The findings of this study may aid in the discovery of novel P. vivax antigens with potential for vaccine development with capacity to prevent the infection 34,47,48. To our knowledge, this is the first time that sera from a PvRAS clinical trial demonstrating significative sterile immunity have been screened for breadth of antibody response to identify parasite proteins associated with P. vivax malaria infection prevention. Although vaccines targeting other stages of the parasite cycle are important, in the case of P. vivax, due to the development of liver hypnozoites, and its consequent relapsing behavior pre-erythrocytic vaccines capable of inducing sterile immunity are of the utmost importance.
Although the original PvRAS aimed at delivering ten immunizations, the complex logistics imposed by the lack of P. vivax in vitro culture and the need for parasites from clinical infections reduced the immunization schedule to seven doses 11. Clinical trials with PfRAS using in vitro adapted parasite clones are highly feasible and reproducible leading to > 77%- >90% sterile protection of volunteers 49,50. Nevertheless, PvRAS vaccination led to 42% (5/12) protection, providing the opportunity to compare the seroreactivity of protected and non-protected volunteers. In both vaccination groups, the reactivity increased as immunization progressed, confirming the dose-response effect.
This study profited from the access to Fy + and Fy- volunteers which were included to attempt dissecting the early and late immune responses elicited during the P. vivax liver cycle. While Fy + individuals support the complete P. vivax cycle, both pre-erythrocytic and erythrocytic phases, Fy- individuals allow the complete parasite liver cycle but not the erythrocytic phase which is arrested upon parasite entry in the blood circulation, due to the lack of the Duffy Antigen Receptor for Chemokines (DARC/Fy receptor) on the erythrocyte surface, required for merozoite invasion 51; therefore, the immune response to P. vivax is expected to differ in these two populations. Additionally, including volunteers subjected to mock immunization allowed the dissection between parasite specific responses and the potential influence of mosquito saliva proteins in the specific immune response to the parasite antigens.
Regarding the human vaccination with Plasmodium RAS, it has been demonstrated that attenuated parasites get arrested early during the liver phase and that therefore, the immune responses elicited by RAS and responsible for sterile-protection presumably target parasite antigens expressed during the early hepatic schizogony. Indeed, the absence of microscopic or parasite DNA detection by qPCR during the immunization period in Fy + volunteers indirectly confirms the complete parasite radiation attenuation, and presumably the early parasite arrest 52; in this context, no responses are expected to arise against blood parasite stages.
The comparison of antibody responses induced by PvRAS sterilely protected and non-protected volunteers revealed a notably lower reactivity in the protected volunteers. However, after probing serum from both groups in the custom P. vivax protein microarray, the antibody response profile associated with sterile immunity revealed ten proteins with the highest reactivity, notably for the PvCSP and the PVX_089630 proteins. Responses to these two antigens were significantly lower in the non-protected volunteers, suggesting them as potentially responsible for protection. Indeed, a PvCSP formulation recently evaluated in naïve and semi-immune volunteers conferred sterile protective efficacy of 35% and 40% and overall protection of 55% and 60%, respectively 53. The extraordinary boost in the immune response induced by the CHMI was particularly intriguing, considering that it was performed with a limited number of parasites (2–4 mosquito bites). Again, this robust antibody boosting was more evident in non-protected volunteers, which might be explained by the parasite growth and multiplication in that group, which obviously did not occur in the sterile protected volunteers.
As expected, in the case of the Fy- control group, despite the massive inoculation of live sporozoites during immunization, did not develop patent microscopic parasitemia. However, parasite DNA was detected in blood by qPCR during the first three vaccination doses, but not after; this observation indicated that these three sporozoite immunization doses were sufficient to induce sterile immunity11. In this group, seroconversion was observed between the second and fifth immunizations, notably to PvCSP and PvMSP-1, but also to other 28 proteins. Overall, the number of reactive antigens was higher than in the PvRAS group, which is reasonable as volunteers were exposed to seven doses of live sporozoites. Surprisingly, a few antigens were only recognized once during immunization or after CHMI (PVX_117680, PVX_091785, PVX_117150, PVX_091970). This group displayed an important antibody boosting upon CHMI despite the significantly lower number of sporozoites inoculated during the 2–4 mosquitoes used for the CHMI.
It has been reported that Fy- individuals from Madagascar, Cameroon, and Ethiopia may develop parasitemias when infected by P. vivax 54,55. However, all Fy- volunteers in this study were refractory to blood infection by P. vivax. Samples from those volunteers allowed the evaluation of antibody response elicited specifically against pre-erythrocytic stages. Notably, after the second immunization, the reactivity against a group of antigens was maintained or increased through the immunization schedule, suggesting that immunization with heterologous parasites induces a similar antibody response. Due to the need to use P. vivax parasites from natural clinical infections, each immunization is potentially a genetically different parasite, despite the limited number of circulating antigenic variants from which the blood samples to infect mosquitoes were obtained 12. In addition, the presence of antibodies against several members of the MSP family and other antigens in the asexual blood stages indicates the presence of asexual parasites in peripheral blood. Although these stages cannot invade and replicate in Fy- erythrocytes, the number of parasites (merozoites) released from the liver is likely enough to elicit the production of detectable antibodies. This agrees with minor symptoms and parasite DNA found in the peripheral blood 8 to 16 days after the initial immunizations (second to fourth). Since several of the antigens identified in this study are hypothetical proteins, further studies should be carried out to better characterize them.
Although counter-intuitive, observing lower reactivity in the PvRAS protected group is not unusual. Recent studies on the response of individuals from malaria-endemic and non-endemic areas to Pf-RTS,S and PvCS showed a similar hypo-responsiveness5,42,43,53 suggesting that individuals from malaria-endemic regions, either actively infected or not, display an altered basal immune status with a paucity of regulatory mechanisms and altered memory cell function leading to lower responsiveness to vaccines. It appears to correspond to an immunological imbalance caused by permanent exposure to malaria parasites, mosquito bites, and potentially other host and environmental factors that may influence the host’s immune response and immunity to malaria 5,6,41,56. A similar trend was present in volunteers subjected to a P. vivax CHMI, in which individuals naturally exposed to P. vivax malaria with parasitemia and no fever (i.e., clinically protected) had lower reactivity compared to those with fever (i.e., clinically not protected) 10. Moreover, this finding is also consistent with those from P. falciparum vaccination studies in humans where protected individuals did not mount a significant antibody response to CHMI, whereas non-protected individuals had elevated signals to many blood-stage antigens 34,47 perhaps indicating a decrease in strength and variety of exposure to antigens due to earlier control and arrest of parasitic development in protected individuals. This may also hint at the importance of the cellular immune response in achieving protection, which was not assessed in this study.
Despite most of the reactive proteins being hypothetical, a high response was observed against PvCSP, a protein initially discovered by induction of the circumsporozoite precipitation reaction by sera from mice immunized with P. berghei-RAS. Since its discovery, PfCSP has been the most extensively studied malaria antigen, leading to the only vaccine approved by the WHO for mass use 57. Moreover, PvCSP has been the subject of extensive studies 29,32,58 and likely represents the most advanced P. vivax vaccine candidate, with important protective efficacy 53. Notably, we observed higher reactivity to several hypothetical proteins compared to PvCSP in protected volunteers, encouraging further characterization, with emphasis on their potential value as P. vivax pre-erythrocytic vaccine candidates as reported by other studies 45,46.
Although most antibodies are short-lived, those found six months after CHMI can be used as markers of recent malaria infection. Indeed, one of those antigens (PVX_083560) was found previously in semi-immune individuals and was higher in those without fever at day 45 after CHMI10, thus it might be related to clinical protection. Nevertheless, whether the antigens found here remain for a longer period or whether they protect against clinical symptoms in future episodes is unknown.
Another interesting finding was the reactivity of sera from the mock-immunized control group, which was only exposed to mosquito saliva during immunization. Despite a low reactivity and no significant difference in the average fluorescence intensity between immunizations, all volunteers presented a robust reactivity with 18 parasite proteins, after volunteers’ exposure to parasite antigens upon CHMI. Whether this corresponds to a primary response to parasite antigens or a secondary (boosting) response to cross-reacting mosquito-parasite antigens remains to be determined. It is known that mosquito saliva activates innate and adaptative immune host immune responses at the biting site by activating neutrophils, monocytes, and dendritic cells and increasing both Th1/Th2 and T cell regulatory responses 39,56. However, little is known about parasite antigens that could be expressed in the mosquito and released in the saliva 59. Therefore, these findings open new avenues to studying the mosquito-parasite interaction.
Apart from the studies reported here on the identification of pre-erythrocytic antigens recognized by vaccinated volunteers, previous studies have focused on the analyses of parasite antigens associated with clinical protection induced by natural exposure to the parasite in endemic areas of Papua New Guinea (PNG) 45,46. In those studies, authors selected the 20 most likely associated with protection from a total of 342 P. vivax antigens, from which four were merozoite surface antigens which were also recognized by sera from our PvRAS trial upon challenge and by Fy- volunteers vaccinated with live sporozoites. In addition, the PvCSP that was consistently recognized from early after sporozoite immunization was also reactive by clinical samples from PNG45,46.
In summary, we identified a group of P. vivax antigens whose antibody responses are elevated after RAS exposure and appear to contribute to sterile immunity. We also identified candidate proteins to detect previous malaria exposure due to the more durable humoral response they elicit. Taken together, these findings contribute to understanding the antibody response to P. vivax infection, particularly to the correlates of protection. Deeper analyses are required for the identification of potential surrogate markers or signatures of immune protection using systems biology 5,30,60,61