Divergent co-transcriptomes of different host cells infected with Toxoplasma gondii reveal cell type-specific host-parasite interactions

Abstract

The apicomplexan parasite Toxoplasma gondii infects various cell types in avian and mammalian hosts including humans. Infection of immunocompetent hosts is mostly asymptomatic or benign, but leads to development of largely dormant bradyzoites that persist predominantly within neurons and muscle cells. Here we have analyzed the impact of the host cell type on the co-transcriptomes of host and parasite using high-throughput RNA sequencing. Murine cortical neurons and astrocytes, skeletal muscle cells (SkMCs) and fibroblasts differed by more than 16,200 differentially expressed genes (DEGs) before and after infection with T. gondii. However, only a few hundred of them were regulated by infection and these largely diverged in neurons, SkMCs, astrocytes and fibroblasts indicating host cell type-specific transcriptional responses after infection. The heterogeneous transcriptomes of host cells before and during infection coincided with ~5,400 DEGs in T. gondii residing in different cell types. Finally, we identified gene clusters in both T. gondii and its host, which correlated with the predominant parasite persistence in neurons or SkMCs as compared to astrocytes or fibroblasts. Thus, heterogeneous expression profiles of different host cell types and the parasites' ability to adapting to them may govern the parasite-host cell interaction during toxoplasmosis.

Figure 1

Host cell type-specific responses of mouse SkMCs, neurons, astrocytes and fibroblasts to infection with T. gondii. Mature C2C12 myotubes, cortical neurons and astrocytes, and NIH/3T3 fibroblasts were infected with T. gondii at a MOI of 5:1 for 24 hours or were left non-infected. Pools of total RNA from four different biological replicates each were used to prepare cDNA libraries which were then sequenced using Illumina technology. Reads were mapped to the Mus musculus reference genome. (A) Numbers of differentially expressed genes (DEGs) which were up- or down-regulated (>2-fold, p < 0.05) after T. gondii infection as compared to the respective non-infected control. (B) Heatmap of DEGs after infection of SkMCs, neurons, astrocytes and fibroblasts with T. gondii. Expression profiles were clustered according to a one-dimensional self-organizing map. (C) Hierarchical cluster analysis of genes which were differentially expressed after T. gondii infection in at least one host cell type. (D) Venn diagram of genes up-regulated (>2-fold, p < 0.05) by T. gondii in at least one of the four host cell types. (E) Venn diagram of genes down-regulated (>2-fold, p < 0.05) by T. gondii in at least one of the four host cell types. (F) Venn diagram of genes expressed in any of the four different non-infected host cell types.

Figure 2

Common transcriptional responses of SkMCs, neurons, astrocytes and fibroblasts after infection with T. gondii as revealed by RNA high-throughput sequencing. Genes that were commonly regulated in at least two cell types after infection with T. gondii (>2-fold; p < 0.05) were clustered according to host cell groups, and expression intensities were visualized using heat maps. (A) Genes commonly up-regulated in multiple cell types after infection. (B) Genes commonly down-regulated in multiple cell types after infection.

Figure 3

Biological processes differentially or commonly enriched in the T. gondii-regulated transcriptomes of mature SkMCs, neurons, astrocytes and fibroblasts. After genome-wide expression profiling, T. gondii-regulated host cell genes (at least 2-fold up- or downregulation; p < 0.05) were screened for annotation terms being overrepresented as compared to the reference genome to identify biological processes being regulated after infection (FDR < 0.05, modified Fisher exact test with Benjamini-Hochberg correction for multiple testing). Redundant annotation terms were omitted. (A) Annotation terms that are overrepresented among genes up- (black bars) or downregulated (cross-hatches bars) in SkMCs by T. gondii-infection. (B) Annotation term that is overrepresented among genes upregulated in astrocytes by T. gondii-infection. (C) Annotation terms that are overrepresented among genes upregulated in fibroblasts by T. gondii-infection. (D) Impact of T. gondii infection on biological processes being enriched in at least two cell types following infection. Up- and downregulation of genes related to the biological processes as indicated are depicted in green or red, respectively. Figures indicate the numbers of genes the expression of which is regulated in the respective cell types.

Figure 4

Differentially expressed genes in non-infected SkMCs, neurons, astrocytes and fibroblasts. Genome-wide expression profiles were obtained by RNA high-throughput sequencing. Genes that were differentially expressed between cell types were extracted (Chi² test, FWER < 0.01). (A) Heat map of DEGs from SkMCs, neurons, astrocytes and fibroblasts. Expression profiles were clustered according to a one-dimensional self-organizing map. Gene clusters that were differentially expressed in SkMCs and neurons as compared to astrocytes and fibroblasts are indicated by cluster numbers. (B,C) Biological processes showing differential gene expression profiles between SkMCs and neurons as compared to astrocytes and fibroblasts. Annotation terms that were overrepresented among the genes from cluster 3 (B) and cluster 24 (C) as compared to the reference genome were identified by GO analysis (FDR < 0.05, modified Fisher exact test with Benjamini-Hochberg correction for multiple testing). Redundant annotation terms were omitted.

Figure 5

Heterogeneous gene expression in T. gondii after infection of mature SkMCs, neurons, astrocytes and fibroblasts at a MOI of 5:1 for 24 hours. Pools of total RNA from four different biological replicates each were used to prepare cDNA libraries which were then sequenced using Illumina technology. Reads were mapped to the T. gondii reference genome. Genes were identified the expression of which was differentially regulated depending on the host cell type (Chi² FWER < 0.01). (A) Heat map of T. gondii DEGs after infection of SkMCs, neurons, astrocytes and fibroblasts. Expression profiles were clustered according to a one-dimensional self-organizing map. Clusters of genes that were differentially expressed after infection of SkMCs and neurons as compared to astrocytes and fibroblasts are indicated by cluster numbers. (B) Hierarchical cluster analysis of genes differentially expressed after infection of the different host cell types. (C) Venn diagram of genes differentially expressed in T. gondii after infection of the different host cell types.

Figure 6

Differential gene expression in T. gondii following infection of either mature SkMCs and neurons or astrocytes and fibroblasts at a MOI of 5:1 for 24 hours. Expression profiles of T. gondii were obtained by high-throughput RNA sequencing and parasite genes were extracted that were differentially expressed depending on the host cell type (Chi² FWER < 0.01). (A) Expression of bradyzoite antigen (BAG)-1 in T. gondii during infection of host cells as indicated. Data are reads per kilobase per million mapped reads (RPKM). Nd: not detected. (B–E) Biological processes showing differential gene expression profiles in T. gondii residing within SkMCs and neurons as compared to those in astrocytes and fibroblasts. Annotation terms that were overrepresented among the genes from cluster 9 (B), cluster 10 (C), cluster 20 (D) and cluster 21 (E) indicated in Fig. 5A as compared to the reference genome were identified by using the GO analysis tool at www.toxodb.org (FDR < 0.05 with Benjamini-Hochberg correction for multiple testing). Redundant annotation terms were omitted.