Persistent Ehrlichia chaffeensis infection occurs in the absence of functional major histocompatibility complex class II genes

Abstract

Human monocytic ehrlichiosis is an emerging tick-borne disease caused by the rickettsia Ehrlichia chaffeensis. We investigated the impact of two genes that control macrophage and T-cell function on murine resistance to E. chaffeensis. Congenic pairs of wild-type and toll-like receptor 4 (tlr4)- or major histocompatibility complex class II (MHC-II)-deficient mice were used for these studies. Wild-type mice cleared the infection within 2 weeks, and the response included macrophage activation and the synthesis of E. chaffeensis-specific Th1-type immunoglobulin G response. The absence of a functional tlr4 gene depressed nitric oxide and interleukin 6 secretion by macrophages and resulted in short-term persistent infections for > or =30 days. In the absence of MHC-II alleles, E. chaffeensis infections persisted throughout the entire 3-month evaluation period. Together, these data suggest that macrophage activation and cell-mediated immunity, orchestrated by CD4(+) T cells, are critical for conferring resistance to E. chaffeensis.

FIG. 1.

Results of in vitro culture isolation and PCR assays for E. chaffeensis-infected mice. Typical results obtained for each mouse strain were presented at each time point postinfection for all 7 mouse strains. Numbers 3, 8, 16, 23, 30, 50, and 92 refer to sample analysis days postinfection. Ninety-two days postinfection samples for C2D mice were not available. All control mice were negative for E. chaffeensis as determined by culture isolation and PCR assay. Culture, culture positives of E. chaffeensis from peritoneal exudate cells (+, positive; −, negative; c, contaminated; n, data not available); PCR, hybridization data from PCR assay for E. chaffeensis-specific ribosomal gene in genomic DNA isolated from liver samples.

FIG. 2.

Liver lesion severity in mice infected with E. chaffeensis Arkansas isolate. Lesions were evaluated at various time intervals and assigned lesion scores (described in Materials and Methods). Data are presented as the median scores based on the presence (black bars) or absence (white bars) of MHC-II and tlr4 gene alleles. MHC-II^+/+ data included 8 mice each, while tlr4^n/n data are from 10 mice each. The MHC-II^−/− data included 12 mice each, and data for tlr4^d/d mice were compiled from 10 mice. An asterisk indicates significant differences between mice with functional and mutant alleles (P < 0.01 using a Mann-Whitney rank sum test).

FIG. 3.

Histopathological (A to C) and immunohistochemical (D to F) analyses of liver. Histology sections were stained with hematoxylin II and eosin, while specific staining of E. chaffeensis-infected mononuclear cells was achieved using alkaline phosphatase-conjugated secondary antibodies with Fast Red TR/napthol AS-MX as the substrate (Sigma Chemical Co.). (A) Representative inflammatory response at day 8 postinfection. Several inflammatory cells and lesions are present (magnification, ×10; H&E staining). (B) Inflammatory response focus containing neutrophils and an apoptotic body (magnification, ×100; H&E staining). (C) Granulomatous inflammation on day 16 (magnification, ×40, H&E staining). (D) Control mouse liver; there is no immunostaining (magnification, ×100; hematoxylin staining). (E) E. chaffeensis-positive red staining in the cytoplasm of inflammation cells and mononuclear cells in hepatic sinusoids (magnification, ×100; hematoxylin staining). (F) E. chaffeensis-positive red staining in macrophage foci and mononuclear cells in hepatic sinusoids (magnification, ×100; hematoxylin staining).

FIG. 4.

Western blot profile showing response to an expressed E. chaffeensis 28-kDa recombinant outer membrane protein. Antibody data are not presented for mice that did not induce the IgG response. Typical results obtained for each mouse strain were presented. Numbers 3, 8, 16, 23, 30, 50, and 92 refer to sample analysis days postinfection.

FIG. 5.

IgG subclass distribution in mice that had an IgG response to the E. chaffeensis infection. Plasma samples from all mice beginning 16 days to 92 days postinfection were analyzed by quantitative ELISA. The data for all mice are presented by days postinfection and the pooled data for mice from all postinfection dates from days 16 to 92 are also presented. Each bar represents a median IgG concentration determined from samples analyzed from eight mice.