Design and Evaluation of Chimeric Plasmodium falciparum Circumsporozoite Protein-Based Malaria Vaccines

Efficacy data on two malaria vaccines, RTS,S and R21, targeting Plasmodium falciparum circumsporozoite protein (PfCSP), are encouraging. Efficacy may be improved by induction of additional antibodies to neutralizing epitopes outside of the central immunodominant repeat domain of PfCSP. We designed four rPfCSP-based vaccines in an effort to improve the diversity of the antibody response. We also evaluated P. falciparum merozoite surface protein 8 (PfMSP8) as a malaria-specific carrier protein as an alternative to hepatitis B surface antigen. We measured the magnitude, specificity, subclass, avidity, durability, and efficacy of vaccine-induced antibodies in outbred CD1 mice. In comparison to N-terminal- or C-terminal-focused constructs, immunization with near full-length vaccines, rPfCSP (#1) or the chimeric rPfCSP/8 (#2), markedly increased the breadth of B cell epitopes recognized covering the N-terminal domain, junctional region, and central repeat. Both rPfCSP (#1) and rPfCSP/8 (#2) also elicited a high proportion of antibodies to conformation-dependent epitopes in the C-terminus of PfCSP. Fusion of PfCSP to PfMSP8 shifted the specificity of the T cell response away from PfCSP toward PfMSP8 epitopes. Challenge studies with transgenic Plasmodium yoelii sporozoites expressing PfCSP demonstrated high and consistent sterile protection following rPfCSP/8 (#2) immunization. Of note, antibodies to conformational C-terminal epitopes were not required for protection. These results indicate that inclusion of the N-terminal domain of PfCSP can drive responses to protective, repeat, and non-repeat B cell epitopes and that PfMSP8 is an effective carrier for induction of high-titer, durable anti-PfCSP antibodies.


Introduction
There have been measurable gains in reducing the global malaria burden through the integration of control programs involving insecticide-treated bed nets, indoor spraying of residual insecticides, rapid diagnosis, treatment with multi-drug combinations, and intermittent preventative therapy [1].Nevertheless, the most recent World Malaria Report indicates that the preceding decline in clinical cases of malaria and malaria-related deaths is leveling off in most regions and is even showing signs of reversal in others [2].An efficacious malaria vaccine would be a valuable tool to add to control efforts, with a multivalent, multistage formulation being most desirable [3][4][5].In addition to the complexity of immune-mediated protection against Plasmodium falciparum, challenges for the subunit malaria vaccine effort relate to scalable production of high-quality recombinant vaccine candidates, polymorphism of T and B cell epitopes, immunogenicity and durability of vaccine-induced immune responses, antigenic competition, and adjuvant selection [6,7].Many of these were encountered during the development of RTS,S, the most advanced P. falciparum subunit malaria vaccine, which partially protected children in a Phase III clinical trial [8][9][10][11][12][13].However, protection elicited by this pre-erythrocytic-stage vaccine waned 2 of 21 and the lack of vaccine-and/or infection-induced immunity to blood-stage parasites left these children susceptible to clinical disease.It is encouraging that increases in efficacy and durability of protection have been achieved with the more recently developed R21 vaccine, a modified version of RTS,S, formulated with Matrix-M as adjuvant [14][15][16].With a positive recommendation by the World Health Organization for vaccination of children with RTS,S and R21, the deployment of these vaccines in several African countries was initiated and is expanding [17,18].
RTS,S and R21 contain a portion of the immunodominant central repeat domain of the P. falciparum circumsporozoite protein (Pf CSP) along with the C-terminus containing T cell epitopes and an adhesive thrombospondin type-I repeat (TSR) [19][20][21].Recent studies mapped additional neutralizing B cell epitopes to the junction of the N-terminal domain and the central repeat domain [22][23][24][25].Earlier work also demonstrated that antibodies which block proteolytic cleavage of Pf CSP at a site in the N-terminal domain prevent infection [26].These protective B cell epitopes are lacking in RTS,S and R21.Immunization with Pf CSP vaccine constructs that include the N-terminus, a portion of the central repeat, and the C-terminus of Pf CSP may drive IgG responses to a broader set of conserved B cell epitopes and increase efficacy.At the same time, earlier data demonstrating that the N-terminus of Pf CSP can fold to mask the adhesive domain of the C-terminus and facilitate the exit of sporozoites out of the dermis [27,28] must also be considered.
Historically, carrier proteins have been of value in overcoming some challenges related to vaccine immunogenicity [29,30].Most commonly, these are heterologous protein carriers unrelated to the pathogen of interest.This is the case with the use of Hepatitis B surface Antigen as a carrier, fused to a truncated Pf CSP molecule in the RTS,S vaccine and more recently in the related R21 construct [19][20][21].Building on extensive in vivo immunogenicity and efficacy data with monovalent Plasmodium yoelii merozoite surface protein 1 42 (MSP1 42 ) and merozoite surface protein 8 (MSP8) vaccines and a chimeric P. yoelii MSP1 19 + MSP8 vaccine [31,32], we developed the highly conserved, highly immunogenic Pf MSP8 as a carrier for malaria vaccine delivery [33,34].By genetically fusing targeted neutralizing B cell epitopes to Pf MSP8, we overcame challenges associated with the production of recombinant antigen (rAg) vaccines (conformation, quality, yield) and suboptimal immunogenicity.We demonstrated its utility as a carrier for blood-stage vaccine candidates including Pf MSP1 and Pf MSP2, and for the structurally complex 25 kDa transmission blocking vaccine candidate, Pf s25.For Pf MSP1 19 , fusion to Pf MSP8 facilitated production of a vaccine in high yield and appropriate conformation that elicited strong CD4+ T cell help for the production of merozoite neutralizing antibodies to linked Pf MSP1 19 epitopes in outbred strains of mice, rabbits, and non-human primates [34,35].Elicited antibodies were highly cross-reactive between FVO and 3D7 alleles of Pf MSP1 19 and potently inhibited the in vitro growth of P. falciparum blood-stage parasites.For Pf MSP2 vaccines, fusion to Pf MSP8 prevented formation of Pf MSP2 fibrils and masking of B cell epitopes in Pf MSP2, which enhanced production of Pf MSP2-specific antibodies in mice and rabbits that opsonized merozoites for uptake by macrophages [36,37].Immunization with chimeric Pf MSP2/8 formulated with glucopyranosyl lipid adjuvant-stable emulsion (GLA-SE, a synthetic TLR4 agonist) as adjuvant elicited high Pf MSP2-specific antibody titers that persisted for at least 12 months.As a carrier, Pf MSP8 facilitated production of a Pf s25-based vaccine in high yield and appropriate conformation using simplified purification protocols and enhanced the immunogenicity of Pf s25 when adjuvanted with GLA-SE [38,39].Pf s25/8induced IgG (rabbit, mouse) was functional, potently reducing transmission of P. falciparum parasites to mosquitoes when measured using the standard membrane feeding assay (SMFA).Critical for the development of multivalent, multistage malaria vaccines, fusion to Pf MSP8 was required to maintain the immunogenicity of Pf MSP1 19 , Pf MSP2, and Pf s25 when formulated in combination [37].
Here, we build on knowledge gained from our own vaccine studies and from the wealth of information revealed by studies (old and new) of protective immune responses to Pf CSP.We tested the hypotheses that (i) immunization with rPf CSP-based vaccines (R/A) rPf CSP (#1) and R/A rPf CSP/8 (#2) were prepared by sequential treatment with 25 mM dithiothreitol overnight at 4 • C and 1 h at 37 • C followed by 125 mM iodoacetic acid for 1 h at 37 • C prior to dialysis as above.rPf MSP8 was purified as reported [33].

Mice and Immunizations
Five-to six-week-old male and female CD1 mice were purchased form Charles River Laboratories (Wilmington, MA, USA).All animals were housed under specific pathogenfree conditions in the Animal Care Facility of Drexel University College of Medicine.Animal use protocols were reviewed, approved, and conducted in compliance with Drexel University's Institutional Animal Care and Use Committee (Protocol #20900).Six comparative immunogenicity and efficacy studies were completed.In each study, groups of outbred CD1 mice (n = 10, five males, five females) were immunized s.c. with 2.5 µg/dose rPf CSPbased vaccine formulated with 5 µg/dose of GLA-SE (glucopyranosyl lipid adjuvant, a synthetic TLR4 agonist, in a stable squalene-in-water emulsion, Access to Advanced Health Institute, Seattle, WA, USA) [43,44].Control groups were immunized with GLA-SE alone.Four weeks following the primary immunization, mice received a second immunization with the same dose of antigen and adjuvant.
IgG Avidity: The avidity of antigen-specific antibodies was estimated by ELISA based on the ability of antibodies to remain bound to rPf CSP (#1) in the presence of increasing concentrations of ammonium thiocyanate (0.5 M to 4.0 M) as previously described [37,45].Serum from each animal was diluted to yield an OD 405 of ~1.0 in the absence of ammonium thiocyanate.OD 405 was plotted against increasing concentrations of ammonium thiocyanate.The line of best fit equation was used to determine an avidity index (AI) for each serum sample based on the concentration of ammonium thiocyanate needed to reduce the amount of bound antibody by 50%.
IgG subclass: To assess the profile of IgG subclasses induced by immunizations, serum samples were titered on ELISA plates coated with rPf CSP (#1) as previously described [31,37,38].Bound antibodies were detected with HRP-conjugated goat antibodies specific for mouse IgG1, IgG2a, IgG2b, IgG2c, and IgG3 (Southern Biotech, Birmingham, AL, USA) with ABTS as substrate.Each plate included a standard curve of purified IgG subclass-specific mouse myeloma proteins.IgG subclass concentration was expressed as units/mL where 1 U/mL is equivalent to 1 µg/mL myeloma standard.
Epitope specificity: A set of 37 N-terminally biotinylated Pf CSP peptides (15-mers, 8 amino acid overlap) spanning the length of rPf CSP (#1) was synthesized (Genscript Biotech Corp., Piscataway, NJ, USA) (Table S1).Individual peptides (0.25 µg/well) were captured on streptavidin-coated ELISA plates (Pierce TM , ThermoFisher Scientific).Mouse sera were assayed at a dilution of 1:1000.Bound antibody was detected with rabbit anti-mouse IgG (H+L)-HRP and ABTS as substrate.On each plate, serial dilutions of a high-titer pool of sera from rPf CSP/8 (#2)-immunized mice were assayed on peptide #22 (3 NANP core repeats) and used for normalization between plates.Immunization sera were considered reactive if the mean OD 405 in peptide coated wells was greater than mean OD 405 + 3 standard deviations of GLA-SE control sera.Following an initial screening assay with pooled sera from immunization studies #1 and #2 (above), 13 informative peptides spanning the N-terminal domain, junctional region, major and minor repeat region, and the C-terminus of Pf CSP were selected to assess the epitope specificity of antibodies in individual sera from immunized mice.

Analysis of Cellular Responses
Four weeks following the second immunization, spleens were harvested, and singlecell suspensions were prepared as described [31,37,38].To assess antigen-specific cytokine production, splenocytes were stimulated with peptide pools covering the C-terminal domain of Pf CSP (3 pools, 6 peptides per pool, 3 µg/mL of each peptide per pool), or with rPf CSP (#1) or rPf MSP8 at 10 µg/mL for 96 h.Concanavalin A (1 µg/mL) and cells alone were used as positive and negative controls, respectively.Cell-free culture supernatants were collected and stored at −80 • C until analysis.IFNγ released by antigen-stimulated T cells was measured using the BD OptEIA Mouse IFNγ ELISA kit according to the manufacturer's protocol (BD Biosciences, Inc., San Diego, CA, USA).Antigen-induced IFNγ was quantified based on a standard curve run on each assay plate.

Vaccine Efficacy
To test vaccine efficacy, mice were challenged with transgenic P. yoelii 17X sporozoites engineered to express P. falciparum CSP from the endogenous locus [46] (provided by Dr. Stephen Hoffman, Sanaria, Inc., Rockville, MD, USA).Cryopreserved sporozoites stored in liquid nitrogen vapor phase were quickly thawed, diluted in RPMI 1640 containing 20 mM MOPS, pH 7.4, and 1% pooled normal mouse sera.Immunized and control mice were infected by i.v.injection of 6250 sporozoites (100 µL) via the retro-orbital plexus.To determine sterile protection, infections were monitored daily for 14 days by microscopy of thin tail-blood smears stained with Giemsa.Mice were considered infected on the day in which two or more blood-stage parasites in 50 fields (1000×) were detected, with progression of infection observed on subsequent days.

Statistical Analysis
Statistical analysis utilized Graph Pad Prism, version 10.1.2(GraphPad Software Inc., Boston, MA, USA).For direct comparison of cytokine and antibody responses of unrelated groups, the Mann-Whitney non-parametric test was used.For comparison of antibody responses in more than two groups, the Kruskal-Wallis test was used followed by Dunn's test to correct for multiple comparisons.For analysis of boosting and the durability of antibody responses using paired serum samples, the Wilcoxon matched pairs signedrank test was used.Differences in sterile protection induced by Pf CSP-based vaccines were determined using the Mantel-Cox log rank test considering time to appearance of blood-stage parasites and the number of animals remaining uninfected.For all analyses, p values < 0.05 were considered statistically significant.

Expression and Purification of rPfCSP-Based Vaccine Antigens
As diagrammed in Figure 1, four recombinant Pf CSP-based proteins were expressed and purified using an E. coli expression system.rPf CSP (#1) consists of the N-terminal domain, 19 NVDP/NANP central repeats, and the C-terminal domain of Pf CSP.rPf CSP/8 (#2) is a chimeric protein containing rPf CSP (#1) fused to the rPf MSP8 carrier protein.rPf CSPN/8 (#3) contains the N-terminal domain and the first nine tetra-amino acid repeats of the central domain, fused to Pf MSP8.Finally, rPf CSPC/8 (#4) consists of the last 10 tetra-amino acid repeats of the central domain followed by the C-terminal domain, fused to Pf MSP8.Recombinant proteins were expressed, purified by nickel chelate affinity chromatography, and analyzed by SDS-PAGE on 10% polyacrylamide gels run under both non-reducing and reducing conditions that were then stained with Coomassie blue (Figure 2A).Under both non-reducing and reducing conditions, rPf CSP (#1) migrated as a doublet of approximately 38-39 kDa with several low molecular weight bands present in low concentration.The full-length product appeared larger than its predicted molecular weight.This is consistent with the aberrant migration of native Pf CSP, which has been attributed to the large, central NANP repeat domain.rPf CSP/8 (#2), rPf CSPN/8 (#3), and rPf CSPC/8 (#4) migrated as distinct bands at ~76, ~61, and ~60 kDa, respectively, close to their predicted molecular weights.
responses in more than two groups, the Kruskal-Wallis test was used followed by Dunn's test to correct for multiple comparisons.For analysis of boosting and the durability of antibody responses using paired serum samples, the Wilcoxon matched pairs signed-rank test was used.Differences in sterile protection induced by PfCSP-based vaccines were determined using the Mantel-Cox log rank test considering time to appearance of bloodstage parasites and the number of animals remaining uninfected.For all analyses, p values < 0.05 were considered statistically significant.

Expression and Purification of rPfCSP-Based Vaccine Antigens
As diagrammed in Figure 1, four recombinant PfCSP-based proteins were expressed and purified using an E. coli expression system.rPfCSP (#1) consists of the N-terminal domain, 19 NVDP/NANP central repeats, and the C-terminal domain of PfCSP.rPfCSP/8 (#2) is a chimeric protein containing rPfCSP (#1) fused to the rPfMSP8 carrier protein.rPfCSPN/8 (#3) contains the N-terminal domain and the first nine tetra-amino acid repeats of the central domain, fused to PfMSP8.Finally, rPfCSPC/8 (#4) consists of the last 10 tetraamino acid repeats of the central domain followed by the C-terminal domain, fused to PfMSP8.Recombinant proteins were expressed, purified by nickel chelate affinity chromatography, and analyzed by SDS-PAGE on 10% polyacrylamide gels run under both non-reducing and reducing conditions that were then stained with Coomassie blue (Figure 2A).Under both non-reducing and reducing conditions, rPfCSP (#1) migrated as a doublet of approximately 38-39 kDa with several low molecular weight bands present in low concentration.The full-length product appeared larger than its predicted molecular weight.This is consistent with the aberrant migration of native PfCSP, which has been attributed to the large, central NANP repeat domain.rPfCSP/8 (#2), rPfCSPN/8 (#3), and rPfCSPC/8 (#4) migrated as distinct bands at ~76, ~61, and ~60 kDa, respectively, close to their predicted molecular weights.Native Pf CSP contains two disulfide bonds in the C-terminal domain while the rPf MSP8 carrier contains a C-terminal double EGF-like domain.The lack of high molecular weight aggregates on gels run under non-reducing conditions suggested that disulfide bonds in the rPf CSP-based antigens formed correctly.To confirm the folding of the Cterminal domain of Pf CSP, vaccine antigens were separated by SDS-PAGE, run under non-reducing conditions, and analyzed by immunoblot with mAb 4B3, which recognizes a disulfide-dependent B cell epitope of Pf CSP.As shown in Figure 2B, mAb 4B3 was highly reactive with rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPC/8 (#4), which all contain the C-terminal domain of Pf CSP, but not with rPf CSPN/8 (#3), which lacks this epitope.Comparable results were obtained with mAb 4C2 (not shown).Furthermore, reduction and alkylation of rPf CSP (#1) to disrupt the folding of this domain eliminated reactivity with mAb 4B3.As such, folding of the rPf CSP-based antigens mimicked the native protein and the folding of the C-terminal domain of Pf CSP was not altered by fusion to the Pf MSP8 carrier.Native PfCSP contains two disulfide bonds in the C-terminal domain while rPfMSP8 carrier contains a C-terminal double EGF-like domain.The lack of high mol lar weight aggregates on gels run under non-reducing conditions suggested that disul bonds in the rPfCSP-based antigens formed correctly.To confirm the folding of th terminal domain of PfCSP, vaccine antigens were separated by SDS-PAGE, run un non-reducing conditions, and analyzed by immunoblot with mAb 4B3, which recogn a disulfide-dependent B cell epitope of PfCSP.As shown in Figure 2B, mAb 4B3 highly reactive with rPfCSP (#1), rPfCSP/8 (#2), and rPfCSPC/8 (#4), which all contain C-terminal domain of PfCSP, but not with rPfCSPN/8 (#3), which lacks this epitope.C parable results were obtained with mAb 4C2 (not shown).Furthermore, reduction alkylation of rPfCSP (#1) to disrupt the folding of this domain eliminated reactivity w mAb 4B3.As such, folding of the rPfCSP-based antigens mimicked the native protein the folding of the C-terminal domain of PfCSP was not altered by fusion to the PfM carrier.

rPfCSP-Based Vaccines Induce High Titers of anti-PfCSP IgG of Multiple Subclasses
Use of PfMSP8 as a carrier has facilitated expression and purification of target a gens but has also provided additional T cell help to enhance antibody responses to tective, but subdominant B cell epitopes [34,[36][37][38][39].In Study #1, the immunogenicit rPfCSP (#1) and rPfCSP/8 (#2) was compared to evaluate the impact of PfMSP8 on overall antibody response to near full-length rPfCSP.In Study #2, the immunogenicit rPfCSPN/8 (#3) and rPfCSPC/8 (#4) was compared to determine if the separation of th terminal and C-terminal domains positively or negatively impacts immunogenicity.rPfCSPC/8 (#4) is closest in design to the fusion protein present in the RTS,S and PfCSP-based vaccines but with the carrier protein switched from HBsAg to PfMSP8.G SE was selected as the adjuvant for these studies based on its overall robust perform when used to formulate rPfMSP1/8, rPfMSP2/8, and rPfs25/8 vaccines [37,38].Outb CD1 mice (five males, five females per group) were immunized and boosted with µg/dose of vaccine antigens at a 4-week interval.

rPfCSP-Based Vaccines Induce High Titers of Anti-PfCSP IgG of Multiple Subclasses
Use of Pf MSP8 as a carrier has facilitated expression and purification of target antigens but has also provided additional T cell help to enhance antibody responses to protective, but subdominant B cell epitopes [34,[36][37][38][39].In Study #1, the immunogenicity of rPf CSP (#1) and rPf CSP/8 (#2) was compared to evaluate the impact of Pf MSP8 on the overall antibody response to near full-length rPf CSP.In Study #2, the immunogenicity of rPf CSPN/8 (#3) and rPf CSPC/8 (#4) was compared to determine if the separation of the N-terminal and C-terminal domains positively or negatively impacts immunogenicity.The rPf CSPC/8 (#4) is closest in design to the fusion protein present in the RTS,S and R21 Pf CSP-based vaccines but with the carrier protein switched from HBsAg to Pf MSP8.GLA-SE was selected as the adjuvant for these studies based on its overall robust performance when used to formulate rPf MSP1/8, rPf MSP2/8, and rPf s25/8 vaccines [37,38].Outbred CD1 mice (five males, five females per group) were immunized and boosted with 2.5 µg/dose of vaccine antigens at a 4-week interval.
To investigate any potential sex-based differences in vaccine-induced immune responses, Pf CSP-specific secondary sera titers were stratified by sex (female, open symbols; male, closed symbols) within each immunization group.There were no significant differences in Pf CSP-specific antibody titers in secondary sera of female vs. male mice immunized with rPf CSP (#1), rPf CSP/8 (#2), or rPf CSPN/8 (#3).However, female mice immunized with rPf CSP/C (#4) had somewhat higher anti-Pf CSP-specific antibody titers than male mice in the same group (Figure 3A, Study #2, p = 0.02).These data indicate that all formulations were immunogenic; inducing high-titer, boostable anti-Pf CSP antibody responses; and that inclusion of rPf MSP8 as carrier did not inhibit or increase the magnitude of the anti-Pf CSP antibody response to a nearly full-length construct (#1 vs. #2).With the truncated constructs, however, the immunogenicity of rPf CSPC/8 (#4) was significantly greater than that of rPf CSPN/8 (#3).
To begin to assess the quality of vaccine-induced antibody responses, the avidity and IgG subclass distribution of anti-Pf CSP-specific antibodies induced by each construct were determined.Avidity indices were calculated based on the resistance of antigen-antibody complexes to disruption by increasing concentrations of ammonium thiocyanate.As shown in Figure 3B (left panel), the avidity of anti-Pf CSP-specific antibodies induced by immunization with rPf CSP (#1) vs. rPf CSP/8 (#2) was comparable (Study #1, p = 0.11, ns).As shown in Figure 3B (right panel), there was a trend for slightly higher avidity antibodies Secondary sera from immunized mice were analyzed for production of IgG1, IgG2a/c, IgG2b, and IgG3 subtypes via ELISA using rPf CSP (#1)-coated wells and quantitated against standard curves of myeloma proteins of each IgG subclass.In general, the subclass distribution reflected a balanced production of antigen-specific antibodies of the IgG1, IgG2a/c, and IgG2b subclasses with more limited production of IgG3 antibodies (Figure 3C).This is consistent with the production of both Th1 and Th2 cytokines characteristic of vaccines formulated with GLA-SE as adjuvant [47][48][49].Mice immunized with rPf CSP/8 (#2) produced slightly more IgG1 and IgG3 antibodies relative to those immunized with rPf CSP (#1) (Figure 3C, Study #1).More notable was the induction of higher levels of anti-Pf CSP IgG2a/c antibodies by immunization with rPf CSPC/8 (#4) in comparison to rPf CSPN/8 (#3) (Figure 3C, Study #2, p < 0.01).Of additional interest, some of the variability in the subclass distribution of vaccine-induced antibodies in these outbred mice could be attributed to sex.Female mice immunized with rPf CSPC/8 (#4) had higher levels of anti-Pf CSP-specific IgG1 and IgG2b antibodies than male mice in the same group (Figure 3C, Study #2, p < 0.02).While small differences were noted, the overall similarity in avidity and IgG subclass profiles suggest that anti-Pf CSP-specific antibodies induced by each of the four Pf CSP-based vaccine constructs were comparable in quality.

rPfCSP-Based Vaccines Induce T Cell Responses of Varied Specificity
In Immunization Study #3, outbred CD1 mice (five males, five females per group) were immunized and boosted as above with rPf CSP(#1), rPf CSP/8 (#2), rPf CSPN/8 (#3), or rPf CSPC/8 (#4) formulated with GLA-SE as adjuvant, or with GLA-SE alone.Analysis of vaccine-induced anti-Pf CSP antibody responses confirmed the results of Study #1 and Study #2.Immunization with rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPC/8 (#4) elicited high and comparable titers of Pf CSP-specific IgG while the response induced by the rPf CSPN/8 (#3) vaccine construct was significantly lower (Supplemental Figure S1, panel A).As antigen-specific T cell responses have been shown to contribute to protection against preerythrocytic stage malaria parasites [50], the magnitude and specificity of vaccine-induced T cell responses was measured, using antigen-specific IFNγ production as a readout.Splenocytes were harvested from mice immunized with each of the four vaccine constructs and stimulated with rPf CSP (#1) or rPf MSP8, representing the constituent domains of the chimeric vaccines, or with pools of 15-mer overlapping synthetic peptides spanning the central repeat region and C-terminal domains of Pf CSP.IFNγ released into culture supernatants was quantified by ELISA.
Immunization with rPf CSP (#1) and rPf CSPC/8 (#4) elicited Pf CSP-specific T cell responses, consistent with the presence of known CD4+ T cell epitopes in the C-terminal domain of Pf CSP [51,52] (Figure 4A).This was confirmed upon stimulation of splenocytes with a pool of six Pf CSP C-terminal peptides (Figure 4B, Table S1, peptides #31, 33-37).In both assays, the range in concentration of IFNγ secreted was fairly large, likely reflecting the genetic heterogeneity of outbred CD1 mice.As expected, Pf CSP-specific T cell responses elicited by immunization with rPf CSPN/8 (#3) were minimal, due to the absence of the C-terminal Pf CSP T cell epitopes in this construct.Somewhat unexpected was the rather low Pf CSP-specific T cell responses elicited by immunization with rPf CSP/8 (#2), which contains all of the same Pf CSP T cell epitopes found in rPf CSP (#1) and rPf CSPC/8 (#4).However, immunization with rPf CSP/8 (#2), as well as rPf CSPN/8 (#3) and rPf CSPC/8 (#4), elicited strong T cell responses to the rPf MSP8 carrier that were relatively consistent among individual animals across the three groups (Figure 4C).There were no significant sex-related differences in antigen-specific T cell responses with the exception that female mice immunized with rPf CSPN/8 (#3) responded to rPf MSP8 better than male mice (p = 0.03).These data show that immunization with the rPf CSP-based vaccines induce strong T cell responses, but that fusion of nearly full-length Pf CSP to Pf MSP8 shifted the specificity of the response away from Pf CSP-specific epitopes resulting in a more dominant Pf MSP8-specific response.
(#4), elicited strong T cell responses to the rPfMSP8 carrier that were relatively consistent among individual animals across the three groups (Figure 4C).There were no significant sex-related differences in antigen-specific T cell responses with the exception that female mice immunized with rPfCSPN/8 (#3) responded to rPfMSP8 better than male mice (p = 0.03).These data show that immunization with the rPfCSP-based vaccines induce strong T cell responses, but that fusion of nearly full-length PfCSP to PfMSP8 shifted the specificity of the response away from PfCSP-specific epitopes resulting in a more dominant PfMSP8-specific response.

Immunization with Chimeric rPfCSP/8 (#2) Protects against Sporozoite Challenge Infection
In Immunization Study #4, the efficacy of the four rPfCSP-based vaccines was assessed.Outbred CD1 mice (five males, five females per group) were immunized and boosted as above with rPfCSP (#1), rPfCSP/8 (#2), rPfCSPN/8 (#3), and rPfCSPC/8 (#4) formulated with GLA-SE as adjuvant, or with GLA-SE alone.Furthermore, a group of mice immunized with the rPfMSP8 carrier only, formulated with GLA-SE as adjuvant, was added to the efficacy study to assess any carrier-dependent effects.Analysis of pre-challenge antibody responses again demonstrated reproducibly high levels of anti-PfCSP-specific IgG with three of the four PfCSP-based vaccine constructs (Supplemental Figure S1, panel B).Three weeks following the second immunization, mice were infected by i.v.injection of 6250 transgenic P. yoelii 17X sporozoites engineered to express P. falciparum CSP from the endogenous locus [46].Sterile protection was assessed by monitoring Giemsastained thin tail-blood smears for the presence of blood-stage parasites for 14 days.
As shown in Figure 5 and Table 1, control mice immunized with GLA-SE alone (9/10) or with the recombinant PfMSP8 carrier (9/10) developed blood-stage malaria within the first 9 days post-challenge.Mice immunized with rPfCSP/8 (#2) showed high and significant protection compared to the control groups, with sterile protection in seven out of nine animals.Despite the induction of high titers of anti-PfCSP antibodies, sterile protection in mice immunized with rPfCSP (#1) or with the truncated rPfCSPC/8 (#4) was limited and not significantly different than controls.No significant protection was observed in animals immunized with rPfCSPN/8 (#3), consistent with lower overall PfCSP-specific T and B cell responses.These data suggest that in addition to the major NANP containing repeat domain and the C-terminus of PfCSP, inclusion of the N-terminal domain, junctional region, and/or minor repeat domains can improve vaccine efficacy.Although the

Immunization with Chimeric rPfCSP/8 (#2) Protects against Sporozoite Challenge Infection
In Immunization Study #4, the efficacy of the four rPf CSP-based vaccines was assessed.Outbred CD1 mice (five males, five females per group) were immunized and boosted as above with rPf CSP (#1), rPf CSP/8 (#2), rPf CSPN/8 (#3), and rPf CSPC/8 (#4) formulated with GLA-SE as adjuvant, or with GLA-SE alone.Furthermore, a group of mice immunized with the rPf MSP8 carrier only, formulated with GLA-SE as adjuvant, was added to the efficacy study to assess any carrier-dependent effects.Analysis of pre-challenge antibody responses again demonstrated reproducibly high levels of anti-Pf CSP-specific IgG with three of the four Pf CSP-based vaccine constructs (Supplemental Figure S1, panel B).Three weeks following the second immunization, mice were infected by i.v.injection of 6250 transgenic P. yoelii 17X sporozoites engineered to express P. falciparum CSP from the endogenous locus [46].Sterile protection was assessed by monitoring Giemsa-stained thin tail-blood smears for the presence of blood-stage parasites for 14 days.
As shown in Figure 5 and Table 1, control mice immunized with GLA-SE alone (9/10) or with the recombinant Pf MSP8 carrier (9/10) developed blood-stage malaria within the first 9 days post-challenge.Mice immunized with rPf CSP/8 (#2) showed high and significant protection compared to the control groups, with sterile protection in seven out of nine animals.Despite the induction of high titers of anti-Pf CSP antibodies, sterile protection in mice immunized with rPf CSP (#1) or with the truncated rPf CSPC/8 (#4) was limited and not significantly different than controls.No significant protection was observed in animals immunized with rPf CSPN/8 (#3), consistent with lower overall Pf CSP-specific T and B cell responses.These data suggest that in addition to the major NANP containing repeat domain and the C-terminus of Pf CSP, inclusion of the N-terminal domain, junctional region, and/or minor repeat domains can improve vaccine efficacy.Although the enhanced efficacy depended on fusion to the Pf MSP8 carrier, the protection could not be correlated with anti-Pf CSP specific antibody titers.
enhanced efficacy depended on fusion to the PfMSP8 carrier, the protection could not be correlated with anti-PfCSP specific antibody titers.

Immunization with PfCSP-Based Vaccines Drives a Strong Antibody Response to Conformational B Cell Epitopes
The C-terminal domain of PfCSP contains two disulfide bonds.mAbs that recognize conformation-dependent epitopes in this C-terminal domain of PfCSP can neutralize P. falciparum sporozoites.To assess the induction of antibodies to linear versus conformational epitopes of PfCSP, sera from mice immunized with each of the four rPfCSP-based vaccines were concurrently assayed by ELISA on wells coated with rPfCSP (#1) or fully reduced and alkylated (R/A) rPfCSP (#1) that lacked disulfide-dependent B cell epitopes.The proportion of antibodies in each sera recognizing linear B cell epitopes was expressed as the ratio of the titer on R/A rPfCSP (#1) divided by the titer on intact rPfCSP (#1).Ratios using sera collected from Immunization Studies #1 through #4 are summarized in Figure 6 (n = 27-30/group).

Immunization with PfCSP-Based Vaccines Drives a Strong Antibody Response to Conformational B Cell Epitopes
The C-terminal domain of Pf CSP contains two disulfide bonds.mAbs that recognize conformation-dependent epitopes in this C-terminal domain of Pf CSP can neutralize P. falciparum sporozoites.To assess the induction of antibodies to linear versus conformational epitopes of Pf CSP, sera from mice immunized with each of the four rPf CSP-based vaccines were concurrently assayed by ELISA on wells coated with rPf CSP (#1) or fully reduced and alkylated (R/A) rPf CSP (#1) that lacked disulfide-dependent B cell epitopes.The proportion of antibodies in each sera recognizing linear B cell epitopes was expressed as the ratio of the titer on R/A rPf CSP (#1) divided by the titer on intact rPf CSP (#1).Ratios using sera collected from Immunization Studies #1 through #4 are summarized in Figure 6 (n = 27-30/group).
As shown in Figure 6, animals immunized with rPf CSPN/8 (#3) produced antibodies that recognized linear epitopes of Pf CSP with the R/A rPf CSP to rPf CSP antibody response ratio near 1 (0.94 ± 0.18).This was as expected, as rPf CSPN/8 (#3) lacks the C-terminal, conformational domain of Pf CSP.Unexpectedly, a considerable proportion of the antibody responses induced by immunization with rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPC/8 (#4) targeted conformational B cell epitopes.The mean R/A rPf CSP to rPf CSP antibody response ratio in rPf CSP (#1)-immunized animals was 0.41 ± 0.14 and comparable to ratios observed in rPf CSP/8 (#2)-and rPf CSPC/8 (#4)-immunized mice (0.48 ± 0.12 and 0.38 ± 0.18, respectively).As such, fusion of Pf CSP sequences to the Pf MSP8 carrier did not impact antibody responsiveness to the C-terminus of Pf CSP.Stratifying these R/A Pf CSP to Pf CSP ratios by sex revealed that there were no significant differences between  7).Sera, assayed at a dilution of 1:1000, were considered reactive (# peptides recognized, y-axis) if the mean OD405 of test sera on a given peptide was greater than mean OD405 + 3 standard deviations of GLA-SE control sera.Significance of differences in mean number (±SD) of peptides recognized by mice in each immunization group was assessed using the Kruskal-Wallis test followed by Dunn's correction for multiple comparisons.* p < 0.05.
As shown in Figure 6, animals immunized with rPfCSPN/8 (#3) produced antibodies that recognized linear epitopes of PfCSP with the R/A rPfCSP to rPfCSP antibody response ratio near 1 (0.94 ± 0.18).This was as expected, as rPfCSPN/8 (#3) lacks the C-terminal, conformational domain of PfCSP.Unexpectedly, a considerable proportion of the antibody responses induced by immunization with rPfCSP (#1), rPfCSP/8 (#2), and rPfCSPC/8 (#4) targeted conformational B cell epitopes.The mean R/A rPfCSP to rPfCSP antibody response ratio in rPfCSP (#1)-immunized animals was 0.41 ± 0.14 and comparable to ratios observed in rPfCSP/8 (#2)-and rPfCSPC/8 (#4)-immunized mice (0.48 ± 0.12 and 0.38 ± 0.18, respectively).As such, fusion of PfCSP sequences to the PfMSP8 carrier did not impact antibody responsiveness to the C-terminus of PfCSP.Stratifying these R/A PfCSP to PfCSP ratios by sex revealed that there were no significant differences between males and females within each group.Overall, these data indicate that approximately half of the antibodies induced by three of four rPfCSP-based vaccine constructs recognize disulfidedependent epitopes in the C-terminal domain of PfCSP.

Variability in the Recognition of Linear B Cell Epitopes by Antibodies Induced by Immunization with PfCSP-Based Vaccines
To assess the fine specificity of vaccine-induced antibodies against linear B cell epitopes of PfCSP, ELISAs were conducted using 37 overlapping 15-mer biotinylated peptides, spanning the length of the rPfCSP (#1) construct (Table S1).For the initial screen, pools of sera from male and female mice immunized with rPfCSP (#1), rPfCSP/8 (#2), rPfCSPN/8 (#3), rPfCSPC/8 (#4), or GLA-SE were screened for reactivity with each peptide.As shown in Figure S2, multiple peptides from the N-terminal junctional region, minor repeat region, the central repeat domain, and to a lesser degree, the C-terminal domain  7).Sera, assayed at a dilution of 1:1000, were considered reactive (# peptides recognized, y-axis) if the mean OD 405 of test sera on a given peptide was greater than mean OD 405 + 3 standard deviations of GLA-SE control sera.Significance of differences in mean number (±SD) of peptides recognized by mice in each immunization group was assessed using the Kruskal-Wallis test followed by Dunn's correction for multiple comparisons.* p < 0.05.

Variability in the Recognition of Linear B Cell Epitopes by Antibodies Induced by Immunization with PfCSP-Based Vaccines
To assess the fine specificity of vaccine-induced antibodies against linear B cell epitopes of Pf CSP, ELISAs were conducted using 37 overlapping 15-mer biotinylated peptides, spanning the length of the rPf CSP (#1) construct (Table S1).For the initial screen, pools of sera from male and female mice immunized with rPf CSP (#1), rPf CSP/8 (#2), rPf CSPN/8 (#3), rPf CSPC/8 (#4), or GLA-SE were screened for reactivity with each peptide.As shown in Figure S2, multiple peptides from the N-terminal junctional region, minor repeat region, the central repeat domain, and to a lesser degree, the C-terminal domain were recognized by sera from immunized mice.Based on these data, a set of 13 informative peptides (Figure 7A) were selected to investigate the diversity of the anti-Pf CSP antibody responses in individual sera obtained from Immunization Studies #1 through #4 (n = 27-30/group).
rPf CSPC/8 (#4) is the construct most similar to RTS,S and R21 vaccines.As shown in Figure 7E, rPf CSPC/8 (#4) induced antibody responses that bound NANP repeat containing peptides in 60-90% of animals (i.e., peptides 20,21,22,23).The limited recognition of peptides derived from the C-terminal regions was expected, given the conformational nature of this domain.As shown in Figure 7D, immunization with rPf CSPN/8 (#3) shifted the response away from NANP-containing epitopes to some degree with increased recognition of linear epitopes associated with the junctional region and minor-repeat regions (i.e., peptides 14,15,16,18,19).Importantly, immunization with the near full-length rPf CSP (#1) increased the breadth of linear B cell epitopes recognized to include peptides derived from the N-terminal domain, the junctional region, the minor NVDP repeat regions, as well as the central NANP repeat domain (Figure 7B).Fusion of the nearly full-length Pf CSP sequence to the Pf MSP8 carrier did not further increase the diversity of epitopes recognized.However, rPf CSPC/8 (#4) immunization increased the overall percentage of animals seroreactive to these linear epitopes with 70-100% of animals recognizing multiple peptides.
Vaccines 2024, 12, x FOR PEER REVIEW 13 of 22 were recognized by sera from immunized mice.Based on these data, a set of 13 informative peptides (Figure 7A) were selected to investigate the diversity of the anti-PfCSP antibody responses in individual sera obtained from Immunization Studies #1 through #4 (n = 27-30/group).rPfCSPC/8 (#4) is the construct most similar to RTS,S and R21 vaccines.As shown in Figure 7E, rPfCSPC/8 (#4) induced antibody responses that bound NANP repeat containing peptides in 60-90% of animals (i.e., peptides 20,21,22,23).The limited recognition of peptides derived from the C-terminal regions was expected, given the conformational nature of this domain.As shown in Figure 7D, immunization with rPfCSPN/8 (#3) shifted the response away from NANP-containing epitopes to some degree with increased recognition of linear epitopes associated with the junctional region and minor-repeat regions (i.e., peptides 14,15,16,18,19).Importantly, immunization with the near full-length rPfCSP (#1) increased the breadth of linear B cell epitopes recognized to include peptides To quantitate differences in breadth of linear B cell epitopes recognized, the number of peptides each animal recognized was determined.Values were then compared across the four immunization groups.As shown in Figure 6B, animals immunized with rPf CSP/8 (#2) recognized a significantly greater number of peptides than animals immunized with either rPf CSPN/8 (#3) or rPf CSPC/8 (#4).While the mean number of peptides recognized by animals immunized with rPf CSP/8 (#2) was elevated compared to animals immunized with rPf CSP (#1), the difference was not statistically significant.Combined, these data show that addition of the N-terminal domain, junctional region, and minor-repeat regions to the RTS,S-like rPf CSPC/8 (#4) construct expands the breadth of antibody responses to include novel, potentially protective, linear B cell epitopes.The magnitude of the antibody response to conformation-dependent epitopes elicited by immunization with rPf CSP (#1) or rPf CSP/8 (#2) was striking.In Immunization Study #5, the contribution of these antibodies to vaccine-induced protection was evaluated.Outbred CD1 mice (five males, five females per group) were immunized and boosted with rPf CSP (#1) and rPf CSP/8 (#2) as above or with rPf CSP (#1) and rPf CSP/8 (#2) that had been reduced and alkylated to disrupt disulfide bonds required for proper conformation of C-terminus of Pf CSP.All vaccines were formulated with GLA-SE as adjuvant.As expected, immunization with rPf CSP (#1) or rPf CSP/8 (#2) elicited high titers of anti-Pf CSP antibodies (Figure 8A), a sizable proportion of which recognized conformational epitopes (Figure 8B).In contrast, anti-Pf CSP antibody responses induced by immunization with R/A rPf CSP (#1) or with R/A rPf CSP/8 (#2) were decreased relative to the corresponding folded antigens (Figure 8A), and highly restricted to recognition of linear determinants (Figure 8B).To assess impact on efficacy, immunized mice were then infected with transgenic P. yoelii 17X sporozoites expressing P. falciparum CSP and sterile protection monitored as above.As shown in Figure 8 and Table 2, the lack of antibodies directed against C-terminal conformational B cell epitopes of Pf CSP did not alter vaccine-induced protection.

Immunization with PfCSP-Based Vaccines Drives Strong, Durable Antibody Responses
In the final Immunization Study #6, the durability of the antibody response induced by the four rPf CSP-based vaccines was evaluated.Outbred CD1 mice (five males, five females per group) were immunized and boosted as above with rPf CSP (#1), rPf CSP/8 (#2), rPf CSPN/8 (#3), or rPf CSPC/8 (#4).Following the boost, serum samples were collected once per month for a period of 6 months and anti-Pf CSP antibody titers in each test bleed were determined by ELISA. Figure 9A shows that immunization with three of the four constructs induced high titers of anti-Pf CSP antibodies.As expected, the anti-CSP antibody titers dropped significantly from month one (M1) to month six (M6).However, except for animals immunized with rPf CSPN/8 (#3), the overall anti-Pf CSP titer was relatively stable, only dropping by ~twofold during the 6 month follow up period.In contrast, anti-Pf CSP antibodies induced by rPf CSPN/8 (#3) immunization dropped markedly, with ~sixfold reduction in titer over the 6 month period.To determine if there was any preferential drop in antibodies that recognized linear versus conformational epitopes, the R/A rPf CSP to rPf CSP antibody response ratio was again determined.Mice immunized with rPf CSPC/8 (#4) had a small but significant reduction in the R/A rPf CSP vs. rPf CSP response ratio when comparing M1 to M6, indicative of a preferential drop in antibodies directed toward linear epitopes (Figure 9B).However, this was not observed in animals immunized with either rPf CSP (#1) or rPf CSP/8 (#2).Overall, these data show that three of the four Pf CSP-based vaccines induce high and durable anti-Pf CSP antibody responses in mice.rPfCSPC/8 (#4) had a small but significant reduction in the R/A rPfCSP vs. rPfCSP response ratio when comparing M1 to M6, indicative of a preferential drop in antibodies directed toward linear epitopes (Figure 9B).However, this was not observed in animals immunized with either rPfCSP (#1) or rPfCSP/8 (#2).Overall, these data show that three of the four PfCSP-based vaccines induce high and durable anti-PfCSP antibody responses in mice.

Discussion
The goal of this study was to consider changes in the design of the pre-erythrocyticstage RTS,S [19,21] and R21 [20] malaria vaccines that could potentially improve efficacy.Both of these vaccines target the central repeat and C-terminal domains of Pf CSP and induce antibodies to the central repeat, which are critical for protection [9,14,15].With this in mind, we first evaluated novel Pf CSP-based constructs designed to include epitopes outside of major NANP repeat epitope in an effort to improve the diversity of B cell epitopes recognized.Second, we evaluated a promising malaria-specific carrier protein, Pf MSP8, as a substitute for the heterologous HBsAg carrier.In general, we focused our comparisons on the immunogenicity and efficacy of rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPN/8 (#3) relative to rPf CSPC/8 (#4), which is most similar in design to R21 and the Pf CSP-specific component of RTS,S.With all four constructs, the avidity and IgG subclass profiles of vaccine-induced antibodies were similar and may reflect the use of GLA-SE as adjuvant for all formulations [37,38,[47][48][49].We noted differences in the magnitude, specificity, durability, and efficacy of vaccine-induced responses.For the most part, these differences in immunogenicity and efficacy were independent of sex and directly related to construct design.
We achieved success in increasing the diversity of B cell epitopes recognized upon immunization by inclusion of the N-terminal domain, the junctional region, and the minor repeats with rPf CSP (#1) and rPf CSP/8 (#2).Antibodies from most animals immunized with rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPC/8 (#4) strongly recognized peptides containing the immunodominant NANP repeat epitope.In addition, animals immunized with near full-length rPf CSP (#1) and rPf CSP/8 (#2) also recognized linear epitopes located near the junction of the N-terminal domain and central repeat region, which included peptides containing the minor NVDP repeats.Although the fine specificity of these antibodies elicited in mice may differ from neutralizing human mAbs such as CIS43 [22] and L9 [24], the data show that this region of Pf CSP is immunogenic in rPf CSP (#1) and rPf CSP/8 (#2).These epitopes were similarly recognized by mice immunized with rPf CSPN/8 (#3), which includes the same N-terminal domain and junctional region.However, the response to rPf CSPN/8 (#3) immunization was associated with a somewhat less consistent antibody response to the major NANP repeat epitope, suggesting that the inclusion of the C-terminal domain of Pf CSP may be beneficial for overall immunogenicity.The C-terminal domain of Pf CSP also contains two disulfide bonds and mAbs that recognize conformational, C-terminal epitopes neutralize P. falciparum sporozoites [42,53].In our hands, the proportion of vaccine-induced antibodies that recognized this conformationally constrained domain was expectedly high in animals immunized with rPf CSP (#1), rPf CSP/8 (#2), and rPf CSPC/8 (#4).This was demonstrated by a drop of ~50% in ELISA titers when assayed using reduced and alkylated rPf CSP (#1) vs. intact rPf CSP (#1) as the coating antigen (R/A Pf CSP-to-Pf CSP response ratio).Additionally, there was minimal recognition of linear peptide epitopes in this C-terminal region of Pf CSP across groups.
In addition to providing B cell help to enhance both the magnitude and the quality of vaccine-induced antibodies, Pf CSP-specific T cells can independently contribute to protection.Using antigen-specific IFNγ production as a metric of responsiveness, we showed that immunization with both rPf CSP (#1)-and rPf CSPC/8 (#4)-induced T cells that recognized the C-terminal domain of Pf CSP, which contains well-characterized T cell epitopes [21,51,52,54].The Pf CSP-specific T cell response in mice immunized with rPf CSPN/8 (#3) was absent, consistent with the lack of this C-terminal domain in this construct.Unexpectedly, immunization rPf CSP/8 (#2), which contains the C-terminal T cell epitopes of Pf CSP, did not induce Pf CSP-specific T cell responses.All constructs containing the Pf MSP8 carrier induced a consistent Pf MSP8-specific T cell response, similar to our previous studies with blood-stage and sexual-stage chimeric vaccines [33,34,37,38].These results suggest that fusion of a nearly full-length Pf CSP sequence to the Pf MSP8 carrier shifted the T cell response away from Pf CSP-specific epitopes toward Pf MSP8 in the rPf CSP/8 (#2) vaccine.
Immunization with the rPf CSP fused to Pf MSP8 allows for cognate recognition of T and B cell epitopes in a single protein upon immunization.Combined, our data indicate that T cell recognition of strong epitopes within the Pf MSP8 carrier is effective in providing help to Pf CSP-specific B cells, driving strong, class-switched, durable B cell/Ab responses.We previously demonstrated this with chimeric MSP1/8 vaccines in both P. yoelii and P. falciparum.The ability of Pf MSP8-specific T cells to provide help to Pf CSP-specific B cells combined with the use of the larger, nearly full-length Pf CSP component in Pf CSP/8 (#2) partially explains the increase in diversity of the B cell response.In addition, we previously showed in studies with Pf MSP2, which is known to form amyloid-like fibrils [55][56][57], that fusion to the Pf MSP8 carrier prevents fibril formation and the masking of B cell epitopes in the Pf MSP2 domain.This increased the induction of Pf MSP2-specific antibodies of greater diversity which were functional in the opsonophagocytosis of merozoites [36,37].A similar influence on the conformation of Pf CSP when fused to Pf MSP8 may also have contributed to an increase in the breadth of the B cell response.
We designed our efficacy studies to accentuate potential differences in protection when comparing our Pf CSP-based vaccines by including only two immunizations (2.5 µg/dose) followed by a robust sporozoite challenge with transgenic P. yoelii sporozoites expressing Pf CSP from the endogenous locus [24].The rPf CSP/8 (#2) vaccine provided a consistently high level of sterile protection against sporozoite challenge.This was in contrast to the corresponding rPf CSP (#1) lacking the Pf MSP8 carrier, in which protection was partial, trending toward statistical significance.This may reflect reduced variability in vaccineinduced antibody responses in these outbred mice immunized with rPf CSP/8 (#2) in comparison to rPf CSP (#1).Considering our T cell data, these results also indicate that T cell responses to epitopes in the C-terminus of Pf CSP, which were lacking in rPf CSP/8 (#2)immunized mice, are not required for protection induced by this construct.By inclusion of an additional group of animals immunized with the Pf MSP8 carrier alone, we demonstrated that protection is dependent on the Pf CSP-specific effector responses, and not the effector activity of antibodies or T cells specific for Pf MSP8.Protection induced by immunization with rPf CSPN/8 (#3) and rPf CSPC/8 (#4), which lack a full complement of potentially protective B cell epitopes, was not significantly different than rPf MSP8 and adjuvant alone control groups.
The reduction and alkylation of rPf CSP (#1) and rPf CSP/8 (#2) impacted immunogenicity, in part due to the loss of antibodies that recognize C-terminal, conformation dependent epitopes.Unexpectedly, the loss of this rather large population of Pf CSP-specific antibodies did not impact protection.At least two explanations seem plausible.First, the vaccine-induced anti-NANP repeat antibody response alone may be of sufficient magnitude to confer significant protection in this vaccine model system.Mounting data from phase II and III clinical trials with the R21 Pf CSP vaccine, lacking the N-terminal and junctional region epitopes, identified the strength of the anti-NANP repeat antibody response as a correlate of protection [14][15][16].However, this does not exclude the possibility that under circumstances when a less robust, NANP-specific response is elicited by immunization, antibodies to conformational epitopes in the C-terminus of Pf CSP may contribute to protection [58].This possibility is certainly supported by studies with mAbs [42,53] and provides a basis for the concept of passive immunization with Pf CSP-specific mAbs to non-NANP epitopes to protect against seasonal malaria [59][60][61].Alternatively, it has been reported that in vivo, the N-terminal domain of Pf CSP folds to mask the C-terminal, TSR-containing adhesive domain of Pf CSP, to facilitate exit of sporozoites from the dermis following a mosquito bite [27,28].In this in vivo setting, conformation-dependent C-terminal epitopes of Pf CSP on the surface of P. falciparum sporozoites may not be accessible for antibody binding, allowing for parasites to escape neutralization as they transit from the skin to the liver.This may in part have contributed to our inability to establish clear correlates of Pf CSP vaccine-induced, polyclonal antibody-mediated protection in this study.
Our data, considered in the framework of our immunization protocols and parameters, clearly show added value of Pf CSP-based vaccines designed to include the N-terminal domain, the central repeat region, and the C-terminal domain, with respect to maintaining a strong anti-repeat antibody response, while driving responses to additional, potentially protective B cell epitopes.Inclusion of Pf MSP8 as a vaccine carrier facilitated expression and purification but most importantly improved efficacy.Based on the data with our near fulllength Pf CSP constructs (#1, #2), a side-by-side comparison of Pf MSP8 and HBsAg as carrier may be of interest.However, we anticipate further benefit from the inclusion of Pf MSP8 as a carrier when formulating a Pf CSP-based vaccine in combination with other targets, which we firmly believe will be required for success.In this regard, we have demonstrated that immunization with rPf MSP1, rPf MSP2, and rPf s25 vaccines required fusion of the domain targeted by each vaccine to Pf MSP8 to maintain the immunogenicity of each component when formulated in a trivalent combination.Effectively adding rPf CSP/8 (#2) to this formulation has potential to achieve the goal of providing strong protection against pre-erythrocytic-stage parasites concurrent with the ability to neutralize breakthrough blood-stage parasites and protect children from life-threatening malaria.

Figure 7 .
Figure 7. Fine specificity of vaccine-induced antibodies against linear B cell epitopes of PfCSP.(A) Schematic of PfCSP (adapted from Ref. [18]) showing the approximate location of 13 synthetic peptides within the N-terminal domain, central repeat region, and C-terminal domain.Numbering is consistent with that shown in Supplemental Figure S2.Mapped B cell epitopes recognized by known neutralizing antibodies, as well as C-terminal CD4 + and CD8 + T cell epitopes, are indicated at the top of the figure.The epitope specificity of antibodies from mice immunized with (B) rPfCSP (#1), (C) rPfCSP/8 (#2), (D) rPfCSPN/8 (#3), or (E) rPfCSPC/8 (#4) (Immunization Studies #1-#4, n = 27−30 animals per group; male (blue bars), female (red bars)) were determined by ELISA on the indicated peptides.Sera, assayed at a dilution of 1:1000, were considered reactive if the mean OD405 of test sera on a given peptide was greater than mean OD405 + 3 standard deviations of GLA-SE control sera.Values are expressed as the percent of animals responding to each peptide (y-axis).

Figure 7 .
Figure 7. Fine specificity of vaccine-induced antibodies against linear B cell epitopes of Pf CSP.(A) Schematic of Pf CSP (adapted from Ref. [18]) showing the approximate location of 13 synthetic peptides within the N-terminal domain, central repeat region, and C-terminal domain.Numbering is consistent with that shown in Supplemental Figure S2.Mapped B cell epitopes recognized by known neutralizing antibodies, as well as C-terminal CD4 + and CD8 + T cell epitopes, are indicated at the top of the figure.The epitope specificity of antibodies from mice immunized with (B) rPf CSP (#1), (C) rPf CSP/8 (#2), (D) rPf CSPN/8 (#3), or (E) rPf CSPC/8 (#4) (Immunization Studies #1-#4, n = 27-30 animals per group; male (blue bars), female (red bars)) were determined by ELISA on the indicated peptides.Sera, assayed at a dilution of 1:1000, were considered reactive if the mean OD 405 of test sera on a given peptide was greater than mean OD 405 + 3 standard deviations of GLA-SE control sera.Values are expressed as the percent of animals responding to each peptide (y-axis).

3. 7 .
Antibody Recognition of Disulfide Bond-Dependent Epitopes of PfCSP Are Not Required for PfCSP-Based Vaccine-Induced Protection

Figure 8 .
Figure 8. Contribution of antibodies to conformation-dependent epitopes of Pf CSP to vaccineinduced protection.(A) Groups of CD1 mice (n = 9-10/group, both sexes) were immunized and boosted with rPf CSP (#1, circles), rPf CSP/8 (#2, triangles), or fully reduced and alkylated R/A rPf CSP (#1, squares) or R/A rPf CSP/8 (#2, diamonds).Animals immunized with rPf MSP8 or GLA-SE alone served as negative controls (hexagon).Antigen-specific IgG titers in secondary sera were measured by ELISA on wells coated with rPf CSP (#1).(B) The proportion of antibodies in each sera recognizing linear B cell epitopes was expressed as the ratio of the titer on R/A rPf CSP (#1) divided by the titer on intact rPf CSP (#1) (y-axis).Violin plots depict group mean (±SD, dashed lines).Statistical significances of difference in (A,B) were evaluated using the Kruskal-Wallis test followed by Dunn's correction for multiple comparisons.* p < 0.05.(C) Vaccine efficacy was determined upon challenge infection with 6250 transgenic P. yoelii 17X sporozoites expressing P. falciparum CSP.Sterile protection was monitored using Giemsa-stained thin tail-blood smears to detect the presence of blood-stage parasites for 14 days.Kaplan-Meyer curves represent the time to appearance of blood-stage parasites (x-axis) and the percentage of animals remaining uninfected (y-axis).Pf CSP (#1)-black dashed line; Pf CSP/8 (#2)-yellow dashed line; R/A rPf CSP (#1)-blue solid line; R/A rPf CSP/8 (#2)-red dashed line; rPf MSP8 + GLA-SE controls-black solid line.

Figure 9 .
Figure 9. Durability of Pf CSP-based vaccine-induced antibodies.Groups of CD1 mice (n = 9-10/group, both sexes) were immunized and boosted with rPf CSP (#1), rPf CSP/8 (#2), rPf CSPN/8 (#3), or rPf CSPC/8 (#4).Following the boost, serum samples were collected once per month for a period of six months.(A) Sera were assessed for antigen-specific IgG responses using ELISA plates coated with rPf CSP.Graph depicts mean (±SD) of antigen specific IgG titer with differences between month one (M1, white solid bars) and month six (M6, black solid bars) assessed by Wilcoxon matched pairs signed-rank test.Sequential shaded bars represent samples collected at months 2 through 5. Fold change in anti-Pf CSP antibody titers between M1 and M6 are indicated at the top of the figure.(B) Secondary sera obtained at M1 (filled symbols) and M6 (open symbols) were assayed by ELISA on plates coated with rPf CSP (#1) or R/A rPf CSP (#1).The proportion of antibodies in each sera recognizing linear B cell epitopes was expressed as the ratio of the titer on R/A rPf CSP divided by the titer on intact rPf CSP (y-axis) with statistical significance of differences in ratios determined the Wilcoxon matched pairs signed-rank test.* p < 0.05.
a Mantel-Cox log rank test-time to appearance of blood-stage parasites, # of animals parasite free.
a Mantel-Cox log rank test-time to appearance of blood-stage parasites, # of animals parasite free.

Table 2 .
Study #5: Sterile protection induced by immunization with rPf CSP-based vaccines.Mantel-Cox log rank test-time to appearance of blood-stage parasites, # of animals parasite free. a