PLASMODIUM ANTIBODIES

Abstract

The present invention relates to an antibody binding to a peptide comprising an amino acid sequence NANP (SEQ ID NO:1) and to at least one peptide comprising an amino acid sequence selected from NVDP (SEQ ID NO:2), NPDP (SEQ ID NO:3), and KQPA (SEQ ID NO:4), to a polynucleotide or polynucleotides encoding said antibody, and to said antibody for use in medicine and for use in prevention of Plasmodium infection. Moreover, the present invention relates to methods, kits, and devices related thereto.

Claims

1. An antibody binding to a peptide comprising an amino acid sequence NANP (SEQ ID NO:1) and to at least one peptide comprising an amino acid sequence selected from NVDP (SEQ ID NO:2), NPDP (SEQ ID NO:3), and KQPA (SEQ ID NO:4), wherein the peptide comprising the amino acid sequence NANP consists of the amino acid sequence NPNANPNANPNANPNANPNANP (SEQ ID NO:44) or NANPNANPNANPNANPNANP (SEQ ID NO:45), wherein the peptide comprising the amino acid sequence KQPA consists of the amino acid sequence KQPADGNPDPNANPN (SEQ ID NO:37), wherein the peptide comprising the amino acid sequence NPDP consists of the amino acid sequence NPDPNANPNVDPNANP (SEQ ID NO:38), and wherein the peptide comprising the amino acid sequence NVDP consists of the amino acid sequence NVDPNANPNVDPNANPNVDP (SEQ ID NO:39).

2. The antibody of claim 1, wherein the dissociation constant K.sub.D for the antibody and the peptide comprising an amino acid sequence NANP is at most 10.sup.−6 M, preferably at most 2×10.sup.−7 M, more preferably at most 10.sup.−7 M, even more preferably at most 5×10.sup.−8 M, still more preferably at most 2×10.sup.−8 M, most preferably at most 10.sup.−8 M, wherein the dissociation constant K.sub.D for the antibody and the peptide comprising an amino acid sequence NVDP or NPDP is at most 10.sup.−6 M, preferably at most 2×10.sup.−7 M, more preferably at most 10.sup.−7 M, even more preferably at most 5×10.sup.−8 M, still more preferably at most 2×10.sup.−8 M, most preferably at most 10.sup.−8 M, and/or wherein the dissociation constant K.sub.D for the antibody and the peptide comprising an amino acid sequence KQPA is at most 10.sup.−5 M, preferably at most 5×10.sup.−6 M, more preferably at most 2×10.sup.−6 M, most preferably at most 10.sup.−7 M.

3. The antibody of claim 1, wherein said antibody comprises complementarity determining regions (CDRs) comprising the sequences of SEQ ID NOs:5 to 10 or SEQ ID NOs:11 to 16 or comprises complementarity determining regions (CDRs) comprising sequences at least 80% identical to the sequences of SEQ ID NOs:5 to 10 or SEQ ID NOs:11 to 16.

4. The antibody of claim 1, wherein said antibody comprises (i) an amino acid sequence of the heavy chain as shown in SEQ ID NO:29 or a sequence at least 50% identical to SEQ ID NO:29; and an amino acid sequence of the light chain as shown in SEQ ID NO:30 or a sequence at least 50% identical to SEQ ID NO:30; or (ii) an amino acid sequence of the heavy chain as shown in SEQ ID NO:31 or a sequence at least 50% identical to SEQ ID NO:31; and an amino acid sequence of the light chain as shown in SEQ ID NO:32 or a sequence at least 50% identical to SEQ ID NO:32.

5. The antibody of claim 1, wherein said antibody is a monclonal antibody or a fragment thereof; and/or is a human or a humanized antibody.

6. (canceled)

7. (canceled)

8. A method of preventing a Plasmodium infection in a subject, comprising contacting said subject with the antibody according to claim 1.

9. A method for detecting a Plasmodium circumsporozoite protein in a sample, comprising a) contacting said sample with an antibody according to any one of claims 1 to 5, and thereby b) detecting said Plasmodium circumsporozoite protein in said sample.

10. The method of claim 9, wherein said method comprises further step al) detecting binding of said antibody to said Plasmodium circumsporozoite protein.

11. A method for detecting an antibody suitable for preventing malaria, comprising a) providing a candidate antibody suspected to be suitable for preventing malaria; b) determining an affinity of said candidate antibody to a peptide comprising the amino acid sequence NANP (SEQ ID NO:1); c) determining an affinity of said candidate antibody to a peptide comprising at least one amino acid sequence selected from NVDP (SEQ ID NO:2), NPDP (SEQ ID NO:3), and KQPA (SEQ ID NO:4); and d) identifying an antibody suitable for preventing malaria based on the results of steps b) and c).

12. The method of claim 11, wherein an antibody suitable for preventing malaria is identified if (i) it is determined in step b) that the dissociation constant K.sub.D for the candidate antibody and the peptide comprising an amino acid sequence NANP is at most 10.sup.−6 M, preferably at most 2×10.sup.−7 M, more preferably at most 10.sup.−7 M, even more preferably at most 5×10.sup.−8M, still more preferably at most 2×10.sup.−8M, most preferably at most 10.sup.−8 M; and (ii) it is determined in step c) that the dissociation constant K.sub.D for the candidate antibody and the peptide comprising an amino acid sequence NVDP, NPDP, and KQPA is at most 10.sup.−6 M, preferably at most 2×10.sup.−7 M, more preferably at most 10.sup.−7 M, even more preferably at most 5×10.sup.−8 M, still more preferably at most 2×10.sup.−8 M, most preferably at most 10.sup.−8 M and/or the dissociation constant K.sub.D for the candidate antibody and the peptide comprising an amino acid sequence KQPA is at most 10.sup.−5 M, preferably at most 5×10.sup.−6 M, more preferably at most 2×10.sup.−6 M, most preferably at most 10.sup.−7 M.

13. (canceled)

14. (canceled)

15. (canceled)

16. (canceled)

17. (canceled)

18. The method of claim 8, wherein said plasmodium infection is a Plasmodium falciparum infection.

19. The method of claim 8, wherein said preventing a Plasmodium infection comprises preventing malaria.

Description

FIGURE LEGENDS

[0106] FIG. 1. Cross-reactivity of human PfCSP-reactive antibodies

[0107] (A) Schematic representation of PfCSP from NF54 comprising the N-terminus, central repeat, and C-terminal (C-CSP) domain. The amino acid sequence downstream of the N-terminal domain including the conserved region 1 (RI), the N-terminal junction, and the central repeat domain is indicated, as well as the NANA epitope in the linker region upstream of the aTSR domain in C-CSP. Amino acid sequences of overlapping peptides in the N-terminal junction containing known epitopes of protective antibodies (Kisalu et al., 2018, Tan et al., 2018), designated as KQPA, NPDP, NVDP, and of a NANP 5.5-mer peptide (designated as NANP) are indicated and shown in different shades of grey.

[0108] (B) Binding strength of anti-PfCSP antibodies (n=200; Murugan et al., 2018) to the indicated overlapping peptides and C-CSP as in (A) is shown as calculated area under curve (AUC) values based on ELISA measurements at different antibody concentrations. The frequency of reactive and non-reactive antibodies is indicated.

[0109] (C and D) Binding profile to the indicated PfCSP peptides and C-CSP as in (B) of representative antibodies with specificity for NVDP, NANP, or C-CSP (C) and cross-reactive antibodies with different binding profiles (D).

[0110] Data in B-D shows mean values from three independent experiments. Horizontal lines in B-D indicate the reactivity threshold.

[0111] FIG. 2. Cross-reactivity of NANP antibodies with the N-terminal junction but not C-CSP is associated with high affinity

[0112] (A-C) Affinity of epitope-specific compared to cross-reactive antibodies as determined by SPR.

[0113] (A) NVDP affinity (left) of NVDP-specific antibodies (n=16) and of NVDP, NPDP cross-reactive antibodies (n=22) and NPDP affinity of NVDP, NPDP cross-binders (right).

[0114] (B) NANP (left), NVDP (center left), NPDP (center right) and KQPA (right) affinities of NANP-specific antibodies (n=41) and of NANP binders with cross-reactivity to NVDP (n=24), NVDP and NPDP (n=19), or NVDP, NPDP and KQPA (n=6).

[0115] (C) C-CSP (left), NANP (center left), NVDP (center right) and NPDP (right) affinities of C-CSP-specific antibodies (n=4) and C-CSP binders with cross-reactivity to NANP (n=9), NANP and NVDP (n=5), or NANP, NVDP and NPDP (n=3).

[0116] (A-C) Black horizontal lines indicate geometric means. P-values were calculated by Mann-Whitney test. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001; ns indicates statistically non-significant differences.

[0117] FIG. 3: The paratope core of cross-reactive PfCSP antibody 4493 preferentially binds NANP motifs.

[0118] (A) Affinity of clonally related antibodies (4142, 4493 and 4560) with different numbers of

[0119] IGH and IGL somatic mutations (aa exchanges) to the indicated peptides determined by Surface Plasmon Resonance (SPR).

[0120] (B) Affinity of mAb 4493 to the indicated peptides as measured by isothermal titration calorimetry (ITC).

[0121] (C-G) mAb 4493 in co-complex with the peptides KQPA (C), NPDP (D), NDN.sub.3 (E), DND.sub.3

[0122] (F) and NANP (G) H-bonds between mAb 4493 and the respective peptide are shown.

[0123] (H) Positioning of the 4-aa motifs in the indicated peptides in the mAb 4493 paratope as observed in the antibody co-complexes (C-G). Three distinct paratope positions (0, 1 and 2) are indicated. Amino acid residues resolved in the X-ray crystal structures are underlined.

[0124] FIG. 4: High-affinity cross-reactive antibodies are potent parasite inhibitors

[0125] (A-C) Capacity of antibodies (n=139, 100 μg/ml) with the indicated binding profiles to inhibit the hepatocyte traversal activity of Pf sporozoites in vitro.

[0126] (A) NVDP-specific antibodies (n=14) and NVDP, NPDP cross-reactive antibodies (n=18).

[0127] (B) NANP-specific antibodies (n=41) and NANP binders with cross-reactivity to NVDP (n=20), NVDP and NPDP (n=18), or NVDP, NPDP and KQPA (n=6).

[0128] (C) C-CSP-specific antibodies (n=5) and C-CSP binders with cross-reactivity to NANP (n=9), NANP and NVDP (n=5), or NANP, NVDP and NPDP (n=3).

[0129] (D) IC.sub.50 values versus NANP (left), NVDP (middle), and NPDP (right) of selected antibodies with >95% Pf hepatocyte traversal-inhibitory activity (B and C).

[0130] (A-C) Black horizontal lines indicate arithmetic mean. (A-C) Data is representative of at least two independent experiments. (D) IC.sub.50 values were calculated from data of at least three independent experiments. P-values were calculated by Mann-Whitney test. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001; ns indicates non-significant statistical differences.

[0131] FIG. 5. In vivo protective activity of cross-reactive IGHV3-33- and non-IGHV3-33-encoded PfCSP antibodies with different binding profiles

[0132] (A) Scheme and time line of the experiments. Antibody potency was assessed in C57BL/6 mice after intra peritoneal (i.p.) passive transfer of mAb 20 h before exposure to bites of mosquitoes infected with Plasmodium berghei transgenic parasites expressing P. falciparum CSP (PbPfCSP). Two to three hours post infection (hpi) blood was collected and serum concentrations of the monoclonal antibodies were determined by ELISA. The time to blood-stage parasitemia was monitored by blood smear between day 3 and day 10 after the mosquito-bite exposure.

[0133] (B-D) In vivo protective activity of cross-reactive IGHV3-33-encoded PfCSP antibodies with different binding profiles.

[0134] (B) Affinity profiles of mAbs 1210 (Imkeller et al., 2018), 2164 and 4476 to KQPA, NPDP, NVDP, NANP and C-CSP as determined by SPR.

[0135] (C) Capacity of the indicated antibodies (300 μg/ml) to protect mice from infection by mosquito bites with PbPfCSP parasites in three independent experiments as determined by the percentage of parasite-free mice. The total number of mice per antibody in three experiments is indicated (n).

[0136] (D) Serum concentration of the transferred monoclonal antibodies in individual mice at the time of parasite challenge.

[0137] (E-G) In vivo protective activity of cross-reactive non-IGHV3-33-encoded PfCSP antibodies with different binding profiles.

[0138] (E) Affinity profiles of mAbs 4493, CIS43 (Kisalu et al., 2018), and 317 (Oyen et al., 2017) to KQPA, NPDP, NVDP, NANP and C-CSP as determined by SPR in comparison to the IGHV3-33-encoded mAb 1210 (Imkeller et al., 2018).

[0139] (F) Capacity of the indicated antibodies (150 μg/ml) to protect mice from infection by mosquito bites with PbPfCSP parasites in two independent experiments as determined by the percentage of parasite-free mice. The total number of mice per antibody in two experiments is indicated (n).

[0140] (G) Serum concentrations of the transferred monoclonal antibodies in individual mice at the time of parasite challenge.

[0141] (H) Affinity profile of mAb 2541 (filled circle) to the indicated peptides compared to the other IGHV3-33-, IGKV1-5-encoded plasmablast antibodies (open circles) as determined by SPR.

[0142] (I) All mAbs as in (F) and mAb 2541 were tested for their capacity to protect from blood-stage parasitemea after passive i.p. mAb transfer of 150 μg/mouse. Pooled data from three independent experiments is shown. The total number of mice per group is indicated (n). (J) Serum concentration of the transferred monoclonal antibodies in individual mice at the time of parasite challenge.

[0143] (B, E) Data represent the mean from three independent measurements. (D, G, J) Data represent the mean from at least two independent measurements. (C, F, I) mAbs showing statistically significant difference (P<0.05) in protection are indicated with different alphabets (a, b, c). (D, G, J) Horizontal black lines indicate means. P-values were calculated by Mantel-Cox log-rank test (C, F, I) and Mann-Whitney test (G). ***P<0.001.

[0144] The following Examples shall merely illustrate the invention. They shall not be construed, whatsoever, to limit the scope of the invention.

EXAMPLE 1: MATERIALS

Cell Lines

[0145] HEK293F cells were cultured according to the manufacturer's instructions. The cells were passaged at 37° C., 8% CO.sub.2 and 180 rpm in a 50 ml Bioreactor in FreeStyle™ 293-F medium. HC-04 cells (MRA-975, deposited by Jetsumon Sattabongkot; Sattabongkot et al., 2006) were cultured at 37° C. and 5% CO.sub.2 using HC-04 complete culture medium (428.75 ml MEM (-L-glu), 428.75 ml F-12 Nutrient Mix (+L-glu), 15 mM HEPES, 1.5 g/l NaHCO.sub.3, 2.5 mM L-glutamine and 10% FCS).

Bacteria

[0146] MAX Efficiency® DH10B™ Competent Cells were cultured at 37° C. and 180 rpm in LB medium for maintenance and Terrific broth for plasmid production.

Plasmodium falciparum Cultures

[0147] Plasmodium falciparum PfNF54 (a kind gift of Prof. R. Sauerwein) were cultured in O+ human red blood cells at 37° C., 4% CO.sub.2 and 3% 02 in a Heracell 150i Tri-gas incubator (Thermo Scientific). For gametocyte production, asynchronous parasite cultures were diluted to 1% parasitaemia and maintained for 15-16 days with daily change of RPMI-1640 medium (Thermo Scientific cat #52400) supplemented with 10% human A+ serum and 10 mM hypoxantine (c-c-Pro) until mosquito infections.

[0148] Pb-PfCSP, a replacement P. berghei line expressing Pf CSP (NF54) under the control of the Pb csp regulatory sequences (Triller et al. 2017), was obtained from Chris J. Janse and Shahid M. Khan and passaged every 3-4 days in CD1 female mice.

Mosquitoes

[0149] All mosquitoes were kept at 28-30° C. and 70-80% humidity. Anopheles coluzzii Ngousso S1 strain (Harris et al, 2010) were used for the production of Pf NF54 sporozoites for in vitro traversal assays. A. gambiae 7b line, an immunocompromised transgenic mosquitoes derived from the G3 laboratory strain (Pompon and Levashina, 2015), were used for the production of Pb-PfCSP sporozoites for in vivo infections.

Mice

[0150] Female C57BL/6 mice (7-9 weeks old) and female CD-1 mice (8-12 weeks old) were bred in the MPIIB Experimental Animal Facility (Marienfelde, Berlin), handled in accordance with the German Animal Protection Law (§ 8 Tierschutzgesetz) and approved by the Landesamt für Gesundheit and Soziales (LAGeSo), Berlin, Germany (project numbers 368/12 and H0335/17).

EXAMPLE 2: METHODS

Ig Gene Cloning and Recombinant Antibody Production

[0151] Ig heavy and light chain genes corresponding to antibody were cloned into human Igγ1 (AbVec2.0-IGHG1, Genbank ID: LT615368.1) and Igκ (AbVec1.1-IGKC, Genbank ID: LT615369.1) or Igλ (AbVec1.1-IGLC2-XhoI) expression vectors, respectively (Tiller et al., 2009). The cloning vectors are available from Addgene (Catalog numbers: 80795, 80796 and 99575). In brief, restriction site-tagged specific V and J-gene primers were used for amplifying Ig genes from single B cells and the amplicons were cloned into the above mentioned vectors. Ig genes of mAbs CIS43 and 317 were obtained by reverse translation of the protein sequences deposited (PDB accession number 6B5M for CIS43 (Kisalu et al., 2018) and 6AXL for 317 (Oyen et al., 2017)). Ig genes of CIS43 were synthesized at MWG Eurofins Genomics with AgeI restriction site at the 5′ end and SalI and BsiWI restriction sites at the 3′end of the heavy and kappa Ig genes, respectively. Ig genes of 317 were synthesized at GeneArt (Thermofisher) and restriction sites were introduced via PCR. Upon successful cloning, recombinant monoclonal antibodies were expressed in HEK293F cells (ThermoFisher Scientific).

Enzyme-Linked Immunosorbent Assay

[0152] Recombinant monoclonal antibodies were purified using Protein G Sepharose beads (GE healthcare) and the IgG concentration was measured by ELISA as described (Tiller et al., 2008). Antigen and serum ELISAs were performed as described (Triller et al., 2017). In brief, high-binding 384 well polystyrene plates (Corning) were coated overnight at 4° C. with KQPA, NPDP, NVDP, C-CSP or Streptavidin at 50 ng/well or PfCSP at 40 ng/well in 25 μl. Streptavidin-coated plates were incubated for 1 h with 200 ng/well biotinylated NANP5.5 in 25 μl. Plates were washed 3 times with 0.05% Tween 20 in PBS, blocked with 50 μl of 1% BSA in PBS for 1 h at room temperature (RT), and washed again prior to incubation with monoclonal antibodies at the indicated concentrations for 1.5 h at RT. Wells were washed and incubated with goat anti-human IgG-HRP (Jackson Immuno Research) in PBS with 1% BSA. One-step ABTS substrate (RT, 20 μl/well; Roche) and 1×KPL ABTS® peroxidase stop solution (RT, 20 μl/well; SeraCare Life Sciences, Inc.) were used for detection. A chimeric version of the murine anti-PfCSP antibody 2A10 (Triller et al., 2017) with human Ig heavy and Ig kappa constant regions and the non-PfCSP-reactive antibody mGO53 (Wardemann et al., 2003) were used as a positive control and negative controls, respectively. ELISA area under curve (AUC) values were calculated using GraphPad Prism 7.04 (GraphPad).

Surface Plasmon Resonance (SPR)

[0153] SPR measurements were performed with a BIACORE T200 (GE Healthcare) as described (Murugan et al., 2018). In brief, the instrument was docked with a series S sensor chip CMS (GE Healthcare). 10 mM HEPES with 150 mM NaCl at pH 7.4 was used as running buffer. All samples were immobilized by amine coupling using the human antibody capture kit (GE Healthcare) according to the manufacturer's instructions. Sample antibodies and the non-PfCSP-reactive negative control antibody mGO53 (Wardemann et al., 2003) were captured in the sample and reference flow cell at equal concentrations, respectively. Flow cells were stabilized with running buffer at 10 μl/min flowrate for 20 min. The respective peptides were dissolved in running buffer and injected at 0, 0.015, 0.09, 0.55, 3.3, and 20 μM concentration. A flow rate of 30 μl/min was maintained, allowing the association and dissociation of the peptides for 60 s and 180 s respectively, at 25° C. For high affinity antibodies (˜10.sup.−10 M), additional measurement at 0, 0.42, 2.57, 15.43, 92.6 and 555.5 nM concentration was performed. After each run, both flow cells were regenerated with 3 M MgCl.sub.2. The data were fit using 1:1 binding model or steady state kinetic analysis using the BIACORE T200 software V2.0.

Pf Sporozoite Hepatocyte Traversal Assay

[0154] Anopheles coluzzii mosquitoes were infected with mature Pf gametocytes (NF54 strain) via artificial midi-feeders (Glass Instruments, The Netherlands) for 15 min and kept at 26° C. and 80% humidity in a controlled S3 facility in accordance with local safety authorizations (Landesamt für Gesundheit and Soziales Berlin, Germany, LAGeSo, project number 411/08). Infected mosquitoes received an additional uninfected blood meal 7-8 days post infection (dpi) and were collected 13-15 dpi to isolate sporozoites. Sporozoites were isolated in HC-04 medium by dissecting and grinding mosquito thoraces containing salivary glands with glass pestles, followed by filtering the extracts with 100 μm and 40 μm cell strainers. The isolated salivary gland sporozoites were enumerated in a hemocytometer (Malassez, Marienfelde) and used for traversal assays as previously described (Triller et al., 2017). Briefly, salivary gland Pf sporozoites in HC-04 medium were pre-incubated with 100 μg/ml or serial dilutions (0.032, 0.16, 0.8, 4 and 20 μg/ml) of monoclonal antibodies in 27.5 μl for 30 min on ice and added to human hepatocytes (HC-04, (Sattabongkot et al., 2006)) for 2 h at 37° C. and 5% CO.sub.2 in the presence of 0.5 mg/ml dextran-rhodamine (Molecular Probes). Cells were washed and fixed with 1% PFA in PBS before measuring dextran positivity using FACS LSR II instrument (BD Biosciences). Data analysis was performed by subtraction of the background (dextran positive cells after incubation with uninfected mosquito salivary gland material) and normalization to the maximum Pf traversal capacity (dextran positive cells after incubation with salivary gland Pf sporozoites) using FlowJo V.10.0.8 (Tree Star). A chimeric humanized version of the PfCSP-reactive monoclonal antibody 2A10 (Triller et al., 2017) and of the non-PfCSP-reactive monoclonal antibody mGO53 (Wardemann et al., 2003) was used as positive and negative control, respectively. IC.sub.50 values were calculated for each antibody by four-parameter logistic curve fitting in GraphPad Prism 7.04 (GraphPad) using the measurements from at least three independent experiments.

Fab Production

[0155] Fabs of mAbs 2243, 4498, 2164, 4476 and 3945 were generated by papain digestion of IgG, purified via Protein A chromatography followed by cation-exchange chromatography (MonoS, GE Healthcare) and size-exclusion chromatography (Superdex 200 Increase 10/300 GL, GE Healthcare). Fabs of mAb 4493 were generated by cloning of the IGH and IGK variable region gene segments into pcDNA3.4 TOPO expression vectors immediately upstream of human IGK and CHI constant regions, respectively, followed by transient expression in HEK293F cells (Thermo Fisher Scientific) and purification via KappaSelect affinity chromatography (GE Healthcare), cation-exchange chromatography (MonoS, GE Healthcare) and size-exclusion chromatography (Superdex 200 Increase 10/300 GL, GE Healthcare).

Crystallization and Structure Determination

[0156] To enhance crystallizability, purified 4493 Fabs were mixed with the anti-kappa V.sub.HH domain (Thermofisher) in a 1:2 molar ratio and excess V.sub.HH domain was removed away via size exclusion chromatography (Superdex 200 Increase 10/300 GL, GE Healthcare). Fab-V.sub.HH co-complexes were concentrated to 6 mg/mL and diluted to 5 mg/mL with the respective peptide. 4493-V.sub.HH-KQPA crystals grew in 0.1 M HEPES pH 7.0, 1 M lithium chloride, 20% (w/v) PEG 6000 and were cryoprotected in 15% (w/v) ethylene glycol. 4493-V.sub.HH-NPDP crystals grew in 20% (w/v) PEG 3350, 0.2 M sodium nitrate and were cryoprotected in 15% (w/v) ethylene glycol. 4493-V.sub.HH-NDN.sub.3 crystals grew in 20% (w/v) PEG 3350, 0.2 M potassium nitrate and were cryoprotected in 15% (w/v) ethylene glycol. 4493-V.sub.HH-DND.sub.3 crystals grew in 20% (w/v) PEG 8000, 0.1 M IVIES pH 6.0 and 0.2 M calcium acetate and were cryoprotected in 20% (w/v) glycerol. 4493-V.sub.HH-NANP3 grew in 40% (w/v) PEG 600, 0.1 M sodium citrate pH 5.5 after microseeding from thin needle-like crystals that grew in 0.1 M HEPES pH 7.5, 20% (w/v) PEG 8000 and were cryoprotected in 15% (w/v) ethylene glycol. Data were collected at the 08ID-1 beamline at the Canadian Light Source (CLS), the 23-ID beamline at the Advanced Photon Source (APS), or the 17-ID-2 beamline at the National Synchrotron Light Source II (NSLS-II) processed and scaled using XDS (Kabsch et al., 2010). The structures were determined by molecular replacement using Phaser (McCoy et al., 2007). Refinement of the structures was carried out using phenix.refine (Adams et al., 2010) and iterations of refinement using Coot (Emsley et al., 2010). Software were accessed through SBGrid (Morin et al., 2013).

Isothermal Titration Calorimetry

[0157] calorimetric titration experiments were performed with an Auto-iTC200 instrument (Malvern) at either 15° C. or 25° C. Proteins were dialyzed against 20 mM Tris pH 8.0 and 150 mM sodium chloride overnight at 4° C. Fabs were concentrated to 10 μM and added to the calorimetric cell, which was titrated with peptide (100 μM) in 15 successive injections of 2.5 μl. Experiments were performed at least twice and the mean and standard error of the mean were reported (FIG. 3). The experimental data were analyzed according to a 1:1 binding model by means of Origin 7.0.

Plasmodium berghei Infections In Vivo

[0158] A. gambiae 7b mosquitoes were fed on female CD-1 mice infected with Pb-PfCSP parasites (0.1-0.8% gametocytemia) and kept at 20° C. and 80% humidity until further usage. Infected mosquitoes were offered an additional uninfected blood meal 7 dpi and 20 mosquitoes were dissected for oocyst counts 17 dpi. Female C57BL/6 mice were passively immunized by i.p. injection of 150 or 300 μg of monoclonal antibodies in 200 μl of PBS. After 20 h (18 dpi), mice were exposed to Pb-PfCSP-infected mosquitoes (infection prevalence between 75% and 100%; Supplemental Tables S9-S12). All blood-fed mosquitoes were collected individually for gDNA extraction (NucleoMag VET, Macherey-Nagel) followed by PCR to determine their Pb-PfCSP infectivity status. Specific primers amplifying P. berghei 18s RNA gene (Friesen et al. 2010) and control primers amplifying A. gambiae AGAP001076 gene (Gildenhard et al. 2019) were used. Mosquitoes positive for both PCR reactions were considered infected. Antibody titers were measured by ELISA in serum samples collected from the submandibular vein 2-3 h post mosquito bite. Blood parasitaemia was assayed by daily tests of a minimum of 100 microscopic fields per Giemsa-stained thin blood smears 3-7 days and 10 days post mosquito bite. Infected mice were sacrificed two days after the detection of parasitaemia.

Statistics

[0159] Statistics was performed on Prism 7.04 (GraphPad) or RStudio (version 3.2.2) using two-tailed Mann-Whitney assuming non-normal distribution or Mantel-Cox log-rank test for in vivo experiments, as described in the figure legends. P values less than 0.05 were considered significant (*P<0.05; **P<0.01; ***P<0.001; ****P<0.0001) as indicated in the figure legends.

Data and Software Availability

[0160] The data that support the findings of this report are available from the corresponding authors upon reasonable request. The crystal structures reported herein have been deposited in the Protein Data Bank, www.rcsb.org (PDB ID codes 6O23, 6O24, 6O25, 6O26, 6O28, 6O29, 6O2A, 6O2B, 6O2C).

EXAMPLE 3: THE MAJORITY OF PFCSP ANTIBODIES ARE CROSS-REACTIVE WITH SEVERAL B CELL EPITOPES

[0161] To identify antibodies with reactivity to the different subdomains, we analyzed a large panel of 200 recombinant PfCSP-reactive human monoclonal antibodies (mAbs) for binding to the junctional and C-CSP epitopes by ELISA (FIG. 1). The antibodies were derived from memory B cells of malaria-naïve volunteers after repeated immunization with live sporozoites under chemoprophylaxis (PfSPZ-CVac). Binding was determined to three overlapping peptides in the N-terminal junction containing highly similar NPDP, NVDP, and NANP amino acid (aa) motifs, to a NP(NANP).sub.5 peptide representative of the central repeat, and to the complete C-terminus (C-CSP) with a unique NANA sequence (FIG. 1A). The N-terminal junction peptides, here abbreviated as KQPA, NPDP, and NVDP according to the first four aa of each peptide, covered aa 95-109 (KQPADGNPDPNANPN, SEQ ID NO:37), aa 101-116 (NPDPNANPNVDPNANP, SEQ ID NO:38), and aa 109-125 (NVDPNANPNVDVNANPNVDP, SEQ ID NO:39), respectively. Nearly 80% of the antibodies (155/200) bound the peptide above our cut-off with an ELISA area under curve (AUC) of >5. The vast majority bound NANP (57%), as previously described (Murugan et al., 2018), and NVDP (50%). Antibodies with reactivity to NPDP (25%) or C-CSP (9%) were less abundant, and only a few recognized KQPA (5%), representing the most N-terminal part of the junction. Several of the 155 antibodies recognized only the NANP repeat (26%), the NVDP peptide (11%), or the C-terminus (3%) but not KQPA, or NPDP and were therefore defined as epitope-specific (FIG. 1C). In contrast, 92/155 antibodies (59%) recognized two or more PfCSP peptides with distinct binding profiles and were therefore referred to as epitope cross-reactive (FIG. 1D). The vast majority (76%) of these cross-binders interacted strongly with the NANP repeat, whereas 24% lacked NANP-reactivity and instead showed preferential binding to NVDP and NPDP. Presumably due to the low degree of sequence similarity between the C terminus and N-terminal junction, C-CSP binding was lower in C-CSP-reactive antibodies with cross-reactivity to the repeat and the N-terminal junction peptides compared to those that only cross-reacted with NANP or compared to C-CSP specific antibodies. The different binding profiles of all antibodies were resolved in a t-distributed stochastic neighbor embedding (t-SNE) analysis by highlighting the ELISA binding strength of each antibody to the individual peptides and to C-CSP. In summary, our analysis identified a high number of PfCSP antibodies with cross-reactivity and a wide spectrum of antigen-binding profiles to the N-terminal junction, the central repeat and C-CSP with highly similar 4-aa NANP, NVDP, NPDP or NANA sequence motifs.

EXAMPLE 4: CROSS-REACTIVITY WITH THE N-TERMINAL JUNCTION CORRELATES WITH ANTIBODY AFFINITY TO NANP

[0162] To delineate the link between cross-reactivity and binding strength, we measured the affinity of antibodies with ELISA-reactivity to the different PfCSP peptides and C-CSP by surface plasmon resonance (SPR; FIG. 2). The data confirmed the high degree of antibody cross-reactivity observed in the ELISA (FIG. 1). Anti-NVDP antibodies with cross-reactivity to the overlapping NPDP peptide showed on average significantly higher NVDP affinities compared to NVDP-specific antibodies (FIG. 2A). Similarly, anti-NANP antibodies with cross-reactivity to the N-terminal junction had higher NANP affinities than NANP-specific antibodies (FIG. 2B). Thus, NANP affinity increased with cross-reactivity to the N-terminal junction and was highest in antibodies that recognized all junctional peptides. The gain in NANP affinity was paralleled by a comparable increase in affinity to NVDP and NPDP. Highest mean affinities to NANP, NVDP and NPDP were observed in cross-reactive antibodies that bound all four peptides, although their KQPA affinity was overall low compared to the other peptides (FIG. 2B). In contrast, strong binding to C-CSP was not associated with cross-reactivity to the repeat and N-terminal junction peptides (FIG. 2C). In summary, high affinity to NANP and NVDP correlated with antibody cross-reactivity to the PfCSP N-terminal junction but not C-CSP.

EXAMPLE 5: THE CORE PARATOPE OF PFCSP ANTIBODIES PREFERENTIALLY BINDS TO NANP MOTIFS

[0163] To understand how a single cross-reactive antibody would bind to the repeat and the N-terminal junction peptides, we investigated mAb 4493, a rare IGHV3-49-, IGKV3-20-encoded antibody that has affinity matured to NANP and the junctional epitopes (FIGS. 3A and 3B). Specifically, we determined the crystal structure of the mAb 4493 Fab fragments in complex with NANP.sub.3 (2.15 Å), two shortened NVDP peptides, NDN.sub.3 (NANPNVDPNANP, SEQ ID NO:40, 2.1 Å) and DND.sub.3 (NVDPNANPNVDP, SEQ ID NO:41, 2.02 Å), NPDP (NPDPNANPNVDPNANP, SEQ ID NO:42, 2.4 Å), and KQPA (KQPADGNPDPNANPN, SEQ ID NO:43, 1.93 Å) (FIG. 3). mAb 4493 contacted all peptides with its HCDR2, HCDR3, and KCDR3. Especially, the IGHV3-49 germline-encoded amino acid H.Arg52 in HCDR2 was strongly involved in direct H-bond contacts with the peptide backbones. The preferential binding of mAb 4493 to NPDP (FIG. 3) correlated with more extensive H-bond interactions compared to the other peptides (11 H-bonds compared to 8 or 9 for DND.sub.3 and NANP.sub.3, respectively), where H.Arg52 alone formed five H-bonds with NPDP.

[0164] Strikingly, mAb 4493 bound all peptides in largely superimposable U-shaped conformations (rmsd<0.2 Å) around H.Arg52, similar to IGHV3-33-encoded antibodies, which bound all peptides in inverted S-shape conformations around H.Trp52. The mAb 4493 paratope engaged with three consecutive 4-aa motifs at three distinct positions, here referred to as position 0, 1, and 2. In five of the six structures, position 1 was occupied by NANP illustrating a strong preference of the core paratope for this motif. Indeed, although NPDP could be accommodated at position 1 as observed in the co-complex of mAb 4493 with the KQPA peptide, this binding mode, which placed the NANP motif in position 2, was associated with overall weaker affinity compared to all other peptides including the high-affinity NPDP peptide interaction with NPDP, NANP, and NVDP motifs at positions 0, 1, and 2, respectively (FIG. 3). Compared to the strong preference for NANP motifs at position 1, mAb 4493 showed more flexibility at position 0 as it accommodated NPDP, NANP, and NVDP motifs in the co-complexes with NPDP, NANP.sub.3, and DND.sub.3 peptides, respectively. Binding flexibility was also observed at position 2, which bound NANP and NVDP motifs but not NPDP. Likely due to the strong preference for binding to NANP at position 1, position 2 was more frequently occupied by NVDP, thereby placing NPDP into position 0 according to the natural order of the NPDP-, NANP-, NVDP-motifs in full-length PfCSP. Strikingly, the antigen recognition mode of mAb 4493 was highly similar to CIS43 (Kisalu et al., 2018), a potent cross-reactive IGHV1-2-, IGKV4-1-encoded PfCSP antibody with preference for binding to the N-terminal junction that had been induced by immunization with irradiated sporozoites. Despite differences in the binding orientation between the two antibodies, the peptides bound by mAb 4493 showed highly similar conformations compared to peptides bound by CIS43, which also recognizes recognized junctional peptides through interactions of the core paratope centered on NANP motifs.

EXAMPLE 6: EVOLUTION OF ANTI-PFCSP ANTIBODY BINDING PROFILES

[0165] To determine whether NANP binding played a role in the development of cross-reactive antibodies, we assessed the evolution of the response over time. The frequency of NANP-binding antibodies with strong cross-reactivity to the junctional peptides, especially NPDP and NVDP, increased with repeated parasite exposure and was higher after the third immunization and challenge compared to the second immunization. The vast majority was also class-switched to IgG, including many IGHV3-33-, IGKV1-5-encoded but also rare antibodies with non-prominent gene combinations such as mAb 4493. Most cross-reactive antibodies belonged to clonally expanded and diversified B cell clusters, but the majority showed no differences in their cross-reactivity profile independently of their binding preference and their absolute affinity. To assess what drove the selection of high-affinity cross-reactive antibodies, we compared the kinetic rates of cross-reactive antibodies to associate with (k.sub.on) or dissociate from (k.sub.off) NANP, NVDP, and NPDP. Despite similar k.sub.on rates for all peptides, the antibodies differed in their k.sub.off, which was significantly lower for binding to the NANP repeat compared to NVDP or NPDP, suggesting that the antibodies had been primarily selected for their ability to bind the repeat and not the junctional peptides. Indeed, although overall infrequent, several of the IGHV3-33- and IGKV1-5-encoded antibodies from PfCSP-reactive memory B cell and also bona-fide plasmablasts carried selected somatic mutations that have been shown to either directly (H.S31, H.V50) or indirectly (H.N56, K.S93) improve NANP binding through homotypic antibody-antibody interactions. Thus in line with our structure analyses, the accumulation of cross-reactive antibodies in response to repeated parasite exposure was likely due to the direct association of antibody cross-reactivity with NANP affinity and continuous selection of high-affinity clones from the naïve repertoire or of B cells that had gained affinity through somatic mutations, especially of IGHV3-33-, IGKV1-5-encoded antibodies.

EXAMPLE 7: ANTIBODY BINDING TO NANP BUT NOT THE N-TERMINAL JUNCTION OR C-CSP CORRELATES WITH POTENT PF INHIBITION IN VITRO

[0166] To determine whether the observed differences in antibody-binding preferences impacted on their Pf-inhibitory capacity, we compared the activity of 139 antibodies with different cross-reactivity profiles to inhibit the sporozoite traversal of hepatocytes in vitro (100 μg/ml). NVDP-specific antibodies were overall poor inhibitors (mean 55%). Although the mean inhibitory activity was significantly higher for NVDP binders with cross-reactivity to NPDP (mean 74%), in line with their higher affinity (FIG. 2), none of the antibodies reached 100% inhibition. NANP-specific antibodies were overall better Pf inhibitors than NVDP-specific antibodies (mean 69%) with two antibodies that conferred complete inhibition (FIG. 4B). The mean potency of NANP-binders increased significantly with cross-reactivity to the N-terminal junction and was higher for antibodies that bound to two or three junctional peptides (mean 92%) than for anti-NANP antibodies with limited cross-reactivity to NVDP only (mean 76%). C-CSP specific antibodies were the weakest inhibitors of all (mean 13%) but their potency was significantly improved with associated cross-reactivity and higher affinity to NANP as previously reported (mean 72%; Scally) and with additional binding to NVDP and NPDP (mean 95%; FIG. 4C). Thus, NANP binding was associated with parasite inhibition, explaining the overall low protective activity of non-NANP-reactive antibodies.

[0167] We next examined whether the affinities (K.sub.D) of cross-reactive antibodies to NANP, NVDP, or NPDP were predictive of low (50-70%), intermediate (>70-90%) or high (>90%) levels of Pf inhibition (FIG. 4D). NANP, but not NVDP or NPDP affinity was strongly associated with anti-Pf activity. The most potent inhibitors (>90%) were almost exclusively antibodies with NANP affinity below 10.sup.−7 M. The NVDP affinity of the most potent inhibitors was overall weaker (mean K.sub.D=1.9×10.sup.−7 M) than their NANP affinity (mean K.sub.D=1.7×10.sup.−8 M), and comparable to antibodies with intermediate Pf-inhibitory activity (mean K.sub.D=3.4×10.sup.−7 M).

[0168] Although rare antibodies with high NPDP affinity were identified in all three groups, NPDP affinity was overall low in the range of K.sub.D 10.sup.−6 M and did not discriminate between antibodies with low, intermediate, or high levels of Pf inhibition. In summary, cross-reactive antibodies with low anti-parasite activity were mostly weak binders, whereas high affinity to NANP but not to NVDP or NPDP discriminated the most potent cross-reactive inhibitors from antibodies with intermediate anti-parasite activity.

[0169] Next, we selected 15 of the most potent non-clonally-related antibodies with >95% inhibition activity including one NANP-specific and 14 cross-reactive antibodies with different binding profiles and determined their IC.sub.50 values. Ten antibodies were encoded by IGHV3-33 in combination with IGKV1-5 (8/10) or IGKV3-20 (2/10), reflecting the strong enrichment of this gene combination among antibodies with >90% Pf inhibitory activity, whereas the other five antibodies were encoded by different IGHV3 and various IGKV and IGLV light chain genes. Regardless of their binding profile, the IC.sub.50 values of all 15 antibodies ranged from 0.2 μg/ml to >20 μg/ml and correlated with their NANP (K.sub.D 10.sup.−7-10.sup.−9 M) but not their NVDP (K.sub.D 10.sup.−6-10.sup.−9 M) or NPDP (K.sub.D 10.sup.−5-10.sup.−9 M) affinities (FIG. 4D). Strikingly, with one exception, the IC.sub.50 values of all ten IGHV3-33-encoded antibodies were <1.0 μg/ml (mean=0.68 μg/ml). In contrast, only 2/5 antibodies with other gene combinations showed IC.sub.50 values <1.0 μg/ml, whereas for the other three, the values ranged from 4.7 μg/ml to >20 μg/ml. Thus, non-IGHV3-33-, encoded antibodies with high Pf-inhibitory activity were rare. The most potent antibodies in this assay all showed NANP affinities <10.sup.−7 M independently of their binding profiles and included cross-reactive and rare epitope-specific antibodies.

EXAMPLE 8: ANTIBODY-MEDIATED PROTECTION FROM THE INFECTION IN VIVO

[0170] To determine the inhibitory activity of the most potent antibodies in vivo, we measured the protective efficacy of the antibodies as time to the development of blood-stage parasites (prepatency) in mice infected with PJCSP-expressing transgenic Plasmodium berghei parasites (Pb-PfCSP) by mosquito bites 20 h after intraperitoneal antibody injection (300 μg/mouse, FIG. 5A). First, we assessed the potency of three cross-reactive IGHV3-33-encoded antibodies with IC.sub.50 values <1.0 μg/ml (mAbs 1210, 2164, 4476) compared to a non-protective C-CSP-specific negative control (mAb 1710, Scally). mAbs 4476 and 1210 had comparable NANP, NVDP, and NPDP affinities, but only mAb 4476 recognized C-CSP. In contrast, mAb 2164 bound NANP, NVDP, and NPDP with up to 2,000-fold higher affinity and also cross-reacted with KQPA but not C-CSP. The highest protection (56% parasitemia-free mice) was observed for mAb 1210, but the differences compared to mAb 2164 (33%) and mAb 4476 (30%) were not statistically significant (Table S8). Thus, affinity and cross-reactivity with the N-terminal junction or C-CSP did not predict the in vivo potency of these IGHV3-33-encoded antibodies.

[0171] To determine whether the same was true for non-IGHV3-33-encoded antibodies, we compared the potency of mAb 4493 (IGHV3-49-, IGKV3-20) with mAb 1210 in the same model (FIGS. 5E-5G). Additionally, we included two potent published non-IGHV3-33-encoded antibodies. mAb CIS43 (IGHV1-3, IGKV4-1; Kisalu et al., 2018) was similar to mAb 4493 in its strong affinity for NPDP, but showed no measurable KQPA reactivity and about 10-fold and 100-fold lower NVDP and NANP affinity, respectively. mAb 317 (IGHV3-30-3, IGKV1-5; Oyen et al., 2017) had an exceptionally high NANP affinity (<10.sup.−10 M) with relatively low cross-reactivity to NVDP and NPDP (FIG. 5E). At a dose of 300 μg per mouse, mAbs 4493, CIS43, and 317 showed similar levels of protection from blood-stage parasitemia of 83%, 91%, and 100%, respectively. To better resolve potential differences in the inhibitory activities of the antibodies, we halved the dose to 150 μg per mouse (FIGS. 5F and 5G). Although the degree of protection was overall lower, all three antibodies retained their high potency compared to mAb 1210. mAb 317 protected 83% of mice compared to 62% for mAb CIS43 and 58% for mAb 4493, but these differences were not statistically significant. Of note, independent of the dose, mAb 4493 consistently showed on average two-fold lower serum concentrations at the time of challenge than the other antibodies (FIG. 5G).

[0172] To determine whether IGHV3-33-encoded antibodies could reach similar levels of potency, we compared the antibodies to mAb 2541, a high affinity cross-reactive IGHV3-33-, IGKV1-5-encoded plasmablast antibody with 8-aa-long KCDR3 and an IC.sub.50<1 μg/ml in the in vitro Pf-traversal inhibition assay. Notably, mAb 2541 showed higher affinity to NANP and NVDP than any IGHV3-33-, IGKV1-5-encoded memory B cell antibody in our panel, as well as additional cross-reactivity to NPDP and KQPA (FIG. 5H). In direct comparison to mAbs CIS43, 4493, and 317, mAb 2541 was as protective as these non-IGHV3-33-encoded antibodies (FIGS. 51 and 5J). Thus, the most potent PfCSP antibodies with high levels of in vivo protection against malaria parasites showed exceptional affinity to the repeat or the junctional epitopes and were encoded by IGHV3-33 (mAb 2541) or other gene combinations (mAbs 4493).

LITERATURE CITED

[0173] Adams et al., 2010, Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 213-221 [0174] Emsley et al., 2010, Acta Crystallogr. D Biol. Crystallogr. 66, 486-501 [0175] Friesen et al. 2010, Science Translational Medicine; 2(40):40ra49 [0176] Gildenhard et al. 2019, Nat Microbiol. 4(6):941-947 [0177] Kabsch et al., 2010, Acta Crystallogr. D Biol. Crystallogr. 66, 125-132 [0178] Kisalu et al., 2018, Nat. Med. 24, 408-416 [0179] McCoy et al., 2007, J. Appl. Cryst. 40, 658-674 [0180] Mordmüller et al., 2017, Nature; 542(7642):445-449 [0181] Morin et al., 2013, eLife 2, e01456 [0182] Murugan et al., 2018, Sci. Immunol. 3, eaap8029 [0183] Oyen et al., 2017, Proc. Natl. Acad. Sci. U.S.A 114, E10438-E10445 [0184] Pompon and Levashina, 2015, PLoS Biol. 13(9):e1002255 [0185] Sattabongkot et al., 2006, Am J Trop Med Hyg., 74(5):708-15 [0186] Tan, et al., Nat. Med. 24, 401-407 [0187] Tiller et al. 2008, J Immunol. Methods 329:112-124 [0188] Tiller et al., 2009, J Immunol Methods. 350:183-193 [0189] Triller et al., Immunity 47, 1197-1209.e10 (2017) [0190] Wardemann et al., 2003, Science 301:1374-1377