A set of glycosylphosphatidyl inositol-anchored membrane proteins of Plasmodium falciparum is refractory to genetic deletion

Abstract

Targeted gene disruption has proved to be a powerful approach for studying the function of important ligands involved in erythrocyte invasion by the extracellular merozoite form of the human malaria parasite, Plasmodium falciparum. Merozoite invasion proceeds via a number of seemingly independent alternate pathways, such that entry can proceed with parasites lacking particular ligand-receptor interactions. To date, most focus in this regard has been on single-pass (type 1) membrane proteins that reside in the secretory organelles. Another class of merozoite proteins likely to include ligands for erythrocyte receptors are the glycosylphosphatidyl inositol (GPI)-anchored membrane proteins that coat the parasite surface and/or reside in the apical organelles. Several of these are prominent vaccine candidates, although their functions remain unknown. Here, we systematically attempted to disrupt the genes encoding seven of the known GPI-anchored merozoite proteins of P. falciparum by using a double-crossover gene-targeting approach. Surprisingly, and in apparent contrast to other merozoite antigen classes, most of the genes (six of seven) encoding GPI-anchored merozoite proteins are refractory to genetic deletion, with the exception being the gene encoding merozoite surface protein 5 (MSP-5). No distinguishable growth rate or invasion pathway phenotype was detected for the msp-5 knockout line, although its presence as a surface-localized protein was confirmed.

FIG. 1.

Possible genetic outcomes of following negative selection of parasites transfected with gene knockout constructs. Four distinct genetic outcomes are possible during drug cycling with double recombination vectors, including retention of the episomal form as a concatemer, 5′ or 3′ single integration of the vector, double integration representing a targeted genetic deletion, and the more rare event of nonhomologous recombination. WT, wild type.

FIG. 2.

Southern blot hybridizations of pulsed-field gels across drug-cycled transfected lines. (A) 3D7-TKMSP4/5 chromosomes probed with hDHFR and pGem probes across four drug cycles. (B) 3D7-ΔMSP5 chromosomes across four drug cycles also probed in a Southern blot with hDHFR and pGem, indicating the loss of msp-5 in late drug cycles. A 3D7 genomic control was loaded in the leftmost lane of every blot, and the position of migration of chromosome 2 is denoted by the bars. Identical blots were hybridized with the hDHFR probe (positive selectable marker) and the pGEM probe (plasmid backbone). Loss of chromosome 2 signal (arrow) indicates double crossover integration. C, cycle.

FIG. 3.

Mapping of the integration events in MSP-5 knockout and MSP-4 truncated 5′ UTR lines. (A) Schematic diagram detailing the strategy employed in targeting of the endogenous msp-5 gene. The black boxes represent sequences encoding the signal peptides and the GPI attachment moieties. The central repeat regions that define the MSP-2 alleles are also indicated by the black boxes toward the center of the MSP-2 coding sequence. Gray boxes represent the EGF-like domains. (B) Southern blot of restricted genomic DNA isolated from P. falciparum 3D7 parental parasites and a number of 3D7-ΔMSP5 clones (cycles 3 and 4 [C3 and C4]). Both blots (ScaI-MscI and BbvI-EcoNi) were probed with the MSP-5 3′ targeting sequence (F2) delineated in panel A. (C) Schematic diagram detailing the outcome of an attempt to target PfMSP-4 and PfMSP-5 in the one transfected line, resulting in a 5′ single recombination event integrating one or more copies of the episome. (D) Southern blot of restricted genomic DNA isolated from P. falciparum 3D7 parental parasites and two clones of 3D7-TKMSP4/5 parasites from late drug cycles (C4). Blot SacII-AflIII was probed with the MSP-4 5′ targeting sequence (F1), and blot XbaI-FauI was probed with the MSP-5 3′ targeting sequence (F2) delineated in panel C. W.T., wild type; Cl, clone.

FIG. 4.

Western blot analysis of MSP-5 knockout and MSP-4 truncated 5′ UTR lines. (A) Western blots of total material from synchronized late-blood-stage parasites comparing the 3D7 wild type with the 3D7-ΔMSP5 line across progressive WR99210 drug cycles, including lines treated with both WR99210 (WR) and ganciclovir (Gan) and three late-drug-cycle 3D7-ΔMSP5 clones probed with rabbit polyclonal anti-MSP-5 C-terminal antibodies (37). The panels below represent a loading control probed with rabbit polyclonal anti-MSP-1₁₉ antibodies (18). (B) Western blots of total material from parental 3D7 (wild type [W.T.]) and 3D7-TKMSP4/5-synchronized parasites at 32, 40, and 48 h postinvasion. Blots were probed with a rabbit polyclonal anti-MSP-4 antibody to confirm MSP-4 expression in 3D7-TKMSP4/5 and with rabbit polyclonal anti-SERA5 antibody (29) as a loading control. C, cycle; cl, clone.

FIG. 5.

Confirmation that P. falciparum MSP-5 localizes to the merozoite surface. Fixed 3D7 (wild type [WT]) or 3D7-ΔMSP5 (ΔMSP5) parasites were analyzed by double-label immunofluorescence with a polyclonal rabbit (r) anti-MSP5 antibody (green) and a mouse (m) monoclonal anti-MSP-1₁₉ antibody (4H9/19 [10]) (red). Colocalization of MSP-1 and MSP-5 is evident upon merging of images. DAPI was used for nuclear staining (blue).

FIG. 6.

Disruption of MSP-5 does not affect parasite growth or result in a switch in invasion pathways. (A) Results of an extended growth rate assay comparing 3D7 wild-type parasites with the 3D7-ΔMSP5 line over a 14-day period. Parasites were subcultured 1 in 5 every 48 h. Here, parasitemia levels are compared on days 1, 7, and 14. (B) Invasion into enzyme-treated erythrocytes by 3D7 wild-type and 3D7-ΔMSP5 parasites. Erythrocyte treatments include NaHCO₃ (0.2%)-buffered RPMI-HEPES as a control or buffered media supplemented with 0.066 U/ml neuraminidase (Nm), 0.9 mg/ml trypsin (Trp), 0.9 mg/ml chymotrypsin (CTrp), 0.066 U/ml neuraminidase plus 0.9 mg/ml trypsin (Nm+Trp), and 0.5 mg/ml trypsin inhibitor (Trp Inh) as a control.