A next-generation genetically attenuated Plasmodium falciparum parasite created by triple gene deletion

Abstract

Immunization with live-attenuated Plasmodium sporozoites completely protects against malaria infection. Genetic engineering offers a versatile platform to create live-attenuated sporozoite vaccine candidates. We previously generated a genetically attenuated parasite (GAP) by deleting the P52 and P36 genes in the NF54 wild-type (WT) strain of Plasmodium falciparum (Pf p52(-)/p36(-) GAP). Preclinical assessment of p52(-)/p36(-) GAP in a humanized mouse model indicated an early and severe liver stage growth defect. However, human exposure to >200 Pf p52(-)/p36(-) GAP-infected mosquito bites in a safety trial resulted in peripheral parasitemia in one of six volunteers, revealing that this GAP was incompletely attenuated. We have now created a triple gene deleted GAP by additionally removing the SAP1 gene (Pf p52(-)/p36(-)/sap1(-) GAP) and employed flippase (FLP)/flippase recognition target (FRT) recombination for drug selectable marker cassette removal. This next-generation GAP was indistinguishable from WT parasites in blood stage and mosquito stage development. Using an improved humanized mouse model transplanted with human hepatocytes and human red blood cells, we show that despite a high-dose sporozoite challenge, Pf p52(-)/p36(-)/sap1(-) GAP did not transition to blood stage infection and appeared to be completely attenuated. Thus, clinical testing of Pf p52(-)/p36(-)/sap1(-) GAP assessing safety, immunogenicity, and efficacy against sporozoite challenge is warranted.

Figure 1

Strategy for targeted gene deletion of Pf SAP1, characterization, and phenotypic analysis of the Pf p52⁻/p36⁻/sap1⁻3KO. (a) Schematic of strategy for deleting the SAP1 gene in a Pf p52^-/p36^- KO clone lacking the HsDHFR marker. Enzymes and probes for Southern blotting and primers for polymerase chain reaction (PCR) are shown. Sizes of genomic DNA fragments and PCR products are indicated in kilobases. (b) Southern blotting (left and middle panels) and PCR (right panel) to show deletion of SAP1 in a Pf p52⁻/p36⁻ KO clone (Pf p52⁻/p36⁻/sap1⁻ 3KO). AflIII-digested genomic DNA was hybridized with a 3′ probe yielding a 4.3 and 3.2 kb band for wild type (WT) and 3KO DNA, respectively. Hybridization of EcoRV-digested genomic DNA with a 5′ probe detected a 5.2 and 0.9 kb band for WT and 3KO DNA, respectively. PCR on genomic DNA using primers p535 and p536 yielded a 3.4 and 4.1 kb band for WT and 3KO DNA, respectively. (c) Comparison of asexual blood stage growth rates, as measured by increase in parasitemia over time, between WT and 3KO. Cultures were initiated at 0.5% parasitemia and analyzed daily by Giemsa-stained thin blood smears until day 5. Growth assays were performed in triplicate for each line and parasitemia plotted as mean ± SEM. Mann–Whitney U-test was used for statistical analysis. (d) Comparison of % gametocytemia between WT and 3KO cultures at the time of feeding in vitro gametocyte cultures to mosquitoes (14–16 days postinitiation of in vitro gametocyte cultures). Gametocytemia was determined three independent times in duplicate cultures for each line and plotted as mean ± SEM. Mann–Whitney U-test was used for statistical analysis. (e) Comparison of number of male gamete exflagellation events between WT and 3KO gametocyte cultures at the time of feeding to mosquitoes. Exflagellation (plotted as mean ± SEM) was analyzed in several microscopic fields four independent times in duplicate and triplicate cultures for WT and 3KO, respectively. Mann–Whitney U-test was used for statistical analysis.

Figure 2

Mosquito stage development, and in vitro sporozoite host cell traversal and invasion assays. (a) Prevalence of oocysts in mosquito midguts infected with WT and 3KO parasites. Oocyst prevalence (plotted as mean ± SEM) was calculated by dividing the number of dissected midguts containing oocycts by the total number of midguts on day 7 postfeeding of mosquitoes with in vitro gametocyte cultures. Prevalence was determined in several mosquitoes four independent times in duplicate and quadruplicate for WT and 3KO, respectively. Mann–Whitney U-test was used for statistical analysis. (b) Comparison of average number of oocysts per mosquito midgut (plotted as mean ± SEM) between WT and 3KO on day 7 postfeeding of mosquitoes. Oocyts numbers were determined in several mosquitoes four independent times in duplicate and quadruplicate for WT and 3KO, respectively. Mann–Whitney U-test was used for statistical analysis. (c) Comparison of average number of sporozoites per mosquito (plotted as mean ± SEM) between WT and 3KO on 14–16 days postfeeding of mosquitoes. Sporozoite numbers were determined three independent times in at least duplicate for each line. Mann–Whitney U-test was used for statistical analysis. (d) Staining of CSP trails using Alexa 488-conjugated anti-PfCS 2A10 antibody in motility assays of salivary gland sporozoites from WT and 3KO. Sporozoites were collected 14–16 days postfeeding of mosquitoes. (e) Average total of HC-04 cells in traversal assays with WT and 3KO salivary gland sporozoites (plotted as mean ± SEM) as measured by the fraction of total HC-04 cells in the sample that had taken up FITC-dextran. Total dextran positive cells were determined four independent times in duplicate for each line. Mann–Whitney U-test was used for statistical analysis. (f) Average total infection of HC-04 cells by WT and 3KO salivary gland sporozoites (plotted as mean ± SEM) as measured by the fraction of total HC-04 cells in the sample that were positive for intracellular parasites as measured by CS staining. Total CS-positive cells were determined four independent times in duplicate for each line. Mann–Whitney U-test was used for statistical analysis CS, circumsporozoite; FITC, fluorescein isothiocyanate.

Figure 3

Phenotypic analysis of Pf p52⁻/p36⁻/sap1⁻3KO (-H) clones. (a) Comparison of asexual blood stage growth rates between wild type (WT) and 3KO (-H) clones. Cultures were initiated at 0.5% parasitemia and analyzed daily until day 5. Growth assays were performed in triplicate for each line and parasitemia plotted as mean ± SEM. Mann–Whitney U-test was used for statistical analysis. (b) Comparison of average number of sporozoites (plotted as mean ± SEM) per mosquito between WT and 3KO (-H) clones 14–16 days postfeeding of mosquitoes. Sporozoite numbers were determined three independent times in at least duplicate for each line. Mann–Whitney U-test was used for statistical analysis. (c) Staining of CS trails using Alexa Fluor 488-conjugated anti-PfCSP 2A10 antibody in motility assays of salivary gland sporozoites of WT and 3KO (-H) clones.