Recent advances in malaria research: new tools to understand an old enemy

Abstract

Scientists who gathered at the most recent Keystone Symposia meeting on malaria described several advances, including applications of new technologies as well as new insights into pathogenesis and immunology. Enhanced tools to visualize the immune response to the Plasmodium pathogen, and systems biology approaches to interrogate host–parasite interactions, are altering our understanding of immunity and disease. At the same time, a partially effective subunit vaccine in human trials, and new models of highly protective human immunity, are raising expectations for immunological control of this ancient pandemic.

Malaria is still a devastating disease, with approximately 3 billion people at risk and 300–500 million clinical cases annually, mostly in developing countries. Therefore, there is tremendous urgency to understand the immune mechanisms involved in this disease and its pathogenesis, as well as to develop effective vaccines that would prevent over a million deaths every year, most of whom are children under 5 years of age [1]. This Keystone Symposia meeting focused on malaria infection and especially on the major advances that have been made towards the understanding of host–parasite immunological interactions using the new technologies available. Cutting-edge research analyzing protection followed by vaccination with the leading subunit candidate (RTS,S) or with live sporozoites showed remarkable results. A wider use of new tools that enable the prediction of protection was a recurrent point of discussion.

In the keynote lecture, Michael Good (Queensland Institute of Medical Research, Australia) presented an overview of the challenges in studying malaria immunology and designing a vaccine. Recent work from his group has shown that ultra-low doses of parasites can in fact lead to protection in both mice and humans. The group are currently testing the efficacy of dead parasites in conjunction with CpG in mice: data so far show that this is protective, and that protection is dependent on CD4 cells, IFN-γ, nitric oxide and IL-2. The challenges that need to be overcome before considering this for human trials will be a source of red blood cells (RBCs) and parasite, and the methods of efficacy testing in humans.

The first interactions between the malaria parasite and host can determine the subsequent immune response: acquisition of immunity and/or immunopathology. Paolo Arese (University of Torino Medical School, Italy) showed that fibrinogen binding to native hemozoin leads to induction of oxidative burst in macrophages, inducing activation of p47 and rapid secretion of preformed TNF-α, which contribute to the inflammatory response and may play a role in the periodic fever seen in Plasmodium falciparum malaria. Using in vivo imaging of mouse malaria, James Brewer (University of Glasgow, UK) confirmed that during infection antigen presentation is not inhibited, but T-cell division and contact with B cells are reduced, leading to less expansion of the follicles. Keeping to the topic of early cellular dynamics during malaria, Jean Langhorne (National Institute for Medical Research, UK) showed that inflammatory monocytes (CD11b^hiLy6C⁺) increase during infection and they seem to be generated from atypical progenitors (IL-7R⁺ c-kit^hi). The emergence of these progenitors is independent of MyD88, TRIF and Toll-like receptor (TLR)9; however, the expression of IFN-γ receptor on their surface is essential for their appearance. Kate Fitzgerald (University of Massachusetts Medical School, MA, USA) focused her talk on the induction of innate responses by malaria. The P. falciparum genome contains more than 6000 ATTTTTAC motifs. Using representative sequences to stimulate different cell types, they studied the NFκB, IL-6 and TNF-α response. The response was dependent on the formation of a stem-loop structure within the oligo, and was independent of TLR signaling. Knockout studies found that AT-rich oligos induce interferon independently of RNA pol III, and that STING/TBK1 and IRF3/7 are critical components of AT-rich DNA sensing. Type I interferon-inducible gene signatures have also been detected in patients infected with P. falciparum in Brazil, and are linked to experimental cerebral malaria in a Plasmodium berghei ANKA model. Eleanor Riley (London School of Hygiene and Tropical Medicine, UK) showed that natural killer (NK) cells produced IFN-γ and perforin, and upregulated lysosomal-associated membrane protein 1 in response to P. falciparum-infected RBCs. The response could be blocked by depleting CD3⁺ T cells or blocking IL-2, suggesting that NK cells can mount a recall response in the presence of IL-2-producing memory CD4⁺ T cells.

Pre-erythrocytic-stage immunity was discussed in several sessions. In an experimental human malaria infection model, Robert Sauerwein (Radboud University, The Netherlands) presented new data on volunteers exposed to three sporozoite infections whilst receiving chloroquine. This led to 100% protection 2 months after the last ‘vaccination’. Six donors were rechallenged 874 days post-immunization, and four were still completely protected. T-cell memory responses were high and antibody responses low, but none of these correlated with protection. Nick Crispe (Seattle Biomedical Research Institute, WA, USA) showed that chronic presence of antigen in the liver deactivates CD8⁺ T cells, as they produce less IFN-γ after 8 weeks. Moreover, ovalbumin delivered by vectors directly into the liver activated CD8⁺ T cells independently of CD4⁺ T cells. This was confirmed when MHC class II knockout mice presented similar CD8⁺ T-cell response as wild-type animals. Ute Frevert (New York University School of Medicine, NY, USA), on the other hand, showed that CD4⁺ T cells may be involved in killing of infected hepatocytes, as they recognize human peptide-primed target cells, reoriented their cytotoxic granules towards the zone of contact, formed stable immunological synapses and killed target cells. Laurent Renia (Singapore Immunology Network, Singapore) explored cellular responses to liver-stage parasites. Using Plasmodium yoelii, he showed that liver CD8⁺ T cells expressing CCR2 after infection led to lower levels of liver parasites. TNF-α also played a role in inhibiting liver-stage development independent of nitric oxide or reactive oxygen intermediates, and mediated by a soluble factor, possibly acute-phase protein. Results varied depending on the hepatocyte cell culture and parasite strain used. Fidel Zavala (Johns Hopkins Bloomberg School of Public Health, MD, USA) showed that the circumsporozoite surface protein (CSP) antigen is presented for at least 2 months after immunization with irradiated sporozoites and prolonged antigen presentation seems to be critical for optimal development of memory CD8⁺ T-cell responses. John Harty (University of Iowa, IA, USA) highlighted the high level of CD8⁺ T-cell responses required for sterilizing immunity to malaria using irradiated sporozoites; CD11a^hiCD8^int CSP-specific CD8 memory cells represented 5–15% of the total response. Using two mouse–parasite combinations, he showed that boosting could greatly increase the CD11a^hiCD8^int cells, and thus increase protection; however, the same numbers of CD11a^hiCD8^int did not correlate with protection in all models. Nathan Schmidt (University of Iowa) showed that the threshold of CSP-specific T cells required for protection varied in different mouse models and parasite strains. Urszula Krzych (Walter Reed Army Institute of Research, MD, USA) demonstrated that, despite irradiated sporozoites undergoing abortive development, protection mediated by liver CD8⁺ T cells is induced. P. berghei-irradiated sporozoites induce an accumulation of liver CD11c⁺NK1.1^- dendritic cells (DCs) of which, CD8a⁺ DCs are the major antigen-presenting cells for CD8⁺ T cells. The survival of memory CD8⁺ T cells is dependent on IL-15 produced by Kupffer cells and this cytokine maintains proliferation of CD8⁺ T-central memory. IL-15 knockout mice develop only short-term immunity and are not protected against re-challenge after 2 months. Sandra March (Massachusetts Institute of Technology, MA, USA) and Sangeeta Bhatia (Massachusetts Institute of Technology) presented data on the establishment of liver-stage cultures of P. falciparum and Plasmodium vivax in a microscale human liver platform. Similarly, Anne-Marie Zeeman (Biomedical Primate Research Centre, The Netherlands) used primary rhesus or human hepatocytes, as well as several hepatoma cell lines, to support the apparent development of the hypnozoite stage of the simian parasite Plasmodium cynomolgi, and used these to screen drugs against this dormant liver-stage parasite. These methods might be used for vaccine validation, drug screening and study of the biology of the parasite during this stage. In order to identify new proteins involved in the invasion and development of the parasite during the liver stage, Audrey Gego (INSERM UMR S 945, France) and colleagues used the siRNA technique, targeting more than 6000 hepatocyte genes. Five genes were studied in detail: two of them were involved in invasion and the other three genes were essential for intra-hepatic development. The role of each gene remains to be determined.

Immunity to erythrocytic stages and transmission-blocking immunity was also discussed by several speakers. Pierre Buffet emphasized that splenomegaly is a common feature during infection and that malaria severity is increased after splenectomy, highlighting the importance of the spleen in immunity. He showed that mature infected cells and, unexpectedly, 50% of the rings, are rapidly retained in human spleens. This process could contribute to disease pathogenesis, causing anemia and splenomegaly, but might also represent an important role for the spleen in reducing the number of circulating parasites. Using a protein microarray chip, Peter Crompton (NIH, MD, USA) analyzed the acquisition of antibodies to malaria in different age groups in Mali, showing an increase in the breadth of the response with age, although most responses were short-lived. Despite intense exposure, many adults did not show significant memory B-cell responses: the memory cell population increased only very slowly, and it seems that repeated exposure is required for a slow stepwise increase in memory B cells and antibodies. In order to decipher invasion pathways and identify new targets for blockage, Julie Healer (Walter and Eliza Hall Institute of Medical Research, Australia) looked at combinations of parasite knockouts and antibody blocking of genes from the erythocyte-binding antigen and P. falciparum reticulocyte-binding homologue families alone or in combination. This work highlighted that several invasion pathways were possible, and concomitant blockade of many might therefore be required to abrogate invasion.

Regarding the development of transmission-blocking vaccines, Takafumi Tsuboi (Ehime University, Japan) emphasized that people vaccinated with post-fertilization proteins (P25/P28) were able to produce specific antibodies against proteins on the ookinete, and showed that these antibodies could significantly reduce transmission of both laboratory strains and field strains from Thailand in a mosquito feeding assay. In addition, in the presence of complement, transmission blocking was significantly increased for antibodies directed against Pfs230. Joanna Kubler Kielb (NIH) presented an interesting method of preparing conjugate vaccine for both the mosquito and the liver stages of the parasite. Cross-linking of Pfs25, a mosquito-stage surface antigen, renders the protein immunogenic and induces long-lasting antibodies. Vaccination of mice with these conjugates of Pfs25 led to 100% transmission blocking immunity. When the asparagine–alanine–asparagine–proline (NANP) repeat region of CSP was covalently bound to this complex, high levels of long-lasting antibodies to both the complex and NANP were induced. This strategy will soon be tested in humans. Philip Bejon (University of Oxford, UK) reported on the latest immunological data from RTS,S, the only malaria vaccine to date to reach Phase III trials. Further analysis revealed that both antibody titer and cellular response were important for sterile protection. Modeling the CSP antibody response suggests that dichotomizing responses into two groups is more predictive of protection than simply using the continuous values. CD4⁺ T cells produced cytokines in response to CSP; however, only the percentage of CD4⁺ cells producing TNF-α correlated with protection. Richard Pleass (University of Nottingham, UK) made the point that development of new adjuvants that are more efficient in stimulating the immune response and less toxic for humans is a necessity. He discussed that antibodies themselves behave like adjuvant molecules, activating the immune system and presented some preliminary findings on the ability of fusion proteins of immunoglobulin Fc with merozoite surface protein (MSP)-1 or apical membrane protein (AMA)-1 without adjuvant to induce high antibody titers in mice carrying a human Fc receptor.

Bali Pulendran (Emory University, GA, USA) presented a novel approach predicting antibody responses to flu vaccines, which may have application for future malaria vaccines. Comparing the inactivated influenza vaccine (flu shot) and the live-attenuated influenza vaccine (FluMist), they showed that people vaccinated with flu shot had higher antibody titers to hemagglutinin and higher B-cell numbers. Using microarrays, they were able to correlate expression signatures to the antibody response: both plasma cell expression signatures and X-box binding protein (XBP)-1 target genes correlated to the maximum hemagglutinin–inhibition response.

The host response to the parasite can also contribute to malaria pathology. Maria Mota (Instituto de Medicina Molecular, Portugal) and Andrew Armitage (Weatherall Institute of Molecular Medicine, UK) highlighted that the level of hepcidin, a major regulator of iron homeostasis, increases ten-times during blood-stage infection, and can lead to anemia by inhibiting iron uptake and decreasing iron in the blood. Its lower levels in the hepatocytes may also contribute to the dramatic inhibition of the parasite growth in the liver during a blood-stage infection. Gareth Turner (Oxford University, UK) presented work from postmortem studies of severe malaria in Vietnam. Jaundice and acute renal failure were the top two correlates of death in individuals with cerebral malaria. Sequestration, axonal injury and parasite biomass correlated with cerebral malaria and fatalities. Pulmonary edema was detected in many cases by histology; the number of host leukocytes in the lung was increased and parasite sequestration was observed. Many patients had acute renal failure, a common complication of severe malaria in Southeast Asia, but not in Africa. Patrick Duffy (National Institute of Allergy and Infectious Diseases, MD, USA) focused on placental malaria and host responses. A novel hypothetical conserved protein, PfCSAL, binds specifically to chondroitin sulfate A (CSA) and may be involved in placental sequestration. Although also present in parasites from children, PfCSAL is only seen on the surface of the RBC infected by maternal parasites, where it colocalizes with VAR2CSA, possibly forming a complex. Microarrays on the placental tissue showed that B-cell-related genes are the most highly upregulated host genes during chronic malaria. Also, an increase in VEGF expression by maternal macrophages and in its receptor by fetal cells was observed. The soluble receptor, called sVEGFR or sFlt1, is a biomarker of pre-eclampsia and has anti-inflammatory effects. sFlt1 may be produced in response to the maternal inflammatory infiltrate in the placenta and might contribute to high blood pressure. Interestingly, cord blood inflammatory responses, in particular increased TNF-α levels, correspond to increased levels throughout early life as well as to malaria outcomes. Nick Anstey (Menzies School of Health Research, Australia) underlined the growing importance of severe P. vivax malaria, especially severe anemia in children, respiratory distress in adults and more rarely cerebral malaria, as well as pregnancy malaria. Unlike P. falciparum malaria, where severity is linked to both parasite biomass and inflammation, severity during P. vivax malaria is linked to proinflammatory cytokines. Some data have suggested that P. vivax parasites may also adhere, although all parasite stages are seen in the blood. The causes of severe P. vivax malaria are not well understood and are probably multifactorial. Anemia may be due to repeated relapses, preferential invasion of reticulocytes, possible increased fragility of erythrocytes, splenic pooling and bone marrow dyserythropoiesis. Lung injury could be due to an inflammatory response secondary to congestion of alveolar capillaries with monocytes, possibly exacerbated by parasite killing following antimalarial treatment. Ana Paula Freitas do Rosario (National Institute for Medical Research) presented work on IL-10, an important immune-modulatory cytokine in malaria infection. In malaria-infected mice, IL-10 was shown to be mainly produced from CD4⁺Foxp3^-IFN-γ⁺ cells. Interestingly, these cells are present throughout the infection, but produce more IFN-γ early in infection and more IL-10 toward the end of infection, pointing to a self-regulation mechanism. Erica Pasini (Biomedical Primate Research Centre) presented her work on parasite sequestration by performing comparative proteomics on PbANKA, a sequestering strain and PbK173, a non-sequestering parasite. Using this approach, the small exported protein SMAC was implicated in sequestration of infected erythrocytes.

Parasite variability and responses were also discussed using some exciting novel technologies. Johanna Daily (Albert Einstein College of Medicine, NY, USA) focused her talk on the effect of the host response on the parasite. Prior studies found that parasites from donors in Senegal displayed three transcriptional patterns: starvation/respiration, growth/glycolysis and stress. To recapitulate the parasite stress transcriptional pattern they co-cultured P. falciparum with peripheral blood mononuclear cells/neutrophils for 24 h. These co-cultured parasites upregulated genes involved in DNA repair, ribosomal proteins and cellular repair, as well as hypothetical proteins. The pattern of heat-shock transcripts matched the in vivo results, suggesting that the in vitro conditions can replicate physiologic conditions. Linkage disequilibrium is directly correlated with the level of transmission and can be used to identify genes under selection. In the case of high transmission areas, where low linkage disequilibrium is observed, more markers are necessary to determine parasite resistance to drugs and the haplotype should not be considered. Dyann Wirth (Harvard University, MA, USA) showed that genes encoding for MSP-1, CSP, AMA-1 and MSP-2 presented relatively high diversity, which makes vaccine design difficult. Using genome-wide analysis it is possible to determine novel drug resistance loci; an increase in the copy numbers of a gene is directly associated with drug resistance.

Joseph Smith (Seattle Biomedical Research Institute) reviewed the involvement of var genes in malaria pathogenesis and binding of infected erythrocytes to host proteins. Using mini-var constructs, the N-terminal sequence of the var protein was shown to be crucial for its transportation to the membrane. Moreover, proteins with truncated head structures were exported to the surface but did not localize into knobs. Parasites containing mini-var genes were able to express two different PfEMP1 proteins on the cell surface but, interestingly, they were unable to bind to two different proteins at the same time, which can possibly be explained by mechanical/physical impairment. Peter Preiser (Nanyang Technological University, Singapore) presented work on STEVOR, one of the multigene families involved in merozoite binding and invasion of RBCs. Trypsin and neuraminidase treatment can inhibit binding. STEVOR proteins are also involved in rosetting: unlike PfEMP1 knockout (KO), STEVOR KO parasites cannot form rosettes. These distinct roles suggest a facilitated process for the parasite, whereby STEVOR on the infected RBC surface rosette with uninfected RBCs, which can be immediately invaded by merozoites bearing the same STEVOR. Naturally acquired immunity (NAI) to blood-stage malaria is mediated by antibody but the antigens necessary to drive this immune response are not known. Lars Hviid (University of Copenhagen, Denmark) mentioned that each parasite can switch expression among different variant surface antigens, which requires the host to develop a large antibody repertoire. Antibodies to PfEMP1, the major target of NAI, prevent cytoadherence and lead to protection. Using samples from patients in Tanzania and analysis for 32 different Duffy binding-like (DBL) domains from 17 PfEMP1 variants, he showed that the acquisition of a specific antibody is structured and hierarchical, with antibodies acquired first to the most commonly recognized DBL domains in the population. Benoit Gamain (Pasteur Institute, France) highlighted that, although several individual DBL domains of VAR2CSA have been synthesized, none of these bound to CSA specifically. His group has succeeded in expressing a full-length VAR2CSA protein that binds specifically to CSA with high affinity. Small-angle x-ray scattering and other biophysical techniques were used to resolve the low-resolution structure of the protein, which possesses a compact organization most likely governed by specific interdomain interactions. These data suggest that a high-affinity, CSA-specific binding site is formed by the higher-order structure of the VAR2CSA extracellular region. In order to identify transcription signatures that could be associated with severity of the disease, Marissa Vignali (Seattle Biomedical Research Institute) and colleagues analyzed P. falciparum-infected RBCs from children using the not-so-random sequencing technique, microarray, RT-PCR and bioinformatics analysis. The children were grouped according to the level of severity and the results show that VAR2CSA expression was higher in severe malaria cases, and the serum samples revealed that 50% of the children recognized the DBL5e domain of VAR2CSA, suggesting a role for VAR2CSA in syndromes other than malaria in pregnancy.

This Keystone Symposia meeting highlighted the recent advances in basic and applied research on malaria immunology and pathogenesis. While the first reproducibly protective subunit malaria vaccine RTS,S proceeds through pivotal multicenter trials, the new evidence that exposure to a limited number of parasites induces sterile protection beyond 2 years will raise expectations for more highly efficacious immunogens in the future, potentially boosted by naturally occurring parasite exposure. Enhanced tools to interrogate and visualize the in vivo immune response, and genome-based platforms to comprehensively assess parasite and host responses, are informing and accelerating the community effort to reach this critical public health goal.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.