The Toxoplasma Dense Granule Proteins GRA17 and GRA23 Mediate the Movement of Small Molecules between the Host and the Parasitophorous Vacuole

Abstract

Toxoplasma gondii is a protozoan pathogen in the phylum Apicomplexa that resides within an intracellular parasitophorous vacuole (PV) that is selectively permeable to small molecules through unidentified mechanisms. We have identified GRA17 as a Toxoplasma-secreted protein that localizes to the parasitophorous vacuole membrane (PVM) and mediates passive transport of small molecules across the PVM. GRA17 is related to the putative Plasmodium translocon protein EXP2 and conserved across PV-residing Apicomplexa. The PVs of GRA17-deficient parasites have aberrant morphology, reduced permeability to small molecules, and structural instability. GRA17-deficient parasites proliferate slowly and are avirulent in mice. These GRA17-deficient phenotypes are rescued by complementation with Plasmodium EXP2. GRA17 functions synergistically with a related protein, GRA23. Exogenous expression of GRA17 or GRA23 alters the membrane conductance properties of Xenopus oocytes in a manner consistent with a large non-selective pore. Thus, GRA17 and GRA23 provide a molecular basis for PVM permeability and nutrient access.

Figure 1. GRA17 associates with the PVM and affects PV morphology

(A) Human foreskin fibroblasts (HFFs) were infected with the indicated Toxoplasma tachyzoite strains for 24 h, fixed and subjected to IF with the antibodies indicated. The WT+GRA17-V5/GRA23-HA-FLAG also expresses cytosolic GFP. The fluorescent images represent a single deconvoluted focal slice. Brightfield images were taken with differential interference contrast (DIC) microscopy (scale bar = 10 μm). (B) HFFs were infected with WT+GRA17-HA tachyzoites for 24 h, mechanically lysed and a membrane-enriched fraction was treated with either PBS or sodium carbonate and centrifuged to obtain pelleted (P) and soluble (S) fractions and subjected to SDS-PAGE. The fractions were Western blotted as indicated. (C) HFFs were infected with the indicated strains for 36 h and imaged live by phase-contrast microscopy (scale bar = 25 μm).

Figure 2. GRA17 is required for the transfer of small molecules into the PV

(A) Indicated tachyzoite strains endogenously tagged with GRA24-HA-FLAG were used to infect HFFs and the samples were fixed at the indicated time-points, stained with Hoechst dye (red), anti-HA (green), and overlaid onto a DIC image. For each channel, the exposure time is the same. This experiment was performed twice. (B) Sub-confluent HFFs were infected with the indicated strains for 24 h after which they were pulse labeled with CDCFDA (green) for 10 min before the cells were imaged live. White arrows indicate intra-vacuolar fluorescence and the red arrow indicates the absence of intra-vacuolar fluorescence (left panel; scale bars = 10 μm). The percentage of CDCF-fluorescent vacuoles was quantified for each strain and averaged across experiments (right panel). A minimum of 50 vacuoles per strain was counted for each replicate. Error bars represent ±SD and the numbers above the bars of the graph indicate the number of experimental replicates for each condition. Unpaired two-tailed t tests compared to WT (*** P=3×10^-7; ** P=0.002) or ΔGRA17 (### P≤0.00001, ##P=0.01) were performed. (C) Confluent HFFs were infected with the indicated strains for 24 h after which either Lucifer yellow or Dextran3000-Alexa 488 were scrape-loaded into the cells and imaged live immediately. Intra-PV fluorescence was scored relative to the fluorescence of the individual host cell as present, intermediate, or absent as a percentage of total vacuoles (Fig. S3D) and averaged across 3 experimental replicates (Error bars = ±SD).

Figure 3. GRA17 affects the stability of Toxoplasma PVs

(A) Sub-confluent HFFs on were infected with ΔGRA17 tachyzoites and imaged live. Images obtained by phase-contrast microscopy are presented at 30 min time points from 19:30 h through 25:00 h post-infection. Matching colored arrowheads indicate the collapse of indicated PVs in consecutive time points (scale bar = 10 μm). (B) Confluent monolayers of HFFs were infected with 100 tachyzoites of each indicated strain for 4 days. For every experiment, the area of at least 30 plaques from each strain was measured (Fig. S5A) and the mean area was normalized to WT. Each bar of the graph represents an average of means across experimental replicates (n=8 for all conditions except for ΔGRA17+GRA17-HA, where n = 3). Error bars represent ±SD and paired two-tailed t tests were performed comparing each condition to WT to assess statistical significance (***; P<0.0001).

Figure 4. GRA17 affects the virulence and in vivo proliferation of Toxoplasma

(A) Cohorts of five female CD-1 mice (aged 6-8 wks) were each infected with 100 tachyzoites (one additional cohort was infected with 1000 tachyzoites of ΔGRA17) intraperitoneally with each type I (RH) strain and survival was monitored. At day 30, the surviving ΔGRA17 mice from the 100 tachyzoites infection were re-infected with 10⁴ WT tachyzoites and survival was monitored (arrow). (B) Infection of mice was carried out as above and survival was monitored for 34 days post-infection (top panel). The mice were weighed throughout the infection and plotted as an average of the change in body weight for each cohort, where 100% body weight is the day before infection (bottom panel). Error bars indicate +SD and single asterisks indicate P<0.05 statistical significance assessed at that time point compared to ΔGRA17 by a student's t test.

Figure 5. GRA17 and GRA23 are synergistically required for growth

(A) Confluent HFFs were infected with ΔGRA17/ΔGRA23+GRA17-HA or ΔGRA17+GRA17-HA parasites that were transiently co-transformed with Cas9-3×FLAG and either non-specific or GRA17-specific gRNAs along with GRA17 donor oligonucleotide templates containing a nonsense mutation. The infected cells were fixed at 24 h, 48 h, or 120 h at different MOIs and subjected to IFA with anti-HA and anti-FLAG antibodies (sample image, top left panel; scale bar = 10 μm) or anti-HA and anti-SAG1 antibodies (sample image, bottom left panel; scale bar = 25 μm). White asterisks indicate vacuoles where HA-signal is absent. The dashed line distinguishes between the two plaques. (B) The percentage of Cas9-3×FLAG-positive vacuoles and GRA17-HA-positive vacuoles out of the total were scored and averaged (top). In total, 750 vacuoles each were scored across three experimental replicates. The percentage of Cas9-3×FLAG-positive plaques and GRA17-HA-positive plaques out of the total were scored and averaged (bottom panel). The number, n, over each bar graph refers to the total number of plaques scored in total for each condition across three experimental replicates. Error bars represent ±SD. Statistical significance was determined using a two-tailed Fisher's Exact Test comparing the total number of HA-negative vacuoles and plaques for both conditions (***, P<0.0001).

Figure 6. GRA17 or GRA23 alter the membrane conductance properties of Xenopus oocytes

(A) Percentage survival of Xenopus laevis oocytes injected with indicated amounts of mRNA or H₂O in a dose-dependent manner 3 days post-injection. Survival was based on a morphological appearance of the oocytes. Fractions on top of each bar graph represent the number of oocytes that survived out of the total injected. (B) Average resting membrane potentials (RMP) of surviving oocytes. (C) Average membrane currents measured from a holding potential of -90 mV potentials of -120 mV to +50 mV were applied for 2 seconds. Peak currents from each applied voltage were used to generate an I-V relationship. (D) Average conductance (G, bottom) and the reversal potential (Vrev, top) were computed by fitting the I-V relationship as a straight line. For all above sub-figures, data are shown as the mean ±SE. Confidence levels were calculated using Student's paired t-test, where * signifies P<0.05 when compared to H₂O-injected oocytes and # signifies P<0.05 when compared to GRA15-injected oocytes.