The Toxoplasma effector TEEGR promotes parasite persistence by modulating NF-κB signalling via EZH2

Abstract

The protozoan parasite Toxoplasma gondii has co-evolved with its homeothermic hosts (humans included) strategies that drive its quasi-asymptomatic persistence in hosts, hence optimizing the chance of transmission to new hosts. Persistence, which starts with a small subset of parasites that escape host immune killing and colonize the so-called immune privileged tissues where they differentiate into a low replicating stage, is driven by the interleukin 12 (IL-12)-interferon-γ (IFN-γ) axis. Recent characterization of a family of Toxoplasma effectors that are delivered into the host cell, in which they rewire the host cell gene expression, has allowed the identification of regulators of the IL-12-IFN-γ axis, including repressors. We now report on the dense granule-resident effector, called TEEGR (Toxoplasma E2F4-associated EZH2-inducing gene regulator) that counteracts the nuclear factor-κB (NF-κB) signalling pathway. Once exported into the host cell, TEEGR ends up in the nucleus where it not only complexes with the E2F3 and E2F4 host transcription factors to induce gene expression, but also promotes shaping of a non-permissive chromatin through its capacity to switch on EZH2. Remarkably, EZH2 fosters the epigenetic silencing of a subset of NF-κB-regulated cytokines, thereby strongly contributing to the host immune equilibrium that influences the host immune response and promotes parasite persistence in mice.

Fig. 1. The export of TEEGR in the host cell nucleus.

The gene TGME49_239010 encoding TEEGR was originally identified in silico together with GRA16, GRA24 and TgIST, as strong parasite candidate genes with attributes to both target the parasite secretory pathway and to reach the host cell nuclei. a, TEEGR is a highly disordered 80 kDa protein (predicted disorder score > 0.5) that carries a signal peptide, a putative nuclear localization sequence (NLS) and multiple repeated domains (R1, R2 and R3) whose repeat number and distribution have evolved through coccidian subclasses and among T. gondii lineages. Tg: Toxoplasma gondii, Nc: Neospora caninum, Hh: Hammondia hammondia. b, HAFlag (HF)-tagged TEEGR (in red) in Pru∆ku80 extracellular parasites is contained in cytoplasmic organelles distinct from the apical micronemes (MIC2, in green) and rhoptries (Toxofilin, in green), and partially co-localizing with the dense granule protein GRA7 (in green). Cells were co-stained with Hoechst DNA-specific dye (in blue). Scale bars, 1 μm. c, Time course of TEEGR secretion and export to the host cell nucleus. Human Foreskin Fibroblasts (HFF) were infected with a Pru∆ku80 TEEGR-HF and stained with anti-HA antibodies (red). Scale bars, 10 μm. d, in situ subcellular localization of TEEGR chimeric proteins (in red) in HFF infected with parasites expressing endogenously HF-tagged TEEGR truncated proteins (TEEGR∆SER and TEEGR∆R3) or TEEGR^Nter-R3 expressed under the control of a TUB8 promoter. Scale bars, 10 μm. All data are representative of 3 independent biological experiments.

Fig. 2. TEEGR activates gene expression in human cells in a E2F3 and E2F4-dependent manner.

Genomewide expression profiling of human fibroblasts (HFF) and astrocytes (HA) infected with the ∆teegr mutant (∆), the parental type II strain Pru∆ku80 (wt) and the trans-complemented strain (C). a, Results of tests for differential expression are presented in a volcano plot which plots statistical significance against fold change for each gene. Genes (HFF, n=784; HA, n=1529, in red) with adjusted p-values (Bonferroni corrected P-value ≤ 0.05) and absolute fold changes of ≥1.5 are deemed of interest for in silico pathway analyses (Supplementary Table 1). b, Heat map of expression values for differentially expressed genes in human cells infected with WT, Δ and C parasites. For the 472 genes (more than fourfold, p < 0.05, unpaired t-test) that were defined as core TEEGR-regulated genes in HFF and HA, mean log2 gene expression values were median centered, genes were clustered by hierarchical clustering based on Pearson correlation, and a heat map with 58 genes is presented. The complete set of genes is listed in Supplementary Table 1. c, HFF cells were left uninfected (ui) or infected for 24 h with WT, Δ, C or TEEGR protein missing R3 (ΔR3). Levels of iNOS and Alox12 mRNAs were determined by RT-qPCR. Values were normalized to the amount of β2-microglobulin. Data are mean value ± SD of three biological replicates. The p-values were calculated using two-tailed unpaired Student’s t-test. d, Gene Set Enrichment Analysis (GSEA) of the 784 genes positively regulated by TEEGR in HFF cells (Supplementary Table 1) highlighted eight gene expression signatures of chemical and genetic perturbations (CGP) that were significantly and selectively enriched (p < 0.05, False Discovery Rate (FDR)-corrected, two-sided Welch t-test). e, Transcription Factor Binding Sites (TFBS) analysis of the 784 TEEGR-upregulated genes was performed by DIRE and the most significant transcription factors are listed. f, Venn diagram illustrating the overlap between the number of genes up-regulated by TEEGR (> twofold, p < 0.05, unpaired t-test) in HFF cells and the number of genes identified as E2F3- or E2F4-bound by Julian et al., (2016) and Marson et al., (2007) respectively.

Fig. 3. TEEGR forms partnership with E2F/DP host transcription factors.

a, IFA of TEEGR-HAFlag ectopically and stably expressed in the 293-TRex cell line. Cells were either left untreated (-) or treated (+) with tetracycline for 20h and processed for IFA using anti-HA antibodies (red) and Hoechst DNA-specific dye (blue). Scale bar, 2.5 μm. b, TEEGR-associated polypeptides were purified from nuclear extracts of 293-TRex cells that had been tetracycline-induced to express TEEGR-HF. Size exclusion chromatography (SEC) of TEEGR-containing complexes after Flag affinity selection. Fractions were analyzed on silver stained SDS PAGE gels and blots processed to detect TEEGR-HF (anti-HA), E2F3, E2F4 and DP-1. c, Mass spectrometry-based proteomic analysis of SEC fraction 24 identified the transcription factors E2F3 and E2F4 associated with DP-1/DP-2 as the only TEEGR relevant partners. Identity of the proteins with their respective number of peptides is indicated on the right. Asterisks indicate degradation products. d, TEEGR-associated proteins were purified by Flag chromatography from protein extracts of murine J774 macrophage (MØ) line infected by Pru∆ku80 TEEGR-HAFlag or RH∆ku80 TEEGR-HAFlag. Fractions were analyzed on gels by silver staining and then by mass spectrometry-base proteomics to detect TEEGR and aforementioned partners. Identity of the proteins with their respective number of peptides and percentage of coverage are indicated. All data are representative of two independent biological experiments.

Fig. 4. E2F3 and E2F4 DNA association to EZH2 promoter is enhanced by T. gondii infection in a TEEGR-dependent manner.

a, Heat map representation of host transcription factors and chromatin-modifying enzymes up-regulated in a TEEGR-dependent fashion in HFF cells infected for 24 h with WT, Δ and C strains. EZH2 is indicated in red. b, Venn diagram illustrating the overlap between the number of genes up-regulated by TEEGR in HFF cells as well as BMDM (> twofold, p < 0.05, unpaired t-test) and the number of genes identified as E2F4-bound by Lee BK et al., (2011). c, IFA of TEEGR-dependent induction of EZH2 (red) in uninfected (ui) or 24 h infected HFFs with WT, Δ and C strains. In situ quantification of nuclear EZH2 using IFA. Horizontal bars represent the mean nuclear EZH2 intensity ± s.d. of three independent experiments (n=100 nuclei per dot). Scale bars, 10 μm. d, Nuclear extracts from HFF cells left uninfected (ui) or infected (24 h) by the indicated strains were analyzed by Western blotting using indicated antibodies. TBP (host-specific) and toxofilin (parasite-specific) are shown as loading controls. Data are representative of three experiments. e, Schematic presentation of the genomic regions corresponding to the promoter of EZH2. The black box in the EZH2 promoter represents a region of four putative E2F sites. HFFs were left uninfected (ui) or infected for 24 h with WT or Δ strains. Samples were analyzed by ChIP assay with specific antibodies to E2F3 and E2F4. IgGs were used as negative control. Bound DNA corresponding to EZH2 promoter was quantified by qPCR-ChIP, and signals were normalized with the input DNA. Data are mean value ± SD of three biological replicates. The p-values were calculated using two-tailed unpaired Student’s t-test, unless otherwise stated in the figure.

Fig. 5. TEEGR-dependent repression of NF-κB-regulated genes is mediated by EZH2.

a, Genomewide expression profiling of HFF and HA cells infected with wt and ∆. Results of tests for differential expression between wt and ∆ are presented in a volcano plot which plots statistical significance against fold change for each gene. Genes colored in blue (HFF, n=494; HA, n=571) have Bonferroni corrected p-values of ≤ 0.05 and absolute fold changes of ≤ 2 and are deemed of interest for in-silico pathway analyses (Supplementary Table 1). b, Heat map of expression values for differentially expressed genes in human cells infected with WT and Δ parasites. For the 69 (more than threefold, P < 0.05, unpaired t-test) that are defined as core TEEGR-repressed genes in HFF and HA, mean log2 gene expression values were median centered, genes were clustered by hierarchical clustering based on Pearson correlation, and a heat map is presented. c, HFF cells were left uninfected (ui) or infected for 24 h with WT, Δ, C or ΔR3 strains. Levels of IL-1β and IL-6 mRNAs were determined by RT-qPCR. Values were normalized to the amount of β2-microglobulin. Data are mean value ± SD of three biological replicates. d, TFBS analysis of the 69 TEEGR-repressed genes was performed by DIRE and the most significant transcription factors are listed. e, THP-1 murine cells were treated with TNFα or infected with Pru∆ku80 WT or Δteegr strains and then followed by determination of NF-kB activity using a reporter gene assay. Data are mean value ± SD of twenty biological replicates. f, HFFs were left uninfected (ui) or infected for 24 h with WT or Δ strains. Samples were analyzed by ChIP assay with specific antibodies to EZH2 and H3K27me3. IgG was used as negative control. Bound DNA corresponding to IL-6 and IL-8 promoters (respectively P1 and P2) was quantified by qPCR-ChIP, and signals were normalized with the input DNA. Data are mean value ± SD of three biological replicates. The p-values were calculated using two-tailed unpaired Student’s t-test, unless otherwise stated in the figure.

Fig. 6. In vivo control of teegr-deficient T. gondii tachyzoite population is likely mediated by NF-kB-regulated pro-inflammatory cytokines.

a, Mice were given intraperitoneally (i.p.) either 10⁵ (n=10) or 10⁶ (n=6) tachyzoites of WT, Δ and C strains and survival was monitored. At 5 days’ post inoculation all mice display clinical signs (weight loss, ruffled fur). Significance was tested using Log-rank (Mantel-Cox) test (p-value = 0.0161) and Gehan-Breslow-Wilcoxon test (p-value = 0.0154). A two-sided P value of <0.05 was considered statistically significant. b, Luminescent parasites were imaged with an IVIS imaging system from day 0 to 12 after the i.p. inoculation to BALB/c of 5.10⁴ tachyzoites per condition. The data are representative of two experiments (three mice per group). c, BALB/c mice were given i.p. a dose of 10⁵ Pru∆ku80 or Pru∆ku80 Δteegr tachyzoites. Peritoneal lavage fluids were collected on days 2, 5, 7, and 9 post inoculation. Number of tachyzoites was estimated within the collected samples by parasite DNA PCR and concentrations of IL-1β, IL-12, and IFN-γ were determined by ELISA. Data shown are means ± SD with n= 3 individual mice per parasite genotype at each time point. d, Schematic representation of the consequences of TEEGR delivery in the host cell by Toxoplasma tachyzoites.