Biological Activities of Anti-Merozoite Surface Protein-1 Antibodies Induced by Adjuvant-Assisted Immunizations in Mice with Different Immune Gene Knockouts

Abstract

Immunizations with Plasmodium falciparum MSP1-42 or MSP1-19 induce antibodies that inhibit parasites in vitro, which correlates with in vivo protective immunity by vaccination. We previously showed that several adjuvant formulations can induce anti-MSP1-19 antibodies in interleukin-6, intercellular adhesion molecule 1, CD80, and CD86 knockout (KO) mice and at levels similar to those obtained in the healthy uninfected hosts. Here, we determine whether these immune gene KOs or the immunopotentiating activities of the adjuvants have a more important influence on the induction of parasite-inhibitory anti-MSP1-19 antibodies. Results showed that the biological activities of the anti-MSP1-19 antibodies induced by these adjuvants were not affected by the immune gene KOs. All adjuvant formulations that induced significant inhibitory antibody responses (i.e., >50% inhibition of parasite growth) contained monophosphoryl lipid A (MPL) in emulsion carriers, whereas MPL or emulsion carriers alone were ineffective. The ability to retain vaccine efficacy by the MSP1-19 and adjuvant formulations in the altered immunological background is a valuable and significant attribute in light of many instances of skewed immune status in the targeted vaccine populations.

Many vaccines against infectious diseases require coadministration of adjuvants to enhance the potency and efficacy of the vaccine. It has also been established that adjuvants can alter the quality of the vaccine-induced immune responses, such as TH1 versus TH2 responses, different effector cell populations, immunoglobulin (Ig) isotypes, etc. (19, 33, 35, 41). Moreover, adjuvants have been shown to have significant influence on the specificity of epitope recognition (6, 15, 38).

While many adjuvants, e.g., monophosphoryl lipid A (MPL) and CpG, act initially by specifically binding with cell surface ligands, such as the Toll-like receptors (reviewed in reference 34), the ensuing immunological cascades, including the production of cytokines and expression of costimulatory molecules, have much broader immunobiological effects or consequences that may or may not overlap among distinct adjuvants. It is, however, less clear what role and/or how important that these secondary events play to specifically augment the potentiation activities of immunological adjuvants.

Using the Plasmodium falciparum merozoite surface protein 1 (MSP1) vaccine as a model, we have recently evaluated the potency of a number of adjuvant formulations in mice rendered deficient (knockout [KO]) in a number of immune molecules, namely, gamma interferon (IFN-γ), interleukin-4 (IL-4), signal transducer and activator of transcription factor 6 (STAT6), CD80, CD86, and intercellular adhesion molecule 1 (ICAM-1) (11, 13, 16). The aim was to determine the relative importance of these secondary signals in the adjuvant's ability to induce antigen-specific antibody and T-cell responses. Some of the adjuvants were shown to retain potency in inducing antibody responses in one or more immune deficient mouse models (11, 13, 16). Since adjuvants influence the specificity of antibody responses to MSP1 (6, 14, 15), and the ability of the anti-MSP1-19 antibodies to inhibit parasites in vitro is a strong correlate of vaccine-induced immunity (4, 17, 42), we sought to address a key question in the present study. The question is what is the relative importance of the host's immune environment changes described above, compared to the immunopotentiating effects of adjuvants, in determining the ability of the MSP1-19 vaccine formulations to induce parasite-inhibitory antibodies? Results showed that in the immune gene knockout models we examined, the induction of parasite growth-inhibitory antibodies was not affected by the selective immunodeficiency; rather, it is influenced only by the adjuvant formulation used. Thus, MSP1 vaccines with the proper adjuvants may also be efficacious in hosts with altered immune status that can arise due to chronic, concurrent, or prior infections (1, 7, 8, 10, 22, 24, 26, 36, 44).

MATERIALS AND METHODS

Mouse anti-MSP1-19 serum samples.

Polyclonal anti-MSP1-19 sera were collected from various mouse strains. The mouse strains were C57BL/6 or BALB/c (wild type [WT]) and the knockout strains, IFN-γ^−/−, IL-4^−/−, IL-6^−/−, ICAM-1^−/−, CD80^−/−, and CD86^−/−, immunized with P30P2MSP1-19 (3), in three previous studies using eight different adjuvant formulations based on aqueous and emulsified MPL, QS21, MF59, and Montanide ISA720 (Table 1) (11, 13, 16). Details of formulating the adjuvant components have been described previously (11, 13, 16). All mice were immunized intraperitoneally with a 10-μg dose of P30P2MSP1-19 and adjuvants, and a total of four immunizations were given at 21-day intervals. Sera collected after the fourth immunization (i.e., quaternary sera) and matched baseline preimmunization sera were used in this study.

TABLE 1.

In vitro parasite growth inhibition by purified Ig of WT and KO mice immunized with P30P2MSP1-19 in different adjuvant formulations

Adjuvant formulationa	Mouse strain in C57BL/6 background	Antibody response of KO mice compared to WT controlsb	In vitro parasite growth inhibition (%)c
			Homologous strain (FVO)	Heterologous strain (FUP)
ISA720	WT	NA	18.0/21.2	12.6/9.8
	IL-6 KO	+	29.8/33.8	ND
	CD80 KO	O	30.2/35.6	ND
MF59	WT	NA	16.4/11.0	ND
	IL-6 KO	+	27.4/20.2	ND
	CD80 KO	O	23.6/17.4	ND
QS21	WT	NA	0.0/9.2	7.6/1.4
	IL-6 KO	O	6.8/7.2	ND
	CD80 KO	O	12.4/10.2	ND
	CD86 KO	O	0.0/0.8	ND
	ICAM-1 KO	O	11.8/5.6	ND
ISA720-QS21	WT	NA	0.0/3.8	5.0/11.4
	IL-6 KO	O	9.6/9.8	ND
	CD80 KO	O	11.2/0.0	ND
	CD86 KO	O	0.0/0.0	ND
ISA720-MPL	WT	NA	52.6/63.2	24.4/19.6
	IL-6 KO	O	48.8/55.2	16.0/10.8
	CD80 KO	O	61.6/53.0	23.2/20.4
	ICAM-1 KO	O	60.8/57.6	17.8/21.2
ISA720-QS21-MPL	WT	NA	62.2/61.2	29.6/20.2
	CD80 KO	O	65.8/56.6	23.4/17.8
	CD86 KO	O	50.0/64.4	15.0/25.8
	ICAM-1 KO	O	53.2/60.0	21.0/12.6
MPL-SE	WT	NA	68.6/66.0	28.8/19.8
	IL-6 KO	−	61.4/58.4	21.4/23.0
	CD80 KO	−	47.8/45.4	10.6/5.8
	CD86 KO	−	57.8/59.0	25.8/14.8
	ICAM-1 KO	O	61.2/60.6	27.8/10.0
MPL-AF	WT	NA	15.4/12.0	8.6/0.0
	IL-6 KO	O	4.6/11.6	ND
	CD80 KO	O	17.0/0.0	ND
	CD86 KO	−	11.2/1.0	ND
	ICAM-1 KO	O	19.8/14.6	ND

^aAdjuvant formulations as in references 11 and 13.

^bELISA antibody titers to MSP1-19. Data were from references 11 and 13. NA, not applicable; +, titers higher than the titers in WT mice; O, titers not significantly different from those of WT mice; −, titers significantly lower than those of WT mice.

^cPercent parasite growth inhibition of test antibodies compared to preimmune serum antibodies from the same mouse group. Results of two independent assays are shown. ND, not done.

Selection of serum samples for studies.

We have previously determined that the ability of anti-MSP1 antibodies to inhibit parasite growth is dependent on both antibody titer and specificity (6, 14, 18). Thus, selection of test sera was based on satisfying the following two criteria: (i) adjuvant formulations that induced a minimum enzyme-linked immunosorbent assay (ELISA) antibody titer of ≥1/20,000 in the wild-type mice and (ii) KO mouse strains that produced anti-MSP1-19 antibody titers similar to those of their wild-type counterparts. Table 1 summarizes the mouse strains and adjuvant groups that were studied. Pooled serum samples from five mice from each group were used. Preimmune serum samples from each mouse group were similarly processed as controls.

Enrichment of Ig fraction from mouse sera.

Mouse sera were heat inactivated at 56°C for 40 min as previously described (6). Individual mouse serum samples from each study group were pooled for further processing as follows. Ig fractions were precipitated in 50% NH₄SO₄, and redissolved in distilled H₂O. Dialyses and concentration were performed by three successive changes of the Ig solution with RPMI 1640 using a spin column, Vivaspin (ICS BioExpress, Kaysville, UT) with a molecular weight cutoff of 100,000, at 14,000 rpm for 5 min. The final volume was adjusted to an equivalent mouse serum concentration of 15% in RPMI 1640. In order to support parasite growth, normal, heat-inactivated human serum was also added to give a final concentration of 5%.

In vitro parasite growth inhibition assays.

Parasite growth inhibition assays using semipurified mouse immunoglobulins were performed as previously described (6). The FVO parasite strain was used, as it shares the homologous MSP1-19 region of the P30P2MSP1-19 vaccine, and the 3D7 strain was used as the heterologous strain, since it does not share the variant amino acids in MSP-19 with the FVO parasites (27). The final concentration of mouse Ig was equivalent to that of a 15% serum concentration. Briefly, parasite cultures were synchronized with 5% sorbitol (25), and late schizont stages were used as starter cultures at an initial parasitemia of approximately 0.2% and 0.8% hematocrit. These cultures were incubated in various purified mouse Ig fractions for 72 h without medium replacements, but the infected red blood cells were resuspended at 24-h intervals. Thin blood smears were prepared, and the percent parasitemia was determined microscopically after Giemsa staining. The degree (percent) of parasite growth inhibition of the test mouse immune Ig was determined by comparison of the percent parasitemia in cultures incubated with the preimmune Ig fractions of the respective mouse groups as described previously (6) and calculated as follows: percent growth inhibition = 100% × [(P − O) − (T − O)]/(P −O), where P is the percent parasitemia of cultures in preimmune Ig at 72 h, T is the percent parasitemia of parasite cultures in test Ig at 72 h, and O is the percent parasitemia of cultures at the start of experiment. Statistical comparisons of the degree of parasite growth inhibition were done by Student's t tests.

RESULTS

Table 1 summarizes the results of the inhibition of in vitro parasite growth using semipurified Ig fractions of mice, WT and KOs, immunized with different adjuvant formulations. In this study, we arbitrarily chose growth inhibition of >50% as significant because in monkey vaccination studies with MSP1, the protected animals have antibodies that inhibited in vitro parasite growth >50% (4, 17, 42), and in human clinical trials of MSP1-42, the vaccinees were not protected, and they had antibodies with <50% inhibition (2). On the basis of this cutoff value, the adjuvant formulations, ISA720-MPL, ISA720-QS21-MPL, and MPL-SE induced significant parasite growth-inhibitory antibodies, whereas ISA720, MF59, QS21, and MPL-AF did not.

An important finding is that those adjuvant formulations that induced parasite-inhibitory antibodies in the WT mice showed little or no reduction of activities in the various KO mouse strains studied. The inhibitory activities of the antibodies from these KO mice remained above the 50% level. The only exception was the MPL-SE-CD80KO group in which the growth inhibition dipped below 50%. Equally important is the finding that for those formulations that failed to induce significant parasite-inhibitory antibodies in WT mice, the various KO counterparts produced similar antibody responses in terms of biological activity. Using the same adjuvant formulations, we have shown that the immunogenicity of the P30P2MSP1-19 in the CD80 CD86 double knockout mice is extremely poor with little or no anti-MSP1-19 antibodies detected (13). The sera from these double KOs were similarly tested, and no parasite-inhibitory activities were detected (data not shown).

Figure 1 plots the ELISA antibody titers versus percent parasite growth inhibition of the anti-MSP1-19 antibodies from all mouse groups (open and closed circles). When all groups were analyzed together, antibody titers alone correlated poorly with growth inhibition (dotted regression line, r² = 0.0464). We reanalyzed the data by separating them into two groups and plotted separately. The first group (solid circles) represents the adjuvant formulations that induced significant parasite growth-inhibitory antibodies (i.e., >50% growth inhibition), while the remaining adjuvant formulations comprised the second group (open circles). The correlations between antibody titers and growth inhibition for these two data sets were r² values of 0.594 and 0.119, respectively. The ranges of antibody titers between these two groups (inhibitory and noninhibitory) were similar, and there were no statistical differences (P = 0.66, Student's t test). There was no correlation between growth inhibition and Ig isotypes, even when the data were similarly stratified (results not shown).

FIG. 1.

Relationship between ELISA anti-MSP1-19 titer and the degree of parasite growth inhibition of anti-MSP1-19 antibodies induced by different adjuvant formulations in WT, IL-6 KO, CD80 KO, CD86 KO, and ICAM-1 KO mice. Data are segregated into two groups represented by filled circles and open circles having parasite growth inhibition of >50% and <50%, respectively. Antibodies induced by the adjuvants ISA720-MPL, ISA720-QS21-MPL, and MPL-SE (filled circles) and antibodies induced by the adjuvants ISA720, MF59, QS21, ISA720-QS21, and MPL-AF (open circles) are shown. Regression lines (solid lines) for each of the two groups of data are shown with 95% confidence limits (dashed lines). A regression line (dotted pattern) for the entire combined data set is also shown.

Adjuvant formulations that induced inhibitory antibodies have the common characteristics that they were all comprised of MPL derivative in emulsion-type carriers (Table 1), but MPL alone (MPL-AF) or emulsion-type formulations alone (ISA720 or MF59 alone) were not effective. Furthermore, the addition of QS21 to ISA720 (ISA720-QS21 formulation) did not enhance efficacy.

The ability of the anti-MSP1-19 antibodies (FVO specific) to inhibit a heterologous parasite strain carrying a completely different set of variant epitopes on MSP1-19 (3D7 strain) was investigated for those antibody samples showing significant parasite inhibition. The results (Table 1, rightmost column) showed that none of the antibody samples were able to inhibit the 3D7 parasites at a significant (>50%) level. Taken as a group, the degrees of inhibition against heterologous parasite were significantly lower than those for the homologous parasites (P < 0.00001). On the other hand, the degrees of inhibition of the heterologous parasite by this group of antibody samples were significantly higher than the degrees of inhibition against homologous parasites by antibody samples from other adjuvant groups (ISA720, MF59, QS21, ISA720-QS21, and MPL-AF) (P = 0.018). In addition, the degrees of heterologous parasite inhibition were also higher than those of WT mice receiving ISA720, MF59, QS21, ISA720-QS21, and MPL-AF (Table 1, rightmost column) (P = 0.00015). This suggests that even though the degrees of heterologous inhibition were below the established significant level (>50%), they were nevertheless biologically active.

DISCUSSION

MSP1 vaccine-induced protective immunity against P. falciparum is strongly associated or correlated with the ability of the induced anti-MSP1-19 or MSP1-42 antibodies to inhibit parasite growth in vitro (4, 17, 42). We have previously shown that adjuvant formulations can affect the induction of parasite-inhibitory antibodies by MSP1 vaccines (14, 18); most recently, we show that the immunological environments of the host can also alter the functions of the adjuvants (11, 13). Thus, it becomes imperative to determine whether changes in the host's immune status, as can result from chronic diseases and infections, will have a significant impact on the adjuvant's ability to potentiate a protective response. From our recent studies in various immune gene knockout mice (11, 13), it is apparent that some adjuvant formulations are able to resist the effects of gene knockout and maintain their potency in potentiating antibody response to the MSP1-19 vaccine, while others cannot. Our results here demonstrated that within the confines of the mouse models, deficiencies of immune genes did not alter the specificity of the anti-MSP1-19 antibodies induced by the adjuvants, as measured by parasite growth inhibition. Accordingly, adjuvant formulations that induced parasite-inhibitory antibodies in WT mice did so in the KO models, and those that did not induce inhibitory antibodies remained ineffective in the knockouts. This has relevant and positive indication from a vaccine deployment point of view. MSP1 vaccine formulations that tested immunogenic and efficacious in healthy subjects will likely have similar performance in individuals whose immune systems are skewed. It is also important to point out that this study focused on the selected adjuvant formulations that have met the criteria of being able to retain their potency in the immune KO models or to induce in the KO mouse groups antibody titers of >1/20,000 with the rationale that these formulations merit further development due to their superior performance in our malaria vaccine model system.

The adjuvant formulations that induced significant parasite-inhibitory antibodies were similar in that they all have MPL derivatives formulated in emulsion-type carriers. However, MPL alone (MPL-AF) or emulsion alone (ISA720 or MF59 alone) were ineffective, despite the fact that they can induce comparable antibody titers (e.g., MPL-AF). This is not surprising since the ability of the anti-MSP1 antibodies to inhibit parasites in vitro is dependent on both antibody titer and specificity (6, 12). The parasite-inhibitory epitopes on MSP1-19 have not been fully defined. Only two conserved inhibitory epitopes have been characterized (29-31). We have recently provided evidence to suggest that additional inhibitory epitopes exist on MSP1-19 (12). Thus, meaningful comparisons of antibody specificity between inhibitory versus noninhibitory mouse sera will be attempted once a more comprehensive categorization of inhibitory epitopes is available.

The inhibitory antibody samples were much less effective against heterologous parasites, but their biological activities are significant compared with the degrees of inhibition of homologous parasites by the noninhibitory antibody samples. We previously demonstrated that when rabbits were immunized with P30P2MSP1-19, there is substantial variability in the individual anti-MSP1-19 serum samples to inhibit heterologous parasites, with some sera inhibiting heterologous parasites as effective as the homologous strain while other sera show much reduced efficacy against the heterologous strain (12). We believe it is likely that the same variability existed among individual mice in each adjuvant group, and pooling the serum samples inevitably lowered the overall inhibitory activities. Unfortunately, the volume of serum samples for each mouse was too small to be tested individually in this study.

When all inhibitory antibody samples were grouped and the relationship between the antibody titer and degree of parasite inhibition was analyzed (Fig. 1), it is evident that the degree of inhibition was relatively insensitive to changes in antibody levels, even though a positive correlation was observed. This suggests that the inhibitory epitopes were weakly immunogenic. This is in line with previous immunization studies in rabbits and monkeys in which hyperimmunizations (four injections) with MSP1 are required to induce parasite-inhibitory antibodies (4, 5, 14, 17, 18). In these scenarios, repeated boosting may result in much slower increases in the levels of inhibitory antibodies relative to noninhibitory antibodies. When all the noninhibitory antibody samples were grouped and analyzed in the same manner, a similar relationship is evident, except that the level of parasite inhibition spans a much lower range (Fig. 1, open circles). However, the ranges of antibody titers between the inhibitory and noninhibitory antibody groups were very similar (Fig. 1, compare closed circles versus open circles). We hypothesize that a critical level of parasite growth-inhibitory antibodies must be reached before significant biological activities can be observed and that this is independent of the development of the total anti-MSP1 antibody responses.

Although QS21 alone did not induce parasite-inhibiting antibodies, its inclusion in the ISA720-MPL-QS21 formulation had no negative effects compared to the effect of ISA720-MPL, but the addition provided no significant advantages. We have previously investigated the ability of adjuvant formulations based on synthetic muramyl dipeptide (MDP) and MPL in multilamellar liposomal carriers to potentiate parasite growth-inhibitory antibodies to MSP1 in mice (14). In the study, only the formulation MPL (LA-15-PH) in liposomes is able to induce significant levels of inhibitory antibodies, whereas MSP1 in liposomes alone is ineffective (14). Thus, for the MSP1 vaccines, it appears that the MPL-formulated adjuvants are efficacious in both emulsion- and liposome-type carriers. The basis for the efficacy of MSP1-19 formulations with MPL as an adjuvant in inducing parasite-inhibitory antibodies is currently under investigation. We hypothesize that the weakly immunogenic nature of the parasite-inhibitory epitopes on MSP1-19 (see above) may additionally benefit from an adjuvant, such as MPL, because it can act as a polyclonal B-cell activator and support short-term CD4⁺ T-cell clonal expansion (23, 43) aside from being able to activate antigen-presenting cells and T cells (20, 21, 28). Expansion of minor B-cell clones specific for an inhibitory epitope(s), even temporarily, may raise the absolute level of inhibitory antibodies above the threshold required for in vitro biological activities to significantly affect parasite growth.

Our studies in mice provided evidence that MPL formulated carrier-type adjuvants may be more efficacious in inducing protective anti-MSP1-19 antibodies. We are unaware of any published work indicating that these adjuvants are equally effective for the P. falciparum MSP1-19 vaccines in a monkey (Aotus) model of this human malaria. An immunogenicity study of the Plasmodium vivax MSP1-19 in marmosets with different adjuvants shows that MPL in combination with trehalose dimycolate is inferior to QS21 in inducing antibody responses, but in vivo protection and antibody-mediated in vitro parasite growth inhibition were not evaluated (40). It is unclear in this study whether mixing the MPL with trehalose dimycolate, a strong immunomodulator, would alter the immunobiological characteristics of MPL. Recent and previous human clinical trials of malaria circumsporozoite protein and blood-stage antigens, including MSP1, have used combination adjuvant formulations containing MPL derivatives (9, 32, 37, 39). Induction of biologically active antibodies was demonstrated in these studies. However, for the MSP1 vaccine, inhibitory antibodies were produced at low levels (32). It is unclear whether these adjuvant formulations were superior to other adjuvants in enhancing functional antibodies or whether parallel observations were made in mice and primates. Thus, our results with P. falciparum MSP1-19 in different MPL-formulated adjuvants would benefit from further validation of efficacy in Aotus monkey vaccination-challenge studies.

In summary, as an extension from previous studies in which we identified several adjuvant formulations that can retain potency of the malaria vaccine, MSP1-19 in selective immune skewed environments, we demonstrated here that a subgroup of these same adjuvants could also induce parasite-inhibitory, anti-MSP1-19 antibodies. Importantly, such efficacy was unaffected by similar immunodeficiencies. This is clearly advantageous as individuals from many of the vaccine-targeted populations in developing countries may suffer from concurrent infections that would skew their immunological environment (1, 7, 8, 10, 22, 24, 26, 36, 44). On the basis of these findings, MSP1 with MPL in oil emulsions merits further investigations as formulated vaccines.