<i>Plasmodium falciparum</i> infection can be distinguished from that of other plasmodia which infect humans by one specific facet of its biology, that of sequestration of mature developmental forms in capillary beds. A variant antigen, <i>Plasmodium falciparum</i> erythrocyte membrane protein 1 (PfEMP-1) is involved in this process and members of this trypsin soluble 210-230 kDa protein family are responsible for cytoadherence of mature infected erythrocytes. A model system has been developed to investigate the molecular basis of infected erythrocyte adhesion to host endothelial receptors. <i>P. falciparum</i> clone 3D7A was selected for adhesion to Chinese Hamster Ovary (CHO) cells resulting in an adhesive phenotype, which is sensitive to trypsin treatment of infected erythrocytes. The nature of the cytoadhesion receptor on the CHO cell surface has been analysed and does not appear to correspond to any of the protein or glycoprotein cytoadherence receptors so far described. It has been demonstrated that unselected 3D7A transcribes at least 27 different <i>var</i> genes. On selection for adhesion to CHO cells this transcription is narrowed down to the transcription of a single gene named <i>var</i>-CHO. This gene has been cloned, sequenced and annotated. <i>Var-</i>CHO consists of 3 DBL domains, DBL-x, DBL-b and DBL- s, two CIDR domains, CIDR-x and CIDR-b, and a C2 domain. A provisional genomic map position of this member of the PfEMP-1 has been identified as chromosome 13 by a combination of pulse field gel and <i>P. falciparum</i> genome project data.

On the 1st January 1970 there commenced in the Solomon Islands a special project for the eradication of malaria. This full Malaria Eradication Programme followed on from a Pilot Project which started in 1962 and a Pre-Eradication Programme which started in 1965. These widely scattered islands have been notoriously malarious at least since the days of the early explorers and probably long before that. In more recent times infants and children have certainly died in their hundreds and most of the people have been weakened with persistent infections. In the late 1950 IS, following on from the suceessful eradication of yaws, the Government and the World Health Organisation turned their attention to the possibilities of eradicating malaria. These early efforts have already been descri bed by Dr. J. D. Macgregor in his M. D. thesis entitled "Malaria in the Island Territories of the South West Pacific". This present thesis expands that work, concentrating particularly on the years 1970-75 which were the first six years of the full Malaria Eradication Programme.

I investigated RBC selectivity as a potential virulence factor further in a rodent malaria model using <i>Plasmodium chabaudi</i>. The experimental design allowed me to look at the selectivity index in congenic lines, which significantly differed in their virulence to the host. Supporting my <i>P. falciparum</i> data, no association between the selectivity index and virulence was seen. This is the first description of the selectivity index in <i>P. chabaudi</i> and RBC invasion was found to be very unselective as expected for a malaria parasite which does not exhibit a reticulocytes preference. The Kenyan field isolates were also typed for their RBC invasion pathway profile by measuring their invasion into enzyme-treated RBC. Although invasion profiles of field isolates have been reported previously, as far as I am aware, this is the first report looking at invasion profiles of field isolates in relation to disease severity. In line with previous results, field isolates invade RBC by multiple invasion pathways. No difference between the invasion profiles of isolates from uncomplicated or severe malaria patients was seen. Overall my thesis presents a detailed study of potential <i>P. falciparum</i> virulence factors related to invasion, working both with clinical isolates and laboratory strains. Severe malaria in Africa was not found to be associated with parasite multiplication rates, RBC selectivity or specific invasion pathways. Other virulence factors such as resetting and platelet-mediated clumping may be of major importance in this region. This work suggests the possibility of differences in malaria virulence factors between sub-Saharan Africa and Asia, which may need to be taken into consideration for drug and vaccine design.

1. A chloroquine-resistant parasite line of Plasmodium chabaudi was obtained by submitting a sensitive line to a continuous and gradually increasing level of drug pressure. The line 411AS survived treatment of 3 mg/kg given for 6 days which eliminates the original drug sensitive line. 2. Cyclical transmission of 411AS as well as stability studies carried out showed that resistance was stable and transmissable through mosquitoes in the absence of drug pressure. 3. The genetic basis of chloroquine resistance was studied by making a cross between 411AS and a sensitive line of different origin, 96AJ, which differed additionally in 3 characters. Seventy clones from this cross were classified. Various recombinant classes were obtained and the results showed that chloroquine resistance in P. chabaudi is a stable character which undergoes genetic recombination with other markers. 4. Another line, 524AJ, resistant to 3 mg of chloroquine/kg administered for 6 days, was obtained by similar low drug pressure method. 5. Cyclical transmission of 524AJ and from clones established from this population showed that this line exhibited two types of resistance: (a) a stable, heritable form detectable after mosquito transmission, and (b) an unstable form which could not be detected after mosquito transmission. 6. Competition studies between chloroquine-resistant and chloroquine-sensitive parasites were made by mixing different proportions of blood forms in mice and by establishing and testing clones at different stages of the infection. Sporozoites from mosquitoes which fed on each mixture were also used to establish new infections in mice and clones were established and tested in a similar way. All the experiments showed an apparent selective advantage of the resistant over the sensitive forms in the absence of chloroquine pressure. 7. A discussion of the selection experiments, crossing experiments and competition studies is included and further research suggested by this work is considered.

A laboratory study with the rodent malaria parasite <i>P. chabaudi </i>and <i>A. stephensi</i> vector indicated that mosquito morality varied with parasite genotype, infection diversity and nutrient availability. In standard conditions, mixed clone infections were the most lethal, but when glucose water was limited, mortality was highest in mosquitoes infected with CR. A second experiment showed that under standard conditions, mixed infections also had the greatest impact on vector fecundity. The virulence of mixed infections could not be explained by parasite load, nor their rate of resource uptake by parasites within the mosquitoes. During the parasite development period, infected mosquitoes had the same amount of three key physiological resources (lipids, glycogen, proteins), as those that were uninfected. Furthermore, mosquitoes infected with the most virulent parasite genotypes had an increased abundance of glucose relative to the controls. This is consistent with <i>Plasmodium </i>manipulating mosquito sugar-feeding behaviour in order to increase its own transmission. Several laboratory studies of malaria parasites and some field observations suggest that <i>Plasmodium</i> virulence in vertebrates is positively correlated with transmission to mosquitoes. A final experiment was undertaken to test whether the transmission advantage of infections that are virulent to vertebrates could be offset by an increased probability of causing death to their vectors. Mice were infected with one of seven distinct genetic clones of <i>P. chabaudi</i> that are known to vary in virulence. Infection virulence in mice (weight loss and anaemia) was positively correlated with mosquito infection rate but not with mosquito survival. Vector survival was influenced only by parasite clone and oocyst burden (negative association). These results suggest that vector fitness should not place an upper limit on malaria virulence. Overall, this research demonstrates that <i>Plasmodium </i>can be virulent to its vector, and that the magnitude of virulence is dependent on parasite genotype, infection diversity and environmental conditions. Although <i>P. chabaudi </i>virulence in vectors was not correlated with virulence in vertebrates, parasite genetic differences do impact vector fitness. Thus differential vector mortality could play an important role in determining the genetic composition of <i>Plasmodium </i>populations.

The aim of this research was to identify the erythrocyte surface antigens of <i>P. falciparum</i> infected erythrocytes (IEs) that are expressed following selection of the parasite for adhesion to the placental receptor chondroitin sulphate A (CSA). These surface antigens are considered potential vaccine targets against pregnancy-associated malaria (PAM). Typically <i>P. falciparum</i> surface antigens are highly sensitive to treatment of the IE with the protease trypsin, however variant surface antigen (VSA) mediated adhesion to CSA has often been found to be relatively trypsin resistant. In this study the protease sensitivity of the VSA of CSA binding IEs was determined with regard to serum IgG recognition and placental receptor binding. Discordance between the protease sensitivity for these phenomena was identified; this may have implications for PAM vaccine design. The completion of several <i>Plasmodium</i> genome sequences and advances in mass spectrometry has opened up the field of proteomics to the study of the malaria parasite. Proteomic methods have been applied to the study of <i>Plasmodium</i> VSAs. However, the initial step of these methods, surface antigen labelling, proved problematic. Therefore, this thesis concentrates on the optimisation of these techniques, during which an antibody to the <i>Plasmodium falciparum</i> erythrocyte protein-1 (PfEMP1) family was raised and characterised. This reagent in combination with the advances made for IE surface labelling will be valuable tools for future studies of VSA expression.

This thesis focuses on the determinants of virulence in single and mixed-clone malaria infections, and the consequent impact of these infections for host- and parasite-fitness. Controlled experiments were conducted using a rodent model of malarial disease, <i>Plasmodium chabaudi. </i>Mice were infected with precise numbers of virulent and avirulent parasites. In the mixtures, known ratios of two clones that differed in virulence were used. Virulence was quantified in terms of host morbidity and mortality. Experiments investigating how virulence is determined in mixtures revealed that both the proportion of virulent clone in the innocula, and the genetic diversity of the infection determine virulence. Replacing virulent parasites with avirulent ones in a mixture was shown to confer protection for the host. These results challenge the various assumptions made in the models of the evolution of virulence about how virulence is determined in mixtures. They also suggest that selection against virulence can be reduced if virulent clones coinfect with avirulent ones, because host mortality is reduced in mixtures when the avirulent clone dominates. In single infections, the inoculating dose of virulent and avirulent parasites affected the virulence of the infection. Larger doses caused greater anaemia. They also caused additional weight loss, and death, but only for the virulent clone. Clone differences in virulence were maintained over the range of doses. Dose effects were manifested through the timing and/or magnitude of peak parasite densities, broadly supporting the idea that disease severity is due to the time the host has to control parasite densities and ameliorate the effects of parasites. To investigate the correlates of mortality, multivariate analyses were conducted. These generally showed that both the initial weight and red blood cell density of mice, and the rate at which they lost red blood cells and weight affected their probability of survival.

As part of a longitudinal study into the humoral immune response to malaria in infants, 52 out of 156 children were found to be chronically infected with <i>P. falciparum</i>. In 15 of these children, the allelic diversity of these infections was analysed by PCR typing of <i>MSP 1</i> and <i>MSP 2</i> alleles followed by sequencing the polymorphic region of <i>MSP 2</i>. The study shows that infants become infected with a number of different allelic types of <i>P. falciparum</i>. These infections were of very low density and were asymptomatic. The number of different genotypes varied over time in each child. Persistent infection in some children was clearly due to multiple new infection, whereas in other children single parasite genotypes persisted for very long periods of time, up to seven months in some children. In order to evaluate serological responses to asexual stage malaria antigens, serum obtained at monthly intervals over the first year of life were tested, by ELISA, against recombinant proteins representing various merozoite antigens, to determine the specific response to MSP 2, novel polymorphic sequences of <i>MSP 2</i> were expressed as GST fusion proteins. GST fusion proteins representing the conserved and dimorphic sequences of <i>MSP 2</i> were also constructed. Responses to specific antigens varied between individuals, different children produced antibodies to different agencies at different time points depending on the presence or absence of infection. There were occasions when a decline in a specific antibody titre was associated with infection or reinfection with parasites carrying the same specific polymorphic antigen, which may be indicative of "strain-specific immunity" operating at the level of MSP 2. Interestingly, antibody responses to several different <i>MSP 2</i> sequences were found to be cross-reactive indicating that the number of serologically identifiable MSP 2 variants may be smaller than the number of sequence variants, intriguingly, we also found that pairs of immunologically cross-reactive variants of msp2 were frequently found together in different children, this data provides evidence in support of the Gupta-Day theory of strain-structuring of malaria parasite populations and indicates that antibody responses to MSP 2 may be able to select for elimination or control of malaria parasites, if so, this would be strong support evidence for the effector role of anti-MSP 2 antibodies and would suggest that MSP 2 might be an important component of a malaria vaccine.

Monoclonal antibodies to the gamete surface antigens of <I>Plasmodium falciparum, </I>Pfs230 and Pfs48/45, can abolish the infectivity of gametes to mosquitoes, hence these antigens have been proposed as candidates for inclusion in a malaria transmission-blocking vaccine. One possible mechanism of antibody-mediated transmission-blocking is complement-mediated gamete lysis. A cohort of human sera, from geographically distinct malaria-endemic regions, was used to investigate whether this may be a mechanism of naturally acquired transmission-blocking immunity to <I>P. falciparum. </I>By immunoprecipitation, it was shown that antibody recognition of Pfs230 and pfs48/45 is limited, despite universal exposure to <I>P. falciparum</I> gametocytes. <I>In vitro</I> complement-mediated lysis of <I>P. falciparum</I> gametes was positively associated with the presence of antibodies to Pfs230, but not with antibodies to the N-terminal region of the precursor molecule (Pfs260) which is shed from the gametocyte surface at the time of gametogenesis. Similarly, antibodies to two other gametocyte-specific proteins, Pfs48/45 and Pfs27/25 were not associated with gamete lysis. All sera which mediated gamete lysis contained IgG1 and/or IgG3 antibodies to gamete surface proteins as determined by ELISA. These data suggest that Pfs230 is a major target of complement-fixing antibodies which may be important for antibody-mediated transmission-blocking immunity. A selection of these malaria-immune human sera were tested for their ability to affect the infectivity of <I>P. falciparum</I> gametocytes to <I>Anopheles</I> mosquitoes. It was found that transmission-reducing effects of the sera were associated with the presence of IgGl antibodies to the gamete surface, specifically against the protein, Pfs230.

Dimorphic and polymorphic regions of <I>MSP-1 </I>and <I>MSP-2 </I>were genotyped by polymerase chain reaction (PCR) in 379 <I>P. falciparum </I>clinical isolates from Malawi. Polymorphisms in the genes were reflected in antigenic diversity of the translated proteins detected by indirect immunofluorescence (IFA) typing. Most <I>MSP-1 </I>alleles were MAD20 dimorphic and K1 block 2 types, whereas <I>MSP-2</I> type A alleles predominated. The effect of <I>P. falciparum</I> genetic polymorphisms on the specificity of immune responses was investigated in children immunised by natural infections typed for <I>MSP-1 </I>and <I>MSP-2.</I> Specific IgG antibodies detected by ELISA were mainly directed to MSP-1 conserved C-terminus fragments (19kDa and 42kDa), whereas antibodies induced by MSP-2 predominantly recognised group-specific regions. Overall, naturally induced human antibody responses to MSP-1 and MSP-2 were short-lived, type-specific, and correlated with PCR typing of the infecting parasites present. The wide spectrum of disease manifestations observed in <I>P. falciparum </I>infections probably reflects a combination of various host and parasite factors. Children with severe malarial anaemia (SMA, n=50) were distinguished by a higher multiplicity of infections than children with uncomplicated malaria (UM, n=92) or cerebral malaria (CM, n=93). SMA patients had higher prevalence of MSP-1 K1/Well and MSP-2 B dimorphic types, whereas block 2 MAD20-type of MSP-1 was more common in UM cases. A differential pattern of antibody responses to defined regions of the proteins was found. Children who developed SMA contained very low levels of antibodies to conserved regions, whereas children with CM had significantly higher levels of antibodies to the conserved regions than children with UM or SMA. CM and SMA differed substantially for all parameters assessed, indicating that they should not be regarded as one severe disease group but as two very distinct syndromes.