Genetic variation of the major surface protein 3 in anaplasma marginale

Material Information

Genetic variation of the major surface protein 3 in anaplasma marginale
Meeus, Patrick Frans
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:


Subjects / Keywords:
Anaplasma marginale ( jstor )
Cattle ( jstor )
Genomes ( jstor )
Infections ( jstor )
Membrane proteins ( jstor )
Open reading frames ( jstor )
Operator regions ( jstor )
Pathogens ( jstor )
Pseudogenes ( jstor )
Ticks ( jstor )
Dissertations, Academic -- Veterinary Medicine -- UF ( lcsh )
Veterinary Medicine thesis, Ph. D ( lcsh )
government publication (state, provincial, terriorial, dependent) ( marcgt )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


ABSTRACT: The family Anaplasmataceae, which includes the genera Anaplasma and Ehrlichia, contains small, gram-negative, obligate intracellular bacteria that cause important tick-transmitted diseases of humans and animals. All typically require a sufficient number of long-lasting circulating organisms in the host to ensure infection of ticks. A common and important mechanism to achieve this is antigenic variation of the proteins recognized by the immune system. This powerful mechanism for generating phenotypic diversity, the focus of this dissertation, is not only an efficient strategy for adapting to rapidly responding immune system defenses, but it also allows for adaptation to a broad range of different environments, hosts and host tissues. Bovine anaplasmosis is a disease of cattle and other ruminants caused by Anaplasma marginale and is of major economic importance to livestock production throughout many areas of the world. A. marginale establishes a persistent infection characterized by sequential cycles of rickettsemia in which new antigenic variants emerge. The two immunodominant outer membrane proteins, major surface protein (MSP)2 and MSP3, are both structurally and antigenically polymorphic. This study, using a combination of expression and genomic analysis, explores the mechanisms by which this polymorphism is generated in MSP3. The results indicate that MSP3 is expressed from a single locus and that variation in the expressed msp3 gene is generated by recombination utilizing msp3 pseudogenes. Each of the msp3 pseudogenes encodes a unique central variable region (CVR) flanked by conserved 5 and 3 regions. Changes in the CVR of the expressed msp3, concomitant with invariance of the pseudogenes, indicate that expression-site variation is generated using gene conversion. This mechanism of msp3 gene conversion is similar to that reported for msp2 variation but each gene family utilizes distinct sets of pseudogenes. Genomic analysis of the two gene families reveals that they originated from a common ancestor. Despite having diverged substantially, both families have maintained similar structural features, an identical C-terminal end in the expressed copies and a close linkage within the genome. The overall organization of the two gene families is also closely linked to a set of recombinase gene homologs and a number of short and long repeats, suggesting a role in recombination. Further examination of these structures and gene products will allow us to better understand the mechanisms involved in antigenic variation and to design novel control measures.
Thesis (Ph. D.)--University of Florida, 2002.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
General Note:
Title from title page of source document.
General Note:
Includes vita.
Statement of Responsibility:
by Patrick Frans Meeus.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Meeus, Patrick Frans. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
Resource Identifier:
029834822 ( ALEPH )
83059829 ( OCLC )


This item is only available as the following downloads:

Full Text




ii ACKNOWLEDGEMENTS “If you wish to succeed in life, make perseverance your bosom friend, experience your wise counselor, cauti on your elder brother, and hope your guardian genius.” Joseph Addison Since graduating as a Doctor in Veterinary Medicine many people have inspired me to do research, each in hi s/her own peculiar way. Prof. Jozef Vercruysse ingrained the importance of publishing in me and showed me the way to Africa. Africa and the late Pro f. Peter Nansen showed me that research can have a huge positive impact on people’s lives and bridge many divides. My good friend Dr. Jan De Bont taught me that a research career, although sometimes precarious, can be meaningful and satisfying. The chair of my PhD committee and mentor, Dr. Tony Barbet, se t an example of research integrity and thoroughness that will stay with me for t he rest of my career. I hope I can live up to his 95% certainty rule when publishing. I am extremely grateful to all the members of my PhD advisory committee: Dr. Kathy Kocan for entrenc hing in me the word “perseverance,” which sounds an awful lot better than stubborn; Dr. Michael Burridge for keeping me in touch with the clinical imp lications of the diseases we work with and for diligently insuring that the practi cal aspects of being a graduate student were taken care of; Dr. Davi d Allred and Dr. Wayne McCormack for contributing their vast knowledge of the subject through many useful


iii suggestions and comments and finally Dr. Bruno Goddeeris for contributing his expertise on tick-borne disease and bovi ne immunology while giving me the opportunity to stay connected with my Belgian research roots. I am very grateful to Dr. Michael Bo wie, Debbie Couch, Anna Lundgren, Annie Moreland, Sally O’Connell, Carlos Sulsona, Jooyoung “J” Yi and Hyun Yi for their friendship and techni cal support. I am also grat eful to Dr. Ellis Greiner for being such a wonderful friend and st udent advocate, and to Dr. John Dame and Dr. Charles Courtney for their advice and financial support. I would like to thank my parents for giving me the best education a person can wish for and for cautiously encouraging me in my slightly off the beaten path endeavors. I would like to t hank my brothers and sisters and their family for always being there for me despite the sometimes considerable physical distances that separ ate us. Last but foremost I would like to thank my wife, Dr. Merijo Jordan, who lived every bit as much through this project as I did. I could not have hoped for a better friend and soul mate.


iv TABLE OF CONTENTS page ACKNOWLEDGEMENTS ...................................................................................ii TABLE OF CONTENTS ....................................................................................iv LIST OF FIGURES ABBREVIATIONS .............................................................................................vii ABSTRACT .......................................................................................................ix CHAPTER 1 INTRODUCTION .....................................................................................1 Justification ..............................................................................................1 Research Objectives ................................................................................3 2 LITERATURE REVIEW ...........................................................................4 Anaplasma marginale ..............................................................................4 Immunity to Anaplasma spp . ..................................................................10 Pathogenesis of Anaplasmosis ..............................................................20 Control of Anaplasmosis ........................................................................23 Cultures of A. marginale ........................................................................27 Antigenic Variation of Anaplasmataceae ...............................................28 3 TICK CELL CULTURES AND CLONING ...............................................45 Introduction ............................................................................................45 Experimental Procedures .......................................................................47 Results ...................................................................................................51 Discussion .............................................................................................55


v 4 MSP3 EXPRESSION AND GENETIC VARIATION ...............................59 Introduction ............................................................................................59 Experimental Procedures .......................................................................62 Results ...................................................................................................70 Discussion .............................................................................................81 5 ANALYSIS OF THE GENOMIC STRUCTURE OF THE MSP3 AND RELATED GENE FAMILIES ..................................................................86 Introduction ............................................................................................86 Experimental Procedure ........................................................................88 Results ...................................................................................................90 Discussion ...........................................................................................102 6 5Â’RACE ANALYSIS OF EXPRESS ION IN DIFFERENT SPECIES OF ANAPLASMATACEAE .........................................................................116 Introduction ..........................................................................................116 Experimental Procedure ......................................................................118 Results .................................................................................................122 Discussion ...........................................................................................128 7 DISCUSSION AND CONCLUSIONS ...................................................132 LIST OF REFERENCES ................................................................................135 BIOGRAPHICAL SKETCH .............................................................................160


vi LIST OF FIGURES Figure page 1 Visualization of A. marginale colonies within infected tick cells. ................52 2 Visualization of individual A. marginale colonies within tick cells. ..............53 3 Variability in msp2E of A. marginale in a population of infected cells or in single cells. ................................................................................................54 4 A polymorphic msp3 expression site. ........................................................71 5 Presence of multiple copies of msp3 , but a single hybridizing band from the msp3 expression site in genomic A. marginale DNA .................................73 6 Schematic representation of the msp3 family of genes and its relationship to msp2 . .....................................................................................................74 7 The generation of diversity in msp3 expression is achieved by insertion of different complete gene copies or segments into an expression site ........77 8 Comparison of the msp3 variable regions from tw o geographically distinct isolates of A. marginale (Florida and St. Maries). ......................................80 9 msp2 and related ORFs. ...........................................................................91 10 Orientation of MSP ex pression sites and pseudogenes. ...........................92 11 BAC G11 repeats and recombinase genes. ..............................................93 12 BAC E6 repeats and recombinase genes. .................................................94 13 msp3 pseudogenes share many features in their coding and flanking regions. ......................................................................................................97


vii 14 Recombination hot-spots surrounding msp3 genes. ................................100 15 Structure and variability of msp2(p44) expression in A. phagocytophilum . ................................................................................................................125 16 Structure and variability of p28 expression in E. chaffeensis . ..................126 17 Structure and variability of msp3 expression in A. marginale . .................127


viii ABBREVIATIONS 5Â’RACE rapid amplific ation of cDNA ends APC antigen presenting cell B bursa BAC bacterial artificial chromosome bp base pair CD cluster of differentiation cDNA complementary deoxyribonucleic acid CVR Central variable region DNA deoxyribonucleic acid GCG genetics Computer group IFN interferon Ig immunoglobulin IL interleukin Kb kilo base pair kDa kilodalton M molar MACAW multiple alignment construction & analysis workbench MAP major antigenic protein Mb mega base pair MHC major histocompatibility complex MSP major surface protein msp gene encoding major surface protein msp2E expressed copy of msp2 genes msp3E expressed copy of msp3 genes msp2P msp2 pseudogene msp3P msp3 pseudogene NCBI national center for biotechnology information PCR polymerase chain reaction OMP outer membrane protein opag operon associated gene ORF open reading frame RBC red blood cells RNA ribonucleic acid RFLP restriction fragm ent length polymorphism RT reverse transcriptase SDS sodium dodecyl sulfate SSC standard sodium citrate T thymus Th T helper U units


ix Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for t he Degree of Doctor of Philosophy GENETIC VARIATION OF THE MAJOR SURFACE PROTEIN 3 IN ANAPLASMA MARGINALE By Patrick F.M. Meeus December, 2002 Chair: Anthony F. Barbet Major Department: Veterinary Medicine The family Anaplasmataceae, which includes the genera Anaplasma and Ehrlichia , contains small, gram-negative, obl igate intracellular bacteria that cause important tick-transmi tted diseases of humans and animals. All typically require a sufficient number of long-lasti ng circulating organisms in the host to ensure infection of ticks. A common and important mechanism to achieve this is antigenic variation of the proteins recognized by the immune system. This powerful mechanism for generating phenotypic diversity, the focus of this dissertation, is not only an efficient stra tegy for adapting to rapidly responding immune system defenses, but it also al lows for adaptation to a broad range of different environments, hosts and host tissues. Bovine anaplasmosis is a disease of cattle and other ruminants caused by Anaplasma marginale and is of major economic importance to livestock


x production throughout many ar eas of the world. A. marginale establishes a persistent infection characterized by sequential cycles of rickettsemia in which new antigenic variants emerge. The two immunodominant outer membrane proteins, major surface protein (MSP) 2 and MSP3, are both structurally and antigenically polymorphic. This study, using a combination of expression and genomic analysis, explores the mechanisms by which this polymorphism is generated in MSP3. The results indicate t hat MSP3 is expressed from a single locus and that variation in the expressed msp3 gene is generated by recombination utilizing msp3 pseudogenes. Each of the msp3 pseudogenes encodes a unique central variable region (C VR) flanked by conserved 5Â’ and 3Â’ regions. Changes in the CVR of the expressed msp3 , concomitant with invariance of the pseudogenes, indicate t hat expression-site variation is generated using gene conversion. This mechanism of msp3 gene conversion is similar to that reported for msp2 variation but each gene family utilizes distinct sets of pseudogenes. Genomic analysis of the two gene families reveals that they originated from a common ance stor. Despite having diverged substantially, both families have mainta ined similar structural features, an identical C-terminal end in the express ed copies and a close linkage within the genome. The overall organization of the two gene families is also closely linked to a set of recombinase gene homol ogs and a number of short and long repeats, suggesting a role in recombinat ion. Further examination of these structures and gene products will allow us to better understand the mechanisms involved in antigenic variation and to design novel control measures.


1 CHAPTER 1 INTRODUCTION Justification Anaplasmosis is an arthropod-bor ne disease of cattle and other ruminants caused by an intraerythr ocytic rickettsia of the genus Anaplasma . The pathogen is known to occur in Africa , Asia, Australia, southern Europe, South America and the former Soviet Union and is believed to be endemic in major areas of the United States. The losses attributable to bovine anaplasmosis, caused by Anaplasma marginale , have been estimated from $ 511 million in California alone (Goodger, Ca rpenter, and Riemann, 1979) to over $ 300 million in the continental United St ates (Palmer, 1989). While ticks have been described as vectors of human diseases for over a hundred years, they only reached notoriety in 1982 with the emergence of Borrelia burgdorferi as the etiologic agent of Lyme disease (Burgdorfer et al. , 1982). Since then, members of the family Anaplasmataceae ( Ehrlichieae ), previously thought to be only animal pathogens, have joined the incr easingly long list of human pathogens transmitted by ticks. Biologically, members of the genus Anaplasma ( A. marginal e, A. (Ehrlichia) platy s, A. (Ehrlichia) bovis and A. phagocytophilum ) are closely related. In their respec tive mammalian hosts these pathogens are most often detected in cells derived from bone marrow precursors, and ticks ensure transmission between hosts. A. marginale is maintained in a host population by infect ed tick vectors (Ge et al. , 1996) and persistent subclinical


2 infection of ruminants, including wild animals such as deer (Woldehiwet, 1983; Kuttler, 1984), and similar persistent infections with A. phagocytophilum , a human pathogen, have become established in ruminants (Woldehiwet, 1983; Siebinga and Jongejan, 2000). Species within the Anaplasma genus are characterized by several immunodominant outer membrane proteins. These antigens share significant homologies at the amino acid and nucleotide leve l and appear to be encoded by paralogous and orthologous genes. A 36 kDa antigen called major surface protein 2 (MSP2) (Palmer and McGuire, 1984; Palmer et al. , 1994a), a 86 kDa antigen MSP3 (Palmer and McGuire, 1984;Palmer et al. , 1986b;Alleman et al. , 1997), and a 31 kDa MSP4 (Oberle and Barbet, 1993;Oberle, Palmer, and Barbet, 1993) were described in A. marginale as well as a 42-45 kDa antigen called p44 in A. phagocytophilum (Dumler et al. , 1995;Zhi et al. , 1998;Zhi, Ohashi, and Rikihisa, 1999). This complex of outer membrane proteins is also encoded in E. chaffeensis , E. canis , E. (Cowdria) ruminantium and, potentially, in other Ehrlichia species (Reddy et al. , 1998). They further share a common feature, with the exception of MSP4, in that they are all encoded by polymorphic multigene families suspected to contri bute to immune evasion and persistence in reservoir hosts (Alleman et al. , 1997; Reddy et al. , 1998; French et al. , 1998; Reddy and Streck, 1999; Barbet et al. , 2000; Barbet et al. , 2001a). The ability to develop persistent infe ctions may be caused by antigenic variation of outer membrane proteins . Indeed, MSP2 and MSP3 are both polymorphic and migrate as a series of spots on 2-dimensional gel


3 electrophoresis, MSP2 at ~36,000 kD a and MSP3 at ~80,000 – 90,000 (Palmer and McGuire, 1984; Alleman and Barbe t, 1996). Unlike MSP2, there is no information on the mechanisms of MSP3 expr ession or the possible role of this multigene family in generation of diversity. Research Objectives The overall goal of this dissertation was to determine the involvement of MSP3 in antigenic variation and to gai n a better understanding of the underlying processes. My hypothesis is that MSP3 diversity, like MSP2, is generated from a single genomic expression site in wh ich polymorphism is generated through gene conversion. This mechanism is conserved in both the genera Anaplasma and Ehrlichia . Understanding the mechanism s generating diversity in A. marginale will create a better understanding of the pathoge nicity of all Anaplasmataceae and will allow us to develop effect ive control strategies. The specific aims to achieve this were the following : 1) To develop methods to clone A. marginale in tick cell cultures. 2) To define the structure of expressed msp3 transcripts and the MSP3 expression site(s). 3) To evaluate the mechanism by which genetic variation is generated during infection in cattle. 4) To analyze the genomic structures of the msp2 and msp3 gene families and evaluate their roles in cr eating diversity from a small genome. 5) To analyze the expression of msp2 homologs in the family Anaplasmataceae.


4 CHAPTER 2 LITERATURE REVIEW Anaplasma marginale The organism Anaplasma marginale , identified by Theiler in South Africa in 1909, was first described as "marginal po ints" in bovine erythrocytes (Theiler, 1910). The scientific name is based on its staining characteristics and location within the host cell, with " Anaplasma " referring to the lack of stained cytoplasm and " marginale " denoting the peripheral location of the organism in the host erythrocyte. Anaplasma marginale has been reclassified numerous times since its first description. It is now clas sified as a gram-negative proteobacterium, subdivision class, in the order of the Rickettsiales and family of the Anaplasmataceae (Dumler et al. , 2001). The family includes the genera Anaplasma , Ehrlichia , Neorickettsia and Wolbachia , and encompasses a group of obligate intracellular bacteria that reside in eukaryotic cells. All members reside, unlike the rickettsial organisms, wit hin a cytoplasmic vacuole inside host cells that include erythrocytes, re ticuloendothelial cells, bone marrow-derived phagocytic cells, endothelial cells and cells of helminth and arthropod reproductive tissues (Dumler et al. , 2001). Biologically, members of the genus Anaplasma ( A. marginale , A. (Ehrlichia) platys , A. (Ehrlichia) bovis and A. phagocytophilum ) are closely related. In th eir respective mammalian hosts these pathogens are most often detect ed in cells derived from bone marrow precursors, and ticks ensure transmission between hosts. For A. marginale , the


5 only site of development that has been de scribed within cattle is the erythrocyte (Kocan, 1992). However, recent in vitro studies indicate that development of the pathogen might additionally involve endothelial cells. Munderloh and coworkers (2002) have been able to continuously propagate A. marginale and A. phagocytophilum in bovine and primate vascu lar endothelial cell lines, respectively. If confirmed in vivo , then the life cycle of Anaplasma would more closely parallel the development of E. ruminantium , another cattle pathogen, for which initial development in macrophages or neutrophils has been described, but typical development occurs in endot helial cells of blood vessels (Prozesky and Du Plessis, 1987; Logan et al. , 1987). Anaplasma marginale enters erythrocytes by endocytosis after which it divides by binary fission (Friedhoff and Ristic, 1966; Simpson, Kling, and Love, 1967). A similar active mode of entry has been described for E. ruminantium , entering endothelial cells through a process resembling phagocytosis (Prozesky and Du Plessis, 1987) The membrane vacuole in which A. marginale resides is derived from the erythrocyte membrane and c ontains 4-8 pathogens (Kocan et al. , 1978; Francis, Kinden, and Buening, 1979). Anaplasma marginale, apparently in contrast to the exit from endothelial cells by E. ruminantium , leaves the host cells again without disrupti ng the erythrocyte (Franc is, Kinden, and Buening, 1979; Prozesky and Du Plessis, 1987). Nevertheless clinical disease is associated with anemia, caused at least in part by removal of erythrocytes from the circulation (Kuttler, 1984). Bovine anaplasmosis remains a major disease endemic in many regions of the United States of Americ a, as well as worldwide,


6 and is an impediment to meat, milk , and fiber production in tropical and subtropical areas (McCallon, 1973). Dr amatic weight loss, abortion, and death often occur during the acute phase (A lderink and Dietrich, 1981). Animals surviving this acute phase develop lifelon g persistent infections and serve as reservoirs for mechanical and biological transmission of A. marginale to new susceptible hosts (McGuire et al. , 1991; Eriks, Stiller, and Palmer, 1993). This persistent infection is characterized by sequential, often microscopically undetectable cycles of rickettsemia that rise to levels of 10 7 bacteria / ml of blood followed by a rapid decline to <10 3 bacteria / ml of blood (French et al. , 1998) when bacteremia is controlled by a specific immune response (Kieser, Eriks, and Palmer, 1990; Frenc h, Brown, and Palmer, 1999). Under natural conditions A. marginale is transmitted to susceptible animals both mechanically and biologically from carrier cattle and a range of wild ruminants (Kocan, 1995). In the United States important wildlife reservoirs are black-tailed deer ( Odocoileus hemion us columbianus ) and mule deer ( Odocoileus hemionus hemionus ). White-tailed deer ( Odocoileus virginianus ) in the eastern United States are susceptible to A. marginale , but they are not thought to be a major reservoir of infect ion (Kuttler, 1979). Some 20 species of ticks were shown to be capable of biol ogically transmitting anaplasmosis. Dermacentor andersoni , D. occidentalis and D. variabilis have been identified as the main tick vectors in the United States (Dikman, 1950; Anthony and Roby, 1966; Ewing, 1981; Kocan et al. , 1981; Stiller et al. , 1989). Ticks become infected when feeding on infected hosts after which extensive and complex


7 multiplication occurs in several tick tissues (Kocan, Hair, and Ewing, 1980; Kocan et al. , 1983; 1988; 1990a; 1990b; 1993; Ge et al. , 1996). Ticks can also acquire infection after feeding on persist ently infected cattle with low bacteremia although the percentage of infected ticks is directly related to the bacteremia during feeding. Conversely, once ticks acquire infection, the biological replication of the organism within the ticks makes up for initial differences in the infecting dose (Eriks, Stiller, and Palmer, 1993). In A. marginale , like in A. ovis and E. ruminantium , the developmental sequence invo lves multiplication in the gut epithelium of the tick, followed by its presence in the gut muscle and hemolymph, and finally development in the salivary glands (Kocan and Bezuidenhout, 1987; Kocan, Bezu idenhout, and Hart, 1987; Kocan et al. , 1987; Kocan, 1992; Kocan et al. , 1992b; Kocan and Stiller, 1992; Ge et al. , 1996). At each site of infection in ticks, A. marginale develops within membrane-bound vacuoles. The first form of A. marginale seen within these vacuoles is the reticulated (vegetative) fo rm that divides by binar y fission, forming large colonies that may contain hundreds of organisms. The reticulated form then changes into the dense form, which is the infective form and can survive outside of cells. Cattle become infected with A. marginale when the dense form is transmitted during tick feeding via t he salivary glands (Kocan, Hair, and Ewing, 1980; Ge et al. , 1996; Blouin and Kocan, 1998). Transmission to susceptible hosts can occur transstadial ly and intrastadially, and transovarial transmission has been described but does not appear efficient (Anthony and Roby, 1966; Bram and Roby, 1970; Kocan et al. , 1981; Stich et al ., 1989;


8 Kocan, 1995). Intrastadial transmission by male ticks occurs and is believed to be an important epidemiologica l factor. Male ticks rema in infected for over 6 months and can serve as a rese rvoir transmitting the pathogen during sequential feeding on multiple susceptible animals (Kocan et al. , 1992a; Eriks, Stiller, and Palmer, 1993). A similarly important role has been described for male Amblyomma hebraeum ticks in the epidemiology of E. ruminantium (Andrew and Norval, 1989). Mechanical transmission of A. marginale can occur via any bloodcontaminated fomites, including cont aminated needles, dehorning saws, nose tongs, tattooing instrument s, ear-tagging devices, and castration instruments (Reeves, III and Swift, 1977). Additionally, in utero transmission during acute infection has been described (Norton, Parker and Forbes-Faulkner, 1983; Zaugg and Kuttler, 1984; Zaugg, 1985). The efficiency of in utero transmission varies greatly, but transmission rate s can be as high as 15.6% (Zaugg and Kuttler, 1984; Potgieter and Van Rensburg, 1987). The role of transplacental transmission of anaplasmosis, therefore, may contribute to the epidemiology of this disease in some regions. Mec hanical transmission by arthropods is important and has been demonstrated or is suspected in horse flies ( Tabanus spp . ), stable flies ( Stomoxys calcitrans ), deer flies ( Chrysops spp . ), horn flies ( Haematobia irritans ) and mosquitoes ( Culicidae ) (Ristic, 1968). There is no evidence of infection or a developmental cycle in these vectors. Horse flies seem especially important in the epidem iology of the disease, presumably because of their rasping bite that can transfer blood together with their frequent


9 change of hosts when distur bed. The period post-infection in which horse flies can transmit the pathogen is as long as 2 hours, and as little as 10 bites are needed for reliable transmission of the pathogen (Hawkins, Love, and Hidalgo, 1982). In the southeaster n United States the primary natural mode of transmission is thought to be mechanica lly through blood-sucking tabanids despite the presence of competent tick vectors (Kuttler, 1979; Morley and Hugh-Jones, 1989a;Morley and Hugh-Jones , 1989c). This coincides with observations of A. marginale isolates that are not tr ansmissible by known vector ticks (Smith et al. , 1986; Wickwire et al. , 1987; de la Fuente et al. , 2001c). A possible explanation for this phenomenon can be found in the absence of adhesion of A. marginale to tick cells in vitro , as observed for the Florida isolate. This is mediated, at least in part, by the adhesion surface protein MSP1a, shown to be different between tick-tr ansmissible and tick non-transmissible isolates (de la Fuente et al. , 2001a). This agrees with in vivo observations that the Florida isolate does not infect D. andersoni (Friedhoff and Ristic, 1966) and that tick-transmissibility is not relat ed to the presence of a tail-like appendage (Smith et al. , 1986). Anaplasma marginale isolates further exhibit morphological differences (Kreier and Ristic, 1972), variability withi n their surface proteins (Barbet et al. , 1983; Oberle et al. , 1988), differences in rest riction endonuclease patterns (Krueger and Buening, 1988; Alleman et al. , 1993) and other genetic markers (Ferreira et al. , 2001), and antigenic differences as illustrated by complementfixing antibody titers (Kuttler and Wi nward, 1984), polyclonal bovine antiserum


10 (Goff and Winward, 1985) and m onoclonal antibodies (McGuire et al. , 1984; Kano et al. , 2002; Goncalves Ruiz et al. , 2002). Importantly these antigenic differences translate into a lack of cross-protection among many isolates (Kuttler, Zaugg, and Johnson, 1984; Tebele and Palmer, 1991; Palmer et al. , 1994b). Immunity to Anaplasma spp. A protective immune response to A. marginale can be elicited using either live pathogens, killed organisms or purified outer membrane proteins, and it appears to involve both antibod y and cell-mediated immune responses (Buening, 1976; Carson, Sells, and Ristic, 1976; Francis, Buening, and Amerault, 1980; Tebele, McGuir e, and Palmer, 1991; Gale et al. , 1996a). The immunity acquired is long-lasting. Ca rrier animals are immune to subsequent re-infection for several years, and this immunity is retained for at least 8 months even after the elimination of t he pathogen by chemotherapy (Roby et al. , 1974; Gale et al. , 1996a; Richey et al. , 1977). Although our knowledge of protective immunity to A. marginale is fragmented and based on few animals and experiments, significant advances are being made in understanding the underlying responses. While reviewing thos e advances it is useful to look at what is known on bovine immune res ponses against intracellular pathogens and the responses against related pat hogens in different animal species. Intracellular Pathogens and Bovine Immunity Anaplasmataceae are obligate intr acellular pathogens and current concepts on immunity suggest the requi rement of a Th1 type response to


11 control these types of infections (reviewed by Ismail et al. , 2002). Antigen presenting cells (APC) of the innate immune system ar e extremely important in the induction and polarization of the adaptive immune response besides their roles in early defenses. Of the diff erent types of APCs-B cells, macrophages and dendritic cells (DC)-only DCs can st imulate naïve CD4 T cells since only they have the necessary co-stimulatory factors (Janeway, Jr., 2001a). After DCs discriminate between infectious nonself and non-infectious self through a series of Toll like and other receptors, NF B is activated (Ismail et al. , 2002). This leads to transcription of many genes involved in host defenses. Naïve T helper cells are subsequently activated through antigen pres entation and the co-stimulatory molecules CD80 and CD86 (Janeway, Jr., 2001b). How these CD4 Th cells evolve towards producing ei ther a Th1 or a Th2 cytokine profile in vivo is complex and not well und erstood. It is influenc ed by (1) the cytokine environment during antigen priming, (2 ) cytokine mediated regulation, (3) antigen dose and affinity for the TCR, and (4) the timing and level of costimulatory signals betw een APC and T cells (Brown, Rice-Ficht, and Estes, 1998; Taylor-Robinson and Phillips, 1998). In mice the activation of the Th1 pat hway in CD4 cells activated by DC results in the synthesis of cytokines that both activate and coordinate the responses to intracellular pathogens. IFNactivates macrophages which permits them to efficiently kill engulfed pathogens, the principal effector action of Th1 cells. IL-2 induces T cell pro liferation and enhances t he release of other cytokines. IL-3 and GM-CSF stimulat e the production of new macrophages by


12 acting on hematopoietic stem cells in the bone marrow (Janeway and Travers, 1997). On the other hand macrophages also act as effector cells by releasing IL-12. IL-12 in turn activates NK cells, which through the production of -IFN are responsible not only for Th1 differentiation of Th cells, but also for IFNresponses by both T cells and macrophages dur ing intracellular infection (Xing, 2000). The second arm of cell-mediated im munity is cytotoxic T cells or CD8 cells. Once inside a cell, pathogens ar e no longer accessible to antibodies or macrophages, and can be eliminated only by the destruction or modification of the infected cells. This role is played by CD8 cells. These cells can be activated by DCs directly or by any APC in the presence of CD4 helper cells. Once activated CD8 cytotoxic cells k ill target cells that display antigenic peptides of cytosolic pathogens bound to MHC class I molecules at the cell surface (Wang et al. , 2001). The Th1-Th2 model provides a useful scaffold for understanding the observed polarized immune responses in mice. However, it is not only an over simplification of a much more comp lex immunoregulatory network, additionally significant differences in other animal species exist. Data suggest that in ruminants the pathways of regulation may include additional subpopulations of and T lymphocytes (Davis and Hamilt on, 1998). Studies with bovine Th cell clones reveal a cytokine-mediated regul ation that differs from that observed in mice. Using cloned cells, Brown et al. (1998) demonstrated that the majority of antigen-specific Th cells are not pol arized and co-express IL-4 (Th2-like) and IFN(Th1-like). Nonetheless, polarized type-1 or type-2 responses were


13 observed in infections when the relati ve levels of cytokines are compared (Brown, Rice-Ficht, and Estes, 1998). T hese results parallel those obtained using real-time PCR to measure bovine IFNand IL-4 gene expression by a population of antigen stim ulated peripheral blood mononuclear cells (Mena et al. , 2002). Overall it appears that the Th1-Th2 paradigm can be applied to ruminants when the cytokines invo lved are analyzed quantitatively. Immune Responses Against A. marginale Anaplasma marginale is distinctive as far as intracellular pathogens go, in that it infects erythrocytes. Erythr ocytes, lacking a nucleus, are incapable of making proteins de novo and hence are unable to present foreign molecules in the context of MHC I molecules needed for the destruction of infected cells by CD8 cytotoxic cells. It is, therefore, not surprising that CD8 cells do not seem to play a major role in immunity against A. marginale or at least have received little consideration. Desp ite the development of A. marginale inside erythrocytes, cell-mediated immune responses and cyt okine profiles have been correlated with immunity in vaccinated and naturally infected cattle (Buening, 1973; Buening, 1976; Brown et al. , 1998a). Although relatively little is known of the pathways inducing protective immunity against A. marginale , results are consistent with a major role for a type-1 response (Gale et al. , 1996a; Brown et al. , 1998a). The central element of this model is the IFNproducing CD4 Th lym phocyte and the activated macrophage (Palmer et al. , 1999). IL-12 may enhance the type-1 cytokine response since it was shown to significantly enhance IFNproduction by lymph


14 node mononuclear and CD4 cells from immunized calves following recall stimulation with A. marginale (Tuo et al. , 2000). Supernatants from lymphocyte and monocyte cultures isolated from infected calves during in vivo control of acute anaplasmosis reduced the proportion of erythroc ytes containing viable A. marginale in vitro , indicating an antibody-independent mechanism of control during acute anaplasmosis with possible di rect or indirect roles for IFNin the elimination of A. marginale (Wyatt et al. , 1996). IFNhas been shown to inhibit growth of E. ruminantium in vitro (Totte et al. , 1993; Totte et al. , 1994; Mahan, Smith, and Byrom, 1994; Barbet et al. , 1994; Mahan et al. , 1996; Totte et al. , 1996). Additionally, possibly in synergy with IL-2, IFNregulates B-cell synthesis of the opsonizing IgG2 subclass (Estes, Closser, and Allen, 1994) and activates the macrophages (Stich et al. , 1998). These macrophages are important in the ultimate remova l of the opsonised pathogens from the circulation. Studies demonstrating t hat immune serum and complement in the presence of mouse peritoneal macr ophages inhibit the infectivity of E. ruminantium , but not in the absence of ma crophages, point out the importance of the activated macrophage in the contro l of these infections (Du Plessis, Berche, and Van Gas, 1991). Macrophages of toll-like receptor 4-deficient mice show a depressed nitric oxide and IL-6 secretion when infected with E. chaffeensis . This reduced macrophage response results in persistent infections for up to 30 days, while wild-type mi ce readily clear the pathogen (Ganta et al. , 2002). Although infecting a different ce ll type and animal species, cytokine studies on the closely related A. phagocytophilum do support the requirement


15 for a type-1 response, and IFN-related mechanisms have been established in the clearance of the early phases of in fection in mice (Akkoyunlu and Fikrig, 2000; Martin, Bunnell, and Dumler, 2000; Martin, Caspersen, and Dumler, 2001; Borjesson et al. , 2002). Splenocytes from in fected mice produced more IFNand less IL-4 than controls, suggesting at least a partially Th1-skewed immune response (Akkoyunlu and Fikrig, 2000). IFN-independent mechanisms have nevertheless been suggested to play a role at later time points (Akkoyunlu and Fikrig, 2000) and mi ght reflect the temporal changes in cytokines profiles observed in other hemoparasites. In mice infected with A. phagocytophilum high IL-10 levels, a Th2 cytokine in mice but not in humans, have been observed throughout infection (Martin, Bunnell, and Dumler, 2000) and an IL-8 up-regulation was observed in infected cells. This IL-8 upregulation seems to play a role more in attracting new neutrophils enhancing the infection, rather than controlling it (Akkoyunlu et al. , 2001). Th1-mediated responses have been implicated in protection against other pathogens infecting erythrocytes; however, initial Th1-responses often evolve towards a Th2 profile during persistent infection. Recovery of mice from acute blood-stage infection with malari a is Th1-response mediated, while clearance of the chronic low-grade infection is Th2 regulated and antibody dependent (Taylor-Robinson and Phillips, 1998). A similar evolution of responses is seen during infection of CBA mice with the purely erythrocytic parasite Babesia microti . Th1 responses (IL-2 and IFN) are predominately activated during the acute phase, wher eas Th2 responses (IL-4 and IL-10)


16 again dominated the recovery phase (Chen et al. , 2000). It is uncertain if a similar dynamic cytokine profile occurs with A. marginale infections, but some data are consistent with a change in imm une response, Th1 to Th2, over time (Buening, 1973; Buening, 1976; Wyatt et al. , 1996). In contrast to responses during natural acute infection, persistent ly infected animals no longer display increased complement-fixing antibodies or response to intradermal antigen testing despite the persistence of infe ction (Buening, 1973; Hallab, Jacobs, and Black, 2000). Conversely a type-1 response has consistently been demonstrated in most experiments c onducted by Brown and colleagues, yet these experiments were not done with natural infections but in calves protected against challenge after immunization with either disrupted outer membranes in saponin (Brown et al. , 1998a;Brown et al. , 1998b), naked DNA using msp1 (Arulkanthan et al. , 1999), or native purified MSP2 in alum with or without IL-12 (Tuo et al. , 2000). It has been shown that the immune response in natural infection and outer membrane vaccination is different even though both can partially protect the cattle (Buening, 1973; Carson, Sells, and Ristic, 1977a). These somewhat conflicting result s are likely the consequence of the different methodologies used, but they pr obably also indicate that the immune response to A. marginale in cattle will be reflected by cytokine kinetics rather than the absolute presence or absence of a given cytokine. Possibly, a shift in the type of response during the course of a natural infection may additionally be at play.


17 A. marginale and Antibodies Immune serum against A. marginale generated in rabbits is able to neutralize infectivity for cattle when added to the pathogens before infection (Palmer and McGuire, 1984). Protection against A. marginale infection has been shown to correlate with antibody tite rs to membrane polyp eptides (Tebele, McGuire, and Palmer, 1991). This adaptive immunity has been shown to involve high titers of opsonizing IgG2 antibody (Brown et al. , 1998a). Other data also suggest a possible role for IgG1 in the control of bovine anaplasmosis (Arulkanthan et al. , 1999; Valdez et al. , 2002). Both bovine IgG1 and IgG2 have been demonstrated to fix complement and could mediate phagocytosis (McGuire, Musoke, and Kurtti , 1979). In contrast, tr ansfer of high levels of A. marginale antibody alone to naïve calves al ters neither the course nor the outcome of infection (Gale et al. , 1992). Moreover, even in the absence of high IgG1 and / or IgG2 A. marginale antibody titers, calves are able to adequately control acute anaplasmosis (Valdez et al. , 2002). Calves borne to immune mothers are not protected by colost ral antibody, and immune carrier animals relapse severely when splenectomized des pite the continued presence of high levels of circulating antibodies (Gale et al. , 1996a). Other researchers found that colostral antibodies and / or other maternal fa ctors lengthen the prepatent period and delay anemia (Zaugg and Kuttler, 1984). Equivalent results were seen in cattle and mice with E. ruminantium where specific antibody responses were detected following recovery from in fection, but serum transfer experiments implied that these were, on their own, of little im portance in protection (Du


18 Plessis, 1970). Different results were obtained for E. chaffeensis , where passive transfer of either immune se rum or antibodies provided long-term protection to susceptible SCID mice . This was true when administration occurred at the time of inoculation and ev en well after intracellular infection was established (Winslow et al. , 2000). Since different antigens and immuniza tion routes were used in the various experiments, it is likely that the observed inconsistencies were due to differences in the experimental procedures as well as differences in the genetic make-up of the experimental animals or animal models. Nevertheless, several studies on other intracellular pathogens have provided evidence that antibodies can protect against intracellular bacteri a, fungi, and protozoa (Casadevall, 1998). The mechanisms of humoral imm unity during intracellular infection might involve affecting the growth of some intracellular pathogens within the host cell, preventing the intercellular transfer and / or su bsequent invasion of cells. Alternatively, the uptake of opsonized pathogens might induce an oxidative burst in phagocytes and promote phagosome lysosome fusion (Li et al. , 2002). It is likely that one or more of these mechanisms are involved in the control of A. marginale infections in the bovine host. Immunogens of A. marginale The characterization of A. marginale immunogens was first reported in 1984 with the comparison of proteins der ived from two different isolates, revealing pathogen proteins with molecula r weights ranging from ~14 to ~200 kDa (Barbet et al. , 1983). This was followed by the characterization of the


19 major surface proteins (MSP), by radi oiodination and detection by rabbit antisera against initial bodies, as pr oducts of 105, 86, 61, 36, and 31 kDa (McGuire et al. , 1984). These MSPs were found to be present in cattle, ticks and, eventually, the cultured fo rms of the pathogen (Palmer et al. , 1985; Barbet et al. , 1999). Monoclonal antibodies revealed that the MSPs from different isolates Israel, Kenya, Zimbabwe and t he United States were antigenically different, but conserved epitopes were found on MSP1 (105 kDa), MSP2 (36 kDa), MSP3 (86 kDa), MSP4 (31 kDa) and MSP5 (19kDa) (Palmer et al. , 1988a; Tebele, McGuire, and Palmer, 1991; Tebele and Palmer, 1991; Visser et al. , 1992). Native purified MSP1, MSP2 and membrane fractions containing all MSPs have been shown to induce some protective immunity against challenge, as measured by a significant delay in onset of rickettsemia and reduction in clinical sympto ms such as anemia (Palmer et al. , 1986a; 1988b; 1989; 1994b; Tebele and Palmer, 1991; T ebele, McGuire, and Palmer, 1991). Sera from animals immunized with the outer membrane fractions and protected against clinical disease recognized t he MSPs (Tebele, McGuire, and Palmer, 1991) and PBMC responded to MSP1, MSP2 and MSP3 several months after the last immunization (Brown et al. , 1998a). Eight out of 12 T-cell clones derived from membrane-immunized and pr otected animals recognized MSP2, MSP3, or both prior to challenge (Brown et al. , 1998b). While it is not surprising that the majority of the immune re sponses in these membrane-vaccinated animals was directed against the MSPs, dat a from carrier animals do support the MSPs as the major immunogens, alt hough other different antigens are also


20 recognized (Palmer et al. , 1986b; McGuire et al. , 1991). Fifteen immunodominant polypeptides were recognized by bovine sera in a Venezuelan isolate of A. marginale , including the MSP previously characterized, but also some new antigens (Leal et al. , 2000). Investigations on the major antigens of related pathogens seem to confirm that th e majority of the immune response is directed against outer membrane proteins. In an extensive analysis of E. ruminantium recombinant antigens recognized by antibody and peripheral blood mononuclear cells from immune ruminants, many new antigens were described, but these ar e still predicted to often encode outer membrane proteins (Barbet et al. , 2001a). In E. chaffeensis , 40 percent of antibodies recovered from three indepen dent hybridoma fusions recognized outer membranes. Antibodies that recogni zed proteins other than MSPs were also recovered, but for t he most part were low a ffinity IgM of undetermined specificity (Li et al. , 2002). Therefore, major surf ace protein recognition by both antibodies and lymphocytes plays a major role in host defense during these intracellular bacterial infections. Several antigenic glycoproteins have been described in A. marginale , but it is still unclear what the impact of this discovery will be on understanding of the protective immune responses (Moens and Vanderleyden, 1997; Leal et al. , 2000). Pathogenesis of Anaplasmosis The losses attributable to bovine anapl asmosis in the United States have been estimated from $ 5-11 million in California (Goodger, Carpenter, and Riemann, 1979) to over $ 300 million in t he continental United States (Palmer,


21 1989). Losses are caused by death, abortion, weight loss and increased veterinary and management costs (Alder ink and Dietrich, 1981). In most regions of the world where anaplasmosis is endemic, it is difficult to determine the losses, mainly due to lack of record s, inability to quantify production losses and concurrent infection with other hem oparasitic and tick-borne diseases. The clinical signs, mainly caused by the anemia due to phagocytosis and removal of erythrocyte, include depr ession, weakness, fever, dehydration and jaundice. Lactating cows have a decrease in milk production (Morley and Hugh-Jones, 1989b) and abortion is a common featur e in advanced cases of pregnancy (Palmer, 1989). Infected bulls sh ow a lack of libido and abnormal sperm morphology, occasionally accompanied by transitory testic ular degeneration during acute infection (Swift and Thomas, 1983). Anemia and fever are the main clinic al features of anaplasmosis. The haemolytic nature of the anemia is indi cated by a significant increase in unconjugated bilirubin during the acute phase (Ajayi, Wilson, and Campbell, 1978). Since it has been shown that A. marginale leaves erythrocytes without concomitant lysis of the host cell and hemoglobinuria (Erp and Fahrney, 1975), the immune system is suspected to play a role in the lysis of erythrocytes and the pathogenesis of anaplasmosis. Soler–Rodriguez et al. (1990) found a significant reduction in bovine serum co mplement activity during the acute phase of infection and an increase in the sensitivity of the erythrocytes to bovine complement lysis in vitro. Also , erythrocyte destruction is delayed in splenectomized calves upon challenge (B uening, 1973). Using erythrocyte


22 fragility and direct antiglobulin tests (C oombs test), Swenson and Jacobs (1986) were able to document regenerative i mmunohemolytic anemia in cows with anaplasmosis. The role of the immune response in the pathogenesis of Ehrlichial diseases is conflicting. On the one hand, the immune response is obviously involved in the elimination of pat hogens, and these pathogens have developed mechanisms to suppress this response. On the other hand, there is clear evidence that the immune response is for a large part responsible for the pathologies involved. Clini cal features of rickettsia l diseases only moderately correspond to damage of the main ta rget cells. The majority of Ehrlichia species selectively inhabit unique intrac ellular niches, yet clinical signs are remarkably similar across both Ehrlichia and host species (Madigan et al. , 1995). The spectrum of clinical illness is also disproportionate to the number and anatomic distribution of organisms detec ted in blood or tissues in humans and in animal models (Martin, Bunnell, and Du mler, 2000). The relative lack of host tissue injury in the absence of significant inflammatory responses during infection of SCID mice clearly sugges ts an immunopathologic component to the disease (Bunnell et al. , 1999). This has been shown for A. phagocytophilum infections which induce an IFNpeak in mice prior to maximal pathologic change when ehrlichial organisms are still abs ent from tissues (Martin, Bunnell, and Dumler, 2000), while another cytokine, IL-10, moderates the pathology (Martin, Caspersen, and Dumler, 2001).


23 Immunosuppression, on the other hand, is also a common feature in many rickettsial diseases and can cont ribute equally to the pathogenesis. In E. chaffeensis infections observations of opportunistic infections in severe and fatal human cases suggest suppression or dysregulation of the immune response (Walker and Dumler, 1996). Immunosuppression by A. phagocytophilum has been well documented. In fections lead to decreased neutrophil adherence, decreased bacterial killing and proliferation by peripheral blood leucocytes of sheep, and dec reased production of antibodies (Woldehiwet, 1987a; Woldehiwet, 1987b). All of these can contribute to the pathogenesis of the disease through the increase and aggravation of concurrent infections. The mechanism s by which ehrlichial organisms impair the host responses are not fully underst ood, but data demons trate respiratory burst inhibition by A. phagocytophilum in vitro in a mouse model and in human patients (Walker and Dumler, 1996; Banerjee et al. , 2000; Wang et al. , 2002). Regardless of all the ev idence on their involvement in the pathology, the immunological reactions against ricke ttsial agents are considered mostly protective (Bunnell et al. , 1999). As so often in immune responses it is likely that the right immune re sponse will be a delicate balance between protection and pathology. Control of Anaplasmosis The eradication of anaplasmosis is not possible in most countries because of the wide range of arthropods which are capable of harboring and


24 transmitting the disease, the long period of infectivity of carrier animals and the presence of carriers in the wild animal population (Blood, 1983). Control of ticks and the blood-sucking insects which transmit A. marginale is difficult to achieve with cattl e in open ranches and farms. Dipping or spraying with insecticides or repell ents is expensive, the development of acaricide resistance likely and the dist urbance of the endemic stability for Anaplasma and other tick-borne diseases are serious drawbacks (Kuttler, 1979). Several chemotherapeutics have been developed and used, but only tetracycline compounds and imidocarb are currently in use (Kuttler, 1979; McHardy et al. , 1980; Kocan, Blouin and Barbet, 2000). Tetracycline and chlortetracycline have been used at different doses and intervals, but the advent of long-acting oxytetracycline pr eparations have made treatment more effective. Imidocarb at 3 mg/kg body weig ht is an effective treatment for clinical cases and has the advantage of not inte rfering with the development of immunity to A. marginale (de Vos et al. , 1987). Chemoprophylaxis using chlortetracycline in feed supplements and mineral blocks was found to be of little or no benefit in controlling infection, but no significant production losses are observed in surviving animals (Thompson et al. , 1978; Kuttler, 1986). Under natural conditions animals ac quire protection against clinical anaplasmosis dependent on primary infe ction and recovery, but without clearance of the pathogen. Based on this phenomenon, several vaccines have been developed, all of which have advantag es and drawbacks. These vaccines


25 are either based on live pathogen s of the related species A. centrale , or on live, attenuated or inactivated A. marginale pathogens, or on recombinant antigens of A. marginale . Vaccines containing the less-pathogenic, live A. centrale are currently used in South Africa, South America, Israel and Austra lia for control of bovine anaplasmosis. A. centrale and A. marginale share immunodominant antibody and CD4+ T cell epitopes that may play a role in the protection induced by A. centrale (Shkap et al. , 1991; Shkap et al. , 2002). The pathogens are derived from infected splenectomized calves and upon infection they provide partial, variable protection against A. marginale . Protection against challenge is adequate in most cases to prevent disease and use of the vaccine appears to be justified (Bock and de Vos, 2001). Prot ection against antigenically diverse, highly virulent stocks of A. marginale in some countries (e.g., Uruguay and Zimbabwe) is, at times, clearly inadequat e and better vaccines are required in situations where the challenge is severe (Brizuela et al. , 1998; Turton et al. , 1998). On the American mainland, attempts have been made to develop inactivated or live vaccines based on A. marginale . In search of a live vaccine based on A. marginale , three different methods we re developed involving the use of a virulent isolate together with treatm ent, a mild isolate, with or without treatment, and an isolate attenuated by passage through non-bovine animals (Carson, Sells, and Ristic, 1977a; Vizcaino et al. , 1980). All three live vaccines afford protection against clinical di sease, but even with the sheep-passaged


26 attenuated live vaccine reversion to viru lence remained a problem (Jorgensen et al. , 1993). Killed vaccines developed in the United States were marketed until 1999 when they were withdrawn due to company restructuring. While there is no risk of reversion to virul ence and the risk of contamination with undesirable infectious agents is low with inactivated vaccines their efficacy is more variable. Further disadvantages of killed vaccines include the need for yearly boosters and the higher cost of purification of A. marginale from erythrocytes. Bovine blood-derived vaccines have been linked to neonatal isoerythrolysis in calves induced by imperfect removal of the erythrocyte antigens (Carson and Buening, 1979). A bovine-derived, “purified” and inactivated vaccine in adjuvant is still in use today in the United States, although only approved for use as an experimental vaccine by the USDA (University Products, L.L.C., Baton Rouge, LA). Recently, a continuous culture system was developed for A. marginale in an embryonic cell line of the tick Ixodes scapularis (Munderloh et al. , 1996). The MSPs characterized on A. marginale in bovine erythrocytes were found to be conserved on the cell cult ure-derived pathogens (Barbet et al. , 1999) and tested as an immunogen for cattle. Ca ttle immunized with the cell culturederived A. marginale were protected against anaplasmosis after challengeexposure to infected blood or by feeding infected ticks (Kocan et al ., 2001; de la Fuente et al ., 2002). The protection was, howev er, partial and the disease is not prevented in all animals. The main effect of the vaccine is similar to the


27 effect observed with erythrocyte-derived A. marginale , resulting predominantly in a less pronounced anemia. Native or recombinant polypeptides have been proposed and tested as an alternative to these vaccines (Palmer, 1989; Palmer and McElwain, 1995). Limited vaccine trials have also been conducted using recombinant vaccinia virus expressing an A. marginale antigen (McGuire et al., 1994) or with naked DNA (Arulkanthan et al., 1999). These re combinant approaches have thus far failed to induce good protection against challenge. Cultures of A. marginale A major obstacle in anaplasmosis research has been the lack of an in vitro culture system, and all research has relied on the use of infected cattle or ticks as a source of A. marginale . Several attempts to grow the pathogen in mammalian host cells have failed. Cultures in erythrocytes, cells derived from bovine lymph nodes or bovine turbinate cells have resulted only in short-term infections (Hidalgo, 1975; Kessler and Ristic, 1979; Mazzola and Kuttler, 1980; Blouin, Kocan, and Ewing, 1992; Blouin et al. , 1993; Waghela et al. , 1997; Leal et al. , 2000). In each system, the organism showed an initial per iod of growth followed by a gradual decrease in the per centage of parasitized cells. Several attempts were made by growing A. marginale in arthropod cells, including mosquito and tick cell lines (Mazzola , Amerault, and Roby, 1976; Samish, Pipano, and Hana, 1988; Hidalgo et al. , 1989). Oddly enough it was only when embryonic tick cells from Ixodes scapularis , not a known vector of A. marginale , were used that continuous propa gation was achieved (Munderloh et al. , 1996).


28 Importantly, the infected cell cultures can be frozen, stored and then recovered by inoculating uninfected cultures. Colonies of A. marginale in this IDE8 cellline are morphologically similar to thos e found in naturally infected tick cells (Blouin and Kocan, 1998). The cell cultur es have remained infective for cattle, they cause clinical anaplasmosis, and D. andersoni ticks can acquire and transmit the organisms from these culture-infected animals (Blouin et al. , 1999). All major antigens characterized in eryt hrocytes are present on culture stages and appear structurally conserved dur ing continuous culture (Barbet et al. , 1999). These characteristics are refl ected in the cultures showing good promise as a source of antigens in the development of vaccines for anaplasmosis (Barbet et al. , 1999; Kocan et al. , 2001) and basic research. The A. marginale tick cell culture system was adapted for short term growth in 24well and 96-well plate formats for use in the development of various assays, including a competitive ELISA (Saliki et al. , 1998), A. marginale tick infectivity studies (de la Fuente et al. , 2001a, 2001b), infectivity excl usion studies, tick cell adhesins identification, dr ug screening tests (Blouin et al ., 2002), and cloning of the pathogen. Antigenic Variati on of Anaplasmataceae Higher eukaryotic multicellular org anisms act as microenvironments for numerous microorganisms. The successful reproduction of a microbial species within such environments requi res competition for nutrients and evasion of host defenses. Additionally, from a microbial standpoint, higher organisms constitute noncontiguous “island” habita ts, requiring between-host transmission strategies


29 to ensure long-term survival of the s pecies. Microbes have evolved powerful mechanisms for generating phenotypic dive rsity as an efficient strategy for adapting to rapidly responding imm une system defenses and the broad range of polymorphisms characteristic of differ ent host tissues. These mechanisms of variation may be of particular relevance in the successful spread of an infection through a host population. While many of these genetic mechanisms of phenotypic variation appear to be involved in evasion of host immunity, they are also found in organisms on which the impact of immune defenses is unknown or absent, such as the intestinal parasite Giardia (Nash et al. , 2001; Singer et al. , 2001) and free-living organisms (Kusch and Schmidt, 2001). Although classical gene regulatio n can produce phenotypic change, the mechanisms of antigenic variation offer an enormous range of diversity while using only a relatively small fraction of the genome. The loci with the potential to generate extensive variation are c onfined to a minority of nucleotide sequences within a genome. This loca lization has apparently evolved to produce a repertoire of variant molecule s that modulate such properties as antigenicity, motility, chemotaxis, atta chment to host cells, acquisition of nutrients and sensitivity to antibiotics, while avoiding the deleterious effects that high mutation rates would impose on other conserved housekee ping functions, (e.g., cell cycle control, metabolism, signaling and synthesis) (Deitsch, Moxon, and Wellems, 1997). The molecular switchi ng and variability provided by these genes in protozoal, bacterial and fungal pat hogens are important in determining


30 whether a particular pathogen is cleared from the host, persists to cause a relatively benign infection, or produces severe or even fatal disease. Many pathogenic bacteria exist in nature as multiple antigenic types or serotypes, variant strains of the same pathogenic species. Although this has advantages at the populatio n level, it is not considered real antigenic variation, but rather diversity. A second more dy namic mechanism of antigenic variation is seen in retroviruses. The rate of mutation occurrence in retroviruses is extremely high because of the lack of proof-reading function of the reverse transcriptase. Many changes may be deleter ious to the virus, but stable aminoacid substitutions may arise. Addi tionally, segmented virus genomes make the interchange of genes from different viruse s possible. This re-assortment can occur when two different viruses are pr esent in the same cell, and new strains with a shift in genetic compositions can be formed. True ant igenic variation, however, arises in a single clone or genotype in a single host (Barbour and Restrepo, 2000). In most cases, this c hange is reversible (i.e., the information for producing the original antigen is arch ived in the cell and can be used in the future). True antigenic variation is often found in vector-borne pathogens. Since they have to “wait” for a vector to be transmitted to a new host, they need to survive in the host for a long time at sufficient levels to be picked up, constantly avoiding the immune system. Four general mechanisms for antigen ic variation have been described : modification of transcript levels, gene conversion, DNA rearrangement, and multiple point mutations (Restrep o and Barbour, 1994; De itsch, Moxon, and


31 Wellems, 1997). The first and the second mechanism are probably the most widespread for replacing expression of one gene with another. Both allow a pathogen to retain a complete repertoir e of variable anti gen genes. In gene conversion, the gene is displaced from the expression site and “replaced” by a copy of the gene from a more stable loca tion in the genome. Gene conversion systems are present in many different pathogens, such as Mycoplasma VlhA variation (Noormohammadi et al. , 2000), trypanosome VSG variation (Borst et al. , 1997) and the malaria var system (Newbold et al ., 1997). The most detailed molecular models have been developed for the mating-type system of Saccharomyces cerevisiae (Haber, 1998). The beststudied bacterial systems are the Neisseria pilus antigenic variation (Seifert, 1996) and the Borrelia variable membrane protein (VMP) and VMPlike system (Vls) antigenic variation (Restrepo and Barbour, 1994; Zhang et al. , 1997). Neisseria gonorrhoeae has evolved several systems for varying t he antigenicity of different surface antigens. A phase variation of surface st ructure expression alters the antigenic characteristics of the cell surface. It involves ~10 full-length opa genes that all can be expressed or turned off using a num ber of 5-mer repeats to cause frame shifts at the beginning of the reading frame. Antigenic variation of the major subunit of the surface pilus occurs by unidirectional, homologous recombination between a truncated gene in a silent locu s and the full-length expressed copy (Seifert, 1996). In Borrelia hermsii several vmp (variable membrane proteins) use the same locus for expression. Gene conversion happens between a linear plasmid containing a collection of s ilent vmp genes and anot her linear plasmid


32 with an active vmp gene downstream from a promoter. The gene conversion may be partial, yielding a chimeric vmp gene. Since the genes are often found in tandem, additional variation can be generated through a DNA rearrangement in which the first member of the tandem pair is deleted, placing the formerly distal vmp in the next to the promot er (reviewed in Barbour and Restrepo, 2000). Gene conversion often involves genes on separate chromosomes or plasmids in the cell (Deitsch, Moxon, and Wellems, 1997). Anaplasmataceae are unique in that they have developed el aborate mechanisms for antigenic variation within the size restrictions of their one small chro mosome (1-1.6 Mb) (Alleman et al. , 1993; de Villiers et al. , 2000) and in the absence of extra chromosomal plasmids, as found in Borrelia species (Wang, van, and Dankert, 2001). Additionally, this involves not one but several families of related genes to generate variation. Within each genus the organisms are antigenically highly cross-reactive and share several homologous surface antigens. The genus Ehrlichia has a 23 to 34 kDa major surface antigen complex, Anaplasma spp. have a 36 to 44 kDa major antigen in common, and Neorickettsia share a 51 to 55 kDa complex. These surface anti gens undergo antigenic variation in an effort for the organism to evade t he immune system, apparently often using similar mechanisms within one genogroup.


33 Antigenic Variation in the Genus Ehrlichia MAP1, OMP1, P28, P30 Ehrlichia spp. possess several immunodominant proteins ranging from 20 to 120 kDa (Rossouw et al. , 1990; Brouqui et al. , 1992; Yu, McBride, and Walker, 1999). A group of antigens of lower molecular weight are among the most immunodominant proteins and have been labeled p28 (OMP1), p28 homologue, p30, and MAP1 in E. chaffeensis , E. ewingii , E. canis and E. ruminantium , respectively (Rossouw et al. , 1990; Nyindo, Kakoma, and Hansen, 1991; Rikihisa, E wing, and Fox, 1994; Gusa et al. , 2001). These antigens share common epitopes and t hey are each encoded by homologous multigene gene families. The genes and t heir products have homology to the MSP2 multigene family identified in the genus Anaplasma (Bakken et al. , 1994; Chen et al. , 1996; Ohashi et al. , 1998a; Ohashi et al. , 1998b; Sulsona, Mahan, and Barbet, 1999). Despite the similarity of the i ndividual genes, the overall genomic structure of the gene families encoding the outer membrane proteins in Anaplasma and Ehrlichia is very different. Msp2 genes, both complete and truncated, are dispersed throughout the genome (Palmer et al. , 1994a; Brayton et al. , 2001), whereas all the map1like genes are located in one or two loci. Recent studies have found increasing numbers of map1 -like genes arranged in tandem, apparently all in the same locus, in E. chaffeensis , E. ewingii , E. canis and E. ruminatium . Yu et al. (2000a) describe a locu s containing 21 p28-genes in E. chaffeensis , and Ohashi et al. (2001) describe a slightly different cluster in


34 another strain containing 22 paralogs. A similar region including 22 paralogs plus an additional region wit h 3 paralogs was found in E. canis (Ohashi, Rikihisa, and Unver, 2001). Inte restingly, and different from msp2 , all genes appear complete and are characterized by four conserved regions separated by three hypervariable regions. Most of the genes are arranged in tandem in one direction, except for one ( E. canis ) or two ORFs ( E. chaffeensis ) at the end of the locus that are on the oppos ite strand. In the first half of the locus, 14 paralogs are linked by short intergenic s paces ranging from 8 bp to 27 bp and the eight remaining paralogs in t he 3'-end half are connected by longer intergenic spaces ranging from 213 to 632 bp. Two different mechanisms to generate diversity have been proposed for this family of genes. One involves a mechanism of “somatic” mutation, the other transcriptional regulati on. Nucleotide identity diffe rences of up to 14.5% in the same map1 -like gene in different strains of the same species are common (Reddy and Streck, 1999; Long et al. , 2002). This system, if active in all ~21 genes of the locus could potentia lly generate substant ial diversity within a population. Frequent point mutations in expressed vmp genes can generate significant antigenic diversification in Borrelia hermsii (Restrepo and Barbour, 1994) . However, while several different map1 -like gene variants can be found in one geographical location, near identical variants can be found at a different location, suggesting a relative stability of the variants (Long et al. , 2002). Since most genes appear to be full length, differential expression of genes from their own promoter is lik ely a second and major mechanism for


35 variation. Transcriptional analysis of the different map1 -like genes reveals differential expression. This ex pression is sometimes dependent on temperature or host cell environment, but importantly little influence is noted from the immune system during infe ction in the mammalian host. In E. canis most genes appear to be transcribed in cultures, but three paralogs are undetectable during canine infection, sugges ting that transcriptional regulation is different during infection (McBride, Yu, and Walker, 2000; Ohashi, Rikihisa, and Unver, 2001). Downregulation of the expression of all except one p30 paralog in ticks, as well as in DH82 cells grown at 25°C, suggests that temperature may also be a factor r egulating the mRNA expression (Unver et al. , 2001). The paralogs with short intergenic spaces seem to be transcribed on a polycistronic message (i.e., together on one mRNA), the remaining genes monocistronically (i.e., on separate m RNAs) (Ohashi, Rikihisa, and Unver, 2001). In contrast, in E. chaffeensis all transcripts appear monocistronic by RTPCR (Long et al. , 2002). Also, only 16 out of 22 paralogs are transcribed in culture and 16, not all the same , during infection in dogs (Long et al. , 2002; Unver et al. , 2002). Reddy et al. (1998) argue that only one gene is transcriptionally active, but they di d not examine all paralogs (Reddy et al. , 1998). Transcription in A. americanum ticks involves a single gene, both before and after transmission feeding, as occurs with E. canis in R. sanguineus , involving the ortholog at the same location in the locus (Unver et al. , 2002). Finally, in isolates of E. ruminantium grown in bovine endothelial cells in two different tick cell lines and in A. variegatum ticks, only one map1 gene, out of


36 three examined, was always transcr ibed. Expression of another gene, map1-2 , was not detected under any condition (Bekker et al. , 2002). Temperature, isolate and passage differences we re observed for a third gene, map1-1 . In the wild-type Senegal stock, expression of the gene was detected in ticks but not in bovine endothelial cells at either 30 or 37°C. Map1-1 was, however, found in different passages of the in vitro -attenuated Senegal isolate grown in bovine endothelial cells, as well as in the Gardel isolate grown in two tick cell lines. When transcribed, this map1-1 gene, was present on a polycistronic message together with map1 (Bekker et al. , 2002). It is essential to point out that, for both E. canis and E. chaffeensis infections of dogs, no change in transcrip tion was found during the course of the infection (Unver et al. , 2001; Unver et al. , 2002). Unlike in A. marginale msp2 , no clear peak of bacteremia or radi cal changes in the compositions of expressed p30 paralogs was observed in E. canis in dogs during a 56-day infection period, with the same paralogs remained transcriptionally active. Since antibodies that develo ped did not appear to clear E. canis which expressed a particular set of p30 paralogs and since E. canis expressing new p30 paralogs did not emerge after t he development of the antibodies, p30 gene expression by E. canis in dogs does not appear to be related to evasion of the humoral immune response (Unver et al. , 2001). 120-kDa immunodominant protein The 120-kDa protein is one of th e immunodominant proteins of E. chaffeensis and E. canis . DNA sequence analysis in E. chaffeensis identified


37 an open-reading frame with five near i dentical 240-bp tandem repeat units, comprising 60% of the entire gene (Yu, Crocquet-Valdes, and Walker, 1997). In E. canis tandem repeat units were also ob served. However, neither the repeat number nor the amino acid sequences in the repeat s are identical in the two Ehrlichia species (Yu et al. , 2000b). These differences account for antigenic variability between these two species, not antigenic variation. However, Popov and co-workers (2000) detected the 120 kDa protein by immunoelectron microscopy in the outer membrane of the cell wall of densecore forms, but not in the cell wall of re ticulated forms. This provides evidence for stage-specific protein expression, as observed for many bacteria, including Borrelia burgdorferi (Schwan et al. , 1995), and is a form of antigenic variation that might be important in the development of vaccines against the pathogen. Antigenic Variation in the Genus Anaplasma MSP1b MSP1 is a dimer of two structural ly unrelated polypeptides, MSP1a and MSP1b (Barbet et al. , 1987; Vidotto et al. , 1994). The complex contains an infection neutralization-sens itive epitope conserved between A. marginale isolates in the United States , Israel, and Af rica (Palmer et al. , 1987). Data suggest that the MSP1a and MSP1b polypeptides have functions as adhesins to tick cells and erythrocytes (McGarey et al. , 1994; de la Fuente et al. , 2001b). MSP1a is encoded by a single copy gene (Allred et al. , 1990) and marked size polymorphisms among diffe rent isolates of A. marginale are due to variations in the numbers of tandemly repeated sequences (Allred et al. , 1990). These


38 repeats are stable over time and in different host environments (Palmer, Rurangirwa, and McElwain, 2001; Bowie et al. , 2002). However, like MSP2 and MSP3, MSP1b is encoded by a polymorphi c multigene family (Barbet and Allred, 1991) and thus has the potential to undergo antigenic variation. The exact structure of the gene family is still under debate. Using primers derived from the 5' and 3' ends of the previously pub lished Florida strain msp1ß sequence. Camacho-Nuez et al. (2000) identified three new complete msp1ß genes in the Florida strain. Each of these polymorphic genes encoded a structurally unique MSP1b, apparently resulting from a unique mosaic of five variable regions. However, using an identical approach, only two unique msp1ß copies were identified in the Ha vana (Cuba) strain. Subsequent RTPCR using short but gene-specific sequences indicated that at least two of the three genes were transcribed. In a separate study Viseshakul et al. (2000) screened a phage Lambda library, of t he Florida isolate, with an msp1ß probe and found that the msp1ß family consists of five par tially homologous copies, of which only two are complete and expressed in A. marginale . All five copies were found on a 156 kb fragment, but no closely linked structure as in the map1 -gene family was present. The di screpancies between the numbers of unique full-length genes present could result from rapi d recombination of the genes in A. marginale . Alternatively “novel” genes might be produced by artificial mosaic formation during PCR, as demonstrated by Bowie et al. (2002). The formations of mosaic structures during PCR amplification is a known shortcoming of the technique (Sinkora, Sun, and Butler, 2000), especially in


39 homologous multigene families. Screeni ng of a library does not have this shortcoming but might not yield all the genomic copies. Recombination between members of the msp1 gene family was examined by comparison of the genes in pathogen po pulations derived from acutely and persistently infected cattle, from ticks that acqui red infection with the pathogen and, finally, after transmission to a new animal during the acute phase (Bowie et al. , 2002). Few amino acid substitu tions were observed in the different A. marginale populations in either msp1 -1 or msp1 -2 . The substitutions observed might represent a recombination event between the fulllength and the closely related truncated msp1 genes. However, it is impossible to exclude the selection of a divergent subpopulation present in the initial inoculum or small amplif ication and sequencing errors (Bowie et al. , 2002). Limited recomb ination of related msp1 genes and / or regulation at the expression level of the fu ll-length genes thus might pl ay a role in generating diversity in this gene and protein family. MSP2 MSP2 is the only antigen of A. marginale or any species of the family Anaplasmataceae for which antigenic variation has been demonstrated. Closely related orthologs have been demonstrated in A. phagocytophila , A. ovis and A. centrale at the protein and nucleic acid level (Shkap et al. , 1991; Dumler et al. , 1995; Palmer et al. , 1998; Zhi et al. , 1998; Murphy et al. , 1998; Ijdo et al. , 1998; Shkap et al. , 2002). In A. marginale, A. ovis and A. phagocytophilum , multiple msp2 gene-related copies are present, widely distributed throughout


40 the chromosome (Palmer et al. , 1994a; Palmer et al. , 1998; Zhi, Ohashi, and Rikihisa, 1999). The presence a nd the expression of polymorphic msp2 genes in A. marginale, A. ovis and A. phagocytophilum was subsequently confirmed (Eid et al. , 1996; Zhi et al. , 1998; Shkap et al. , 2002). Expressed msp2 genes are characterized by a large centra l region of amino acid polymorphism, including additions, deletions, and substi tutions flanked by highly conserved Nand Cterminal regions (fig. 6). The Nand C-terminal thirds of MSP2 are not identical between A. centrale and A. marginale , yet they are highly conserved among MSP2 variants within each species (Shkap et al. , 2002). In A. marginale and A. centrale several variant msp2 types are present in each rickettsemia cycle (French et al. , 1998; Shkap et al. , 2002). Importantly, in A. marginale , the variable region contains exposed su rface epitopes that induce antibody subsequent, but not prior, to that ricke ttsemia cycle and, ther efore, represents true antigenic variation caused by repeated emergence of MSP2 variants in the bovine host followed by immune control (French, Brown, and Palmer, 1999). A unique operon, containing f our open reading frames, is the only site in the genome for significant expression of full-length msp2 transcripts in A. marginale and A. ovis (Barbet et al. , 2000; Shkap et al. , 2002). Two of the upstream ORF products, OpAG-2 and OpAG-3 (operon asso ciated gene), also localize to the bacterial surface; while opag-1 does not appear to be translated (Lohr et al. , 2002). MSP2 transcripts from this ex pression site are polymorphic in bloodstream populations of A. marginale (Barbet et al. , 2000; Barbet et al. , 2001b). At least four diffe rent variants of the cent ral hypervariable region are


41 found in each rickettsemia cycle of a persistent infection (Eid et al. , 1996; French et al. , 1998). The remaining msp2 copies are truncated pseudogenes, most of which contain a central hyperva riable region flanked by shorter portions of the 5Â’ and 3Â’ conserved regions (Brayton et al. , 2001). These pseudogenes serve as a source of variants for t he expression operon, through a mechanism of segmental gene conversion (Barbet et al. , 2000; Brayton et al. , 2002). More distantly related msp2 copies are also present in the genome but their function remains unclear (Meeus and Barbet, 2001). In A. phagocytophilum a similar mechanism with one polymorphic expression site for msp2 , or p44 , exists, although only one upstream ORF is present upstream from the msp2 gene (Barbet et al. , unpublished 2002). It remains unclear if this expression site is the only one within the genome, although ribonuclease protection assays and 5Â’RACE results are consistent with it being at least the ma jor expression site (Barbet et al. , unpublished 2002). Using variant-specific probes in a genomic Southern blot and monoclona l antibodies specific for expressed copies Zhi et al. (1999) argue for the presence of several different active expression sites with a potentially different mechanism for generating diversity in A. phagocytophilum by transcriptional regulation. However, a single but polymorphic expression site in a population can account as well, if not better, for the re sults obtained. While progress is being made in understanding the mechanism involved in generating polymorphic M SP2 expression and variat ion, little is known concerning the influences that driv e these events. Is the observed gene conversion in A. marginale merely a random continuous event with variants


42 arising simply by selection or do certain stimuli induce this switching? Diversity in the expression site has been observed in all environments in which the pathogen thrives, but the immuno-com petent bovine host appears to be the environment in which most new variants arise (Barbet et al. , 2001b). Recent studies have demonstrated a rapid and si multaneous replacement of dominant msp2 and msp3 variants during bovine infection (Brayton et al ., unpublished 2002). Increased diversity over time has been observed in long-term infection of ticks for some A. marginale isolates (Rurangirwa, Stiller, and Palmer, 2000; de la Fuente and Kocan, 2001), although a re striction of variants in ticks has been described for another isolate (Rurangirwa et al. , 1999). A similar situation exists in A. phagocytophilum where a polymorphic expression site is described in all populations examined. However, different culture environments (i.e., HL60 human monocytes or ISE6 tick cells) are consistently characterized by specific different dominant variants of msp2 (Barbet et al ., unpublished 2002). It is unclear whether these different environments favor the selection of a subpopulation, a selection occurs of newly randomly created vari ants, or if the environments trigger a specific recomb ination event. Nevertheless, this observation is important since it highlight s that antigenic variation might not play a role only in immune evasion but also in adaptation to the environment (e.g., different receptors or nutri ents). Together the data ar e consistent with a slow but continuously generated diversity, r apidly selected for by the immune system or different environmental pressures.


43 The simultaneous occurrence of variation in msp2 and msp3 in A. marginale could be the result of either independent recombination mechanisms at similar rates for the two genes or a coordinating mechanism for both genes. The latter is likely since msp2 and msp3 pseudogenes are often found in a “pseudogene complex”, in a ta il-to-tail orientation flanked by a long conserved repeat at the 5’ and / or 3’ end (see chapt er 3). This structure might play a critical role in the mechanism ge nerating the observed diversity. MSP3 MSP3 is structurally and antigenica lly polymorphic among strains of A. marginale (Alleman and Barbet, 1996; Kano et al. , 2002). Multiple, partially homologous gene copies have been i dentified and they appear widely distributed throughout the chromosome. Sequence analysis of three unique msp3 genes revealed both conserved and variant regions within the open reading frames. Interestingly, the msp3 ORF contained amino acid blocks related to MSP2 at the 5’ end and two available 3’ flanki ng regions contained msp2-related sequences (Alleman et al. , 1997). This structure was confirmed by data from the ongoing A. marginale genome sequencing project (Brayton et al. , 2001). The presence of msp2 -related sequence in the msp3 genes suggests that the msp3 genes could be involved in varying both MSP3 and MSP2. The concerted appearance and the similar flanking sequences could indicate that the regulation of variation could be linked or at least involve similar mechanisms. msp3 has been described in A. ovis , but no clear indications of its presence in A. centrale or A. phagocytophilum have been published.


44 Alleman and Barbet (1996) used Western blots with A. marginale antigens and sera from animals infected with either A. ovis , A. centrale , E. risticii , E. equi or E. ewingii to demonstrate that each of t hese cross-reacted with the MSP3 antigen, as well as with some smaller product s. However, the cross-reactivity in all those species is more likely due to the close relationship of MSP3 with MSP2, for which the MAP1 -like homolog in the genus Ehrlichia has been demonstrated. Database searches agai nst the unfinished genome sequences of E. ruminantium and E. chaffeensis do not yield any significant hits when searched with msp3 -specific sequences. A Southern blot of A. phagocytophilum genomic DNA with ol igonucleotide probes derived from msp3 conserved regions did not produce any hybridizing bands (Meeus, unpublished 2002). However, Western blot resu lts and Southern blot hybridization established the presence of MSP3 in A. ovis and revealed that it was also encoded by a multi-gene family (A lleman and Barbet, 1996; Palmer et al. , 1998). For A. centrale , long used as a live vaccine strain to protect against anaplasmosis (Bock and de Vos, 2001), no cl ear evidence of the presence of MSP3 has been published. Western blot s reveal a highly immunogenic group of proteins ranging from 80120 kDa in size, but since A. centrale immune serum does not cross-react with A. marginale MSP3 it is unclear what these proteins are or what their role in antigenic variation might be (Bowles et al. , 2000; Molloy et al. , 2001).


45 CHAPTER 3 TICK CELL CULTURES AND CLONING Introduction A common and signific ant feature of A. marginale and closely related pathogens is their ability to evade the immu ne system and persist in their host. The carrier state is of gr eat importance in the maint enance of the life cycle by alternate tick/host challenge and may be necessary for the maintenance of immunity. Possible mechanisms of persi stence include antigenic variation of the proteins recognized by the imm une system, immuno-suppression by the parasite and, survival of stages in hos t locations with low levels of immune defense such as the brain. Antigenic vari ation, especially, is a problem for the development of vaccines against the disease. There is at present considerable proof for antigenic diversity and variation in A. marginale . Several investigators have demonstrated extensive polymorphism s in major surface proteins between A. marginale isolates from different geographica l regions, within isolates from the same geographical region and within th e same isolates at different times post infection (Oberle et al. , 1988; Palmer et al. , 1994a; Eid et al. , 1996; French et al. , 1998; French, Brown, and Palmer, 1999). It is unclear how the data have been complicated by evolutionarily divergent subpopulations existing in the original isolate. All research has employed naturally derived isolates t hat have been passaged for varying times


46 in ticks and cattle. Analysis of these populations provides clear evidence for substantial MSP2 diversit y at the population level in both environments (Barbet et al. , 2001b). In order to assess the capacity of A. marginale to vary antigenically one must characterize the progeny of a single organism. That is, isolate a clone of A. marginale and characterize antigenic variation in subsequent infection cycles. A major impediment in cloning the pathogen was the lack of an efficient in vitro culture system for A. marginale . Recently, A. marginale was successfully propagated in a tick cell line derived from the black-legged tick, Ixodes scapularis , using infected bovine blood as the inoculum (Munderloh et al. , 1996). Erythrocytic stages invaded the tick cells and multiplied in membrane-lined vacuoles to form colonies typical of those observed in naturally infected ticks as demonstrated by light and electron microscopy (Blouin and Kocan, 1998). Antigens present in A. marginale from tick cell culture were recognized by bovine immune serum against the blood stages of A. marginale and the cell-culture-derived organism retai ned its infectivity for cattle (Barbet et al., 1999). The cultures therefore not only provi de a suitable source of pathogens for basic research and vacci ne development, but also provide a prime environment to clone t he organism. In this chapter I describe different methods developed that can be us ed to obtain a clonal line of A. marginale by serial dilution and micromanipulation and a strategy to demons trate the clonality of the obtained population.


47 Experimental Procedures Propagation of A. marginale in IDE8 Cells. The IDE8 (ATCC CRL 11973) uninfected tick cells were maintained at 31ºC in L-15B medium, pH 7.2, supplem ented with 5% heat inactivated fetal bovine serum (FBS; Sigma), 10% tryp tose phosphate broth (Difco) and 0.1% bovine lipoprotein concentrate (ICN) (Munderloh and Kurtti, 1989; Munderloh et al. , 1994; Munderloh et al. , 1996). Cultures were grown in 25-cm2 plastic flasks (Nunc) with 5 ml of medium, and the medi um replaced weekly. The cells were subcultured at 1:5 to 1:20 depending on the need for c onfluent monolayers. The cells become tightly adherent to the culture substrate and multiply with a population doubling time of 3 to 5 days to a dens ity of about 5 x 106 cells/ml. Tick cell cultures infected wit h the Oklahoma strain of A. marginale were maintained in L15B medium supplement ed with 5% fetal bovine serum, 0.01% bovine lipoprotein concentrate, 10% tryptose phosphate broth and maintained at 34ºC with a pH of 7.4 using NaHCO3 and MOPS buffer as described (Munderloh et al. , 1996; Blouin et al. , 2000). Culture medium was replaced weekly until desired infection levels we re reached. Infected cultures were monitored by phase contrast microsc opy and Giemsa-stained cytospin samples of culture medium containi ng detached cells and/or cells washed from the flask. When clumps of cells began to detach fr om infected monolayers, the flasks were subcultured or materi al collected for analysis.


48 Motility of Initial Bodies Inside Tick Cell Vacuoles A characteristic motility of A. marginale initial bodies inside tick cell vacuoles, potentially a virulence factor and method to assess viability of the pathogens, was examined. When the colonies in infected tick cell cultures were sufficiently developed to be easily det ected, the flasks were thoroughly examined to determine the presence of motility inside vacuoles using an inverse research microscope with up to 1, 000-fold magnification. When motility was detected, the cells were washed from the flasks and 100 µl of the total 25 ml added to 6-well tissue culture plates (N unc). Five wells contained 1 ml of medium with bacterial energy pathways inhibitors, one well contained medium without glucose, one well medium with ox ytetracycline at a concentration of 5 µg/ml and one well contained normal medium as a negative control. Potassium cyanide (KCN) was used at a concentra tion of 0.01 M, sodium arsenate (NaAsO2) and sodium fluoride (NaF) at 0.0001 M, and dinitrophenol (C6H4O5N2) and dicyclohexylcarbodiimide (C13H22N2) at 0.001 M. These concentrations have been shown to inhibit bacterial motilit y within seconds after addition to the medium (Daniels, Longland, and Gilbart, 1980). Motility of the initial bodies in the vacuoles was followed continuously for three hours starting immediately after addition of the cells to the wells. The cells in medium lacking glucose or containing tetracycline were again observed after 12 and 24 hours.


49 PCR Amplification and Restricti on Fragment Length Polymorphism Analysis of the msp2 Expression Site. The genomic expression site for msp2 was amplified using the SuperTaq system (Ambion) in semi-nested PCR’s . Oligonucleotide primers AB750 and AB752 were used in the first round, and AB688 and AB752 in the second round of amplification. All primers lo calize to flanking regions of the msp2 expressed gene and have been described previously (Barbet et al. , 2000). DNA template from A. marginale infected cultured tick cells was extracted with NucleoSpin nucleic acid purification kits (Clont ech). Single cell PCR was conducted similarly but a single tick ce ll containing one colony of A. marginale selected by micromanipulation was used directly as a source of DNA template. The total volume of the cell and medium added was estimated at 5 µl. Secondary PCR used 1/100 diluted products of the first round of amp lification as template. Control reactions were conducted similarly, but without DNA template. The PCR conditions used were an initial den aturation cycle of 5 min. at 93ºC, followed by 35 cycles of 30 se c. at 93ºC, 1 min. at 55ºC, 3 min. at 72ºC and final extension of 25 min. at 72ºC. PCR amplified products were analyzed by gel electrophoresis and cloned into the pCR-XL-Topo vector (Invitrogen). Clones were grown under kanamycin selection in a 96-well system and plasmids isolated, digested with EcoR I, to cut out the insert, and HinfI, to analyze the diversity of the inserts using electrophoresis in 1.5% agarose gels.


50 Primers The following primers and oligonucleotide probes were used : AB688 (GGACTGCTTGCCTTCACGCTGTT), AB750 (GGATTTTGTGGTCGGGTTTGTAT); AB752 (CACCGGTTGATGAAGTTTGC), Micromanipulation The micromanipulation wo rkstation consists of an inverse research microscope with up to 1,000-fold magni fication (phase contrast), an MMR110 micromanipulator 110 as well as a 0.1 ml gas-tight glass syringe with a Luertaper incorporated in a SR10-100L Mycrolyte Microi njector System. A Luertaper hypodermic needle is mounted on the taper of the syringe, the syringe is connected by hard polythene tubing to a micr opipette, which in turn is fitted to the micromanipulator. Monolayers of infe cted tick cell cultures were split during their geometric growth phase using a syri nge and needle. This results in small clumps of cells, but also individual ce lls. In suspension the cells become rounded and do not form attachments to adj acent cells. Individual cells were isolated by diluting them in medium until they contain approximately one infected tick cell per field at 200x magnification on a microscope coverslip. Single cells were inspected individually from different angles to ensure they contained only one vacuole and picked with the micromanipul ator, transferred to a clean coverslip and a drop of medi um added. The cells were picked again using a new micropipet and transferred to uninfected tick cell cultures for cloning or PCR tubes for am plification. The PCR t ubes were stored at -20ºC until used in PCR.


51 Serial Dilution When the tick cell monolayers showed an infection of 70 to 100%, the supernatants of four 25 ml flasks we re collected and centrifuged at 500 x g for 10 min. to pellet cell debris. The s upernatants obtained, containing free organisms, were serially diluted twofold in culture medium. One ml of each dilution was inoculated into each of 4 flasks of 25 ml coated with monolayers of IDE8 cells. The principle of the method is to infect IDE8 monolayers at a high multiplicity of infection with free organism s without clumps of infected cells and cells containing several colonies. When the colonies were sufficiently developed to be easily detected, the flasks were thoroughly examined to determine the percentage of infected ce lls and verified for the presence of mainly one colony per infected cell. Cultures with 80 to 100% of the cells infected by A. marginale were used for cloning. S upernatants of the cultures were diluted serially in medium and added to uninfected tick cell cultures in 96well microtiter plates. Addi tionally the infected cells were washed from the flask and counted to calculate the number of in fected cells per milliliter. These cells were seeded at a dilution of 0.3 infect ed cells per well (final volume, 100 µl / well) in 96-well microplates containing IDE8 cells. Results Motility Study: A. marginale in IDE8 Cell Lines No difference was observed in initial body motility inside vacuoles between infected cells kept in wells c ontaining medium with energy pathway inhibitors or normal medium over a th ree hour observation period. Medium


52 lacking glucose and medium containing oxytetracycline equally failed to halt the motility, even after 24 hours. Cloning of A. marginale A method to identify infected cells without staining and micromanipulate single tick cells infected with a single colony of A. marginale was developed (fig.1 & 2). Individual colonies of A. marginale are visible inside tick cells as a highly characteristic motile granular st ructure using either phase contrast or Nomarski optics. DNA staining with SYTO13 (Molecular Probes), which can be used on live cells, confirmed that the gr anules are mainly composed of DNA, as expected in a prokaryote (fig .1). This is the first de scription of visualization of A. marginale and its vacuoles in live cultured tick cells. A B C A A B B C C Figure 1 Visualization of A. marginale colonies within infected tick cells. The different panels represent the same cluster of A. marginale infected IDE8 tick cells at 1000x. The structures are similar to those described using thick sections with Mallory's stain under light microscopy or using ultrathin sections under a transmission electr on microscope. (Blouin et al ., 2000) A. Nomarski optics reveal the typical granu lar structure of the large colonies of A. marginale . The boundaries of the vacuoles are easily distinguished. B. The vital cell DNA stain SYTO13 conf irms that the gr anular structures contain DNA, as expe cted for prokaryotes. C. The same structures as seen in B c an be visualized in infected cells using phase contrast.


53 Individual cells containing single vacuoles were picked and transferred multiple times using a micromanipulator . Twenty two attempts were made to clone A. marginale by micromanipulation and only on one occasion was some propagation of the organism observed. This infected cell line was, however, lost due to contamination wit h a multiple antibiotic resi stant bacterial strain. None of the attempts to clone A. marginale by serial dilution either using the supernatant containing in itial bodies or infected cells were successful. N C V V V V CAB N C V N C V V V V C V V V CAB Figure 2 Visualization of individual A. marginale colonies within tick cells. Non-stained IDE8 tick cells infected with 3 or 1 colony of A. marginale as seen under Nomarski optics at 1000x. A. A partially ruptured IDE8 cell re veals protruding intact membrane lined vacuoles (V) filled with rickettsial organisms. Anaplasma initial bodies are seen as a granular structure, but under the microscope a di stinct and characteristic motility can be observed. In intact and ruptured cells the membranes lining the different colonies can be clearly seen (arrows). B. In this panel the parasitic vacuole (V) fills almost the entire IDE8 cell, marginalizing the cytoplasm (C) and the nuc leus (N). Simila r structures can be visualized using phase contrast. Using a micromanipulator it is possible to rotate the infected cells and make sure, l ooking from different angles, that only one colony is present. Single Cell PCR and the Reduction of Complexity PCR of the msp2 expression site was successfully carried out using whole infected tick culture DNA or a single infected cell as a template. The


54 single cell RFLP analysis did not reveal a consistent reduction in complexity when compared with the whole population which contained at least 7 variants in 17 plasmid clones analyzed. One singl e cell had only two RFLP variants and the major pattern was found in 13 out of 14 plasmid clones (Cell 4, fig. 3). One single cell contained only one RFLP pattern in 23 different amplicons tested (Cell 5, fig.3). Figure 3 Variability in msp2E of A. marginale in a population of infected IDE8 cells or in single cells. Variability in msp2E ( msp2 gene in expression site) was compared between a cell culture population of A. marginale (Total) and the popul ation of pathogens found in a single cell containing a si ngle membrane lined colony (Cell). Variability was determined by PCR amplification of msp2E , using whole cultures or single cells as a source of template , and RFLP analysis of cloned amplicons on agarose gels. The first lane in each panel is a 100bp DNA standard, the adjacent lanes represent the four mo st dominant RFLP patterns in each population of pathogens, wit h percentages given below and the number of amplicons used (n) above the panels.


55 Discussion Genetic heterogeneity of natural isolates of A. marginale and phenotypic dominance of subpopulations contribute to the difficulty of analyzing antigenic variation. In order to asse ss the capacity and mechanisms of A. marginale to vary antigenically, one must characteri ze the progeny of a single organism. That is, isolate a clone of A. marginale and characterize antigenic variation in subsequent infection cycles. For many bacteria, growing on plates of solid medium, cloning is fairly straight forw ard using the classi cal plating process invented by Robert Koch. For many intracellular protozoa and bacteria, growing in a tissue culture system, the solu tions are less straight forward. Two different approaches have been described. The first involves dilution to a precalculated probability of obtaining a single infected cell. A second method consists of microscopic ex amination of diluted suspens ions containing infected cells and selecting those with only one cell (Oduola et al. , 1988). Although cloning by micromanipulat ion requires expensive equipment and the development of special techniques it avoids some of the potential pitfalls inherent to limiting dilution tec hniques. A major obstacle in cloning A. marginale by micromanipulation is the small 0.5-2.5 µm size of the bacteria (Blouin et al. , 2000), making it impossible to vi sualize the individual organism under a light microscope. Our data demons trated, however, that it is possible to characterize cells containing one or multiple membrane lined colonies of initial bodies in live cells without staining. Since each vacuole is thought to originate from a single A. marginale initial body, cells containing only one colony


56 can be considered clonal and used for cloning by micromanipulation (Perez et al ., 1997). Micromanipulations of these single vacuole containing cells was carried out repeatedly, allowing one to “was h” the cells in fresh medium, before transferring them to new uni nfected tick cell cultures. Nevertheless none of the cloning efforts were successful in yi elding propagated cont inuous cultures. Success rates of cloning Plasmodium falciparum by micromanipulation have been shown to be as low as 15% (Beale, Thaithong, and Siripool, 1991). It is likely that the process for A. marginale is equally inefficient. The selection and manipulation of the infe cted tick cells, which often takes more then one hour, could further affect the viability of the pathogens. Alternatively the viability of pathogens from these extremel y large vacuoles might be inherently poor. The selection of these large colonies allo wed for a better identification of single colonies within cells, but they might be the phenotypic expression of a defective pathogen population. During the development of the micr omanipulation cloning procedure it was observed that A. marginale pathogens were visible in side their vacuole as a granular structure and these granules di splayed a characteristic motility. Previously motility of A. marginale had been described (Ferris, 1972; Kreier and Ristic, 1972) and active motility is an im portant virulence factor in several pathogens (Heinzen et al., 1999). Active motility could possibly also be used to assess the viability of the pathogens during cloning. Most motile bacteria move by the use of flagella, threadlike loco motor appendages protruding form the cell wall (Namba and Vonderviszt, 1997). But no such structures have been


57 described for any Anaplasmataceae. Ot her forms of motility are less well defined but often also involve energ y dependent bacterial motors. These motors have been shown to be inhibited within seconds by several energy metabolism blocking chemicals (Daniels, Longland, and Gilbart, 1980). None of the products tested here or the removal of glucose from the medium resulted in reduced motility, even after several hours of observation, suggesting another mechanism. A mechanism to generat e motility through continual polar polymerization of an actin tail can be found in Listeria monocytogenes , Shigella flexneri and Rickettsia rickettsii (Heinzen et al. , 1999; Rutenberg and Grant, 2001). A tail-like inclusion appendage, co mposed of multiple actin filaments, has been described in some isolates of A. marginale (Kocan et al. , 1978; Stich et al. , 1997). Actin based motility depends on continuous recruitment and polymerization of protein. Since oxytet racycline, which kills the pathogen inside vacuoles at the tested dose (Blouin et al ., 2002), was unable to block the motility even after 24 hours it was c oncluded that the observed movement within the vacuoles was likely Brownian motion. PCR and RFLP analysis of the msp2 operon confirms earlier observations that this site is polymorphic in an in vitro A. marginale population (Barbet et al. , 2001b). Micromanipulation and si ngle cell PCR analysis failed to produce consistent reduction in this comp lexity. It is possible that rapid recombination generates the observed co mplexity within one vacuole. This would appear to disagree with observa tions at the protein and DNA level indicating that MSP2 and other surface proteins are consistently expressed


58 over time in culture (Barbet et al. , 1999; Barbet et al. , 2001b). This contradiction could, however, indicate that while rapid recombination occurs within one cell the overall env ironment of the cu ltures exerts a strong selective pressure leading to a uniform and stable population. It has indeed been shown for A. phagocytophilum that different culture env ironments can exert strong selective pressure on the expression of the msp2 ortholog p44 , creating a nearly uniform population of pathogens (Barbet et al. , unpublished). These selective pressures could be at the base of our current inability to propagated clones selected by micromanipulation. Since single cell PCR suggests that single vacuoles can contain variants di fferent from the pr edominant population, but theses variants might not be fit for pr oliferation in the culture environment.


59 CHAPTER 4 MSP3 EXPRESSION AND GENETIC VARIATION1 Introduction The Family Anaplasmataceae includes tick-borne pathogens of the genera Anaplasma and Ehrlichia (Dumler et al. , 2001). Anaplasma phagocytophilum and Ehrlichia chaffeensis cause the recently emergent human diseases, human granulocytic ehrlichiosis (HGE) and human monocytic ehrlichiosis (HME), respectively, and A. marginale is the cause of a highly prevalent infection of cattle, anaplasmo sis. Common to all of the tick-borne pathogens in these genera is the requireme nt for persistent infection of a mammalian reservoir host to provide a sufficient number of circulating organisms to allow ticks to acquire the infection (Dumler et al ., 2001). Pathogen persistence within an imm unocompetent host requires immune evasion and studies on A. marginale within persistently infected cattle have demonstrated antigenic variation of outer membrane proteins (French, Brown, and Palmer, 1999). Our intere st is in understanding how A. marginale and other pathogens in the Family Anaplasma taceae generate antigenic variation of their surface antigens. 1 Patrick F.M. Meeus, Kelly A. Brayton, Gu y H. Palmer and Anthony F. Barbet (2003); Conservation of a gene conversion mechanism in two distantly related paralogues of Anaplasma marginale . Molecular Microbiology 47 .


60 The A. marginale outer membrane contains two highly immunodominant proteins, designated major surface protei n (MSP) 2 and MSP3. Eight out of 12 CD4+ T lymphocyte clones derived from ani mals immunized with purified outer membranes of A. marginale recognized MSP2, MSP3, or both (Brown et al ., 1998), and sera from outer membrane-imm unized or infected calves have the highest titers to MSP2 and MSP3 (Palmer et al. , 1986; McGuire et al. , 1991). Notably, expressed MSP2 and MSP3 are polymorphic, both structurally as assessed by using two-dimensional gel electrophoresis and antigenically (Alleman and Barbet, 1996; Blouin et al. , 2000; Kano et al. , 2002). Results obtained by Alleman et al . (Alleman et al ., 1997) and analysis of data obtained from the A. marginale genome project reveal t hat MSP2 and MSP3 likely originated from a common ancesto r (Meeus and Barbet, 2001), but the encoding genes have since diverged substant ially. The average identity at the amino acid level between MSP2 and M SP3 pseudogenes is 38%, within MSP3 pseudogenes 68% and within MSP2 ps eudogenes 78%. MSP2, which has closely related orthologs in A. phagocytophilum , A. centrale , and A. ovis , is a ~40 kDa protein with variants defined by a central hypervariable region flanked by highly conserved Nand C-termini (Eid et al. , 1996; Ijdo et al. , 1998; Palmer et al. , 1998; Murphy et al. , 1998; Zhi et al. , 1998; Shkap et al. , 2002). The hypervariable region spans ~100 ami no acids and bears surface exposed epitopes that are unique to the emergent variant and are not recognized by antibody prior to emergence (French et al. , 1998; French, Brown, and Palmer, 1999). Thus, MSP2 represents true ant igenic variation with sequential


61 emergence of unique MSP2 variants in the bovine host followed by immune control. MSP3 on the other hand is an 80-90 kDa protein with individual variants shown to differ by up to 15 kDa in molecular size, in charge, and by reactivity with specific antibodie s (Alleman and Bar bet, 1996; Alleman et al. , 1997; Blouin et al. , 2000; Kano et al. , 2002). Despite this difference at the protein level msp3 genes have maintained a general structure similar to msp2 genes with a variable region flanked by highly conserved sequence, although the length and the composition of the sequences are different. How does A. marginale generate the thousands of surface protein variants needed for long-term persistenc e? Gene conversion of a single, operon-linked, full-length msp2 using one of 9-10 truncated msp2 pseudogenes generates a new expressed variant contai ning the hypervariable region of the recombined pseudogene (Barbet et al. , 2000; Brayton et al. , 2002). Furthermore, short segments derived fr om the variable regions of the pseudogenes recombine, also through gene conversion, into the expression site msp2 to create novel hypervariable region mosaics (Brayton et al. , 2002). These combinatorial mechanisms c an generate the >10,000 MSP2 variants needed for persistence. While there must be a mechanism to generate MSP3 variants, how this is accomplishe d is a significant gap in understanding A. marginale persistence. Interestingly, t here are several observations that support a common mechanism for generating variation in both msp2 and msp3 , despite the sizeable differences at the pr otein level. First, the genome contains multiple msp3 coding sequences with hypervariabl e regions flanked by highly


62 conserved segments (Alleman et al. , 1997; Brayton et al. , 2001), the same general structure observed for msp2 . These msp3 copies identified to date are putative pseudogenes since they are trunc ated to varying degrees at their 5 and 3 ends, and are unlikely to encode an M SP3 product of ~86 kDa (Brayton et al ., 2001). Second, the msp3 conserved regions contain sequences homologous to msp2 (Alleman et al. , 1997, Meeus and Bar bet, 2001). Third, the msp2 and msp3 pseudogenes are often found in close proximity to each other and a similar ~600 bp region flanks many copies of both pseudogene families (Brayton et al. , 2001; Meeus and Barbet, 2001). In this chapter, we report the identification of a single MSP3 expression site and demonstrate the recombination of pseudogenes into the full-length expressed msp3 gene during A. marginale infection using a mechanism similar to msp2 . These two paralogous families have thus diverged s ubstantially, but nevertheless retained similar mechanisms to generate variation, each using a specific subset of paralogous genes as a source of variants. Experimental Procedures RNA Isolation, RT-PCR and 5 RACE A splenectomised calf (2153) was in fected intramuscularly (IM) with a stabilate (B442) of t he Florida strain of A. marginale . Total RNA was isolated from whole blood on Day 24 post infection (PI) during acute rickettsemia (42% infected RBCÂ’s) by extraction with 6 M urea-3M LiCl (Van der Ploeg et al ., 1982). Isolated RNA was digested with DNas eI (DNA-free; Ambion) before use in reverse transcription (RT)-PCRÂ’s and 5 RACE.


63 For RT-PCR msp3 mRNA transcripts were reverse transcribed into cDNA using the RETROscript kit (Ambi on) according to the manufacturer’s protocol and primer AB928 wh ich anneals to the conserved 3 region of msp3. Primary PCRs used oli gonucleotide primers AB754 and AB929, annealing in the 5 and 3 conserved regions, respectively , 2.5 U of SuperTaq polymerase (Ambion) and other reagents obtained fr om Ambion. Secondary PCR used nested primers in the conserved regi ons, AB927 and AB925, and 1/100 diluted products of the first round of amplificat ion. Control reactions were conducted similarly, but without reverse transcriptase in the initial RT reaction. The PCR conditions used were initial denaturation cycl e of 2 min. at 93ºC, followed by 35 cycles of 30 sec. at 93ºC, 30 sec. at 55.5 (1ºPCR) or 52ºC (2ºPCR), 2 min. at 72ºC and final extension of 10 min. at 72ºC. For 5 RACE reactions mRNA was reverse transcribed into cDNA as above but using a primer in the 5 conserved region, AB755. After first strand cDNA synthesis, the original mRNA template was removed by treatment with an RNase Mix and the cDNA purified using a GLASSMAX Spin Cartridge according to the manufacturers re commendation (Life Technologies). A homopolymeric tail was then added to the 3 end of the cDNA using TdT and dCTP. An aliquot of the reaction was directly amplified by PCR using SuperTaq (Ambion), a nested gene-specific pr imer (AB878) and a deoxyinosine and guanine containing anchor pr imer (Life Technologies). An additional round of PCR using the AUAP (Abridged Univer sal Amplification Primer, Life Technologies), in conjunction with a progr essively nested primer (AB877) and a


64 size-selected product from the first r ound of amplification, was required to obtain an adequate amount of cDNA to permit cloning. Products of RT-PCR and 5 RACE were analyzed by agarose gel electrophoresis and Southern blotting using msp3 -specific DNA probes. PCR products were cloned into the pCR-XL-T OPO vector (Invitrogen) and plasmid DNA was isolated, digested with EcoRI and analyzed by agarose gel electrophoresis and Southern bl otting. Plasmids with msp3 inserts were subsequently digested with EcoRI, to cut out the insert, and RsaI, to analyze the diversity of the inserts per size group. Selected plasmids were sequenced to verify the structure of the amplified cDNA. Southern Blotting of A. marginale Genomic DNA Fluoresceinated oligonucleoti de probes specific to the msp3 expression site (AB973) and the 5 and 3 conserved regions (AB934, AB892) were used in Southern blotting of digested A. marginale genomic DNA (Florida strain). A. marginale genomic DNA was isolated from hi ghly rickettsemic bovine blood (calf 115) by lysis with sodium dodecyl sulfate (SDS) and lysozyme, treatment with proteinase K and RNase, phenol-c hloroform extraction, and ethanol precipitation. DNA was digested usi ng EcoRI, FspI, KpnI and SphI (Gibco BRL), separated on a 0.5% agarose gel, and subsequently transferred to a nylon membrane. The blot was hybridized at 42ºC in 6X SSC, 10mM Na3PO4 pH 6.8, 1mM EDTA, 10X Denhardt's solution, 100ug/ml salmon sperm DNA, 0.5% SDS overnight and 5 fluorescein end-labeled oligonucleotides. The membrane was washed twice for 5 min. at room temperature in 1X SSC/1%


65 SDS, then with an additional four washes of 10 min. at 30ºC in 1X SSC/1% SDS, and followed by another two 5 min. washes in 0.1X SSC/1% SDS and by two 5 min. washes in Tris buffered saline. The probes were detected by chemiluminescence (Illuminator c hemiluminescent detection system; Stratagene). Library Construction and Screening Splenectomised calf 198 was infected with a Florida strain stabilate (2588) and DNA from the acute population was isolated, as above, on Day 18 PI when rickettsemia had reached 55%. Day 15 blood of calf 198 had a rickettsemia of 8.3% and was used to infect a non-splenectomised calf, 196. This calf was allowed to develop acut e anaplasmosis and recover in a large animal isolation facility. On day 107 PI animal 196 was splenectomised and developed relapse anaplasmosis. DNA from this persistent population of A. marginale was collected on Day 126 PI. Genomic A. marginale DNA libraries of acute ( 198) and persistent (196) populations were constructed in the co smid vector SuperCos 1 (Stratagene). Libraries were prepared by digesting DNA with Sau3AI to an average size of 35kbp and cloning into the BamHI site of SuperCos 1. Individual clones were picked into 96-well microtiter plates . Cosmid DNA was prepared, dot-blotted and fixed to nylon membranes, and hybrid ized with the expression site and the 3 conserved region probes as described for the Southern blots. Three cosmids positive for both probes, one for the acut e and two for the persistent population, were sequenced directly using primer walking. Sequencing of Cosmid 1C,


66 acute population, was suppl emented with PCR amplif ication and sequencing for a segment of the msp3 variable region. An A. marginale St. Maries strain Bacterial Artificial Chromosome (BAC) library was constructed as previously de scribed (Brayton et al., 2001). Briefly, isolated A. marginale cells were embedded in agaros e blocks, and lysed using proteinase K and SDS. A. marginale genomic DNA was partially digested with HindIII, size selected on pulse field gels, ligated into vector pBELOBAC11 and electroporated into E. coli strain DH10B . 1,536 BAC clones were arrayed into 384 well plates with an average insert si ze of 110Kb. BAC 16D3 was selected for sequence analysis after probing with a digoxigenin labeled (Roche) AB973. PCR Amplification and Restricti on Fragment Length Polymorphism Analysis of Size Selected msp3 Loci Genomic DNA, obtained as described above, was digested with EcoRI and run in two adjacent lanes on a 0.3% agarose gel as for the Southern blot analysis. One lane was used in a Souther n blot protocol as above, but with only a two hour hybridization. The other gel half was kept at 4º C. Southern blot exposures were aligned with t he agarose gel and DNA fragments corresponding with msp3 hybridization bands were excised the same day. An agarose gel segment below the smallest hybrization product was excised to be used as a negative control. DNA wa s extracted using Freeze and Squeeze extraction (fragments F1-F5) (Biorad) or S.N.A.P. purific ation (Invitrogen) (fragments F6-F7). Each DNA sample was used as a template in a PCR reaction with primers AB815 and AB984, located in the 5 and 3 conserved regions of msp3 respectively. Primer-d imer products were removed from the


67 PCR reaction using the QIAquick PCR Puri fication Kit (Qiagen) and amplicons cloned using the TOPO TA Cloning Kit for Sequencing or the TOPO XL PCR Cloning kit (acute and persistent populations respectively). For each amplified locus as many colonies as possible we re picked with a maximum of 96 for the loci outside the expression site and up to 186 for the expression site locus. Clones were grown up under kanamycin selection in a 96-well system and plasmids isolated, digested with EcoRI, to cut out the insert, and RsaI, to analyze the diversity of the inserts using 1.8% agarose gels. Primers and Probes The following primers and oligonucleotide probes were used : AB754 tggggagatggtaggagttga; AB755 tctaaccacttaagttctccaagt; AB815 gtggggagatggtagg; AB877 taattctt ctggtctgctcagttc ; AB878 gctttgctagttcatgttgtag; AB 925 tttaaataatcagtccaaccc; AB927 tacaagaactgagcagaccagaag; AB928 cgttcatcagcatccaag; AB929 gcccattttctaatcttctcag; AB984 ttgcagcattagctctcacttac; AB973 Fcagcttgcacactggagctataggacaagttaca (F = Fluorescein); AB934 Ftggggagatggtaggagttgatgaaggactagtta; AB892 Ftgttgtagaaggggttaggaaaagggttaggg. Western Blot Analysis Specific antiserum against the amino acid domain predicted to be common to MSP2 and MSP3 was generated in Balb/C mice by immunization with the synthetic peptide CSDYTGTAGKNK DT. The amino terminal cysteine was included for conjugation and 2 mg of peptide were cross-linked to


68 maleimide activated keyhole limpet hemocyanin (KLH) using the Imject Maleimide Activated Imm unogen Conjugation Kit (Pierce) according to the manufacturer’s recommendations. Balb/C mice were immunized by subcutaneous inoculation of 100 µg anti gen emulsified with complete Freund’s adjuvant. The mice were boosted using three immunizations of 100 µg antigen in incomplete Freund’s adjuvant. The pos itive control antibodies for MSP2 and MSP3 expression were, respectively, monoclonal antibodies ANAF19E2 and 43/23 (Palmer et al., 1994; Alleman et al., 1997). The negative control antibodies were normal mouse serum and antiTrypanosoma brucei monoclonal antibody Tryp1E1. A. marginale (Florida strain) was isolated from infected erythrocytes as previously descr ibed (Palmer and McGuire, 1984) using proteinase inhibition buffer (50 mM Tris [pH 8.0], 5 mM EDTA, 5 mM iodoacetamide, 0.1 mM N-p-tosyl-L-lysine chlorome thyl ketone, and 1 mM phenylmethylsulfonyl fluoride). Elec trophoresis was carried out on precast SDS-containing 4-20% polyacrylamide gel s (Biorad) for 30 min. at 200 V. Following transfer to nitrocellulose, expr ession was detected with the antibodies described above using the Western-St ar Chemiluminescence Immunoblot Detection System (Tropix) according to manufacturer’s instructions. Sera were tested at a final dilution of 1:200 and monoclonal antibodies at a final concentration of 2 µg/ml. Sequencing Sequencing was performed at the University of Florida DNA Sequencing Core Laboratory using ABI Prism dye terminator cycle sequencing protocols


69 developed by Applied Bio systems (Perkin-Elmer Corp.). The fluorescently labeled extension products were analyzed on a model 373 Stretch DNA Sequencer (Applied Biosyst ems). Nucleotide sequences were analyzed using the GCG programs (Genetics Computer Group, University of Wisconsin) available through the Biologica l Computing core facilities of the Interdisciplinary Center for Biotechnology Research at the University of Florida. For BAC sequencing, random shotgun libraries were constructed from partially digested BAC 16D3 DNA. The randomly generat ed fragments were size selected, cloned into pCRScript (Stratagene) and electroporated into E. coli strain XL1Blue. Insert DNA was sequenced us ing BigDye terminator chemistry on an ABI 377XL-96 instrument (Applied Bio systems). Data were assembled and analyzed using Sequencher (Genec odes) and PHRED & PHRAP software (University of Washington). When r equired, gene walking or direct BAC sequencing was performed to ensure a minimu m of 2X coverage with an overall average of 3X coverage. Databases URL addresses for unfinished genome s equences used were as follows : Ehrlichia ( Cowdria ) ruminantium : ; Wolbachia sp. : ; Ehrlichia chaffeensis : : Ehrlichia / and Anaplasma marginale : research_vmp/anagenome.


70 Results Detection of Variant msp3 Transcripts During Acute Infection Transcripts expressed in vivo during acute A. marginale infection were detected by RT-PCR using pr imers derived from the 5 and 3 sequences conserved among msp3 pseudogenes (fig. 4 & 5). Si ze differences among the resulting amplicons (~1350bp, ~1050bp an d ~850bp) were consistent with variation in molecular size previo usly observed among expressed MSP3 proteins (McGuire et al. , 1991; Alleman and Barbet, 1996). Subsequent cloning and analysis by RFLP revealed additional polymorphism within amplicons of the same size (fig. 7B, and results not shown). Examination of the cDNA sequences confirmed that the transcripts encoded msp3 with a central variable region (CVR) flanked by highly conserved 5 and 3 segments. These results establish that multiple msp3 transcripts are expressed in a population of A. marginale during acute infection. Identification of the msp3 Expression Site The chromosomal msp3 expression site was identified using 5 RACE (rapid amplification of cDNA ends ) combined with genomic sequencing and verified by detection of expressed protei n. To obtain information extending upstream of the CVR, RACE was us ed with primers located in the 5 conserved region (fig. 4). The largest amplicons obtained were cloned and four inserts sequenced. The 3 end of the transcripts was nearly identical to the previously described msp3-12 gene and flanking region obtained by screening an A. marginale expression library with a m onoclonal antibody against MSP3


71 (Alleman et al ., 1997). The transcripts contai ned the MSP3 N-terminal coding region and also two additional open reading frames (ORF) 5 of the ATG initiation codon for msp3 (fig. 4). Figure 4 A polymorphic msp3 expression site. Sequence information from three cosmid s (Florida) and one BAC (St. Maries), represented diagrammatically, revealed a similar but polymorphic msp3 expression site. In the msp3E block green represents conserved regions found in all available copies, red are regions with homology to msp2 , solid for near identity and dotted for less homology. Light blue is remaining ORF and the large grey block is the central variable region of msp3E . The relative positions of RT-PCR and 5 RACE products generated from msp3 mRNA and cosmids from genomic DNA from the Florida strain of A. marginale , are indicated. It also shows the location of probes AB973, AB934 and AB892, used in the genomic Southern blot analysis of the Florida stra in (fig. 5). The gray ellipses mark additional regions of sequence variability in the 5 flanking region, at the beginning and end of ORF X. ORF Y and ORF X are represented as two dark blue arrows and are direct repeats in tandem found flanking many msp2 and msp3 copies. Black arrows are ot her significantly long ORFÂ’s. Using the sequence from the 5 RACE analysis, specif ic oligonucleotides were designed to probe the genome. Southern blot analysis of A. marginale genomic DNA (Florida strain) using probes against the 5 and 3 conserved regions revealed an estimated 8 loci containing msp3 copies (fig. 5). In contrast, a single fragment was detected with the expression site probe AB973, derived from unique sequence at the 5 end of the msp3 polycistronic mRNA transcript. This fragment also hybridiz ed, as expected, with the probes against the 5 and 3 conserved regions confi rming that this expressed msp3 ( msp3E ) represents a genuine genomic structure. In addition , three cosmid clones


72 generated from different time-points (one acute and two persistent) during A. marginale Florida strain infection and a 118k b BAC clone, 16D3, derived from an acute infection with the St. Maries stra in hybridized with the expression site probe AB973. Sequencing of these genomic clones revealed a similar expression site with a full-length msp3E gene copy. Consistent with the RTPCR data showing transcription of msp3 genes that differed in the CVR, the expression site msp3E in each of the genomic clones contained a unique CVR (fig. 7). Interestingly, the 413 bp at the 3 terminal coding region of all the examined msp3E genes was >99% identical to the 3 end of the expressed msp2 . Antibody against a peptide in this C-terminal shared domain bound both MSP2 and MSP3 in A. marginale from an acute Florida strain infection (fig. 6), demonstrating that this domain was expressed in the MSP2 and MSP3 proteins. The binding to two MSP3 prot eins of different size (fig. 6B) is consistent with the CVR polymorphism ident ified by the transcr iptional analysis and the genomic sequences. Structure of msp3 Pseudogenes and msp3 Variation The expression site msp3 genes, encoding full-length MSP3 proteins, were compared with the msp3 pseudogenes in other chromosomal loci using A. marginale DNA obtained from eit her acute or persistent infection. The CVR from 7 different genomic loci of the Florida strain was obtained by PCR amplification and cloning of Eco RI fragments (fig. 5) us ing primers derived from 5 and 3 conserved regions. The 11kb Eco RI fragment (fig. 5) represents the full-length msp3E gene in the expression site and the remaining six loci contain


73 msp3 pseudogenes defined by the conserved 5 and 3 regions flanking a unique CVR. An additional pseu dogene only hybridized with the 3 conserved region probe and was not cloned. Clones fr om PCR amplificat ion of the 20.5kb Eco RI fragment displayed two distinct RFLP patterns, msp3-F2A and msp3F2B , indicating the presence of two msp3 copies, consistent with the intensity of this band on the Southern blots (fig. 5). 23130 6557 4351 2322 9416 bp124 3124 3124 3ABC 23130 6557 4351 2322 9416 bp124 3 124 3124 3 124 3124 3 124 3ABC Figure 5 Presence of multiple copies of msp3 , but a single hybridizing band from the msp3 expression site in genomic A. marginale DNA. Southern blots were done on rest riction enzyme digests of A. marginale DNA (Florida strain) using four different en zymes: (1) EcoRI , (2) FspI, (3) KpnI and (4) SphI. Blot A was hybridized with a probe identifying the msp3 expression site (AB973), blot B and C with o ligonucleotides specific for the 5 and 3 conserved region (AB934, AB892), respectively (see fig. 4 for location of probes). Approximately eight msp3 copies can be distinguished if bands of significant higher intensity are assumed to contain two copies.


74 Figure 6 Schematic r epresentation of the msp3 family of genes and its relationship to msp2 . A. Diagram of an expressed (E) and pseudogene (P) msp2 and msp3 copy. Both msp2 and msp3 feature a central variable region (gray) flanked by 5 and 3 conserved regions, with additional full le ngth features found in the expression site. Color legends are as fo r fig. 4-. A bp reference is given at the bottom. A * indicates the position of the peptide us ed to generate anti-MSP2/3 serum used in panel B. B. Western Blot of A. marginale showing the conser ved C-terminal end of MSP2 and MSP3 antigens. The antigen is Florida A. marginale and the antibodies are as indicated. The ant i-MSP2 MAb is ANAF19E2 and the antiMSP3 MAb is 43/23. The mouse anti -MSP2/3 antibody is polyclonal serum (dilution 1:200) against the C-terminal shared peptide. NMS is normal mouse serum, also at 1:200. The extra lane is a 5 fold increase of protein loaded to show that both MSP3 polypeptides rec ognized by the MAb are also recognized by the serum. The size difference is consistent with the size differences observed between the different msp3 copies at the nucleotide level.


75 Each of the six pseudogene loci, including msp3-F2A and msp3-F2B , contained a unique CVR sequence, with identity ranging from 68-79%. However, there was no change in the pseudogene sequences when compared between the acute and persistent A. marginale populations, indicating that the pseudogenes are invariant over time (fig. 7). In contrast, the expression site msp3E differed and increased in variability between the acute and persistent populations. The CVR sequences found in msp3E were identical to, or were a combination of, segments from the msp3 pseudogenes. A region of near sequence identity, ranging from 18 to 146 bp in length, was observed at the cross-over points between any two sequence segments found in the expression site. This observed heterogeneity in msp3E derived by PCR am plification from the Eco RI fragments agrees with information obtained by direct sequencing of the expression site cosmid clones deriv ed from the same acute and persistent A. marginale populations. The cosmid clone obtained from acute infection ( msp3E-cosAc1C , fig. 7A) had a CVR identical to the most common CVR found by PCR of the same A. marginale population. The two cosmid clones obtained from persistent A. marginale infection ( msp3E-cosPers9D and msp3EcosPers2E , fig. 7A) revealed, like the PCR amplicons, a mosaic CVR apparently derived from several msp3 pseudogenes (fig. 7A). The 3 end of the expression site CVR in cosmid msp3E-cosPers2E was only partially related to pseudogene msp3-F4 (fig. 7A). However, a msp3 pseudogene ( msp3-SM4 ) found in the St. Maries strain was identical to the msp3E-cosPers2E CVR 3 end and could represent a msp3 pseudogene not ident ified in the Florida strain.


76 Analysis of cloned RT-PCR products from msp3 mRNA similarly revealed identity with the CVR derived from one or more of the msp3 pseudogenes (fig. 7B). Comparison of msp3 Loci Between A. marginale Strains The sequence of five msp3 pseudogene loci and the msp3E expression site locus were identified in the St. Maries strain of A. marginale as part of a genome sequencing project (Brayton et al. , 2001; Brayton et al. , 2002). As in the Florida strain, the CVR in the St. Maries strain expression site msp3 ( msp3E-SM16d3 , fig. 4 and 7C) appeared to represent a mosaic derived from several of the msp3 pseudogenes. The CVR sequences of the five St. Maries strain msp3 pseudogenes ( msp3-SM1 to msp3-SM5 , fig. 7C) were compared with those of the Florida strain ( msp3-F1 to msp3-F6 , fig. 7A). With the exception of msp3-SM1 , which was identical to msp3-F6 , the St. Maries strain msp3 pseudogenes were different than those of the Florida strain (fig. 8A). However, an amino acid similarity plot comparing all CVR regions revealed a domain that is relatively more conserved flanked on both sides by smaller conserved regions (fig. 8B). These mo re conserved segments were maintained in the expression site msp3 and may represent s equences important for biological function or, alter natively, they may represent sites of recombination.


77 Figure 7 The generation of diversity in msp3 expression is achieved by insertion of different complete gene copies or segments into an expression site. A. Diagrammatic representati on of the different genomic msp3 loci identified by PCR amplification of size selected fr agments (F1-F6, see fi g. 5) or from genomic libraries ( msp3E-cos ) in populations from ac ute and persistent bovine infections with the Florida strain of A. marginale . msp3 pseudogenes were invariable between the two populati ons and are presented once. The msp3E locus was polymorphic and changed between acute and persistent infections, and is represented for each population with the percentages of clones with a given RFLP-pattern indicated. Simila r colors indicate identical sequences. B. Diagram of RT-PCR products generated from msp3 mRNA isolated from an unrelated population of A. marginale (Florida strain) during acute infection of a bovine host. RT-PCR-E is identical to genomic copy msp3-F2B flanked by conserved regions, but a significant segm ent of the variable region is deleted. The deleted fragment is delineated by two (3A) and one diamond (3B). C. Diagram of t he CVR of 5 genomic msp3 pseudogenes and an expression site copy ( msp3E ) in the St. Maries strain (SM). The shading in the msp3E reflects the sequence in the pseudogene copies. msp3-SM1 is identical to msp3-F6 and was given the same color code. The dark grey box in msp3E represents a region on ly distantly similar to the available msp3 CVRÂ’s.


78 Figure 7 (continued). Putative Genes Associated with the msp3 Expression Site Sequencing of the Florida strain cosm id clones and the St. Maries strain BAC clone containing the msp3 expression site revealed several associated ORFÂ’s. The two upstream ORFÂ’s, X and Y, are predicted to encode proteins of approximately 90 and 130 amino acids, respec tively. ORFÂ’s X and Y are similar to each other and form a tandem repeat in the expression site separated by 75 bp in the Florida strain and 72 bp in t he St. Maries strain. Both ORF X and ORF Y were shown to be present within the msp3 expression site transcript by 5 RACE and thus have the potential to be expressed. Interestingly, the 5 end of ORF X encodes an amino acid sequenc e similar to the predicted signal peptide of MSP3 and, like msp3E , ORF X is polymorphic in this locus among the Florida 5 RACE products, cosmids and the St. Maries BAC. The first variable region is immediately past t he predicted signal peptide and the second is at the 3 end of the ORF (fig. 4). These regions are characterized by the deletion, insertion, and r eplacements of several codon s that do not change the predicted reading frame. Neither ORF X nor Y had significant homology to


79 known genes or proteins. A putative promoter (-35 and -10 underlined) was found upstream of the start of the transcript (capitali zed), as defined by three out of four 5 RACE products: tttaca aaattttagcaactctgtgctgtaat gctgcc Agc. This promoter was also predict ed using the Neural Netw ork Promoter Prediction Tool ( oter.html) with a score of 0.97, but with the double-underlined cytos ine as a predicted transcription start point. Downstream from the msp3 expression locus, two large ORFÂ’s (1100 bp and 872 bp) were identified (fig. 4) that encoded putative homologs of membrane proteins HflC and HflK described in Rickettsia conorii (TBLASTX E value e -138 ) and Rickettsia prowazekii (TBLASTX E value e -126 ), as well as in other bacteria.


80 Figure 8 Comparison of the msp3 variable regions from two geographically distinct isolates of A. marginale (Florida and St. Maries). A. Dendrogram showing the clustering relationship between all available msp3 pseudogene copies of the Flori da and St. Maries isolates of A. marginale . Distance along the horizontal axis is proportional to the difference between sequences. Distance along the vertical axis has no significance. The dendrogram shows that certain msp3 pseudogene copies have been conserved between the two geographical isolates, wh ile other copies only have more distant homology to copies in the other is olate or are more closely related to a copy in the same isolate. B. A similarity score plot of an amino ac id alignment of all available variable regions of msp3 pseudogene copies from the Fl orida (6) and St. Maries (5) isolates of A. marginale . The plot used the translated pseudogene sequences depicted in fig. 7 and reveals the conserved 5 and 3 regions flanking the CVR as well as more conserved domai ns within the variable region.


81 Discussion A. marginale expresses antigenically and st ructurally distinct MSP3 proteins during infection of cattle, the mammalian reservoir host (Alleman and Barbet, 1996; Alleman et al. , 1997), and in tick cell culture (Blouin et al. , 2000). The data presented in this study identify recombination of msp3 pseudogenes into a single expression site as a mec hanism to generate MSP3 variants. Only the expression site msp3 encodes the complete MSP3 protein, including the Cterminus common to both MSP2 and MSP3. Most notably, ex amination of the expression site msp3E at different time points in infection with the A. marginale Florida strain and between the Florida and St. Maries strains clearly demonstrated variation in the CVR. The variation in the expressed msp3 was characterized by size differences as great as 543 bp and CVR sequence identity as low as 68%. This variation is consistent with the up to 15 kDa size difference observed in expressed MSP3 proteins (Alleman and Barbet, 1996; Barbet et al. , 1999; Blouin et al. , 2000; Kano et al. , 2002). An increase in variability of the expression site msp3 was observed between the acute and the persistent populations of Florida strain A. marginale as the predominant msp3E (87% of the analyzed clones) in acute infection ( msp3E-AcB , fig. 7A) was replaced by a heterogeneous mixture of va riants in the persistent population with no individual variant representing more than 24% of the analyzed clones. The remaining msp3 loci are pseudogenes, each with a unique CVR and most often flanked by conserved r egions. None of t he pseudogenes, with known flanking sequences, encode the N-terminal end and only one ( msp3-


82 SM2 ) encodes the C-terminal end (Brayton et al. , 2001; Brayton et al. , 2002). Southern blot and PCR analysis i ndicate the presence of 6-7 msp3 pseudogenes in the Florida strain genome. This is consistent with data from sequencing the St. Maries strain that has identified six pseudo genes in the 86% of the genome completed to date. Fo r five of the St. Maries strain msp3 pseudogenes both 5 and 3 flanking regions are known (fig. 7C), there is only partial sequence of the sixt h pseudogene, which is not s hown. In contrast to the variation in the expression site msp3E , examination of the msp3 pseudogenes from different populations of the same strain revealed no changes. These observations support a mechanism of gene conversion whereby invariant msp3 pseudogenes recombine into the single expression site to create a unique, expressed msp3E . Recombination appears to involve either the entire CVR, as illustrated by pseudogene msp3-F4 recombining into the expression site clone msp3E-cosAc1C (fig. 7A), or segments within the CVR to create unique mosaics in the expression site msp3 (e.g., msp3E-AcC , fig. 7A). Comparison of the msp3 pseudogenes with the sequence of the expression site msp3EÂ’s support both mechanisms and both have been demonstrated in msp2 variation (Brayton et al. , 2001; Brayton et al. , 2002). The recombination appears to be mediated by the regions flanking the CVR and conserved among all gene copies and, for recombination of segments, through shorter stretches of homology within the variable regions. The presence of a single msp3 expression site accompanied by multiple truncated pseudogenes dispersed th roughout the chromosome appears


83 common among all examined A. marginale strains. Southern blot analysis of the South Idaho and Oklahoma strains al so indicated the presence of 5-8 msp3 genes per genome (data not shown). However, alignment of the msp3 pseudogene sequences of the Florida and St. Maries strains indicates that, with the exception of the identical msp3-F6 and msp3-SM1 pseudogenes, the pseudogenes differ between strains (fig . 8A). This indicates that msp3 pseudogenes do undergo genetic changes, although most likely at a low frequency. A similar divergence in pseudogenes encoding surface antigens was observed between two strains of Neisseria gonorrhoeae (Hamrick et al ., 2001). While general features, such as the presence of conserved domains and the organization of the pil loci, were similar, none of the sequences were identical between the two N. gonorrhoeae strains throughout the entire length of a pseudogene copy. In A. marginale , alignment of MSP3 CVR amino acid sequences encoded by the different pseudo genes reveals short stretches of semi-conserved regions (fig. 8B). Al though not absolutely conserved, there may be structural constraints related to MS P3 function that lim it the degree of variation in these CVR domains. As previously noted (Palmer et al. , 1994; Alleman et al. , 1997; Brayton et al. , 2001; Meeus and Barbet, 2001), both immunodominant A. marginale surface proteins, MSP2 and MSP3, ar e encoded by multi-gene families with shared structural features. The present study further defines the similarity in msp2 and msp3 genetic structure and indicate s that variation in both is generated by gene conversion. Although t he presence of a single expression


84 site and multiple truncated pseudogenes dispersed throughout the genome is common to msp2 and msp3 , each uses distinct se ts of pseudogenes. The shared C-terminal region in MSP2 and M SP3 is due to the presence of this coding region in each expression site ra ther than recombi nation between the two gene families and there is no evidence to date that msp2 pseudogenes recombine into the msp3 expression site or vice-versa (fig. 6). Together this makes these paralogous gene families truly unique. After an original duplication event the genes diverged s ubstantially, creating antigenically different proteins of 40 and 80 kDa. At least one, possibly both, gene families underwent additional duplic ation events creating the two gene families. Each gene family, however, retained a similar mechanism to create diversity and a highly conserved C-terminal end in the expression site. Wi thin a population of organisms, whether in cultur e, erythrocytes or in ti ck vectors, MSP2 and MSP3 are always co-expressed (Barbet et al. , 1999). In situ surface protein crosslinking experiments indicate t hat MSP2 and MSP3 are nearest neighbors in the pathogenÂ’s outer membrane and are thus co-expressed in individual pathogens (Vidotto et al. , 1994). Yet both proteins seem to undergo simultaneous variation during the course of a bovine infection. The same populations used here to demonstrate msp3 variation have revealed a similar degree of msp2 variation during the course of an infection (Brayton et al. , 2002). Whether there is any regulatory linkage between msp2 and msp3 recombination remains to be determined, but this is not unlikely. The msp2 and msp3 pseudogenes are often found in close pr oximity to each other, in a tail to


85 tail orientation (Brayton et al ., 2001). In addition, many of the msp2 and msp3 copies, including the expression site msp3 , are flanked by a similar tandem repeat composed of ORF’s X and Y (fig. 6A ). This tandem repeat could play a role in recombination or, alternativel y, ORF’s X and Y could be part of yet another polymorphic multigene family ex pressed on the same transcript as msp3 . Interestingly, TBLASTN s earches against the unfinished Ehrlichia ( Cowdria ) ruminantium , Wolbachia endosymbiont of Drosophila melanogaster , and Ehrlichia chaffeensis genome databases revealed several ORF’s homologous to ORF X in all species (E values : 1.2e -27 –4.7e -23 ; 2.2e -12 –1.1e 12 ; 4e -33 –6e -28 respectively), suggesting its conservation and importance within this family of pathogens. Uncove ring how these multigene families, such as msp2 and msp3 , are regulated and allow the surface antigenic variation required for immune evasion will be a ke y step in understanding persistence in the mammalian reservoir host.


86 CHAPTER 5 ANALYSIS OF THE GENOMIC STRUCTURE OF THE MSP3 AND RELATED GENE FAMILIES Introduction Organisms infecting mammals enjoy abundant nutrients in a temperature controlled, buffered environment. This luxury comes at a price: continuous assault from a formidable immune system. Many pathogens use one or more elaborate strategies to evade these defenses . Antigenic variation is often found in vector-borne pathogens (Barbour and Rest repo, 2000). In protozoa, variation may be achieved by switching expre ssion between one of many different complete genes encoding a variable anti gen. Bacteria must make more efficient use of their small genomes if t hey are to have an extensive repertoire of variants. Some organisms that cause disease in animals and humans have been receiving much attention lately. These pathogens of the family Anaplasmataceae include Anaplasma marginale , the causative agent of anaplasmosis in cattle, Ehrlichia (Cowdria) ruminantium , which causes heartwater disease of ruminants, and E. canis and A. platys causing infections of dogs. Also included are emerging ti ck-borne infectious diseases of humans caused by E. chaffeensis (human monocytic ehrlichiosis) and A. phagocytophilum (human and animal granulocytic ehrlichiosis) (Dumler et al. , 2001). Despite their small, circular genom es of 1.2 to 1.6 Mb these organisms


87 nevertheless have the ability to persist for long time periods in their mammalian hosts by expression of different surface antigens. The increas ing availability of sequence from Anaplasma and Ehrlichia genome sequencing projects has provided us with leads as to how th is can be achieved from these small genomes. Following infection of naive cattle with A. marginale , an acute infection develops characterized by an increase in number of erythr ocytes containing pathogens and typically, a severe anemia. If the animal survives, the anemia resolves, clinical symptoms disappear and infected erythrocytes become difficult to detect. However, more sensitiv e indicators of infection, such as PCR, continue to detect A. marginale and show that the number of organisms cycles, with a periodicity of 4-8 weeks (Palme r, Brown, and Rurangirwa, 2000). The reasons for the persistent infection are becoming clear following investigations into the structural diversity of MSP2 and MSP3. MSP2 and MSP3 are polymorphic proteins with a molecular weight of about 40,000 Da (MSP2) and 80-90,000 (MSP3). Two families of genes, each comprising >1% of the rickettsial genome, encode these proteins (Palmer et al. , 1994a; Alleman et al. , 1997). It was observed earlier that the msp2 and msp3 gene families are related, with one partial genomic copy of msp3 containing an msp2 gene segment (Alleman et al. , 1997). Also, T-helper cell clones recognizing A. marginale often respond to both MSP2 and MSP3, showing shared T cell epitopes (Brown et al. , 1998b). The publication of 44.5 kbp genomic information by Brayton et al. not only confirmed this relationship, as a similar


88 msp2 segment was found in two additional msp3 pseudogenes, but also shows that the structure of the two gene families is similar and their genomic organization closely linked (Brayton et al. , 2001). Finally, characterization of the MSP3 expression site revealed t hat the entire MSP2 C-terminal end was conserved in the expressed msp3E copy. Both systems seem to employ a similar mechanism to create genetic divers ity despite the substantial differences at the nucleotide and amino acid leve l between these two major antigens. While we begin to understand the genetics of these processes little is known about how the processes are mediated. The availability of several hundred kb of sequence information from the ongoing A. marginale genome sequencing project provided the opportunity for furt her investigation into the mechanisms involved and the development of testable hypotheses. Experimental Procedure Data Three complete BAC sequences, G11, E6 and A10, were obtained from the A. marginale genome sequencing project webpage : mp/anagenome. BAC 16D3 contains the msp3E expression site locus for the St. Maries strain and >8000 bp surrounding msp3E was available for analysis as discussed in the previous chapter. Different sequences from the Florida strain msp3E locus were obtained, again as described previously, from four different 5Â’RACE RT-PCR clones and three different genomic cosmid clones.


89 Analysis Repeats within the BACÂ’s were characterized using the REPuter program (http://bibiserv.te STARTUP) (Kurtz et al. , 2001), MACAW (Multiple Alignment Construction & Analysis Work bench) (Schuler, Altschul, and Lipman, 1991) and BLAST searches on a local A. marginale server ( Further analysis was done using PILEUP, PRETTY, BESTFIT, GAP, DI STANCES, REPEAT and STEMLOOP programs on the Wisconsin GCG package Version 10.3 (Accelrys Inc, San Diego, CA) available through the Biol ogical Computing Facility of the Interdisciplinary Center for Biotechnology Re search at the University of Florida. Promoter prediction was done using the Neural Netw ork Promoter Prediction program (http://www.fruitfly.or g/seq_tools/promoter.html). Gene Homologs To find gene homologs involved in DNA repair and recombination annotated sequences from the Rickettsia conorii and R. prowazekii genomes were used (http://www.ncbi.nlm.nih. gov/entrez/query.fcgi?db=Genome). Additional orthologs discu ssed here originated from Escherichia coli obtained from the NCBI protein database ( entrez/query.fcgi?db=Protein).

PAGE 100

90 Results Genomic Structure of msp2 and msp3 Gene-Families Database mining of finished and unfinished BAC sequences of the A. marginale genome project reveals several BACÂ’s containing sequence with significant homology to known msp2 and msp3 genes. The database contains 12 continuous sequences, totaling 909,765 bases out of an estimated ~1.2 Mb in the genome or > 75%. The actual sequence of BACÂ’s G11, E6, A10 and a section of 16D3 were available for furt her analysis. BACÂ’s G11 and E6 contain many of the msp2 and msp3 genes. The msp2 and msp3 expression sites are found on E6 and 16D3 respectively. BAC A10 does not contain any sequence with close relationship to the msp2-msp3 gene families. Both G11 and E6 contain multiple msp2 and msp3 related genes; some of these genes are found in a pseudogene complex with msp2 and msp3 in a tail to tail orientation, while other g enes are found independently from the other gene family (fig. 11 & 12). A sequence sim ilar to the 5Â’ flanking region and the N-terminal coding region of the expressed msp3 gene is found downstream from the msp2E operon on the opposite strand. In addition to the classical msp2 pseudogenes, characterized by unique CVR flanked by specific conserved regions, several clusters of msp2 -like sequences ( msp2lk ) can be found distributed over both BACÂ’s. T hese sequences have a more distant relationship to msp2E and its associated pseudogenes. Most msp2lk ORFÂ’s have, however, a struct ure similar to the msp2 gene family, with a CVR flanked by conserved regions. These conserved regions, although often conserved

PAGE 101

91 within a cluster of msp2lk genes, are different from those found in msp2 . Many of the msp2lk regions have long ORFÂ’s with Cterminal regions similar to msp2E and msp3E , they have putative start site s and no frame shifts. Other ORFÂ’s, e.g. msp2lk-5 and msp2lk-8 , have homology to msp2 , but are not characterized by the same structural features and some have frame shifts within their reading frames (fig. 9). Figure 9 msp2 and related ORFÂ’s. This amino acid MACAW alignment represents the relationship of msp2E with its pseudogenes and several related msp2lk ORFÂ’s found in the A. marginale genome. Shaded blocks indicate sequence sim ilarity of the different ORFÂ’s with msp2E . Darker colors indicate a more frequent occurrence. msp2 pseudogenes, msp2-sm1 through sm5 , are clearly characterized by truncated ORFÂ’s with highly conserved 5Â’ and 3Â’ regions flanking a CVR. msp2lk ORFÂ’s have conserved a large part but not all of the 5Â’ conserved region and none of the 3Â’ conserved region. On the other hand most msp2lk ORFÂ’s appear fulllength and have retained homology to t he MSP2 Nand C-terminal ends. Finally msp2lk5 and msp2lk-8 are more distantly related ORFÂ’s with only short stretches of homology to MSP2.

PAGE 102

92 While closely linked msp2-msp3 pseudogenes are persistently encoded on opposite strands there is no obvious overa ll bias as to which strand either of the gene families encodes (fig. 10). In BAC E6 the two msp2 pseudogenes are found on the opposite strand compared to the expressed copy and the msp3 pseudogene is on the same strand as msp2E . This trend is, however, not continued in BAC G11. Two msp2 and one msp3 pseudogenes are on one strand, while one msp2 and three msp3 are found on the other. MSP4, another MSP2 paralog, is also encoded on the latter. Database analysis reveals that the MSP2 and MSP3 expression sites ar e separated by 64,829 bp (from stop to start codon) and the ORFÂ’s are ori ented in the same direction. Several genes that could play a role in DNA repair and recombination of msp2 and msp3 are also present on BAC G11 and E6. RecA, RecF, RecJ and RecG orthologs are found on G11 and E6. No homologs to the genes of the RecBCD pathway were found, however, in the A. marginale database. G11 also contains a XerC/D related gene, an integrase / recombinase associated with site specific recombination. Figure 10 Orientation of MSP expr ession sites and pseudogenes. The relative orientation of the MSP expression sites and msp3 pseudogenes. msp3 is represented in green, the expre ssion site by a full arrow and the pseudogenes by a triangle indicating thei r respective orientations. Also represented are flanking recombinase homologs (light blue) known to be involved in DNA repair and recombination.

PAGE 103

93 Each pseudogene or pseudogene complex is flanked by a similar conserved region, ORFX, at the 5Â’ or 3Â’ end. The reading frame of this ORF is always on the opposite strand, N-terminal to N-terminal, as the pseudogene to which it is closest. The msp3 expression site is an exception to this since ORFX forms direct tan dem structure with the msp3E gene. Often, but not always, ORFX is repeated in tandem to form the ORFX-Y flanking complex, also found in the expression site. Figure 11 BAC G11 repeats and recombinase genes. BAC G11 contains multiple msp2 (red) and msp3 (green) related genes, as well as homologs to recombinase genes (purpl e). ORFX-Y is represented by a blue arrow. A bp size references is given under the sequence.

PAGE 104

94 Figure 12 BAC E6 repeats and recombinase genes. BAC E6 contains the msp2 operon (light blue operon associated genes (opag) and red arrows), multiple msp2 and msp3 pseudogenes and recombinase related ORFÂ’s (purple). Also present is a duplication of the 5Â’ end and flanking sequence of msp3E (dotted green). ORFX-Y is r epresented by a blue arrow. msp2-msp3 Pseudogene Structure A total of five msp3 pseudogenes, five msp2 pseudogenes and at least 10 msp2lk ORFÂ’s were discovered on BACÂ’s G11 and E6. Alignment of all msp3 pseudogenes reveals a similar st ructure with a CVR flanked by conserved regions unique for msp3 . Four pseudogenes have a ~170 bp 5Â’ region in their ORF that is near identic al (97% identity) to sequence found in msp2E and msp2 pseudogenes. One pseudogene, msp3P-SM5 does not have this region. Additionally, the msp3E sequences on BAC 16D3 and those identified in the Florida st rain have a similar but more divergent 5Â’ shared region with msp2 with only 67% identity. The average identity at the amino acid level between MSP2 and MSP3 pseudogenes is 38%, within MSP3 pseudogenes 68% and within MSP2 pseudogenes 78%. There is, however, a wide range of

PAGE 105

95 amino acid identities, mainly based on the presence or absence of msp2 related sequence at the 5’ and / or 3’ end. In the absence of the obviously msp2 related regions, as for msp3-SM5 , the percentage identity with msp2 pseudogenes is as low as 22%. In msp3-SM2 , which has an ORF containing the entire C-terminal end of the expressed msp2 and msp3 genes, the identity with msp2 pseudogenes is as high as 60%. Other msp3 pseudogenes – SM1 , SM2 and SM4 , also share a homologous sequence with msp2 at the 3’ end of their ORF’s. This sequence, encoding t he amino acids VAGA(F), characterizes the transit of the CVR to the 3’ conserved region in msp2 . Within the msp3 pseudogenes the CVR can be as much as 624 bp different in size, while only minimal changes are noted in the msp2 pseudogenes. Within the msp2 and msp3 pseudogenes it is apparent that certain genes are more closely related to each other than to other genes, both in the ORF and in the flanking sequence. msp3-SM4 and msp3-SM5 are the closest related msp3 pseudogenes with an overall nucleotid e identity of >89%, and a nearly identical 3’ CVR. Their 3’ flankin g sequence has however maintained a >99% identity over 1311 bp, whic h includes the flanking msp2 pseudogene. The region upstream from these related msp3 pseudogenes does not have any significant homology. An inverse situati on is seen in two other closely related msp3 pseudogenes, i.e. msp3-SM1 and msp3-SM2 . While the genes again share a very similar 3’ CVR (90% identit y at nucleotide leve l), the 3’ flanking region is different, but the 5’ flanking r egion is near identical for over 815 bp. Also msp3-SM2 and SM3 do not share sequence imm ediately upstream of

PAGE 106

96 their ORF, but they do have homologo us sequence more upstream in their flanking regions. Finally, all msp3 pseudogene CVRÂ’s are characterized by a more conserved region in the center (fig. 13). Recombination Hot Spots and Chi-like Sequences Several recombination hot spots are present in the msp3 expression locus. Not only are different CVR pr esent in the expressed msp3E copy, additional sites in the 5Â’ flanking regi on undergo recombination. Two regions, one immediately past the predicted signal peptide in ORFX and one at the end of ORFX, are characterized by substi tution, deletions and insertions when clones obtained from DNA and RNA are com pared. While recombination within the Florida strain is substantial, the difference with the St Maries locus is characterized by a ~190 bp insertion between ORFX and msp3E. Sequence alignment of these different expression sites reveals a Chi-like sequence close to this 3Â’ recombination hot spot (fig. 14). Analysis and database searching reveals that this short sequence is al so repeated numerous times in regions flanking the msp2 and msp3 pseudogenes. In contrast the Chi-like sequence is not found in BAC A10, which does not have any msp2-msp3 pseudogenes. Alignment of all regions containing this Ch i-like sequence shows that it is part of a larger 15 nucleotide repeat, agagagc TGTGGTGG, with the small cap nucleotides not always present. The st atistical chance that the conserved decamer occurs is ~1/106 bp. An alternative version containing a thymidine at position 12 occurs (underlined). Thes e Chi-like sequences often are found where stretches of similar or hom ologous sequences diverge (fig. 14).

PAGE 107

97 A B 1 50 msp3p-sm1 SMLTALEGSI GYSIGGARVE VEVGYERFVI KGGKKSNEDT ASVFLLGKEL msp3p-sm2 SMLTALEGSI GYSIGGARVE VEVGYERFVI KGGKKSNEDT ASVFLLGKEL msp3p-sm3 YRSGRARVE VGIGHERFVI KGG....DDT A..FLLGREL msp3p-sm4 GSI GYRIGGARVE VGIGHERFVI KGG....DDA A..FLLGREL msp3p-sm5 51 100 msp3p-sm1 AYDTARGQLL SSALGRMSMG DVRRLKKEVV GSIGRGTASP VRAMFSRKIS msp3p-sm2 AYDTARGQLL SSALGRMSMG DVRRLKKEVV GSIGRGTASP VRAMFSRKIS msp3p-sm3 ALDTARGQLL SSALGRMSMG DVHRLKKEVV GSIGRGTASP VRAMFSRKIS msp3p-sm4 ALDTARGQLL SSALGRMSMG DVHRLKKEVV GSIGRGTASP VRAMFSRKIS msp3p-sm5 SIGRGTASP VRAMFSRKIS 101 150 msp3p-sm1 DGDTLLAGEM VGVDEGLVIQ ELSRPEELEK LQHELAKQVS KLAELGELKW msp3p-sm2 DGDTLLAGEM VGVDEGLVIQ ELSRPEELEK LQHELAKQVS KLAELGELKW msp3p-sm3 DGDTLLAGEM VGVDEGLVIQ ELSRPEELEK LQHELAKQVS KLAELGELKW msp3p-sm4 DGDTLLAGEM VGVDEGLVIQ ELSRPEELEK LQHELAKQVS KLAELGELKW msp3p-sm5 DGDTLLAGEM VGVDEGLVIQ ELSRPEELEK LQHELAKQVS KLAELGELKW Figure 13 msp3 pseudogenes share many features in their coding and flanking regions. A. A diagrammatic representatio n of the St Maries strain msp3 pseudogenes reveals that msp3 pseudogenes are characterized by highly conserved blocks of sequence (dark green boxes) flanking a CVR (grey block), which has a more conserved region (dotted green box) in the middle. Additional but more variable shared features include a flanking ORFX repeat (blue arrows), regions of msp2 homology at the 5Â’ and / or 3Â’ end (red boxes) and flanking msp2 pseudogenes (red and grey boxes) or msp2lk ORFÂ’s (red dotted arrows). Blue boxes are remaining ORFÂ’s. Other regions of homology are represented by color dotted diamonds. A gap was introduced in the CVR of msp3-SM1 and msp3-SM4 to illustrate the homol ogy of their 3Â’ CVR with pseudogenes msp3-SM2 and msp3SM5 respectively, as highlighted by si milar xand t-hashed CVR segments. The msp3-SM5 5Â’ flanking region contains a sequence near identical to the 5Â’ coding and flanking region of msp3E in opposite orientation (blue, red and green box). B. The amino acid alignment of the ORFÂ’s of msp3 pseudogenes reveals the extent of the homology in t he conserved regions at the edges and in the middle of the CVR. Red shaded areas have homology to MSP2, dark green are the conserved flanking regions and li ght green is the conserved region in the center of the CVR.

PAGE 108


PAGE 109


PAGE 110

100 A B msp3P-SM1 ttggtaccac ctacaaaccc aagtcccaac ttctggcacc tcctgccctt msp3P-SM2 tacccctggc acctcagccc aacttggcac ga CCACCACA GCTCTCT gca msp3P-SM1 ggcactcaac ccaactcatt acccagaccg CCACCACAGC TCTCT gcacc msp3P-SM2 ccatcactac ccagcaccgg gcttgaccgc tgtatgctgt gaacagtcaa msp3P-SM1 atcactaccc agcagcacct aatcttctat cggttacttt gggcagtgcc msp3P-SM2 acttactggc atcacctatg ccttcaccac taagtccggc ctccttgccc msp3P-SM1 ttt atgactg gcacagtcaa acttaccagc attctccatg ccttcaccac msp3P-SM2 agt atgactg gcacagtcaa acttatcggc atccttcaca ccttcaccac 1165 bp (99% identity) -> C <176 bp (97 % identity) msp3P-SM2 ctataaactc ctggcatccc agtctaac tt ggtaccacac agcatctccc msp3P-SM3 ctataaactc ctggcatccc agcccaac ct g CCACCACAG CTCTCT gcac msp3P-SM2 aacctctggt acctcctcct gcccttggca ctctcccaa msp3P-SM3 cggtatcacc gacataacat aaccgccagc acctcatct Figure 14 Recombination hot-spots surrounding msp3 genes. A. Diagram of regions of homology and divergence surrounding the St Maries msp3 pseudogenes. The boxes and letters designate the position of the sequences in panels B,C and D. Green arrows represent msp3 , red msp2 , blue ORFX and black the Chi-like sequence. Dotted diamonds represent region of homology outside the pseudogenes and ORFX repeats. B. Homology (shaded boxes) in the 5’ coding and 5’ flanking region of msp3 pseudogenes msp3-SM1 and –SM2 diverges towards the end of the ORFX-Y repeat. A Chi-like sequence (underlined and bold) can be found just upstream. C. A similar Chi-like sequence is pres ent down stream fr om a region of homology in the msp3-SM2 and msp3-SM3 5’ flanking region. D. Similar features are observed in the 3’ flanking region of msp3-SM4 and – SM5 . E. Interestingly this is also a feature found in the msp3E expression locus. ORFX, the second part of the ORFX-Y repeat, is characterized by recombination at the 5’ (not shown) and 3’ end. Again, homology diverges towards the end of the open reading fram e of ORFX and a Chi-like sequence is present in all expression site sequences obtained from the Florida and St Maries isolate despite the variation in that region.

PAGE 111

101 D <2312 bp (98% identity) msp3P-SM4 ttacagctat attctttggt gctggtaag g cttctactca tgctactcat msp3P-SM5 ttacagctat attctttggt gctggtaag c ttatggctaa atttgctgct msp3P-SM4 actggcaagg atggtattgg cagcaaccct tacggttgct gcgtgttgta msp3P-SM5 gggttgagtg gtgacggtat ggataaggcc gagaactttg actgtgccac msp3P-SM4 gggccagaag gttttggtgg cgcggatgtg ctgtgtttgg tcgcgcggtg msp3P-SM5 agtgcgtaat aagagtacca agtaggtgat ggtgcagg GA GCTGTGGTGG msp3P-SM4 agtggcttgc gacaagtctg tgcaaaattc ctttgtatac ggcgagtatt msp3P-SM5 tgccaggagt ttggagatat gggcttgggt tattgtcggg tggtaccaag msp3P-SM4 gttatatgct tgactctcgc ttttcgtggc cagggttttg agaggagttg msp3P-SM5 agcttcgaat ggtgc AGAGA GCTGTGGTGG cattgtagta agttgggctg E _________________________ORFX_________________________ 1 50 msp3E-F5r9 TTTGCTGCTG GGTTGAGTGG TGAGGGTGTG AAGGA...TG CCGGTAGCTT msp3E-F5r16 TTTGCTGCTG GGTTGAGTGG TGAGGGTGTG AAGGA...TG CCGGTAGCTT msp3E-F5r15 TTTGCTGCCG GGTTGAGTGG TGAGGGTGTG AAGGA...TG CCGGTAGCTT msp3E-F5r14 TTTGCTGCTG GGTTGAGTGG TGAGGGTGTG AAGGA...TG CCGGTAGCTT msp3E-FcosAc1C TTTGCTGCTG GGTTGAGTGG TGAGGGTGTG AAGGA...TG CCGGTAGCTT msp3E-FcosPers9D TTTGCTGCTG GGTTGAGTGG TGAAGGCATG GAGAA...TG CTGGTAAGTT msp3E-FcosPers2E TTTGCTGCCG GACTTAATGG TGAAGGCATA GGTGA...TG CCAGTAAGTT msp3E-SM16d3 TTTGCTGCTG GATTGGGCGA TGTGGGAGGA GGAACGGATG CCAGTAAGTT _______________________________________________ _ _ _ 51 100 msp3E-F5r9 TGACTGTTCC AGTTATAAAG G......... .......... .......... msp3E-F5r16 TGACTGTTCC AGTTATAAAG G......... .......... .......... msp3E-F5r15 TGACTGTTCC AGTTATAAAG G......... .......... .......... msp3E-F5r14 TGACTGTTCC AGTTATAAAG G......... .......... .......... msp3E-FcosAc1C TGACTGTTCC AGTTATAAAG G......... .......... .......... msp3E-FcosPers9D TGACTGTGCA AGTTATAAGG GTA....... .......... .......... msp3E-FcosPers2E TGACTGTGCC AGTCATAAAG GTAGCAGTGG TGCTGCTGGG TAGTGATGGT msp3E-SM16d3 TGACTGTGCC AGTCATAAAG GTAGCAGTGG TGCTGCTGGG TAGTGATGGT 101 150 msp3E-F5r9 .......... .......... .......... ........CA CTGCTAGGCA msp3E-F5r16 .......... .......... .......... ........CA CTGCTAGGCA msp3E-F5r15 .......... .......... .......... ........CA CTGCTAGGCA msp3E-F5r14 .......... .......... .......... ........CA CTGCTAGGCA msp3E-FcosAc1C .......... .......... .......... ........CA CTGCTAGGCA msp3E-FcosPers9D .......... .......... ...CTGCCGC TGGGAAGTAG CTTCTGGGTA msp3E-FcosPers2E GCAGAGAGCT GTGTTGGCCT TGGTTAGGAG ATGCTGGTTA CCGTTGAAGT msp3E-SM16d3 GC........ .......... .......... .......... ...AGAGAGC Chi-like sequence 151 200 msp3E-F5r9 GTGATGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-F5r16 GTGATGGTGC AGAGA GCTGT GTTGG TCGTG CCAAGTTAGA CTG....... msp3E-F5r15 GTGATGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-F5r14 GTGATGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-FcosAc1C GTGATGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-FcosPers9D GTGATGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-FcosPers2E GAACCGGTGC AGAGA GCTGT GGTGG TCGTG CCAAGTTAGA CTG....... msp3E-SM16d3 TGTGGTGGTG GCAGA GCTGT GGTGG TCGTG CCAAGTTGGG CTGAGTGCTG Figure 14 (continued).

PAGE 112

102 351 400 msp3E-F5r9 .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-F5r16 .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-F5r15 .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-F5r14 .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-FcosAc1C .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-FcosPers9D .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-FcosPers2E .......... .......... ...GGATGCC AGGGGTTTAT AGATAGTGGC msp3E-SM16d3 AAGAGGTGCT GTGAGGATGT AGGCACCAGG TTATGTCGGT GGTACCAAGA 401 450 msp3E-F5r9 TTTGTTATGT CGGTGGTACC AAGAACTTCA AGT...GTGT TGGTGCCCAG msp3E-F5r16 TTTGTTATGT CGGTGGTACC AAGAACTTCA AGT...GTGT TGGTGCCCAG msp3E-F5r15 TTTGTTATGT CGGTGGTACC AAGAACTTCA AGT...GTGT TGGTGCCCAG msp3E-F5r14 TTTGTTATGT CGGTGGTACC AAGAACTTCA AGTTGATGGT GACCTGGGTA msp3E-FcosAc1C TTTGTTATGT CGGTGGTACC AAGAACTTCA AGTTGATGGT GACCTGGGTA msp3E-FcosPers9D TTTGTTATGT CGGTGGTACC AAGAACTTCA AGT...GTGT TGGTGCCCAG msp3E-FcosPers2E TTTGTTATGT CGGTGGTACC AAGAACTTCA AGT...GTGT TGGTGCCCAG msp3E-SM16d3 GCTTTGACTG TGCCGGTTAT AAGGAGGCAC GAATGAGGTG TTGTGCCCAG 451 476 msp3E-F5r9 TAGAGTTGGA AGTGTGAGTG ATGTTT msp3E-F5r16 TAGAGTTGGA AGTGTGAGTG ATGTTT msp3E-F5r15 TAGAGTTGGA AGTGTGAGTG ATGTTT msp3E-F5r14 GTAAGTTGGA AGTGTGAGTG ATGTTT msp3E-FcosAc1C GTAAGTTGGA AGTGTGAGTG ATGTTT msp3E-FcosPers9D TAGAGTTGGA AGTGTGAGTG ATGTTT msp3E-FcosPers2E TAGAGTTGGA AGTGTGAGTG ATGTTT msp3E-SM16d3 TAGAGTTGGG AGTGTGAGTG ATGTTT Figure 14 (continued). Discussion Genomic Structure It is interesting not only that the structure of msp2 and msp3 genes is similar, but that the ps eudogenes also appear in concert in the genome. Msp2msp3 genes and pseudogenes do not appear totally random in the genome as previously suggested (Palmer et al. , 1994a ; Alleman et al. , 1997). Most paralogs are confined to ~25% of the genome. Earlier observations, in which msp2 and msp3 pseudogenes were found in a close tail to tail position, are confirmed with the new genes identified by the A. marginale genome sequencing project (Alleman et al. , 1997; Alleman et al. , 1997; Brayton et al. , 2001). This suggests a possible coordinated control of expression or

PAGE 113

103 recombination of the two gene families. Ho wever this tail to tail structure is not universal since for both msp2 and msp3 gene families pseudogenes exist that are not flanked by a member of the other gene family. Th is indicates that this structure is probably not a prer equisite for re combination. On the other hand, a conserved flanking region, ORFX, is present next to each pseudogene or ps eudogene complex ( msp2 + msp3 ). Often this ORF is repeated in tandem to form the ~600bp ORFX-Y flanking repeat. Sequences or repeats of such lengths can hardly be a ttributed to chance and this suggests a role in the recombination of both msp2 and msp3 genes (Brayton et al. , 2001). In N. gonorrhoeae , flanking repeats, called the Sm a/Cla repeat, at the 3' end of pilin genes have been shown to be involv ed in recombination (Wainwright, Pritchard, and Seifert, 1994). In Borrelia hermsii three imperfectly repeated segments of 2 kb precede the vmp (variable membrane pr otein) gene promoter (Barbour et al. , 1991). Similarly, the ORFX-Y repeat is also found in the msp3 expression site. In contrast to its orientation in rela tion to most pseudogenes, in the msp3 expression site ORFX-Y is encoded on the same strand as msp3E . The repeat has been shown to be present on msp3E transcripts by 5Â’ RACE and thus has the potential to be expressed. Like msp3E , ORFX is polymorphic in this locus among the Florida 5Â’ RACE products, cosmids and the St Maries BAC. The first variable region is in t he 5Â’ end of the ORF and the second at the 3Â’ end of the ORF (fig. 4). These vari able regions are characterized by the deletion, insertion, and r eplacements of several codon s that do not change the predicted reading frame. ORFX is, however, not found in the msp2E operon

PAGE 114

104 and this suggests that its presence in t he expression site might not be essential for recombination. An alternative explan ation is that ORFX is yet another multigene family that underwent duplication together with msp2 and msp3 , and is expressed from the msp3E locus. Interestingly, the ORFX in the msp3 expression site contains a predicted si gnal peptide similar (18/26 identical amino acids) to t he signal peptide of msp3E . This could indicate that the protein products of msp3E and ORFX are targeted for similar cellular localization. Nevertheless, its systematic presence near msp2 and msp3 genes is intriguing. It will be important to discover whether the variation in the msp3 expression site ORFX-Y repeats is temp lated, and whether similar variation occurs in or around the pseudogenes. Recombination often involves exchanges between segments of DNA molecules of nearly identical sequenc e. Although repeated sequences are more scarce in bacterial genomes t han in eukaryotic genomes, enough exist to produce extensive rearrangement, deletion or multiplication of genetic material (Rocha and Blanchard, 2002). These re combination events may conflict with genome stability by causing chromosoma l rearrangements. Probably to avoid such events, Mycoplasma spp. strongly avoid inverse repeats in comparison to co-oriented repeats (Rocha and Blanchard, 2002). One might expect that orientation of the msp2 and msp3 genes plays a role in facilitating recombination between the expre ssion site and the pseudogenes, while impeding recombination between diffe rent pseudogenes and between the two families. Analysis of the A. marginale genome sequence indicates that msp2E

PAGE 115

105 and msp3E are encoded on the same strand, separated by ~64kbp. Also encoded on the same strand is pseudogene msp3-SM2 , which contains the entire conserved C-terminal end. This orientation potentially prevents fatal recombination between these different loci characterized by extensive similarity. On the other hand, there is no apparent or ientation bias in the pseudogenes of either families. Additionally, no bias c ould be detected in the flanking ORFX-Y repeat. It is fascinating to observe that the “true” genes identif ied in the context of this analysis, including msp2E , msp3E and the different rec -genes, all are encoded on the same strand, with the exception of msp4 . This is a trend reported for E. coli and many other prokaryotes (McLean, Wolfe, and Devine, 1998). Pseudogene Structure A total of 5 msp2 and 5 msp3 pseudogenes, out of an estimated total of ~7 each, were available for analysis. Both pseudogene families reveal a similar structure with a CVR flanked by highly cons erved regions at the 5’ and 3’ end. From the analysis it is clear that both families originate from a common ancestor and are thus paralogs. This is su pported by the stretches of sequence homology at the 5’ and / or 3’ end. While this is most obvious in the expressed copies additional evidence can be found in the pseudogenes. Four out of five msp3 pseudogenes share a near identic al 5’ ORF (125-176 bp) with msp2 , two pseudogenes ( msp3-SM1 & -SM4 ) share a short 3’ end with msp2 (encoding VAGAF, fig. 13) and the msp3-SM2 pseudogene has retained the entire conserved C-terminal coding region also f ound in the expression sites. While

PAGE 116

106 the 5’ ORF sequence in msp3 pseudogenes is near identical to msp2 sequence it has diverged more in the msp3E sequence. On could therefore hypothesize that conservation of this sequence is important in the pseudogenes. However the absence of this stretch of homology in pseudogene msp3-SM5 conflicts with this. Despite their common ancestry, not able differences are present between the two gene families, with specific msp3 genes featuring additional conserved regions flanking a much longer CVR. Al so, the total length of the CVR in the different msp2 pseudogenes stays within a narrow range (up to 27 bp difference) while the msp3 CVR length is more variable (up to 624 bp difference). Moreover, the msp2 gene family has remained more homogeneous (73-99 % identity) than the msp3 family (55-85% identity), hinting at different selection pressures or bi ological constraints. Msp2 and msp3 genes must have undergone a number of duplicatio n events since they originated from their common ancestor. Within the msp3 gene family it becomes clear that not only the msp3 pseudogenes were duplicated, but equally their flanking regions, specifically ORFX-Y. Evidence of th is can be found by comparisons of msp3SM1 with –SM2 , and of msp3-SM4 with –SM5 . The first pair shares a more conserved 3’ CVR as well as a conserved 5’ flanking sequence. The second pair has an almost identical 3’ CV R and 3’ flanking region, including an msp2 pseudogene. Interestingly, while msp3-SM4 and –SM5 have diverged substantially (only 85% identity at the amino acid level), the flanking msp2 pseudogenes are entirely conserved. This, together with the higher degree of

PAGE 117

107 diversity in the msp3 pseudogenes overall, agai n indicates that both pseudogene families are under different select ion pressures or constraints. If true, this would result in msp2 genes being more conserved between different strains of A. marginale than msp3 genes. Some evidence for this can be found in the comparison of msp3-19 , a pseudogene and flanking sequence characterized in the Florida strain (Alleman et al. , 1997), with the St. Maries strain pseudogene msp3-SM3 . Comparison of the two msp3 pseudogenes does not reveal any specific close relati onship, but again the 3Â’ flanking region is highly conserved and this includes an ORFX and an msp2lk ORF. A similar relationship is found between the St Maries msp3-SM1 and the Florida msp3-11 (Alleman et al. , 1997), which have both retained a similar 3Â’ flanking region, including an msp2 pseudogene (96% nucleoti de identity), but the msp3 pseudogenes have diverged (50% nucleotide identity). The homologies in the CVR of the two pairs of related St. Maries strain msp3 pseudogenes (SM1-SM2 and SM4SM5) diverge around the same sequence, i.e. the central, more cons erved region within the CVR. This indicates that this sequence might play a role as a cross-over point in pseudogene and expression site recombination. Alternatively, the 5Â’ end of the CVR might again be under different select ive pressures than the 3Â’ end. Both pairs of flanking regions diverge at the location of ORFX. This indicates that recombination possibly hinges around this sequence. This role of ORFX, as a recombination hot-spot, agrees with observations from the msp3 expression locus when this is compared within and bet ween isolates. Comparison of other

PAGE 118

108 homologous regions in the St Maries BACÂ’s additionally reveals a short stretch of conserved nucleotides in close proximity (agagaGCTGTGGTGG). This sequence shows similarity with the E. coli (GCTGGTGG) or Haemophilus influenzae (GNTGGTGG or GG/CTGGAGG) Chi-sequences, which have been shown to be involved in the RecBCD pathway of homologous recombination (Sourice et al. , 1998; Kowalczykowski, 2000). This Chi-like sequence is also present in all available msp3E 5Â’ flanking sequences. The Chi-like sequences do not occur outside these regions of hom ology, the conserved decamer is only predicted to occur by chance every ~1/106, and it is often r epeated 2 or 3 times near msp2-msp3 pseudogenes (fig. 14). Pseudogenes, or Only Pretending? Analysis of the different msp3 genes indicates that outside the expression locus the msp3 genes are more than likely pseudogenes, i.e. no MSP3 protein is made directly from these sites. With the exception of msp3SM2 , none of the pseudogenes have the region encoding the C-terminal end of MSP3 as determined by Western blot (s ee previous chapter). The ORFÂ’s are also too short to encode the ~80-90 kD a MSP3 protein and clear promoter structures are missing within r easonable distance upstream (Brayton et al. , 2001). A final and conclusive argument can be found in msp3-SM5 which contains a stop codon within the CVR, confirmed multiple times (Brayton, personal communication 2002) and can t herefore not encode a genuine MSP3 product.

PAGE 119

109 The situation within the msp2 and msp2related family of genes is more complicated. msp2 related ORF’s found in the genomic data fall roughly into three groups : truncated msp2 pseudogenes, msp2 -like ORF’s ( msp2lk number) and ORF’s with limited relationship to msp2 ( msp2lk ) (fig. 11 & 12). The msp2 pseudogenes (SM1 through – SM5 ) are characterized by conserved regions, flanking a CVR, which are near identical to those found in the msp2E expression site operon (Barbet et al ., 2000). The pseudogenes are, however, truncated at both the 5’ and 3’ ends, and could not encode the mature ~40 kDa MSP2. The msp2 pseudogenes act as sequence donors to alter the msp2E expression site. The msp2lk ORF’s still share a si milar structure with msp2E , with a variable central region flanked by conserved regions, but the latter have diverged substantially. Nonetheless many msp2lk ORF’s have retained similarity to the expressed MSP2 N-terminal end and all of them have maintained a similar C-terminal end. msp2lk ORF’s all have genuine start codons at the beginning of their ORF and could encode proteins of roughly the same size as MSP2. Arguing in fa vor of at least some of these msp2lk genes being pseudogenes is the presence of nonsense mutations (stop codons). Some regions containing msp2 like sequences do not have clear ORF’s, e.g. in msp2-(3-19) of the Florida strain (Alleman et al. , 1997; Meeus and Barbet, 2001). Whether msp2lk ORF’s could play a role in msp2E recombination is unclear since they lack the highly c onserved regions flanking the CVR. Recombination could nevertheless be m ediated through shorter stretches of conserved sequences within the ORF’s. Alternatively these ORF’s could be

PAGE 120

110 transcribed independently from the msp2E operon and form a separate cluster of paralogs. Some msp2lk ORFÂ’s form a tandem structure (e.g. msp2lk-1 through -4 ) resembling the locus structur e of the MAP1-like genes in E. chaffeensis , E. ruminantium and E. canis (Ohashi et al. , 1998b; Reddy et al. , 1998; Sulsona, Mahan, and Barbet, 1999; Yu et al. , 2000a;). A final group of msp2 related ORFÂ’s have only short stretches of homology to msp2E , e.g. msp2lk5 and msp2lk8 at the C-terminal end and in the centre respectively. Their function is unclear, but given their di stant relationship it is not likely that they play a major role in msp2E recombination. Mechanism of Antigenic Variation Three types of recombination events occur in bacteria : illegitimate, sitespecific and homologous (Dybvig and Voelker, 1996). Illegitimate recombination occurs between sequences with little or no sequence homology and does not require special sites or the ac tivity of recombinases. Site-specific recombination occurs at highly pref erred sites and is catalyzed by a recombinase that specifically recognizes these preferred site s. The Xer sitespecific recombination system correc ts the potential damage of homologous recombination between circular DNA mo lecules by converting dimers back to monomers. In E. coli this system uses two related recombinases, XerC and XerD, which catalyze exchanges at a family of sites presents on several plasmids and on the chromosome (Spi ers and Sherratt, 1999). This system appears to be highly conserved amongst bacteria and A. marginale sequence analysis reveals the presence of two hom ologous ORFÂ’s in the genome. One

PAGE 121

111 xer is found on the same BAC G11 on which many of the msp2 and msp3 pseudogenes are found. This indicate s a possible functional relationship, maybe in resolving dimers formed duri ng gene conversion of the expression site operons. Homologous recombination occurs bet ween virtually any sites that share extensive stretches of nucl eotide homology and is mediat ed by a specific set of enzymes. It thus always involves repeats, in a specific orientation to each other. Repeats in general play an import ant role in creating diversity from a small genome. There is a strong negative correlati on between the genome size of a bacterium and the densit y of repeats. This means that smaller genomes have relatively more repeats, apparently to off-set their reduced coding capacity (Rocha, Danchin, and Viari, 1999). Homologous recombination within the msp2 and msp3 gene families is likely. T he different members of both msp2 and msp3 gene families are repeated numerous times in the genome and they are characterized by blocks of highly conserved sequence. These conserved regions could serve as anchors in changi ng the variable region in the expressed copy. However msp2 and msp3 genes are paralogs and share a 5Â’ and / or 3Â’ region of sequence identity. Specific targeting of msp3 pseudogenes to the MSP3 expression site therefore may resu lt from additional blocks of conserved sequence unique to msp3 genes. The potential role of these conserved regions becomes especially clear in pseudogenes msp3-SM2 and msp3-SM5 . msp3SM2 is very similar to the msp2 expressed copy, both at the 5Â’and at the 3Â’end of the gene, but no recombinat ion of this pseudogene with the msp2

PAGE 122

112 expression site has yet been observed. Such an event, if it occurred, would have been noticed given the substant ial size difference between the msp2 and msp3 genes. On the other extreme, pseudogene msp3-SM5 has no sequence homology with msp2 genes (22-23% overall identity at amino acid level) and therefore demonstr ates that the msp3 -specific conserved regions are sufficient to mediate recombination. In addition to the importance of sequence similarity in homologous recombination, the orientat ion of the repeats plays a vital role. Sequences that are co-oriented have been shown to recombine less than those in opposite orientation (R ocha and Blanchard, 2002). While there is no clear orientation bias in t he pseudogenes of either msp2 or msp3 , both expression site operons and msp3-SM2 are on the same strand. This orientation pattern makes it less likely that these regions recombine despite their significant homology at the C-terminal end. To explain the extensive diversity observed in msp2E and msp3E it is necessary that not only full-length pseudogenes, but also partial gene segments, involving s horter stretches of homology within the CVR, recombine into the expression site. This has been confirmed to occur in msp2E and msp3E (Brayton et al. , 2002). The conserved regions flanking the CVR might still play an impor tant anchoring role in thes e events. Nevertheless, because recombination through short stretc hes of homology can occur, many of the msp2lk genes could be involved in msp2E recombination. More research needs to be done to confirm or rule out this hypothesis. Besides repeats, homologous recombinat ion typically involves a specific set of enzymes orchestrating the whole process. Many of the enzymes have

PAGE 123

113 first been characterization in Escherichia coli . The major recombination pathway in E. coli is RecBCD, which is involved in conjugal and transductional recombination, DNA repair, and degradat ion of foreign DNA. RecA, RecB, RecC, RecD and SSB (single strand bind ing) proteins, and a specific DNA sequence called Chi (GCTGGTGG), are essential to this pathway (Smith, 2001). The RecBCD enzyme complex is both a DNA helicase and a nuclease that loses its nuclease activity, but not its helicase activity, when reaching a Chi sequence. These properties effectively create a single strand of DNA important for recombination. The Chi-like sequence flanking many msp2 pseudogenes, msp3 pseudogenes and msp3E might similarly be invo lved in the creation of single strands needed for recombination. However no such sequence is found near the msp2E operon. When a single str and of DNA is created in E. coli , RecBCD loads the RecA protein onto this ssDNA. RecA then pairs homologous DNA molecules and promotes the subsequent exchange of DNA strands (Skaar, Lazio, and Seifert, 2002) . An alternative pathway of homologous recombination, called RecF , confers UV resistance and plasmid recombination to E. coli . Proteins involved include aga in RecA in addition to RecF, RecO, RecR, RecQ, and RecJ (Skaar, Lazio, and Seifert, 2002; Kowalczykowski, 2000). Gene conversi on, the unbalanced exchange of DNA sequences between gene homologues, exists as a subset of this homologous recombination process (Howell-Adam s and Seifert, 2000). Gene conversion has been proposed as the mechanism of genetic variation in the msp2E and msp3E operons (Barbet et al. , 2000; Brayton et al. , 2001; Brayton et al. , 2002).

PAGE 124

114 Neisseria gonorrhoeae pilus antigenic variation is probably the best defined bacterial gene conversion system and it has many structural features in common with msp2 and msp3 of A. marginale . Antigenic variation of the N. gonorrhoeae pilus, a surface exposed organelle mainly made of pilin, occurs when a variant pilin pseudogene from a silent locus recombines into the expression locus, pilE , by gene conversion. In Neisseria , pseudogenes are transferred to the expression site by reco mbination events at limited regions of homology of 9-30 bp, not un like the observations for msp2 , 76 bp (Brayton et al. , 2002), or msp3 , 14 bp. Like in A. marginale , the pilin pseudogenes do not encode an entire pilin gene and are characterized by conserved regions flanking gene specific variable regions (Howell-Adams and Seifert, 2000). At the 3' end of all pilin loci lies a cons erved DNA sequence, called the Sma/Cla repeat, that has been shown to be essentia l for pilin recombination (Wainwright, Pritchard, and Seifert, 1994). ORFX-Y found flanking all msp2 and msp3 pseudogenes and msp3E could play a similar role. In N. gonorrhoeae , a RecFlike pathway is central to pilus gene conversion and involves RecA and RecJ (Skaar, Lazio, and Seifert, 2002). The formation of ssDNA intermediates, by RecJ, and pairing of homologous DNA by RecA , is postulated to play a role in all current models describing antigenic variation at pilE (Howell-Adams and Seifert, 2000). The presence of RecA, RecJ, RecF and RecG homologs in the same region of the A. marginale chromosome that encodes the mps2 and msp3 expression sites and most pseudogenes is lik ely not coincidental. One can thus envision a mechanism in which an exonuc lease, like RecJ, removes the CVR of

PAGE 125

115 msp3E up to a specific sequence, possibly the Chi-like sequence or alternatively ORFX-Y, to create a ssDNA. At the same time this exonuclease could create a single strand at an msp3 pseudogene but conserve the genetic information when stopped at ORFX-Y or the Chi-like sequence, which is oriented in the opposite direction in ps eudogenes. This hypothesis fits in with the N. gonorrhoeae models predicting a mechanism for gene conversion. All models assume that DNA replication firs t produces two chro mosomes. In the preferred model, an intrac hromosomal recombination , triggered by ssDNA, between the pilE locus and a silent pilS copy on the donor chromosome produces an episomal circle carrying a hy brid locus. Similar recombination could be achieved through a single strand msp3 pseudogene and the msp3E locus on the same chromosome. In Neisseria these hybrid circles are predicted to then recombine with a recipient pilE on the second chromosome (HowellAdams and Seifert, 2000). Finally the ensuing dimer could be resolved by the Xer-C and Xer-D enzymes, in A. marginale encoded close to the msp2-msp3 genes on BAC G11. A tantalizing but s peculative suggestion by Howell-Adams and Seifert is that the formation of diplococcal forms in Neisseria could be the result of this inter-chromosomal mechanism of pilE variation with a sacrificial donor and a recipient chromosome (How ell-Adams and Seifert, 2000). In A. marginale and related organisms this coul d be the mechanism behind dense forms transforming into re ticulated forms and evolving back to smaller and infective dense pathogens (Kocan, Be zuidenhout, and Hart, 1987; Blouin and Kocan, 1998;).

PAGE 126

116 CHAPTER 6 5Â’RACE ANALYSIS OF MAJOR S URFACE PROTEIN EXPRESSION IN DIFFERENT SPECIES OF ANAPLASMATACEAE Introduction Antigenic variation in prokaryoti c organisms often involves multigene families. These repeated gene sequences are used to generate diversity through either recombination of a singl e or few expression site(s) or through differential expression from several diffe rent expression site s. In order to unravel the mechanisms involved it is essential to determine the genomic locus or loci from which the genes are tran scribed. While the coding regions of multigene families are often too similar to positively identify the genomic origin of a certain transcript, the flanking regi ons can offer distinguishing markers. Several methods have been devised to obtain these unknown flanking sequences. Traditional methods for eukaryotic organisms include constructing and screening of poly-T selected cDNA libr aries. Prokaryotic mRNA, however, provides a number of unique challenges. First, the 3' ends of both prokaryotic and eukaryotic mRNA are polyadenylated, but the poly-A tracts of prokaryotic mRNA are generally shorte r, ranging from 15 to 60 adenylate residues and associated with only 2-60% of the molecule s of a given mRNA species (Sarkar, 1997). In addition, polyadenylation in bac teria is highly indiscriminate and is even associated with ribosomal RNA. Secondly, bacterial mRNA transcripts are highly unstable with half-lives in the or der of a few minutes . These features

PAGE 127

117 make the detection of transcripts and especially low abundance transcripts challenging using cDNA libraries (Sar kar, 1997; Rauhut and Klug, 1999; Tillett, Burns, and Neilan, 2000). Rapid amplif ication of cDNA ends, or RACE, is a powerful PCR-based technique which over comes some of these problems. Primers add specificity to the process and the logarithm ic amplification ensures the acquisition of even low numbers of transcripts. In general, PCR amplification requires two sequence-specific primers th at flank the region of sequence to be amplified. 5'RACE, or anchored PCR, is a technique that facilitates the isolation and characteriza tion of unknown regions 5Â’ of a known cDNA sequence. It uses a set of tw o or three gene specific reverse primers (GSP) based on sequences conserved am ong the different members of the gene families. mRNA is reverse transcr ibed into cDNA using a gene-specific antisense oligonucleotide (G SP1). TdT (terminal deox ynucleotidyl transferase) is then used to add homopolymeric tails, a ru n of similar nucleotides, to the 3' ends of all cDNA, which is than amp lified by PCR. The PCR uses a combination of nested gene-specific pr imers (GSP2 and GSP3), which anneal 3' to GSP1; and complementary homopol ymer-containing anchor primers which permit amplification from t he homopolymeric tail. This allows amplification of unknown sequences between the GSP2/3 and the 5'-end of the mRNA (Reviewed by Schaefer, 1995). We have used this methodology to analyze the transcription of msp2-msp3 related homologs in three different species of the Anaplasmataceae. The results indicate the usefulness of this technique in

PAGE 128

118 detecting expression sites and illustrate the differences in expression patterns between the genera Anaplasma and Ehrlichia . Experimental Procedure Anaplasmataceae Isolates Two A. phagocytophilum strains (NY-18 and Webster) isolated from human patients were supplied by M.E. Aguero-Rosenfeld (New York Medical College, Valhalla, NY) and cultured in HL-60 cells with RPMI 1640 medium containing 2mM L-glutamine (Cellgro ) supplemented with 10% fetal bovine serum (Cellgro) at 37°C in 5% CO2 as described previously (Walls et al ., 1999). The cells were harvested when the in fectivity reached 90-100%. 2-4 x 107 HL60 cells infected with A. phagocytophilum were centrifuged fo r 10 min. at 150xg, washed in phosphate buffered saline (PBS) and then resuspended to 2x107 cells/ml in PBS followed by 7-8 volumes of RNAlater (Ambion) for storage at 80ºC. RNA was isolated from stored aliquots using the RNAqueous kit (Ambion). E. chaffeensis (Arkansas isolate) was kindly provided by J. E. Dawson and J. G. Olsen (Centers for Disease Cont rol, Atlanta, Ga) and grown in the canine macrophage cell line DH82 in Eagl e's minimum essential medium containing 10% fetal bovine serum, 26 mM sodium bicarbonate, and 2 mM Lglutamine at 34°C. Cells were harves ted when 90-100% of t hem were infected, and ehrlichiae were purified as previously described (Chen et al ., 1996). RNA was isolated as above, but immediately after harvesting the pathogens without storage.

PAGE 129

119 A. marginale of the Florida strain was obtained from a splenectomised calf (2153) infected intramuscularly wit h a stabilate (B442). Total RNA was isolated from whole blood on Day 24 post infection during acute rickettsemia (42% infected RBCÂ’s) by extraction with 6 M urea-3M LiCl (Van der Ploeg et al ., 1982). Rapid Amplification of cDNA Ends (RACE) For all samples, isolated RNA was digested with DNaseI (DNA-free; Ambion) before use in 5 RACE. mRNA transcripts were reverse transcribed into cDNA using the RETROscript kit (A mbion) according to the manufacturerÂ’s protocol and a species specific GSP1. After first strand cDNA synthesis, the original mRNA template was removed by treatment with an RNase Mix and the cDNA purified using a GLASSMAX Spin Cartridge according to the manufacturers recommendation (Life Tec hnologies). A homopolymeric tail was then added to the 3 end of the cDNA using TdT and dCTP. An aliquot of the reaction was directly amplified by P CR using SuperTaq (Ambion), a nested gene-specific primer, GSP2 , and a deoxyinosine-containi ng anchor primer (Life Technologies). An additional r ound of PCR using the AUAP (Abridged Universal Amplification Primer, Life Technologies), in conjunction with a progressively nested primer, GSP3, was used when required to obtain an adequate amount of cDNA to permit cloning. Products of the 5 RACE were analyzed by agarose gel electrophoresis and Southern blotting using gene specific probes. PCR products were cloned into the pCR-XL-TOPO vector (Invit rogen) and plasmid DNA was isolated,

PAGE 130

120 digested with EcoRI and analyzed by agaros e gel electrophoresis and Southern blotting. Plasmids with inserts were subsequently digested with EcoRI, to cut out the insert, and RsaI, to analyze the dive rsity of the inserts per size group. Selected plasmids were sequenced to veri fy the structure of the amplified cDNA. For A. phagocytophilum the GSPs AB943, AB970 and AB947, were located in the region downstream from the CVR and conserved among the available copies of p44 . Therefore the 5’RACE products should contain the CVR and the 5’ end of expressed msp2(p44) genes . The PCR conditions were initial denaturation cycle of 2 min. at 93ºC, followed by 35 cycles of 30 sec. at 93ºC, 1 min. at 55.3ºC and 5 min. at 72º C, and final extension at 72ºC for 25 min.. Two control reactions were conducted similarly, one without reverse transcriptase in the initial RT-reaction and another with no template in the PCR. For E. chaffeensis paralog-specific GSP’s had to be used given the sequence divergence among the different p28 related genes. GSP1 used in the RT-reaction, AB980, is complementar y to the 3’ end of the paralog p28 , but has only 1 or 2 nucleotides diffe rence with 3’ sequence in omp1c, -d, -e, and –f . Anchored PCR was carried out on two specific genes, p28 and omp1f , using GSP’s AB996 / AB997 and AB993 / AB994 respectively. The PCR conditions were initial denaturation cycl e of 2 min. at 93ºC, followed by 40 cycles of 30 sec. at 93ºC, 1 min. at 50ºC ( p28 ) or 56.7ºC ( omp1f ) and 3 min. at 72ºC, and final extension at 72ºC fo r 25 min.. Annealing in t he nested reactions was done at 55ºC. Control reactions were included as above.

PAGE 131

121 5Â’RACE for the A. marginale msp3 gene was carried out as described in Chapter 4 using AB75 5, AB878 and AB877. Primers and Probes The following primers and oligonucleotide probes were used : AB755 tctaaccacttaagttctccaagt; AB841 atgtttcgaggggagagtgg; AB877 taattcttctggtctgctcagttc ; AB878 gctttgctagttcatgttgtag; AB980 ccaagttctattccaaagt; AB993 tggcatacacttagtgttactata; AB994 tgtgagagttgaggtactggg; AB996 t ccagtattactattgcaggg; AB997 cctgaagtgttgatccagta; AB943 aagaagatcataacaagcatt; AB970 ccttcaatagtagtyttagctagtaaccc and AB 947 taacaacatcataagctaactcc. Data Analysis Genomic information to analyze the origin of the obtained transcripts was available from three ongoing genome s equencing projects and published loci. For E. chaffeensis the entire p28 locus was obtained from the NCBI database (U72291) and was published previously (Ohashi et al ., 1998b). BLAST searches were done on http://ti for A. phagocytophilum and E. chaffeensis and on http://genomics.vetmed.w for A. marginale . Further analysis was done usi ng programs on the Wisconsin GCG package Version 10.3 (Accelrys Inc.) available through the Biological Computing Facility of the Interdisciplinary Center for Biotechnology Research at the University of Florida.

PAGE 132

122 Results Amplicons, shown to be specific by Southern blot, were obtained for each pathogen tested. For A. phagocytophilum six different clones from the NY18 strain and two from the Webster strain were sequenced. All but one clone extended 5Â’ of the predicted msp2(p44) initiation codon. Six of these clones appeared to originate from t he same genomic locus. Database searching suggests that this flanking s equence occurs only once in the genome and therefore the transcripts probably or iginated from this unique genomic expression locus. Database searches also provided information downstream from the obtained transcripts and rev ealed a genomic locus with a full-length msp2(p44) gene flanked by several ORFÂ’s (fig. 15A ). Upstream restriction-site PCR (Barbet et al ., unpublished) and the database re vealed the presence of several p44 related ORFÂ’s upstream and downstream from msp2E(p44). Interestingly, despite the apparent comm on genomic origin of these six 5'RACE amplicons their CVR were different, suggesting that this expression locus undergoes genetic recombination (fig. 15B & C). It was also interesting to observe that several transcripts with di fferent CVR and from different strains ended at the same location in the 5Â’ flanking region (e.g. msp2E-Web vs. msp2E-5RA5 and msp2E-5RC10 vs. msp2E-5RC9 ). One NY strain amplicon, msp2E-5RC11 , had sequence similar to msp2E(p44) starting bp21 downstream of its first ATG initiation codon. T he N-terminal end was however different, as was the 5Â’ flanking sequence. BLAST search results localized this flanking sequence to another p44 gene on contig31 of the unfinished A.

PAGE 133

123 phagocytophilum genome. This 16 kbp contig contains two p44 genes, one of which, msp2Ap-contig31 (fig. 15D), has an identical CVR and 5Â’ flanking region as those found in the 5Â’RACE amplicon msp2E-5RC11 . For E. chaffeensis four clones for each primer combination, based on the p28 and omp1f genes respectively, were sequenced (fig. 16). Within each combination the amplicons revealed a similar sequence, both in the coding region as well as in the 5Â’ flanking r egion. The sequences obtained were near identical to those described by Ohashi et al. (Ohashi et al ., 1998b). The results suggested that the transcr ipts originated from these two earlier described expression loci and database searching sugg ests that these loci and flanking regions are unique within the genome. All four transcripts from gene omp1f stopped at the same position in the flanking sequence. The published data reveal a run of 12 GÂ’s at this positi on in the genome which likely served as an annealing place for the anchor primer (fig. 16B). A similar run of GÂ’s was subsequently detected upstream from omp1d . Three transcripts originating from gene p28 ended ~26 bp upstream fr om the ATG initiation codon, but no poly-G stretch was present at th is position. The one remaining p28 transcript ended within the ORF. The 5Â’RACE results for the msp3 gene family in A. marginale have been described in chapters 4 and 5. Briefly, four different clones , obtained by cloning the largest generated amplicon, were s equenced. The data suggested that the transcripts originated from a single, but polymorphic genomic locus (fig. 17). This was confirmed by Southern blot analysis of genomic A. marginale DNA

PAGE 134

124 (chapter 4) and by database searching. In A. marginale the CVR was not present in the 5Â’RACE products since cloning of full length gene and flanking regions had proven difficult. The pol ymorphic character of this unique expression site was determined by analysi s of a genomic cosmid library as described earlier. In contra st to the two other specie s analyzed, the 5Â’ flanking regions of msp3E were characterized by several va riable regions. Similar to the other pathogens analyzed three out of four transcripts terminated at the same nucleotide. This is not entirely surp rising since the amplicons were size selected before cloning and multiple sm aller amplicons were not analyzed. Nevertheless, the substitutions, delet ion and insertions in the sequences suggest that again all amplicons originat ed from different transcripts and that transcripts of this specific length are therefore not random events.

PAGE 135

125 msp2E-5RA11 msp2E-5RA5 msp2E-5RC10 msp2E-5RC9 msp2E-5RF3 msp2E-5RC11 msp2Ap-contig31 msp2E-Web msp2E(p44) p44Esup1tRNA-synthetasep44P ftsY p44P A B C D msp2E-5RA11 msp2E-5RA5 msp2E-5RC10 msp2E-5RC9 msp2E-5RF3 msp2E-5RC11 msp2Ap-contig31 msp2E-Web msp2E(p44) p44Esup1tRNA-synthetasep44P ftsY p44P msp2E-5RA11 msp2E-5RA5 msp2E-5RC10 msp2E-5RC9 msp2E-5RF3 msp2E-5RC11 msp2Ap-contig31 msp2E-Web msp2E(p44) p44Esup1tRNA-synthetasep44P ftsY p44P A B C D Figure 15 Structure and variability of msp2(p44) expression in A. phagocytophilum . A. A diagrammatic re presentation of the msp2E(p44) expression site and flanking regions as determined by 5Â’RA CE, restriction-site PCR and database searching. Red arrows represent msp2 related ORFÂ’s, including the expression site msp2E(p44) and two additional truncated ORFÂ’s p44P . Other ORFÂ’s are represented in light blue and the neares t homologs as determined by BLAST search indicated above. The orange arrow is an ORF homologous to a gene found in the msp2 operon of A. marginale . A nucleotide scale is provided under the diagram. B. This panel represents the locati on and structure of t he sequenced 5Â’RACE amplicons. 5Â’RACE products are repr esented by red blocks and the CVR by colored blocks or x-hashed black boxes. Si milar colors indicate similar CVRÂ’s. The black boxes represent the polyG tail added at the beginning of the transcript. C. A PLOTSIMILARITY graph comparing vari ability of this expression site in seven amplicons from the above expression site drawn to the same scale. A similarity score of 1.0 indicates identic al sequence in a sliding window of 10 nucleotides and a decreasing score from 1.0 to 0.0 indicates increasing variability. D. This diagram represents the location of the msp2 related gene, msp2Apcontig31 , on contig 31 of the unfinished A. phagocytophilum genome. This gene is the origin of the 5Â’RACE amplicon msp2E-5RC11 represented below. Both have a similar CVR.

PAGE 136

126 p28-5R98 p28-5R109 p28-5R108 p28-5R99 p28-5R98 omp1F-5R105 omp1F-5R104 omp1F-5R102 omp1F-5R101 omp1F-5R103MNQPTUVWXYSHZACDEFp2812 BA B C p28-5R98 p28-5R109 p28-5R108 p28-5R99 p28-5R98 omp1F-5R105 omp1F-5R104 omp1F-5R102 omp1F-5R101 omp1F-5R103MNQPTUVWXYSHZACDEFp2812 B p28-5R98 p28-5R109 p28-5R108 p28-5R99 p28-5R98 omp1F-5R105 omp1F-5R104 omp1F-5R102 omp1F-5R101 omp1F-5R103 p28-5R98 p28-5R109 p28-5R108 p28-5R99 p28-5R98 omp1F-5R105 omp1F-5R104 omp1F-5R102 omp1F-5R101 omp1F-5R103MNQPTUVWXYSHZACDEFp2812 BA B C Figure 16 Structure and variability of p28 expression in E. chaffeensis . A. A diagrammatic re presentation of the p28 locus and the location of paralogs omp1f and p28 as described by Ohashi et al. (Ohashi et al. , 1998b). Red arrows represent p28 related ORFÂ’s with their letter (for genes given the omp1 nomenclature) or number (for p28 ) indicated above the arro w. Other ORFÂ’s are represented in light blue. A nucleoti de scale is provided under the diagram. B. This panel represents the locati on and structure of t he sequenced 5Â’RACE amplicons. 5Â’RACE products are repres ented by red blocks. The black boxes represent the poly-G tail added at the begi nning of the transcrip t or the run of GÂ’s present in the genome in front of omp1f . C. A PLOTSIMILARITY graph comparing vari ability of the transcripts originating from these expression sites drawn to the same scale. A similarity score of 1.0 indicates identical sequence in a s liding window of 10 nucleotides and a decreasing score from 1.0 to 0.0 indicates increasing variability.

PAGE 137

127 msp3E-F5R9 msp3E-F5R16 msp3E-F5R15 msp3E-F5R14 msp3E YX C B A msp3E-F5R9 msp3E-F5R16 msp3E-F5R15 msp3E-F5R14 msp3E YX msp3E-F5R9 msp3E-F5R16 msp3E-F5R15 msp3E-F5R14 msp3E YX C B A Figure 17 Structure and variability of msp3 expression in A. marginale . A. A diagrammatic representation of msp3E and flanking regions as determined by 5Â’RACE and genomic library screening and sequencing. The msp3E gene is represented by a mult i-colored block with msp2 related segments in red, the conserved regions in green, the CVR in grey and additional ORF in light blue. ORF Y and ORF X are represented as dark blue arrows and other ORFÂ’s in light blue or black. A nucleotide sca le is provided under the diagram. B. This panel represents the locati on and structure of t he sequenced 5Â’RACE amplicons. 5Â’RACE products contain the msp3 N-terminal region, green blocks, as well as upstream ORFÂ’s Y and X, dark blue arrows. The black boxes represent the poly-G tail added at the beginning of the transcript. C. A PLOTSIMILARITY graph comparing va riability of the obtained transcripts drawn to the same scale. A similarity score of 1.0 indicates identical sequence in a sliding window of 10 nucleotides and a decreasing score from 1.0 to 0.0 indicates increasing variability.

PAGE 138

128 Discussion Anchored-PCR or 5Â’RACE is a powerful technique to obtain unknown upstream flanking sequence. In our studies the technique revealed the expression sites for two msp2 homologs, msp3 and msp2(p44) , in two different species of Anaplasma and confirmed the expression of msp2 paralogs in E. chaffeensis from different loci. In each ca se the results have been confirmed by several other techniques including Southern blot analysis of genomic DNA, Western blot on native proteins, RT -PCR (results not shown and Barbet et al ., unpublished; Unver et al ., 2002; Long et al ., 2002), RNA protection assays (Barbet et al. , unpublished) and database analysis, validating the strength and robustness of the 5Â’RACE technique. The data obtained reveal the differenc es in expression and the creation of phenotypic variability between the tw o genera of Anaplasmataceae. Within the genus Anaplasma at least one, A. phagocytophilum , or two, A. marginale , highly variable genomic expression sites for msp2 paralogs exist that are used for expression of multiple mRNAs with di verse CVRÂ’s. In contrasts, the genus Ehrlichia appears to transcribe the paralogs from multiple different genomic expression sites containing relatively st able and full-length copies. It is intriguing that two such closely rela ted genera would have evolved towards two very different systems within the same or thologs. It is however likely that neither system is the only available mechanism in these genera. Indeed within Anaplasma several full-length msp2like genes exist that are organized in similar tandem repeats as found in Ehrlichia. These genes could be transcribed and

PAGE 139

129 play a role in the creation of additi onal diversity through transcriptional regulation. The independent transcription of the msp2 like gene msp4 in A. marginale and results obtained by 5Â’RACE on msp2(p44) genes in A. phagocytophilum provide some support for this hypothesis. The msp2E-5RC11 transcript originated from a different locus than the major characterized polymorphic expression site and could repr esent an alternative expression site. Identical CVRÂ’s found in the 5Â’RACE product and in the genomic database from a different isolate suggest that this locu s is relatively stable over time as seen for the p28 genes. In contrast, no data are av ailable that sugge st the existence of gene conversion mechanisms in the Ehrlichias like those identified in Anaplasma . More research will be needed to fu lly elucidate the involvement of the different orthologs and paral ogs in the creation of diversity in this family of pathogens. 5Â’RACE can uncover many insights in the expression patterns of multigene families. It can, as demonstrated , reveal the origin and structure of the majority of transcripts and it can ex pose variability within an expression site and its flanking regions. The 5Â’RACE technique was, however, originally developed to obtain the very 5Â’ end of trans cripts. This piece of information is often needed to characterize the pr omoter and fully understand the transcriptional regulation of a gene. In our experiments the different transcripts obtained from the different genera are most often of different length, suggesting that the true 5Â’ ends were not obtai ned. Did the technique thus fail in characterizing the true beginning of the transcripts or are there alternative

PAGE 140

130 explanations? Sometimes the reason for ear ly termination is fairly obvious as in the presence of a poly-G st retch of nucleotides in omp1f to which the adaptor primer will anneal, a well recognized draw back of the technique (Schaefer, 1995). Other transcriptional ends are less easy to explain, but closer analysis reveals some clues. Within A. phagocytophilum two sets of transcripts containing different CVR and sometimes orig inating from different strains end at the same nucleotide position in the flanki ng region. This indicates that either different transcription start sites exist or, maybe more likel y, that some posttranscriptional regulation protects thes e mRNA ends. Indeed, while RPA and / or RT-PCR analysis suggest that the tr anscripts obtained by 5Â’RACE for msp2E(p44) and msp3E constitute a major fractions of transcripts from that site, they also reveal that longer transcrip ts exist (results not shown, Barbet et al. , unpublished). While the am ount of full-length mRNA available for protein synthesis is dependent on its transcription, the stability of diffe rent segments of the message also plays an im portant role. Metabolic instability is a hallmark property of mRNAs in most if not all organisms and plays an essential role in facilitating rapid responses to regula tory cues (Rauhut and Klug, 1999). Both stabilizing and destabilizing RNA struct ures exist that can influence the availability of the different genes on a pol ycistronic message. In polycistronic transcripts, as found in msp2E , msp2E(p44) and msp3 (Barbet et al. , 2000; Barbet et al ., unpublished), mRNA degradation make s a significant contribution to differential gene expression and could reflect the different stoichiometric amounts needed of the different protein products (Rauhut and Klug, 1999). The

PAGE 141

131 ends of the transcripts found here by 5 Â’RACE therefore likely represent RNase cleavage sites or stabilizing structures, rather then transcription start sites. In summary, the data show the useful ness of the 5Â’RACE technique in determining the expression site or sites in multigene families. It also revealed a fundamental difference betw een the expression of the msp2 orthologs in two genera of the family Anaplasmataceae. Anaplasma species generate diversity from a few expression sites which are kept polymorphic in a population of pathogens by homologous recombination us ing truncated pseudogenes present elsewhere in the genome. The Ehrlichia species on the other hand use transcriptional regulation from a series of more divergent full-length genes.

PAGE 142

132 CHAPTER 7 DISCUSSION AND CONCLUSIONS Despite the importance of the Anaplasmataceae for livestock production and human health no satisfactory cont rol methods are available. The development of effective and safe vaccines has been hampered by our lack of knowledge on the interactions betw een immune responses and bacterial virulence factors. In this dissertation, I have uncovered some of the mechanisms involved in A. marginale Â’s ability to rapidly and continuously alter its phenotype. MSP3 expression, one of the major antigens re cognized by the bovine host, was found to be polymorphic during infection in ca ttle. Subsequent experiments revealed that these different transcr ipts originated from a singl e expression site within the genome. Diversity in expressed msp3 variants was created through a mechanism of segmental gene conver sion, using invariant pseudogenes present elsewhere in the genome. Anal ysis of the pathogenÂ’s genome exposed the presence of a large pat hogenicity island dedicated to this process. The locus not only contained two related multigene families, msp2 and msp3 , but was further characterized by the pr esence of numerous recombinase genes and recombination hot spots. Conservative estimates of the potential to create combinatorial diversity from six pseudoge nes far outstrip the number of variants needed by the pathogen for life-time persist ence in its bovine host or to

PAGE 143

133 circumvent the specific immune res ponses induced by killed vaccines. Additionally, comparison of msp3 pseudogenes between two different isolates of A. marginale reveals extensive differences , adding yet another challenge to the development of widely crossprotective vaccine strategies. Many persistent pathogens have evolved families of closely related variant epitopes (Plebanski et al ., 1999). Experiments presented in this dissertation reveal that despite the conservation of orthologous msp3 multigene families in the genera Anaplasma and Ehrlichia , each has developed distinct strategies to use them. Anaplasma species use gene conversion of single expressed gene to the evade the immune responses. Ehrlichia species, on the other hand, express mult iple different full-length msp3 orthologs simultaneously from different expression site s. It is, neverthel ess, possible that this divergent mechanism still plays a role in the creation of persistent infections. Closely related polymorphi c epitopes can generate impaired T-cells when presented by the same APC duri ng the induction phase of the immune response (Daniel, Grakoui, and Allen, 1998; Plebanski et al ., 1999) and cause a transitory inhibition of memory T-ce ll effector functions (Plebanski et al ., 1999). The msp3 orthologs in Ehrlichial pathogens could contribute to immunomodulation rather than evasion, but mo re research needs to be done to confirm this. The differences observed between Anaplasma and Ehrlichia spp. not only offer us new insights into the pathogenesis of the two genera, but also provide a clear indication that control strategi es, specifically recombinant vaccine technologies, will have to be tailored to each unique mechanism.

PAGE 144

134 Given the availability of such co mbinatorial mechanisms for achieving diversity in A. marginale , are there any possibilities for development of vaccines? MSP3 does, at first, not appear to be a good candidate for inclusion in any vaccine strategy given its substantia l antigenic diversity. However, it is one of the major antigens recognized by the immune system and could be crucial to success. The conservati on of the entire Cterminal end between MSP2 and MSP3, as demonstrated in these st udies, indicates that this region is essential to the pathogen. The C-termi nal region therefor e is an excellent candidate to be included in recombinant vaccine strategies aimed at inducing memory CD4 T-cell responses. My studies have further helped uncover the potential Achilles' heel of Anaplasma spp.: antigenic variation. Since antigenic variation appears to be one of the major virulence factors in Anaplasma infections, one can envision that any stra tegy aimed at influencing it will have a tremendous impact on the ability of the immune response to control the pathogen. I have identified a number of recombinase enzymes which could be the target of such an approach. In fu ture studies we need to investigate the effects of pharmaceuticals inhibiting thes e enzymes and the effects of different gene knock-outs. Knock-out and transformation models in Neisseria and Rickettsial pathogens have shown that this approach is feasible and they might hold the key to success.

PAGE 145

135 LIST OF REFERENCES Ajayi,S.A., Wilson,A.J ., and Campbell,R.S. (1978) Experimental bovine anaplasmosis: clinico-pathological and nutritional studies Res.Vet.Sci. 25 : 7681. Akkoyunlu,M., Fikrig,E. (2000) Gamma interferon dominates the murine cytokine response to the agent of human gr anulocytic ehrlichiosis and helps to control the degree of early rickettsemia Infect. Immun. 68 : 1827-1833. Akkoyunlu,M., Malawista,S.E., Anguita,J., and Fikrig,E. (2001) Exploitation of interleukin-8-induced neutrophil chemotaxis by the agent of human granulocytic ehrlichiosis Infect. Immun. 69 : 5577-5588. Alderink, F. J. and Dietrich, R. (1981) Anaplasmosis in Texas: epidemiologic and economic data from a questionnaire survey Proc. Nat. Anaplasmosis Conf. 7 , 27-44 . Alleman,A.R., Barbet,A.F. (1996) Evaluation of Anaplasma marginale major surface protein 3 (MSP3) as a diagnostic test antigen J. Clin. Microbiol. 34 : 270-276. Alleman,A.R., Kamper,S.M., Viseshakul ,N., and Barbet,A.F. (1993) Analysis of the Anaplasma marginale genome by pulsed-field electrophoresis . J Gen. Microbiol. 139 : 2439-2444. Alleman,A.R., Palmer,G.H., McGuire,T. C., McElwain,T.F., Perryman,L.E., and Barbet,A.F. (1997) Anaplasma marginale major surface protein 3 is encoded by a polymorphic, multigene family Infect. Immun. 65 : 156-163. Allred,D.R., McGuire,T.C., Palmer ,G.H., Leib,S.R., Harkins,T.M., McElwain,T.F., and Barbet,A.F. (1990) Mole cular basis for surface antigen size polymorphisms and conservation of a neut ralization-sensitive epitope in Anaplasma marginale Proc. Natl. Acad. Sci. U.S.A. 87 : 3220-3224. Andrew,H.R., Norval,R.A. (1989) The role of males of the bont tick (Amblyomma hebraeum) in the transmission of Cowdria ruminantium (heartwater) Vet. Parasitol. 34 : 15-23. Anthony,D.W., Roby,T.O. (1966) The ex perimental transmission of bovine anaplasmosis by 3 species of North American ticks Am.J.Vet.Res. 27 : 191-198.

PAGE 146

136 Arulkanthan,A., Brown,W. C., McGuire,T.C., and Kno wles,D.P. (1999) Biased immunoglobulin G1 isotype responses induced in cattle with DNA expressing msp1a of Anaplasma marginale Infect. Immun. 67 : 3481-3487. Bakken,J.S., Dumler,J.S., Chen,S.M., Eckman,M.R., Van Etta,L.L., and Walker,D.H. (1994) Human granulocytic ehrlichiosis in the upper Midwest United States. A new species emerging? J.A.M.A. 272 : 212-218. Banerjee,R., Anguita,J., Roos,D., and Fikr ig,E. (2000) Cutting edge: infection by the agent of human granulocytic ehrlichiosis prevents the respiratory burst by down-regulating gp91phox J. Immunol. 164 : 3946-3949. Barbet,A.F., Allred,D.R. (1991) Th e msp1 beta multigene family of Anaplasma marginale : nucleotide sequence analysis of an expressed copy Infect. Immun. 59 : 971-976. Barbet,A.F., Anderson,L.W., Palm er,G.H., and McGuire,T.C. (1983) Comparison of proteins synthesiz ed by two different isolates of Anaplasma marginale Infect. Immun. 40 : 1068-1074. Barbet,A.F., Blentlinger,R., Yi,J., Lundgr en,A.M., Blouin,E.F., and Kocan,K.M. (1999) Comparison of su rface proteins of Anaplasma marginale grown in tick cell culture, tick salivary glands, and cattle Infect. Immun. 67 : 102-107. Barbet,A.F., Lundgren,A., Yi,J., Ru rangirwa,F.R., and Palmer,G.H. (2000) Antigenic variation of Anaplasma marginale by expression of MSP2 mosaics Infect. Immun. 2000.No v.;68.(11.):6133.-8. 68 : 6133-6138. Barbet,A.F., Palmer,G.H., Myler,P.J., and McGuire,T.C. (1987) Characterization of an immunoprotective protein complex of Anaplasma marginale by cloning and expression of the gene codi ng for polypeptide Am105L Infect. Immun. 55 : 2428-2435. Barbet,A.F., Semu,S.M., Chigagur e,N., Kelly,P.J., Jongejan,F., and Mahan,S.M. (1994) Size variation of the major immunodominant protein of Cowdria ruminantium Clin. Diagn. Lab. Immunol. 1 : 744-746. Barbet,A.F., Whitmire,W.M., Kamper ,S.M., Simbi,B.H., Ganta,R.R., Moreland,A.L., Mwangi,D.M., McGuire,T.C., and Mahan, S.M. (2001a) A subset of Cowdria ruminantium genes important for imm une recognition and protection Gene 275 : 287-298. Barbet,A.F., Yi,J., Lundgren,A., McEw en,B.R., Blouin,E.F., and Kocan,K.M. (2001b) Antigenic variation of Anaplasma marginale : major surface protein 2 diversity during cyclic transmission between ticks and cattle Infect. Immun. 69 : 3057-3066.

PAGE 147

137 Barbour,A.G., Carter,C.J., Burman,N ., Freitag,C.S., Garon,C.F., and Bergstrom,S. (1991) Tandem inserti on sequence-like elements define the expression site for variable antigen genes of Borrelia hermsii Infect. Immun. 59 : 390-397. Barbour,A.G., Restrepo,B.I. (2000) An tigenic variation in vector-borne pathogens Em. Infect. Dis. 6 : 449-457. Beale,G.H., Thaithong,S., and Siripool,N. (1991) Isolation of clones of Plasmodium falciparum by micromanipulation Trans. R. Soc. Trop. Med. Hyg. 85 : 37. Bekker,C.P., Bell-Sakyi,L., Paxton,E .A., Martinez,D., Bensaid,A., and Jongejan,F. (2002) Transcriptional analys is of the major antigenic protein 1 multigene family of Cowdria ruminantium Gene 285 : 193-201. Biswas,B., Vemulapalli,R., and Dutta,S.K . (1998) Molecular basis for antigenic variation of a protective strain-specific antigen of Ehrlichia risticii Infect. Immun. 66 : 3682-3688. Blood D.C., Radostitis, O.M., Henderson, J. A. (1983) In : Veterinary Medicine, 6th Ed., Bailliere Tindall, London, U.K.. Blouin,E.F., Barbet,A.F., Yi,J., Kocan,K. M., and Saliki,J.T. (2000) Establishment and characterization of an Oklahoma isolate of Anaplasma marginale in cultured Ixodes scapularis cells Vet. Parasitol. 87 : 301-313. Blouin,E.F., Kocan,K.M. (1998) Morphology and development of Anaplasma marginale (Rickettsiales: Anaplasmataceae) in cultured Ixodes scapularis (Acari: Ixodidae) cells J. Med. Entomol. 35 : 788-797. Blouin,E.F., Kocan,K.M., de la Fuente,J ., and Saliki,J.T. (2002) Effect of tetracycline on development of Anaplasma marginale in cultured Ixodes scapularis cells Vet. Parasitol. 107 : 115-126. Blouin,E.F., Kocan,K.M., and Ewing,S.A. ( 1992) Preliminary attempts to infect bovine erythrocytes in vitr o with a tick-derived stage of Anaplasma marginale Ann. N. Y. Acad. Sci. 653 : 72-77. Blouin,E.F., Kocan,K.M., Murphy,G.L., and Ge,N. (1993) Persistence of tickderived Anaplasma marginale in cultured bovine turb inate and endothelial cells Rev. Elev. Med. Vet. Pays Trop. 46 : 49-56. Bock,R.E., de Vos,A.J. (2001) Immunity following use of Australian tick fever vaccine: a review of the evidence Aust.Vet.J. 79 : 832-839.

PAGE 148

138 Borjesson,D.L., Simon,S.I., Hodzic,E ., Ballantyne,C.M., and Barthold,S.W. (2002) Kinetics of CD11b/CD18 up-regulati on during infection with the agent of human granulocytic ehrlichiosis in mice Lab. Invest. 82 : 303-311. Borst,P., Rudenko,G., Blundell,P. A., van Leeuwen,F., Cross,M.A., McCulloch,R., Gerrits,H., and Chaves,I.M . (1997) Mechanisms of antigenic variation in African trypanosomes Behring Inst.Mitt. : 1-15. Bowie,M.V., de la Fuente.J., Kocan,K. M., Blouin,E.F., and Barbet,A.F. (2002) Conservation of major surf ace protein 1 genes of Anaplasma marginale during cyclic transmission between ticks and cattle Gene 282 : 95-102. Bowles,P.M., Molloy,J.B., Blight,G.W., Singh,S., and M abikacheche,L.G. (2000) Evaluation of a commercially available ELISA kit for detection of antibodies to Anaplasma marginale and Anaplasma centrale in cattle in Australia and Zimbabwe Onderstepoort J. Vet. Res. 67 : 83-86. Bram,R.A., Roby,T.O. (1970) Atte mpted hereditary transmission of Anaplasma marginale Theiler with Dermacentor albipictus (Packard) J.Parasitol. 56 : 620621. Brayton,K.A., Knowles,D. P., McGuire,T.C., and Palm er,G.H. (2001) Efficient use of a small genome to generate antigeni c diversity in tick-borne Ehrlichial pathogens. Proc. Natl. Acad. Sci. U.S.A. 98 : 4130-4135. Brayton,K.A., Palmer,G.H., Lundgren,A., Yi,J., and Barbet,A.F. (2002) Antigenic variation of Anaplasma marginale msp2 occurs by combinatorial gene conversion Mol. Microbiol. 43 : 1151-1159. Brizuela,C.M., Ortellado,C. A., Sanabria,E., Torres,O ., and Ortigosa,D. (1998) The safety and efficacy of Australian ti ck-borne disease vaccine strains in cattle in Paraguay Vet. Parasitol. 76 : 27-41. Brouqui,P., Dumler,J.S., Raoult,D., and Walker,D.H. (1992) Antigenic characterization of Ehrlichia e: protein immunoblotting of Ehrlichia canis , Ehrlichia sennetsu , and Ehrlichia risticii J. Clin. Microbiol. 30 : 1062-1066. Brown,W.C., Rice-Ficht,A.C., and Estes, D.M. (1998) Bovine type 1 and type 2 responses Vet. Immunol. Immunopathol. 63 : 45-55. Brown,W.C., Shkap,V., Zhu,D., McGu ire,T.C., Tuo,W., McElwain,T.F., and Palmer,G.H. (1998a) CD4(+) T-lymphocyt e and immunoglobulin G2 responses in calves immunized with Anaplasma marginale outer membranes and protected against homologous challenge Infect. Immun. 66 : 5406-5413.

PAGE 149

139 Brown,W.C., Zhu,D., Shkap,V., McGuir e,T.C., Blouin,E.F., Kocan,K.M., and Palmer,G.H. (1998b) The repertoire of Anaplasma marginale antigens recognized by CD4(+) T-lymphocyte clones from protectively immunized cattle is diverse and includes major su rface protein 2 (MSP-2) and MSP-3 Infect. Immun. 66 : 5414-5422. Buening,G.M. (1973) Cell-mediated i mmune responses in calves with anaplasmosis Am. J. Vet. Res. 34 : 757-763. Buening,G.M. (1976) Cell-mediated i mmune response in anaplasmosis as measured by a micro cell-mediated cytot oxicity assay and leukocyte migrationinhibition test Am. J. Vet. Res. 37 : 1215-1218. Bunnell,J.E., Trigiani,E.R., Srinivas,S.R ., and Dumler,J.S. (1999) Development and distribution of pathologic lesions ar e related to immune status and tissue deposition of human granulocytic ehrlichios is agent-infected cells in a murine model system J. Infect. Dis. 180 : 546-550. Burgdorfer,W., Barbour,A.G., Hayes, S.F., Benach,J.L., Grunwaldt,E., and Davis,J.P. (1982) Lyme dis ease-a tick-borne spirochetosis? Science 216 : 13171319. Camacho-Nuez,M., de Lourdes,M., Suar ez,C.E., McGuire,T.C., Brown,W.C., and Palmer,G.H. (2000) Expression of polymorphic msp1beta genes during acute Anaplasma marginale rickettsemia Infect. Immun. 68 : 1946-1952. Carson,C.A., Buening,G.M. (1979) The imm une response of cattle to live and inactivated Anaplasma vaccines and response to challenge J. S. Afr. Vet. Assoc. 50 : 330-331. Carson,C.A., Sells,D.M., and Ristic,M. (1976) Cutaneous hy persensitivity and isoantibody production in cattle in jected with live or inactivated Anaplasma marginale in bovine and ovine erythrocytes Am. J. Vet. Res. 37 : 1059-1063. Carson,C.A., Sells,D.M., and Ristic,M. (1977a) Cell-mediated immune response to virulent and attenuated Anaplasma marginale administered to cattle in live and inactivated forms Am. J. Vet. Res. 38 : 173-179. Carson,C.A., Sells,D.M., and Ristic,M. (1977b) Cell-mediated immunity related to challenge exposure of cattle inoculat ed with virulent and attenuated strains of Anaplasma marginale Am. J. Vet. Res. 38 : 1167-1172. Casadevall,A. (1998) Antibody-mediat ed protection against intracellular pathogens Trends Microbiol. 6 : 102-107. Chen,D., Copeman,D.B., Burnell,J., and Hu tchinson,G.W. (2000) Helper T cell and antibody responses to infection of CBA mice with Babesia microti Parasite Immunol. 22 : 81-88.

PAGE 150

140 Chen,S.M., Popov,V.L., Feng,H.M., and Walker,D.H. (1996) Analysis and ultrastructural localization of Ehrlichia chaffeensis proteins with monoclonal antibodies Am. J. Trop. Med. Hyg. 54 : 405-412. Corrier,D.E., Vizcaino,O., Ca rson,C.A., Ristic,M., Kuttler ,K.L., Trevino,G.S., and Lee,A.J. (1980) Comparison of three me thods of immunization against bovine anaplasmosis: an examination of postvaccinal effects Am. J. Vet. Res. 41 : 1062-1065. Daniel,C., Grakoui,A., and Allen,P.M. ( 1998) Inhibition of an in vitro CD4(+) T cell alloresponse using altered peptide ligands Journal of Immunology 160 : 3244-3250. Daniels,M.J., Longland,J.M., and Gilbar t,J. (1980) Aspects of Motility and Chemotaxis in Spiroplasmas Journal of General Microbiology 118 : 429-436. Davis,W.C., Hamilton,M.J. (1998) Compar ison of the unique characteristics of the immune system in different species of mammals Vet. Immunol. Immunopathol. 63 : 7-13. de la Fuente.J., Garcia-Garcia,J.C ., Blouin,E.F., and Kocan,K.M. (2001a) Differential adhesion of major su rface proteins 1a and 1b of the Ehrlichia l cattle pathogen Anaplasma marginale to bovine erythrocytes and tick cells Int. J. Parasitol. 31 : 145-153. de la Fuente.J., Garcia-Garcia,J.C., Blouin,E.F., McEwen,B.R., Clawson,D., and Kocan,K.M. (2001b) Major surface protein 1a effects tick infection and transmission of Anaplasma marginale Int. J. Parasitol. 31 : 1705-1714. de la Fuente.J., Kocan,K.M. (2001c) Expression of Anaplasma marginale major surface protein 2 variants in persistently infected ticks Infect. Immun. 69 : 51515156. de la Fuente, J., Kocan,K.M.,Garcia-Garci a,J.C, Blouin,E.F., Claypool,P.L. and Saliki,J.T. (2002) Vaccination of cattle with Anaplasma marginale derived from tick cell culture and bovine erythrocytes followed by challenge-exposure by infected ticks Vet. Microbiol. 89 : 239-251. de Villiers,E.P., Brayton,K.A., Zweygar th,E., and Allsopp,B.A. (2000) Genome size and genetic map of Cowdria ruminantium Microbiology 146 : 2627-2634. de Vos,A.J., McGregor,W., Stewart,N.P., Dalgliesh,R.J., and Barry,D.N. (1987) Influence of imidocarb prophylaxis on the pathogenicity and immunogenicity of Anaplasma marginale infections in cattle Aust. Vet. J. 64 : 83-85. Deitsch,K.W., Moxon,E.R., and Welle ms,T.E. (1997) Shared themes of antigenic variation and virulence in bac terial, protozoal, and fungal infections. Microbiol. Mol. Biol. Rev. 61 : 281-293.

PAGE 151

141 Dikman,G. (1950) The trans mission of anaplasmosis Am. J. Vet. Res . 11 :5-16. Du Plessis,J.L. (1970) Imm unity in heartwater. I. A preliminary note on the role of serum antibodies Onderstepoort J. Vet. Res. 37 : 147-149. Du Plessis,J.L., Berche,P., and Van Gas, L. (1991) T cell-mediated immunity to Cowdria ruminantium in mice: the protective role of Lyt-2+ T cells Onderstepoort J. Vet. Res. 58 : 171-179. Dumler,J.S., Asanovich,K.M., Bakk en,J.S., Richter,P., Kimsey,R., and Madigan,J.E. (1995) Serologic cross-reactions among Ehrlichia equi , Ehrlichia phagocytophila , and human granulocytic Ehrlichia J. Cli n. Microbiol. 33 : 10981103. Dumler,J.S., Barbet,A.F., Bekker,C.P., Dasch,G.A., Pa lmer,G.H., Ray,S.C., Rikihisa,Y., and Rurangirwa,F.R. (2001) Reorganization of genera in the families Rickettsiaceae and Anaplasmataceae in the order Rickettsiales: unification of some species of Ehrlichia with Anaplasma , Cowdria with Ehrlichia and Ehrlichia with Neorickettsia , descriptions of six new species combinations and designation of Ehrlichia equi and 'HGE agent' as subjective synonyms of Ehrlichia phagocytophila Int. J. Syst. Evol. Microbiol. 51 : 2145-2165. Erratum in Int. J. Syst. Evol. Microbiol. (2002) 52 : 5-6. Dybvig,K., Voelker,L.L. (1996) Mo lecular biology of mycoplasmas Annu. Rev. Microbiol. 50 : 25-57. Eid,G., French,D.M., Lundgren,A.M., Barbet,A.F., McElwain,T.F., and Palmer,G.H. (1996) Expression of major surface protein 2 antigenic variants during acute Anaplasma marginale rickettsemia Infect. Immun. 64 : 836-841. Eriks,I.S., Stiller,D., and Palmer ,G.H. (1993) Impact of persistent Anaplasma marginale rickettsemia on tick infection and transmission J. Clin. Microbiol. 31 : 2091-2096. Erp,E., Fahrney,D. (1975) Exit of Anaplasma marginale from bovine red blood cells Am. J. Vet. Res. 36 : 707-709. Estes,D.M., Closser,N.M., and Allen,G.K. (1994) IFN-gamma stimulates IgG2 production from bovine B cells co stimulated with anti-mu and mitogen Cell Immunol. 154 : 287-295. Ewing,S.A. (1981) Transm ission of Anaplasma marginale by arthropods Proc. Nat. Anaplasmosis Conf . 7 . 395-423. Ferreira,A.M., Suzart,S., Vidotto,O., K nowles,D.P., and Vidotto,M.C. (2001) Use of repetitive DNA elements to de fine genetic relationships among Anaplasma marginale isolates FEMS Microbiol.Lett. 197 : 139-143.

PAGE 152

142 Ferris,D.H. (1972) Cinem atographic recording of Paranaplasma caudata and Plasmodium gallinaceum : in vitro kinetics Exp. Parasitol. 31 : 60-66. Francis,D.H., Buening,G.M., and Ameraul t,T.E. (1980) Characterization of immune responses of cattle to erythrocyte stroma, Anaplasma antigen, and dodecanoic acid-conjugated Anaplasma antigen: humoral immunity Am. J. Vet. Res. 41 : 362-367. Francis,D.H., Kinden,D.A., and Buening,G.M. (1979) Characterization of the inclusion limiting membrane of Anaplasma marginale by immunoferritin labeling Am. J. Vet. Res. 40 : 777-782. French,D.M., Brown,W. C., and Palmer,G.H. (1999) Emergence of Anaplasma marginale antigenic variants duri ng persistent rickettsemia Infect. Immun. 67 : 5834-5840. French,D.M., McElwain,T.F., McGuire,T. C., and Palmer,G.H. (1998) Expression of Anaplasma marginale major surface protein 2 variants during persistent cyclic rickettsemia Infect. Immun. 66 : 1200-1207. Friedhoff,K.T., Ristic,M. (1966) Anaplasm osis. XIX. A preliminary study of Anaplasma marginale in Dermacentor andersoni (Stiles) by fluorescent antibody technique Am. J. Vet. Res. 27 : 643-646. Frohlich,J., Konig,H. (1999) Rapid isolation of single microbial cells from mixed natural and laboratory populations with the aid of a micromanipulator Syst. Appl. Microbiol. 22 : 249-257. Gale,K.R., Gartside,M .G., Dimmock,C.M., Zakr zewski,H., and Leatch,G. (1996a) Peripheral blood lym phocyte proliferative responses in cattle infected with or vaccinated against Anaplasma marginale Parasitol. Res. 82 : 551-562. Gale,K.R., Leatch,G., DeVos,A. J., and Jorgensen,W.K. (1996b) Anaplasma marginale : effect of challenge of cattle with varying doses of infected erythrocytes Int. J. Parasitol. 26 : 1417-1420. Gale,K.R., Leatch,G., Gartside, M., and Dimmock,C.M. (1992) Anaplasma marginale : failure of sera from immune cattle to confer protection in passivetransfer experiments Parasitol. Res. 78 : 410-415. Ganta,R.R., Wilkerson,M. J., Cheng,C., Rokey,A.M. , and Chapes,S.K. (2002) Persistent Ehrlichia chaffeensis infection occurs in the absence of functional major histocompatibility complex class II genes Infect. Immun. 70 : 380-388.

PAGE 153

143 Ge,N.L., Kocan,K.M., Blouin,E.F., and Murphy,G.L. (1996) Developmental studies of Anaplasma marginale (Rickettsiales:Anaplasmataceae) in male Dermacentor andersoni (Acari:Ixodidae) infected as adults by using nonradioactive in situ hybridization and microscopy . J Med. Entomol. 33 : 911920. Goff,W.L., Winward,L. D. (1985) Detection of geographic isolates of Anaplasma marginale , using polyclonal bovine ant isera and microfluorometry Am. J. Vet. Res. 46 : 2399-2400. Goncalves Ruiz,P.M., Passos,L.M., Martins,M.S., Patarroyo,J.H., and Ribeiro,M.F. (2002) Antigenic characteri zation of morphologically distinct Anaplasma marginale isolates using a panel of monoclonal antibodies Vet. Parasitol. 107 : 169-177. Goodger,W.J., Carpenter,T ., and Riemann,H. (1979) Estimation of economic loss associated with anaplasmosis in California beef cattle J. Am. Vet. Med. Assoc. 174 : 1333-1336. Gusa,A.A., Buller,R.S., St orch,G.A., Huycke,M.M., Ma chado,L.J., Slater,L.N., Stockham,S.L., and Massung,R.F. (2001) Identification of a p28 gene in Ehrlichia ewingii: evaluation of gene for use as a target for a species-specific PCR diagnostic assay J. Clin. Microbiol. 39 : 3871-3876. Hallab,N., Jacobs,J.J., and Black,J. (2000) Hypersensitiv ity to metallic biomaterials: a review of leuko cyte migration inhibition assays Biomaterials 21 : 1301-1314. Hawkins,J.A., Love,J.N., and Hidalgo,R.J. (1982) Mechanical transmission of anaplasmosis by tabanids (Diptera: Tabanidae) Am. J. Vet. Res. 43 : 732-734. Heinzen,R.A., Grieshaber,S.S., Van Kirk ,L.S., and Devin,C.J. (1999) Dynamics of actin-based movement by Rickettsia rickettsii in vero cells Infect. Immun. 67 : 4201-4207. Hidalgo,R.J. (1975) Propagation of Anaplasma marginale in bovine lymph node cell culture Am. J. Vet. Res. 36 : 635-640. Hidalgo,R.J., Jones,E.W., Brown, J.E., and Ainswort h,A.J. (1989) Anaplasma marginale in tick cell culture Am. J. Vet. Res. 50 : 2028-2032. Howell-Adams,B., Seifert,H .S. (2000) Molecular models accounting for the gene conversion reactions mediating gonoc occal pilin antigenic variation Mol. Microbiol. 37 : 1146-1158.

PAGE 154

144 Ijdo,J.W., Sun,W., Zhang,Y., Magnarelli,L.A., and Fikrig ,E. (1998) Cloning of the gene encoding the 44-kilodalton antigen of the agent of human granulocytic ehrlichiosis and characterization of the humoral response Infect. Immun. 66 : 3264-3269. Ismail,N., Olano,J.P., Feng,H.M., and Walk er,D.H. (2002) Current status of immune mechanisms of killing of intracellular microorganisms FEMS Microbiol. Lett. 207 : 111-120. Janeway,C.A. and Travers, P. (1997) In : Immunobiology, The immune system in health and disease. 3rd Ed., Garland Publishing Inc., New York. Janeway,C.A., Jr. (2001a) How the i mmune system protects the host from infection Microbes. Infect. 3 : 1167-1171. Janeway,C.A., Jr. (2001b) How the imm une system works to protect the host from infection: a personal view Proc. Natl. Acad. Sci. U.S.A. 98 : 7461-7468. Jorgensen,W.K., Bock,R.E., de Vos,A.J. , and Shiels,I.A. (1993) Sheep-adapted Anaplasma marginale maintains virulence for cattle Aust. Vet. J. 70 : 192-193. Kano,F.S., Vidotto,O., Pacheco,R.C ., and Vidotto,M.C. (2002) Antigenic characterization of Anaplasma marginale isolates from different regions of Brazil Vet. Microbiol. 87 : 131-138. Kessler,R.H., Ristic,M. (1979) In vitro cultivation of Anaplasma marginale : invasion of and development in noninfected erythrocytes Am. J. Vet. Res. 40 : 1774-1776. Kieser,S.T., Eriks,I.S., and Palmer,G.H . (1990) Cyclic rickettsemia during persistent Anaplasma marginale infection of cattle Infect. Immun. 58 : 11171119. Kocan,K.M. (1992) Recent advances in the biology of Anaplasma spp. in Dermacentor andersoni ticks Ann. N. Y. Acad. Sci. 653 : 26-32. Kocan,K.M. (1995) Targeting ticks fo r control of selected hemoparasitic diseases of cattle Vet. Parasitol. 57 : 121-151. Kocan,K.M., Bezuidenhout,J.D. (1987) Morphology and development of Cowdria ruminantium in Amblyomma ticks Onderstepoort J. Vet. Res. 54 : 177182. Kocan,K.M., Bezuidenhout,J.D., and Hart,A . (1987) Ultrastructural features of Cowdria ruminantium in midgut epithelial cells and salivary glands of nymphal Amblyomma hebraeum Onderstepoort J. Vet. Res. 54 : 87-92.

PAGE 155

145 Kocan,K.M., Blouin,E.F., and Barbet,A.F . (2000) Anaplasmosis control. Past, present, and future Ann. N. Y. Acad. Sci. 916 : 501-509. Kocan,K.M., Goff,W.L., Stiller,D., Cla ypool,P.L., Edwards,W., Ewing,S.A., Hair,J.A., and Barron,S.J. (1992a) Persistence of Anaplasma marginale (Rickettsiales: Anaplasmataceae) in male Dermacentor andersoni (Acari: Ixodidae) transferred successively from infected to susceptible calves J. Med. Entomol. 29 : 657-668. Kocan,K.M., Goff,W.L., Stiller,D., Ed wards,W., Ewing,S.A., Claypool,P.L., McGuire,T.C., Hair,J.A., and Barron, S.J. (1993) Development of Anaplasma marginale in salivary glands of male Dermacentor andersoni Am. J. Vet. Res. 54 : 107-112. Kocan,K.M., Hair,J.A., and Ewing, S.A. (1980) Ultrastructure of Anaplasma marginale Theiler in Dermacentor andersoni Stiles and Dermacentor variabilis (Say) Am. J. Vet. Res. 41 : 1966-1976. Kocan,K.M., Hair,J.A., Ewing,S.A., and Stratton,L.G. (1981) Transmission of Anaplasma marginale Theiler by Dermacentor andersoni Stiles and Dermacentor variabilis (Say) Am. J. Vet. Res. 42 : 15-18. Kocan,K.M., Halbur,T., Blouin,E.F., Onet,V., de la Fuente.J., GarciaGarcia,J.C., and Saliki,J.T. ( 2001) Immunization of cattle with Anaplasma marginale derived from tick cell culture Vet. Parasitol. 102 : 151-161. Kocan,K.M., Morzaria,S.P., Voigt,W.P. , Kiarie,J., and Irvin,A.D. (1987) Demonstration of colonies of Cowdria ruminantium in midgut epithelial cells of Amblyomma variegatum Am. J. Vet. Res. 48 : 356-360. Kocan,K.M., Oberst,R.D., Ewing,S.A ., Hair,J.A., and Barron,S.J. (1983) Demonstration of Anaplasma marginale in hemolymph of Dermacentor andersoni by animal inoculation and by fluorescent-antibody technique Am. J. Vet. Res. 44 : 798-801. Kocan,K.M., Stich,R.W., Claypool,P.L., Ewing,S.A., Hair,J.A., and Barron,S.J. (1990) Intermediate site of development of Anaplasma marginale in feeding adult Dermacentor andersoni ticks that were infected as nymphs Am. J. Vet. Res. 51 : 128-132. Kocan,K.M., Stiller,D. (1992a) Development of Anaplasma ovis (Rickettsiales: Anaplasmataceae) in male Dermacentor andersoni (Acari: Ixodidae) transferred from infected to susceptible sheep J. Med. Entomol. 29 : 98-107. Kocan,K.M., Stiller,D., Goff,W.L., Cla ypool,P.L., Edwards,W., Ewing,S.A., McGuire,T.C., Hair,J.A., and Barron, S.J. (1992b) Development of Anaplasma marginale in male Dermacentor andersoni transferred from parasitemic to susceptible cattle Am. J. Vet. Res. 53 : 499-507.

PAGE 156

146 Kocan,K.M., Venable,J.H., Hsu,K.C., and Brock,W.E. (1978) Ultrastructural localization of anaplasmal antigens (Paw huska isolate) with ferritin-conjugated antibody Am. J. Vet. Res. 39 : 1131-1135. Kocan,K.M., Wickwire,K.B ., Ewing,S.A., Hair,J.A., and Barron,S.J. (1988) Preliminary studies of the development of Anaplasma marginale in salivary glands of adult, feeding Dermacentor andersoni ticks Am. J. Vet. Res. 49 : 10101013. Kocan,K.M., Yellin,T.N., Claypool,P.L., Barron,S.J., Ewing,S.A., and Hair,J.A. (1990b) Development and infectivity of Anaplasma marginale in Dermacentor andersoni nymphs Am. J. Vet. Res. 51 : 1292-1294. Kowalczykowski,S.C. (2000) Initia tion of genetic recombination and recombination-depend ent replication Trends Biochem. Sci. 25 : 156-165. Kreier,J.P., Ristic,M. (1972) Definition and taxonomy of Anaplasma species with emphasis on morphologic and immunologic features Z. Tropenmed. Parasitol. 23 : 88-98. Krueger,C.M., Buening,G.M. (1988) Isol ation and restriction endonuclease cleavage of Anaplasma marginale DNA in situ in agarose J. Clin. Microbiol. 26 : 906-910. Kurtz,S., Choudhuri,J.V., Ohlebusch,E., Schleiermacher,C., Stoye,J., and Giegerich,R. (2001) REPuter: the manifold applications of repeat analysis on a genomic scale Nucleic Acids Res. 29 : 4633-4642. Kusch,J., Schmidt,H.J. (2001) Genetically controlled expression of surface variant antigens in free-living protozoa. J. Membr. Biol. 180 : 101-109. Kuttler,K.L. (1979) Current anaplasmosi s control techniques in the United States J. S. Afr. Vet. Assoc. 50 : 314-320. Kuttler,K.L. (1984) Anaplasma infections in wild and domestic ruminants: a review J. Wildl. Dis. 20 : 12-20. Kuttler,K.L. (1986) Dynamics of Anaplasma marginale in splenectomised calves treated with either imidoc arb or oxytetracycline Trop. Anim. Health. Prod. 18 : 91-96. Kuttler,K.L., Winward,L.D. (1984) Serologic comparisons of 4 Anaplasma isolates as measured by t he complement-fixation test Vet. Microbiol. 9 : 181186.

PAGE 157

147 Kuttler,K.L., Zaugg,J.L., and Johnson,L. W. (1984) Serologic and clinical responses of premunized, vaccinated, and previously infected cattle to challenge exposure by two different Anaplasma marginale isolates Am. J. Vet. Res. 45 : 2223-2226. Leal,M., Noda,A., Reyna-Bello,A., Ca sas,B., Precigout,E., Aso,P.M., Gorenflot,A., and Gonzatti,M.I. (2000) I dentification and characterization of corpuscular, soluble and secreted ant igens of a Venezuelan isolate of Anaplasma marginale Vet. Parasitol. 94 : 1-15. Li,J.S., Chu,F., Reilly,A., and Winslow,G.M. (2002) Antibodies highly effective in SCID mice during infection by the intracellular bacterium Ehrlichia chaffeensis are of picomolar affinity and exhibit preferential epitope and isotype utilization J. Immunol. 169 : 1419-1425. Logan,L.L., Whyard,T.C., Quintero,J .C., and Mebus,C.A. (1987) The development of Cowdria ruminantium in neutrophils Onderstepoort J. Vet. Res. 54 : 197-204. Lohr C.V., Brayton K.A., Shkap V., Mol ad T., Barbet A.F., Brown W.C., and Palmer G.H. (2002) Expression of Anapl asma marginale major surface protein 2 operon-associated proteins during mammalian and arthropod infection Infect. Immun. 70 : 6005-6012 Long,S.W., Zhang,X.F., Qi,H., Standaert,S ., Walker,D.H., and Yu,X.J. (2002) Antigenic variation of Ehrlichia chaffeensis resulting from differential expression of the 28-kilodalton protein gene family Infect. Immun. 70 : 1824-1831. McCallon,B.R. (1973) Pr evalence and economic aspects of anaplasmosis Proc. Nat. Anaplasmosis Conf. 6 , 1-3. Madigan,J.E., Richter,P.J.J., Kimsey,R .B., Barlough,J.E., Bakken,J.S., and Dumler,J.S. (1995) Transmission and pass age in horses of t he agent of human granulocytic ehrlichiosis J. Infect. Dis. 172 : 1141-1144. Mahan,S.M., Sileghem,M., Smith,G.E., and Byrom,B. (1996) Neutralization of bovine concanavalin-A T ce ll supernatant-mediated antiCowdria ruminantium activity with antibodies specific to inte rferon gamma but not to tumor necrosis factor Parasite Immunol. 18 : 317-324. Mahan,S.M., Smith,G.E., and Byrom,B. (1994) Conconavalin A-stimulated bovine T-cell supernatant s inhibit growth of Cowdria ruminantium in bovine endothelial cells in vitro Infect. Immun. 62 : 747-750. Martin,M.E., Bunnell,J.E., and Dumler,J .S. (2000) Pathology, immunohistology, and cytokine responses in early phases of human granulocytic ehrlichiosis in a murine model J. Infect. Dis. 181 : 374-378.

PAGE 158

148 Martin,M.E., Caspersen,K., and Du mler,J.S. (2001) Immunopathology and Ehrlichial propagation are regulated by interferon-gamma and interleukin-10 in a murine model of human granulocytic ehrlichiosis Am. J. Pathol. 158 : 18811888. Mazzola,V., Amerault,T.E., and R oby,T.O. (1976) Survival of Anaplasma marginale in Aedes albopictus cells Am. J. Vet. Res. 37 : 987-989. Mazzola,V., Kuttler,K.L. (1980) Anaplasma marginale in bovine erythrocyte cultures Am. J. Vet. Res. 41 : 2087-2088. McBride,J.W., Yu,X.J., and Walker,D.H. (2000) A conserved, transcriptionally active p28 multigene locus of Ehrlichia canis Gene 254 : 245-252. McGarey,D.J., Barbet,A.F., Palmer,G.H., McGuire,T.C., and Allred,D.R. (1994) Putative adhesins of Anaplasma marginale : major surface polypeptides 1a and 1b Infect. Immun. 62 : 4594-4601. McGuire,T.C., Davis,W.C., Brassfie ld,A.L., McElwain,T.F., and Palmer,G.H. (1991) Identification of Anaplasma marginale long-term carrier cattle by detection of serum antibody to isolated MSP-3 J. Clin. Microbiol. 29 : 788-793. McGuire,T.C., Musoke,A.J., and Kurtti,T. (1979) Functi onal properties of bovine IgG1 and IgG2: interaction with comp lement, macrophages, neutrophils and skin Immunology 38 : 249-256. McGuire,T.C., Palmer,G.H., Goff,W.L ., Johnson,M.I., and Davis,W.C. (1984) Common and isolate-restricted antigens of Anaplasma marginale detected with monoclonal antibodies Infect. Immun. 45 : 697-700. McHardy,N., Berger,J., Taylor,R.J., Farebrother,D ., and James,J.A. (1980) Comparison of gloxazone, an effective but toxic Anaplasmacide, with imidocarb dihydrochloride Res. Vet. Sci. 29 : 198-202. McLean,M.J., Wolfe,K.H., and Devine,K.M. (1998) Base composition skews, replication orientati on, and gene orientation in 12 prokaryote genomes J. Mol. Evol. 47 : 691-696. Meeus,P.F., Barbet,A.F. (2001) Ingenious gene generation Trends Microbiol. 9 : 353-355. Mena,A., Ioannou,X.P., Van Kessel,A ., Van Drunen Little-Van Den Hurk, Popowych,Y., Babiuk,L.A., and Godson,D.L. (2002) Th1/Th2 biasing effects of vaccination in cattle as determined by real-time PCR J. Immunol. Methods 263 : 11-21. Moens,S., Vanderleyden,J. (1997) Gl ycoproteins in prokaryotes Arch. Microbiol. 168 : 169-175.

PAGE 159

149 Molloy,J.B., Bock,R.E., Templeton, J.M., Bruyeres,A.G., Bowles,P.M., Blight,G.W., and Jorgensen,W.K. (2001) Identification of antigenic differences that discriminate betwe en cattle vaccinated with Anaplasma centrale and cattle naturally infected with Anaplasma marginale Int. J. Parasitol. 31 : 179-186. Morley,R.S., Hugh-Jones,M.E. (1989a) Seroepidemiology of Anaplasma marginale in white-tailed deer ( Odocoileus virginianus ) from Louisiana J. Wildl. Dis. 25 : 342-346. Morley,R.S., Hugh-Jones,M.E. (1989b) The cost of anaplasmosis in the Red River Plains and south-east areas of Louisiana Vet. Res. Commun. 13 : 349358. Morley,R.S., Hugh-Jones,M.E. (1989c ) The effect of management and ecological factors on the epidemiology of anaplasmosis in the Red River Plains and south-east areas of Louisiana Vet. Res. Commun. 13 : 359-369. Munderloh,U.G., Blouin,E.F., Kocan, K.M., Ge,N.L., Edwards,W.L., and Kurtti,T.J. (1996) Establishment of the tick (Acari:Ix odidae)-borne cattle pathogen Anaplasma marginale (Rickettsiales:Anaplasmataceae) in tick cell culture J. Med. Entomol. 33 : 656-664. Munderloh,U.G., Kurtti,T.J. (1989) Formula tion of medium for tick cell culture Exp. Appl. Acarol. 7 : 219-229. Munderloh,U.G., Liu,Y., Wang,M., Chen,C., and Kurtti,T.J. (1994) Establishment, maintenance and descrip tion of cell lines from the tick Ixodes scapularis J. Parasitol. 80 : 533-543. Munderloh, U.G., Linch M.J., Palmer, A. T. (2002) Infection of endothelial cells with human and bovine Anaplasma spp. IX International Conference on Lyme Borreliosis and other tick-borne diseases Abstract O-11. Murphy,C.I., Storey,J.R., Recchia,J., Doros-Richert,L.A., Gingrich-Baker,C., Munroe,K., Bakken,J.S., Coughlin,R.T., and Beltz,G.A. (1998) Major antigenic proteins of the agent of human granulocytic ehrlichiosis are encoded by members of a multigene family Infect. Immun. 66 : 3711-3718. Namba,K., Vonderviszt,F. (1997) Molecular architecture of bacterial flagellum Quarterly Reviews of Biophysics 30 : 1-65. Nash,T.E., Lujan,H.T., Mowatt,M.R., and C onrad,J.T. (2001) Variant-specific surface protein switching in Giardia lamblia Infect. Immun. 69 : 1922-1923. Newbold,C.I., Craig,A.G., Kyes,S., Be rendt,A.R., Snow,R.W ., Peshu,N., and Marsh,K. (1997) PfEMP1, polymorphism and pathogenesis Ann. Trop. Med. Parasitol. 91 : 551-557.

PAGE 160

150 Noormohammadi,A.H., Markham,P.F., Kanci,A., Whithear,K.G., and Browning,G.F. (2000) A novel mechanism fo r control of antigenic variation in the haemagglutinin gene family of Mycoplasma synoviae Mol. Microbiol. 35 : 911-923. Norton,J.H., Parker,R.J., and Forbes -Faulkner,J.C. (1983) Neonatal anaplasmosis in a calf Aust. Vet. J. 60 : 348. Nyindo,M., Kakoma,I., and Hansen,R. (1991) Antigenic analysis of four species of the genus Ehrlichia by use of protein immunoblot Am. J. Vet. Res. 52 : 12251230. Oberle,S.M., Barbet,A.F. (1993) Deri vation of the complete msp4 gene sequence of Anaplasma marginale without cloning Gene 136 : 291-294. Oberle,S.M., Palmer,G.H., and Barbe t,A.F. (1993) Expression and immune recognition of the conserved M SP4 outer membrane protein of Anaplasma marginale Infect. Immun. 61 : 5245-5251. Oberle,S.M., Palmer,G.H., Barbet,A.F., and McGuire,T.C. (1988) Molecular size variations in an immunoprotective protein complex am ong isolates of Anaplasma marginale Infect. Immun. 56 : 1567-1573. Oduola,A.M., Weatherly,N.F., Bowd re,J.H., and Desjardins,R.E. (1988) Plasmodium falciparum: cloning by si ngle-erythrocyte mi cromanipulation and heterogeneity in vitro Exp. Parasitol. 66 : 86-95. Ohashi,N., Rikihisa,Y., and Unver,A. (2001) Analysis of transcriptionally active gene clusters of major outer membr ane protein multigene family in Ehrlichia canis and E. chaffeensis Infect. Immun. 69 : 2083-2091. Ohashi,N., Unver,A., Zhi,N., and Rikihisa,Y. (1998a) Cloning and characterization of multigenes enc oding the immunodominant 30-kilodalton major outer membrane proteins of Ehrlichia canis and application of the recombinant protein for serodiagnosis J. Clin. Microbiol. 36 : 2671-2680. Ohashi,N., Zhi,N., Zhang,Y., and Rikihis a,Y. (1998b) Immunodominant major outer membrane proteins of Ehrlichia chaffeensis are encoded by a polymorphic multigene family Infect. Immun. 66 : 132-139. Palmer GH. (1989) Anaplasma vaccines In : Veterinary Protozoan and Hemoparasites, Editor Wright IG., CRC Press, Boca Raton: 1-29. Palmer,G.H., Abbott,J.R., French,D.M., and McElwain,T.F. (1998) Persistence of Anaplasma ovis infection and conserva tion of the msp-2 and msp-3 multigene families within the genus Anaplasma Infect. Immun. 66 : 6035-6039.

PAGE 161

151 Palmer,G.H., Barbet,A.F., C antor,G.H., and McGuire,T. C. (1989) Immunization of cattle with the MSP-1 surface protei n complex induces protection against a structurally variant Anaplasma marginale isolate Infect. Immun. 57 : 3666-3669. Palmer,G.H., Barbet,A.F., Davis,W.C., and McGuire,T.C. ( 1986a) Immunization with an isolate-common surf ace protein protects cattle against anaplasmosis Science 231 : 1299-1302. Palmer,G.H., Barbet,A.F., Kuttler,K.L., and McGuire,T.C. (1986b) Detection of an Anaplasma marginale common surface protein pr esent in all stages of infection J. Clin. Microbiol. 23 : 1078-1083. Palmer,G.H., Barbet,A.F., Mu soke,A.J., Katende,J.M., Ru rangirwa,F., Shkap,V., Pipano,E., Davis,W.C., and McGuire,T.C. (1988a) Rec ognition of conserved surface protein epitopes on Anaplasma centrale and Anaplasma marginale isolates from Israel, K enya and the United States Int. J. Parasitol. 18 : 33-38. Palmer,G.H., Brown,W.C., and Rurangirwa, F.R. (2000) Antigenic variation in the persistence and tr ansmission of the Ehrlichia Anaplasma marginale Microbes. Infect. 2 : 167-176. Palmer,G.H., Eid,G., Bar bet,A.F., McGuire,T.C., and McElwain,T.F. (1994a) The immunoprotective Anaplasma marginale major surface protein 2 is encoded by a polymorphic multigene family Infect. Immun. 62 : 3808-3816. Palmer,G.H., Kocan,K.M., Barron,S.J., Ha ir,J.A., Barbet,A.F., Davis,W.C., and McGuire,T.C. (1985) Presence of co mmon antigens, including major surface protein epitopes, between the cattle (int raerythrocytic) and tick stages of Anaplasma marginale Infect. Immun. 50 : 881-886. Palmer,G.H., McGuire,T.C. (1984) Immune serum against Anaplasma marginale initial bodies neutralizes infectivity for cattle J. Immunol. 133 : 10101015. Palmer,G.H., Munodzana,D., Tebele,N., Ushe,T., and McElwain,T.F. (1994b) Heterologous strain challe nge of cattle immunized with Anaplasma marginale outer membranes Vet. Immunol. Immunopathol. 42 : 265-273. Palmer,G.H., Oberle,S.M., Barbet,A .F., Goff,W.L., Davis,W.C., and McGuire,T.C. (1988b) Imm unization of cattle with a 36-kilodalton surface protein induces protection against homologous and heterologous Anaplasma marginale challenge Infect. Immun. 56 : 1526-1531. Palmer,G.H., Rurangirwa,F.R., Kocan,K. M., and Brown,W.C. (1999) Molecular basis for vaccine development against the Ehrlichial pathogen Anaplasma marginale Parasitol. Today 15 : 281-286.

PAGE 162

152 Palmer,G.H., Rurangirwa,F.R., and McElwa in,T.F. (2001) Strain composition of the Ehrlichia Anaplasma marginale within persistently infected cattle, a mammalian reservoir fo r tick transmission J. Clin. Microbiol. 39 : 631-635. Palmer,G.H., Waghela,S.D., Barbet,A.F., Davis,W.C., and Mc Guire,T.C. (1987) Characterization of a neutralization-s ensitive epitope on the Am 105 surface protein of Anaplasma marginale Int. J. Parasitol. 17 : 1279-1285. Plebanski,M., Lee,E.A.M., Hannan,C. M., Flanagan,K.L., Gilbert,S.C., Gravenor,M.B., and Hill,A. V.S. (1999) Altered pept ide ligands narrow the repertoire of cellular immune responses by interfering with T-cell priming Nature Medicine 5 : 565-571. Perez,J.M., Martinez,D., Debus,A., Sheikboudou,C., and Bensaid,A. (1997) Development of an in vitro cloning method for Cowdria ruminantium Clin. Diagn. Lab. Immunol. 4 : 620-623. Popov,V.L., Yu,X., and Walker,D.H. (2000) The 120 kDa outer membrane protein of Ehrlichia chaffeensis : preferential expression on dense-core cells and gene expression in Escherichia coli associated with attachment and entry Microb. Pathog. 28 : 71-80. Potgieter,F.T. and Van Rensburg,L. (1987) The persistence of colostral Anaplasma antibodies and incidence of in utero transmission of Anaplasma infections in calves, under laboratory conditions Onderstepoort J. Vet. Res. 54 : 557-560. Prozesky,L., Du Plessis,J.L. (1987) H eartwater. The development and life cycle of Cowdria ruminantium in the vertebrate hos t, ticks and cultured endothelial cells Onderstepoort J. Vet. Res. 54 : 193-196. Rauhut,R., Klug,G. (1999) mRNA degradation in bacteria Fems Microbiology Reviews 23 : 353-370. Reddy,G.R., Streck,C.P. (1999) Variability in the 28-kDa surface antigen protein multigene locus of isolates of the emerging disease agent Ehrlichia chaffeensis suggests that it plays a role in immune evasion Mol. Cell. Biol. Res. Commun. 1 : 167-175. Reddy,G.R., Sulsona,C.R., Barbet,A.F ., Mahan,S.M., Burridge,M.J., and Alleman,A.R. (1998) Mole cular characterization of a 28 kDa surface antigen gene family of the tribe Ehrlichiae Biochem. Biophys. Res. Commun. 247 : 636643. Reeves,J.D., III, Swift,B.L. (1977) Iatrogenic transmission of Anaplasma marginale in beef cattle Vet. Med. Small Anim. Clin. 72 : 911-914.

PAGE 163

153 Restrepo,B.I., Barbour,A.G. (1994) An tigen diversity in the bacterium B. hermsii through "somatic" mutations in rearranged vmp genes Cell 78 : 867-876. Richey,E.J., Brock,W.E., Kliewer,I.O., and Jones,E.W. (1977) Resistance to anaplasmosis after elimination of latent Anaplasma marginale infections Am. J. Vet. Res. 38 : 169-170. Rikihisa,Y., Ewing,S.A., and Fox,J.C. (1994) Western immunoblot analysis of Ehrlichia chaffeensis , E. canis , or E. ewingii infections in dogs and humans J. Clin. Microbiol. 32 : 2107-2112. Ristic,M., Sibinovic,S., Welter,C.J. (1968) Proc. Annu. Meet. U.S. Anim. Health Assoc. 72 , 56. Ristic,M., Carson,C.A. (1977) Methods of immunoprophylaxis against bovine anaplasmosis with emphasis on use of the attenuated Anaplasma marginale vaccine Adv. Exp. Med. Biol. 93 : 151-188. Roby,T.O., Amerault,T.E., Mazzola,V ., Rose,J.E., and Ilemobade,A. (1974) Immunity in bovine anaplasmosis after elimination of Anaplasma marginale infections with imidocarb Am. J. Vet. Res. 35 : 993-995. Rocha,E.P., Blanchard,A. (2002) Genomic repeats, genom e plasticity and the dynamics of Mycoplasma evolution Nucleic Acids Res. 30 : 2031-2042. Rocha,E.P., Danchin,A., and Viari,A. (1999) Analysis of long repeats in bacterial genomes reveals alternative evolutionary mechanisms in Bacillus subtilis and other competent prokaryotes Mol. Biol. Evol. 16 : 1219-1230. Rossouw,M., Neitz,A.W., de Waal,D.T., Du Plessi s,J.L., Van Gas,L., and Brett,S. (1990) Identification of the antigenic proteins of Cowdria ruminantium Onderstepoort J. Vet. Res. 57 : 215-221. Rurangirwa,F.R., Stiller,D., French,D.M., and Palmer,G.H . (1999) Restriction of major surface protein 2 (MSP2) variant s during tick transmission of the Ehrlichia Anaplasma marginale Proc. Natl. Acad. Sci. U.S.A. 96 : 3171-3176. Rurangirwa,F.R., St iller,D., and Palmer,G.H . (2000) Strain diversity in major surface protein 2 expression during tick transmission of Anaplasma marginale Infect. Immun. 68 : 3023-3027. Rutenberg,A.D., Grant,M. (2001) Curved tails in polymer ization-based bacterial motility Physical Review 6402 : art-021904. Samish,M., Pipano,E., and Hana,B. (1988) Cultivation of Anaplasma marginale from cattle in a Dermacentor cell line Am. J. Vet. Res. 49 : 254-256.

PAGE 164

154 Saliki,J.T., Blouin,E.F., Rodgers,S.J., and Kocan,K.M. (1998) Use of tick cell culture-derived Anaplasma marginale antigen in a competitive ELISA for serodiagnosis of anaplasmosis Ann. N. Y. Acad. Sci. 849 : 273-281. Sarkar,N. (1997) Polyadenylation of mRNA in prokaryotes Annual Review of Biochemistry 66 : 173-197. Schaefer,B.C. (1995) Revolutions in Rapi d Amplification of Cdna Ends New strategies for polymerase chain-reac tion cloning of full-length cDNA ends Analytical Biochemistry 227 : 255-273. Schuler,G.D., Altschul,S.F., and Lipman,D. J. (1991) A workbench for multiple alignment construction and analysis Proteins 9 : 180-190. Schwan,T.G., Piesman,J., Golde,W.T. , Dolan,M.C., and Rosa,P.A. (1995) Induction of an outer surface protein on Borrelia burgdorferi during tick feeding Proc. Natl. Acad. Sci. U.S.A. 92 : 2909-2913. Seifert,H.S. (1996) Questions about gonococcal pilus phaseand antigenic variation Mol. Microbiol . 21 : 433-440. Shkap,V., Molad,T., Brayton,K.A., Brown,W.C., and Palmer,G.H. (2002) Expression of major surface protein 2 va riants with conserved T-cell epitopes in Anaplasma centrale vaccinates Infect. Immun. 70 : 642-648. Shkap,V., Pipano,E., McGuire,T.C., and Pa lmer,G.H. (1991) I dentification of immunodominant polypeptides common between Anaplasma centrale and Anaplasma marginale Vet. Immunol. Immunopathol. 29 : 31-40. Siebinga,J.T., Jongejan,F. (2000) Tick-borne fever ( Ehrlichia phagocytophila infection) on a dairy farm in Friesland Tijdschr. Diergeneeskd. 125 : 74-80. Simpson,C.F., Kling,J.M., and Love,J.N. (1967) Mor phologic and histochemical nature of Anaplasma marginale Am. J. Vet. Res. 28 : 1055-1065. Singer,S.M., Elmendorf,H.G., Conrad,J.T. , and Nash,T.E. (2001) Biological selection of variant-specific surface proteins in Giardia lamblia J. Infect. Dis. 183 : 119-124. Sinkora,M., Sun,J., and Butler,J.E. (2000) Antibody repertoire development in fetal and neonatal piglets. V. VDJ gene chimeras resembling gene conversion products are generated at high frequency by PCR in vitro Mol. Immunol. 37 : 1025-1034. Skaar,E.P., Lazio,M.P., and Seifert,H.S . (2002) Roles of the recJ and recN genes in homologous recombination and DNA repair pathways of Neisseria gonorrhoeae J. Bacteriol. 184 : 919-927.

PAGE 165

155 Smith,G.R. (2001) Homologous recombi nation near and far fr om DNA breaks: alternative roles and contrasting views Annu. Rev. Genet. 35 : 243-274. Smith,R.D., Levy,M.G., Kuhlenschmidt,M.S., Adams,J.H., Rzechula,D.L., Hardt,T.A., and Kocan,K.M. (1986) Isolate of Anaplasma marginale not transmitted by ticks Am. J. Vet. Res. 47 : 127-129. Soler-Rodriguez,A.M., Romano,E., Ar anguren,Y., and Soyano,A. (1990) A new hemolytic assay for bovine serum co mplement and its application during experimental bovine anaplasmosis Vet. Immunol. Immunopathol. 24 : 347-360. Sourice,S., Biaudet,V., El Karoui,M., Ehrlich,S.D., and Gruss,A. (1998) Identification of the Chi site of Haemophilus influenzae as several sequences related to the Escherichia coli Chi site Mol. Microbiol. 27 : 1021-1029. Spiers,A.J., Sherratt,D.J. (1999) C-terminal interactions between the XerC and XerD site-specific recombinases Mol. Microbiol. 32 : 1031-1042. Stich,R.W., Kocan,K.M., Pa lmer,G.H., Ewing,S.A., Ha ir,J.A., and Barron,S.J. (1989) Transstadial and attempted tr ansovarial transmission of Anaplasma marginale by Dermacentor variabilis Am. J. Vet. Res. 50 :1377-1380. Stich,R.W., Kocan,K.M., Damian,R.T., and Fechheime r,M. (1997) Inclusion appendages associated with t he intraerythrocytic rickettsial parasite Anaplasma marginale are composed of bundl ed actin filaments Protoplasma 199 : 93-98. Stich,R.W., Shoda,L.K., Dr eewes,M., Adler,B., Jungi,T.W., and Brown,W.C. (1998) Stimulation of nitric ox ide production in macrophages by Babesia bovis Infect. Immun. 66 : 4130-4136. Stiller,D., Kocan,K.M., Edwards,W., Ewing,S.A., and Ba rron,J.A. (1989) Detection of colonies of Anaplasma marginale in salivary glands of three Dermacentor spp infected as nymphs or adults Am. J. Vet. Res. 50 : 1381-1385. Sulsona,C.R., Mahan,S. M., and Barbet,A.F. ( 1999) The map1 gene of Cowdria ruminantium is a member of a multigene fam ily containing both conserved and variable genes Biochem. Biophys. Res. Commun. 257 : 300-305. Swenson,C., Jacobs,R. (1986) Spherocytos is associated with anaplasmosis in two cows J. Am. Vet. Med. Assoc. 188 : 1061-1063. Swift,B.L., Thomas,G.M. (1983) Bovine An aplasmosis Elimination of the carrier state with injectable long-acting oxytetracycline J. Am. Vet. Med. Assoc. 183 : 63-65. Taylor-Robinson,A.W., Phillips,R.S. (1998) Infective dose modulates the balance between Th1and Th2-regulat ed immune responses during bloodstage malaria infection Scand. J. Immunol. 48 : 527-534.

PAGE 166

156 Tebele,N., McGuire,T.C., and Palmer,G.H . (1991) Induction of protective immunity by using Anaplasma marginale initial body membranes Infect. Immun. 59 : 3199-3204. Tebele,N., Palmer,G.H. (1991) Crossprote ctive immunity between the Florida and a Zimbabwe stock of Anaplasma marginale Trop. Anim. Health Prod. 23 : 197-202. Theiler, A. (1910) Anaplasma marginale. The marginal points in the blood of cattle suffering from a specific disease. In : Report of the Government Veterinary Bacteriologist (1908-1909). Transvaal Department of Agriculture, Transvaal, South Africa . Thompson,K.C., Todorovic,R.A., Mateus ,G., and Adams,L.G. (1978) Methods to improve the health of cattle in the tropics: immunisation and chemoprophylaxis against haemoparasitic infections Trop. Anim. Health Prod. 10 : 75-81. Tillett,D., Burns,B.P., and Neilan,B.A. ( 2000) Optimized rapid amplification of cDNA ends (RACE) for mapping bacterial mRNA transcripts Biotechniques 28 : 448-456. Tilney,L.G., Tilney,M.S. (1993) The wily wa ys of a parasite: induction of actin assembly by Listeria Trends Microbiol. 1 : 25-31. Totte,P., Blankaert,D., Zilimwabagabo,P ., and Werenne,J. (1993) Inhibition of Cowdria ruminantium infectious yield by interferons alpha and gamma in endothelial cells Rev. Elev. Med. Vet. Pays Trop. 46 : 189-194. Totte,P., Jongejan,F., De Gee,A.L., and Werenne,J. (1994) Production of alpha interferon in Cowdria ruminantium -infected cattle and its effect on infected endothelial cell cultures Infect. Immun. 62 : 2600-2604. Totte,P., Vachiery,N., Ma rtinez,D., Trap,I., Ball ingall,K.T., MacHugh,N.D., Bensaid,A., and Werenne,J. (1996) Re combinant bovine interferon gamma inhibits the growth of Cowdria ruminantium but fails to induce major histocompatibility complex class II following infection of endothelial cells Vet. Immunol. Immunopathol. 53 : 61-71. Trachtenberg,S. (1998) Mollicutes-wall-less bacteria with internal cytoskeletons J. Struct. Biol. 124 : 244-256. Tuo,W., Palmer,G.H., McGuire,T.C ., Zhu,D., and Brown,W.C. (2000) Interleukin-12 as an adjuvant promotes immunoglobulin G and type 1 cytokine recall responses to major surface pr otein 2 of the Ehrlichial pathogen Anaplasma marginale Infect. Immun. 68 : 270-280.

PAGE 167

157 Turton,J.A., Katsande,T.C., Matingo, M.B., Jorgensen,W.K., UshewokunzeObatolu,U., and Dalgliesh,R.J. (1 998) Observations on the use of Anaplasma centrale for immunization of cattle against anaplasmosis in Zimbabwe Onderstepoort J. Vet. Res. 65 : 81-86. Unver,A., Ohashi,N., Tajima,T., Stich, R.W., Grover,D., and Rikihisa,Y. (2001) Transcriptional analysis of p30 major out er membrane multigene family of Ehrlichia canis in dogs, ticks, and cell cultur e at different temperatures Infect. Immun. 69 : 6172-6178. Unver,A., Rikihisa,Y., Stich,R.W., Ohas hi,N., and Felek,S. (2002) The omp-1 major outer membrane multigene family of Ehrlichia chaffeensis is differentially expressed in canine and tick hosts Infect. Immun. 70 : 4701-4704. Valdez,R.A., McGuire,T.C., Brown, W.C., Davis,W.C. , Jordan,J.M., and Knowles,D.P. (2002) Select ive in vivo depletion of CD4(+) T lymphocytes with anti-CD4 monoclonal antibody during acute infection of calves with Anaplasma marginale Clin. Diagn. Lab. Immunol. 9 : 417-424. Vidotto,M.C., McGuire,T.C ., McElwain,T.F., Palmer, G.H., and Knowles,D.P.J. (1994) Intermolecular relationships of major surface proteins of Anaplasma marginale Infect. Immun. 62 : 2940-2946. Viseshakul,N., Kamper,S., Bowie,M.V., and Barbet,A.F. (2000) Sequence and expression analysis of a surface an tigen gene family of the rickettsia Anaplasma marginale Gene 253 : 45-53. Visser,E.S., McGuire,T.C., Palmer,G.H., Davis,W.C., Shkap,V., Pipano,E., and Knowles,D.P.J. (1992) The Anaplasma marginale msp5 gene encodes a 19kilodalton protein conserved in all recognized Anaplasma species Infect. Immun. 60 : 5139-5144. Vizcaino,O., Carson,C.A., Lee,A.J., and Risti c,M. (1978) Efficacy of attenuated Anaplasma marginale vaccine under laboratory and field conditions in Colombia Am. J. Vet. Res. 39 : 229-233. Vizcaino,O., Corrier,D.E., Terry,M.K., Carson,C.A., Lee,A.J., Kuttler,K.L., Ristic,M., and Trevino,G.S. (1980) Comparison of three methods of immunization against bovine anaplasmosis : evaluation of protection afforded against field challenge exposure Am. J. Vet. Res. 41 : 1066-1068. Waghela,S.D., Cruz,D., Droleskey,R. E., DeLoach,J.R., and Wagner,G.G. (1997) In vitro cultivation of Anaplasma marginale in bovine erythrocytes cocultured with endothelial cells Vet. Parasitol. 73 : 43-52. Wainwright,L.A., Pritchard,K.H., and Seifert,H.S. (1994) A conserved DNA sequence is required for efficient gonococcal pilin antigenic variation Mol. Microbiol. 13 : 75-87.

PAGE 168

158 Walker,D.H., Dumler,J.S. (1996) Emer gence of the ehrlichioses as human health problems Emerg. Infect. Dis. 2 : 18-29. Walls,J.J., Aguero-Rosenfeld,M., Ba kken,J.S., Goodman,J.L., Hossain,D., Johnson,R.C., and Dumler,J.S. (1999) Inte rand intralaborat ory comparison of Ehrlichia equi and human granulocytic ehrlich iosis (HGE) agent strains for serodiagnosis of HGE by the im munofluorescent-antibody test J. Clin. Microbiol. 37 : 2968-2973. Wang,B., Norbury,C.C., Greenwood,R ., Bennink,J.R., Yewdell,J.W., and Frelinger,J.A. (2001) Multiple paths fo r activation of naive CD8+ T cells: CD4independent help J. Immunol. 167 : 1283-1289. Wang,G., van Dam, A.P.., and Dankert,J. (2001) Analysis of a VMP-like sequence (vls) locus in Borrelia garinii and Vls homologues among four Borrelia burgdorferi sensu lato species Fems Microbiol. Letters 199 : 39-45. Wang,T., Malawista,S.E., Pal,U., Grey,M., Meek,J., Akkoyunlu,M., Thomas,V., and Fikrig,E. (2002) Superox ide anion production during Anaplasma phagocytophila infection J. Infect. Dis. 186 : 274-280. Wickwire,K.B., Kocan,K.M., Barron,S.J., Ewing,S.A., Smith,R. D., and Hair,J.A. (1987) Infectivity of three Anaplasma marginale isolates for Dermacentor andersoni Am. J. Vet. Res. 48 : 96-99. Winslow,G.M., Yager,E., Shilo,K., Volk ,E., Reilly,A., and Chu,F.K. (2000) Antibody-mediated elimination of the ob ligate intracellular bacterial pathogen Ehrlichia chaffeensis during active infection Infect. Immun. 68 : 2187-2195. Woldehiwet,Z. (1983) Tickborne fever: a review Vet. Res. Commun. 6 : 163175. Woldehiwet,Z. (1987a) Depression of ly mphocyte response to mitogens in sheep infected with tick-borne fever J. Comp. Pathol. 97 : 637-643. Woldehiwet,Z. (1987b) The effects of ti ck-borne fever on some functions of polymorphonuclear cells of sheep J. Comp. Pathol. 97 : 481-485. Wyatt,C.R., Davis,W.C., Knowles,D. P., Goff,W.L., Palmer,G.H., and McGuire,T.C. (1996) Effect on intraerythrocytic Anaplasma marginale of soluble factors from infected ca lf blood mononuclear cells Infect. Immun. 64 : 48464849. Xing,Z. (2000) Current understanding of macrophage type 1 cytokine responses during intracellular infections Histol. Histopathol. 15 : 199-205.

PAGE 169

159 Yu,X., McBride,J.W., Zhang,X., and Walk er,D.H. (2000a) Characterization of the complete transcriptionally active Ehrlichia chaffeensis 28 kDa outer membrane protein multigene family Gene 248 : 59-68. Yu,X.J., Crocquet-Valdes,P., and Walker ,D.H. (1997) Cloning and sequencing of the gene for a 120-kDa i mmunodominant protein of Ehrlichia chaffeensis Gene 184 : 149-154. Yu,X.J., McBride,J.W., Dia z,C.M., and Walker,D.H. ( 2000b) Molecular cloning and characterization of the 120-kilodalton pr otein gene of Ehrlichia canis and application of the recombi nant 120-kilodalton protein for serodiagnosis of canine ehrlichiosis J. Clin. Microbiol. 38 : 369-374. Yu,X.J., McBride,J.W., and Walker,D.H. (1999) Genetic dive rsity of the 28kilodalton outer membrane protein gene in human isolates of Ehrlichia chaffeensis J. Clin. Microbiol. 37 : 1137-1143. Zaugg,J.L., Kuttler,K.L. (1984) Bovine anap lasmosis: in utero transmission and the immunologic significance of ingested colostral antibodies Am. J. Vet. Res. 45 : 440-443. Zaugg,J.L. (1985) Bovine anaplasmosis: Tr ansplacental transmission as it relates to stage of gestation Am. J. Vet. Res. 46 :570-572 Zhang,J.R., Hardham,J.M., Barbour,A.G ., and Norris,S.J. (1997) Antigenic variation in Lyme disease borreliae by promiscuous recombination of VMP-like sequence cassettes Cell 89 : 275-285. Zhi,N., Ohashi,N., and Rikihisa,Y. (1999) Multiple p44 genes encoding major outer membrane proteins are expressed in the human granulocytic ehrlichiosis agent J. Biol. Chem. 274 : 17828-17836. Zhi,N., Ohashi,N., Rikihisa,Y., Horowit z,H.W., Wormser,G.P., and Hechemy,K. (1998) Cloning and expression of the 44-kilodalton major outer membrane protein gene of the human granulocytic ehrlichiosis ag ent and application of the recombinant protein to serodiagnosis J. Clin. Microbiol. 36 : 1666-1673.

PAGE 170

160 BIOGRAPHICAL SKETCH Patrick F.M. Meeus was born on September 19th, 1968, in Leuven, Belgium. He received a Bachelor of Veterinary Science degree in 1990 and a Doctor of Veterinary Medicine degree in 1993 from the University of Gent, Faculty of Veterinary Medicine, Belgiu m. During his veterinary training he chose clinical parasitology as one of hi s elective courses, and it was here that he was first introduced to research. With his interest sparked, he stayed on one more year in the Department of Parasi tology as a research assistant under Prof. Jozef Vercruysse, before taking on a position as lecturer and large animal clinician at the School of Veterinary M edicine, University of Zambia. His responsibilities included in-patient and ambulatory care of livestock and the development and management of a rural vete rinary field station. These duties were carried out as a service to t he farming community and as a means to teach the veterinary students. Du ring this period he managed a Danish government research capacit y-building project involv ed in helminth research under Prof. Peter Nansen from The Royal Veterinary and Agricultural University, Denmark. This includ ed epidemiological studies and the field evaluation of recombinant vaccines against Fasciola and Schistosoma . In 1997 he joined, as the project veterinarian, a Belgian government sponsored project in Zambia managed by the Tropical Institut e of Antwerp. The work focused on the development, testing, and producti on of a vaccine against the protozoal

PAGE 171

161 bovine parasite Theileria parva . During this period he realized a need for further specialized personal study to develop as an independent clinical research scientist. This would allow him to contribute to the field in a significant way. In August of 1998 he entered the In fectious Disease and Experimental Pathology graduate program at the University of Florida, College of Veterinary Medicine. This program and department enjoy world wide recognition in the field of vector borne diseases and molecu lar parasitology. He was admitted to candidacy for the PhD and has since been involved in research on vaccine development and molecular biology of rickettsial organisms, including both human and animal pathogens. After graduati on, Patrick will remain at the University of Florida working on the mo lecular biology of tick born pathogens with a NIH K08 grant he secured.