Secreted immunodominant Mycobacterium tuberculosis antigens are processed by the cytosolic pathway

Abstract

Exposure to Mycobacterium tuberculosis can result in lifelong but asymptomatic infection in most individuals. Although CD8(+) T cells are elicited at high frequencies over the course of infection in both humans and mice, how phagosomal M. tuberculosis Ags are processed and presented by MHC class I molecules is poorly understood. Broadly, both cytosolic and noncytosolic pathways have been described. We have previously characterized the presentation of three HLA-I epitopes from M. tuberculosis and shown that these Ags are processed in the cytosol, whereas others have demonstrated noncytosolic presentation of the 19-kDa lipoprotein as well as apoptotic bodies from M. tuberculosis-infected cells. In this paper, we now characterize the processing pathway in an additional six M. tuberculosis epitopes from four proteins in human dendritic cells. Addition of the endoplasmic reticulum-Golgi trafficking inhibitor, brefeldin A, resulted in complete abrogation of Ag processing consistent with cytosolic presentation. However, although addition of the proteasome inhibitor epoxomicin blocked the presentation of two epitopes, presentation of four epitopes was enhanced. To further examine the requirement for proteasomal processing of an epoxomicin-enhanced epitope, an in vitro proteasome digestion assay was established. We find that the proteasome does indeed generate the epitope and that epitope generation is enhanced in the presence of epoxomicin. To further confirm that both the epoxomicin-inhibited and epoxomicin-enhanced epitopes are processed cytosolically, we demonstrate that TAP transport and new protein synthesis are required for presentation. Taken together, these data demonstrate that immunodominant M. tuberculosis CD8(+) Ags are processed and presented using a cytosolic pathway.

FIGURE 1

Presentation of M. tuberculosis Ags requires ER-Golgi transport but not acidification. DCs were pretreated with BFA (0.1 μg/ml) or bafilomycin (0.2 μM) for 1 h prior to infection with H37Rv-eGFP (MOI = 20) or addition of peptide Ag (1 μg/ml). After 15–16 h in the presence of inhibitor, DCs were fixed, washed, and used as stimulators in an IFN-γ ELISPOT assay (25,000 M. tuberculosis-infected DCs/well, 2,000 peptide-pulsed DCs/well) where CD8⁺ T cell clones are effectors (10,000 cells/well). A, Representative experiment for three clones. Each bar represents the mean ± SEM of triplicate wells. B, Data have been normalized to the untreated control, and each bar reflects the mean ± SEM of at least three experiments per clone (two for D443 H9). *p < 0.05; **p < 0.01 using two-tailed Student t test compared with untreated controls. Mtb, Mycobacterium tuberculosis.

FIGURE 2

Presentation of M. tuberculosis epitopes is either inhibited or enhanced by epoxomicin. DCs were treated with epoxomicin (1–10μM), infected with H37Rv-eGFP or pulsed with peptide, and used as APCs in an IFN-γ ELISPOT assay as described in Fig. 1. A, Data have been normalized to the untreated control, and each bar represents the mean ± SEM of at least three experiments (two for D443 H9). *p < 0.05; **p < 0.01 using two-tailed Student t test compared with untreated controls. B, DCs were treated with increasing concentrations of epoxomicin, infected with M. tuberculosis (M. tuberculosis-DCs) or pulsed with cognate peptides (Ag-DCs), and used as stimulators as described. One representative experiment of three is shown. Mtb, Mycobacterium tuberculosis.

FIGURE 3

CFP10_2–12 is generated by proteasomes and enhanced by epoxomicin. CFP10_2–21 (50 μg) was incubated with 5 μg purified 20S immunoproteasomes or chymotrypsin for 20 h, and peptide digests were analyzed by LC-MS/MS and IFN-γ ELISPOT. A, Representative LC-MS/MS analysis of digested peptide. After incubation with proteasomes, 15–25% of the input peptide is degraded, whereas almost complete digestion is seen with chymotrypsin. Peaks are labeled with the appropriate peptide species detected, with the calculated masses in parentheses. Data are representative of two experiments. B and C, Serial dilutions of proteasome digests (B) or synthetic peptides (C) were incubated with CD8⁺ T cell clones (20,000/well) in the presence (left panels) or absence (right panels) of LCLs (20,000/well) in an IFN-γ ELISPOT assay. Error bars represent the SEM of duplicate wells, and data are representative of two experiments. Ox, oxidized.

FIGURE 4

TAP transport is required for presentation of all epitopes. DCs were infected with either adenoviral ICP47 or empty vector using Lipofectamine 2000. After 6–26 h, DCs were washed and infected with H37Rv-eGFP (A) or pulsed with Ag (B). Following overnight incubation, T cell clones were added, and IFN-γ production was assessed by intracellular cytokine staining. For each clone, the mean response to mock-treated, M. tuberculosis-infected DCs was at least 4-fold higher than the response to uninfected DCs. Data have been normalized to the untreated controls, and each bar represents the mean ± SEM of at least three independent experiments. *p < 0.05; **p < 0.01 using two-tailed Student t test compared with untreated controls. Mtb, Mycobacterium tuberculosis.

FIGURE 5

Protein synthesis is required for presentation of epoxomicin-enhanced epitopes. DCs were treated with cycloheximide (10 μg/ml), infected with H37Rv-eGFP or pulsed with peptide, and used as APCs in an IFN-γ ELISPOT assay as described in Fig. 1. Data have been normalized to the untreated control, and each bar represents the mean ± SEM of at least three experiments. *p < 0.05; **p < 0.01 using two-tailed Student t test compared with untreated controls. Mtb, Mycobacterium tuberculosis.