High-throughput identification of Toxoplasma gondii effector proteins that target host cell transcription

Abstract

Intracellular pathogens and other endosymbionts reprogram host cell transcription to suppress immune responses and recalibrate biosynthetic pathways. This reprogramming is critical in determining the outcome of infection or colonization. We combine pooled CRISPR knockout screening with dual host-microbe single-cell RNA sequencing, a method we term dual perturb-seq, to identify the molecular mediators of these transcriptional interactions. Applying dual perturb-seq to the intracellular pathogen Toxoplasma gondii, we are able to identify previously uncharacterized effector proteins and directly infer their function from the transcriptomic data. We show that TgGRA59 contributes to the export of other effector proteins from the parasite into the host cell and identify an effector, TgSOS1, that is necessary for sustained host STAT6 signaling and thereby contributes to parasite immune evasion and persistence. Together, this work demonstrates a tool that can be broadly adapted to interrogate host-microbe transcriptional interactions and reveal mechanisms of infection and immune evasion.

Keywords: CRISPR screening; STAT6; Toxoplasma gondii; effector proteins; host-microbe interactions; host-pathogen interactions; immune evasion; perturb-seq; single-cell RNA sequencing.

Figure 1.. Dual perturb-seq transcriptional profiles recapitulate phenotypes of known T. gondii effectors.

A. Schematic of perturb-seq plasmid vector and capture sequence sgRNAs. B. Knockout efficiency of perturb-seq vectors. T. gondii parasites were transfected with perturb-seq vectors targeting the UPRT gene and selected for integration of the plasmid using the HXGPRT marker. Loss of function of the UPRT gene was measured by plaque assay in the presence/absence of 5-fluorodeoxyuridine, with the knockout efficiency calculated as the percentage of plaque-forming-units resistant to 5-fluorodeoxyuridine. Differences were tested by two-sided t-test with Bonferroni correction. C. Percentage of infected cells with a detectable sgRNA in all dual perturb-seq samples. Difference tested by two-sided t-test. D. Mean number of sgRNA UMIs in cells with a detectable sgRNA in all dual perturb-seq samples. Difference tested by two-sided t-test. E. Schematic of T. gondii effector protein export into the host cell. Effector proteins may be secreted into the host cell from the rhoptry organelles during parasite invasion of the host cell, or may be secreted into the parasitophorous vacuole of intracellular parasites and subsequently translocated into the host cell by a complex that contains the MYR1 protein. F. Experimental strategy for dual perturb-seq experiment with two sgRNAs. T. gondii parasites were transfected with perturb-seq vectors targeting either UPRT or MYR1 and selected for integration of the vector into the genome. The transfected parasites were used to infect HFFs for 24 h, following which the infected cells were purified by fluorescence-activated cell sorting and analysed by droplet-based scRNA-seq with capture of both polyadenylated host and parasite mRNAs and the sgRNA transcripts. G. UMAP of single cell transcriptomes based on host cell gene expression and coloured according to the sgRNA species detected or by expression of the most significantly up- and down-regulated host cell genes in sg(MYR1)-expressing cells. H. UMAP of single cell transcriptomes based on T. gondii gene expression and coloured according to the sgRNA species detected or by expression of T. gondii cell cycle marker genes. Expression of glycolytic enzymes (e.g. GAPDH2) peaks in early G1 phase (G1a), while apicoplast-localised proteins (e.g. ACP) peak in late G1 phase (G1b), rhoptry proteins (e.g. ROP1) in S phase, and microneme proteins (e.g. MIC1) in M phase. Note that T. gondii lacks a discernible G2 phase. I. Experimental strategy for dual perturb-seq experiment with 24 sgRNAs. T. gondii parasites were transfected with one of 24 perturb-seq vectors and selected for integration of the vector into the genome. The knockout parasites were pooled together to infect HFFs for 24 h and infected cells purified and analysed by scRNA-seq as above. J. For each sgRNA in the 24-sgRNA experiment, the number of significantly differentially expressed host cell genes was determined relative to sg(UPRT) (adjusted p-value < 0.01 and average log₂ fold-change > 0.5). See also Table S2. K. Single cell transcriptomes in the 24-sgRNA experiment were scored for expression of marker genes regulated by established effector proteins using VISION. A higher score indicates greater concordance with published bulk RNA-seq data for a given effector protein. Differences were tested by two-sided Wilcoxon rank-sum test.

Figure 2:. A dual-perturb-seq screen identifies T. gondii effectors that target host cell transcription.

A. Schematic of dual perturb-seq screen. ssDNA oligonucleotides encoding protospacer sequences were selected from an arrayed library and incorporated into the perturb-seq vector by pooled Gibson assembly. T. gondii parasites were transfected with the perturb-seq vector pool and selected for integration of the plasmid into the genome. HFFs were infected with the resulting pool of knockout parasites for 24 h with or without treatment with interferon-gamma. Infected cells were purified by FACS and analysed by scRNA-seq. B. Perturbation of host cell transcriptomes by T. gondii effectors. Single-cell transcriptomes were analysed by principle component analysis (PCA) using the host cell gene expression, and each T. gondii effector was tested for an altered distribution of PCA embeddings using Hotelling’s t²-test with Benjamini Hochberg adjustment. See also Table S4. C. PCA embeddings of pseudo-bulk host cell transcriptomes. D. Pearson correlations between pseudo-bulk host cell transcriptomes of significant effectors. E. Host cell pathways regulated by putative effectors. Single cell transcriptomes were scored for expression of gene sets from the Pathway Interaction Database using VISION. Gene sets that were significantly differentially regulated by at least one putative effector (p < 0.01, two-sided Wilcoxon rank-sum test with Benjamini-Hochberg adjustment) are plotted with the mean VISION signature score transformed to a Z-score. See also Figure S5 and Table S8.

Figure 3.. GRA59 (TGGT1_313440) contributes to dense granule effector protein export.

A. Scoring of single cell transcriptomes for a ΔMYR1-infected phenotype using VISION. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. B. Immunofluorescence localisation of GRA59-HA within the T. gondii vacuole at 24 h post-infection (hpi). Scale bar = 10 μm. C. Model of dense granule effector export into the host cell. D. Quantification of nuclear c-Myc immunofluorescence in the nuclei of infected cells at 24 hpi. Scale bar = 10 μm. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. E. Quantification of nuclear phospho-p38 immunofluorescence in the nuclei of infected cells at 24 hpi. Scale bar = 10 μm. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. F. Quantification of nuclear EZH2 immunofluorescence in the nuclei of infected cells at 24 hpi. Scale bar = 10 μm. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. G. Quantification of nuclear IRF1 immunofluorescence in the nuclei of IFNγ-stimulated infected cells at 24 hpi. Scale bar = 10 μm. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. H. Ratio of GRA16-HA immunofluorescence in the host cell nuclei compared to the vacuole at 24 hpi. Scale bar = 10 μm. Differences tested by paired two-sided t-test with Bonferroni adjustment. I. Ratio of IST-HA immunofluorescence in the host cell nuclei compared to the vacuole at 24 hpi. Scale bar = 10 μm. Differences tested by paired two-sided t-test with Bonferroni adjustment. J. Percentage of vacuoles containing accumulations of IST-HA immunofluorescence at 24 hpi. White arrowheads indicate accumulations. Scale bar = 10 μm. Differences tested by paired two-sided t-test with Bonferroni adjustment.

Figure 4.. SOS1 (TGGT1_222100) is required for sustained STAT6 signalling in infected cells.

A. Differentially expressed host cell genes for sg(SOS1)-expression cells compared to all other (non-control) cells (two-sided Wilcoxon rank-sum test with Benjamini-Hochberg adjustment). See also Table S9. B. Differentially expressed Pathway Interaction Database gene sets for sg(SOS1)-expression cells compared to all other (non-control) cells (two-sided Wilcoxon rank-sum test with Benjamini-Hochberg adjustment). See also Table S8. C. Immunofluorescence localisation of SOS1-HA in intracellular T. gondii parasites at 24 hpi. Scale bar = 10 μm. D. Quantification of nuclear phospho-STAT6 immunofluorescence in the nuclei of infected HFFs at 1 h and 24 h post-infection. Scale bar = 10 μm. Differences tested by two-sided Wilcoxon rank sum test with Bonferroni adjustment. E. Arginase activity of infected cell lysate. Differences tested by two-sided t-test with Bonferroni adjustment. F. Model of ROP16 and SOS1-dependent regulation of STAT6 signalling.

Figure 5.. SOS1 maintains efficient bradyzoite cyst formation in neurons.

A. Phospho-STAT6 immunofluorescence in HFFs infected with the indicated T. gondii strains at 1 h and 24 h post-infection. Scale bar = 10 μm. B. Quantification of phospho-STAT6 relative to total STAT6 from Western blots of samples of primary murine neurons infected for 24 h with indicated T. gondii strains. Differences tested by two-sided t-test with Bonferroni adjustment. C. Percentage of DBA-positive bradyzoite cysts out of total vacuoles with ≥2 parasites in primary murine neurons infected for 1, 2 or 3 days. Differences test by two-sided t-test with Bonferroni adjustment. D. Model of the dynamics of ROP16- and SOS1-dependent phenotypes.