PIMMS43 is required for malaria parasite immune evasion and sporogonic development in the mosquito vector

Significance Malaria is transmitted among humans through mosquito bites. Here, we characterize a protein found on the surface of mosquito stages of malaria parasites and reveal that it serves to evade the mosquito immune system and ensure disease transmission. Neutralization of PIMMS43 (Plasmodium Infection of the Mosquito Midgut Screen 43), either by eliminating it from the parasite genome or by preincubating parasites with antibodies that bind to the PIMMS43 protein, inhibits mosquito infection with malaria parasites. Differences in PIMMS43 detected between African malaria parasite populations suggest that these populations have adapted for transmission by different mosquito vectors that are also differentially distributed across the continent. We conclude that targeting PIMMS43 can block malaria parasites inside mosquitoes before they can infect humans.

RT-PCR analysis of the expression of Pbc43 transcripts in blood stages, in vitro ookinetes and A. coluzzii mosquito stages. Gametocyte expressed gene P28 and constitutive expressed GFP served as a stage-specific and loading controls, respectively. Δc43 parasites were used as a negative control. Abbreviations: ABS, asexual blood stages; NGP, non-gametocyte producing; MBS, mixed blood stages; Gc, gametocytes; Gc (+), activated gametocytes; Ook, ookinetes; pbf, post blood feeding; wt, wild-type (c507); ko, Δc43 knockout. Figure S3. Generation of P. berghei Δc43 knock out mutant in the c507 reference line (A) Schematic representation of Pbc43 locus disruption by double crossover homologous recombination. The disruption vector design allows for 50% ko of the coding DNA sequence. Arrows indicate binding sites for primers P5, P6 and P7 used in diagnostic PCR. P5 and P7 were used to detect integration and P5 and P6 that bind to the endogenous PbPIMMS43 locus were used to confirm absence of the endogenous gene in the ko line.
(B-C) Genotypic analysis of Δc43 mutants following transfection and dilution cloning by (B) southern blot analysis on pulsed field gel electrophoresis separated transgenic chromosomes and (C) PCR on mix parasite (MP) and Δc43 clonal populations (CP). Used primer combinations are shown below each detected band. Pfc43 (A) Schematic diagram of the endogenous Pbc43 locus, targeting construct and resulting transgenic Pb Pfc43 locus. Arrows P6, P20 and P21 indicate binding sites for primers used in diagnostic PCR. P20 and P21 were used to detect integration, and P20 and P6 that bind to the endogenous Pbc43 locus were used to confirm absence of the endogenous gene in the transgenic line. (B) Diagnostic PCR to determine integration of the Pfc43 targeting construct into the endogenous Pbc43 locus. The c507 wt parasite served as a control. (C) RT-PCR in ookinetes to confirm expression of Pfc43 in the Pb Pfc43 transgenic line. Amplifications of Pbc43 and CTRP were used as positive control for the endogenous locus and stage-specific control, respectively. (D) Western blot analysis under reducing and no-reducing conditions of whole c507, Δc43 and Pb Pfc43 parasite cell lysates using the α-Pbc43 opt and α-Pfc43 opt antibodies. Pfc43 protein bands are indicated with asterisks. (E) Oocyst density in naïve and LRIM1 kd A. coluzzii infected with the Pb Pfc43 transgenic line. The c507 wt and Δc43 parasites served as controls in naïve mosquito infections. Red horizontal lines indicate median. ***, P<0.0001 with Mann-Whitney test. Figure S5. PIMMS43 localization on the surface of young oocysts and midgut sporozoites in A. coluzzii infected midgut epithelia (A) Immunofluorescence assays of NF54 P. falciparum oocysts found in the mosquito midgut epithelium at 2 dpbf (left) and midgut/oocyst sporozoites (spz) at 11 dpbf (right), stained with α-Pfc43 opt (green), α-Pfs25 (red) and/or α-PfCSP (purple) antibodies. DNA was stained with DAPI. Staining with pre-immune serum was used as a negative control. (B) Immunofluorescence assays of P. berghei c507 parasites found in the mosquito midgut epithelium at 2 dpbf (left) and oocyst sporozoites at 15 dpbf (right), stained with α-Pbc43 opt antibody (green), α-P28 (red) and/or α-PbCSP (purple) antibodies. DNA was stained with DAPI. Staining with pre-immune serum was used as a negative control. Images in both panels are de-convoluted projection of confocal stacks. BF denotes bright field and scale bars correspond to 5 μm. Figure S6. Asexual blood stage (ABS) growth of Δc43 and control c507 parasites ABS growth is measured every 24 hours upon parasite inoculation for 3 consecutive days and expressed as percentage (%) of infected red blood cells (iRBC). The arithmetic mean of two biological replicates and the standard error of the mean (SEM) are shown. Figure S7. Complementation of Δc43 mutant with endogenous PbPIMMS43 wt allele (A) Schematic diagram of the Δc43 ko locus, targeting construct and transgenic Δc43::c43 wt complemented locus. Arrows indicate binding sites of primers used in diagnostic PCR. P20, P6 and P26 detect integration and are used to confirm re-introduction of the wt Pbc43 in the Δc43::c43 wt transgenic line. P20 and P7 bind to the Δc43 ko locus. (B) Diagnostic PCR to determine integration of the targeting construct into the Δc43 ko locus. The c507 wt and Δc43 parasites served as PCR controls. (C) Oocyst density in A. coluzzii infected with the Δc43::c43 wt transgenic line. The c507 wt and Δc43 parasites served as controls. Red horizontal lines indicate the median; ns, not significant; ***, P<0.0001 with Mann-Whitney test. The design allows for 74% removal of the gene coding sequence. Arrows indicate binding sites of primers P8-11 used in diagnostic PCR assays. P10 and P11 were used to detect integration, and P8 and P9 bind to the endogenous locus and were used to confirm absence of the endogenous gene in the ko parasite. (B) Diagnostic PCR to determine integration of the targeting construct into the endogenous locus. (C) Female gametocyte to ookinete conversion rate in vitro (left), and in vivo in the A. coluzzii midgut (middle left) of P. berghei c1804 and mutant Δc43 red lines. Error bars indicate SEM. Representative images of in vivo invading and in vitro produced Δc43 red ookinetes are shown (middle right). The graph on the right shows numbers of c1804 and mutant Δc43 red oocysts in the midguts of A. coluzzii mosquitoes 10 dpbf. (D) Gametocyte to ookinete conversion rate in the midgut bolus (left) and oocyst numbers (right) of c1804 and Δc43 red P. berghei lines in A. stephensi mosquitoes. (E) Sporozoite numbers of c1804 and Δc43 red P. berghei lines in the midgut of A. stephensi mosquitoes. In all the graphs, red lines indicate median, n is the number of independent experiments, ns and *** denote nonsignificant P values and P<0.0001, respectively, assessed with the Mann-Whitney test. Volcano plot of A. coluzzii midgut transcriptional responses to Δc43 vs. c507 wt parasites. X-axes show log2 fold change and y-axes show log10 p-value calculated using one-way ANOVA. Blue and orange filled circles indicate genes that are at least 2-fold down downregulated and 2-fold upregulated, respectively. Black circles show with no significant differential regulation. Known gene names are indicated.

Figure S10. Transcriptional profiles of TEP1, LRIM1 and APL1C in infected A. coluzzii tissues
Relative abundance of TEP1, LRIM1 and APL1C transcripts in the midgut and whole body of A. coluzzii mosquitoes infected with the c507, Δc43 or Δpbp47 parasite lines, measured by qRT-PCR at 24 hpbf. Whole body refers to mosquito tissues after the removal of wings, legs and head. Data are derived from two independent replicates, normalized to the abundance of mosquito S7 transcripts in each of the two tissues and referenced to data obtained at 1 hpbf that was used as baseline. Error bars indicate SEM.    0 0 0 0 0 of 3 Mean oocyst and salivary gland sporozoite numbers from three biological replicates of A. coluzzii infections with Δc43 or c507 parasite lines. For each biological replicate, sporozoite numbers was determined from two or three pools (technical replicates) of ten homogenized mosquito midguts or salivary glands at 15 and 21 days post infection respectively. Data from pooled and independent biological replicates are presented. The data from each technical replicate are also presented in brackets. Infectivity of sporozoites was assessed by infected mosquito bite back experiments with at least 30 mosquitoes on C57/BL6 mice at 21 dpi. Following this, parasitaemia was monitored until 14 days post mosquito bite. SEM represents standard error of mean.   ND ND 0 of 1 I. Mean oocyst and salivary gland sporozoite numbers in LacZ and LRIM1 knockdown A. coluzzii mosquitoes infected with c507 or Δc43 parasites, obtained from 2 to 4 biological replicates. In each replicate, the mean number of sporozoites was calculated from three pools of ten homogenised midguts or salivary glands at 15 and 21 days post blood feeding respectively. Parasite infectivity to mice was assessed upon blood feeding of at least 30 mosquitoes on C57BL/6 mice at 21-22 days post blood feeding. Parasitaemia was monitored until 14 days post mosquito bite. SEM represents the standard error of the mean. II. The mean number of midgut sporozoites in LacZ and LRIM1 knockdown A. coluzzii mosquitoes infected with Pb Pfc43 or c507 parasites was calculated from two replicates obtained from suspensions of 10-12 homogenised midguts at day 15 post blood feeding. Parasitaemia was monitored until 14 days post mosquito bite. 0 (0/0/0) 0 0/6 Mean salivary gland sporozoites at 21 days post A. coluzzii haemocoel inoculation with c507 wt or ∆c43 ookinetes, obtained from 3 biological replicates. Data from independent biological replicates are presented in brackets. Infectivity of sporozoites was assessed by infected mosquito bite back experiments of C57/BL6 mice at day 21 post haemocoel inoculation. Parasitaemia was monitored for 14 days post mosquito bite. SEM represents the standard error of the mean. Oocyst data at 7 days post blood feeding from P. falciparum SMFAs using the α-Pfc43 opt antibodies. P values for infection prevalence were calculated using the Fisher's exact test and P values for infection intensities on the basis of the median number of oocysts was calculated using the Mann-Whitney U test.   CACATATATAATACAGTATGCTACTTTTAC Diagnostic primer Pbc43 WT complement Where appropriate, target restriction sites are shown as underlined italics and restriction site overhangs are also shown. The appropriate restriction enzyme is presented in the description column. F, forward; R, reverse; LIC, ligation independent cloning; IF, in-fusion; INT, integration; WT, wild-type; KO, knockout; UTR, untranslated region. All primers are listed in a 5 to 3 direction.

RT-PCR and quantitative RT-PCR
Total RNA was extracted from P. falciparum and P. berghei parasites using Trizol reagent (ThermoFisher) according to the manufacturer's instructions. cDNA was synthesized using the Primescript Reverse Transcription Kit (Takara) after Turbo DNase (ThermoFisher) treatment. For RT-PCR, the resulting cDNA was used in the PCR of P. falciparum and P. berghei PIMMS43 using gene specific primers (Table S10). P. falciparum Pfs25 and PfCSP, and P. berghei P28 and CTRP served as stage specific controls. Constitutively expressed GFP in P. berghei was used as an internal control. For qRT-PCR, SYBR green (Takara) and gene specific qRT-PCR primers (Table S10) were used according to the manufacturer's guidelines. Expression of PbPIMMS43 was normalized against GFP and expression of TEP1, LRIM1 and APL1C was normalized against S7 using the ∆∆Ct method.

Expression and purification of recombinant PIMMS43 in E. coli
PfPIMMS43 and PbPIMMS43 was codon optimized for expression in E. coli (GeneArt, ThermoFisher) and termed Pfc43 opt and Pbc43 opt , respectively. Pfc43 opt (aa 25-481) and Pbc43 opt (aa 22-327) fragments that both exclude the signal peptide and the C-terminal hydrophobic domain were amplified with primers (Table S10) and cloned into a NotI digested protein expression vector plasmid, pET-32b (which carries N and Cterminal 6xHistidine tags) (Novagen) by In-Fusion Cloning (Takara). Shuffle T7 E. coli cells (NEB) containing the recombinant protein expression plasmid were grown at 30°C and induced with 1 mM isopropyl-1-thio-βd-galactopyranoside at 19°C for 16 h. Cells were harvested and lysed using bugbuster-lysonase (Novagen) containing protease inhibitors (cOmplete EDTA-free, Roche). Cell debris were removed by centrifugation. Both proteins were expressed as His-fusion proteins and hence purified by cobalt affinity chromatography using TALON® metal affinity resin (Takara). The Pfc43 opt recombinant protein was soluble and purified under native conditions in phosphate buffered saline (PBS), pH 7.4. The Pbc43 opt recombinant protein was extracted from inclusion bodies using the Inclusion Body Solubilization Reagent (ThermoFisher). The solubilized protein was then purified under denaturing conditions in 8M urea in PBS, pH 7.4. Refolding of Pbc43 opt was carried out in decreasing concentrations of urea in PBS. Protein samples were analyzed by SDS-PAGE to determine purity prior to their use for immunization in rabbits for the generation of the affinity purified polyclonal antibodies α-Pfc43 opt and α-Pbc43 opt (Eurogentec).

Expression and purification of recombinant PIMMS43 in Sf9 cells
PbPIMMS43 (aa 22-331) that excludes the signal peptide and includes four amino acids of the C-terminal hydrophobic domain was amplified from 24 h in vitro ookinetes cDNA using primers in Table S10. This fragment was cloned by ligation independent cloning into the linearized pIEX-10 EK-LIC vector which carries a C-terminal 10xHis tag (Novagen) to generate pIEX-10: Pbc43-SP/TM. A stable line expressing the recombinant protein was generated by co-transfection of pIEX-10: Pbc43-SP/TM and pIEX-10:Neo plasmid(7) using the Cellfectin® II Reagent (ThermoFisher) according to the manufacturers' guidelines. pIEX-10:Neo plasmid carries the neomycin resistance cassette and provides resistance to the antibiotic G418 (Sigma). The recombinant protein was extracted from cells using lysis buffer (1XPBS, 1% v/v Triton X-100, pH 7.4) containing benzonase (Novagen) and protease inhibitors. The His-tagged recombinant PbPIMMS43 protein was insoluble and extracted by solubilization in 8M urea in PBS, pH 7.4. The protein was purified using TALON® metal affinity resin under denaturing conditions in 8M urea in PBS, pH 7.4. Bound proteins were eluted using denaturing elution buffer. Refolding of Pbc43 was carried out in decreasing concentrations of urea in PBS. Protein samples were analyzed by SDS-PAGE to determine purity prior to their use for immunization in rabbits in the generation of the affinity purified polyclonal antibody α-Pbc43 Sf9 (Eurogentec).

Generation of transgenic parasites
Partial ko of P. berghei c43 CDS was carried out by double crossover homologous recombination in the c507 and 1804cl1 lines. For partial disruption in the c507 line, a 765 bp ApaI/HindIII 5' homology arm and a 528 bp EcorI/BamHI 3' homology arm was amplified from P. berghei 2.34 genomic DNA using the primer pairs P1/P2 and P3/P4 respectively. These fragments were cloned into the pBS-TgDHFR vector which carries a modified Toxoplasma gondii dihydrofolate gene (TgDHFR/TS) cassette that confers resistance to pyrimethamine (12). The targeting cassette was released by ApaI/BamHI digestion and it allows ko of 50% of P. berghei c43 CDS at the 5' region. For partial disruption in the 1804cl1 line, the target vector containing the human DHFR selection cassette was used (kindly provided by plasmoGEM, vector design ID, PbGEM-042760; http://plasmogem.sanger.ac.uk/). hDHFR confers resistance to the drugs pyrimethamine and WR92210. The targeting cassette was released by NotI digestion and allows ko of 74% of PbPIMMS43 CDS leaving a small part of the 3' region of the CDS. To express P. falciparum c43 in P. berghei, the transgenic parasite Pb Pfc43 was created in the c507 line. The Pfc43 replacement construct was generated using the plasmid pL0035 which carries the hDHFR selection cassette (13). A 1.7 kb HindIII/ApaI fragment upstream of the PbPIMMS43 coding DNA sequence was amplified using primer pair P12/P13 and cloned upstream of the pL0035 selectable marker cassette. The 1.5 kb Pfc43 CDS was amplified from cDNA using the ApaI and SacII primer pair P14/P15 and cloned downstream of the 1.7kb fragment. A 518 bp region corresponding to the 3'UTR, downstream of the PbPIMMS43 stop codon, was amplified as a SacII fragment using primers P16/P17 and cloned downstream of the 1.5 kb Pfc43 CDS. Downstream of the pL0035 selectable marker, a 700 bp region corresponding to part of the PbPIMMS43 coding region and part of the 3'UTR was amplified using primers P18/P19 and cloned as a XhoI and SmaI fragment. This replacement vector was linearized by HindIII and SmaI digestion. To re-introduce PbPIMMS43 into the Δc43 ko parasite, the transgenic parasite Δc43::c43 wt was created. A 3.5 kb upstream region that includes the PbPIMMS43 ORF and its 5'UTR and 3'UTR was amplified as a HindIII and SacII fragment using primers P22/P23. A 518 bp downstream region corresponding to the PbPIMMS43 3'UTR was amplified as a XhoI and SmaI fragment using primers P24/P25. These fragments were cloned into the pL0035 vector and served as homology regions for homologous recombination at the Δc43 ko locus in the c507 line. This replacement vector was linearized by HindIII and SmaI digestion. Transfection of linearized constructs, selection of transgenic parasites and clonal selection was carried out as described previously (3).

Genotypic analysis of transgenic parasites
Purified blood stage parasites were obtained after white blood cells removal using hand packed cellulose (Sigma) columns and red blood cell lysis in 0.17M NH4Cl on ice for 20 min. Genomic DNA was extracted from parasites using the DNeasy kit (Qiagen). Detection of successful integration events or maintenance of the unmodified locus was performed by PCR on genomic DNA using primers listed in Table S10. Blood stage parasites within agarose plugs were lysed in lysis buffer (1XTNE, 0.1 M EDTA pH 8.0, 2% (v/v) Sarkosyl, 400μg/mL proteinase K) to release nuclear chromosomes. Southern blot analysis on pulsed field gel electrophoresis separated chromosomes (Run settings: 98 volts, 1-5 mins pulse time for 60 h at 14 °C) was carried out with a probe targeting the TgDHFR/TS-P. berghei DHFR 3'UTR, obtained by HindIII and EcoRV digestion of the pBS-TgDHFR plasmid.

Phenotypic assays
Exflagellation assays were performed as described previously (14). Briefly blood from a high gametocytemia mouse was added in a 1:40 ratio to ookinete medium (RPMI 1640, 20% v/v FBS, 100 μM xanthurenic acid, pH 7.4), and exflagellation was counted in a standard haemocytometer under a light microscope.In vitro conversion assays were performed as previously described (14). For in vivo conversion assays, the blood bolus of 10 mosquitoes at 17-18 hpbf was pelleted, washed in PBS and resuspended in 50 µL of fresh ookinete medium. The suspension was then incubated with a Cy3-labelled 13.1 mouse monoclonal α-P28 (1:50 dilution) for 20 min on ice. The conversion rate was calculated as the percentage of Cy3 positive ookinetes to Cy3 positive macrogametes and ookinetes. Ookinete motility assays were performed as described (15). Briefly, 24 h in vitro ookinete culture was added to Matrigel (BD biosciences), dropped onto a slide, sealed with nail varnish, and allowed to set at RT for 30 min. Time-lapse microscopy (1 frame every 5 seconds, for 10 min) of ookinetes were taken on a Leica DMR fluorescence microscope and a Zeiss Axiocam HRc camera controlled by the Axiovision (Zeiss) software. The speed of individual ookinetes was measured using the manual tracking plugin in the Icy software package (http://icy.bioimageanalysis.org/).

Ookinete injections in mosquito haemocoel
24 h in vitro ookinetes was adjusted with RPMI 1640 to achieve an injection concentration of 800 ookinetes per mosquito as described previously (16). This was injected into the thorax of A. coluzzii mosquitoes using glass capillary needles and the Nanoject II microinjector. Salivary gland sporozoites were counted at 21 dpbf as described below.
Gene silencing cDNA was prepared from total RNA extracted (as described above) from A. coluzzii midgut infected with P. berghei c507, at 24 hpbf. The cDNA was used in the amplification of CTL4, LRIMI and TEPI using primers with T7 overhangs as reported in (17,18). The resulting T7 PCR products and the T7 high yield transcription kit (ThermoFisher) was used to produce dsRNA. DsRNA was purified using the RNeasy kit (Qiagen) and 0.2 μg in 69 nL was injected into the thorax of A. coluzzii mosquitoes using glass capillary needles and the Nanoject II microinjector (Drummond Scientific). Injected mosquitoes were left for 2-3 days before P. berghei infection.

RNA-sequencing library preparation
Three replicate infections of A. coluzzii mosquitoes with the Δc43 and c507 P. berghei lines were performed and infected midguts were dissected at 1 and 24 hpbf. Total RNA was extracted as described elsewhere and was used for RNA sequencing by Genewiz (New Jersey, US) using the NEB Ultra prep kit and an Illumina

NGS RNA-sequencing-Data processing and analysis
RNA-Seq reads were mapped using HiSat2 v2.0.5 (19) with default parameters to the A. gambiae genome (AgamP4 assembly) (20) and the P. berghei ANKA (21). Transcript abundance was quantified as fragments per kilobase per million reads (FPKM) using Cufflinks v2.2.1 (22) on the A. gambiae (Anopheles-gambiae-PEST_BASEFEATURES_AgamP4.9.gtf) and P. berghei (PlasmoDB-39_PbergheiANKA.gff) annotation sets. Differential expression analysis was performed using Cuffdiff v.2.2.1 (23). The sequencing data were uploaded to the Galaxy web platform (an open source, web-based platform for data intensive biomedical research), and we used the VectorBase Galaxy server (https://galaxy.vectorbase.org) to analyze the data (24). Data are derived from three independent biological replicates, each of which included three technical replicates. To filter out the biological or technical noise from the actively expressed genes, an FPKM cutoff was selected that is based on an implementation of the zFPKM normalization method described previously (25). Functional classification of P. berghei differential regulated genes were performed in PlasmoDB (http://plasmodb.org/plasmo/) using the P. berghei full genome as a reference genome. PANTHER (v13.1; http://pantherdb.org) (26) was used for functional classification of A. gambiae differentially regulated genes. The RNA sequencing data were deposited to and can be downloaded from the European Nucleotide Archive with experiment codes ERX3197375-410.

Population genetics analysis
The genome sequences of 1,509 African P. falciparum samples determined in the context of the P. falciparum Community Project were obtained from the MalariaGen website (http://www.malariagen.net/data). They include samples from 11 African countries including Gambia (73)

P. berghei mosquito infections
Mosquitos were infected with P. berghei by utilizing direct feeding assays (DFAs) or standard membrane feeding assays (SMFAs). DFAs were carried out by mosquito feeding on mice with 5-6% parasitaemia and 1-2% gametocytaemia. SMFAs were carried out as described previously (28). Briefly, for each feed, 350 μL of heparanized P. berghei ANKA 2.34 infected blood from mice with 5-6% parasitaemia and 2-3% gametocytaemia was mixed with 150 μL of PBS containing either α-Pbc43 Sf9 or the isotopic monoclonal UPC10 (negative control; Sigma) antibodies to yield final antibody concentrations of 50, 100 and 250 µg/mL. Infected mosquitoes were maintained at 19-21°C, 70-80% humidity, 5% sucrose and 12/12 hours light/dark cycle. Midguts tissues were dissected at 4-10 dpbf and fixed in 4% PFA in PBS. Fixed midguts were mounted in Vectashield® (VectorLabs) and oocysts or melanised ookinetes were enumerated using light and/or fluorescence microscopy. Oocyst images and sizes were also analyzed using fluorescence microscopy. Oocyst and salivary gland sporozoite numbers at 15 and 21 dpbf respectively were counted using a standard haemocytometer, in 3 technical replicates of homogenates of 10 P. berghei infected A. coluzzii midguts or salivary glands. Finally, in mosquito to mouse transmission assays, at least 30 P. berghei infected mosquitoes were allowed to feed on 2-3 anaesthetized C57/BL6 mice at 20-22 dpbf. Parasitaemia was monitored up until 14 days post mosquito bite by Giemsa stained tail blood smears.

P. falciparum mosquito infections
Mosquitos were infected with P. falciparum by SMFAs as described previously (6). Briefly, gametocytemia and density of viable mature stage V gametocytes at day 14 post-induction were assessed by Giemsa stained blood smears and by testing in vitro exflagellation of male gametocytes respectively. Day 14 stage V gametocytes cultures were pooled in a pre-warmed tube containing 20% v/v uninfected serum-free hRBCs and 50% v/v heat-inactivated human serum to a final volume of 300 μL. The mixture was immediately transferred to pre-warmed glass feeders kept a constant temperature of 37°C and mosquitoes were allowed to feed. For transmission blocking assays, α-Pfc43 opt antibodies were added to the gametocytemic blood mix to final antibody concentrations of 50, 125 and 250 µg/mL, in a final volume of 300 μL. Blood fed mosquitoes were maintained at 27°C, 70% humidity, 5% sucrose and 12/12 hours light/dark cycle. Infected midgut tissues at 10 dpbf were dissected and analyzed as above.

Statistical analysis
Statistical analyses for exflagellation, ookinete conversion and motility assays were performed using a twotailed, unpaired Student's t-test. For statistical analyses of the oocyst or melanized parasite load (infection intensity) and presence of oocysts (infection prevalence), p values were calculated using the Mann-Whitney test and the Fisher's exact test, respectively. Statistical analyses were performed using GraphPad Prism v7.0 and v8.0. The generalized linear mixed model (GLMM) was used to also determine statistical significance in oocyst infection intensity in transmission blocking assays. GLMM analyses were performed in R (version 2.15.3) using the Wald Z-test on a zero-inflated negative binomial regression (glmmADMB). The various treatments were considered as covariates and the replicates as a random component. Fixed effect estimates are the regression coefficients.

Supplementary Dataset legends
Dataset S1 (separate file). RNAseq transcription profiling of P. berghei Δc43 and c507 lines in the A. coluzzii midgut Transcription profiles are assessed at (A) 1 hpbf and (B) 24 hpbf. FPKM is transcript abundance quantified as fragments per kilobase per million reads in each of the Δc43 and c507 parasite lines. The log2transformed ratio of Δc43 to c507 FPKM data for each gene is shown (log2 (Δc43/c507)). P values (P value), adjusted P values (q value) and statistical significance (q<0.005; yes or no) are indicated. Data are derived from three independent biological replicates, each of which included three technical replicates.
Dataset S2 (separate file). RNAseq transcription profiling of A. coluzzii midgut responses to P. berghei Δc43 and c507 infections Transcription profiles are assessed at (A) 1 hpbf and (B) 24 hpbf. FPKM is transcript abundance quantified as fragments per kilobase per million reads in each of the A. coluzzii genes infected with Δc43 (Ac Δc43) and c507 (Ac c507) parasite lines. The log2-transformed ratio of FPKM values for each A. coluzzii gene in infections with Δc43 to infections with c507 is shown (log2 (Ac Δc43/Ac c507)). P values (P value), adjusted P values (q value) and statistical significance (q<0.005; yes or no) are indicated. Data are derived from three independent biological replicates, each of which included three technical replicates.