Cytokine-dependent and-independent gene expression changes and cell cycle block revealed in Trypanosoma cruzi-infected host cells by comparative mRNA profiling

Abstract

Background: The requirements for growth and survival of the intracellular pathogen Trypanosoma cruzi within mammalian host cells are poorly understood. Transcriptional profiling of the host cell response to infection serves as a rapid read-out for perturbation of host physiology that, in part, reflects adaptation to the infective process. Using Affymetrix oligonucleotide array analysis we identified common and disparate host cell responses triggered by T. cruzi infection of phenotypically diverse human cell types.

Results: We report significant changes in transcript abundance in T. cruzi-infected fibroblasts, endothelial cells and smooth muscle cells (2852, 2155 and 531 genes respectively; fold-change > or = 2, p-value < 0.01) 24 hours post-invasion. A prominent type I interferon response was observed in each cell type, reflecting a secondary response to secreted cytokine in infected cultures. To identify a core cytokine-independent response in T. cruzi-infected fibroblasts and endothelial cells transwell plates were used to distinguish cytokine-dependent and -independent gene expression profiles. This approach revealed the induction of metabolic and signaling pathways involved in cell proliferation, amino acid catabolism and response to wounding as common themes in T. cruzi-infected cells. In addition, the downregulation of genes involved in mitotic cell cycle and cell division predicted that T. cruzi infection may impede host cell cycle progression. The observation of impaired cytokinesis in T. cruzi-infected cells, following nuclear replication, confirmed this prediction.

Conclusion: Metabolic pathways and cellular processes were identified as significantly altered at the transcriptional level in response to T. cruzi infection in a cytokine-independent manner. Several of these alterations are supported by previous studies of T. cruzi metabolic requirements or effects on the host. However, our methods also revealed a T. cruzi-dependent block in the host cell cycle, at the level of cytokinesis, previously unrecognized for this pathogen-host cell interaction.

Figure 1

Differential gene expression in T. cruzi-infected HFF, HMVEC and VSMC at 24 hours post-infection. A. Heat map of genes modulated during T. cruzi infection in each cell line identified by hybridization to HG_U133 2.0 arrays and analyzed with Rosetta Resolver. Genes modulated (≥ 2-fold, p < 0.01) in at least one of the three cell types are shown. B. Venn diagrams comparing up or downregulated genes (≥ 2-fold, p < 0.01) and GO functions (p < 0.05) in each cell type. C. qPCR analysis of the expression of interferon beta (IFN-β) mRNA normalized to GAPDH. D. Canonical pathways identified by Inguenuity Pathway Analysis™ software as significantly altered (p < 0.05) following T. cruzi infection of HFF, HMVEC and VSMC.

Figure 2

Identification of cytokine-dependent and – independent transcriptional responses in T. cruzi-infected HFF and HMVEC. A. Schematic showing experimental design where T. cruzi-infected cells on the bottom of the transwell plate are separated from uninfected cells on the top layer and mock-infected controls. B. Heat maps showing combined information for two replicate experiments for changes in transcript abundance (≥ 2-fold, p < 0.01) for cells plated in the bottom and top the top Transwell insert and for the genes remaining after subtraction of the genes present in both top and bottom (Methods). Black bars represent cytokine-dependent genes removed by filtering the top response from the bottom. C. qPCR confirmation of microarray data for selected non-differentially expressed genes (C), IFN-stimulated genes (ISG) and presumptive cytokine-independent genes (cytokine-independent) identified by microarray analysis of samples obtained from HFF cells in Transwell experiments. Where more than one sequence corresponding to the same gene was modulated in the microarray data, the perfect match (_at) sequence was chosen. Tropomodulin 3 (TMOD 3), IFNβ (IFN-β), radical S-adenosyl methionine domain-containing protein 2 (RSAD2), signal transducer and activator of transcription 1 (STAT-1), cholesterol-25-hydroxylase (CH25H), kruppel-like transcription factor-4 (KLF4), purine nucleoside phosphorylase (PNP), solute carrier family 39 member 8 (SLC39A8), sphingosine kinase-1 (SPHK1), huntingtin-interacting protein 1-related protein (HIP1R), phosphofructokinase (PFK), myocardin (MYOCD). cDNA generated from three independent infections was analyzed and the mean fold-change ± s.d. are reported for the qPCR.

Figure 3

Cellular processes predicted by microarray data to be modulated during T. cruzi infection. A. Canonical pathways in T. cruzi-infected cells that are significantly modulated in a cytokine-independent manner. B. Venn diagrams comparing genes and representative GO functions for all up or downregulated genes (p < 0.05) in VSMC or following subtraction of the cytokine-dependent genes identified in the transwell experiments for HFF and HMVEC.

Figure 4

Cell cycle progression of T. cruzi-infected cells. A. Quantitation of BrdU-stained HFF cells mock- or T. cruzi-infected in the presence or absence of 5 ng/ml EGF for 24, 48 or 72 hours. Within infected cultures, parasite-containing and parasite-free cells were scored separately. B. Phospho-H3 staining (green) in T. cruzi-infected cells at 48 hours post-infection where host and parasite DNA are stained with DAPI (blue). C. Images of co-cultured T. cruzi-infected HFF cells preloaded with Cytotracker™-green (green) or Cytracker™-orange (red). Host cell nuclei (N) and parasite DNA (smaller blue) are stained with DAPI (blue).

Figure 5

Summary of common and cell specific responses elicited in human cells infected with T. cruzi. Common transcriptional responses are shown inside the circle, while individual responses are shown in the outside, next to each cell type. Transwell data was not available for VSMC. Arrows indicate up or downregulation.