Malaria parasite clag3 genes determine channel-mediated nutrient uptake by infected red blood cells

Abstract

Development of malaria parasites within vertebrate erythrocytes requires nutrient uptake at the host cell membrane. The plasmodial surface anion channel (PSAC) mediates this transport and is an antimalarial target, but its molecular basis is unknown. We report a parasite gene family responsible for PSAC activity. We used high-throughput screening for nutrient uptake inhibitors to identify a compound highly specific for channels from the Dd2 line of the human pathogen P. falciparum. Inheritance of this compound's affinity in a Dd2 × HB3 genetic cross maps to a single parasite locus on chromosome 3. DNA transfection and in vitro selections indicate that PSAC-inhibitor interactions are encoded by two clag3 genes previously assumed to function in cytoadherence. These genes are conserved in plasmodia, exhibit expression switching, and encode an integral protein on the host membrane, as predicted by functional studies. This protein increases host cell permeability to diverse solutes.

Figure 1. High-throughput screen and a Dd2-specific inhibitor.

(A) All-points histogram of mean % block at 2h for 48,157 compounds screened against sorbitol uptake by red blood cells infected with HB3 or Dd2 lines, calculated according to Eq. S2. Red shading reflects compounds having % block ≥ 50, corresponding to inhibitory K_0.5 values ≤ 1 μM (Pillai et al., 2010). (B) All-points histogram of the normalized statistic WDS for these compounds (Eq. S3). Most compounds have WDS < 1.0, indicating negligible difference in activity against channels on HB3 and Dd2. (C) Structure of ISPA-28, a Dd2-specific inhibitor (WDS = 5.6; see also Figure S1). (D) Continuous recordings of osmotic lysis kinetics in sorbitol for indicated parasites without or with 7 μM ISPA-28 (black and red traces, respectively). (E) Dose responses for inhibition of organic solute permeability (P) into HB3- and Dd2-infected cells (circles and triangles, respectively). Symbols represent the mean ± S.E.M. of up to 8 measurements each. Solid lines represent best fits to Eq. S1.

Figure 2. ISPA-28 interacts directly with PSAC to block transport.

(A) Whole-cell patch-clamp recordings on erythrocytes infected with indicated parasite lines without or with addition of 10 μM ISPA-28 to bath and pipette solutions. Ensemble traces represent responses to pulses of −100 to +100 mV (10 mV increments). Horizontal and vertical scale bars represent 20 ms and 5 nA, respectively. (B) Current-voltage relationships showing mean ± S.E.M. whole-cell currents determined from up to 28 cells each. ** P < 10⁻⁹. (C) Single PSAC recordings on HB3- or Dd2-infected erythrocytes without inhibitor (left and right columns, respectively). (D) Single channel recordings from these lines with 10 μM ISPA-28 in bath and pipette solutions. (E) Resolved intrinsic gating in the presence of ISPA-28, expanded from the bottom traces in panel D as indicated with red dashes. Black dashes adjacent to each trace represent the closed channel level. Horizontal scale bar = 100 ms, 2 s, or 74 ms (panels C, D, and E respectively); vertical scale bar = 3 pA for all panels. All whole-cell and cell-attached recordings used symmetric bath and pipette solutions of 1 M choline chloride, 115 mM NaCl, 10 mM MgCl₂, 5 mM CaCl₂, 20 mM Na-HEPES, pH 7.4. (F) Mean ± S.E.M. single channel open probability for indicated lines in the absence or presence of 10 μM ISPA-28. ** P < 0.0002.

Figure 3. Inheritance of ISPA-28 affinity in a genetic cross and QTL analysis.

(A) Mean ± S.E.M. block of sorbitol uptake by 7 μM ISPA-28 for indicated parental lines and progeny clones at a 60 min timepoint according to Eq. S2. (B) Logarithm of odds (LOD) scores from a primary scan of QTL associated with ISPA-28 efficacy. The peak (LOD of 12.6) maps to the 5′ end of chromosome 3. Dashed line is the 0.05 significance threshold calculated from 1000 permutations.

Figure 4. Complementation implicates two clag genes.

(A) Schematic shows the plasmid maps for the translocon vector into which genes from the HB3 line were cloned (left) and the helper plasmid (right). (B) Mean ± S.E.M. ISPA-28 K_0.5 values for Dd2 without or after complementation with the indicated genes from HB3. * indicates confirmed expression of the HB3 allele in the transfectant. (C) Osmotic lysis kinetics showing significant decrease in ISPA-28 efficacy after transfection with PFC0120w. [ISPA-28] in micromolar as indicated.

Figure 5. Allelic exchange yields an intermediate phenotype.

(A) Schematic shows the integration plasmid (top) and the result of single homologous recombination between the plasmid’s Dd2 clag3.1 fragment and genomic clag3.2 of HB3. (B) Sorbitol-induced osmotic lysis kinetics for the allelic exchange clone HB3^3rec with indicated [ISPA-28]. (C) Mean ± S.E.M. ISPA-28 dose-response for HB3^3rec (circles). This dose response is intermediate between those of HB3 and Dd2 (top and bottom solid lines, respectively; taken from fits in Figure 1E).

Figure 6. Relative expression of clag3 alleles determines ISPA-28 affinity.

(A) ISPA-28 dose responses for indicated progeny (symbols) and the expected efficacy based on inheritance of the locus from Dd2 (solid line). (B) Ethidium-stained gels showing relative expression of clag3 alleles in indicated lines. Amplicons represent full-length cDNA. (C-D) Transport inhibition dose responses after in vitro selections with sorbitol and ISPA-28. Selection does not change HB3’s dose response, but improves that of HB3^3rec and indicated progeny to levels matching Dd2’s. Solid lines are from Figure 1E. (E) Expression patterns for clag genes in indicated lines before and after in vitro selection, determined by quantitative RT-PCR. Mean ± S.E.M. of replicates. (F) clag3.1: clag3.2 expression ratio after selections of indicated lines, showing that lines isogenic with Dd2 at the clag3 locus are strongly purified, while HB3 is not. Mean ± S.E.M. values calculated from up to 4 RNA harvests each. (G) Kinetics of decrease in ISPA-28 efficacy for HB3^3rec upon continuous in vitro culture after end of ISPA-28 selection. (H) Kinetics of decreasing expression of the chimeric clag3.2_HB3-3.1_Dd2 transgene in HB3^3rec after cessation of ISPA-28 selection, presented as a ratio relative to that of the endogenous clag3.1. (I) Structure of ISPA-43. See also Figure S6. (J) Reversed selection using ISPA-43. 7C20 culture previously selected for clag3.1 using ISPA-28 (panel C) was subjected to sorbitol synchronizations in the presence of ISPA-43. Graphic shows ISPA-28 dose response (circles, mean ± S.E.M.) and that of the parental Dd2 (solid line). (K) qRT-PCR expression profile for 7C20 after reversed selection using ISPA-43.

Figure 7. clag3 mutant analysis and host membrane localization.

(A) Sequence chromatograms from wild-type HB3 and mutant HB3-leuR1 showing the point mutation (red arrow) that changes A1210 to threonine. (B) Immunoblot of total lysate from HB3-infected cells without or with protease treatment of intact cells. Blot was probed with a polyclonal antibody to a recombinant clag3 fragment. (C) Control immunoblot showing that the intracellular parasite protein KAHRP is not hydrolysed by protease treatment. (D) Osmotic lysis kinetics without or with protease treatment, showing that PSAC activity is degraded by pronase E. Addition of protease inhibitors prevents loss of channel activity (pronase E + pi, green trace). (E) Immunoblot showing that the protein is membrane-associated, partially extractable with carbonate treatment (CO₃⁼), and that the protease hydrolysis fragment is an integral membrane protein (membr). (F) Immunoblot of carbonate-extracted membranes from HB3 and HB3^3rec without or with protease treatment as indicated. Blot was probed with anti-FLAG tag antibody, which does not recognize HB3 and is specific for the recombinant clag3 product. (G) Schematic shows two models for the role of the clag3 product in PSAC formation. A pronase E cleavage site, a known hypervariable domain, and the site of the HB3-leuR1 mutation are indicated. Our studies support an intracellular C-terminus; the transmembrane topology shown is speculative and based on computational analysis. The green box (right panel) represents additional parasite or host protein(s) that may contribute to channel activity.