Role for parasite genetic diversity in differential host responses to Trypanosoma brucei infection

Abstract

The postgenomic era has revolutionized approaches to defining host-pathogen interactions and the investigation of the influence of genetic variation in either protagonist upon infection outcome. We analyzed pathology induced by infection with two genetically distinct Trypanosoma brucei strains and found that pathogenesis is partly strain specific, involving distinct host mechanisms. Infections of BALB/c mice with one strain (927) resulted in more severe anemia and greater erythropoietin production compared to infections with the second strain (247), which, contrastingly, produced greater splenomegaly and reticulocytosis. Plasma interleukin-10 (IL-10) and gamma interferon levels were significantly higher in strain 927-infected mice, whereas IL-12 was higher in strain 247-infected mice. To define mechanisms underlying these differences, expression microarray analysis of host genes in the spleen at day 10 postinfection was undertaken. Rank product analysis (RPA) showed that 40% of the significantly differentially expressed genes were specific to infection with one or the other trypanosome strain. RPA and pathway analysis identified LXR/RXR signaling, IL-10 signaling, and alternative macrophage activation as the most significantly differentially activated host processes. These data suggest that innate immune response modulation is a key determinant in trypanosome infections, the pattern of which can vary, dependent upon the trypanosome strain. This strongly suggests that a parasite genetic component is responsible for causing disease in the host. Our understanding of trypanosome infections is largely based on studies involving single parasite strains, and our results suggest that an integrated host-parasite approach is required for future studies on trypanosome pathogenesis. Furthermore, it is necessary to incorporate parasite variation into both experimental systems and models of pathogenesis.

FIG. 1.

Measurement of cell populations by FACS. (Panel i) Forward scatter versus side scatter. Gated cell populations are shown. A, trypanosomes; B, calibration beads; C, red blood cells; D, white blood cells. (Panel ii) Forward scatter versus FL1 channel. X, reticulocytes stained by thiazole orange; Y, mature red blood cells.

FIG. 2.

Phenotypic parameters of strain 927-infected (black bars), strain 247-infected (gray bars), and uninfected (white bars) mice at days 3, 6, 10, and 12 p.i. (mean value ± 1 standard error, n = 6). For statistical analysis of phenotypes, see Table S1 in the supplemental material.

FIG. 3.

The number of genes significantly differentially expressed (FDR of <1) between trypanosome-infected and uninfected mouse spleens. The overlap represents commonality in strain 247- and 927-induced changes relative to uninfected controls, and numbers to the left and right are the strain-specifically regulated genes at this FDR level. Panel A represents genes upregulated in strain 927- or 247-infected mice relative to control mice, and panel B represents genes downregulated in strain 927- or 247-infected mice relative to control mice.

FIG. 4.

Significantly differentially regulated biological functions (A) and canonical pathways (B) between strain 247-infected and uninfected mice (247 versus C) and between strain 927-infected and uninfected mice (927 versus C), as identified by IPA. The y axis is the negative log of the Fischer exact test P value for the genes identified with the respective biofunction or canonical pathway. The dashed line indicates the threshold of significance (P < 0.05).

FIG. 5.

Canonical pathways significantly differentially regulated in strain 247- and strain 927-infected mice. The bar chart shows the Fischer exact test P values determining the probability that the association between the genes in the data set and the canonical pathway is explained by chance alone (the significance threshold of P < 0.05 is indicated by the dashed line).

FIG. 6.

Real-time PCR of the Ym1 (A) and Arg2 (B) genes. Real-time results are presented as relative quantification (RQ) with respect to control sample 1. Control reactions 1 to 3 were carried out with RNA from uninfected control mice, strain 927 reactions 1 to 3 used RNA extracted from mice infected with strain 927, and strain 247 reactions 1 to 3 used RNA from mice infected with strain 247. Negative controls (absence of reverse transcriptase and no-template controls) were all negative in each case (data not shown).