Host cell-specific metabolism of linoleic acid controls Toxoplasma gondii growth in cell culture

Abstract

The obligate intracellular parasite Toxoplasma gondii can infect and replicate in any warm-blooded cell tested to date, but much of our knowledge about T. gondii cell biology comes from just one host cell type: human foreskin fibroblasts (HFFs). To expand our knowledge of host-parasite lipid interactions, we studied T. gondii in intestinal epithelial cells, the first site of host-parasite contact following oral infection and the exclusive site of parasite sexual development in feline hosts. We found that highly metabolic Caco-2 cells are permissive to T. gondii growth even when treated with high levels of linoleic acid (LA), a polyunsaturated fatty acid (PUFA) that kills parasites in HFFs. Caco-2 cells appear to sequester LA away from the parasite, preventing membrane disruptions and lipotoxicity that characterize LA-induced parasite death in HFFs. Our work is an important step toward understanding host-parasite interactions in feline intestinal epithelial cells, an understudied but important cell type in the T. gondii life cycle.

Keywords: Toxoplasma gondii; enterocytes; linoleic acid; lipid droplet; lipotoxicity; polyunsaturated fatty acid.

Fig 1

Linoleic acid slows parasite growth in HFF cells but not in Caco-2 cells. (A) T. gondii abundance after 4 days of treatment with 350 µM OA, 350 µM LA, 500 nM DHC, or an equal volume of bovine serum albumin (BSA) (final concentration 75 µM) in confluent HFFs or Caco-2 cells. After fixing and staining parasites red using immunofluorescence, 16 technical replicate images for each of three biological replicate wells were collected at 20 × magnification on the IncuCyte imaging system. The red fluorescence area was calculated for each technical replicate and averaged to obtain a value for each biological replicate. The mean of biological replicates is displayed on the y-axis ± SE. Two-way ANOVA was used to detect differences of means due to treatment, cell type, and the interaction of the treatment and cell type (see Table S1). Post-hoc Dunnett’s test was used to compare each treatment’s mean to the control (BSA) mean. Significance is shown in the figure. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. One representative experiment of two is shown. (B) T. gondii replication assessed by parasitophorous vacuole (PV) size after 24 hours of treatments noted in 1 a. Number of parasites per vacuole is shown as mean percentages of total PVs ± SE. At least 100 vacuoles were counted per biological replicate. One representative experiment of two is shown, with three biological replicate wells per condition. A mixed-effects Poisson regression was used to compare mean vacuole sizes of treated parasites to the control treatment within each host cell type. Means and 95% confidence intervals are shown in Table S2. Significance values of post-hoc Dunnett’s test comparisons are shown in the figure. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Fig 2

Linoleic acid is parasiticidal in HFFs. (A) RH-mCherry T. gondii abundance after 4.5 days of treatment with LA at indicated doses in confluent HFFs. Time 0 images were collected immediately following addition of LA at indicated micromolar doses. Sixteen technical replicate images for each of four biological replicate wells were collected at 20 × magnification on the IncuCyte imaging system every 12–24 hours. The red fluorescence area was calculated for each technical replicate and averaged to obtain values for each biological replicate. The mean of the log₂-transformed biological replicate values is displayed on the y-axis ± SE. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by Student’s t-test. (B) Red fluorescent images from 2 a are shown at 24-hour increments up to 4 days post-treatment, and phase images from 2 a are shown at 4 days post-treatment. (C) Representative images of tachyzoite-stage (left, red) and bradyzoite-stage (right, green) EGS DoubleCat T. gondii, IncuCyte imaging system, 20 × magnification. (D) EGS DoubleCat T. gondii abundance after 6 days of treatment with LA at indicated doses (µM) in confluent HFFs. 37°C = tachyzoite-promoting conditions with CO₂ and neutral pH. 33°C = bradyzoite-promoting conditions without CO₂ and pH 8.0 growth media. Sixteen technical replicate images for each of four biological replicate wells were collected at 20 × magnification on the IncuCyte imaging system every 12–24 hours. Areas of either the sum of red and green (top) or green divided by the sum of red and green (bottom) were computed and averaged by biological replicate. Either log₂-transformed means ± SE (top) or mean proportions ± SE (bottom) of biological replicates are displayed on the y-axis. *, P < 0.05; **, P < 0.01; ****, P < 0.0001 by Student’s t-test.

Fig 3

In HFFs, linoleic acid reduces parasite size and disrupts parasite membranes and increases parasite lipid droplets. (A) Representative immunofluorescence microscopy images of ME49 ∆hpt luciferase, Pru Cre-mCherry, and RH-GFP parasite vacuoles after 24 hours of treatment with 350 µM OA, 350 µM LA, or an equal volume of BSA (final concentration 75 µM) in confluent HFFs.. Monolayers were fixed with methanol to quench endogenous fluorescence and stained with a polyclonal anti-T. gondii antibody (magenta) and a mouse monoclonal anti-SAG1 antibody (yellow), then counterstained with DAPI (cyan). (B) Quantification of parasite membrane disruption as assayed by percent of parasites with anti-T gondii staining outside the SAG1-demarcated parasite membrane. (C) Quantification of parasite length and width in microns. One-way ANOVA was used to detect differences of means due to treatment. Post-hoc Tukey’s test was used to compare treatments. Significance is shown in the figure. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Fig 4

In HFFs, linoleic acid (LA) induces lipid droplet formation in T. gondii vacuoles. (A) ME49 ∆hpt luciferase tachyzoites were treated as in Fig. 3, except cultures were fixed in paraformaldehyde (PFA) to preserve lipid droplet integrity. Magenta = T. gondii polyclonal antibody, yellow = lipid droplets, BODIPY 493/503, and cyan = DAPI. Scalebars are applicable to all photos, as all images were collected at the same magnification. (B) Lipid droplets per parasite from (A) were counted and plotted as percent of total parasites ± SE. At least 100 parasites were counted for each of three biological replicate wells. One representative experiment of two is shown. ***, P < 0.001 by Student’s t-test.

Fig 5

High linoleic acid metabolism in Caco-2 cells influences parasite load. (A) LA quantitation in HFF and Caco-2 supernatants, 24 hours post-treatment with 350 µM LA. Values (n = 4 per condition, shown as dots) are expressed in parts per million (ppm). Two preparations of LA were tested in separate experiments on the same mass spectrometry instrument: Sigma (left) and Nu-Chek Prep (right). (B) Nu-Chek Prep LA-treated host cell pellets were sent for lipidomics analysis (n = 4 per condition). Raw metabolite intensities were log₁₀-transformed. Individual dots represent outliers. ***, P < 0.001; ****, P < 0.0001 by Student’s t-test. (C) Old and young cultures of Caco-2 and A549 cells and HFFs (see Materials and Methods) were infected with Pru Cre mCherry tachyzoites and grown in the presence of indicated doses of LA for 8 days. Sixteen technical replicate images for four biological replicate wells per condition were collected at 8 days post-treatment at 20 × magnification on the IncuCyte imaging system. The red fluorescence area was calculated for each technical replicate and averaged to obtain a value for each biological replicate. The y-axis shows the mean ± SE of biological replicates, expressed as the percent of the mean of BSA-treated control wells from the same host cell plate. (D) Pru Cre mCherry tachyzoites were grown in the presence of indicated doses of LA in Caco-2 cells that were treated with either DMSO or mitomycin C (MMC) 1 week prior to infection. Sixteen technical replicate images for four biological replicate wells per condition were collected at 9 days post-treatment at 20 × magnification on the IncuCyte imaging system. The red fluorescence area was calculated for each technical replicate and averaged to obtain a value for each biological replicate. The y-axis shows the mean ± SE of biological replicates, expressed as the percent of the mean of BSA-treated control wells. One representative experiment of two. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by Student’s t-test.