Current approaches to malaria vaccines

Abstract

The complex Plasmodium life cycle offers different vaccine approaches with distinct parasitological and clinical effects. The approaches and their rationales were established decades ago: vaccines targeting pre-erythrocytic (sporozoite and liver-stage) parasites prevent infection, those to blood-stage parasites reduce disease, and those to sexual-stage parasites or mosquito vector reduce transmission and eliminate malaria through herd immunity. The pre-erythrocytic RTS,S vaccine (Mosquirix, GlaskoSmithKline (GSK)), recommended by WHO in 2021, reduces clinical malaria in children. Knowledge of parasite biology, host–parasite interactions, and immune mechanisms is informing new concepts to improve on RTS,S and to target other parasite stages. This review emphasizes vaccine approaches and candidates currently in the clinic or likely to enter clinical testing soon.

Introduction

In October 2021, WHO recommended RTS,S/AS01 vaccine (hereafter RTS,S) to reduce clinical malaria in young children in moderate-to-high transmission zones. RTS,S is a subunit virus-like particle (VLP) pre-erythrocytic vaccine (PEV) targeting immunodominant Plasmodium falciparum circumsporozoite protein (CSP) on sporozoites. WHO approval culminated 3 decades of development, making RTS,S the first parasite vaccine recommended for human use.

New regimens (NCT03276962) [1], deployment strategies [2••], or biosimilars [3•] seek to improve RTS,S partial efficacy. Other concepts are advancing (Supplemental Tables S1, S2), with several recent malaria vaccine ‘firsts’: whole-sporozoite PEV yielded field efficacy; blood-stage vaccines (BSV) reduced parasite multiplication during controlled human malaria infection (CHMI); BSV designed for pregnant women entered trials; novel BSV antigens/concepts emerged from seroscreening expression libraries; transmission-blocking vaccines (TBV) against mosquito-stage parasites advanced to Phase-2 trials. In this update, we highlight progress with candidates in the clinic or concepts poised to enter trials soon.

Pre-erythrocytic vaccines

PEV target clinically silent sporozoite and liver-stage parasites (Figure 1) and were inspired by radiation-attenuated sporozoites (RAS) shown to confer sterilizing immunity in animals and humans decades ago. RAS immunity in humans prevents infection with homologous (identical strain) [4] and heterologous P. falciparum sporozoites (PfSPZ) [5] but not blood-stage parasites [4]. Antibodies, CD8 T cells, and CD4 T cells contribute to RAS-protective immunity in animals: antibodies against surface antigens clear sporozoites from skin or bloodstream or block their invasion of hepatocytes, while CD8 T cells attack infected hepatocytes (Figure 1).

Figure 1.

Pre-erythrocytic Plasmodium development and immune mechanisms of leading vaccine candidates. RTS,S and R21 are similar subunit vaccines that induce antibodies and CD4 T cells against the CSP displayed on sporozoite surface and expressed by liver-stage parasites. CSP monoclonal antibodies may target sporozoites in skin (as seen for rodent parasites [71]), bloodstream [72], during hepatocyte traversal [72] and invasion [73], and liver-stage development [74]. RTS,S serology studies suggest roles for repeat- and C-terminus-specific antibodies in protection, involving antibody-dependent phagocytosis and NK cell activation [11]. Efficacy of PfSPZ vaccines is associated with CSP titers and Vδ2 γδ T-cell levels [28,30••,31•]; Vd2 cells may be required to induce protective CD8 T cells that kill intrahepatocytic parasites [37].

RTS, S and circumsporozoite protein-based vaccines

To produce RTS,S, hepatitis-B surface antigen (‘S’) coassembles with a fusion of P. falciparum CSP fragment (containing repeat and C-terminal regions) to S (‘RTS’), forming VLPs that display the CSP fragment on a minority of capsid subunits [6]. Whereas RTS,S was initially hoped to confer sterilizing immunity [7], field studies revealed anti-infection immunity waned over several weeks, and development focused on reducing clinical malaria in children. In Phase 3, the efficacy of RTS,S in liposomal adjuvant AS01 was ~36% against clinical malaria in young children receiving 4 vaccine doses (over 20 months), with significant protection against severe malaria in children but not infants [8].

Strategies are being pursued to increase RTS,S protective benefits. Whereas RTS,S has been administered contemporaneous to other childhood vaccines, Phase 3 studies in Burkina Faso and Mali administered RTS,S to maximize CSP titers during the transmission season (referred to here as seasonal RTS,S). These countries’ under-5 children routinely receive monthly antimalarial chemoprevention during rainy season (seasonal malaria chemoprevention or SMC). Seasonal RTS,S together with SMC reduced clinical malaria by 62.8%, severe malaria hospitalization by 70.5%, and malaria mortality by 72.9%, compared with SMC alone [2••].

New RTS,S regimens are also being examined. In malaria-naive adults, a fractional third dose (one-fifth volume), and delayed dosing schedule (0, 1, and 7 months) significantly increased efficacy against CHMI [9]. In Systems Serology studies, fractional dosing elicited a balanced antibody response to repeat and C-terminal regions of CSP versus a repeat-specific response with regular regimen, and identified C-terminal-specific antibody-dependent phagocytosis [10] and repeat-specific antibody-dependent phagocytosis and natural killer (NK) cell activation as antibody features related to RTS,S efficacy [11•]. Fractional and delayed RTS,S regimens are in field trials (NCT03276962) [1].

Another approach to increase efficacy is improved vaccine design. A ‘next-generation RTS,S-like vaccine’ called R21 expresses the CSP–HBsAg fusion protein in Pichia pastoris yielding VLP in which all capsid subunits display the CSP–HBsAg fusion [12]. In Phase 2, R21 in saponin-based adjuvant Matrix-M^™ (Novavax, Inc.) yielded 77% efficacy against clinical malaria among Burkinabe children. Recently, CSP human monoclonal antibody CIS43 (whose N-terminal/repeat-region epitope is absent from RTS,S and R21) conferred sterile immunity to CHMI. The CSP N-terminus is involved in sporozoite attachment and invasion of hepatocytes [13], and N-terminal seroreactivity associated with protection from clinical malaria in Tanzanian children [14]. Full-length CSP candidates containing CIS43 epitope have been developed [15,16] and entered the clinic ([17]; NCT03589794) (Figures 2,3).

Figure 2.

Blood-stage Plasmodium development and immune mechanisms of leading vaccine candidates. Vaccine candidates targeting merozoite antigens (PfRH2, AMA1–RON2 complex, PvDBPII) generate antibodies that inhibit nonredundant erythrocyte-invasion pathways: RH5 forms a complex with two other secreted parasite proteins (CyRPA and Rh5-interacting protein (RIPR)) targeted by growth-inhibitory antibodies [75,76], and binds to membrane-anchored P113 on the merozoite surface [77] from whence it interacts with the essential erythrocyte receptor basigin; AMA1 forms a complex with parasite antigen RON2 that has transferred to the erythrocyte plasma membrane, thereby committing the parasite to invasion [78]; PvDBP forms a dimer on the merozoite surface upon contact with the Duffy antigen receptor that is then sandwiched between two region-II (DBPRII) moieties, and neutralizing antibodies can target both the dimer and the ligand-receptor interfaces [79]. Antibodies to PfGARP (that localizes to the exofacial membrane surface of early-to-late-trophozoite-infected erythrocytes) induce programmed cell death of intraerythrocytic trophozoites [49]. Antibodies to SEA-1A (the SEA-1 N-terminal fragment upstream of its centromeric protein (CCENP-C) and dimerization domains required for chromosomal segregation [80], merozoite viability, and normal egress [81]) decreased parasite replication in one study [50] but not another [51] and SEA-1A vaccination delayed mortality in mice after challenge with lethal P. berghei parasites [50]. VAR2CSA (an integral membrane protein displayed on the exofacial surface of infected erythrocyte protrusions called knobs) uses conserved channels formed by multiple domains from its extracellular region to bind its placental receptor CSA and mediate parasite sequestration in the placenta [57]; antibodies that block parasite adhesion to CSA and opsonize CSA-binding infected erythrocytes are naturally acquired by women as they become resistant to placental malaria, forming the conceptual basis for a placental malaria vaccine.

Figure 3.

Mosquito sexual-stage Plasmodium development and immune mechanisms of leading transmission-blocking vaccine candidates. TBV targeting gamete and zygote surface antigens elicit antibodies to kill or arrest parasite stages in the mosquito, thereby reducing community transmission to achieve improved malaria control or elimination through herd immunity. P. falciparum gamete surface antigens Pfs230 and Pfs48/45 and zygote surface antigen Pfs25 have all entered clinical testing in the field, while female gamete surface antigen Pfs47 showed vaccine activity in preclinical studies. In a comparative trial using the same conjugated-protein-in-adjuvant platform, Pfs230D1 delivered superior immunogenicity and serum transmission-blocking activity versus Pfs25. Pfs230D1 serum activity is enhanced by complement and lyses gametes.

Whole-sporozoite vaccines

While long known to be protective, whole-sporozoite vaccines were thought impractical for manufacture [18]. Recently, Sanaria, Inc. established a platform to purify PfSPZ from aseptic mosquitoes suitable for human injection [19]. PfSPZ vaccines are attenuated by different approaches: radiation (called PfSPZ vaccine), chemoattenuation with antimalarial drugs (called PfSPZ–CVac for chemoprophylaxis vaccination), gene deletions that arrest liver-stage development (called genetically attenuated parasites (GAP)), or PfSPZ–GA1 for the first PfSPZ candidate (NCT03163121) [20,21].

PfSPZ vaccine efficacy increases with dosage [22–24] up to an optimal dose [25]. In malaria-naive adults, PfSPZ vaccine or PfSPZ–CVac conferred high levels of sterile homologous immunity [22,23,26–28] and, at higher doses, durable high-level sterile heterologous immunity with PfSPZ–CVac using coadministered chloroquine [29,30••] or pyrimethamine [30••]. In Malian adults, 5-dose and 3-dose PfSPZ vaccine regimens administered after presumptive antimalarial treatment reduced infection risk by ~50% [24,31•], exceeding RTS,S efficacy reported in African adults [32,33]. In Kenyan infants, 3-dose PfSPZ vaccine regimens administered without prior presumptive antimalarial treatment showed no significant overall efficacy, although the highest dose regimen yielded some evidence of efficacy [34]. PfSPZ–CVac is being assessed in field trials (Supplementary Table 1). GAP vaccines have shown some efficacy in malaria-naive individuals: PfSPZ–GA1 parasites (slarp and b9 gene deletants) conferred sterile immunity in 3/25 vaccine recipients [35•]; PbVAC parasites (PfCSP gene inserted into rodent parasite P. berghei) were nonpathogenic but conferred no sterile immunity albeit patent parasitemia was delayed [36].

PfSPZ vaccines express numerous antigens stimulating a broad immune response, albeit hepatic T-cell responses may not be detectable in peripheral blood and hence difficult to monitor and study. In humans, Vδ2 γδ T-cell levels and CSP antibodies have correlated with protection [28,31•,37]. In mice, γδ T cells are required during vaccination to induce protective CD8+ T cells, but not required during sporozoite challenge for protection [37]. Among Kenyan infants, baseline Vδ2 levels were low and failed to expand during PfSPZ vaccination, prompting speculation that their dearth impaired T-cell responses and efficacy [34]. CSP titers consistently correlate with efficacy, but vary widely between studies and between malaria-naive (high titers) and malaria-experienced PfSPZ vaccinees (low titers). Thus, antibodies may contribute to protection but not be primary mediators. Altogether, evidence in mice, monkeys [38], and humans suggests that CD8₊ T cells largely mediate sporozoite(SPZ)-induced sterile immunity.

Blood-stage vaccines

BSV targets asexual parasites that cause disease and death. In the 1960s, passive transfer of African adult IgG to malarious children cleared parasitemia and symptoms. In monkeys, immunization with parasite preparations rich in merozoites conferred protection from P. falciparum infection [39], and BSV development has since primarily focused on merozoites. Merozoites egress erythrocytes as 1–2-dozen progeny (depending on species) and within seconds reinvade new erythrocytes, offering a brief window of access for neutralizing antibodies. Over 30 BSV trials were completed between 2000 and 2015 (clinicaltrials.gov), mainly targeting merozoite antigens MSP1 and AMA1. Interest in BSV waned after disappointing results: no CHMI efficacy [40] and minimal field efficacy with MSP3-containing vaccine GMZ2 [41]. In addition to limited antibody accessibility, challenges for merozoite targets included antigen polymorphisms, redundant invasion pathways, and high blood-stage parasite bioburden.

Merozoite vaccines that focus on nonredundant invasion pathways have since made progress. P. falciparum reticulocyte-binding protein homolog 5 (PfRH5) binds essential receptor basigin, is highly conserved, and induced broadly neutralizing antibody in animals [42]. PfRH5 vaccination controlled parasitemia in monkeys [43]. PfRh5 formulated in AS01 reduced parasite multiplication during CHMI after primary 3-dose series, and again after boosting 3–4 months later [44••]. Meanwhile, a protein/adjuvant formulation (PvDBPII/Matrix-M^™) of recombinant region II of Duffy-binding protein (that binds Duffy antigen, a common if not essential invasion protein [45,46]), reduced P. vivax multiplication 51% during CHMI, but only using a delayed (0, 1, and 14 months) and not monthly (0, 1, and 2) immunization schedule [47]. These vaccines are the first to significantly reduce parasite multiplication during falciparum and vivax CHMI, respectively.

AMA1 is an essential invasion protein, but conferred no efficacy in clinical trials. However, AMA1 complexed with rhoptry neck protein RON2 (the AMA1-binding partner during invasion) induces more potent antibodies than AMA1 alone [48], and superior protection in monkeys [48]. Unlike Rh5, AMA1 displays extensive sequence variation, which will need to be addressed if AMA1–RON2 complex advances to trials.

The search for novel BSV antigens includes stages other than invasive merozoites. Differential library screening using sera from resistant and susceptible children has generated novel BSV concepts, including programmed cell death of trophozoites as well as impaired schizogony and failure to egress. Antisera to PfGARP, an infected erythrocyte surface antigen, induced programmed cell death of trophozoites in vitro, and PfGARP vaccines protected monkeys from virulent P. falciparum challenge [49•]. Antibody against P. falciparum Schizont Egress Antigen 1 (PfSEA-1) impaired P. falciparum replication in vitro in one study [50] but not another [51], and P. berghei SEA-1 vaccine reduced parasitemia in mice and delayed mortality [50]. PfGARP and PfSEA-1 vaccines may benefit from a longer window of antibody access to the antigens and antibody boosting during natural infections.

Placental malaria vaccines

Placental malaria vaccines are a distinct BSV approach to protect pregnant women. Placental malaria (PM) is caused by chondroitin sulfate A (CSA)-binding parasites that express VAR2CSA. Women acquire antibodies against CSA-binding parasites (and VAR2CSA) over successive pregnancies as they become resistant to PM [52]. VAR2CSA is a variant antigen that binds CSA and encodes a large (> 300 kD) ectodomain with usually 6 but as many as 7–8 Duffy-binding-like (DBL) domains [53].

Two vaccines (PRIMVAC( and PAMVAC) that target N-terminal VAR2CSA fragments completed human testing as protein-in-adjuvant candidates (in alum or in a stable emulsion containing TLR4 ligand GLA). Both induced adhesion-blocking antibodies to homologous parasites, with little activity described against heterologous parasites [54],[55•]. Similar results have been seen in monkeys, in which PM episodes boosted VAR2CSA titers but not heterologous functional activity [56].

Recent VAR2CSA structural studies reveal that CSA binds to major and minor channels formed by a core encompassing DBL1-4 domains, indicating that neither PRIMVAC nor PAMVAC antigens recreate the CSA-binding site [57•]. CSA-binding channels but not flanking regions are highly conserved, offering a blueprint for immunogens that focus on variant-transcending functional epitopes.

Transmission-blocking vaccines

TBV target surface antigens of gametes and zygotes (mosquito/sexual stages) to interrupt transmission through mosquitoes [58,59]. Among clinically advanced candidate antigens, Pfs230 forms a complex with membrane-anchored Pfs48/45 on the gamete surface and plays a role in male gamete fertility [60], while Pfs25 is the major surface protein of zygotes. Female gamete surface antigen Pfs47 was recently recognized as another potential vaccine candidate [61]. These cysteine-rich highly conformational antigens are difficult to produce as properly folded recombinant protein using traditional expression platforms. Pfs25 was the first candidate prepared as recombinant protein and yielded functional antisera in rodent studies [62,63], where activity is measured by feeding P. falciparum gametocytes to Anopheles mosquitoes in antisera versus naive sera, then comparing oocyst burdens that develop. Pfs25 candidates however have yielded only weak or short-lived antibody responses and serum functional activity in malaria-naive [64–66] and malaria-experienced vaccinees [67].

Pfs230 and Pfs48/45 candidates have now entered the clinic. In a head-to-head comparison of Pichia pastoris-expressed antigens conjugated to ExoProtein (EPA) and formulated in Alhydrogel^®, Pfs230 domain 1 (D1) generated higher transmission-blocking activity than Pfs25 in humans and monkeys but not mice [65•]. Pfs230D1 activity depends on complement, and a potent complement-dependent human mAb has identified a large highly conserved neutralizing Pfs230D1 epitope on gametes [68]. Pfs230D1-EPA has now advanced through Phase-1 and −2 field trials using adjuvants such as Alhydrogel^® (NCT02334462), AS01 (NCT02942277, NCT03917654), or Matrix M^™ (NCT05135273).

Pfs48/45 domain-3 (called ‘6C’) candidates have been expressed in Lactococcus lactis as fusions with malaria antigens such as GLURP R0 region (‘R0.6C’) [69] or Pfs230 prodomain and CSP repeat sequences (‘ProC6C’). ProC6C generated higher serum functional activity than R0.6C in animals [70]. These candidates formulated with Matrix M^™ are in Phase-1 testing in Europe (NCT04862416) and Burkina Faso (PACTR202201848463189), and in Mali (ISRCTN136 49456) are being compared with Pfs230D1.

Conclusions

RTS,S regulatory approval is historic, but its antibody-mediated protection can be improved. Whole-sporozoite vaccines are generating T-cell responses and conferring protection, improved merozoite vaccines have reduced parasite growth in vivo, and a Pfs230 transmission-blocking candidate has exceeded benchmark Pfs25 vaccine activity. Progress is being made with all approaches, making this an era of optimism for malaria vaccines following earlier decades of disappointment.

Data Availability

All relevant data are included in the submitted paper. No data were used for the research described in the article.