The Impact of Established Immunoregulatory Networks on Vaccine Efficacy and the Development of Immunity to Malaria

Despite decades of effort, the development of clinically effective parasite vaccines has proven elusive. There are no licensed vaccines for humans that protect against any parasite, although the European Medicines Agency recently recommended the RTS,S/AS01 malaria vaccine for use, a ruling aimed at helping African regulators reach a decision on licensure of a vaccine which provides partial protection against clinical disease for a relatively short time (1). One of the disappointing features of parasite vaccine development, and malaria vaccines in particular, has been the relatively poor results obtained in disease endemic settings (2, 3). This has prompted the establishment of clinical trial facilities in malaria-endemic countries for vaccine testing aimed at more rapid evaluation and selection of promising vaccine candidates in target populations. However, this does not address the underlying cause of the poor performance of vaccines in disease endemic areas. This problem is not unique to parasite vaccines. For example, bacillus Calmette–Guérin–mediated protection against pulmonary tuberculosis varies geographically and appears to be much less effective in areas with a high incidence of previous infection with Mycobacterium tuberculosis or sensitization with environmental Mycobacteria (4, 5). Although there are likely to be many reasons for reduced efficacy of vaccines in disease endemic areas, we will discuss the hypothesis that the early establishment of pathogen-specific immunoregulatory networks is an important factor contributing to this problem. Furthermore, we will examine the idea of incorporating inhibitors of specific immune checkpoints into vaccine formulations or drug treatment protocols as a way to transiently reduce immune suppression to allow the generation of robust, long-lasting, antiparasitic immunity.

Malaria remains an important global health problem with over 200 million cases reported in 2015 by the World Health Organization, resulting in 438,000 deaths, mainly as a result of Plasmodium falciparum infection (6). The complex, multistage life cycle of malaria parasites typically induce suboptimal immune responses that develop slowly over time (7). The development and maintenance of immunity is complicated by the presence of multiple mechanisms of immune evasion employed by malaria parasites (8–10). When some level of immunity is achieved, it is thought to rapidly wane in the absence of parasite exposure (7, 8, 11–13). However, other studies indicate that once established, clinical immunity can be long-lasting and/or may vary depending on the parasite Ag the immune response was generated against (14–18). This is clearly an area that requires further investigation and clarification.

Convincing evidence for the feasibility of malaria vaccines exists. In the early twentieth century, prior to the discovery of the curative effect of penicillin, neurosyphilis patients were treated by deliberate induction of malaria by parenteral injection of infected erythrocytes (19, 20). The subsequent fever killed the temperature-sensitive Treponema pallidum spirochaetes responsible for disease (21), but repeated treatments were less effective due to suppressed parasite growth and reduced fever, signifying the development of some degree of immunity (21). Pioneering studies in the 1960s demonstrated that adoptive transfer of Plasmodium-specific γ-globulin from semi-immune adults into children could transfer protection against severe malaria (22), pointing to a key role for Abs in protection against disease. Thus, these observations provided indications that a malaria vaccine is achievable.

Early efforts to immunize volunteers with attenuated P. falciparum sporozoites using irradiated mosquitoes produced protective immunity, albeit short-lived (23–25), and this idea has now been further developed by producing and successfully testing a cryopreserved, irradiated sporozoite vaccine (26–28). The most advanced malaria vaccine is RTS,S/AS01 (Mosquirx), which consists of 37–49 NANP (N, asparagine; A, alanine; P, proline) repeats along with the entire C-terminal region of the circumsporozoite protein (CSP; clone 3D7) fused to the hepatitis B surface Ag and coexpressed with free hepatitis B surface Ag (to form virosomes). It is adjuvanted with AS0 series adjuvants, where the liposome formulation of AS01 was found to be best at inducing CD4⁺ T cell responses (29). This vaccine has completed phase III clinical trials with a reported protective efficacy against clinical malaria of 28% in children aged 5–17 mo living in malaria-endemic regions of sub-Saharan Africa; this increased to 36% if a fourth dose of vaccine was administered (30). These results were somewhat disappointing given the higher levels of protection achieved in healthy volunteers in the United States participating in controlled human malaria infection studies with either earlier versions of the vaccine (2, 31, 32) or even the same vaccine formulation (33). Of note however, this reduced efficacy was similar to that achieved in healthy adults living in a high malaria transmission region (3). The limited efficacy of the vaccine has been attributed to factors such as the use of a P. falciparum CSP variant in the vaccine not commonly found in Africa (34), polymorphisms within described T cell epitopes of CSP (35), as well as a suboptimal immunization schedule (30). However, another possibility is that early exposure to malaria parasites promotes the development of antiparasitic immunity aimed at ensuring host survival in malaria-endemic areas, but also antagonizes vaccine-induced protective immunity. This may not explain why RTS,S/AS01 is more immunogenic in older children (5–7 mo of age at first vaccination) than in infants (6–12 wk at first vaccination) (30), and suggests the involvement of other confounding factors, one of which may be the influence of protective maternal Abs present in the latter group that influence parasite Ag exposure and subsequent response to infection. This is an important area warranting further investigation.

Preclinical studies in mouse models of malaria show that parasite-specific Abs can prevent sporozoite invasion of hepatocytes following a bite from an infected mosquito, thereby preventing establishment of the erythrocytic stage of infection (36–38). Parasite-specific CD8⁺ T cells can also develop to recognize and kill infected liver cells, thus stopping parasites reaching the blood (39–42). During the erythrocytic stage of infection, CD4⁺ T cells and Abs are important for control and resolution of infection, respectively (43–49). However, in all stages, CD4⁺ T cells play critical roles in coordinating immune responses. These roles include providing help to B cells for high affinity Ab production, CD8⁺ T cells to kill infected cells, and innate immune cells to recognize and remove parasites from the circulation (50, 51). Malaria, like most diseases caused by intracellular protozoan parasites, requires the generation of IFN-γ–producing Tbet⁺ CD4⁺ T (Th1) cells for the activation of phagocytes to kill captured or resident pathogens, such as Plasmodium-parasitized RBCs, and allow dendritic cells (DCs) and macrophages to present Ags that prime or expand CD4⁺ T cell responses (52, 53). However, Th1 cell cytokines that mediate these processes can also stimulate the expression of integrins on vascular endothelial cells that allow the sequestration of P. falciparum-parasitized RBCs in vital organs, and the associated generation of localized inflammation. Hence, parasite-specific CD4⁺ T cell responses need to be tightly regulated so they themselves do not cause disease (54).

Specialized CD4⁺ T cell subsets are major regulators of inflammation during parasitic diseases (55, 56), and can be broadly divided into FoxP3-expressing regulatory T (Treg) cells produced in the thymus (57) and inducible Treg cells that emerge from the thymus as conventional CD4⁺ T cells, but develop regulatory functions in the periphery following exposure to inflammatory stimulation (55). These latter cells include IL-10–producing Th1 or type 1 Treg (Tr1) cells (58, 59), that are increasingly recognized as a critical regulatory CD4⁺ T cell subset that protects tissue from damage caused by excessive inflammation (56, 60, 61). Tr1 cells lack FoxP3 expression and can develop from Th1 cells in environments where there is chronic exposure to Ags (55, 56). There are no clear cell lineage-defining markers for these cells, although LAG-3 and CD49b can be used to identify both mouse and human Tr1 cells under certain conditions (61). IL-10 produced by these and other immune cells acts as a major regulatory cytokine to suppress inflammation by directly inhibiting T cell functions, as well as upstream activities initiated by APCs (60). Initially, IL-10 production was identified in Th2 cells (62), but has since been described in Th1 (59, 63, 64), Treg (65, 66) and IL-17–producing CD4⁺ T (Th17) (67) cell populations. Thus, CD4⁺ T cell–derived IL-10 production is emerging as an important mechanism to dampen T cell activation in the face of intractable infection and/or persistent Ag exposure, as may occur following Plasmodium infection. In mice infected with protozoan parasites, Th1 cells are an important source of IL-10 that promote parasite survival, but also limit pathology (68–76). These Tr1 cells have also been identified in African children with P. falciparum malaria, and they appear to be Ag-specific because they expand after stimulation of PBMCs with parasite Ags (77–79). One of their proposed functions is to protect tissue from damage caused by excessive inflammation (56, 60, 61). Thus, resistance to malaria may involve the development of specialized, parasite-specific CD4⁺ T cells that suppress control of parasite growth, but also prevent host death caused by an excessive inflammatory response.

In malaria-endemic areas, immunity develops slowly, such that adults generally harbor fewer parasites than children. However, children develop resistance to severe disease relatively quickly, and can be asymptomatic while carrying significant parasite burdens (7, 11). Inflammatory responses generated against parasites early in life are aimed at controlling parasite growth, but, as mentioned above, can also contribute to the clinical symptoms of malaria and associated disease. Hence, malaria-induced inflammation must be tightly controlled soon after first exposure to parasites, and although this may cause delayed parasite clearance, it may be necessary to prevent host death. One way that inflammation is controlled is through the establishment of immunoregulatory networks, which appear to be key adaptations of host-parasite interactions (56) (see below).

The blood-stage of infection is responsible for all the clinical symptoms associated with malaria, and Abs are critical for clearance of parasites from the blood and disease resolution (80–82). However, not all Abs protect and the generation of high avidity, cytophilic IgG1 and IgG3 isotypes that recognize parasites and activate complement are critical for preventing parasite invasion of RBCs (83). A specialized subset of CD4⁺ T cells called T follicular helper (Tfh) cells are essential for the selection and maturation of B cells that produce these Abs (84). B cell responses to P. falciparum in children living in malaria-endemic areas are short-lived (85, 86) and dysregulated (87). This altered B cell function has been associated with the generation of atypical memory B cells which are hyporesponsive to stimulation and exhibit an exhausted cell phenotype (88), thus providing evidence for B cell intrinsic factors impacting upon Ab generation. In addition, Tfh cell types with markedly different capacities for providing B cell help have also been reported, including those with similar characteristics to Th1, Th2 and Th17 cell subsets (89). A recent study of Malian children living in an area of seasonal malaria transmission found that Th1-like Tfh cells (Tbet⁺ PD-1⁺ CXCR5⁺ CXCR3⁺) were preferentially expanded following P. falciparum infection, but, of note, these cells had a diminished capacity to help B cells produce protective Abs, compared with Th2-like Tfh cells (GATA3⁺ PD-1⁺ CXCR5⁺ CXCR3⁻) (90) (Fig. 1). Recently, the dysregulated production of IFN-γ and TNF by Th1 cells in a mouse model of malaria was shown to impede development of Tfh cells, resulting in suppressed germinal center reactions, as well as inefficient production of extrafollicular plasmablast responses (91). These findings indicate that inflammation generated during malaria not only contributes to disease, but also impacts upon development of immunity by suppressing the generation of protective antiparasitic Abs. Thus, not only do Tfh cells polarize to a phenotype that is not ideal to help B cells produce high avidity, cytophilic Abs, but malaria-associated inflammation also promotes the development of tissue environments that are not conducive to efficient Ab generation. The reason that regulatory T cells generated early in infection do not prevent this inflammation-mediated, dysregulated Ab response is unknown, but clearly represents a knowledge gap needing further investigation.

Many regulatory molecules and cell populations have been identified in preclinical models of malaria, as well as in malaria patients. These include cytokines such as IL-10 (76, 92) and TGF-β (93, 94), as well as immune checkpoint molecules such as CTLA-4 (95, 96), LAG-3 (97, 98), PD-1 (97, 99) and TIM-3 (100, 101). These latter molecules are often reported on T cell populations following chronic Ag exposure, so it is not surprising that they are expressed by T cells from individuals living in malaria-endemic areas. Nevertheless, the availability of Abs to manipulate the function of T cells expressing these molecules provides significant opportunity to investigate their impact on developing immunity. In particular, it will be important to establish which T cell subsets express these molecules and determine the consequences for manipulating their function on both antiparasitic immunity, as well as pathogenesis. If antiparasitic immunity can be improved without promoting pathology, then strategies to manipulate relevant cell signaling pathways as part of vaccination and/or drug treatment can be considered. Treg cells with increased suppressive function have been reported in adults with malaria (94, 102). When PBMCs from healthy individuals were cultured in the presence of P. falciparum Ag, there was an expansion of Treg cells and an increase in their ability to suppress T cell responses (102), supporting the presence of parasite Ag-reactive Treg cells in humans. However, studies in African children have shown that neither Treg cell number nor function differ between patients with uncomplicated and severe malaria (78). Studies in Ugandan children from areas of different malaria exposure indicated that burden of disease may have an important impact on number and function of Treg cells (103). Other findings from malaria patients in the Peruvian Amazon have shown that neither Treg cell frequency nor number was associated with the risk of malaria-related symptoms (104), suggesting that alternative T cell-mediated mechanisms of immune regulation may be important for controlling inflammation and thus preventing disease.

As mentioned above, the development of immunoregulatory mechanisms to control inflammation during P. falciparum infection may explain why individuals living in malaria-endemic areas have reduced numbers of severe malaria episodes with age (105, 106). Studies on individuals from a malaria-holoendemic area of Western Kenya showed that the proportion of individuals producing IL-10 in response to stimulation of PBMCs with parasite Ags increased with age, and was inversely correlated with the clinical severity of illness (107, 108). Interestingly, increased IL-10 levels in adult Papuans with uncomplicated P. falciparum and P. vivax malaria were linked to increased DC apoptosis and an associated impairment in the ability of these cells to activate T cells (109). In addition, PBMCs isolated from individuals from a hypoendemic area of Thailand who had P. falciparum and/or P. vivax infection in the previous 6 y maintained their ability to produce IL-10 following stimulation with parasite Ags, but rapidly lost the ability to produce IFN-γ (110). This suggests that parasite-specific regulatory responses are long-lived, whereas inflammatory responses are not. However, there is also evidence that these responses are not long-lived (see below), thus making this an important area requiring clarification. The emergence of Tr1 cells following Plasmodium infections has now been reported by several groups in areas of different malaria exposure (77–79, 111). Tr1 cells, but not Treg cells, were more prevalent in Gambian children with uncomplicated malaria compared with those with severe disease (78). A feature of Tr1 cells is that they emerged relatively early in life in children living in malaria-endemic areas (77, 79), suggesting a role for these cells in rapid control of inflammation during disease. Although the studies described above relate to immunoregulatory networks for blood stages of Plasmodium infections, several Ags, such as AMA-1 and MSP-1, are expressed in both liver and blood stages of the parasite lifecycle (112–115), making it possible that regulatory mechanisms generated against blood stage Ags, may also impact on immune responses to sporozoite Ags. Again, this is an important area requiring further research.

In a high malaria transmission area of Uganda, children had reduced frequencies of Treg cells (103). Nevertheless, cellular immune responses were dominated by parasite-specific Tr1 cells, and although their emergence was largely dependent upon recent malaria exposure, their frequency was not associated with protection from future disease, after controlling for prior malaria incidence (79). Significantly, in an area of seasonal malaria, children acquired exposure-dependent Tr1 cell responses that regulated inflammation, while at the same time acquiring antiparasitic immunity (77), indicating that this immunoregulatory pathway may not completely suppress developing immune responses. Other studies have reinforced the observation that the effector phenotype of P. falciparum-specific CD4⁺ T cells is largely influenced by age and transmission intensity (116). During early exposures, children in low transmission areas mount a predominantly inflammatory response governed by higher frequencies of IFN-γ⁺ TNF⁺ CD4⁺ Th1 cells (Fig. 1A) (116). In contrast, children in high transmission areas mount a response dominated by Tr1 cells (Fig. 1B) (79, 116). These Tr1 cells appear to develop from Th1 cells in experimental models of protozoan parasitic infections (117, 118). Given the immunoregulatory potential of these Tr1 cells, it is logical that they may impede the development of cellular and humoral immunity, while protecting against disease (Fig. 1B, 1C) (77, 90). Interestingly, in areas of both low and high malaria transmission, adults had higher frequencies of IFN-γ⁺ TNF⁺ CD4⁺ Th1 cells, relative to Tr1 cells, a situation that has been thought to contribute to their reduced parasite burdens (Fig. 1D) (116). Collectively, these findings suggest that the kinetics of the emergence of immunoregulatory networks depends on the level of exposure to malaria parasites, with high exposure early in life requiring the establishment of these networks to protect against life-threatening inflammation. Thus, in young children living in a high transmission setting, where the increased availability of parasite Ags favors the induction and maintenance of Tr1 cells, these cells dominate antiparasitic CD4⁺ T cell responses. As children get older and their ability to respond to infection improves, parasite-induced CD4⁺ T cell IFN-γ and TNF production is increased (116, 119). Adults in low or high transmission areas (Fig. 1D) (116) respond to infection with inflammatory responses. These can control parasite growth, but may also cause life-threatening parasite tissue sequestration and possibly severe anemia (80). Studies on transmigrant populations in Indonesia show that although adults can develop severe malaria syndromes, and may in fact be more susceptible to severe disease than children initially (80, 120), they are able to acquire protective immunity against chronic exposure to infection relatively quickly, compared with their children (80, 120). Thus, these findings suggest that cell intrinsic factors that change with age may be dependent on the conditions of exposure (acute versus chronic) (80). It would be useful from a vaccine development perspective to know how many and how frequently infections are required to transition from a disease-promoting to a protective immune response in children living in malaria-endemic areas, and furthermore, the number of infections before antiparasitic immunity becomes established.

As mentioned earlier, there are likely to be multiple mechanisms of immune regulation during malaria, and many of these involve increased expression of immune checkpoint molecules on key immune cells. In addition, type I IFNs have emerged as important immune regulators in infectious disease (121). They are required for antiviral defenses, but can also promote pathogen survival by suppressing immunity during infections, such as those caused by lymphocytic choriomeningitis virus (122, 123), influenza (124), Mycobacteria (125–129), Listeria (130), Staphylococci (131), Leishmania (132) and Plasmodium (133, 134). Type I IFNs inhibited the ability of DCs to activate CD4⁺ T cells in experimental malaria models; this has been associated with changes in costimulatory molecule expression, as well as the promotion of DC IL-10 production (135, 136). Others have also shown that Plasmodium infection drives the expansion of DCs that promote T cell IL-10 production (137). TGF-β is another important regulatory cytokine during malaria acting to suppress Th1 cell responses, as well as to promote Treg cell development (94, 138–140). There is also evidence that macrophage polarization to alternatively activated phenotypes, which produce high levels of anti-inflammatory IL-10, TGF-β, and glucocorticoids (141), can modulate inflammatory responses in experimental malaria (142). Hence, it is likely that multiple mechanisms of immune regulation are activated and persist after exposure to Plasmodium parasites.

Accumulating evidence suggests that parasite-specific immunoregulatory networks quickly establish in people living in malaria-endemic areas. A possible consequence of this adaptation is that malaria parasites persist and the development of long-lasting immunity is slow. Remarkably, many of the immunosuppressive pathways reported in people with malaria have also been described in cancer patients. In this area, these observations have resulted in revolutionary developments of immune checkpoint inhibitors for cancer treatment (143–147). Not surprisingly, potential side-effects of this treatment are immune-related adverse events, which can be life threatening (148). Thus, extensive efforts are underway to better understand how to harness the potential of this immunomodulatory strategy. Nevertheless, these promising outcomes highlight the potential of blocking or stimulating specific cellular pathways involved in immune regulation to clinical advantage. Therefore, as we gain a detailed understanding of the regulatory molecules that emerge during malaria to suppress parasite-specific CD4⁺ T cell responses that activate phagocytes to clear infected RBCs or provide help for B cells to produce appropriate, antiparasitic Abs, then we should consider the possibilities of transiently targeting these molecules to allow the rapid emergence and establishment of potent antiparasitic immunity. This could be accomplished by incorporating appropriate immunomodulatory reagents either into vaccine adjuvants or into antimalarial drug treatment protocols. As mentioned above, the immunomodulatory effects of such an approach would have to be temporary to avoid adverse consequences, and careful selection of targets and dosing schedules would be required. However, we require strategies to unlock the antiparasitic potential of the host immune response when immunosuppressive networks have been established, and such approaches may represent a new way forward. Without an awareness of these networks and efforts to shape them accordingly for the development of malaria vaccines that will be deployed in disease-endemic communities, we may continue to struggle to achieve suitable protective efficacy rates.

Recent data indicate that the deployment of insecticide-treated bed nets, more effective artemisinin combination therapy, and improved diagnostics, has led to a significant fall in the incidence of malaria in most disease-endemic regions (6). Consequently, this is changing the current patterns of morbidity and acquisition of immunity (149). As discussed above, our understanding of the temporal pattern of acquisition and maintenance of immunoregulatory responses is incomplete. Therefore, as this epidemiologic shift from hyperendemic to hypoendemic malaria takes place, an improved understanding of the development of such networks with first and subsequent episodes of clinical malaria should provide leads to minimizing morbidity and mortality of malaria.

The administration of antimalarial chemotherapy provides an opportunity to potentially alter the development of these immunoregulatory networks. The administration of high dose corticosteroids to dampen the immune response and secondary pathology has been shown to reduce mortality in some specific settings (such as tuberculous meningitis, and Pneumocysitis jirovecii pneumonia) (150, 151). However, adjunctive therapy in severe malaria has not been demonstrated to have benefit (152, 153). The development of small molecules or biologicals that modulate the immune response may pave a way for more targeted intervention strategies during the acute stages of infection. Such therapies may allow for the establishment of an immune environment that fosters the development of protective immune responses during vaccination.

Developing vaccines against parasitic diseases, including malaria, has proven difficult. Many reasons, including both parasite factors such as antigenic variation and host factors such as age, nutritional status, coinfections, and host genetics are likely to contribute to this disappointing outcome, and most will be difficult to modify. However, there is increasing evidence that antiparasitic immunoregulatory mechanisms rapidly emerge to prevent infection-mediated inflammation causing severe disease and even death. Despite research showing that these regulatory networks, and in particular parasite-specific Tr1 cells, have an impact on antiparasitic immunity, evidence that they directly affect vaccine efficacy is currently lacking and needs further investigation. Some of the key cell populations and molecules suppressing natural antiparasitic immune responses have been identified and, consequently, the potential exists for transiently targeting them to generate improved antiparasitic immunity through either vaccination or drug treatment. These key cell populations include Tr1 cells, and in some situations, possibly Treg cells. Potential molecular targets include cytokines such as IL-10, TGF-β and type I IFNs and/or immune checkpoints such as CTLA-4, LAG-3, PD-1, and TIM-3. These manipulations could be achieved via Abs or small molecule inhibitors. However, to achieve this we first need to understand what types of immune responses are needed for vaccine-mediated protection. This will need to extend beyond the identification of critical effector molecules, such as complement-fixing or invasion-blocking Abs, and include the identification of molecular and cellular mechanisms needed to generate these protective immune responses. We also require better knowledge of developing acquired immune responses in infants and young children living in malaria-endemic regions, and how the immune mechanisms needed to generate protective immunity are subverted in these individuals. Finally, we will need to develop strategies that can target selective immunosuppressive mechanisms safely. This latter challenge needs to be undertaken with the understanding that any improved vaccines or therapies will have to be deployed in resource-poor settings and, as such, will likely need to avoid expensive biological reagents, and instead rely on cheaper small molecules or alternative strategies. Nevertheless, the first important steps in generating the above knowledge have started, and there is no reason that the significant strides in manipulating immune responses for clinical benefit in cancer cannot be made for infectious diseases like malaria.

We thank Susanna Ng for preparing the figure in this article.

The authors have no financial conflicts of interest.