Antigenic Diversity in Eimeria maxima and the Influence of Host Genetics and Immunization Schedule on Cross-Protective Immunity

Eimeria spp. are a group of highly successful intracellularprotozoan parasites that develop within enterocytes. Eimeriamaxima from the chicken is characterized by high immunogenicity(a small priming infection gives complete immunity to subsequenthomologous challenge) and naturally occurring antigenicallyvariant populations that do not completely cross-protect. Inthis study we examined the expression of antigenic diversityin E. maxima, as manifested by cross-strain protection in aseries of inbred chicken lines. The IAH line of Light Sussexchickens and all lines of inbred White Leghorns were susceptibleto primary infections with either of two strains (H and W) ofE. maxima and were protected completely against challenge withthe homologous strain of parasite. The extent of cross-protectionagainst the heterologous parasite strain varied from 0 to almost100% depending on host genetics. Interestingly, in one inbredline of chickens (line 15I) the cross-protective phenotype wasdirectional and intensely influenced by the infection historyof the host. The basis for the observed variation in cross-protectionis not known, but our results suggest that the major histocompatibilitycomplex is not a major genetic component of the phenotype. Theseresults are discussed in relation to the number of protectiveantigens presented by complex pathogens and the developmentof immunoprotective responses in hosts of different geneticbackgrounds.

INTRODUCTION

Top

Introduction

Antigenic diversity is a feature of many classes of pathogens,including viruses (1, 9, 37, 72, 81), bacteria (3, 24, 32, 53,56, 85), protozoa (2, 25, 30, 36, 57, 60, 61), and helminths(86). There are two types of antigenic diversity. The firstarises by switching between variant proteins encoded by largepolymorphic gene families (12, 13, 20, 52, 57, 59, 62, 69) andfacilitates pathogen persistence within a single vertebratehost in the face of immune system attack (20, 26, 62, 80). Thesecond involves genetic diversity within the pathogen so thatan infection with one subpopulation may fail to induce cross-protectionagainst challenge with another (1, 2, 3, 9, 10, 23, 43). Immunologicallyrelevant genetic diversity within a pathogen population allowsdifferent strains of a pathogen to infect hosts that are immuneto other antigenic variants and is most likely driven by theadaptive immune response of the host population (8, 10). Theappearance in the field of pathogens that are antigenicallyvariant and immunologically non-cross-protective poses a significantcomplication to the design and successful application of vaccines.

Experimentally, studies of antigenic polymorphism by parasiteshave generally focused on the specificity of the immune responseinduced by infection rather than the specificity of immune protection.Studies that have examined protection have focused on a singlehost genotype rather than a broader investigation of any variationin protection associated with host genetics. In this paper wedemonstrate that host genetics and infection history are importantparts of the host-pathogen relationship and act in concert todetermine the level of cross-protective immunity generated byinfection.

Eimeria spp. are obligate intracellular apicomplexan protozoanparasites and are the cause of intestinal coccidiosis. Infectionsof the vertebrate host with these parasites result in the inductionof potent protective immune responses (63), and protection againstchallenge infections is confined absolutely to homologous species(63), even though both T- and B-cell responses exhibit cross-speciesreactivity (15).

Eimeria maxima is the most immunogenic of the seven speciesof Eimeria that infect the domestic chicken, and ingestion ofonly a few oocysts induces almost complete (>99.99%) protectionagainst challenge with homologous parasites (44, 65). However,different populations, including field isolates, exhibit substantialimmunological diversity (23, 43, 47, 48, 54, 75). The antigenicityof a strain of E. maxima appears to be stable with time (74),and this stability is distinct from the dynamic process of antigenicvariation that occurs with some other protozoa.

In this report we examined the contribution of host geneticsand infection history to the expression of cross-strain immuneprotection. We used two reference laboratory strains of E. maximathat cross protect incompletely in Light Sussex (LS) chickens(48) to assess their capacity to induce cross-protection insix inbred lines of White Leghorn chickens. While all hostswere susceptible to the two strains of E. maxima and were protectedcompletely against homologous challenge, we show for the firsttime that some hosts developed no immunity to challenge witha heterologous strain. Our results demonstrate the complexityof the immunological relationship, in terms of antigenic diversity,in eucaryotic pathogens and the genetics of their hosts. Theseissues are highly relevant to the development of vaccine strategiesagainst coccidiosis and parasitic organisms where polymorphismsin protective antigens have been identified.

MATERIALS AND METHODS

Top

Materials and Methods

Parasites. The Houghton (H) reference strain of E. maxima was isolatedin 1954 at the Houghton Laboratory (formerly the Houghton PoultryResearch Station) of the Institute for Animal Health. The Weybridge(W) reference strain was kindly provided by Janet Catchpoleof the Veterinary Laboratories Agency (Weybridge, Surrey, UnitedKingdom). Both strains are the progeny of a single oocyst. Methodsfor maintaining the parasites, preparing infective doses, andquantifying infections by detecting or counting the numbersof oocysts in feces between 5 and 10 days postinfection havebeen described elsewhere (45). The two strains of E. maximawere passaged at frequent intervals and used within 21 daysof sporulation for each experiment. All passages of the twostrains of parasites in chickens were undertaken in separateisolated rooms, each fitted with two antelobbies and fumigatedbefore and after use with methyl bromide. Furthermore, oocystsof the two strains of E. maxima were recovered from fecal materialin purpose-built rooms which were disinfected afterwards bya minimum of 7 h of exposure to gaseous ammonia. All experimentswere conducted in methyl bromide-treated rooms to avoid theingress of coccidial parasites.

Animals. The IAH strain of outbred LS chickens and six inbred lines ofWhite Leghorn chickens, lines 6₁, 7₂, 15I, N, C.B4, and C.B12,were used because of their different susceptibilities to infectionswith E. maxima (18). Lines C.B4 and C.B12 have the same geneticbackground but are different at the B locus (major histocompatibilitycomplex [MHC]) (18). Lines 6₁ and 7₂ have the same B locus haplotype(B2) but differ at many other non-B genetic loci.

A hemagglutination assay using blood collected by venipuncturefrom the brachial vein in heparinized tubes (Vacutainer SystemsEurope, Becton Dickinson) (17) was used to determine the MHC(B locus) of the progeny of a backcross between line 15I andC.B12 chickens; individuals were either homozygous for B15 MHCor heterozygous for B15 and B12 MHC, with 75% of the non-MHCgenes derived from the line 15I background.

Chickens were raised free from coccidia and used at 3 to 4 weeksold. For experiments they were housed either individually (experiment7) or in groups of three or four in metal cages with wire floorsthat were covered with plastic mesh inserts for birds up to6 weeks old. Throughout they had unlimited access to unmedicatedfood and water.

Design of experiments. For all experiments chickens were given a primary infectionof 100 oocysts of E. maxima at 3 or 4 weeks of age and, togetherwith a group of previously uninfected control birds housed ina separate room, were challenged 3 weeks later with 250 oocystsof the same or a heterologous strain. In all experiments themean values given in the figures were calculated from the resultsof experiments with four subgroups of at least three (usuallyfour) chickens. Percent cross-protection was calculated as 100x [1 - (number of oocysts per bird produced by immunized birdsafter challenge/number of oocysts per bird produced by unimmunizedbirds after challenge)].

RESULTS

Top

Results

Two reference strains of E. maxima do not completely cross protect in LS chickens. In experiment 1 (Fig. 1) groups of 3-week-old LS chickens weregiven 100 oocysts of either the H or W strain of E. maxima andthen challenged 3 weeks later with 250 oocysts of the homologousor heterologous strain.

View larger version (16K):
[in this window]
[in a new window]
FIG. 1. Outputs of oocysts of H and W strains of E. maxima in LS chickens arising from primary infections and challenge with either homologous or heterologous strains. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

Primary infection with either strain induced complete protection(>99.99%) against homologous challenge (Fig. 1; Table 1),whereas substantial numbers of oocysts were produced after challengewith the heterologous parasite to give cross-protection valuesof 63 (H, primary, W, challenge) and 79% (W, primary, H, challenge).Significantly more oocysts were produced by the W strain inthe 3-week-old birds used at the start of the experiment, butno differences were found when the birds were 6 weeks of age(Fig. 1).

View this table:
[in this window]
[in a new window]
TABLE 1. Influence of host genetics and infection history on cross-protective immunity between two strains of E. maxima

Host genetics and the order of immunization affect the level of cross-protection between the H and W strains of E. maxima. Six inbred lines of White Leghorn chickens were immunized witheither the H or W strain of E. maxima and then challenged witheither the homologous or heterologous strain 3 weeks later.Line 15I birds were consistently more susceptible to primaryinfection with either strain of E. maxima than lines N, 6₁,7_2, C.B12, and C.B4 (Fig. 2 3 to 4). Complete protection againstthe homologous strain of parasite was found in all six inbredlines (Fig. 2 and 3).

View larger version (20K):
[in this window]
[in a new window]
FIG. 2. Outputs of oocysts of H and W strains of E. maxima in four lines of White Leghorn chickens arising from primary infections and challenge with the homologous or heterologous W strain. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

View larger version (21K):
[in this window]
[in a new window]
FIG. 3. Outputs of oocysts of H and W strains of E. maxima in four lines of White Leghorn chickens arising from primary infections and challenge with the homologous or heterologous H strain. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

View larger version (24K):
[in this window]
[in a new window]
FIG. 4. Outputs of oocysts of H and W strains of E. maxima in lines C.B4 and C.B12 of White Leghorn chickens arising from primary infections and challenge with either homologous or heterologous strains. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

In contrast to complete protection against reinfection withthe homologous strain of E. maxima, the level of protectionagainst heterologous challenge infection varied dramatically,depending on the line of chicken (Fig. 1 2 3 4 5 to 6 and Table1). Line C.B4 and C.B12 chickens failed to develop any measurabledegree of cross-immunity (Fig. 2 and 4; Table 1), whereas lineN chickens always developed approximately 70 to 80% cross-strainimmunity (Fig. 2, 3, and 6; Table 1). Line 6₁ and 7₂ chickensusually expressed no, or low levels of, cross-protective immunity(Fig. 2, 3, and 5; Table 1).

View larger version (19K):
[in this window]
[in a new window]
FIG. 5. Outputs of oocysts of H and W strains of E. maxima in lines 6₁ and 7₂ of White Leghorn chickens arising from primary infections and challenge with either homologous or heterologous strains. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

View larger version (18K):
[in this window]
[in a new window]
FIG. 6. Outputs of oocysts of H and W strains of E. maxima in lines 15I and N of White Leghorn chickens arising from primary infections and challenge with either homologous or heterologous strains. The values are means calculated from the results of experiments with four subgroups of at least three chickens. Standard errors of the group means are indicated. Groups within columns annotated with different letters differ significantly (P < 0.05) from each other. ND, not done.

In most cases, the choice of immunizing and challenge strainsof parasite had little, if any, effect on the level of cross-protectiveimmunity generated by a particular chicken genotype. For exampleline N chickens were always 70% cross-protected and line C.B4or C.B12 chickens were never cross-protected. However, withline 7₂ and line 15I the order of infection affected the levelof cross-protection against challenge with the heterologousstrain of parasite. This feature was most dramatic with line15I chickens, whereby strong cross-protection was obtained afterimmunization with the H strain of E. maxima (95 [Fig. 6] and98% [Fig. 2]) and substantially less protection was obtainedafter immunization with the W strain (0 [Fig. 3] and 60% [Fig.6]). The experiment with line 15I chickens was repeated a thirdtime with different immunizing doses of the W strain, and nocross-protection was observed after challenge with the H strain(data not shown).

Exposure of line 15I chickens to the W strain of E. maxima does not prevent the development of immunity to the H strain. To test whether the complete lack of cross-protection betweenthe W and H strains in line15I chickens (after immunizationwith the W strain) was due to specific immune tolerance, wechallenged chickens a third time with both strains of E. maxima.Line N chickens, which exhibit reasonably strong cross-protection,were also examined in this experiment. Outputs of oocysts duringthe primary infections and the degrees of cross-protection againstsecondary challenges were consistent with results of earlierexperiments, viz., line N chickens were well protected againstheterologous challenge (approximately 90 to 95%), whereas inline 15I birds the outcomes of protection to secondary heterologouschallenge were determined by the order in which the parasitestrains were given (Fig. 6; Table 1). Thus, line 15I birds givena primary infection with the H strain were completely protectedagainst challenge with the W strain but only 60% protected whenthe W strain preceded a challenge with the H strain.

In comparison to the outputs of oocysts from an age-matchedcontrol group, all line N and line 15I birds given both theH and W strains serially were immune to a tertiary infection(Fig. 6). However, line 15I chickens given two prior infectionswith only the W strain remained partially susceptible to challengewith the H strain.

MHC haplotype does not influence the degree of cross-protection. Since the cross-protection phenotype is a consequence of adaptiveimmunity and since T cells have been demonstrated to be themajor mediator of immunity to E. maxima, it was reasonable totest for an association with the MHC haplotype. The influenceof the MHC (B locus) on the expression of cross-strain immunitywas examined in three ways.

First, the extent of cross-protection between the two strainsof E. maxima in both combinations of immunizing and challengeinfections was examined in C.B4 and C.B12 sublines of line Cchickens (Fig. 4, experiment 4), which differ at the B locusbut which have similar genetic backgrounds. Each subline wasjudged to be equally susceptible to primary infection with eitherstrain of E. maxima; both were completely immune to homologouschallenge (Fig. 4), and in neither was there evidence of cross-protectionbetween the H and W strains (Fig. 4; Table 1).

Second, a comparison of the levels of cross-protection achievedby lines of chickens that have the same B locus but that areknown to differ at many non-MHC loci was made. Lines 6₁ and7₂ (both B2) were equally susceptible to primary infection witheither strain of E. maxima and were completely protected againstchallenge with the homologous strain (Fig. 5, experiment 5).The H strain conferred partial cross-protection against challengewith the W strain in both lines (65 and 58% for lines 6₁ and7₂, respectively). However, in the reciprocal experiment (inwhich the W strain preceded challenge with the H strain) theline 7₂ chickens were substantially protected (81%) but line6₁ chickens were not significantly protected (Fig. 5; Table1).

Third, an examination of the level of cross-protection obtainedin line (15I x C.B12) x 15I backcrossed chickens given a primaryinfection with the H strain and challenged with the W strainwas made (Fig. 7) and then a test for an association betweenoocyst production and the expressed MHC haplotype was performed.

View larger version (20K):
[in this window]
[in a new window]
FIG. 7. Outputs of oocysts of individual line 15I (15I x C.B12) backcross chickens arising from primary infections with the H strain and challenge with the W heterologous strain. Circles, individual cumulative oocyst counts; numbers identify individual birds. The birds were divided into groups based on the expression of the MHC (B-locus) haplotype.

The outputs of oocysts from both the B15/B15 homozygote andthe B12/B15 heterozygote chickens showed normal and highly overlappingdistributions, and levels of cross-protection within each groupwere similar. The majority of challenged birds excreted relativelyfew oocysts (high levels of cross-protection) although two birdsrepresenting each of the MHC genotypes (4 of 23) excreted largenumbers. Since the birds were tracked individually through bothinfections, the experiment also demonstrated that there wasno association between the numbers of oocysts produced duringa primary infection and the subsequent extent of cross-protectionafter challenge with the heterologous strain.

Taken together, the results of the three experimental approachessuggest that the level of cross-protection was not associatedwith the expression of any one of four different MHC or B locus(B2, B4, B12, or B15) alleles.

DISCUSSION

Top

Discussion

Although there are many examples in the literature of diversitywithin antigen-encoding loci of pathogens, it appears that onlya few studies have examined host-pathogen interactions in relationto host genetics, protective immunity, and antigenic diversity.Using two antigenically distinct strains of an apicomplexanparasite (E. maxima) and both outbred and inbred lines of chickens,we demonstrated the dramatic influence of host genetics on theoutcomes of challenge infections with heterologous strains ofpathogen. The Eimeria-chicken host combination proved to bean excellent model system for this study for reasons that include(i) the highly regulated and self-limiting parasite life cycle,(ii) the high immunogenicity of the parasite, whereby completeprotection against homologous challenge is induced, (iii) pooror no protection against heterologous challenge, (iv) a convenientand accurate readout of numbers of excreted oocysts, and (v)the availability of lines of chickens with different geneticbackgrounds. As far as we are aware, no natural disease modelor host-pathogen relationship that offers such a combinationof phenotypes that are so readily measurable experimentallyhas been reported.

It is clear from our data that line 15I chickens were more susceptibleto primary infections with either strain of E. maxima than anyother line of chicken tested, and this is in concordance withpreviously published data (18). All lines of chickens were completelyimmune to challenge with homologous parasites, which confirms,first, that both strains of E. maxima were equally capable ofinducing protective immunity and, second, that none of the linesof chickens had any major defect in immune function. However,the level of protection in the face of heterologous challengevaried considerably, indicating that the capacity to cross-protectis influenced greatly by the genetics of the host. Indeed, withline C.B4, C.B12, and 7₂ chickens the dichotomy between protectionagainst homologous or heterologous challenge was absolute, i.e.,complete protection against the immunizing strain of E. maximabut no protection against challenge with the heterologous strain.Other inbred lines of chickens were associated with levels ofcross-protection that were intermediate; e.g., line N chickenswere consistently protected to a level of 70 to 90% againstheterologous challenge.

Intriguingly, with line 15I chickens the degree of cross-strainimmunity was directional and dependent on infection history.Greater than 95% cross-protection was obtained when the H strainof E. maxima preceded the W strain, but no cross-strain protectionwas observed when the immunizing and challenge strains werereversed (Fig. 2 and 3). Although immunization with the W strainoffered no cross-protection to challenge with the H strain,it did not interfere with the capacity of these chickens todevelop immunity to tertiary challenge. This result indicatesthat previous exposure of the chickens to the W strain did notprevent the host from responding appropriately to H strain protectiveantigens.

The incomplete cross-protection against strains of E. maximawas recently examined with respect to the migration of sporozoitesreleased in the gut from oocysts representative of two otherparasite strains that induce incomplete cross-protection toeach other (6). Following homologous challenge, an accumulationof sporozoites occurred within the lamina propria of the intestineand only a few were found in the crypts (for asexual development)by 72 h after challenge. In contrast, in birds challenged withthe heterologous strain at least four times as many sporozoitesmigrated to the crypts. Interestingly, the degree of cross protectionas measured by the numbers of sporozoites transported to thecrypts was not reciprocal. Similarly, there may be some slightdifferences in the capacities of these strains to protect againsteach other in terms of pathology (47), but it is unclear whetherthese differences are significant. These results suggest thatthe sporozoite may be an important target, but it is not knownwhether the cross-protective mechanisms also affect later stagesof the life cycle. Further work to determine the target stagesin different host genetic backgrounds and immune cell typesthat mediate different types of cross-protection is indicated.

The influence of host genetics on the resolution and/or controlof primary infections is well established, and there are manyexamples with viral (14, 50), bacterial (16, 58), and parasiticpathogens (51, 55), including murine (33, 46) and avian Eimeriaspp. (17, 18, 39). Although the loci involved have not beendefined, there is strong evidence that the major genetic determinantsof resistance lie outside the MHC (41). The biological natureof host resistance to Eimeria spp. is best understood for themurine parasite Eimeria vermiformis (considered a good modelfor E. maxima infections in the chicken); resistance is mediatedby gamma interferon (IFN- ${gamma}$ )-producing T-cell receptor ${alpha}$ ß⁺(TCR ${alpha}$ ß⁺) CD4⁺ T cells (66-68, 78). T cells are essentialfor immunity against E. maxima, and IFN- ${gamma}$ is strongly stimulatedby infection with eimerian parasites (67, 87). B lymphocytesplay a minor role in limiting primary infection with E. vermiformisand E. maxima, but whether their contribution is mediated byimmunoglobulin production or is due to their role as antigen-presentingcells for CD4⁺ T cells is unknown (77, 78, 84).

The T-cell compartment is essential for the expression of immunityto secondary infection with Eimeria spp. (64, 70, 71, 78, 79).The contribution of host genetics in the expression of a recall(memory) response is much less well defined. For the studiesdescribed herein, the level of protection to challenge withheterologous parasites is a measure of the degree of cross-protectiveimmunological memory.

The MHC (B locus) dependence for generating T-cell responsesagainst Eimeria spp. in the chicken in the context of immunizationwith recombinant antigens has been reported (19, 39, 40). TheMHC linkage of vaccination-induced responses with a single recombinantantigen may be ascribed to a lack of peptides that bind to everyMHC haplotype. Nonetheless, in these studies we immunized chickensby infection with a live, antigenically complex, pathogen thatpresents large numbers of MHC-binding peptides to the immunesystem. The lack of cross-protection indicates that immunologicalprotection may be based on relatively few of the antigens expressedby Eimeria spp. Indeed, T cells and antibodies induced by onespecies of Eimeria exhibit substantial cross-reactivity withantigens prepared from heterologous species (J. Bumstead, personalcommunication; P. Hesketh, unpublished observations) that donot cross-protect during natural infections. These observationsand the dramatic differences in cross-strain protection reportedhere reinforce the view that only a few of the estimated 5,000to 10,000 genes in the Eimeria genome (73) encode antigens thatare protective.

Antigenic diversity also characterizes infections with the bovineparasite Theileria parva, with a close association between immunodominance,the strain specificity of a cytotoxic T lymphocyte response,and protection (82, 83). Similarly, a series of CD8 T-cell clonesfrom an individual infected with the malarial parasite differentiallyrecognized naturally occurring immunological variants of Plasmodiumfalciparum (11). The directional nature of cross-protectionseen in our studies with the two strains of E. maxima in line15I chickens has also been reported for T. parva infectionsin cattle (49, 82) and Salmonella enterica infections in mice(29).

The MHC of chickens is much less complex than that of mammalssuch that each haplotype expresses single dominant class I andclass II molecules (31, 35). The restricted repertoire of MHCclass I-binding peptides in the avian immune response dramaticallyinfluences the capacity of chickens to develop protective immunityto a small RNA virus (Rous sarcoma virus) (34). Interestingly,for an antigenically complex DNA virus (Marek's disease virus),the efficacies of different vaccines have been linked to theexpression of certain MHC haplotypes (4, 5).

Since all TCR ${alpha}$ ß T-cell responses require presentationof a peptide in the context of MHC, antigenic diversity betweenthe strains of E. maxima might be expressed by a change in thecapacity of some antigens to bind to a particular MHC. For thisreason, we assessed the cross-protective phenotype in variouslines of chickens that either share most non-MHC genes but differat the MHC locus (C.B4 and C.B12 sublines of line C) or haveidentical MHCs but differ at many non-MHC-linked loci (lines6₁ and 7₂; both B2). The lack of cross-protection between theH and W strains of E. maxima in C.B4 and C.B12 chickens, differencesseen with line 6₁ and 7₂ chickens, and the results of a limitedstudy with a backcross population (line 15I x [15I x C.B12])support the hypothesis that the MHC is not a major genetic componentof the cross-protective phenotype (Fig. 4, 5, and 7). Some ofthe most susceptible birds in the backcross study were B15 homozygous(Fig. 7); if the MHC haplotype were a major genetic componentof the phenotype, these birds would have been highly protectedagainst heterologous challenge. While it is difficult to proposecandidate non-MHC genes that might influence the developmentof cross-protective immunity, the response is dependent on TCR ${alpha}$ ßT cells. Therefore, differences in the TCR ${alpha}$ ß T-cellrepertoire could be responsible for the differences in the abilitiesof different inbred lines of chickens to mediate cross-protection.The repertoire could be restricted by differences in the genomiccomplement of TCR loci, as has been documented in mice (7, 27,28). Alternatively, the differences could reflect negative selectionof T cells in the thymus by, for example, expressed endogenousretrovirus elements (22, 76) that are known to differ amongchicken lines (38). Consequently, detailed mapping of the hostgenetic loci that influence cross-protection is strongly indicated.

In the field, host populations are generally outbred geneticallyand the induced immune pressures on apicomplexan parasites arelikely to be diverse. However, in areas of high parasite endemicitythe immune selection pressures on parasites will be more intenseand substantial polymorphisms have been found in genes thatencode, for example, the circumsporozoite antigen (42) and themerozoite surface protein in P. falciparum (21).

The generation of strong cross-species, nonprotective responsesindicates a need to examine the relationship between immunodominanceand immunoprotection in relation to the avian responses to E.maxima. Moreover, the influences of host genetics and infectionhistory on the outcome of challenge infection between immunologicalvariants of E. maxima indicate the complexity of the protectiveimmune response. Although further work to determine the functionaland genetic basis of phenotypes reported here is necessary,the E. maxima-avian host system has enabled the complexity ofthe host-pathogen relationship in relation to immune protectionand antigenic diversity to be identified. Understanding theseaspects of the immunological relationship would provide a theoreticaland practical basis for rational development of anticoccidialvaccine strategies. Antigenic diversity is a feature of manypathogens, and similar analysis would support the rational developmentof vaccines against other antigenically complex pathogens ina wide range of host species.

FOOTNOTES

* Corresponding author. Mailing address: Institute for Animal Health, Compton Laboratory, Compton, Newbury, Berks RG20 7NN, United Kingdom. Phone: 44 1635 577293. Fax: 44 1635 577263. E-mail: adrian.smith{at}bbsrc.ac.uk.

Editor: J. M. Mansfield

REFERENCES

Top