAGTTGGAGTTCCTGGAGTTGGAGTTCCTGGAGTTGGAGTACCCGGTGTTGGGGGTGCAACAACTTCATCATCATCAA 1440 GVPGVGVPGVGVPGVGVPGVGGATTSSSST 1441 CAACTTCAACTACTACTACTACTACAACATCATCTTCACCTGGAAAACC'ITCAAACCAAGGAAGCCATGGTACTTCTCCAAGAAAGCTAG 1530 TSTTTTTTTSSSPGKPSNQGSHGTSPRKLV 1531 TAACCAGACAAACTGACTCAATATCAGGACCCATACCATCACCAGGAGATCCAAGAGCAATTACTGGACAAATGGGTTTGTTATATTCAA 1620 TRQTDSISGPIPSPGDPRAITGQMGLLYSS 1621 GTAAATCTTTTGTAGGTGAAGGAGAAAGGTTTGCTGCACAGTTCTTGGGAGATTTTAAACCAAAACCAAGGAGATATGAAGGACAAGAAA 1710 KSFVGEGERFAAQFLGDFKPKPRRYEGQET 1711 CAGATGCAGTAAAACTAAAACAATTCATTTTTGAAGAGGTCAAATCGCTGGTGCAAACGTTAATAAACCTTAAATTAGCAATTGCAAACG 1800 DAVKLKQFIFEEVKSLVQTLINLKLAIAND 1801 ACTTTGTTGAAATCAGTGAAAAGTTGAAAAAGAAAAATCAAAATTACGTACCGAAATTAAAGTTGTTAAAAGGAGAACAATTTGACACCA 1890 FVEISEKLKKKNQNYVPKLKLLKGEQFDTK 1891 AACAGAAGGTAGCCAACGTGCTAAAGGGGTTCAATTCTCTGTACTTCGTATTTTTTATGAACCTTAACCTAGCGAAAGAAGTAAACAAAC 1980 QKVANVLKGFNSLYFV F- FMNLNLAKEVNKP 1981 CGGAAGAATTGGCAGAATTTCTTTC'. GAAACTAAATACAATCCCAGATAAAGTAGGAAGAGAATTTGAGTTAGCAATAGAAAAAACTAAAG 2070 EELAEFLWKLNTIPDKVGREFELAIEKTKG 2071 GTTCAGAGAAAAAGAAGGAATTAGAAGAAGCATTTAATTCAATA000TTAGGTITCAAAATAGCACAGTACGCAACAAATGACATCCTCT 2160 SEKKKELEEAFNSIGLGFKIAQYATNDILS 2161 CAAGTATAACAAATTCAGTCTACTCCCTGATAAAACTAAAGAATTTZGGAGATGATTTTATTACCGAAGTAAGAAAGTCGCTGCAAATGG 2250 SITNSVYSLIKLKNFGDDFITEVRKSLQMV 2251 TTCCACACCAAAAGAACCTAAACGGATCAGCGTTTATAGTCAAAATCTCAGAAATAATCAACAAAAAAGGAACAGAAGATGAGGATCAAA 2340 PHQKNLNGSAFIVKISEIINKKGTEDEDQT 2341 CATCAGGAAGTGGGTCAAAAGGGACAGAAGGAGTATCACTAAGGGGGCAAGATTTGACAGAAGAAGAAGTTTIGAAAGTITTGGATGAAC 2430 SGSGSKGTEGVSLRGQDLTEEEVLKVLDEL 2431 TAGTGAAGGATGTAAGCGAAGAACAGGTTGGAATAGGAGATTTAAGTGA000GAATAGCAGAACACCAAATGCAAAACCAGCCGAACTTG 2520 VKDVSEEQVGIGDLSDPNSRTPNAKPAELG 2521 GACCTTCACTAGTGATACAAAATGTACCATCAGACCCCTCAAAAGTGACACCAACACAGCCTTCAAATTTGCCACAAGTACCAACAACAG 2610 PSLVIQNVPSDPSKVTPTQPSNLPQVPTTG 2611 GGCCGGC'"GAACGGGACGGATGGAACAACAACAGGACCAGGTGGAAACGGGGAAGGAGGCAAAGATTTGAAGGAAGGAGAAAAGAAAGAAG 2700 PGNGTDGTTTGPGGNGEGGKDLKEGEKKEG 2701 GATTATiTCAAAAGATCAAAAACAAACTCTTGf', GCTCAGGATTCGAAGTCACAAGTATTATGATACCAATGACAACAATCATATTCAGTA 2790 LFQKIKNKLLGSGFEVTSIMIPMTTIIFSI 2791 TAGTCCACTAAAACTAAAAACACAACTAACCACACTAATTTATAATATACACAATAAAT 2849 VH' Figure 9: DNA sequence of the gH3.4 allele of SPAG"1. The predicted amino acid sequence is shown below. 103 3.2.3. The comparison of the cDNA (cH) and genomic (gH3.4) sequence of SPAG"1. The amino acid sequences predicted from the gH3.4 and cH copies of the SPAG-1 gene are compared in Figure 10a. These sequences are clearly different due to multiple point mutations, deletions and/or insertions. Therefore it is apparent, unless one invokes unusual post- transcriptional processing mechanisms, that the cH sequence is not the product of the RNA transcript of the gH3.4 gene. There is evidence, presented in chapter 4, consistent with one of the gaps (marked as a series of asterisks in Figure 10a) being an intron. The extent of the polymorphism is shown diagrammatically in Figure 10b. Overall, the degree of identity is 92% with the most conserved regions lying in the C terminal half (97% identity) and the N terminal quarter (92% identity). The second quarter of the molecule is the most variable region and can be divided into three sub-regions with values of 81%, 60% and 86% identity. A number of features highlighted in the original cH sequence (Hall et al., 1992) are conserved. Thus the two blocks of striking PGVGV pentapeptide elastin repeats (Raju and Anwar, 1987) are present; as are the glutamate/aspartate (D/E) and the threonine/serine rich (T/S) motifs. Interestingly the VGVAPG hexapeptide, which is identical to the elastin receptor ligand (Mecham et al., 1989), found three times in the cH sequence, is absent from the gH3.4 sequence (see Discussion and Chapter 5). The putative N terminal signal peptide (amino acids 1-18) and C terminal membrane anchor are well conserved, howerver. 104 qH3.4 cH 1 MNILHFLLTIPVIFVSGADKMPAGESSRTSKPSPLVTLESAVTQPSKDPFKTISAISKATKVM(SAVSVSCDSKTVPTPVSEPIITRSFQ III 1111111 IIIIIIIIIIIIIII111111IIIIIIIIIIIII111111IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 111111 MNIIHFLLTIPAIFVSGADKMPACESSRTSKPSPLVTLESAVTQPSKDPFKTISALSKATKVMKSAVSVSCDSKTVPTPVSEPMITRSFQ 90 EPVSQELEFQSDTE2NESGSGSDED. DDDDEEEEEEDDGSGSSICGAKGSPKAQAAVSSSSTSTASPTSPTTTSSQPGLGSSGSHGQQVPG Illllllllllllllllllllllll 1111 1111111 1 111 1111111 1111111 111111111 11 111 MI 11 11 91 EPVSQELEFQSDTEINESGSGSD£DEDDDDDEEEEEDDKSTSSKNGKGSPKAQPGVSSSSTSSASPTSPTTTLSQTGLGPSGSHAQQDPG 180 VCVPGVGVPCVGVPGVGVFGVCVPCVGVPCVGVPGVGVPGVGVPGVGVPGVGVPGVGAVPGVGCLCDDSSALPGSGGLGAGAKAGKGPCS IIIIIIIIIIIIIIIIII111I1111111111 1111111 111111111 111 11 III IIIIIIIIlil11111 II 282 VGVPCVGVPGVGVPGVGVPGVGVPGVGVPGVG.... CVPCVCVA.... PGVGVPGZIY=V. GVGADSSGLPGSGCLCAGAKAGKGQGS 261 GLQGPGCVGPCVPGVGDAASSSLPCKPPCVG.......... VPGAVEPGLPGAAGVPGGKAGKSNKSSDPELDFGDESDCSGSGSACDDD 111111111 11111 11111 11111111 111 11 1111111 11 11 1 111 1I 11 262 GLQGPGCVG. WPGVGVAASSSSPGKPPGVGACVNPGVGVRAQGGVIIGAPCVAGVPGGKPGQP. VSQELELKSDTEINESCSSSECEDD 349 DNDDEEEEDDKSTSSKCAGAKAGKGQGSVSPGGCSSASQ'TSSTTTSSQSGLAPSCSHPQQVPQQDPALSQPSGGCVPCVGVPGVGVPGVG 11 1111 111111111 111111111111111111111 III II111 1111 11 111111 I II111111111111111111 350 D. DEEEEEENKSTSSKGAGGKAGKGQGSVSP000SSASQTSPTTT. PQSGLASSGSHAQQSPQQDPAPSKPSGGGVPGVCVPGVGVPGVG 437 VPGVGV. PGVG. VPGVGGATTSSSST.. TS: TPITITSSSPGKPSNQGSHGTSPRKLVTRQTDSISGPIPSPGDPRAITWMGLLYSSKS II1111 1111 IIIIII11111111 III Iilll I II11 IIIIIIIII 11IIIIIIIIiIIIIIIiII1Ii11 438 VPGVGVAPGVGWPGVGGATTSSSSTTSTSTS2ZZTPZTSSGKPSDQGSHGTSPRNAVTRQTDSISGPIPSPGDPRAITGQM"""""""" S19 FVGEGERFAAQFLGDFKPKPRRYEGQETDAVXLKQFIFEEVKSLVQTLINLXLAIANDFVEISEXLXKKNQNYVPKLKLLXGEQFDTKQK 1111111 IIIIIIIIIIIIIIl1 IIlI1111111I11111I11I111II11II11IlII1111111111II1I11IIlllllllll S20 ""GEGERFAVQFLGDFKPKPRRYEGQGTDAVKLKQFIFEEVKSLVQTLINLKLAIANDFVEISEKLKKKNQNYVPKLKLLKGEQFDTKQK 607 VANVLKGFNSLYFVFFl4JLNLAKEVNKPEELAEFLWKLNTIPDKVGREFELAIEKTKGSEKKKELEEAFNSIGLGFKIAQYATNDILSSI IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII111111111111111111111111111111111111111 608 VANVLKGFNSLYFVFFIWLNIAKEVNKPEELAEFLWKLNTIPDKVGREFELAIEKTKGSEKKKELEEAFNSIGLGFKIAQYATNDILSSI 697 TNSVYSLIKLKNFGDDFITEVRKSLQlIVPHQKNLNCSAFIVKISEIINKKCTEDEDQTSGSGSKGTEGVSLRGQDLTEEEVLKVLDELVK IIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIilllllll IIIIIIIII1111 IIIIIlI1111111II1I111 698 TNSVYSLIKLKNFGODFVTEVRKSLQMVPHQKNLNGSAFIVKISEIINKKGTEDQDQTSGSGSKCTEGGSLRGQDLTEEEVLKVLDELVK 787 DVSEEQVGIGDLSDPNSRTPNAKPAEI. GPSLVIQNVPSDPSKVTPTQPSNLPQVPTTGPGNGTDGTI'IGPGCNCEGGKDLKEGEKKEGLF III11 IIII11111 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII11111IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 788 DVSEEHVGIGDLSDPSSRTPNAKPAELGPSLVIQNVPSDPSKVTPTQPSNLPQVPTrGPGNGTDGTI'PGPGGNGEGGKDLKECEKKEGLF 877 QKIKNKLLGSGFEVTSIMIPMTTIIFSIVH" 11111111111111 II IIIIIIIIIIIII 878 QKIIONKLLGSGFEVASIIIPMTTIIFSIVH" 907 Figure 10: Comparison of SPAG-1 protein sequences, derived from the DNA sequences of alleles gH3.4 and cH. a) The polypeptide sequences are compared using the GAP programme on the UWGCG analysis software package (Devereux, 1989). The VGVAPG motifs in cH are underlined and the region spanned by the putative intron (see chapter 4) is shown with asterisks. b) Schematic representation of the comparison shown in Figure 10a. The figures above the blocks represent the percentage identity over each segment calculated by scoring a gap as one change. The numbers below designate the amino acid residues at the boundaries of the sequence blocks based on the cH molecule shown in Figure 10a. 105 106 3.2.4. Further analysis of variable and constant regions of SPAG-1 alleles. The above sequence analysis of the cH and the gH3.4 alleles identified a highly polymorphic region towards the second quarter of the SPAG-1 molecule (Figure 10b). The C-terminus containing the 1A7 epitope (Williamson et al., 1989; Boulter et al., 1994) is highly conserved between these two alleles (see Figure 10b). Therefore a PCR based strategy was adopted to study the extent of sequence variation and conservation across these regions. The template DNA used for these PCR reactions was extracted from the macroschizont-infected cell line clones containing the identified segregating Eco RI RFLPs (see Table 8). 3.2.4.1. Sequence analysis of the most polymorphic region of SPAG-1. A DNA fragment of about 1860 bp (positions 53-1919 according to the cH sequence, Hall et at 1992), containing the polymorphic region, was amplified by PCR using primers 420 and Y3 (Materials & Methods and Figure 11) from the DNA of all 12 available cell line clones (see Table 8). Figure 12 shows a representative gel of four of these PCR products. The relevant bands were purified by Geneclean and cloned into pGEM-T. DNA minipreps were prepared using the Wizard reagents (Promega) and clones containing inserts were identified by agarose gel electrophoresis. A representative gel containing i the DNA of 11 such pGEM-T clones and a negative control is shown in Figure 13. The series of clones generated in this way are called "series 1" clones, as depicted in Figure 11. Using these clones as templates, I sequenced the most polymorphic region from bases 850 to 1107 (corresponding to amino acids 284 to 369 of SPAG-1). The primers used for sequencing are 815, Y2 and 647 as shown in Figure 11. Since Taq polymerase is error-prone, I sequenced three independent pGEM-T clones from each PCR reaction. The sequence comparison of all "series in clones, based upon 12 cloned macroschizont cell lines, revealed 4 DNA sequences which are highly polymorphic. These four different sequences, as well as the cH sequence and the gH3.4 sequence are shown in Figure 14a and their corresponding translated amino acid sequences are shown in Figure 14b. 107 Y3 Y2 IPCR CLONING Y3 IN- Y3 I 81_ S t! t-X2 ý SERIES 1 CLONES -000 646 647 ago "09 2 I 420 711 '- Y2 I SERIES 2 CLONES I 932 -"0932 Figure 11: Schematic representation of the PCR based cloning and sequencing strategy. The location of the primers used to generate the series 1 and series 2 fragments is shown by the boxed arrows. whilst those used for sequencing are shown by the smaller, underlined arrows. For precise details for each primer see Table 6 in the Materials and Methods section 2.2.4. 108 I Li '>0* "" 4 ý- 5 1860 bp I Figure 12: Agarose gel with PCR products, stained with ethidium bromide. The PCR products, shown in lanes 1-5, were obtained using the primers 420 and Y3 and the following cloned macroschizont-infected cell line DNA as template: no DNA (lane 1), TaA 46A (lane 2), TaA 139D7 (lane 3), TaA 139E3 (lane 4) and TaA 139E5 (lane 5). 1234S6789 101112 E III 0 4.8 Kb 3.0 Kb Figure 13: Agarose gel with DNA from pGEM-T clones, stained with ethidium bromide. DNA of three putative pGEM-T clones was extracted, using the Wizard miniprep method, for the PCR products of the following parasite clones : TaA 46A (lane 1-3), TaA 139D7 (lane 4-6), TaA 139E3 (lane 7-9) and TaA 139E5 (lane 10-12). The pGEM-T clone in lane 6 has no insert. The bars mark the size of the bands in kb. 109 a) cH CCAGGAAAACCTCCAGGAGTAGGAGCAGGAGTTATGCCTGGAGTTGGTGTACGAGCA A 3.4 --------------------------------------------------------- A 4.8 ---------- A---------------------------------------------- gH 3.4 ------------------------....... ................ -TT H 3.4 ------------------------.............................. -TT H 6.0 ----------A---T-C-----------------------------------A---- cH A 3.4 A 4.8 gH 3.4 H 3.4 H6.0 CAAGGAGGAGTAATAATTGGTGCGCCAGGAGTAGCAGGTGTG......... CCA... ------------------------------------------......... ---... -------------------------------C---------A......... ---... -CT---- C---- GA-CC ---ATTA------- C--------- A............... -CT---- C---- GA-CC ---ATTA------- C--------- A............... --------- C--------------------- C---- A-A... TCTGGATTA---AGG cH ............... GGAGGAAAGCCAGGACAACCA... GTATCTCAAGAACTTGAA A 3.4 ............... ---------------------... ------------------ A 4.8 ........................ G----- A--T-GAATAA---- TCC--C-C---- gH 3.4 ........................ G----- A--T-GAATAA---- TCC--C-C---- H 3.4 ............... ---------G----- A--T-GAATAA---- TCC--C-C---- H6.0 AGGAGCAGGTGTACC--GA----- G-----A--T--AAAAA---- TT---C-C---- cH CTGAAATCAGACACTGAAATTAATGAGTCAGGTTCCAGTTCAGAAGGGGAAGACGAT A 3.4 --------------------------------------------------------- A 4.8 T-A... -T--GAGA---- TC-G---- C----- A---------- G------ C---- T- gH 3.4 T-AG-T-TT-GAGA---- TC-G---G------ A---G------ C------ C------ H 3.4 T-AG-T-TT-GAGA---- TC-G---G------ A---G------ C------ C------ H6.0 TCAG-T-TT-GAGA----TC-G----C-----A----------G------C------ cH GAC ... GATGAAGAAGAGGAAGAAGAAAATAAATCTACCTCATCTAAAGGAGCAGGAGGA A 3.4 ---... ------------------------------------------------------ A 4.8 ---AAT--C--T-------------- CG-------------------------------- gH 3.4 ---AAT--C--T-------------- CG-------------------------------- H 3.4 ---AAT--C--T-------------- CG-------------------------------- H6.0 ----------------- A--C--------------------------------------- Figure 14: Comparison of the SPAG-1 alleles over the most polymorphic region (bases 850.1107, amino acids 284-369). PCR derived sequences are shown for each allele from series 1. Templates were derived from the same macroschizont clones as used in Figure 6. In addition the same region from alleles cH and gH3.4 is presented for further comparison. The data are given in (a) as the DNA sequence and in (b) as the derived peptide sequence. All alignments are made to the cH sequence and - represents identity, . represents a gap and any substitutions are shown by the relevant letter. 110 b) cH PGKPPGVGAGVMPGVGVRAQGGVIIGAPGVAGV ... P ...... GGKPG A 3.4 ---------------------------------... -...... ----- A 4.8 ---H-------------------------A---... -...... ---A- gH 3.4 H 3.4 --------.......... VP-A-EP-L--A---... -...... ---A- H6.0 ---H-A----------- QAQ--L------ A-R. SGL-RRSRCT-R-A- CH QP. VSQELELKSDTEIIQESGSSSEGEDDD. DýTSSKGAGG A 3.4 --. -------------------------- . ------------------ A 4.8 KSNK-SDP--. LGD-SDD----- G-D-V-N-D---- DD---------- gH 3.4 KSNK-SDP--DFGD-SDG---G-A-D---N-D---- DD--------- A H 3.4 KSNK-SDP--DFGD-SDG---G-A-D---N-D---- DD--------- A H6.0 KSKK-LDP-SDFGD-SDD-----G-D---. ----D------------- 111 Six cloned macroschizont cell lines of Ankara origin with the 3.4-kb Eco RI fragments all possessed the same sequence which therefore defines the A3.4 allele (Figure 14a and Table 8). Interestingly and unexpectedly this A3.4 allele matches the cH sequence totally over this stretch (see Discussion). The 3 cell line clones of Hisar origin with the 3.4-kb Eco RI fragment were identical to each other but distinct from A3.4 and thus define the allele H3.4 (Figure 14a and Table 8). The sequence of H3.4 is identical as expected to gH3.4, Two cloned macroschizont-infected cell lines of Ankara origin with the 4.8-kb Eco RI fragment are identical to each other and different to any other sequence and thus define the A4.8 allele (Figure 14a and Table 8). The sequence derived from the only cloned macroschizont-infected cell line of Hisar origin with an 6.0-kb Eco RI RFLP is named H6.0 and is also shown in Figure 14a. 3.2.4.2. Sequence analysis of the C-terminal constant region of SPAG-1. The C-terminal constant region of SPAG-1 was amplified by PCR using DNA obtained from the cloned cell lines, TaA 139D4, TaH 46.2, TaA 139D6 and TaHBL3b. These cell lines were selected as they represent the four different sequence variants identified in section 3.2.4.1. The PCR primers used were Y2 and 932 (Table 7) and the fragments of about 1780 bp generated were cloned into pGEM-T giving rise to the 'series 2' clones (see Figure 11). Three independent pGEM-T clones for each of the cell line clones were sequenced using the primers 711 and 932 (Table 7 and Figure 11). The region of the "series 2" clones which was sequenced corresponds to most of the original SRI region. The observed DNA sequences and the predicted amino acid sequences are shown in Figures 15a and 15b respectively. The comparison of the "series 2" clones revealed that this region is highly conserved with very few base pair changes which resulted in even fewer amino acid changes. Again four distinct sequences can be seen, where cH is identical to A 3.4, gH3.4 is identical to H 3.4, and A 4.8 and H 6.0 are different to any other sequence. Thus the analyses of the polymorphic and SRI regions are in agreement in that they indicate the existence of at least 4 SPAG-1 alleles. 112 a) cH ATAGGAGATTTAAGTGACCCAAGTAGCAGAACACCAAATGCAAAACCAGCCGAACTT A 3.4 --------------------------------------------------------- A 4.8 --------------------------------------------------------- gH 3.4 --------------------G-A---------------------------------- H 3.4 -------------------- G-A---------------------------------- H6.0 --------------------------------------------------------- p67 C-C----- C--------------- G-A---T--T-... C-A--G--A-C--TC---C cH GGACCTTCACTAGTGATACAAAATGTACCGTCAGACCCCTCAAAAGTGACACCAACA A 3.4 --------------------------------------------------------- A 4.8 ------------------------------------ T-------------------- gH 3.4 -----------------------------A--------------------------- H 3.4 ----------------------------- A--------------------------- H6.0 ------------------------------------T-------------------- p67 ----------- T--A---ACTG---G--AAG---GA---A---T---AT-T------ cH A 3.4 A 4.8 gH 3.4 H 3.4 H6.0 p67 cH A 3.4 A 4.8 gH 3.4 H 3.4 H6.0 p67 cH A 3.4 A 4.8 gH 3.4 H 3.4 H6.0 p67 CAGCCTTCAAATTTGCCACAAGTACCAACAACAGGGCCGGGGAACGGGACGGATGGA --------------------------------------------------------- --------------------------------------------------------- --------------------------------------------------------- --------------------------------------------------------- ------------------------------T----A--------------------- GG---CA---TAGCAG-TGG--G-GA-CA-C--CCTTC...... A-CTC-TA----- ACAACAACAGGACCAGGTGGAAACGGGGAAGGAGGCAAAGATTTGAAGGAAGGAGAA --------------------------------------------------------- --------------------------------------------------------- --------------------------------------------------------- --------------------------------------------------------- --CG----G--G-- ............... --C---A-C-C-ACCTG---A------G AAGAAAGAAGGATTATTTCAAAAGATCAAAAACAAACTCTTGGGCTCAGGATTCGAAGTC ------------------------------------------------------------ ------------------------------------------------------------ ------------------------------------------------------------ ------------------------------------------------------------ -----------------------------------------------------------A -------------- GA-A--G---C---- G--A------ C---- G--T------------ Figure 15: Comparison of the SPAG-1 alleles over the most constant region (bases 2419-2706, amino acids 806.902). PCR derived sequences are shown for each allele from series 2. Templates were derived from the same macroschizont clones as used in Figure 6. In addition the same region from alleles cH and gH3.4 and the p67 gene of T. parva is presented for further comparison. The data are given in (a) as the DNA sequence and in (b) as the derived peptide sequence. The 16 amino acids containing the 1A7 epitope (Boulter et al., 1994) are underlined. All alignments are made to the cH sequence and - represents identity, represents a gap and any substitutions are shown by the relevant letter. 113 b) cH IGDLSDPSSRT_F? NAKPAFT f, PST uIONVPSDPSKVTPTQPSNLPQVPT A 3.4 ------------------------------------------------ A 4.8 ------------------------------- S---------------- gH 3.4 -------N---------------------------------------- H 3.4 ------- N---------------------------------------- H6.0 -------------------------------S---------------- p67 L-------G-SS. ERQPS-------TDGQAG-TI-S--G-TIAAGGEQ CH TGPGNGTDG'ITTGPGGNGEGGKDLiEGF. KiCF. GLFQKIiQQKLLGSGFEV A 3.4 ------------------------------------------------ A 4.8 ------------------------------------------------ gH 3.4 ------------------------------------------------ H 3.4 ------------------------------------------------ H6.0 S----------------------------------------------- p67 PPSAPN.. --A---A-TQPEG..... -------I--L-K--------- 114 3.2.5. Comparison of SPAG-1 with the p67 antigen of Theileria parva: implications for the structure of the 1A7 epitope. The sequence comparison of SPAG-1 and p67 (Nene et al., 1992) reveals that overall they are 47 % identical at the amino acid level (Figurel6). Interestingly, the least similar region between these two antigens corresponds to the most polymorphic region within the SPAG-1 alleles. The most conserved parts are located at the N terminus (49% identity) and the C terminus (56% identity) again correlating well with the interallelic SPAG-1 comparison. The elastin repeat motif, PGVGV, only occurs once in p67 whereas as previously noted it is extensively repeated in SPAG-1. The SRI homologous regions of SPAG-1 and p67 are compared in Figure 15. The 16 amino acid sequence containing the neutralising 1A7 epitope (Boulter et al., 1994) is shown underlined. All alleles of SPAG-1 share this sequence but p67 has only 8 identical amino acids of which seven consecutive. This and the further observation that 1A7 and p67 weakly cross react (Knight, 1993) indicate that the 1A7 epitope is at least partially located in these shared residues. 3.2.6. The Eco RI RFLP pattern is based on the loss of an internal Eco RI site. It was shown that both the sequences of cH and gH3.4 contain the internal EcoRI site but both sequences were matched to the 3.4 kb EcoR I RFLP band, as cH is identical to A 3.4 and gH3.4 is identical to H 3.4. Both A 3.4 and H 3.4 have an EcoRI RFLP band of 3.4 kb on a Southern blot probed with SRI. No information was available about the internal EcoR I site of A 4.8 and H 6.0 and whether the EcoRI RFLP is located in the coding region of the SPAG-1 gene. Therefore the "series 1" clones for the cell line clones TaA 139D4, TaH 46.2, TaA 139D6 and TaHBL3b were sequenced using the primers Y3 and 646. The resulting sequence information is shown in Figure 17a and the predicted amino acid sequence is shown in Figure 17b. The EcoRI site is underlined in the cH sequence. It can be seen that cH, gH3.4, A 3.4 and H 3.4 have an intact EcoRI site while no EcoRI site is present at this site for A 4.8 and H 6.0. 115 49% 25% OVERALL IDENTITY 47% 56% Figure 16: Schematic representation of the comparison of the cH allele of SPAG-1 and p67 of T. parva. The figures above the blocks represent the % identity over each segment calculated by scoring a gap as one change. The numbers below designate the amino acid residues at the boundaries of the sequence blocks based on the cH molecule shown in Figure 10. 116 Therefore it can be concluded that the EcoRI RFLP pattern of 4.8-kb, seen in A 4.8, is due to a lack of the internal EcoRI site and that H 6.0 lacks this as well as a further Eco RI site outside the coding region of SPAG-1. The predicted amino acid sequence for this region shown in Figure l7b clearly shows that the change of the EcoRI site at DNA level also affects the predicted amino acid sequence. Therefore, the EcoRI RFLP is in this case a marker for polymorphism of the SPAG-1 gene at both the DNA and protein level. 277 312 93 104 cH GTATCTCAAGAACTTGAATTCCAATCAGATACTGAA VSQEL. EF-QSDTE A 3.4 ------------------------------------ A 4.8 --------------- ---C-GA---T---CC----- ------LKLCP- gH3.4 ------------------------------------ ------------ H 3.4 --------------- ---------------- ------------ H 6.0 -------- T--TGC------... --T--GAGA---T --HDA-L. LGDD Figure 17a Figure 17b Figure 17. Comparison of the SPAG-1 alleles across the internal Eco RI site. (a) Sequence comparison at DNA level (bases 277-312); the Eco RI site is underlined. (b) Sequence comparison at amino acid level (amino acids 93-104); the amino acids encoded by the bases involved in the Eco RI site are underlined. 117 3.3 Discussion. 3.3.1. SPAG-1 is a single copy gene. It has previously been shown that the piroplasm stocks of T. annulata contain a mixture of parasite strains (Wilkie et al., 1986). I have been able to confirm this by RFLP analysis. Further I provided evidence that the RFLPs behave as alternative forms of the same gene at a single locus and thus they behave as alleles. My data provides strong evidence in favour of there being one SPAG-1 gene per haploid genome. 3.3.2. Implications of the cDNA and genomic SPAG-1 sequence comparison. I completed the sequencing of a genomic SPAG-1 clone and compared this to the cH allele (see Figure 10a). This comparison clearly indicates that these two sequences represent two different SPAG-1 alleles. Further, the comparison of their predicted amino acid sequence shows that the two protein sequences are 92% identical (see Figure 10b). More interestingly, the C-terminal half and the N-terminal segment of the molecule are most conserved (97% identity and 92% identity respectively). The polymorphism is greatest in the central region of the molecule (only 60% identity). The variation is due to multiple gaps/insertions and amino acid substitutions. The presence of conserved and polymorphic parts are important for the elucidation of the molecules function as well as for sub-unit vaccine development. These implications are discussed below. Both alleles contain the PGVGV repeats which are homologous to the repeats found in bovine elastin (Raju and Anwar, 1987). Interestingly, unlike the cDNA allele cH, the gH3.4 allele product does not contain any VGVAPG motifs which considerably weakens the hypothesis that this represents a ligand for host cell recognition (Hall et al., 1992). The lack of the VGVAPG motifs in one allele and a lack of correlation between cells expressing the elastin receptor and their susceptibility to infection by sporozoites (Campbell et al., 1994) indicate that the elastin receptor is an unlikely candidate receptor for host cell 118 recognition. This suggests that the main function of the VGVAPG and the PGVGV motifs is most likely attributed to mimicry of the host and thereby the process of immune evasion (Hall et al., 1992: Hall, 1994). 3.3.3. Implications of the sequence comparison of the four SPAG-1 alleles. A number of conclusions can be drawn from the sequence analysis across the most polymorphic region of SPAG-1 from 12 macroschizont-infected cloned cell lines. First of all it is clear that at least 4 distinct SPAG-1 alleles exist. These are A3.4 (this is identical to cH), H3.4 (which is identical to gH3.4), A4.8 and H6.0. The second observation is that the 3.4-kb RFLP is composed of two distinct alleles, A3.4 and H3.4. Unfortunately no cloned macroschizont cell lines were available of Hisar origin with an 4.8-kb EcoRI restriction fragment or of Ankara origin with an 6.0-kb EcoRI restriction fragment. These would indicate the degree of polymorphim between the two isolates. Interestingly, the sequence of the Ta Hisar cH allele is identical to the Ta Ankara allele A3.4 over all regions investigated. Since the cDNA is of T. annulata Hisar origin (Hall et al., 1992) this observation implies that the Hisar stock also contains the A3.4 allele but tests to confirm this are still required. 3.3.4. Implications for vaccine development. The finding that the C terminal and N terminal regions of SPAG-1 are the most conserved regions is important information from the point of view of sub-unit vaccine design. This is reinforced by the recent epitope mapping studies which have located neutralising determinants recognised by bovine sera between residues 784 and 892 of the molecule (Boulter et al., 1994). Therefore, it can be predicted that these regions of the protein are likely, if included in a sub-unit vaccine, to induce cross protective immune responses. It is also of considerable interest that the most conserved region between T. parva p67 and SPAG-1 is in the C- terminal half of the molecule with 56% identity over amino acids 504-907 (Nene et al., 1992) as depicted in Figure 16. In addition the sequence 119 comparison of the four SPAG-1 alleles and p67 showed that a continuous stretch of 7 out of 16 residues containing the 1A7 epitope is conserved between SPAG-1 and p67. Since the monoclonal antibody 1A7 reacts weakly with recombinant p67 (Knight, 1993) it can be concluded that the epitope is at least partially located in these shared residues. This indicates that the 1A7 epitope is a cross reacting neutralising epitope for all isolated SPAG-1 alleles. This raises the idea of a common vaccine to T. annulata and T. parva designed on homologous regions in the C- terminus. In the future these results might assume more importance once more information about immunodominant structures on SPAG-1 is accrued. When T cell epitopes are located it will be possible to assign some form of rating to them, according to their general usefulness depending upon whether they locate to polymorphic or conserved segments. Further studies investigating the optimal sub-unit vaccine delivery system must be conducted to form a basis for the rational design of future vaccine constructs. 3.3.6. Implications for the functional importance of SPAG-1. The comparison of the the cH and gH3.4 allele indicated that the region towards the middle of the SPAG-1 molecule is highly polymorphic with only 60% identity at the amino acid level. While both the C terminus and the N terminus are highly conserved, 97% and 92% identity respectively between the two alleles. This might indicate that these regions are conserved for functional purposes. Since the 1A7 epitope is located in the C terminus of SPAG-1 and 1A7 blocks invasion of sporozoites into host cells in vitro as well as the fact that the C terminus is highly conserved might indicate that this part of SPAG-1 is involved either in the recognition event or in the subsequent invasion events into host cells. The predicted C-terminal membrane anchor (Hall et al., 1992) is highly conserved with only two conservative changes. The N terminus might be conserved for other purposes such as to facilitate the processing and folding of the protein and the predicted leader peptide (Hall et al., 1992) encoded by the first 18 amino acids is also found in 120 gH3.4 with only two conservative changes. Although the middle part of the protein is highly polymorphic, it might still have a very important function, but this does not involve sequence conservation. The middle part of SPAG-1 is most likely to be involved in the immune evasion process. This might also explain the finding of the PGVGV repeats which are conserved between gH3.4 and cH, as these mimic the host's own elastin molecule which has 11 such repeats. It was shown by Boulter (personal communication) that cattle infected with T. annulata or vaccinated with SPAG-1 do not develop antibodies against the PGVGV repeat structure. It can therefore be postulated that the elastin-like repeats make the molecule less immunogenic. 121 Chapter 4 Regulation of the expression of SPAG-1. 4.1. Introduction. The expression of the SPAG-1 gene is stage-specifically regulated. mRNA transcription is initiated when an infected tick starts feeding on its bovine host. No SPAG-1 mRNA is detectable in any of the later life cycle stages of the parasite (Williamson et al., 1989). Since a clone was available containing the 5' DNA sequence of SPAG-1 it was thought that this gene might make a good candidate to study transcriptional regulation during the sporozoite stage in Theileria. 4.2. Results. 4.2.1. Sequencing of the 5' untranslated region of SPAG-1. A pUC 18 clone containing a 1437 bp insert, containing the first 297 bp of the SPAG-1 gH3.4 allele and 1140 bp of the 5' untranslated region of SPAG-1 was given to me by Dr. R. Hall. The origin of this clone and the sequencing strategy of the SPAG-1 component is described in chapter 3. In order to investigate how SPAG-1 is regulated I decided to sequence the 5' region of SPAG-1 to identify sequences that might be involved in the stage-specific gene regulation of this gene. The insert of the pUC 18 clone was cut out of the vector and cut with Hind III which revealed two fragments. The first of these is 601 bp long and contains the first 297 bp of the SPAG-1 gH3.4 allele, and the other fragment of 836 bp in length as depicted in Figure 7 in chapter 3. Both these fragments were cloned into M13 mpl8 and M13 mpl9 for single stranded DNA sequence 122 d26 d2 d15 dt8 y50 f24 f15 f18 y3a b24 y51a g18 g15 g26 1 836 1141 1437 1437 bp insert of 5' Sequence of SPAG-1 SPAG-1 pUC 18 clone Eco RI Hind III Eco RI 5' to 3' Sequence information 1 574 750 1437 3' to 5' Sequence information 412 1437 -M Figure 18: Diagrammatic representation of the sequencing information for the S' untranslated region of SPAG-1. The thin arrows represent the length and position of individual sequencing reactions and their names are on the left hand side of the figure. The block in the middle of the figure represents the insert of a pUC 18 clone containing the first 1141 bp of the 5' untranslated region of SPAG-1 and the first 297 bp of the SPAG-1 gene. The large arrows at the bottom of the figure represent the sequence information obtained for each of the two DNA orientations. 123 analysis. The process of the sequence analysis is summarised in Figure 18. The first part of the figure shows the result of individual sequence reactions of the M13 clones and also the resulting three sequences using specific primers (Y3, Y50 and Y51) and double-stranded template DNA. The sequence was aligned using the UWGCG (Devereux, 1989) package on the SERC facility at the Daresbury laboratory. The summary of sequence information obtained for both DNA orientations is shown at the bottom of this figure. In the 5' to 3' DNA orientation there is a sequencing gap of 175 bases towards the middle of the 5' untranslated region. In the 3' to 5' DNA orientation there is a sequencing gap of 411 bases which are the most 5' bases of the pUC 18 clone. The resulting DNA sequence is shown in Figure 19 and the ATG start codon as well as the Hind III site are highlighted. 4.2.2. Sequence comparison of p67 and gH3.4. The 5' sequence of the gH3.4 allele of SPAG-1 was compared to the 5' sequence of p67, the T. parva homologue of SPAG-1, and this comparison is depicted in Figure 20. This comparison revealed that the 5' untranslated region of both genes is 75.6% identical. It is noteworthy that the level of identity in the 5' untranslated region between p67 and SPAG-1 is higher than in the coding region, since the level of identity at DNA level between the two genes is 52%. Since these two regions are so highly conserved it might be an indication that these regions have regulatory functions. Both genes are transcribed in a stage-specific manner, both during sporozoite development, and therefore it can be assumed that the DNA sequences for regulation of their expression are conserved. However, since the whole of the upstream region is very conserved it is not possible to pinpoint particular regions which might be involved in gene regulation. It has been proposed, since the genome of Theileria is very A/T rich, that regulatory elements might contain a higher percentage of G's and C's. One of the larger stretches which is conserved between p67 and SPAG-1 also contains a higher than average percentage of G's and C's. It is located 490 to 535 bp upstream from the the ATG start codon. This region contains two interesting features 1) a sequence, GCTAGC, which is repeated twice and 2) GAGCTC. Both sequences are palindromic. Since both these sequences are G/C rich and 124 1 AAAAGGAAGAGTTCATTTAAACTGTCAACCTCCATCTCGCTACAATGTAAACTTATAGAA 61 ATAACCAAGTGTTCATGACATATAAAAATATTTTAGAAACTTAAAAGACATATCCGAATT 121 CGTTGGAATCTAGAACGCTAATCAATGTTATTTTTACTAAAATTTGTAACATTTAATCTC 181 ATACCTAATATGAACTTCATGTAAACATAGGATTCACTTCAATAATTACAGTTATAGAAT 241 GAATAATTAACCAAAATTGATTTCTATAAAATAAATATTTAACTATTTAGATTAGATTTT 301 CGAATTTATGGTTGTGTTCACAAAAGGCATATCATAACATTTTTTCTAACGGCGAAAAAA 361 TTATTAAAAAATTATAAAACTAAAGTTTI"TAATGAAATATATGAATAATATGTAATTAAA 421 GTATAATTATGGTTAATTAGACTTTTAGATTCAAATTTATTAAATTGTGTGTAATCCTAT 481 ACTTAAAAATCTCATTTTTTGAATAATTCAAATTCTACATTTAAATTCATATTATATATG 541 TTAATTTAGATATATAAATGATGAGTTTAAGTGAAAATTGAGGGGAATTTAATATGCGAT 601 AGATTAATTGTTZGGCTAGCAGAATGAGCTCAAATAAAGAGCTAGCCAATGTCATTAAAG 661 CTCCAATAAACCCAAGATTAGACTCGTTCACTAGATATACTAATAAAATCCATCATTATT 721 TTTGTGCATTTATCATCGAGTTATTTTCACAAAAAATTATACTAACACACAACTGTATAA Y51 781 GACCGTGTTTAGTCTTTTTCCTTAAATTTGCTACCTATATCTGTAAAACA ft- 10 841 TTTTACAAATTACTATAAAAAACAAATAAAACAAAAACACACTCGTTTTGAGATAATTTT 901 CTTAATAACAAGTGTTTTAAAAAGGTAGGACTTTATCACCTAAAAGCAGAATAGTCTCAA 961 ATGCATTGAGATTAGAGGCTCCCTGATAATTGACTAAAAATGTTATATTACACAGCTTTT 1021 TATTCGATTAATTTATAACAATTATTTATAAATATTACTTATCGAGATAGTTTCTTTAAA 1081 ACTTCATCTTATGCTTAGGAGTAACATTGAAATTTAAATTTCATTTTCCAAAACTCAACG 1141 AT CATTTTACACTTTCTGTTGACCATTCCGGTCATTTTTGTATCTGGAGCGGACAAG --mF K111 1201 ATGCCTGCGGGAGAAAGTTCTAGAACCTCTAAACCCAGTCCCCTAGTAACCCTAGAATCA 1261 GCGGTAACACAACCTTCAAAAGACCCATTCAAGACAATTAGTGCCTTGTCAAAAGCAACA 1321 AAAGTATGGAAGTCAGCGGTATCAGTATCAGGTGACTCTAAGACTGTTCCTACTCCAGTT 1381 TCGGAACCAATTATTACTCGATCTTTTCAAGAACCAGTATCTCAAGAACTTGAATTC Figure 19: The sequence of 1141 bp of the S' untranslated region of SPAG-1 and the first 297 bp of the SPAG-1 gene. The ATG start codon is boxed and the Hind III restriction site for sub-cloning the two fragments for sequencing is circled. The arrows represent primers which have been used for primer extension analysis and sequencing. The arrow heads are at the 3' end of the primers and the name of each primer is written at the 5' end of the arrow. 125 Figure 20: Sequence comparison of the S' untranslated sequence of SPAG"1 of T. annzilara and p67 of T. parra. g13.4 represents the SPAG"I sequence. where g stands for genomie DNA. H indicates that this DNA is of Hisar origin and the 3.4 shows that this allele has a 3.4 kb Eco RI restriction fragment associated with it. The p67 marks the DNA sequence of the T. pan'a sequence. The figures at the end of each row represents the sequence position with reference to the ATO start codon. A line indicates that the bases compared are identical and a space indicates that they are different and a dot represents a sequence gap. There are two GCTAGC sequences which are boxed and one GAGCTC sequence which is circled. these sequences are palindromic and conserved between SPAG"I and p67 and their significance is discussed in the text. gH3.4 AAAAGGAAGAGTTCATTTAAACTGTCAACCTCCATCTCGCTACAATGTAAACTTATAGAAATAACCAAGTGTTCA -1053 III IIIIIIIIIIIIII IIIii 11 Illillll 1 IIIIIIIIIIIII IIIIM I 11 1( p67 AAAGGGAAGAGT'! 'c; AZTTGAACTGCCAGCCTCCATCGCAATACAATGTAAACTCATAGAATCTACTCTAAAATGA -1071 gH3.4 TGACATATAAAAATATTTTAGAAACTTAAAAGACATATCCGAATTCGTTGGAATCTAGAACGCTAATCAATGTTA -978 ý (ý Ilil IIIIIIIIIIIIIIilliliililli ýý (IIIIIIII ý Illiillllll ý ýý ý p67 CCAAATCTAAACATATTTTAGAAACTTAAAAGACATATTCGTCTTCGTTGGAGTTTAGAACGCTAAATATTGAAA -996 gH3.4 TTTTTACTAAAATTTGTAACATTTAA"""""""""TCTCATACCTAATATGAACTTCATGTAAACATAGGATTCA -912 III III I IIIIIIIIIII (II I IIIIIIIIII III IIIIIIII II III) p67 TTTACACTCACATTTGTAACATATAAATCACATAATTTCATACCTAAAATGTGCTTCATGTTATTACTCAATTCT -921 gH3.4 CTTCAATAATTACAGTTATAGAATGAATAATTAACCAAAATTGATTTCTATAAAATAAATATTT. AACTATTTAG -838 1 Illilllll 11111 111 111 11111 IIIIII 1 HIM I IIIIIIIIII 11 11111 ( p67 AATTCATAATTACATTTATATAATTAATTATTAATTAAAATTCAATTCTAT. ATATAAATATTTTAATTATTTGG -847 gH3.4 ATTAGATTTTC. GAATTTATGGTTGTGTTCACAAAAGGCATATCATAACATTTTTTC. TAACGGCGAAAAAATTA -765 .1 11 111 IIIII 11 11 11 IIII (1 11 11 IIIIIiI 1111 IIIIII I P67 GCCATATATTCCGAATT. ATAAACGTATTAACAATGGATGTGCCACAGCGTTTTTTCATGAATGAGTAAAAATAA -773 gH3.4 TTAAAAAATTATAAAACTAAAGTTTTTAATGAAA.. TATATGAATAATATGTAATTAAAGTATAATTATGGTTAA -692 -- . ..... ............ ... .......... ... ...... ..... . .... .. p67 TTAAF. AACGTGTAAAATCAAAGTTTTTAATAAAAKTaATATIGAATA'i"iATACAAATAAACTATAAATTTGG? AAA -öyes gH3.4 TTAGACTTTTA. GATTCAAATTTATTAAATTGTGTGTAATCCTATACTTAAAAATCTCA. TTTTTTGAATAATTC -619 III III III IIII IIIIIIII 11 Ilillil 1 111 1 IIIIiI 111 IIIIII 11 1 111 p67 TTAAACTATTATGATTACAATTTATTCAAGTGTGTGTTAACCTCTGGTTAAAATTTTTAATTTTTTTAAAATTTC -623 qH3.4 AAATTIIACA1 111 11 TTTAAATTCAIITTiTITATiTT. AATTTAGIIATiTAAAiGITGAGTTTAAIIIIIAPI, A4ll'TiGIGIGI -545 p67 AG. TTCTTTAACTAATTTTATATAAAAATAGATCAAAATCTAT. GAAAAATCAGATTTTTAAGTGAAAATT. AGG -551 aH3.4 GGAATTTAA. TATGCGATAGATTAATTGTTT I p67 GGA. AATTGAGCAAATCGTAAATTAATTGTTTI -1 GCTAGCAGAAYGAGCTCAAATAAAGAGCTAGCxAATGTCATTA -471 II, 111111 111 111111 11 111 111111 1111 11 1 GCTAGCGAA GAGCT AAAAAAC STAG AATGCTATGA -476 qH3.4 AAGCTCCAATAAACCCAAGATTAGACTCGTTCACTAGATATACTAATAAAATCCATCATTATT. TTTGTGCATTT -397 III IIIIII IIIIII Iilll 11 IIII IIII IIIIIIIII 11 1 11 11 IIIIIIIIIII P67 AAGGGGCAATAACCCCAAGGTTAGAGTCATTCAGTAGAAATACTAATACGATTAAGCAAATTTATTTGTGCATTT -401 gH3.4 ATCATCGAGTTATTTTCACAAAAAATTATACTAACACACAACTGTATAAGACCGTGTTT. AGTC. TTTTTCCTTA -324 1 11 11 Hill I 111 IIII 1 lill III1 IIIIII 11 11 1111 HIM III p67 ACTGTCAAATCATTTTGATAAATAATTGTTTTAACCTATAGCCCTGTAAGACTGTATTCAAGTCATTTTTCGTTA -326 gH3.4 AATT: 'GCTACCTP. TATCTGTAP. AACACPAGCTTATGTTTTACAAATTACTATAP. P. AAACAAATAAAFiCAAAAP. CA -249 111111 II 111 Illllllll II II Iliillllll IIIII IIII 1111111 lill III p67 AATTTGTCACATTTTTATGTAAAACAGAAACTCATGTTTTACACATTACCGTAAATAACAAATTAAACCAAAGTG -251 gH3.4 CACTCGTTTTGAGATAATTTTCTTAATAACAAGTGTTTTAAAAAGGTAGGACTTTATCACCTAAAAGCAGAATAG -189 p67 TACTCGTTTTGAGATAATTTTCTTATTAACAAGTGTTTCAAAAAGGTAGGACTTAATTAGTTAAGACCAGATTTA -191 gH3.4 TCTCAAATGCATTGAGATTAGAGGCTCCCTGATAATTGACTAAA...... AATGTTATATTACACAGCTTTTTAT -120 IIIIIIII I IIIIIIII 1 11 111 11 1 11 1 1111111 111 1 1111111 p67 TCTCAAATTCTCTGAGATTAAGCGGTCTTGGATTATAAAGTAGAATCACTAATGTTA... TACGAA.. TTTTTAT -121 9H3.4 TCGATTAA. TTTATAACAATTATTTATAAATATTACTTATCGAGATAGTTTCTTTAAAACTTCATCTTATGCTTA -47 p67 TCACTTAATTTTATATAAATTATTTATAAATATCACCTATCAAACTAATTTCTTTACACCTTCATCCTATGCTTA -47 gH3.4 GGAGTAACATTGAAATTTAAATTTCATTTTCCAAAACTCAACGATG p67 GGTGTAACATTGAAATTTGAATTTCACTTTCCGACACTGAACGATG 1111111 1 11111 ýýýýýýýýýýýý 111 ýýýýýýýýýý 111 11 1111 Hill I lill 11 126 are conserved between the two species they are candidate sequences which might be involved in the transcriptional regulation of both SPAG- 1 and p67. 4.2.3. Sequence comparison of the 5' untranslated region of 4 alleles of SPAG-1. In chapter 3, four cloned macroschizont infected cell lines were identified which contain 4 different SPAG-1 alleles. DNA from these cell line clones was used to PCR amplify a 374 bp fragment which maps to the position of -350 to +24 of the gH3.4 sequence. The primers used for this analysis are Y11 and Y51 and are shown in Figure 19. The PCR products were cloned into pGEM-T and sequenced. The sequence comparison over the first 350 bp upstream of the ATG start codon are shown in Figure 21. The 5' region of H3.4 is identical to that of gH3.4 which is in accordance with the results shown in chapter 3, but of interest is that the upstream sequence of alleles A4.8 and H6.0 is also identical. Only the upstream region of A3.4 is different from that of the other alleles by 8 bases. This sequence comparison indicates that the upstream region of the 4 alleles is highly conserved, probably for gene regulatory purposes. Since this region is conserved at such a high level no DNA sequence motifs could be identified which might be involved in the transcriptional regulation of SPAG-1. 4.2.4. Mapping the beginning of the SPAG-1 mRNA. To gain more information about the stage-specific regulation of SPAG-1, the beginning of the mRNA transcript of SPAG-1 was mapped using the primer extension method. For this method two primers were designed which are located around the ATG start codon. The first primer, FK1, maps to position bp 13 to -13 and the second primer Yll maps to bp 46 to 24, as shown in Figure 19. These primers were end-labelled with 32P, annealed to mRNA extracted from infected tick salivary glands and were reverse transcribed. The products were separated by acrylamide gel electrophoresis and visualised by autoradiography. The result is shown in Figure 22. The samples were run adjacent to 32P labelled 100 bp 127 gH3.4 -350 AGTCTTTITCCTTAAATTfGCTACCTATATCTGTAAAACACAAGCTTATGTTTTACAAAT A3.4 ------------------------------------------------------------ A4.8 ------------------------------------------------------------ H3.4 ------------------------------------------------------------ H6.0 ------------------------------------------------------------ gH3.4 -291 TACTATAAAAAACAAATAAAACAAAAACACACTCGTTTTGAGATAATTTTCTTAATAACA A3.4 ------------------------------------------------------------ A4.8 ------------------------------------------------------------ H3.4 ------------------------------------------------------------ H6.0 ------------------------------------------------------------ gH3.4 -231 AGTGTTTTAAAAAGGTAGGACT'ITATCACCTAAAAGCAGAATAGTCI'CAAATGCATTGAG A3.4 ------------------------------------------------------- C---- A4.6 ------------------------------------------------------------ H3.4 ------------------------------------------------------------ H6.0 ------------------------------------------------------------ gH3.4 -171 ATTAGAGGCTCCCTGATAATTGACTAAAAATGTTATATTACACAGCTTTTTATTCCATTA A3.4 -C--------------------------------------- G------------------ A4.8 ------------------------------------------------------------ H3.4 ------------------------------------------------------------ H6.0 ------------------------------------------------------------ gH3.4 -117 ATTTATAACAATTATTTATAAATATTACTTATCGAGATAGTTTCTTTAAAACTTCATCTT A3.4 ---------------------- A---------- T------------------------ C- A4.8 ------------------------------------------------------------ H3.4 ------------------------------------------------------------ H6.0 ------------------------------------------------------------ gH3.4 -58 ATGCTTAGGAGTAACATTGAAATTTAAATTTCATTTTCCAAAACTCAACG= A3.4 ------------------- G--------------------- C----------- A4.8 ----------------------------------------------------- H3.4 ----------------------------------------------------- H6.0 ----------------------------------------------------- Figure 21: Sequence comparison of the first 350 bp of the 5' untranslated sequence of SPAG-1 from 4 different alleles. gH3.4 indicates the DNA sequence from Figure 19. A3.4. A4.8, H3.4 and H6.0 represent the 5' DNA sequence obtained from four cloned macroschizont cell lines, representing 4 different SPAG-1 alleles. The number represents the DNA position with reference to the ATG start codon. A line indicates that the base at a given position is identical to the gH3.4 sequence and if a base is different the appropriate base is given. The ATG start codon is underlined. 128 3 330 bp sm Figure 22: Autoradiograph of the primer extension analysis. Lane 1 shows one band representing the primer extension product using primer FKI. Lane 2 contains molecular weight markers and lane 3 is the primer extension product using primer Yll. The bars next to lanes indicate the size of the observed bands in bp. 0 300 bp 129 markers, to allow an estimation of size of the resulting bands. For each of the reverse transcription reactions there is only one band visible. The size of the band for the reaction containing the primer FK1 is about 290 bp and the band for the reaction containing the primer Y11 has a size of about 330 bp. Taking into account the starting position of each primer, the beginning of the mRNA for SPAG-1 was mapped to aC at 278 bp 5' of the ATG start codon. The transcription initiation site is highlighted by an arrow in Figure 23. 4.2.5. Putative promoter binding sites in the 5' untranslated region of SPAG-1. Once the beginning of the mRNA for SPAG-1 was mapped it became of interest to look for putative DNA binding sites of proteins which might be involved in the regulation of SPAG-1 expression. The 5' untranslated region of SPAG-1 was searched for an array of DNA sequences which have been shown to be DNA binding sites of proteins and which are involved in the regulation of transcription in other organisms. A list of these sequences and their DNA binding proteins with relevant references is shown in Table 3. Three of these sequences were found in the 5' untranslated region of SPAG-1. A total of 8 TATA boxes were found in the region investigated. The presence of these is not very surprising due to the A/T richness of the Th eile ri a genome, but of interest is the finding of the sequence, CAACTG. This sequence in chicken has been shown to be the site to which the Myb oncogene product can bind (Biedenkapp et al., 1988) and it has been shown that the Myb protein is a transcriptional activator which functions by binding to this sequence (Weston and Bishop, 1989). Another two identical sequences, ATGAATAA, were also found in the 5' untranscribed region of SPAG-1. ATGAATAA has been shown to be the site to which Pit-1 binds (Ingraham et al., 1988). The Pit-1 sequence is found in two genes which are expressed specifically in the anterior pituitary gland and it has been shown that the Pit-1 protein binds to this sequence (Ingraham et al., 1988). Subsequently, transformation experiments showed that the Pit-1 sequence is an effective promoter sequence in pituitary cells (Esholtz et al., 1990). The 5' sequence of p67 was searched for the existence of these 130 -1140 AAAAGGAAGAGTTCATTTAAACTGTCAACCTCCATCTCGCTACAATGTAAACTTATAGAA -1080 ATAACCAAGTGTTCATGA AT TATTTTAGAAACTTAAAAGACATATCCGAATT -1020 CGTTGGAATCTAGAA000TAATCAATGTTATTTTTACTAAAATTTGTAACATTTAATCTC -960 ATACCTAATATGAACTTCATGTAAACATAGGATTCACTTCAATAATTACAGTTATAGG =T -900 GAATAATTAACCAAAATTGATTT AT 3OQrAAATATTTAACTATTTAGATTAGATTTT -840 CGAATTTATGGTTGTGTTCACAAAAGGCATATCATAACATTTTTTCTAACGGCGAAAAAA -780 TTATTAAAAAA ATAAAA TAAAGTTTITAATGAAATA AýTATGTAATTAAA -720 GTATAATTATGGTTAATTAGACTTI'TAGATTCAAATTTATTAAATTGTGTGTAATCCTAT -660 ACTTAAAAATCTCATTTTTTGAATAAZTCAAATTCTACATTTAAATTCATA ATATA G -600 TTAATTTA TA AT T ATGAGTTTAAGTGAAAATTGAGGGGAATTTAATATGCGAT -540 AGATTAATTGTTTGGCTAGCAGAATGAGCTCAAATAAAGAGCTAGCCAATGTCATTAAAG -480 CTCCAATAAACCCAAGATTAGACTCGTTCACTAGATATACTAATAAAATCCATCATTATT -420 TTTGTGCATTTATCATCGAGTTATTTTCACAAAAAATTATACTAACA -360 GACCGTGTTTAGTCTTTTTCC, -i7. r TTTGCTACCTATATCTGTAAAACACAAGCTTATG -300 TTTTACAAATTA TAT CAAATAAAACAAAAACACACTCGTTTTGAGATAATTTT -240 CTTAATAACAAGTGTTTTAAAAAGGTAGGACTTTATCACCTAAAAGCAGAATAGTCTCAA -180 ATGCATTGAGATTAGAGGCTCCCTGATAATTGACTAAAAATGTTATATTACACAGCTTTT -120 TATTCGATTAATTTATAACAATTA ATAAA TTACTTATCGAGATAGTTTCTTTAAA -60 ACTTCATCTTATGCTTAGGAGTAACATTGAAATTTAAATTTCATTTTCCAAAACTCAACG CATTTTACACTTTCTGTTGACCATI'CCGGTCAZT'I'TTGTATCTGGAGCGGACAAG Figure 23: Sequence motifs in the S' untranslated region of SPAG-1. The 5' untranslated region of SPAG-1 gH3.4 is shown. The numbers represent the sequence position with reference to the ATG start codon. The ATG start codon is in a thick lined box. The transcription initiation site is marked by an arrow head. 8 sequences which fulfil the criteria of a TATA box are boxed with thin lines. The sequence which is in accordance to the consensus sequence of the Myb binding site is circled and two sequences identical to the consensus sequence for a Pit-1 binding site are underlined with a thick line. The two sequences which are underlined with a thin line and the sequence which is underlined with a medium thickness line were identified in a sequence comparison of 5' region of p67 and SPAG-1, their significance is discussed in the text. 131 sequence motifs. None of the Pit-1 sequences were intact nor was the Myb binding site and only one of the eight TATA boxes was conserved between p67 and SPAG-1. This TATA box lies between the transcription initiation site and the ATG start codon, therefore it is unlikely to be involved in the regulation of SPAG-1 expression. Attempts to isolate DNA binding proteins that bind to the first 371 bases of the 5' untranslated region of SPAG-1 by screening a %g t11 expression library containing T. annulata genomic DNA failed. Therefore no novel or known DNA binding proteins were identified which bind to the 5' untranslated region of SPAG-1 and which could be involved in the regulation of SPAG-1 transcription. The regulatory mechanism of SPAG-1 and p67 transcription therefore remains unknown. 4.2.6. Evidence for an intron in SPAG-1. When the SPAG-1 sequence, gH3.4, which is derived from a genomic T. annulata Hisar DNA stock, was compared to the published cDNA sequence, cH, of SPAG-1 in the previous chapter, a putative intron was recognized and marked as such in Figure 7. The putative intron is based on the observation of a sequence gap of 30 bp in the cH sequence in the comparison between the two sequences. The presence of these 30 bp in the gH3.4 could of course be explained by an insertion in the gH3.4 sequence or a deletion in the cH sequence. As the length of this sequence difference is a multiple of 3 base pairs, it does not interrupt the open reading frame, even if it is an intron. If this sequence difference is due to an intron, it is therefore a cryptic intron. It was decided to investigate further whether this sequence difference is due to a cryptic intron. Four different SPAG-1 alleles were identified in chapter 3 and one of these, allele A3.4, was shown to be identical to the published cDNA sequence, cH, over the two regions investigated. The pGEM-T clone, a series 1 clone from chapter 3, containing genomic DNA for the A3.4 allele was sequenced across the region containing the putative intron 132 using site specific primers, 420 and 710. The sequence of the putative intron from the A3.4 allele was aligned to the cH sequence and is shown in Figure 24. The series 1 pGEM-T clones for the other 3 alleles were also sequenced across the putative intron and their sequences are also shown in Figure 24. It can clearly be seen that all sequences, based on genomic DNA, contain the sequence of the putative intron and that cH does not contain this sequence. A3.4 is identical to cH in all sequence comparisons made so far apart from the presence of the putative cryptic intron. It therefore provides strong supporting evidence that the 30 bp comprises a cryptic intron. Interestingly, this intron is found in all 4 alleles and they are identical. This sequence conservation indicates that most bases in this short cryptic intron are probably needed for the excision of the intron during mRNA processing. cH 1595 GGACAAATGGI ...................... ........ IGTGAAGGAGA 1614 gH3.4 1596 GGACAAATGGIgtttgttatattcaagtaaatcttttgtaglGTGAAGGAGA 1645 A3.4 GGACAAATGGIgtttgttatattcaagtaaatcttttgtaglGTGAAGGAGA A3.8 GGACAAATGGIgtttgttatattcaagtaaatcttttgtaglGTGAAGGAGA H3.4 GGACAAATGGIgtttgttatattcaagtaaatcttttgtaglGTGAAGGAGA H6.0 GGACAAATGGIgtttgttatattcaagtaaatcttttgtaglGTGAAGGAGA Figure 24: The SPAG-1 intron. A sequence comparison of the cDNA sequence cH of SPAG-1 and the corresponding sequence from 4 SPAG-1 alleles (A3.4, A4.8, H3.4 and H6.0) and the genomic SPAG-1 sequence gH3.4. The dots represent sequence gaps in the cDNA sequence and the sequence in lower case letters represents the sequence of the intron. The line indicates the start and the end of the intron. The numbers represent the position of the beginning and the end of the sequence shown in the cDNA and the genomic sequence respectively. 133 4.2.7. Confirmation of intron by Si mapping. To provide more supporting evidence that the 30 bp are a cryptic intron, the method of Si mapping was chosen. A BamHI-Accl fragment of the pUC 18 plasmid, described in chapter 3, containing most of the genomic copy of the gH3.4 SPAG-1 gene was cloned into M13 mp18 and M13 mp19. Both these clones were grown in the presence of organic 32P in an otherwise normal overnight culture. Subsequently, the single stranded DNA was extracted for both M13 clones. The M13 inserts map to the position 1201 to 2179 on the gH3.4 sequence and the putative intron maps to the position 1606-1635. The single stranded DNA was then annealed to sporozoite mRNA extracted from infected tick salivary glands. The RNA-DNA hybrid was treated with the S1 nuclease which digests single stranded DNA and single stranded RNA but not RNA-DNA hybrids. The products were denatured and separated on a 6% polyacrylamide sequencing gel next to a labelled marker track and the results were visualised by auto radiography. A picture of the autoradiograph is shown in Figure 25. In Figure 25 it can clearly be seen that there are 2 bands in the track containing the S1 treated RNA-M13 mpl8 hybrids while no bands were observed in the track containing the negative control. S1 treated RNA-M13 mpl9 hybrids. If no intron was present in the DNA fragment one would expect only a single band but two bands indicate the presence of one intron. If one subtracts the total size of the two bands from the size of the initial DNA fragment one gets an estimated intron size of about 30 bp which coincides with the predicted intron size from the sequence comparison described above. Furthermore, by matching the sizes of the bands with the starting and finishing position of the initial DNA fragment one can predict the position of the intron to be in either of two sites. From this the intron is either at position 1606-1635 or at 1744-1773. The first of these is in accordance with the predicted intron found by sequence comparison of the cDNA and genomic DNA sequence of SPAG-1 as well as the predicted position from the DNA sequences of the macroschizont clones. 134 123 1000 bp . V`r' ý ý, . 71 a db " ý e» ) ,1't;;, ý 4Q5bp Figure 25: Autoradiograph of the Si nuclease intron mapping experiment. Lane 1 contains molecular weight markers. lane 2 contains the product of SI nuclease treated DNA-RNA hybrid and lane 3 contains the negative control. The bars next to lane 1 and 3 represent the molecular weight of the observed bands in bp. 135 4.2.8. Comparison of the SPAG-1 intron to other Theileria introns. In total only three introns have been identified in Theileria. The reason for this is that only a few genes have been cloned and fully sequenced and in most cases the sequence information is based only on a cDNA sequence, and introns would not therefore have been detected. Two further genes have been analysed for introns, the hsp 70.1 gene (Mason et al., 1989) and the cysteine protease gene of T. annulata (Baylis et al., 1992) and it has been shown that neither contain introns. The sequence of the SPAG"1 intron has been aligned to the intron sequences of the two other identified Theileria and the cysteine protease sequence comparison can short, varying from 28 to allow the prediction of a and excision of introns in is Glgttngtt ... ttntagIG. be identified in Theileria existing sequences. introns, from the p67 gene (Nene et al., 1992) intron of T. parva (Nene et al., 1990). T he be seen in Figure 26. All three introns are 32 bp. The comparison of these three introns short consensus sequence for the recognition Theileria. The predicted consensus sequence To verify this prediction more introns need to genes and these also need to be aligned to the T. a. SPAG-1 AAATGGIgtttgttatattcaagta. aat. cttttgtaglGTGAAG T. p. p67 --TCA-I--------- t--. g----. t--. ggg--t---I-CA-TA T. p. cysteine protease ---A--I---a---- c-ca--ct--at--ta---- a---1----T- Consensus Gigttngtt ttntaglG Figure 26: Sequence comparison of introns found in Theileria. The sequence in capital letters represents the coding sequence flanking the intron, the bar marks the start and end of the intron and the sequence of the intron is written in lower case letters. The . represents a gap, the - represents identical bases, T. a. and T. p. represent T. annulata and T. parva respectively. 136 4.3. Discussion 4.3.1. Stage-specific regulation of SPAG"1. It has been shown that SPAG-1 is transcriptionally regulated and is only expressed during sporozoite development when an infected tick is feeding on its host (Williamson et al., 1989). The SPAG-1 gene provides an opportunity for investigation of stage-specific gene regulation in the sporozoite stage of Theileria, since this gene is under tight stage-specific transcriptional control and no factors involved in regulation are known. Studying the 5' untranslated region of SPAG-1 was thought to be likely to reveal the identity of one or more of these factors. The sequence comparison between the first 350 bases of the 5' untranslated region of the four SPAG-1 alleles did not reveal any motifs which might be involved in the transcriptional regulation of SPAG-1. The region analysed contains only about 60 bases upstream of the transcription initiation site since this has subsequently been mapped to position -278. Both SPAG-1 and p67 are under transcriptional control and are expressed during the same life cycle stage, therefore it can be hypothesised that homologous transcription factors are involved in the regulation of both. If homologous transcription factors are involved, then the DNA sequences they bind to must also be conserved in the 5' untranslated region of both genes. The comparison of the 5' sequences of the two genes led to the identification of two G/C rich, palindromic sequences. The sequences are GCTAGC and GAGCAC. Unfortunately these sequences have not been identified previously to be sites which are involved in protein-DNA interactions. Therefore one can only speculate that these sequences might be involved in the stage specific regulation of SPAG-1 and p67. Further experiments are needed to investigate the role of these sequence motifs in the regulation of both genes. It was thought to be likely that by identifying the site for transcriptional initiation in SPAG-1 one might gain more information about the process of stage specific transcriptional regulation during the sporozoite stage. The site of transcriptional initiation was mapped to 277 137 bases 5' of the ATG start codon and is in accordance with the transcription initiation site consensus sequence Py Py CA Py Py Py Py Py (Corden et al., 1980). The immediate 5' region to the initiation site does not bear any homology to other known promoter binding sites. A search for sequence motifs, which in other organisms have been shown to be sites recognized by transcription factors resulted in the identification of only one conserved TATA box between SPAG-1 and p67. As this TATA box is located between the transcription initiation site and the ATG start codon, it seems unlikely that even this motif is involved in the regulatory process. Other TATA boxes are however present in the p67 5' untranslated region which are absent in SPAG-1. Since the genome of Theileria is A/T rich it is not surprising that such sequences would occur frequently and because of this it is very unlikely that any of these are involved in stage specific gene regulation. Similarly there are 7 TATA boxes, two Pit-I and one Myb binding sites in SPAG-1 which are not conserved in p67. It therefore seems likely that none of these sites are of importance for the regulation of SPAG-1 unless p67 and SPAG-1 employ different transcription factors for the regulation of the two genes. This seems unlikely since these genes are closely related. This leaves only the two sequence motifs GCTAGC and GAGCTC as putative regulatory sequences. In Plasmodium, a sequence comparison of the 5' region of the CS gene and that of the GBP130 gene revealed that these sequences contain a region which is related to the SV40 enhancer region (Lanzer et al., 1990; Lanzer et al., 1992). When the 5' region of the SPAG-1 gene was searched for this sequence no region with significant homology could be found. It can be concluded that it is most likely that one or more novel transcription factors are involved in the regulation of SPAG-1 transcription. To identify which parts of the upstream region of the SPAG-1 gene are of importance in its regulation, one would ideally perform transfection studies with a range of deletions of the 5' untranslated region of SPAG-1 linked to a reporter gene. Unfortunately no transfection system is yet available for Theileria. Another possible way to identify binding regions is to use nuclear extracts from 138 sporozoites and conduct band shift assays to identify specific sequences to which these nuclear extracts bind. The limitation of these experiments is the supply of sporozoite material. To determine whether the two motifs, GCTAGC and GAGCTC, identified by sequence comparison of the 5' region of p67 and SPAG-1, are transcription factor binding sites, these could be used to screen an expression library to isolate sequence specific DNA binding proteins. The disadvantage of this approach is that if no proteins are isolated, no conclusion can be drawn from the experiments since the transcription factor may need processing, folding or polyadenylation to make it functional. Even if a protein which can bind to a certain motif has been isolated it does not provide evidence that it is actually a transcription factor. Ultimately, one has to rely on a transfection system to provide this evidence. Since stable transformation has been achieved in Plasmodium (Wu et al., 1995), it might not be much longer until stable transformation of Theileria will be achieved. 4.3.2. The SPAG-1 intron. In this thesis I provide three independent pieces of evidence for the existence of a single, 30bp long, cryptic intron in the SPAG-I sequence. The intron was mapped to the position 1616 to 1635 on the genomic DNA sequence gH3.4. The first indication for the existence of the intron is based on a sequence comparison of the published cDNA sequence of SPAG-1 to the gH3.4 sequence of SPAG-1. One of these experiments is the S1 nuclease mapping. The result of this indicates that the intron is not just only present in the gH3.4 allele but also in the H6.0 allele and if there is an H4.8 allele then also in this allele. The reason for this is based on the observation that there was no full-length band visible in Figure 25, therefore one can conclude that all SPAG-1 RNA species from this mixed parasite stock contain an intron at the predicted position. Analysis of the flanking regions of the only three introns so far identified in Theileria resulted in a consensus sequence which is in accordance of that for other eukaryotes but is still more elaborate than 139 that of Perlman et al. (1984). The consensus sequence is in accordance with the GT-AG rule for splice sites found in introns isolated in Plasmodium (Ravetch et al., 1984; Wesseling et al., 1989; Adams et al., 1992). Therefore it can be concluded that it is unlikely that novel factors are involved in intron splicing in the apicomplexan parasites Thelleria and Plasmodium. 140 Chapter 5 Binding studies with recombinant sporozoite antigens. 5.1 Introduction. In order to develop vaccines or treatments for tropical theileriosis which are sporozoite based, it is of great importance to understand what interactions occur between the sporozoite and host cells and how the parasite invades a host cell. Thus by studying the function of sporozoite antigens one might provide some information which could be used in the prevention of host cell invasion. 5.1.1. Are SPAG-1 and SPAG-2 ligands for host cell recognition/invasion? The first indication that SPAG-1 and SPAG-2 might play a role in the process of host cell recognition and invasion is based on the observation that sporozoite entry into host cells can be blocked In vitro if the parasite is incubated with either of the monoclonal antibodies 1A7 or 4B11 prior to invasion (Williamson, 1988). There are two possible explanations as to how the monoclonal antibodies could block the invasion of the sporozoite into the host cell: 1) the effect is purely steric in the sense that the sporozoite surface is covered by so many antibodies that it prevents its membrane binding closely to the host cell surface and thus inhibits the invasion process or 2) the antibody binds directly to the molecule involved in the binding process, either directly at the binding site or very close to it and thus blocks invasion. The observation that other monoclonal antibodies reacting with the middle region of SPAG-1 do not block the invasion process (Williamson, 1988) seems to indicate that SPAG-1 is directly involved in either the recognition or the invasion process. Bovine serum raised against the C-terminus of SPAG-1. 141 but which does not react with the 1A7 epitope. also inhibits sporozoite invasion of host cells (Boulter et al.. 1994). This observation indicates that, if SPAG-1 is involved in the process of host cell recognition or invasion, its C terminus is important. A similar experimental approach. using monoclonal antibodies to block invasion, has highlighted the importance of the MHC class I molecule on the host cell surface in the invasion of T. parva sporozoites (Shaw et al., 1991). Further support for the theory that SPAG-1 is a ligand for host cell recognition was based on the fact that there are three VGVAPG peptides in the predicted amino acid sequence of the SPAG-1 cH allele (Hall et al., 1992 and Hall, 1994), at least before the results of chapter 3 were obtained. The VGVAPG peptide is also found in bovine clastin (Raju and Anwar, 1987), in which it has been identified as the ligand which binds specifically to the elastin receptor (Blood et at., 1988; Mecham et al., 1989; Robert et al., 1989). This therefore led to the speculation that the sporozoite might recognise a host cell via the VGVAPG peptide in the SPAG-1 molecule (Hall et al., 1992). Interestingly, the elastin receptor has been found on macrophages (Huard et al., 1986) and monocytes (Senior et al., 1984; Blood et at., 1988; Mecham et at., 1989) which are the main target cells for T. annulata sporozoites (Glass et al., 1989 and Spooner et al., 1989), and the receptor is also found on fibroblasts (Ilinek et al., 1988 and Wrenn et al., 1988) which can also be infected by T. annulata (Brown and Gray, 1981 and Morrison et at., 1986). However the suggested role of the VGVAPG motif was largely invalidated once the gH3.4 sequence was obtained and found not to contain it (Katzer et al., 1994; chapter 3). SPAG-1 is stage-specifically regulated (Williamson ct al., 1989). Its mRNA is only transcribed during the early stages of sporozoite development and transcription stops once the sporozoite is mature. In the mature sporozoite the processed SPAG-1 is found on the surface (Williamson et al., 1989 and Knight, 1993). Unfortunately nothing is known about the fate of SPAG-1 either during the entry process or once the entry process has been completed. The T. parva homologue, p67, has been shown to be shed from the sporozoite surface during the invasion process (Dobbelaere et al., 1985). But one can speculate that a gene which is under such tight regulation and is expressed on the surface of 142 the sporozoite might function in host cell recognition or invasion. Little is known about SPAG-2; as stated above, it has only been partially sequenced and one can only speculate about its function. But since the monoclonal antibody 4B11, which reacts with SPAG-2, blocks the invasion of sporozoites into host cells, this antigen might also be involved in the process of host cell recognition and/or invasion. To investigate whether either SPAG-1 or SPAG-2 are involved in the recognition of host cells one has to ascertain if either molecule binds specifically to cells which can be invaded by T. annulata sporozoites but not to those which are not a target for invasion. In order to study whether SPAG-1 or SPAG-2 are involved in host cell recognition or invasion, recombinant proteins have been expressed and employed in binding studies. 5.2 Results 5.2.1. Cloning and sequence data for SPAG"1 and the SPAG-2 constructs. To study the involvement of the SPAG molecules in the process of host cell recognition and invasion, recombinant protein was expressed in bacteria using the pGEX expression system (Smith and Johnson. 1988). The advantage of this system is that large quantities of protein can be expressed and easily purified. A further advantage for these particular studies was that the two genes coding for SPAG"1 and SPAG-2 were already cloned into this vector system for a different purpose. The pGEX-2.7 plasmid clone was kindly given to me by Dr. Hall. This plasmid contains a 2667 bp insert derived from the SPAG-1 gent. pGEX-2.7 was cloned by introducing an artificial Eco RV site. via site directed mutagenesis, into the cDNA sequence of the SPAG"1 gene. The gene was then cut with Eco RV and Eco RI and cloned into a Sma I- Eco RI cut pGEX-3X vector (Boulter et al.. 1994). The resulting gene product 143 lacks the first 19 amino acids of SPAG-1 and the twentieth has been mutated from lysine to isoleucine but is otherwise the complete protein. The first 18 amino acids which are not encoded by the pGEX-2.7 construct are predicted to be a signal peptide (Hall et al.. 1992). The junctions of this construct have been sequenced in order to confirm the validity of the construct (Boulter et al., 1994). Four plasmids containing parts of the SPAG-2 gene were given to me by Dr. Knight. These plasmid constructs are presented diagrammatically in Figure 27. The first plasmid contains the KP8 insert. which is the original SPAG-2 fragment (Knight et al.. 1993), cloned into pGEX-IXT. KP8 is not the full length gene for SPAG-2 since it consists only of 980 bp while the whole gene product is predicted to be 150 kDa. The second plasmid contains the Hinc II fragment of SPAG-2. called KP27 and is cloned into Bluescript SKf. This fragment maps to the middle of KP8 and is 515 bp long but due to the cloning procedure has lost its flanking restriction sites. The third plasmid contains the Eco RI - Xba I fragment (B14) of KP8 and is cloned into pGEM SZff. This fragment consists of the first 900 bp of KP8. Finally the last plasmid given to me contains the last 260 bp of KP8 (A9), and was made from a deletion of KP8. and is cloned into pGEM 5Zf*. At the time when KP8 was given to me, it had not been fully sequenced and only the first 252 bp were available (Knight. 1993). As a priority I thus decided to sequence completely KP8 using double stranded DNA from the four plasmids as templates. The sequence information obtained from these plasmids is shown diagrammatically in Figure 28. Apart from 95 bases within the Hinc lI fragment the KP8 5' to 3' DNA strand was sequenced fully. Four hundred and seventy-eight bases were obtained from the 3' to 5' DNA direction. These overlapped with the sequence from the 5' to 3' direction. The KP8 sequence I obtained was compared to the KP8 sequence obtained subsequently on both strands by Dr. Knight at the University of Glasgow. The finalised DNA sequence based upon both sets of data and its predicted amino acid sequence is shown in Figure 29. At the same time as the sequence information of KP8 was obtained the junctions of the KP8 insert in pGEX"17XT were sequenced. The sequence information of the junctions of the KP8 Insert 144 pGEX-KP8 I P"'Y82 Y83 "Oll 'f, R Hi c il Y82 Yl H cI XýI E RI Biuescript-KP27 pGEM-B14 pGEM-A9 lost, site lost, isrte Eýq RI X ý1 lost ºtý e___Ecp R1 Bam HI pGEX-C350 C 19 RI Figure 27: Diagrammatic representation of the SPAG"2 constructs. The name and vector for each of the constructs is shown at the left. KP-8 in pGEX is the largest isolated fragment of the SPAG-2 gene. Some of its restriction sites are marked. The other SPAG-2 constructs are shown according to their position, and their flanking restriction sites are shown. 'Lost site' Indicates that these restriction sites were lost during the cloning event. The arrows, marked either Y82 or Y83, indicate the location of the primers used to PCR amplify the DNA fragment which was used to clone pGEX-C350. 145 pGEX-KP8 E RI HIC 11 H8 IiE W Sequence from KP8 Sequence from KP27 Sequence from KP27 Sequence 5' to 3' Sequence 3' to 5' -Sequence fron 814 -00- Sequence Sequenuf fort' KP8 from A9 Figure 28: Diagrammatic representation of the SPAG"2 sequence data. KP-8 represents the largest isolated fragment of SPAG"2. The thin arto«s beneath indicate the length and orientation of the sequence information obtained from the construct mentioned to its left. The dark arrows summarise the total sequence information obtained for each orientation. 146 1 OAATTCOOCTCOAOAOATAOCOCCAAOAAAACAAOTOACCATOAOACAAAAOAAAOTAAA ZYOSRDSAMITSDUSTIN29 61 CACCATACACAAAOACAOTACAAATACAATAACAAAOATCATAATTCCAAAOATTACOAA DHRZRZYXTNNZ DONS 9DT! 121 TGCATCGACTCTGAAGCAATCAA000AGTAGTGGAAAAGGCAGTTATAGAAGCATTTG&C CZD8tAI IA VV= RA V2 sA 70 181 AAOT000TOTCAOAAAAAATTAA000? OAAOAAACA? OTC? AACAOACTCAO? AAACCOA ZCL2ZIZ1022TCL? D"r91 241 GAOTACACTTTTGATAAAAATOACACAACACTAAOCTATCAOOAOCAAAOTCAACTTTAT 8YTTDIYDTTL2T01Q8QLT 301 T000OTOATTATCTCCAAAOACTCACAAA? OACCAAAATAACCAOOOAACTTCOOATATO SRD7L0RL? X0QXX00? 1DM 361 AA000ATATOOAAAAAATOATTTTAAAOAAAATAATOAAAATOATACCCA? OO? C? AAAA NRY0ENDtx! RNIKDTROLR 421 CCATTTOAAAAAOATTTCTCAAACOTOOAOAATCOAOAATATTCTAAAOATAAAACAOTT PFZKDFSWVXNRZTS* DRY? 491 QAAAGCAATAACTTTOAOTTTAAA000AAOATAATTTCAAATTCAACAOCATACOCAAAC ESXNFZYKOKXEaTATAN 541 ACACAAGAAAA000ATATGTAACTGATCATATOAOCCAOAA? OTOTATOAOOAOAT?? C4 T0ENPYVTDöx80: VYtt18 601 TATOaOAATaAAaCTOATOAATTTOATAOTOAATCOTTTAATTTTCAATTCAACAA0000 YZNZTDZTD2Z2YN! 0tX10 661 A000AATTTOAATACAATTTTOATOOAOTTOATTTOTTTTCAAATAOAOACCAAACAOAT 80! 2YNTD0VDLTSNAD0TD 721 AOTTTATTCTA000AOATOAAOAACAATTCAATAOTTZOAATAATAATO? AOC? O? AAOA 8LTYPDZZQtxsLXXXVAVA 781 CATCACTTCCCCCTCCCACAAATCTATCATB, ACCAAC8. AC8, AQATAACTTCl000AACAQ DD7AV0ZXTDKZQQDXlSZ0 841 TTGACCAACCAAGATOATAAATCOOAACTCTTATOCTCACAAACOTTTOAAOOAAATOAT LTN0DDK2sLLC2QTT=O 81 0 901 CTAOATOATAAAAATTTTOATCAAACATACACACTTOQAAACTOT? TCOCCAAT? TCAC? LDDAutDQTYTLOxCTAWFT 961 CGAGCCGCTCGAGCCG7UITTC 961 RAARAE7 x Figure 29: DNA sequence of SPAG. 2 (KP8). The predicted amino acid sequence is shown below. The sequence shown is based on the compuisan of sequence information obtained by Dr. Knight and F Kat: er. 147 in the pGEX vector confirmed that the predicted open reading frame is intact. Furthermore, by sequencing pGEX"KP8 using double stranded DA the validity of the construct was confirmed. Finally, another SPAG-2 construct was created for expression In the pGEX system using a PCR approach. This construct Is called pGEX"C330 and was created to determine which part of SPAG"2 harbours the ligand responsible for binding to peripheral blood mononuclear cells. The locations of the primers used to PCR amplify bases 10 to 361 of the Kill sequence are shown in Figure 27 and the sequence of the primers is shown in Table 7 in the Material and Methods section. Bases 10 to 361 of KP8 were amplified by PCR, cloned into pGEM"T. checked by sequencing. cut out of pGEM-T via Bam HI and Eco RI digestion (these restriction silts were built into the primers) and cloned Into pGEX"2T. The junctions of the resulting expression plasmid pGEX"C350 were sequenced In order to confirm that the plasmid contains the correct Insert and that the predicted open reading frame is intact. 5.2.2. Expression of SPAG"1 and the SPAG"2 constructs. The three pGEX expression plasmids pGEX-2.7. pGEX"KPS and pGE: X" C350 were expressed in Escherichia coll. producing three GST fusion proteins, GST-SPAG"1. GST-KP8 and GST-C350 respectively. The fusion proteins were purified on glutathione 4B sepharosc columns and %crc analysed by electrophoresis on SDS polyacrylamide gels. Gels showing the fusion proteins are shown in Figure 30. The GST"SPAG"1 fusion protein has an apparent molecular mass of 145 kDa although the expected size is only 112.8 kDa (Boulter et al.. 1994). The GST"S1'AG"2 protein expressed by pGEX-KP8 has a molecular mass of 58 kDa and the product of pGEX-C350 consisting of GST and the first 116 amino acids of the known SPAG-2, has a molecular mass of 41 kDa. It can be seen that all three constructs produce one major protein product. as visualised by SDS PAGE, at a size slightly larger than expected and they also contain smaller protein bands which may be degradation products. In all lanes there is another protein band visible at about 70 kDa which I beliere Is due to bacterial contaminantion. 14$ 145 KDa ýý .M mosommom 58 KDa 41 KDa 26 KDa Figure 30: Coomassie stained SDS-PAGE gel of the GST-SPAG proteins. SPAG-1 and the two SPAG-2 fragments were expressed using the pGEX expression system and the purified proteins are shown. GST in lane 1, GST- SPAG-1 in lane 2, GST-KP8 in lane 3 and the GST-C350 in lane 4. The bars mark the size of the fusion proteins in kDa. 149 The validity of the protein construct GST-SPAG-1 was confirmed by Western blotting. A nitrocellulose membrane to which the GST-SPAG- 1 fusion protein was transferred from SDS PAGE gels was probed with the monoclonal antibodies 1A7 (Williamson, 1988), BA4 (Wrenn et al.. 1986) and as a negative control with 1E11 (anti Thellerla annulata macroschizont antibody (Shiels et al., 1986)). 1A7 reacts with SPAG-1 and therefore is a positive control. BA4 is an anti-clastin antibody which has been shown to react with the peptide VGVAPG; as this peptide is found three times in the predicted SPAG-1 amino acid sequence, BA4 Is another positive control for the GST-SPAG-1 fusion protein. The Western blots of GST-SPAG-1 probed with 1A7, BA4 and 1E11 (negative control) are shown in Figure 31.1A7 and BA4 both react with GST-SPAG-1 as predicted. and 1E11 does not react with GST-SPAG-1; it can therefore be concluded the pGEX-2.7 vector expresses the correct protein. These results are in accordance with those of Boulter et al. (1994). Of note is that 1A7 reacts with some of the smaller SPAG-1 products. This is surprising. since it would expected that if these products are due to degradation this would occur from the C terminus as the GST moiety must be present because of the purification procedure. One possible explanation is that SPAG-1 oligomerizes through its C terminal region and that N terminally deleted GST fusion proteins co-purify via association with intact molecules (Boulter et al., 1994). Another possible explanation for some of the observed fragments, although extremely unlikely is that SPAG-1 is internally processed, so that N terminal and central regions of this molecule are deleted, leaving the GST moiety and the C terminal region of SPAG-1. The validity of the pGEX-C350 construct was confirmed by sequencing the whole clone and its junctions. The sequence of the 5' junction of the pGEX-C350 clone is shown in Figure 32a and matches that obtained by Knight (1993). The expressed fusion protein has a molecular weight of 41 kDa as would be expected from the DNA sequence. The sequence identity of pGEX-KP8 was also confirmed by sequencing the 5' junction, as shown in Figure 32b. This sequence matches that of Knight (1993 and personal communication) and it has been shown that the pGEX-KP8 fusion protein product reacts with 4ßl1 (Knight, 1993). Therefore I believe that both pGEX-KP8 and pGEX-C350 express the correct fusion proteins as predicted. 150 A) 12 B ) C) 12 12 ý 145 KDa Figure 31: Western blot of GST-SPAG-1. GST (lane 1) and GST-SPAG-1 (lane 2) were run onto PAGE gels, transferred onto a nitrocellulose blots and probed with monoclonal antibodies 1A7 in (a). BA4 in (b) and with 1E11 as negative control in (c). The bar marks the size of the fusion protein in kDa. 151 a) pGEX-C350 CTG GTT CCG CGT OCA TCC TCC ACA CAT ACC CCC AAC AAA ffttttt! f!!!!!!! UP !!!!! t!!!!!!! LVPRO8SRDSAKK b) pGEX-KP8 CTG CTT 000 CCT CCA TCC CCC GAA TTC CCC TCC ACJI LVPRCSP11CSR Figure 32: Sequence data of the junctions of the SPAG"2 clones in the pGEX vector system. The top sequences represent the DNA sequence of the vectors and inserts. The sequences below are the predicted amino acid sequences. The sequence of the vectors is underlined and the bold sequences shows the restriction sites used in cloning these fragments. (a) The 5' junction of the C350 insert in pGEX-2T. validating the expected open reading frame. The double underlined sequence represents the sequence of the primer Y82. (b) The 5' junction of the KP8 insert in pGEX- la. T, validating the expected open reading frame. 5.2.3. Cleavage of SPAG-1 and the SPAG"2 constructs. The recombinant expressed fusion proteins were cleaved from their fusion partner, GST, for further studies. The protein samples needed to be as free from GST as possible, so that the results obtained in subsequent studies could be attributed fully to the parasite protein rather than leaving doubts about the involvement of GST. The method for cleaving the three fusion proteins was adapted from Knight (1993) and is described in 2.2.6. To cleave GST"SPAG"l. activated factor Xa was used while thrombin was used to cleave the SPAG. 2 fusion protein constructs. The method for separating the cleaved protein from GST and any remaining uncleaved protein was improved from that of Smith and Johnson (1988) so that the cleaved protein is run through a glutathione 4B sepharose column three times. The Figure 33 shows a coomassie stained SDS"PAGE gel. showing the GST"KP8 fusion protein in lane 2. In lane 3 is the cleaved protein with cleaved fusion 152 123456 58 ýý ,,,. - - WAN. ý` ý 36 26 Figure 33: Coomassie stained SDS-PAGE gels of the cleaved SPAG-2 during different stages of purification. The GST control is shown in lane 1, the GST-SPAG-2 fusion protein (lane 2), the cleavage product (lane 3), cleaved SPAG-2 after the first purification step (lane 4), cleaved SPAG-2 after the second purification step (lane 5), cleaved SPAG-2 after the third purification step (lane 6). 153 partner, lane 4 shows the protein after the first purification step. I=t 3 shows the protein after the second purification step and finally laze 6 shows the protein after the third purification step. From this it is ckat that there is still some cleaved GST present after the first purificttioz step and although none is visible after the second purification step a third purification step was used to be absolutely sure that there is no GST or uncleaved protein present in the final sample. All three protciA constructs were purified with this adapted method after cleavage. The purified cleaved SPAG-1. KP8 and C350 were run out onto SDS PAGE gsf t next to their uncleaved partners to check the purity and the Valid$ty of the constructs. A SDS-PAGE gel containing the three constructs Math cleaved and uncleaved is shown in Figure 34. 5.2.4. Do SPAG-1 and SPAG-2 bind to host tells? To test whether SPAG-1 and SPAG-2 are involved in the proccss of host cell recognition two different types of binding studies %crc conducted to ascertain whether either molecule binds specifically to CCII types which can be infected by T. annulara sporozoites. The first binding assay involved the binding of biotinylated protein to purified borinc peripheral blood mononuclear cells. The degree of binding was determined by flow cytometry. First, the three purified and cleaved proteins were labelled with biotin as described in 2.2.7. Successful biotinylation was confirmed by Western blotting. To do this, the proteins were run onto SOS-PAGE gels. transferred to a nitrocellulose membrane and probed with Catraridant (Sigma), which binds to the biotin of the blotinylated protein. Extravidine is visualised using the alkaline phosphata, c detection method as described in 2.2.7. A Western blot confirming the svrte, tful biotinylation of the three proteins can be seen in Figure 35. Once it laaj been shown that all three molecules were biotinylated. they Could then be used in the binding studies. Peripheral blood mononuclear cell, were purified from the blood of Friesian calves. Aliquots of Sx lOS of it. tst cells were incubated with a dilution series of either of the three molecules or biotinylated GST which was used as a negative control. Arta 30 minutes incubation and three washes, a streptavidin phycoer)thtia 154 145 kDa ý119 kDa ýý ,ý ýý ý 58 kDa 41kDa 36 kDa 19 kDa 1 23 Figure 34: Coomassie stained SDS-PAGE gels of the cleaved SPAG proteins. GST-SPAG-1 is shown in lane I and the cleaved SPAG-1 in lane 2, GST-KP8 in lane 3 and the cleaved KP8 in lane 4 while GST-C350 is shown in lane 5 and the cleaved purified C350 is shown in lane 6. The bars mark the size of the proteins in kDa. 45 6 155 a) b) 12 um ý 119 ý 36 .  . 19 Figure 35: Western blot of biotinylated protein detected with Extravidine. The labelling of the protein with biotin was checked by western blotting using Extravidine as a detection agent. (a) Biotinylated cleaved SPAG-1 is shown and the size of the largest fragment is 119 kDa. (b) Biotinylated cleaved KP8 is shown in lane 1 and the biotinylated cleaved C350 is shown in lane 2. The bars mark the size of the proteins in kDa. 156 conjugate was bound to the biotinylated protein and was detected using a Facscan flow cytometer. The results for the binding of the SPAG-1 dilution series is shown in Figure 36a, that of the GST dilution series in Figure 36b, the KP8 dilution series in Figure 36c, and the dilution series of C350 is shown in Figure 36d. From these results it can be deduced that all three sporozoite molecules bind to a sub-population of the bovine PBM cells since for these three proteins there is a two-peak histogram present in which the peak nearest to the Y-axis represents the cell population which is negative for protein binding and the peak further away from the Y-axis represents the cells which are positive for protein binding. The binding of the GST dilution series is different in the sense that no clear two-peak histogram could be observed. Furthermore, there is almost no binding observed at a protein concentration of 250 ng and 125 ng, although a clear positive and negative cell population could still be seen for the three SPAG protein constructs at these concentrations. It can therefore be concluded that GST does not bind specifically to a sub- population of peripheral blood mononuclear cells. This experiment using the SPAG constructs indicates that all three molecules bind specifically to some of the PBM cells but not to all of them. The results of the dilution series for all three recombinant protein preparations indicated that a working concentration of 500 ng of protein results in a good separation of positive and negative cells. This working dilution was therefore adopted for further two-colour flow cytometry. 157 Figure 36: Single colour flow cytometry using a dilution series of blotinylated proteins and PBM cells. A method to establish the best protein concentration for two colour flow cytometry. (A) The dilution series of biotinylated cleaved SPAG"1 on PBM cells. Reagent control. cells incubated without protein or primary antibody (panel 1), positive control, anti MHC class II antibody (panel 2), 1.25 µg SPAG"1 (panel 3), 500 ng SPAG"1 (panel 4), 250 ng SPAG"1 (panel 5). 125 ng SPAG-1 (panel 6). 25 ng SPAG"1 (panel 7) and 12.5 ng SPAG"1 (panel 8). (B) The dilution series of biotinylated GST on PBM cells. Reagent control (panel 1). positive control, anti MHC class II antibody (panel 2), 1.25 µg GST (panel 3). 500 ng GST (panel 4), 250 ng GST (panel 5) and 125 ng GST (panel 6). (C) The dilution series for biotinylated cleaved KP8 on PBM cells. The reagent control (panel 1). positive control, anti MHC class II antibody (panel 2). 1.25 µg KP8 (panel 3), 500 ng KP8 (panel 4), 250 ng KP8 (panel 5). 125 ng KP8 (panel 6). 25 ng KP8 (panel 7) and 12.5 ng KP8 (panel 8). (D) The dilution series for biotinylated cleaved C350 on PBM cells. The reagent control (panel 1), positive control, anti MHC class II antibody (panel 2). 1.25 pg C350 (panel 3). 500 ng C350 (panel 4), 250 ng C350 (panel 5). 125 ng C350 (panel 6). 25 ng C350 (panel 7) and 12.5 ng C350 (panel 8). Number of cells 94 : . II ä ý Y L 7 I Reagent control (no protein) Number of cells 94 : . u 3 C ] ü 3 100 101 104 f04 104 1.25 ug SPAG-1 Number of cells too lot 1 02 to' 5 250 ng SPAG-1 Number of cells 7 Number of cells 2 e" _ P1 aý , RPQ ,, NPQ ,ý FP4 ý IPP4,111 020 I 0 u ä ý r C a S too Number of cells Number of cells 125 ng SPAG-1 4 500 ng SPAG-1 6 Igo sat 30= i02 Number of cells 94 8 : O u ä 1. Y L i ü 25 ng SPAG-1 12.5 ng SPAG-1 158 B) Number of cells I Number of cells 94 : II S I. r C S ü Is iSI iYZ sea 104 Reagent control (no protein) Number of cells 64 : I 3 ý Y c 7 I Number of cells 64 : a 7 V I M [ U ; os 10t tale tale 894 250 ng GST : . u s C ý IS Number of cells I 4 S r Y [ ] U U 4 ; lag 101 1 O= 1 O9 194 Number of cells : . {/ i L C i S 1 a2 ILA 21 ýýý - 92 500 ng GST 6 125 ng GST 159 C) Number of cells 1 Reagent control (no protein) Number of cells Number of cells F) 2*4 Number of cells 3 1.25 ug KP8 5 250 ng KP8 7 ýýl 25ma 25 ng KP8 Number of cells Number of cells Number of cells I 12* Number of cells :h 1154 2 CC108 4 500 ng KP8 6 125 ng KP8 8 12.5 ng KP8 160 D) Number bf cells m m1 tin 'iWj Reagent control (no protein) Number of cells Number of cells Number of cells ! gL 1'1 Do 3 1.25 ug C350 5 250 ng C350 7 7.59: ml 25 ng C350 1 Number of cells 2 ^I co Number of cells Number of cells Ch 1254 Number of cells ý) i1Nq CC108 4 500 ng C350 6 8 12.5 ng C350 161 5.2.5. Which are the target cells for SPAG-1 and SPAG-2 binding? The previous results do not indicate to which cell types the three sporozoite protein preparations bind. Two-colour flow cytometry has therefore been used to identify these sub-population of cells. Aliquots of bovine PBM cells were stained with a panel of monoclonal antibodies. These antibodies react with surface molecules of specific PBM subpopulations (such as B cells, T cells, and monocytes) as listed in Table 9. Each of the three biotinylated protein constructs was then incubated with these PBM cells. The binding of the monoclonal antibodies was visualised using the flow cytometer with green fluorescence while the protein binding was visualised by red fluorescence. The binding results for the three proteins is shown in Figures 37-39. Each of the squares represents one of the initial PBM aliquots. The X-axis reflects the binding of the monoclonal antibodies and the Y-axis represents the binding of the protein. The results for SPAG-1 binding are shown in Figure 37, the results for KP8 in Figure 38 and those for C350 are shown in Figure 39. Two-colour flow cytometry using biotinylated cleaved SPAG-1 and a panel of 6 monoclonal antibodies, as described in Table 9, shows that SPAG-1 strongly binds to a subpopulation of 50% of peripheral blood mononuclear cells as shown in Figure 37A. In the next panels, Figure 37B-F, it is shown that SPAG-1 binds to 76% of CD2 positive T cells, 44% of y/S T cells, 19% of B cells, 27% of monocytes and 27% of MHC class II positive cells. Name of antibody Target molecule Target cell type CC15 WCl /S T cells CC21 WC3 B cells CC42 CD2 T cells CC94 CD lib mainly Monocytes ILA 19 MHC class I all PBM cells ILA 21 MHC class II Monocytes and B cells Table 9: Monoclonal antibodies used in the 2 colour flow cytometry., The table lists the monoclonal antibodies used, the surface molecules they react with and the cell types on which those molecules are found. 162 Figure 37: Two colour flow cytometry of PBM cells using blotinylated cleaved SPAG"1 and a panel of monoclonal antibodies. 500 ng SPAG-1 were used for each of the results shown in panel A-F and of a series of monoclonal antibodies. Panel A shows the result for ILA-19, anti MHC class I; CC 42, anti CD2 (T cells) in (B); CC15, anti WCI (yib T cells) in (C); CC21, anti WC3 (B cells) in (D); CC94, anti CD11b (mainly monocytes) in (E) and ILA-21 anti MHC class U in M. SOOng SPAG-1 ILA-19 500ng SPAG-1 T M5 500ng SPAG-1 T CD si .. ý .. ý : ý:; ürl7ý". 1 '"ýýfý: " " ý. , +: n ý" -- .... CD11b is1r J J2 CC94 E s. ý5ý'ýý: M. "_. . ý"`_ t'. 500nq SPAG-1 7 Oý' °' . _r ;. t ý 7i. j, I . ý" -if-al 121 CC42 ,:. ý; -. .. ý. ?I QL. 7 ýI ý" iýý i>>ý SOOng SPAG-1 162...... 16' ' --- 11-4 ILA-21 163 CC21 Figure 38: Two colour flow cytometry of PBM cells using biotinylated cleaved KP8 and a panel of monoclonal antibodies. 500 ng KP8 were used for each reaction of panel C-H. Panel A shows the cell population investigated. The reagent control is shown in (B). A combination of KP8 and one of a series of monoclonal antibodies were used for panels C-H. ILA-19 anti MHC class I in (C); CC42. anti CD2 (T cells) in (D); CC15, anti WCI (yi6 T cells) in (E); CC21, anti WC3 (B cells) in (F); CC94, anti CD11b (mainly monocytes) in (G) and ILA-21 anti MHC class II in M. O Aý ý,. ýý Oh ZO ýý 6 A -20 40 60 666 10001 FSC-H%FSC-Height-), in O ýt. Cell population analysed 5ö B G' ö( ' 500 ng KP-8 MHC I E WC1 CD2 500 ng KP-8 H CD11b N0 ýO F WC3 lÖ0 101 102 103 104i FLI-MFLI-Height-3ý I "- -"--' MHC 11 164 Figure 39: Two colour flow cytometry of PBM cells using the blotinylated cleaved 3S0 and a panel of monoclonal antibodies. 500 ng of C350 were used for each reaction of panel C-H. Panel A shows the cell population investigated. The reagent control is shown in (B). A combination of C350 and one of a series of monoclonal antibodies were used for panels C-H. ILA-19 anti MHC class I in (C) CC42, anti CD2 (T cells) in (D); CC15, anti WC1 (y1S T cells) in (E); CC21, anti WC3 (B cells) in (F); CC94, anti CD11b (mainly monocytes) in (G) and ILA-21 anti MHC class II in M. A 2"'66-46'0' 6ÖÖ 8b0 1000 ^V, , a 500 ng C350 9I Reagent control (no protein, no antibody) 500 ng C350 r.. ý^i H ýö ti iýö D lbo --ib' ----102 . --ßo3 ----x, 04 FLl -UXFL1-Height-a C MHC I 500 ng C350 E WC1 500 ng C350 rö ý ^. ýý. I C.!! G 6°1. CC41 . M1 . FSC-IAFSC-Height-ii Cell population analysed . ': Y äý'.: ': ý .. r J 1i ij`' 20°I. =r S'ý"ý"i:.. ýf" ý 100 1 t:;, 102 103 104" FL1-IAFL1-Height-i- I 4 CD11b CD2 WC3 MHC II 165 From Figure 38 and Figure 39 it becomes apparent that KP8 and C350 behave in an identical manner, and that the binding results of the SPAG-2 constructs are different to those of SPAG-1. It can be seen in Figure 38 that KP8 binds to approximately 2.4% of CD2 positive T cells, 5% of Y/S T cells, 16% of B cells, 22% of monocytes, 10% of MHC class II positive cells and 7% of MHC class I positive cells. For C350 the figures are as follow: 5% of T cells, 5% of y/S T cells, 25% of B cells, 23% of monocytes, 12% of MHC class II positive cells and 8% of MHC class I positive cells. Interestingly the total number of cells to which both KP8 and C350 bind reduces dramatically if the cells were incubated with CC21, CC94, ILA-19 or with ILA-21 before the cells were incubated with the recombinant proteins. This might indicate that these cell populations contain the target cells for SPAG-2 binding but that competition with these monoclonal antibodies prevents the SPAG-2 molecule from binding. In summary it can be concluded that both SPAG-2 constructs behave in a very similar manner to each other and that the main targets for SPAG-2 binding are B cells and monocytes, while only a very small number of T cells were bound by SPAG-2, which is of considerable interest since B cells and monocytes are the target for T. annulata sporozoite invasion. SPAG-1, on the other hand, binds a large proportion of T cells, which are very unlikely to be the target for sporozoite invasion. SPAG-1 also binds to B cells and monocytes. 5.2.6. SPAG-1 binds to BL3 cells, but SPAG-2 does not. It would be of interest to find out whether either of the SPAG molecules bind specifically to a cell line which is known to be susceptible to invasion by T. annulata sporozoites, and BL3 is such a cell line. Flow cytometry has therefore been employed to test whether recombinant SPAG-1 and KP8 bind specifically to these cells. The concentration of the recombinant proteins was chosen to be the same as for the two-colour flow cytometry in the above experiment (500 ng). The result for SPAG-1 binding is shown in Figure 40. Here it can be seen clearly that, at a concentration of SPAG-1 which revealed specific binding to a subpopulation of PBM cells, almost all BL3 cells are positive 166 A) Number of BL3 cells C4 U ý A 0 u .. 3 1. U I C 3 ý 11213 1123 112-2 3122 1124 SPAG-1 binding B) Number of BL3 cells C4   A 0 u 3 4 q v c 3 U -FiIIII Aill- Yi -I iI III I141,, Iq iD0 1 D= iDR 1D2 1 D4 Reagent control (no protein) Figure 40: One colour flow cytometry of BL3 cells using biotinylated cleaved SPAG-1. The combination of 500 ng cleaved SPAG-1 and BL3 cells are shown in (a) and the reagent control is shown in (b). 167 for SPAG-1 binding. This is clear evidence that recombinant SPAG-1 binds specifically to cells which can be invaded by T. annulata sporozoites. This observation indicates that the SPAG-1 molecule might be of importance in recognising BL3 cells as host cells. On the other hand KP8 did not bind to BL3 cells at this concentration. This indicates that SPAG-2 or at least the KP8 part of SPAG-2 has no importance in recognising BL3 cells as host cells. 5.2.7. BL3 cells do not express the elastin receptor. It has been proposed that the elastin receptor might be one of the target receptors on the surface of host cells to which SPAG-1 binds. Through this binding the host cell would be specifically recognised and the invasion process of the T. annulata sporozoite be initiated. It is therefore of interest to find out whether the BL3 cell line, which can be invaded by T. annulata sporozoites, express the elastin receptor. If these cells expresses the elastin receptor it would support the theory that both SPAG-1 and the elastin receptor are involved in the recognition or invasion process of host cells. On the other hand, if the BL3 cells do not express the elastin receptor it will show- that the clastin receptor is probably not of importance for the invasion of BL3 cells, and that SPAG- 1 binds to these cells via a different receptor. Single-colour flow cytometry has been employed to study whether BCZ, an anti-elastin receptor antibody (Mecham et al., 1989), binds to BL3 cells. The antibody concentration was chosen to be the same as for antibodies used in the two-colour flow cytometry and the negative control was TRT1, a mouse monoclonal antibody, which reacts with a surface molecule of a turkey virus (Howard, personal communication). The flow cytometry result is shown in Figure 41. It can clearly be seen that neither of the two monoclonal antibodies bind to BL3 cells. It can therefore be concluded that BL3 cells do not express the elastin receptor. 168 B) Number of BL3 cells 94 I I a u ý 4 a v [ .] 8 BCZ binding A) Number of BL3 cells sa _19ý, ýý12P4, , FQ4, I r-Q4, I FV4I I a ä 0 U 4 a y [ ] 8 ý-__ _ I I I i I i i I f I TRT1 binding Figure 41: One colour flow cytometry of BL3 cells using the anti elastin receptor antibody BCZ. The combination BCZ and BL3 cells is shown in (a) and a negative control using the TRTI monoclonal antibody is shown in (b). 169 5.2.8. How many SPAG-1 receptors are there on BL3 cells? Once it had been established that SPAG-1 binds specifically to BL3 cells, it led to the question of how many receptors for SPAG-1 binding are expressed on the surface of each BL3 cell. In order to answer this question, a different binding assay was used. SPAG-1 and KP8 were labelled with 1251. An autoradiograph of an SDS-PAGE gel containing both labelled SPAG-1 and KP8 is shown in Figure 42, confirming that the proteins were labelled successfully. The labelled protein was incubated with an aliquot of BL3 cells for 30 minutes at room temperature. A control reaction was performed with a 50-fold excess of unlabelled protein included in addition to the 1251-labelled protein. Each sample point was done in triplicate in order to reduce the potential for error. After incubation the unbound protein was washed off and a count of the gamma radiation emitted from the cells was determined. The samples with the 50-fold excess of unlabelled protein give a value for non- specific binding and if this is subtracted from the total binding observed for each given protein concentration it gives the level of specific binding. The results for SPAG-1 binding are shown in Figure 43. Figure 43a shows three plots, where the the top graph represents the total counts read for each protein concentration, as shown on the X-axis. The bottom graph reflects the non-specific binding and the solid line (centre) shows the specific binding which was calculated by subtracting the values used to plot the bottom graph from those of the top graph. In the graphs in Figure 43b the counts were converted into ng of SPAG-1 bound to the BL3 cells and again the dark graph in the middle reflects the specific binding of SPAG-1 to the BL3 cells. Finally the graph representing the specific binding in ng, from Figure 43b, was converted into molecules of SPAG-1 bound specifically to each BL3 cell in each reaction. From the observed plateau in the graph of Figure 43c, one can estimate that 70,000 to 75,000 SPAG-1 molecules bind to each BL3 cell. Assuming that only one SPAG-1 molecule binds to each receptor, there are 70,000 to 75,000 SPAG-1 receptors on each BL3 cell. 170 1 2 119kDa II 36 kDa Figure 42: Autoradiograph of 125I labelled SPAG-1 and KP8 on a SDS PAGE gel. 1251 labelled cleaved SPAG-1 is shown in (A), 1251 labelled cleaved KP8 is shown in (B) and the bars mark the size of the proteins in kDa. 171 a)E 8 400000 300000 200000 100000'1 j, Aý. "`... "«r.. ý l" f ----...... p . -ý . 0 Rý ng of SPAG-1 O b) ng of SPAG-1 added -C - specific count ...... ý0 . ý. '.. total count --'-0---- non-specific count Si ý- ngbound specifically ng bound in total ""0 " ng bond non-specifically D- Number of SPAG-I molect bound per cell Y CCO NC ng of SPAG-1 added Figure 43: Binding curves of SPAG-1 to BL3 cells. The binding curves of SPAG"1 to BL3 cells with the observed counts are shown in (a). The X axis reflects the amount of labelled protein added and the Y-axis shows the observed counts of labelled protein bound to BL3 cells for each point. The top curve represents the total counts for each point, the bottom curve shows non-specific binding. The curve in the middle represents specific binding of SPAG-1 to BL3 cells and is the difference between total count and non- specific counts. (b) The same data as in (a) but the observed counts are converted into ng of protein bound to BL3 cells. (c) The same data as shown in (a and b) but the ng of protein bound were converted into SPAG-1 molecules bound specifically to each BL3 cell. 172 When the binding assay was repeated using 1251 labelled KP8, no specific binding by this molecule could be detected. The readings obtained were almost as low as the experimental background level and the same readings were obtained with a 50 fold excess of unlabelled KP8. These results are not shown and I conclude that 1251 labelled recombinant KP8 does not bind to BL3 cells. This result is in accordance with the binding data of biotinylated KP8 and BL3 cell binding using flow cytometry. 5.3. Discussion 5.3.1. Confirmation of the validity of the protein constructs. Ideally experiments would have been conducted using live sporozoites of T. annulata to study the process of host cell recognition and invasion, but the laboratory in York has no license to keep the sporozoite stage and only a very limited supply of sporozoites is available at the CTVM in Edinburgh. Recombinant proteins were therefore the basis for my investigations of the role of SPAG-1 and SPAG-2 in the process of host cell recognition and invasion by sporozoites. Recombinant proteins, expressed in a prokaryotic expression system, present an obvious limitation since the proteins may not be folded correctly, nor are they processed or glycosylated. Therefore the molecules are only likely to be successful in demonstrating binding to host cells if linear regions are responsible for the binding process. The work presented here demonstrates specific binding of the recombinant proteins showing that linear binding regions may indeed be responsible for the ligand receptor interaction. This has previously been shown to be the case for the CSP-1 protein of Plasmodium falciparum in which the recombinant protein, and subsequently peptides derived from it, bind specifically to host cells (Cerami et al., 1992). 173 5.3.2. Are SPAG-1 and SPAG-2 involved in host cell recognition and invasion? Binding assays using recombinant SPAG-1 revealed that SPAG-1 binds specifically to a large population of T cells, but more interestingly, to a sub-population of monocytes and B cells. Monocytes and B cells are the main target cells for the sporozoites of T. annulata (Spooner et al., 1989 and Glass et al., 1989). It has been argued that T cells are not invaded by T. annulata sporozoites, but it has been shown that they can be invaded at a low frequency in vitro (Spooner et al., 1989). Another observation was that T cells transformed by T. annulata sporozoites lose their T cell surface markers rapidly (Inner et al., 1989). It is therefore difficult to assess which cells are infected and transformed in vivo by T. annulata sporozoites. Some evidence has been found suggesting that T cells might be a target for T. annulata sporozoites in vivo since an in vivo-derived transformed cell line was obtained which expressed three different T cell markers (Howard et al., 1993). Although these results give no clear evidence that SPAG-1 binds to cells which can be invaded by T. annulata sporozoites, there is some correlation of SPAG-1 binding to definite target such as B cells and monocytes. However, SPAG-1 also binds to T cells and one can only speculate whether these are the target cells for the invasion of T. annulata sporozoites in vivo. The observation that SPAG-1 binds to some cells that are target for invasion and the fact that it is expressed only on the surface of the sporozoite stage, in addition to the blocking data, make SPAG-1 a serious candidate for involvement in the host cell invasion process. Recombinant KP8 binds to a low proportion of cells, some monocytes and B cells, which have been shown to be the main target for T. annulata sporozoite invasion (Spooner et al., 1989 and Glass et al., 1989) and only to a small number of T cells. The role of T cells as putative targets is still questionable. This demonstration of selective binding to the main target cell types for T. annulata sporozoite invasion gives support to the theory that SPAG-2 is also involved in the process of host cell recognition and invasion. SPAG-2, however, binds to such low numbers of target cells that it is unlikely that this antigen is soley responsible for the recognition of target cells. Another very interesting finding is that C350 reacts in an almost identical manner to the 174 recombinant KP8. This indicates that the region which is responsible for SPAG-2 binding must be located within the first 116 amino acids of SPAG- 2, as only this region of the protein is common to the two constructs. The results of the binding assay using SPAG-1 and BL3 cells are of particular interest. These give direct evidence that SPAG-1 binds specifically to a cell line which can be invaded by T. annulata sporozoites. This further supports the theory that SPAG-1 is involved in the process of host cell recognition and invasion. The observation that KP8 does not bind to BL3 cells may indicate that the sporozoite might express several different ligands which are involved in the invasion process of different cells. This result does not rule out SPAG-2 as a ligand for host cell recognition and invasion, especially since the binding results obtained by KP8 on peripheral blood mononuclear cells indicate that SPAG-2 binding shows a closer correlation to target cells than that of SPAG-1. Taking into account the blocking data and binding data for both molecules it seems likely that neither of the SPAG molecules alone is sufficient for the sporozoite to identify a host cell. It seems far more likely that both molecules play a part in the recognition and invasion process, but that there are probably other molecules on the surface of the sporozoite which are also involved. Therefore I would propose that a combination of surface molecules of the sporozoite have to bind to a putative host cell in order to initiate the invasion process. 5.3.3. Is the elastin receptor involved in the process of host cell recognition? The finding that SPAG-1 contains three VGVAPG motifs (Hall et at., 1992) gave rise to the theory that SPAG-1 might bind to the elastin receptor on host cells to initiate invasion by the parasite. Bovine elastin contains VGVAPG in its amino acid sequence and the elastin receptor binds to this peptide. My work provided no further evidence that the elastin receptor is involved in sporozoite invasion. In contrast, evidence was found to suggest that the elastin receptor is very unlikely to be involved in the invasion process. The first indication that the elastin receptor and the VGVAPG molecule of SPAG-1 are not involved in host 175 cell recognition is based on the finding that BL3 cells which can be infected by T. annulata sporozoites do not express the elastin receptor but recombinant SPAG-1 still binds to them. The binding mechanism must therefore involve a different receptor on the surface of the BL3 cell. This, of course, is not conclusive evidence that the elastin receptor is not involved; it might indicate that sporozoites utilise different receptors on the surface of different cell types for invasion. Further evidence suggesting that the elastin receptor is not involved in host cell invasion is based on an observation that PBM cells sorted into a population which is negative for the expression of the elastin receptor could be infected by T. annulata sporozoites equally well as PBM cells which express the receptor (Campbell et al., 1994). More evidence that the elastin receptor is unimportant for the invasion of host cell is derived from infection studies in which the monoclonal antibodies BCZ (anti-elastin receptor (Mecham et al., 1989)) and BA4 (anti VGVAPG (Wrenn et al., 1986)), as well as an excess of elastin peptides, failed to block invasion of sporozoites into host cells in vitro (Brown, personal communication). Finally, in chapter 3. I have shown that one of the SPAG-1 alleles does not contain the elastin receptor ligand VGVAPG. I therefore conclude that the elastin receptor is not involved in SPAG-1 binding. It is far more likely that the purpose of the elastin repeats, including VGVAPG, found in SPAG-1 is centred around the evasion of the host immune response (Hall, 1994). 5.3.4. Future work and unanswered questions. Although I have been able to provide evidence to support the theory that SPAG-1 and SPAG-2 are involved in the process of host cell recognition and invasion, there are still many unanswered questions which might be investigated in the future. These investigations could provide a better understanding of the involvement of these molecules during host cell recognition and invasion of bovine cells by T. annulata sporozoites. For example it has been shown that both 1A7 and 4B11 block invasion of sporozoites into host cells, but how exactly do they achieve this? Do they prevent the sporozoite from attaching to the host cell, or can the sporozoite still bind to host cells but the invasion process is 176 blocked at a later stage? Further questions which are of interest are; how and when are the proteins processed, how many of the processed SPAG-1 and SPAG-2 products are expressed on the surface of the sporozoite and what happens to SPAG-1 after the sporozoite has entered the host cell. To answer these questions a plentiful supply of live sporozoites is needed. The answers might reveal which parts of the SPAG molecules are involved and further elucidate the mechanism of invasion. Recombinant constructs of various parts of the SPAG-1 molecule might be employed in binding studies to map the region of the protein which is involved. Another major area of research lies in the cloning and sequencing of the full length SPAG-2 gene, which I believe is currently being addressed by Dr. Knight. This might yield more information about the involvement of SPAG-2 in the invasion process through homology to other molecules. Other major areas to follow up are the identification of further molecules expressed on the sporozoite surface which might also be involved in the invasion process, and molecules which are expressed by the target cell and allow sporozoite invasion. Such sporozoite molecules might be included in future sub-unit vaccines which could be developed to protect against tropical theileriosis. Information gained about the ligand-receptor interactions which are involved in the invasion process might, in future, yield the basis for the development of preventative treatments which inhibit sporozoites from binding and invading host cells. 177 Chapter 6. General Discussion. The work described in this D. Phil thesis is aimed to further the understanding of the biology and practical applications of SPAG-1 in particular and, to a lesser extent, SPAG-2. The three main areas of study attempt to elucidate: a) information about SPAG-1 which is of importance for sub-unit vaccine development, b) the process of stage-specific regulation of the expression of the SPAG-1 gene and c) the involvement of the SPAG molecules during host cell invasion. The results obtained in these areas are discussed below. 6.1. SPAG-1 and sub-unit vaccine development. It has been suggested that SPAG-1 is a candidate for the inclusion in a sub-unit vaccine (Hall et al., 1992). This prediction is based on two observations. The first is that the monoclonal antibody 1A7 blocks sporozoite invasion into host cells in vitro (Williamson et al., 1989). The 1A7 epitope has subsequently been mapped to the SRI region of SPAG-1 (Boulter et al., 1994). Secondly, a trial using recombinant SRI indicated that vaccination with SPAG-1 induces partial immunity in cattle in vivo (Boulter et al., 1995). Further support for the potential of SPAG-1 as a vaccine component is gained from its homology to p67, a sporozoite antigen of T. parva. I have shown that p67 and SPAG-1 are 47 % identical at the protein level and that they share parts of the 1A7 epitope. p67 has also shown promising results in vaccination trials providing partial and full protection from T. parva infection. The high homology between these two antigens may allow the development of a combined T. annulata and T. parva vaccine. 178 If SPAG-1 is to be included as a component in a sub-unit vaccine, then it is important to know as much about the protein as possible. It has been shown that the SPAG-1 gene is linked to an EcoRI RFLP (Williamson et al., 1989), but the extent of the SPAG-1 polymorphism was not known. My studies identified the existence of four polymorphic SPAG-1 alleles from two distinct geographical isolates. Analysis of the protein sequence revealed that the C- and N-termini are highly conserved (92 and 97 % identity respectively) and that the central region of the antigen is highly polymorphic. This finding suggests that SPAG-1 sub-unit vaccine components should encompass the conserved N- and C-terminal regions as well as a cocktail of sequences from the polymorphic region of the middle of SPAG-1. The inclusion of this SPAG-1 cocktail would increase the probability of inducing cross-protective responses in the immunised cattle. It might also be important to eliminate the elastin repeats from the vaccine components. This is because they might induce auto-immune responses of the host, although to date, no humoral auto-immune responses against these regions have been reported either in immunised or naturally infected cattle (Boulter, personal communication). The immunogenicity of the SPAG-1 antigen still need further characterisation as only one protective B-cell epitope has been found in SPAG-1. To optimise the SPAG-1 component in a sub-unit vaccine it is necessary to identify more B-cell and T-cell epitopes. A recent vaccination trial using recombinant SPAG-1 has indicated that there is at least one T-cell suppressor epitope located in the SRI region of SPAG-1 (Boulter et al., 1995). It is thus important to locate such suppressor epitopes in order to eliminate these from the sub-unit vaccine components. Furthermore, it is also of importance to identify protective T-cell epitopes from polymorphic SPAG-1 antigens in order to optimise the vaccine components and thus to maximise the immune responses. Another important area for further study is the optimisation of the vaccine delivery system. This includes research on antigen expression systems as well as adjuvants. 179 6.2. Stage-specific regulation of the SPAG-1 gene. It has been shown that SPAG-1 is only found during the sporozoite stage of the T. annulata life cycle (Williamson, 1988). It was also shown that SPAG-1 expression is regulated at the transcriptional level (Williamson et al., 1989). This finding is based on Northern blot analysis of RNA extracted from infected tick salivary gland, macroschizonts and piroplasms. SPAG-1 specific mRNA was only observed during sporozoite development. In order to investigate the mechanism of transcriptional regulation of the SPAG-1 gene, the 5' region of the gene was sequenced and subsequently the beginning of the mRNA was mapped. The analysis of the 5' region of the mRNA initiation site of SPAG-1 and the homologous region of p67 showed that no conserved sites for known transcription factors are present. But the analysis revealed two palindromic nucleic acid sequences, 6 bp in length, which were conserved. The function of these sequences is unknown but it can be speculated that they might be involved in the stage-specific transcription initiation of both SPAG-1 and p67. To prove this theory one would have to conduct band-shift assays. Attempts to isolate DNA binding proteins which bind to the 5' untranslated region of SPAG-1 failed. If the band-shift assays reveal the involvement of this region, attempts could be made to isolate transcription factors from sporozoite material using affinity columns. The analysis of the mRNA, using Sl nuclease, confirmed the existence of a 30 bp cryptic intron in the genomic SPAG-1 gene sequence. The existence of this intron was predicted during the comparison of the two full length SPAG-1 gene sequences, cH and gH3.4. The Si nuclease analysis confirmed the existence of the intron as well as its position. The intron also follows the GT-AG rule for intron splice sites found in eukaryotes (Perlman et al., 1984). 180 6.3. Functional importance of the SPAG molecules. It has been suggested that SPAG-1 and SPAG-2 are ligands which are involved in host cell invasion (Hall et al., 1992; Knight, 1993). These suggestions are based on results from sporozoite inhibition assays. For SPAG-1, these assays indicate that the C-terminus is of importance during sporozoite invasion of host cells. Further data indicating the importance of the C-terminus are presented in chapter 3. Sequence analysis of four SPAG-1 genes has indicated that the C-terminus is highly conserved in the alleles investigated, while the middle part of the gene and its predicted amino acid sequence are highly polymorphic. This indicates that the C-terminus is probably conserved for functional purposes such as host cell invasion. The polymorphism of the middle region suggests that the antigen is under selection pressure to change. A possible explanation for the polymorphism is the evasion of the immune responses of the bovine host. It is probable that the PGVGV elastin-like repeats, which are found in the four alleles investigated, have the same function, i. e. immune evasion (Hall, 1994). The sporozoite inhibition assays however do not indicate during which stage of the invasion process SPAG-1 is involved, and even less is known about the role of SPAG-2. The results of my binding studies using recombinant SPAG molecules have shown that both SPAG-1 and SPAG-2 bind specifically to a sub-population of bovine peripheral blood mononuclear cells. Neither, however, showed enough specificity to recognise only cells which have been identified as targets for T. annulata sporozoites. I would thus predict that both antigens are involved in the process of host cell invasion. Further, I suggest that both antigens, and possibly other sporozoite antigens which have not yet been isolated, act together in the recognition of a host cell and are subsequently involved in the invasion process. The process of host cell recognition and invasion needs to be further investigated in order to substantiate this prediction. For example, more data is needed about the fate of SPAG-1 during the invasion process, i. e. is it shed during invasion like p67 in T. parva, and which regions of SPAG-1 are involved in the invasion process? 181 Another very important area of research is the identification and isolation of receptors on the host cell to which the sporozoite ligands bind during the invasion process. The elastin receptor has been suggested to fulfil this role as SPAG-1 contains three VGVAPG hexapeptides (Hall et al., 1992). Data presented in this thesis, however, suggest that the elastin receptor is not of importance during sporozoite invasion. The evidence for this is two-fold. Firstly, the genomic SPAG-1 allele, gH3.4, does not contain any VGVAPG hexapeptides. Since SPAG-1 is a single copy gene, sporozoites expressing the gH3.4 SPAG-1 allele would not be able to invade host cells as they do not express the amino acid sequence recognised by the elastin receptor. The second line of evidence is based on flow cytometry data using BL3 cells and the anti-elastin receptor antibody BCZ. BL3 cells are targets for sporozoite invasion in vitro (Baylis et al., 1992b), but my data shows that these cells do not express the elastin receptor. The two sets of data, therefore, suggest that the elastin receptor cannot be the sole receptor for sporozoite invasion, but does not prove that the receptor is not involved in the invasion process at all. However, there is evidence suggesting that the elastin receptor is not involved, provided independently by Campbell et al. (1994) and Wilkie (personal communication). Campbell and co-workers sorted bovine peripheral blood mononuclear cells into two populations; one that expresses the elastin receptor and one that does not. Both these populations could be infected by sporozoites equally well. The second set of experiments by Wilkie showed that sporozoite invasion of host cells could not be blocked by the addition of the anti-elastin receptor antibody. BCZ, nor with BA4, the anti-VGVAPG antibody (Wilkie, personal communication). These four independent experiments show that the elastin receptor is not of importance in sporozoite invasion of host cells. This means, therefore, that no host cell receptor for the invasion of host cells by T. annulata sporozoites has been identified. In contrast, the MHC class I molecule has been shown to be involved during the invasion process of T. parva (Shaw et al., 1991; Shaw et al., 1995). However, the MHC class I molecule is found on every blood 182 cell and therefore cannot be responsible for specificity of T. parva sporozoites for T cell invasion. As a result, it seems more likely that the MHC class I molecule is involved in the invasion process subsequent to host cell recognition. 183 References Achbarou, A., Mercereau-puijalon, 0., Autheman, J. M., Fortier, B., Camus, D. and Dubremetz. J. F. (1991) Characterization of microneme proteins of Toxoplasma gondii. Molecular and Biochemical Parasitology 47,223-233. Adams, J. H., Lee-Sim, B. K., Dolan, S. A., Fang, X., Kaslow, D. C. and Miller, L. H. (1992a) A family of erythrocyte binding proteins of malaria parasites. Proceedings of the National Academy of Sciences, U. S. A. 89, 7085-7089. Adams, J. H., Fang, X., Kaslow, D. C. and Miller, L. H. (1992b) Identification of a cryptic intron in the Plasmodium vivax Duffy binding protein gene. Molecular and Biochemical Parasitology 56,181-184. Aikawa, M., Miller, L. H., Johnson, J. and Rabbege, J. (1978) Erythrocyte entry by malarial parasites. The Journal of Cell Biology 77,72-82. Aikawa, M., Yoshida, N., Nussenzweig, R. S. and Nussenzweig, V. (1981) The protective antigen of the rodent malaria sporozoites (Plasmodium berghel) is a differentiation antigen. Journal of Immunology 126,2494- 2495. Allsopp, B. A., and Allsopp, M. T. E. P. (1988) Theileria parva : genomic DNA studies reveal intra-specific sequence diversity. Molecular and Biochemical Parasitology 28,77-84. Allsopp, B., Carrington, M., Baylis, H., Sohal, S., Dolan, T. and lams, K. (1989) Improved characterisation of The leria parva isolates using the polymcrase chain reaction and oligonucleotide probes. Molecular and Biochemical Parasitology 35,137-148. Baldwin, C. L. Black, S. L. Brown, W. C., Conrad, P. A., Goddeeris, B. M., Kinuthia, S. W., Lalor, P. A., MacHugh, N. D., Morrison, W. I., Mozaria, S. P., Naessens, J. and Newson, J. (1988) Bovine T cells, B cells, and Null cells are transformed by the protozoan parasite Theileria parva. Infection and Immunity 56,462-467. Bannister, L. H. and Dluzewski, A. R. (1990) The ultrstructure of red cell invasion in malaria infections: a review. Blood Cell 16,257-292. Barnwell, J. W., Nichols, M. E. and Rubinstein, P. (1989) In vitro evaluation of the role of the Duffy blood group erythrocyte invasion by Plasmodium vivax. Journal of Experimental Medicine 169,1795-1802. Barta. J. R., Jenkins. M. C. and Danforth, H. D. (1991) Evolutionary relationships of avian Eimeria species among other Apicomplexan protozoa: monophyly of the Apicomplexa is supported. Molecular Biology of Evolution 8,345-355. Baylis, H. A., Megson, A., Mottram, J. C. and Hall, R. (1992a) Characterisation of a gene for a cysteine protease from Theileria annulata. Molecular and Biochemical Parasitology 54,105-108. Baylis, H. A.. Megson, A., Brown, C. G. D., Wilkie, G. F. and Hall, R. (1992b) Thelleria annulata -infected cells produce abundant proteases whose activity is reduced by long-term cell culture. Parasitology 105,417-423. Baylis, H. A., Megson, A. and Hall, R. (1995) Infection with Theileria annulata induces expression of matrix mettaloproteinase 9 and transcription factor AP-1 in bovine leucocytes. Molecular and Biochemical Parasitology 69,211-222. Beckers, C. J. M., Dubremetz, J. -F., Mercereau"Puijalon, 0. and Joiner, K. A. (1994) The Toxoplasma gondil rhoptry protein ROP2 is inserted into the parasitophorous vacuole membrane, surrounding the intracellular parasite, and is exposed to the host cell cytoplasm. Journal of Cell Biology 127,947.961. 185 Ben Miled, L., Dellagi, K., Bernardi, G., Melrose, T. M., Darghouth, M., Bouttour, A., Kinnaird, L. Shields, B., Tait, A. and Brown, C. G. D. (1994) Genomic and phenotypic diversity of Tunisian Theileria annulata isolates. Parasitology 108,51-60. Bermudes, D., Dubremetz, J. F., Achbarou, A. and Joiner, K. A. (1994) Cloning of a cDNA encoding the dense granule protein GRA3 from Toxoplasma gondil. Molecular and Biochemical Parasitology 68,247-257. Biedenkapp, H., Borgmeyer, U., Sippel, A. E. and Klempnauer, K. -H. (1988) Viral myb encodes a sequence-specific DNA-binding activity. Nature 335, 835-837. Bienz, M. and Pelham, H. R. B. (1986) Heat shock regulatory elements function as an inducible enhancer when linked to a heterologous promoter. Cell 45,753-760. Bishop, R. P., Sohanpal, B. K., Allsopp, B. A., Spooner, P. R., Dolan, T. T. and Morzaria, S. P. (1993) Detection of polymorphism among Theileria parva stocks using repetitive, telomeric and ribosomal DNA probes and anti- schizont monoclonal antibodies. Parasitoligy " 107,19-31. Blackman, M. L. Heidrich, H. G., Donachie, S., McBride, J. S. and Holder, A. A. (1990) A single fragment of a malaria merozoite surface protein remains on the parasite during red cell invasion and is the target of invasion-inhibiting antibodies. Journal of Experimental Medicine 172, 379-382. Blackman, M. L. Whittle, H. and Holder, A. A. (1991a) Processing of the Plasmodium falciparum major surface protein-1: identification of a 33- kilodalton secondary processing product which is shed prior to erythrocyte invasion. Molecular and Biochemical Parasitology 49,35-44. Blackman, M. J., Ling, I. T., Nicholls, S. C. and Holder, A. A. (1991b) Proteolytic processing of Plasmodium falciparum merozoite surface protein-I produces a membrane-bound fragment containing two epidermal growth factor-like domains. Molecular and Biochemical Parasitology 49,29-34. 186 Blackman, M. J., Chappel, J. A., Shai, S. and Holder, A. A. (1993) A conserved parasite serine-protease processes the Plasmodium falciparum merozoite surface protein-l. Molecular and Biochemical Parasitology 62,103-114. Blood, C. H., Sasse, J., Brodt, P. A. and Zetter, B. R. (1988) Identification of a tumor cell receptor for VGVAPG, a elastin derived chemotactic- peptide. Journal of Cell Biology 107,1987-1992. Bloom, B. R. (1979) Games parasites play: how parasites evade the immune surveillance. Nature 279.21-26. Borst. P. (1991) Molecular genetics and antigenic variation. Immunology Today 12, A29-A33. Boulter, N., Knight, P. A., Hunt, P. D., Hennessey, E. S., Katzer, F., Williamson, S., Brown, D., Baylis, H. A. and Hall, F. R. (1994) Theileria annulata sporozoite surface antigen (SPAG-1) contains neutralizing determinants in the C terminus. Parasite Immunology 16,97-104. Boulter, N. R., Glass, E. L. Knight, P. A., Bell-Sakyi, L., Brown, C. G. D. and Hall, R. (In press) Theileria annulata sporozoite antigen fused to hepatitis B core antigen used in vaccination trial. Vaccine. Bradford, M. (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utilising the principle of protein-dye binding. Analytical Chemistry 72,248-254. Braunbreton, C., Blisnick, T., Moralesbetoulle, M. E., Barale, J. C. and Langsley, G. (1994) Malaria parasites - enzymes involved in red-blood- cell invasion. Brazilian Journal of Medical and Biological Research 27, 363-367. Brown, C. G. D., Stagg, D. A., Purnell, R. E., Kanhai, G. K. and Payne, R. C. (1973) Infection and transformation of bovine lymphoid cells in vitro by infective particles of Theileria parva. Nature 245,101-103. 187 Brown, C. G. D. and Gray, M. A. (1981) Infection of bovine fibroblastic cell lines with Theileria parva and Theileria annulata. Transaction of the Royal Society of Tropical Medicine and Hygiene 75,323-324. Brown, C. G. D. (1987) Thelleridae. In: In vitro methods for parasite culture. (Taylor, A. E. R. and Baker, J. R., eds. ) Academic Press London. pages 230-250. Brown, C. G. D. (1990a) Control of Tropical Theileriosis (Theileria annulata infection) of cattle. Parasitologia 32,23-31. Brown, S. D. and Van der Ploeg, L. H. T. (1994) Single-stranded DNA- protein binding in the procyclic acidic repetitive protein (PARP) promoter of Trypanosoma brucel. Molecular and Biochemical Parasitology 56,109-122. Brown, S. J. and Askenase, P. W. (1983) Immune rejection of ectoparasitic (ticks) by T cell and IgGI antibody recruitment of basophils and eosinophils. Federation Proceedings 42,1744-1749. Brown, W. C., Sugimoto, C., Conrad, P. A. and Grab, D. J. (1989) Differential response of bovine T-cell lines to membrane and soluble antigens of Theileria parva schizont infected cells. Parasite Immunology 11,567- 583. Brown, W. C., Lonsdale-Eccles, J. D., deMartini, J. C. and Grab, D. J. (1990b) Recognition of soluble Theileria parva antigen by bovine helper T cell clones: characterization and partialpurification of the antigen. The Journal of Immunology 144,271-277. Buratowski. S., Hahn, S., Guarcnte, L. and Sharp, P. A. (1989) Five intermediate complexes in transcription initiation by RNA polymerase II. Ccll 56.549-561. Campbell. J. D.. Brown, DJ.. Glass, E. L. Hall, F. R. and Spooner, R. L. (1994) Theilerla annulata sporozoite targets. Parasite Immunology 16,501-505. 188 Carey, M., Lin, Y. -S., Green, M. R. and Ptashne, M. (1990) A mechanism for synergistic activation of a mammalian gene by GAL4 derivatives. Nature 345,361-364. Cerami, C., Frevert, U., Sinnis, P., Takacs, B., Clavijo, P., Santos, M. J. and Nussenzweig, V. (1992) The basolateral domain of the hepatocyt plasma membrane bears receptors for the circumsporozoite protein of Plasmodium falciparum sporozoites. Cell 70,1021-1033. Cerami, C., Frevert, U., Sinnis, P., Takacs, B. and Nussenzweig, V. (1994) Rapid clearance of malaria circumsporozoite protein (CS) by hepatocytes. Journal of Experimental Medicine 179,695-701. Cesbron-Delauw, M. F., Guy, B., Tropier, G., Pierce, R. J., Lenzen, G., Cesbron, J. Y., Charif, H., Lepage, P., Darcy, F., Lecocq, J. P. and Capron, A. (1989) Molecular characterization of a 23-kilodalton major antigen secreted by Toxoplasma gondii. Procedings of the National Academy of Science of the USA 86,7537-7541. Chappel, J. A. and Holder A. A. (1993) Monoclonal antibodies that inhibit Plasmodium falciparum invasion in vitro "recognise the first growth factor-like domain of the merozoite surface protein-1. Molecular and Biochemical Parasitology 60,303-312. Chaudhri, S. S., and Subramanian, G. (1992a) In vitro infectivity of sporozoites of Theileria annulata to viable lymphocytes of cattle and buffalo. Indian Journal of Animal Sciences 62,225-226. Chaudhri, S. S., and Subramanian, G. (1992b) Cell-mediated immune responses to sporozoites and macroschizonts of Theileria annulata. Veterinary Parasitology 41,23-34. Chitnis, C. E. and Miller, L. H. (1994) Identification of the erythrocyte binding domain of the Plasmodium vivax and Plasmodium knowlesi proteins involved in erythrocyte invasion. Journal of Experimental Medicine 180,497-506. 189 Clarke, L. E., Tomley, F. M., Wisner, M. H., Foulas, I. J. and Boursell, M. E. G. (1990) Regions of an Eimeria teneila antigen contain sequences which are observed in circumsporozoite proteins from Plasmodium spp and which are related to the thrombospondin gene family. Molecular and Biochemical Parasitology 41,269-280. Clarke, R. M., Donachie, S., Anand, R., Wilson, C. F., Heidrich, H. G. and McBride, J. S. (1989) 46-53 kda glycoprotein from the surface of Plasmodium falciparum merozoites. Molecular and Biochemical Parasitology 32,15-24. Clayton, C. E. and Mowatt, M. (1989) The procyclic acid repetitive proteins of Trypanosoma brucel. Journal of Biological Cemistry 264,15088-15093. Conrad, P. A., Baldwin, C. L., Brown, W. C. Sohanpal, B., Dolan, T. T., Goddeeris, B. M., Demartini, J. C. and Ole-Moiyoi, O. K. (1989) Infection of bovine T cell clones with genotypically distinct Theileria parva parasites and analysis of their cell surface phenotype. Parasitology 99, 205-213. Cooper, J. A., Cooper, L. T. and Saul, A. J. (1992) Mapping of the region predominantly recognized by antibodies to the Plasmodium falciparum merozoite surface antigen MSA-1. Molecular and Biochemical Parasitology 51,301-312. Corden, J., Wasylyl, B., Buchwalder, A., Sassone-Corsi, P., Kedinger, C. and Chambon, P. (1980) Promoter sequences of eukaryotic protein-coding genes. Science 209,1406-1414. Courey, A. J. and Tjian R. (1988) Analysis of Spl in vivo reveals multiple transcriptional domains including a novel glutamine-rich activation motif. Cell 55,887-898. Cowdry, E. V. and Ham, A. W. (1932) Studies on East Coast Fever. I. The life cycle of the parasites in the ticks. Parasitology 24,1-43. 190 Demelo, E. J. T., Decarvalho, T. U. and Desouza, W. (1992) Penetration of Toxoplasma gondii into host cells induces changes in the distribution of the mitochondria and the endoplasmic-reticulum. Cell Structure and Function 17.311-317 Devereux, -J., (1989) Genetics computer group sequence analysis package. Proceedings of the National Academy of Science, USA 74,5463-5467. Dickson, J. and Shiels, B. R. (1993) Antigenic diversity of a major merozoite surface molecule in Theileria annulata. Molecular and Biochemical Parasitology 57,55-64. Dobbelaere, D. A. E., Spooner, P. R., Barry, W. C. and Irvin, A. D. (1984) Monoclonal antibody neutralizes the sporozoite stage of different Theileria parva stocks. Parasite Immunology 6,361-370. Dobbelaere, D. A. E., Leitch, B. L., Voigt, W. P. and Irvin, A. D. (1985) Theileria parva : Expression of a sporozoite surface coat antigen. Experimental Parasitology 60,90-100. Dolan, T. T. (1989) Theileriasis: a comrehensive review. Revue Scientifique et Technique (Paris) 8,11-36. Dolan, S. A., Miller, L. H. and Wellems, T. E. (1990) Evidence for a switching mechanism in the invasion of erythrocytes of Plasmodium falciparum. Journal of Clinical Investigation 86,618-624. Dolan, S. A., Proctor, J. L., Ailing, D. W., Okubo, Y., Wellems, T. E. and Miller, L. H. (1994) Glycophorin-B as an EBA-175 independent Plasmodium falciparum receptor of human erythrocytes. Molecular and Biochemical Parasitology 64,55-63. Dschunkowsky, E. and Luhs, J. (1904) Die Piroplasmosen der Rinder Zentralblatt fur Bakteriologie, Parasitenkunde, Infektionskrankheiten und Hygiene 35,486-492. 191 Dubremetz, J. F. and Schwartzman, J. D. (1993) Subcellular organelles of Toxoplasma gondii and host cell invasion. Research in Immunology 144, 31-33. Dyer, M. and Tait A. (1987) Control of lymphoproliferation by Theileria annulata. Parasitology Today 3,309-311. Dyer, M., Shiels, B. and Tait, A. (1992) Theileria annulata : Alterations in phosphoprotein and protein kinase activity profiles in infected leukocytes of the bovine host, Bos taurus. Experimental Parasitology 74, 216-227. Dynan, W. S. and Tjian R. (1985) Control of eukaryotic mRNA synthesis by sequence-specific DNA-binding proteins. Nature 316,774-778. Eichhorn, M. and Dobbelaere, D. A. E. (1994) Induction of Signal Transduction pathways in lymphocytes infected by Theileria parva. Parasitology Today 10,469-472. Ellis, J., Ozaki, L. S., Gwadz, R. W., Cochrane, A. H., Nussenzweig, V. and Nussenzweig, R. S. (1983) Cloning and expression in E. coli of the malarial sporozoite surface antigen gene from Plasmodium falciparum. Nature 302,536-538. Elmoudni, B., Philippe, M., Monsigny, M. and Schrevel, J. (1993) N- acetylglucosamine-binding proteins on Plasmodium falciparun: merozoite surface. Glycobiology 3,305-312. Elsholtz, V., Albert, V. R., Treacy, M. N. and Rosenfeld, M. G. (1990) A two- base change in a POU factor binding site sitches pituitary-specific to lymphoid specific gene expression. Genes and Development 4,43-51. Emery, D. L., MacHugh, N. D. and Morrison, W. I. (1988) Theileria parva (Muguga) infects bovine T-lymphocytes in vivo and induces expression of BoT4 and BoT8. Parasite Immunology 10,379-391. Evans, R. M. (1988) The steroid and thyroid hormone receptor gene superfamily. Science 240,889-895. 192 Falkner, F. G., Moickat, R. and Zachau H. G. (1986) Sequences closely related to an immunoglobulin promoter/enhancer element occur also upstream of other eukaryotic and prokaryotic genes. Nucleic Acid Research 13,7847-7863. Fawcett, D. W. and Doxsey S. (1982) The entry of sporozoites of Theileria parva into bovine lymphocytes in vitro. Electron microscopic observations. European Journal of Cell Biology 27,10-21. Fawcett, D. W., Buscher, G. and Doxsey, S. (1982) Salivery gland of the tick vector of East Coast Fever. III. The ultrastructure of sporogeny in Theileria parva. Tissue and Cell 14,183-206. Fawcett, D., Musoke, A. and Voigt, W. (1984) Interaction of sporozoites of Theileria parva with bovine lymphocytes in vitro . I. Early events after invasion. Tissue and Cell 16,873-884. Fenton, B., Clark, J. T., Wilson, C. F., McBride, J. S. and Walliker, D. (1989) Polymorphism of a 35-48 kDa Plasmodium falciparum merozoite surface antigen. Molecular and Biochemical Parasitology 34,79-86. Fidock, D. A., Bottius, E., Brahimi, K., Moelans, I. I., Aikawa, M., Konings, R. N., Certa, U., Olafsson, P., Kaidoh, T. and Asavanich, A. (1994a) Cloning and characterization of a novel Plasmodium falciparum sporozoite surface antigen, STARP. Molecular and Biochemical Parasitology 64,219- 232. Fidock, D. A., Sallenave-Sales, S., Sherwood, J. A., Gachihi, G. S., Ferreira-da- Cruz, M. F. Thomas, A. W. and Druilhe, P. (1994b) Conservation of the Plasmodium falciparum sporozoite surface protein gene, STARP, in field isolates and distinct species of Plasmodium. Molecular and Biochemical Parasitology 67,255-267. 193 Florent, I., Lebonniec, S., Carcy, B., Grellier, P., Mercereaupuijalon, 0., Bonnefoy, S., Dhermy, D., Monsigny, M., Mayer, R. and Schrevel, J. (1994) Plasmodium falciparum proteases- cloning of a putative gene coding for the merozoite proteinase for erythrocyte invasion (MPEI) and determination of hydolysis sites of spectrin by PF37 proteinase. Memorias do Instituo Oswaldo Cruz 89, S2,47-49. Flores, 0., Lu, H. and Reinberg, D. (1992) Factors involved in specific transcription by mammalian RNA polymerase II. Journal of Biological Chemistry 267,2786-2793. Frevert, U., Sinnis, P., Ccrami, C., Shreffler, W., Takacs, B. and Nussenzweig, V. (1993) Malaria circumsporozoite protein binds heperan sulfate proteoglycans associated with the surface membrane of hepatocytes. Journal of Experimental Medicine 177,1287-1298. Frevert, U. (1994) Malaria sporozitehepatocyte interactions. Experimental Parasitology 79,206-210. Foussard, F., Leriche, M. A. and Dubremetz, J. F. (1991) Characterization of the lipid content of Toxoplasma gondii rhoptries. Parasitology 102,367- 370. Furtado, G. C., Cao, Y. and Joiner, K. A. (1992) Laminin on Toxoplasma gondii mediates parasite binding to ß1 integrin receptor a6ß1 on human foreskin fibroblasts and chinese hamster ovary cells. Infection and Immunity 60,4925-4931. Gajadhar, A. A., Maequardt, W. C., Hall, R., Gunderson, J., Ariztia-Carmona, E. V. and Sogin, M. L. (1991) Ribosomal RNA sequences of Sarcocystis muris, Theileria annulata and Crypthecodinium cohnii reveal evolutionary relatioships among apicomplexans, dinoflagellates, and ciliates. Molecular and Biochemical Parasitology 45,147-154. Gaston, K. and Fried, M. (1992) The isolation of transcription factors from 71gt11 cDNA expression libraries: human steroid 5a-reductase 1 has sequence-specific DNA binding activity. Nucleic Acid Research 20,6297- 6301. 194 Gibson, W. C., Schwinkels, B. W. and Borst, P. (1988) Post-transcriptional control of the differential expression of phosphoglycerate kinase genes in Trypanosoma brucei. Journal of Molecular Biology 201,315-325. Gill, B. S., Bhattacharyulu, Y. and Kaur, D. (1976) Immunization against bovine tropical theileriosis (Theileria annulata infection). Research in Veterinary Science 21,146-149. Gill, B. S., Bansal, G. C., Bhattacharyulu, Y., Kaur, D. and Singh, A. (1980) Immunological relationship between strains of I Dschunkowsky and Luhs 1904. Research in Veterinary Science 29,93-97. Gilmore, T. D. (1990) NF-kappa B, KBF1, dorsal and related matters. Cell 62, 841-843. Glascodine, J., Tetley, L., Tait, A., Brown, D. and Shiels, B. (1990) Developmental expression of a Theileria annulata merozoite surface antigen. Molecular and Biochemical Parasitology 40,105-112. Glass, E. J., Innes, E. A., Spooner, R. L. and Brown, C. G. D. (1989) Infection of bovine monocytc/macrophage populations with Theileria annulata and Theileria parva. Veterinary Immunology and Immunopathology 22,355- 368. Glass, E. J., Spooner, R. L. (1990) Parasite-accessory cell interactions in Theileriosis. Antigen presentation by Theileria annulata -infected macrophages and production of continuously growing antigen- presenting cell lines. The European Journal for Immunology 20,2491- 2497. Godowski, P. J., Picard, D. and Yamaoto, K. (1988) Signal transduction and transcriptional regulation by glococorticoid receptor: regulation by glucocorticoid receptor-Lex A fusion proteins. Science 241,812-816. Goebl, M. G. (1990) The PU"1 transcription factor is the product of the putative oncogene Spi-I. Cell 61, 195 Good, M. F., Pombo, D., Quakyi, I., Riley, E. M., Houghten, R. A., Menon, A., Ailing, D. W., Berzofsky, J. K. and Miller, L. H. (1988b) Human T-cell recognition of the circumsporozoite protein of Palsmodium falciparum Immunodominant T-cell domains map to the polymorphic regions of the molecule. Proceedings of the National Academy of Science, USA 85,1199- 1203. Good, M. F., Pombo, D., Maloy, W. L. de la Cruz, V. F., Miller, L. H. and Berofsky, J. A. (1988c) Parasite polymorphism present within minimal T- cell epitopes of Plasmodium falciparum circumsporozoite protein. Journal of Immunology 1645-1650. Goundis, D. and Rcid, K. B. M. (1988) Properdin, the terminal complement components, thrombospondin and the circumsporozoite protein of malaria parasites contain simmilar sequence motifs. Nature 335,82-85. Gray, M. A. and Brown, C. G. D. (1981) In vitro neutralisation of theilerial sporozoite infectivity with immune serum. In "Advances in the Control of Theileriosis. Eds. A. D. Irvin, M. P. Cunningham and A. S. Young. Martinus Nijhoff Publishers, The Hague. pp 127-129. Grellier, P., Precigout, E., Valentin, A., Carc, B. and Schrevel, J. (1995) Characterization of a new 60-kDa apical protein of Plasmodium falciparun: merozoite expressed in late schizogeny. Biology of the Cell 82, 129-138 Grimwood, J. and Smith, J. E. (1992) Toxoplasma gondii: The role of a 30- kDa surface protein in host cell invasion. Experimental Parasitology 74, 106-111. Gunderson, J. H., Sogin, M. L. Wollett, G., Hollingdale, M., de la Cruz, V. F., Waters, A. P. and McCutchan, T. F. (1987) Structurally distinct, stage- specific ribosomes occur in Plasmodium. Science 238,933-937. Guttinger, M., Caspers, P., Takacs, B., Trzeciak, A., Gillessen, D., Pink, J. R. L. and Sinigaglia, F. (1988) Human T cells recognize polymorphic and non- polymorphic regions of the Plasmodium falciparum circumsporozoite protein. EMBO 7,2555-2558. 196 Hadani, A., Pipano, E. and Senf, Z. (1963) Studies on the transmissoin of Theileria annulata by ticks (Ixodoides, Ixodidae), I. Hyalomma excavatum. Journal of Protozoology 10, Supplement, 35. Hadley, T. J., Klotz, F. W. and Miller, L. H. (1986) Invasion of erythrocytes by malaria parasites: A cellular and molecular overview. Annual Reviews in Microbiology 40,451-477. Hall, R., Osland, A., Hyde, J. E., Simmons, D. L., Hope, I. A. and Scaife, J. G. (1984) Processing, polymorphism, and biological significance of P190, a major surface antigen of the crythrocitic forms of Plasmodium falciparum. Molecular and Biochemical Parasitology 11,61-80. Hall, R., Hyde, J. E., Goman, M., Simmons, D. L., Hope, I. A., MacKay, M., Scaife, L. Merkli, B., Richie, R. and Stocker, J. (1984) Major surface- antigen gene of a human malaria parasite cloned and expressed in bacteria. Nature 311,379-382. Hall, F. R., (1988) Antigens and immunity in Theieleria annulata. Parasitology Today 4,257-261. Hall, R., Coggins, L., McKellar, S., Shields, B. and Tait, A. (1990) Characterisation of an extrachromosomal DNA element from Theileria annulata. Molecular and Biochemical Parasitology 38,253-260. Hall, R., Hunt, P. D., Carrington, M., Simmons, D., Williamson, S., Mecham, R. P. and Tait, A. (1992a) Mimicry of the elastin repetitive motifs by Theileria annulata sporozoite surface antigen. Molecular and Biochemical Parasitology 53,105-112. Hall, R. (1994) Molecular mimicry. Advances in Parasitology 34,81-132. Hanes, S. D. and Brent, R. (1989) DNA specificity of the bicoid activator protein is determined by homeodomain recognition helix 9 Cell 57,1275- 1283. 197 Haynes, J. D., Dalton, J. P., Klotz, F. W., McGinnis, M. H., Hadley, T. J., Hudson, D. E. and Miller, L. H. (1988) Receptor-like specificity of a Plasmodium knowlest malaria protein that binds to Duffy antigen ligands on erythrocytes. Journal of Experimental Medicine 167,1873-1881. Headstorm, R. C., Campbell, J. R. and Leef, M. L. (1990) A malaria sporozoite surface antigen distinct from the circumsporozoite protein. World Health Organization Bulletin 68,152-157. Hesselgesser, J., Chitnis, C. E., Miller, L. H., Yansura, D. G., Simmons, L. C., Fairbrother, W. J., Kotts, C., Wirth, C., Gillececastro, B. L. and Horuk, R. (1995) A mutant of melanoma growth -sti mmul ati ng activity does not activate neutrophils but blocks erythrocyte invasion by malaria. Journal of Biological Chemistry 270,11472-11476. Hilder, A. A., Sanghu, J. S., Hillman, Y., Davey, L. S., Nicholls, S. C., Cooper, H. and Lockyer, M. J. (1987) Processing of the precursor to the major merozoite surface antigen of Plasmodium falciparum. Parasitology 94, 199-208. Hinek, A., Wrenn, D. S., Mecham, R. P. and Barondes, S. H. (1988) The elastin receptor: A galactoside-binding protein. Science 239,15391541. Holder. A. A. (1994) Proteins on the surface of the malaria parasite and cell invasion. Parasitology 108, S5-S18. Hollcnberg, S. M. and Evans, R. M. (1988) Multiple and cooperativetrans- acctivation domains of the human glucocorticoid receptor. Cell 55,899- 906. Holloway, S. P., Sims, P. F. G., Delves, C. J., Scaif, J. G. and Hyde, J. E. (1989) Isolation of a-tubulin genes from the human malaria parasite Plasmodium falciparum : sequence analysis of a-tubulin I. Molecular Microbiology 3,1501-1510. 0 198 11olloiiay. S. p., Gesousis, M.. Delis. CJ.. Sima. P. F. G.. Scaife, J. G. and Hyde, J. E. (1990) The tubulin genus of the human malaria parasite Plasmodium fatelpariren. their chromosomal location and sequence analysis of the a- tubulin 11 gene. Molecular and Biochemical Parasitology 43,257-270. llooituaaad-Rsd. P. W 11awa. NJ. (1973a) Rlaligncact theileriosis of shccp and pats,. Tropical Animal health and Production S. 97-102. 1looshmaad"Rad. P. aad llawa. NJ. (1973b) Transmission of Thtiltria htrict In shccp by 1lyalo+um. 7 aaatollcuu anatolicum. Tropical Animal llcatth ar4 P ducuon 5.103.109. Hope. 1. A. W Strunl. K. (1986) Functional dissection of a cukaryotic uanuriptiooal actiruot GCi: j of )cast. Cell 46.885"894. tton: k. R.. CAicais. CE.. Dufiorrzc. w'. C.. Colby. TJ.. Rybichi, A., Nadlcy, TJ. aod Mi11ct. L. it. (1993) Sticacc : 61.1182"1184. Ilowud. C L. Sopp. P.. Prtston. P. M., 3 chson. L. A. and Brown. C. G. D. (1992) i'bcnotrpic anslysls of bovine leukocyte cell lines infected with Tbt/lerla a*R1laU. `'ctcrinsry Immunology and lmmunopathology 39. 273.282. lluud. T. K.. Malinoff. 11. F. aa.! Wichs. MS. (1986) Macrophages express a plasma membrane rcccptor for basement membrane laminin. American Journal of Pathology 1: 3.36S"370. 11udwo-Taylor. D. L. Doisa. S. A.. KW. F. W» Fujioka. If.. Aikawa, rt.. tioonin. E. V. aa. i NGllcr. L. 11. (1993) Plasmodium jalciparum protein associated with the inrision junction contains conserved oxidoreductase domain. Molecular Alicrubiolo$7 13.463-471. 1lullig, cr. L,. Vr`il. Sc. 1. x. 11.. tirnrºa- CO. D. and Turner. L. (1964) Alodc of taultiplicatioa of 7Attttrta in cutwrcs of bovine lymphocyte cclls. hatVrc 20). 72E-730. 199 Hunt, P. D. and Hall, R. (1993) Immunoblotting & Enzyme linked immunosorbent assay in: "Protocols in Molecular Parasitology" Chapter 30. (Editor Hyde, D. J. E. ) Human Press, New Jersey Hyde, J. E. and Sims, P. F. G. (1987) Anomalous dinucleotide frequencies in both coding and non-coding regions from the genome of the human malaria parasite Plasmodium falciparum. Gene 61,177-187. Iams, K. P., Hall, R.. Webster, P. and Musoke, A. J. (1990) Identification of kgt11 clones encoding the major antigenic determinants expressed by Theileria parva sporozoites. Infection and Immunity 58,1828-1834. Imagawa, M., Chiu, R. and Karin, M. (1987) Transcription factor AP-2 mediates induction by two different signal-transduction pathways: Protein kinase C and cAMP. Cell 51,251-260. Ingraham, H. A., Chen, R., Mangalam, H. J., Elsholtz, H. P., Flynn, S. C., Linn, C. R., Simmons, D. M., Swanson, L. and Rosenfeld, M. G. (1988) A tissue specific factor containing a homeo domain specifies a pituitary phenotype. Cell 55,519-529. Innes, E. A., Millar, P. and Glass, E. J. (1989a) In vitro infection of bovine alloreactive cytotoxic T cell lines with sporozoites of Theileria annulata and T. parva. Research in Veterinary Science 46,367-374. Innes, E. A., Millar, P., Brown, C. G. D. and Spooner, R. L. (1989b) The development and specificity of cytotoxic cells in cattle immunized with autologous or allogeneic Theileria annulata -infected lymphoblastoid cell lines. Parasite Immunology 11,57-68. Irvin, A. D. and Boarer, C. D. H. (1980) Some implication of a sexual cycle in Theilerla. Parasitology 80,571-579. Irvin, A. D. (1987a) Characterization of species and strains of Theileria. Advances in Parasitology 26,145-197. 200 Irvin, A. D. and Morrison, W. I. (1987b) Immunopathology, immunology, and immunoprphylaxis of Theileria infections. Immune responses in parasitic infection: immunology, immunopathology, immunoprophylaxis. Volume III, Protozoa. Soulsby, E. J. L. (Editor), CRC Press Inc., Boca Raton, Florida., Pages 223-274. Ivanov, V., Stein, B., Bauman, I., Dobbelaere, D. A. E., Herrlich, P. and Williams, R. O. (1989) Infection with the intracellular protozoan parasite Theileria parva induces constitutively high levels of NF-KB in bovine T lymphocytes. Molecular and Cellular Biology 9,4677-4686. Ivanov, V. N. and Williams, R. O. (1991) The activation of trans-acting transcription factors in T-lymphocytes infected by intracellular protozoan parasite Theileria parva. Genetika 27,197-209. Jaynes, J. B. and O'Farrell, P. H. (1988) Activation and repression of transcription by homeodomain-containing proteins that bind a common site Nature 336,744-749. Johnson, P. F. and McKnight, S. L. (1989) Eukaryotic transcriptional regulatory proteins. Annual Review of Biochemistry 58,799-839. Joiner , K. A., Fuhrman, S. A., Miettinen, H. M., Kasper, L. H. and Mellman, (1990) Toxoplasma gondii: fusion competence of the parasitophorous vacuoles in Fc receptor-transfected fibroblasts. Science 249,641-646. Joiner, K. A. and Dubremetz, J. F. (1993) Infection and Immunity 61,1169- 1172. Jura, W. G. Z. 0., Brown, C. G. D. and Kelly, B. (1983) Fine structure and invasive behaviour of the early developmental stages of Theileria annulata in vitro. Veterinary Parasitology 12,31-44. Jura, W. G. Z. 0. (1984) Factors affecting the capacity of Theileria annulata sporozoites to invade bovine peripheral blood lymphocytes. Veterinary Parasitology 16,215-223. 201 Kadonga, J. T., Camer, K. R., Masiarz, F. R. and Thian, R. (1987) Isolation of cDNA encoding the transcription factor Spi and functional analysis of the DNA binding domain. Cell 51,1079-1090. Kain, K. C., Orlandi, P. A., Haynes, J. D., Sim, B. K. L. and Lanar, D. E. (1993) Evidence for 2-stage binding by the 175-kD erythrocyte binding antigen of Plasmodium falciparum. Journal of Experimental Medicine 178,1497- 1505. Katende, J. M., Goddeeris, B. M., Morzaria, S. P., Nkonge, J. M. and Musoke, A. J. (1990) Identification of Theileria mutans-specific antigen for use in an antibody and antigen detection ELISA. Parasite Immunology 12,419- 433. Katzer, F., Carrington, M., Knight, P., Williamson, S., Tait, A., Morrison, I. W. and Hall, R. (1994) Polymorphism of SPAG-1, a candidate antigen for inclusion in a sub-unit vaccine against Theileria annulata. Molecular and Biochemical Parasitology 67,1-10. Kasper, L. H., Mineo, J. R. (1994) Attachment and invasion of host cells byToxoplasma gondii. Parasitology Today 10,184-188. Kawamoto, S., Takahashi, K., Onuma, M., Kurosawa, T. and Sonoda, M. (1990) Invasion of Bovine erythrocytes by Theileria sergenti piroplasms in vitro. Japanese Journal for Veterinary Science 52,1261-1263. Kimber, C. D., Purnell, R. E. and Sellwood, S. A. (1973) The use of fluorescent antibody techniques to detect Theileria parva in salivary glands of the tick Rhipicephalus appendiculatus. Research in Veterinary Sciences 14,126-127. Klotz, F. W., Orlandi, P. A., Reuter, G., Cohen, S. J., Haynes, J. D., Schauer, R., Howard, R. J., Palese, P. and Miller, L. H. (1992) Binding of Plasmodium falciparum 175-kilodalton erythrocyte binding antigen and invasion of murine erythrocytes requires N-acetylneuraminic acid but not its O- acetylated form. Molecular and Biochemical Parasitology 51,49-54. 202 Knight, P. A. (1993) The characterisation of Theileria annulata sporozoite surface antigens. Ph. D. Thesis, University of Glasgow. Koch, R. (1906) Beitraege zur Entwicklungsgeschichte der Piroplasmosen. Zeitschrift fuer Hygiene und Infektionskrankheiten 54, 1-9. Kumar, V. and Chambon, P. (1988) The estrogen receptor binds tightly to its responsive element as a ligand-induced homodimer. Cell 55,145-156. Kurtti, T. J., Munderloh, U. G., Irvin, A. D. and Buscher, G. (1981) Thelleria parva : Early events in the development of bovine lymphoblastoid cell lines persistently infected with macroschizonts. Experimental Parasitology 52,280-290. Laemmli, U. K. (1970) Cleavage of the structural proteins during the assembly of the head of the bacteriophage T4. Narure 227,680-682. Laiblin, C. (1978) Klinische Untersuchungen zur Theileria annulata - Infektion des Rindes. Berliner und Munchener Tierarztliche Wochenschrift 91,48-50. Lalor, P. A., Morrison, W. I. and Black, S. J. (1986) Monoclonal antibodies to bovine leukocytes define heterogeneity of target cells for in vitro patrasitosis by Theileria parva . Edited by Morrison W. I. In: The ruminant system in health and disease. Cambridge University Press. pp 72-87. Landschultz, W. H., Johnson, P. F. and McKnight, S. L. (1989) The DNA binding domain of the rat liver nuclear protein C/EBP is bipartite. Science 243,1681-1688. Langsley, G. and Roth, C. (1987) Antigenic variation in parasitic protozoa. Microbial Sciences 4,280-285. 203 Lanzer, M., de Bruin, D. and Ravetech, J. V. (1992a) Transcritpion mapping of a 100 kb locus of Plasmodium falciparum identifies an intergenic region in which transcription terminates and reinitiates. EMBO Journal 11,1949-1955. Lanzer, M., de Bruin, D. and Ravetech, J. V. (1992b) A sequence element associated with the Plasmodium falciparum KAHRP gene is the site for developmentally regulated protein-DNA interactions. Nucleic Acids Research 12,3051-3056. Lanzer, M., Wertheimer, S. P., de Bruin, D. and Ravetch, J. V. (1993) Plasmodium : Control of gene expression in malaria parasites. Experimental Parasitology 77,121-128. Latchman, D. S. (1990) Gene Regulation, A eukaryotic perspective. Unwin Hyman, London Latchman. D. S. (1991) Eukaryotic Transcription Factors. Academic Press Limited. London Latchman, D. S. (1993) Transcripton Factors: A practical Approach. (Editor Latchman, D. S., Series editors: Rickwood, D. and Harnes, B. D. ). IRL Press, Oxford University Press. Lecordier, L., Mercier, C., Torpier, G., Tourvieille, B., Darcy, F., Liu, J. L., Macs, P., Tartar, A., Capron, A. and Cesbron_Delauw. (1993) Molecular structure of a Toxoplasma gondii dense granule antigen (GRA-5) associated with the parasitophorous vacuole membrane. Molecular and Biochemical Parasitology 59,143-154. Lee, W., Mitchell, P. and Tjian, R. (1987) Purified transcription factor AP- 1 interacts with TPA-inducible enhancer elements. Cell 49,741-752. Lee, M. G. -S. and Van der Ploeg, L. H. T. (1990) Homologous recombination and stable transfection in parasitic protozoan Trypanosoma brucei. Science 250,1583-1587. 204 Lee, M. G: S. (1995) A foreign transcriptioon unit in the inactivated VSG gene expression site of the procyclic form of Trypanosoma brucei and formation of large episomes in stably transformed trypanosomes. Molecular and Biochemical Parasitology 69,223-238. Lenardo, M. J. and Baltimore, D. (1989) NF-kappa B: a pleotropic mediator of inducible and tissue-specific gene control. Cell 58,227-229. Lewis, A. P. (1989) Cloning and analysis of a gene encoding the 230- kilodalton merozoite surface antigen of Plasmodium yoelii. Molecular and Biochemical Parasitlogy 36,271-282. Lewis, A. P. (1990) Sequence analysis upstream of the gene encoding the precursor to the major merozoite surface antigen of Plasmodium yoelii. Molecular and Biochemical Parasitology 39,285-288. Lockyer, M. J., Marsh, K. and Newbold, C. I. (1989) Wild isolates of Plasmodium falciparum show extensive polymorphism in T-cell epitopes of the circumsporozoite protein. Molecular and Biochemical Parasitology 37,275-280. Maniatis, T., Goodbourn, S. and Fischer, J. A. (1987) Regulation of inducible and tissue specific gene expression. Science 236,1237-1245. Mash, K. and Howard, R. J. (1986) Antigens induced erythrocytes by Plasmodium falciparum , expression of diverse and conserved domains. Science 231,150-153. Mason, P. J., Shiels, B. R., Tait, A., Beck, P. and Hall, R. (1989) Sequence and expression of a gene from Theileria annulata coding for a 70-kilodalton heat-shock protein. Molecular and Biochemical Parasitology 37,27-36. Mathews, G. V. and Vanderberg, J. P. (1994) Plasmodium sporozoites possess both positively and negatively charged sites accessible on their surface. Journal for Parasitology 80,526-532. 205 McAndrew, M. B., Read, M., Sims, P. F. G. and Hyde, J. E. (1993) Characterisation of the gene encoding an unususally divergent TATA- Binding Protein (TBP) from the extremely A+T rich human malaria parasite Plasmodium falciparum. Gene 124,165-171. McBride, J. S. and Heidrich, H. G. (1987) Fragments of the Polymorphic Mr 185000 glycoprotein from the surface of isolated Plasmodium falciparum merozoites from an antigenic complex. Molecular and Biochemical Parasitology 23,71-84. McCutchan, T. F., Lal, A. A., do Rosario, V. and Waters, A. P. (1992) Two types of sequence polymorphism in the circumsporosoite gene of Plasmodium falciparum. Molecular and Biochemical Parasitology 50,37- 46. McGhee, R. B. (1953) The infection by Plasmodium lophurae of duck erythrocytes in the chick embryo. Journal of Experimental Medicine 97, 773-779. McHardy, N. (1978) In vitro studies on the action of menoctone and other compounds on Theileria parva and Theileria annulata. Annals of Tropical Medicine and Parasitology 72,501-511. McHardy, N., Wekesa, L. S., Hudson, A. T. and Randall, A. W. (1985) Antitheilerial activity of BW720C (buparvaquone): a comparison with parvaquonc. Research in Veterinary Sciences 39,29-33. McHardy, N., (1991) Butalex (Buparvaquone) -a new therapeutic for theileriosis: with comments on its value in an integrated control strategy. In "Proceedings of the Second EEC Workshop on Orientation and Co-ordination of Research on Tropical Theileriosis". Eds. D. K. Singh and B. C. Varshney. Anand Press, Anand. pp 133-140 McKean, P. G., O'Dea, K. and Brown, K. N. (1993) A single amino acid determines the specificity of a monoclonal antibody which inhibits Plasmodium chaudi AS in vivo. Molecular and Biochemical Parasitology 62,211-222. 206 Mecham, R. P., Hinek, A., Entwistle, R., Wrenn, D. S., Griffin, G. L. and Senior, R. M. (1989) Elastin binds to a multifunctional 67 kilodalton membrane protein. Biochemistry 28,3716-3722. Melhorn, H. and Schein, E. (1984) The Piroplasm: Life cycle and sexual stages. Advances in Parasitology 23,37-103. Melrose, T. M. Brown, C. G. D., Morzaria, S. P., Ocama, J. G. R. and Irvin, A. D. (1984) Glucose Phosphate Isomerase polymorphism in Theileria annulata and T. parva. Tropical Animal Health Production 16,239-245. Mendis, K. N., David, P. H., Hommel, M., Carter, R. and Miller, L. H. (1983) Immunity to malaria antigens on the surface of Plasmodium falciparum infected erythrocytes. American Journal of Tropical Medicine and Hygiene 32,926-930. Mendis, K. N., Ihalamulla, R. I. and David, P. H. (1988) Diversity of Plasmodium vivax induced antigens on the surface of infected human erythrocytes. American Journal of Tropical Medicine and Hygiene 38, 42-46. Mendis, K. N., David, P. H. and Carter, R. (1991) Antigenic polymorphism in malaria: is it an important mechanism for immune evasion? Immunology Today 12, A34-A37. Mercier, C., Lecordier, L., Darcy, F., Deslee, D., Murray, A., Tourvieille, B., Maes, P., Capron, A. and Cesbron-Delauw, M. F. (1993) Molecular characterization of a dense granuole antigen (GRA2) associated with the network of the parasitophorous vacuole in Toxoplasma gondii. Molecular and Biochemical Parasitology 58,71-82. Mermond, N. 0., Kelly. T. 7. and Tjian, R. (1989) The prolin-rich transcriptional activator of CTF/NF-1 is distinct from the replication and DNA binding domain. Cell 58,741-753. 207 Mevelec, M. -N., Chardes, T., Mercereau-Puijalon, 0., Bourguin, I., Achbarou, A., Dubremetz, J: F. and Bout, D. (1992) Molecular cloning of GRA4, a Toxoplasma gondii dense granule protein, recognized by mucosal IgA antibodies. Moelcular and Biochemical Parasitology 56,227-238. Miller, L. H., Mason, S. J., Dvorak, J. A., McGinnis, M. H. and Rothman, I. K. (1975) Erythrocyte receptors for (Plasmodium knowlesi) malaria: Duffy blood group determinants. Science 189,561-563. Miller, L. H., Aikawa, M., Johnson, J. G. and Shiroishi, T. (1979) Interaction between cytochalasin B-treated malaria parasite and erythrocytes. Attachment and junction formation. Journal of Experimental Medicine 149,172-184. Miller, L. H., Roberts, T., Shahauddin, M. and McCutchan, T. (1993) Analysis of sequence diversity in the Plasmodium falciparum merozoite surface protein-1 (MSP-1). Molecular and Biochemical Parasitology 59,1- 14. Minami, T., Spooner, P. R., Irvin, A. D., Ocama, J. G. R., Dobbelaere, D. A. E. and Fujinaga, T. (1983) Characterisation of stocks of Theileria parva by monoclonal antibody profiles. Research in Veterinary Science 35,334- 340. Mineo, J. R., McLeod, R., Mack, D., Smith, J., Khan, I. A., Ely, K. H. and Kasper, L. H. (1993) Antibodies to Toxoplasma gondii major surface protein (SAG-1. P30) inhibit infection of host cells and are produced in murine intestine after peroral infection. Journal of Immunology 150, 3951-3964. Miller, L. H., Good, M. F. and Milon, G. (1994) Malaria pathogenesis. Science 264,1878-1883. Mitchell, G. H., Hadley, T. J., Klotz, F. W., McGinnis, M. H. and Miller, L. H. (1986) Invasion of erythrocytes by Plasmodium falciparum malaria parasites: evidence for receptor heterogeniety and two receptors. Blood 67,1519-1521. 208 Moelans, I. I., Meis, J. F., Kocken, C., Konings, R. N. and Schoenmakers, J. G. (1991) A novel protein antigen of the malaria parasite Plasmodium falciparum, located on the surface of gametes and sporozoites. Molecular and Biochemical Parasitology 45,193-204. Morisaki, J. H., Heuser, J. E. and Sibley, L. D. (1995) Invasion of Toxoplasma gondll occurs by active penetration of the host cell. Journal of Cell Science 108,2457-2464. Morrison, W. I., Lalor, P. A., Goddeeris, B. M. and Teale, A. J. (1986) Theileriosis: Antigens and host-parasite interactions. In: Parasite antigens: Towards new stratagies for vaccines (ed. by T. W. Pearson) 167- 213, Marcel Dekker, New York. Morrison, W. I., Goddceris, B. M., Brown, W. C., Baldwin, C. L. and Teale, A. J. (1989a) Thelleria parva in cattle: Characterization of infected lymphocytes and the immune responses they provoke. Veterinary Immunology and Immunopathology 20,213-237. Morrison, W. I. and Goddeeris B. M. (1989b) The role of the MHC in the immunity of cattle to the protozoan parasite 'Theileria parva. Ann. Rech. Vet. 20,385-588. Mozaria, S. P., Spooner, P. R., Bishop, R. P., Musoke, A. J. and Young J. R. (1990) SC I and Not I polymorphisms in Theileria stocks detected by pulse field gel elecrophoresis. Molecular and Biochemical Parasitology 40,203.212. Mozaria, S. P. and Nene, V. (1990) Bovine theileriosis: progress in immunisation methods. International Journal on Animal Science 5,1-14. Muller, H. M., Reckmann, L. Hollingdale, M. R., Bujard, H. and Robson, K. J. H. (1993) Thrombospondin related anonymous protein (TRAP) of Plasmodium falciparum binds specifically to sulfated glycoconjugAtes and to HEPG2 hepatoma-cells suggesting a role for this molecule in sporozoite invasion of hepatocytes. EMBO Journal 12,2881-2889. 209 Murre, C., McCaw, P. S. and Baltimore, D. (1989) A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD and myc proteins. Cell 56,777-783. Musisi, F. L. (1990) Methods currently used for the control of East Coast fever: their validity and proposals for future control strategies. Parasitologia 32,15-22. Musoke, A. J., Nantulya, V. M., Buscher, G., Masake, R. A. and Otim, B. (1982) Bovine immune responses to Theileria parva : neutralizing antibodies to sporozoites. Immunology 45,663.668. Musoke, A. J., Nantulya, V. M., Rurangirwa, F. R. and Buscher, G. (1984) Evidence for a common protective antigen determinant on sporozoites of several Theileria parva strains. Immunology 52,231-238. Musoke, A., Moraria, S., Nkonge, C., Jones, E. and Nene, V. (1992) A recombinant sporozoite surface antigen of Theileria parva induces protection in cattle. Proceedings of the National Academy of Sciences, U. S. A. 89,514-518. Musoke, A., Nene, V. and Morzaria, S. P. (1993) A sporozoite-based vaccine for Theileria parva. Parasitology Today 9,385-388. Myler, P. J. (1989) Transcriptional analysis of the major merozoite surface antigen precursor (GP195) gene of Plasmodium falciparum. In "Parasites: Molecular Biology, Drugs, Vaccine Design" (N. Agabian and A Cerami, Eds. ), pp. 123-137, Wiley-Liss, New York. Nardelli, J., Gibson, T. J., Vesque, C. and Charnay, P. (1991) Base spequence discrimination by zinc-finger DNA binding domains. Nature 349,175-178. Ndengele, M. M., Messineo, D. G., Samyellowe, T. and Harwalkar, J. A. (1995) Plasmodlum falciparum - effects of membrane modulating agents on direct binding of rhoptry proteins to human erythrocytes. Experimental Parasitology 81,191-201. 210 Neitz, W. 0. (1957) Theileriosis, gonderioses and cytauzzoonoses. The Onderstepoort Journal of Veterinary Research 27,275-430. Nene, V., Gobright, E., Musoke, A. J. and Lonsdale-Eccles, J. D (1990) A single exon codes for the enzyme domain of a protozoan cystein protease. Journal of Biological Chemistry 265,18047-18050. Nene, V., lams, K. P., Gobright, E. and Musoke, A. J. (1992) Characterisation of a gene encoding a candidate vaccine antigen of Theileria parva sporozoites. Molecular and Biochemical Parasitology 51,17-28. Neuberg, M., Adamkiewicz, J., Hunter, J. B. and Muller, R. (1989) A Fos protein containing the Jun leucine zipper forms a homeodimer which binds to the API binding site. Nature 341,243-245. Norman, C., Runswick, M., Pollock, R. and Treisman, R. (1988) Isolation and properties of cDNA clones encoding SRF, a transcription factor that binds to the c-fos serum response element. Cell 55,989-1003. Nussenzweig, V. and Nussenzwcig, R. S. (1989) Rationale for the development of an engineered sporozoite malaria vaccine. Advances in Immunology 45,283-224. Odea, K. P., McKean, P. G., Harris, A. and Brown, K. N. (1995) Processing of the Plasmodium chabaudi chabaudi asmerozoite surface protein-1 in vitro and in vivo. Molecular and Biochemical Parasitology 72,111-119. Orlandi, P. A., Sim, B. K., Chulay, J. D. and Haynes, J. D. (1990) Characterization of the 175 kDa erythrocyte binding antigen of Plasmodium falciparum. Molecular and Biochemical Parasitology 40,285- 294. Orlandi, P. A., Klotz, F. W. and Haynes, J. D. (1992) A malaria invasion receptor, the 175 kDa erythrocyte binding antigen of Plasmodium falciparum recognizes the terminal Neu5Ac(alpha2-3)Gal-sequence of glycophorin A. Journal of Cell Biology 116,901-909. 211 Ossorio, P. N., Schwartzman, J. D. and Boothroyd, J. C. (1992) A Toxoplasma gondil rhoptry protein associated with host cell penetration has unusual charge asymmetry. Molecular and Biochemical Parasitology 50,1-16. Ossorio, P. N., Dubremetz, J. F. and Joiner, K. A. (1994) A soluble secretory protein of the intracellular parasite Toxoplasma gondii associates with the parasitophorous vacuole membrane through hydrophobic interactions. Journal of Biological Chemistry 269,15350-15357. Pancake, S. L. Holt, G. D., Mellouk, S. and Hoffman, S. L. (Malaria sporozoite and circumsporozoite proteins bind specifically to sulfated glycoconjugates. Journal of Cell Biology 117,1351-1357. Parsons, M. (1990) Current concepts in stage-regulated gene expression in kinetoplastida. Parasitology Today 6,244-245. Pays, E., Tebabi, A., Pays, A., Coquelet, H., Revelard, P., Salmon, D. and Steinert, M. (1989) Trypanosoma brucei : post-transcriptional control of the variable surface glycoprotein gene expression site. Cell 57,835-845. Perkins, M. (1988) Stage-dependent processing and localization of a Plas