Toxoplasma gondii inhibits the expression of autophagy-related genes through AKT-dependent inactivation of the transcription factor FOXO3a

Abstract

The intracellular parasite Toxoplasma gondii induces host AKT activation to prevent autophagy-mediated clearance; however, the molecular underpinnings are not fully understood. Autophagy can be negatively regulated through AKT-sensitive phosphorylation and nuclear export of the transcription factor Forkhead box O3a (FOXO3a). Using a combination of pharmacological and genetic approaches, herein we investigated whether T. gondii hinders host autophagy through AKT-dependent inactivation of FOXO3a. We found that infection by type I and II strains of T. gondii promotes gradual and sustained AKT-dependent phosphorylation of FOXO3a at residues S253 and T32 in human foreskin fibroblasts (HFF) and murine 3T3 fibroblasts. Mechanistically, AKT-sensitive phosphorylation of FOXO3a by T. gondii required live infection and the activity of PI3K but was independent of the plasma membrane receptor EGFR and the kinase PKCα. Phosphorylation of FOXO3a at AKT-sensitive residues was paralleled by its nuclear exclusion in T. gondii-infected HFF. Importantly, the parasite was unable to drive cytoplasmic localization of FOXO3a upon pharmacological blockade of AKT or overexpression of an AKT-insensitive mutant form of FOXO3a. Transcription of a subset of bona fide autophagy-related targets of FOXO3a was reduced during T. gondii infection in an AKT-dependent fashion. However, parasite-directed repression of autophagy-related genes was AKT-resistant in cells deficient in FOXO3a. Consistent with this, T. gondii failed to inhibit the recruitment of acidic organelles and LC3, an autophagy marker, to the parasitophorous vacuole upon chemically or genetically induced nuclear retention of FOXO3a. In all, we provide evidence that T. gondii suppresses FOXO3a-regulated transcriptional programs to prevent autophagy-mediated killing. IMPORTANCE The parasite Toxoplasma gondii is the etiological agent of toxoplasmosis, an opportunistic infection commonly transmitted by ingestion of contaminated food or water. To date, no effective vaccines in humans have been developed and no promising drugs are available to treat chronic infection or prevent congenital infection. T. gondii targets numerous host cell processes to establish a favorable replicative niche. Of note, T. gondii activates the host AKT signaling pathway to prevent autophagy-mediated killing. Herein, we report that T. gondii inhibits FOXO3a, a transcription factor that regulates the expression of autophagy-related genes, through AKT-dependent phosphorylation. The parasite's ability to block the recruitment of the autophagy machinery to the parasitophorous vacuole is impeded upon pharmacological inhibition of AKT or overexpression of an AKT-insensitive form of FOXO3a. Thus, our study provides greater granularity in the role of FOXO3a during infection and reinforces the potential of targeting autophagy as a therapeutic strategy against T. gondii.

Keywords: AKT; FOXO3a; Toxoplasma gondii; autophagy; host response; host-pathogen interactions; transcriptional regulation.

Fig 1

T. gondii induces phosphorylation of host FOXO3a and promotes its own replication within HFF in an AKT-dependent fashion. (A) HFF cultures were inoculated with either RH or ME49 T. gondii tachyzoites or left uninfected for the indicated times and then processed for Western blot analyses. (B, C) HFF cultures were pretreated with 2 µM MK-2206 or an equivalent volume of vehicle (i.e., DMSO) for 1 h and then inoculated with RH T. gondii parasites. Cells were cultured up to 32 h following infection in the presence or absence of MK-2206. (A, B) Phosphorylation and expression levels of indicated proteins were monitored by Western blotting. Total amounts of β-actin were used as a loading control, and an antibody raised against T. gondii profilin-like protein was used to assess infection of HFF cultures. Total protein extracts from extracellular tachyzoites (Tg only) were used to control for any cross-reactivity of the antibodies against T. gondii proteins. Data are representative of at least three independent experiments (i.e., performed on different days). (C) Cultures were fixed post infection and then processed for epifluorescence microscopy analyses. The number of parasites per PV in at least 50 infected cells in different fields of view was enumerated. Data collected from two independent experiments were compiled. **P < 0.01.

Fig 2

T. gondii infection leads to AKT-dependent FOXO3a export from the host nucleus. (A, B) HFF cultures were inoculated with RH T. gondii tachyzoites or left uninfected and fixed at the indicated times and processed for confocal immunofluorescence microscopy. (C, D) HFF cultures were pretreated with 2 µM MK-2206 or an equivalent volume of vehicle (i.e., DMSO) for 1 h and then inoculated with RH T. gondii parasites. Cells were cultured up to 32 h following infection in the presence and absence of MK-2206. (A, C) Samples were stained with DAPI (shown in blue), used as a nuclear marker, and for total FOXO3a (shown in white). Images are representative of at least two independent experiments. Original magnification (left panels) and four times enlarged insets (right panels). Parasitophorous vacuoles (PVs) are outlined with dashed lines to indicate the presence of parasites within infected cells. (B, D) Co-localization of FOXO3a and DAPI was quantified using the Pearson R coefficient. Data are compilated from two independent experiments (n = 2) in which at least 25 cells were analyzed in different fields of view. Each data point represents the Pearson R coefficient of a single cell. ****P < 0.0001; ns, not significant.

Fig 3

T. gondii-driven phosphorylation and nuclear export of FOXO3a require live infection and EGFR-independent PI3K-AKT signaling. (A) HFF cultures were inoculated with live or heat-killed (HK) RH or ME49 T. gondii tachyzoites, treated with 50 µg/mL STAg concentration, or left uninfected and untreated for 24 h. Phosphorylation and expression levels of indicated proteins were monitored by Western blotting. (B) HFF cultures were pre-treated with 1 µM AG1478 or an equivalent volume of vehicle (i.e., DMSO) for 1 h. Then, cultures were inoculated with RH T. gondii tachyzoites or left uninfected. Samples were collected at the indicated times following inoculation and processed for Western blotting analyses. As a positive control for the induced phosphorylation of EGFR, cells were treated with 100 ng/mL recombinant human EGF for 10 min. (C) HFF cultures were inoculated with RH or ME49 T. gondii tachyzoites or left uninfected for 4 h. Then, cultures were treated with the indicated inhibitors (1 µM AG1468, 20 µM LY294002, 2 µM MK-2206, 1 µM Gö6976, and 20 nM rapamycin) or an equivalent volume of DMSO (i.e., vehicle) for 20 h. Phosphorylation and expression levels of indicated proteins were monitored by Western blotting. (D) HFF cultures were inoculated with RH T. gondii tachyzoites or left uninfected for 4 h. Then, cultures were treated with 1 µM AG1468 or an equivalent volume of DMSO (i.e., vehicle) for 20 h. Samples were processed for confocal immunofluorescence microscopy. Fixed cells were stained with DAPI (shown in blue), used as a nuclear marker, and for total FOXO3a (shown in white). The data in this panel were obtained at the same time as those in Fig. 2C. DMSO images are repeated here to better depict and compare the various treatment conditions across figures. Images are representative of two independent experiments. Original magnification (left panels) and four times enlarged insets (right panels). PVs are outlined with dashed lines to indicate the presence of parasites within infected cells. Data are representative of at least two independent experiments (i.e., performed on different days).

Fig 4

AKT-FOXO3a-sensitive transcription of autophagy-related host genes induced upon serum starvation is impeded following T. gondii infection. (A) HFF cultures were pretreated with 2 µM MK-2206 or an equivalent volume of vehicle (i.e., DMSO) for 1 h and then inoculated with RH T. gondii parasites or left uninfected. To induce autophagy, cultures were deprived of serum throughout the course of the infection (24 h). (B) Knockdown levels of FOXO3 mRNA and FOXO3a protein in lentivirus-transduced HFF were monitored by qPCR and Western blotting, respectively, and compared to cells transduced with scrambled (Scr.) shRNA. (C) Scr. or FOXO3 shRNA-transduced HFF cultures were pretreated with 2 µM MK-2206 or an equivalent volume of vehicle (i.e., DMSO) for 1 h and then inoculated with RH T. gondii parasites or left uninfected. Cultures were deprived of serum throughout the course of the infection (24 h). (A; B, left panel; and C) Relative expression of indicated genes was normalized to ACTB and was calculated as a percentage of (A) serum-starved DMSO-treated uninfected control (i.e., “Control”), (B) Scr. HFF cultures, or (C) serum-starved DMSO-treated uninfected Scr. HFF cultures. Data are presented as mean (SD) in technical triplicates and are representative of at least two independent experiments (i.e., performed on different days). ****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant.

Fig 5

AKT inhibition leads to the recruitment of autophagolysosomal markers at the parasitophorous vacuole. (A, B) HFF cultures were treated with 2 µM MK-2206 or an equivalent volume of vehicle (i.e., DMSO) for 1 h. Cells were then inoculated with RH T. gondii parasites or left uninfected for 24 h. Cultures were serum-starved for the entire length of the experiment. Samples were processed for confocal immunofluorescence microscopy. As shown here, cells were stained with DAPI, for FOXO3a, and (A, B) LysoTracker Red DND-99 (2 h prior to fixation) or (C, D) LC3. Original magnification (left panels) and four times-enlarged insets (right panels). Data are representative of two biological replicates. (A, C) PVs are outlined with dashed lines to indicate the presence of parasites within infected cells. Recruitment of (B) lysosomal/autophagolysosomal structures (Lysotracker-stained) and (D) LC3 to the PV was quantified. In DMSO-treated cells, 62 and 73 PVs were analyzed for lysosomal/autophagolysosomal structure and LC3 recruitment, respectively. In MK-2206-treated cells, 37 and 43 PVs were analyzed for lysosomal/autophagolysosomal structure and LC3 recruitment, respectively. Data are presented as mean (SD) and are representative of three independent experiments (i.e., performed on different days). ****P < 0.0001.

Fig 6

Recruitment of autophagy-related effectors to the parasitophorous vacuole is, in part, dependent on FOXO3a activity. (A, B) HFF were transduced to express either N-terminal Myc-tagged FOXO3a WT or TM forms. Cell cultures were inoculated with RH T. gondii parasites or left uninfected for 24 h. Cultures were deprived of FBS (i.e., starved) for the entire length of the experiment. Samples were processed for confocal immunofluorescence microscopy. Cells were stained with DAPI, for Myc-tagged FOXO3a, and (A, B) LysoTracker Red DND-99 (2 h prior to fixation) or (C, D) LC3. Original magnification (left panels) and four times-enlarged insets (right panels). Data are representative of two biological replicates. (A, C) PVs are outlined with dashed lines to indicate the presence of parasites within infected cells. Recruitment of (B) (Lysotracker-stained) and (D) LC3 to the PV was quantified. In Myc-tagged FOXO3a WT-expressing cells, 128 and 132 PVs were analyzed for lysosomal/autophagolysosomal structure and LC3 recruitment, respectively. In Myc-tagged FOXO3a TM-expressing cells, 99 and 118 PVs were analyzed for lysosomal/autophagolysosomal structure and LC3 recruitment, respectively. Data are presented as mean (SD) and are representative of three independent experiments (i.e., performed on different days). ***P < 0.0001.

Fig 7

T. gondii represses FOXO3a-driven transcriptional programs to hamper autophagic targeting of the PV (Proposed Model). (A) Upon establishment and replication within the PV, T. gondii tachyzoites (shown in pink forming a rosette) activate the host cell PI3K-AKT signaling pathway independently of EGFR, PKCα, and mTOR. Phosphorylation of AKT (S473 and T308) leads to its activation and in turn to the phosphorylation of FOXO3a at AKT-sensitive residues (S253 and T32). Phosphorylation of FOXO3a at these residues leads to its nuclear exclusion and inactivation. As such, transcription of a subset of FOXO3a-dependent autophagy-related genes (i.e., ULK1, BECN1, NBR1, PINK1, and GABARAPL2) is downregulated (as indicated by the red “X” and the downward red arrows). Proteins encoded by this subset of transcripts are reported to participate in distinct steps of the autophagic response (right panel). Consequently, autophagic targeting of the PV is prevented, favoring parasite survival and replication. (B) Pharmacological inhibition of the PI3K-AKT pathway (i.e., treatment with LY294092 or MK-2206) precludes AKT-dependent phosphorylation and nuclear export of FOXO3a, thus promoting the transcription of autophagy-related genes (as indicated by upward black arrows) despite infection by T. gondii. Exogenous expression of an AKT-resistant form of FOXO3a harboring phosphosite mutations (S253A, T32A, and S315A) phenocopies chemical activation of FOXO3-driven autophagic targeting of T. gondii.