Bacterial and host factors involved in the major histocompatibility complex class Ib-restricted presentation of Salmonella Hsp 60: novel pathway

Abstract

Previously, a peptide epitope derived from the Hsp 60 molecule of Salmonella that is presented by the major histocompatibility complex (MHC) class Ib molecule Qa-1 to CD8(+) cytotoxic T cells (CTLs) was described. In the present study we investigated the Salmonella-induced processing and presentation pathway for generating this Qa-1-restricted epitope. Live bacteria and, to a lesser extent, opsonized heat-killed bacteria are able to sensitize target cells for lysis by Salmonella-specific CTL. In contrast, heat-killed bacteria cannot sensitize target cells. Presentation of the Hsp 60 epitope appears independent of bacterial internalization, because cytochalasin D does not affect presentation. Moreover, Salmonella strains defective in the InvA or InvE operon, two critical components of the type III secretion pathway, are as efficient as wild-type Salmonella enterica serovar Typhimurium in sensitizing infected targets to lysis. Collectively, these results suggest the existence of a novel antigen-processing pathway in which exogenous antigens gain access to the cytosolic MHC class I processing machinery. Considering the abundant nature of bacterial Hsp 60 and the upregulation of this protein after Salmonella infection of eukaryotic cells, this mode of antigen presentation may be particularly relevant to understanding the host defense mechanisms against gram-negative bacteria.

FIG. 1.

The Salmonella-specific Qa-1^b-restricted CTL clone SalTc 1.69 recognizes opsonized heat-killed bacteria. J774 macrophages were pulsed with HKST for 1 h (B) or overnight (A) or with HKST opsonized in normal mouse serum (NMS) (B) for 30 min. At the end of coincubation, J774 cells were labeled with ⁵¹Cr and used as targets in a CTL assay employing SalTc 1.69 as the effector at the indicated effector:target ratios. The viability of the bacteria was confirmed by plating an aliquot of the heat-killed bacteria on LB plates. For comparison, live C5-infected J774 cells were included in these assays.

FIG. 2.

SalTc 1.69 recognizes infected target cells of various lineages. EL-4 (A), A20 (B), B78Hi and B78Hi-Qa-1^b (C), and J774 (D) cells were infected with S. enterica serovar Typhimurium C5 and tested for recognition by SalTc 1.69. (E) The average numbers of intracellular bacteria for J774, EL-4, A20, and B78Hi-Qa-1^b were 3.2, 1.2, 1.5, and 2.5, respectively. The average number of viable bacteria, determined 2 h after infection, represents the mean taken from three separate samples each of 10⁶ cells.

FIG. 3.

Sensitization of target cells to CTL lysis does not correlate with bacterial invasion. (A) J774, P815, and L-Qa-1^b cells infected with C5 were lysed with 1% Triton X-100 for quantification of intracellular bacteria 2 h following infection. The average number of viable bacteria, determined 2 h after infection, represents the mean of three separate samples, each of 10⁶ cells. (B) Cellular lysates prepared from infected (C5) and uninfected (−) target cells were probed with immune sera against bacterial Hsp 60 as described in Materials and Methods.

FIG. 4.

Pretreatment of target cells with cytochalasin D does not affect bacterial Hsp 60 antigen processing. J774 (A) and P815 (B) cells were left untreated (squares) or exposed to cytochalasin D at 50 μg/ml (diamonds) for 1 h at 37°C. The cells were then washed and infected with C5. (C) The average number of viable bacteria, determined 2 h after C5 infection, represents the mean taken from three separate samples, each of 10⁶ cells.

FIG. 5.

Susceptibility of infected target cells does not depend on the type III protein secretion system of S. enterica serovar Typhimurium. J774 (A) and P815 (B) cells were infected with parental strain SR11 (squares), the invA⁻ mutant (diamonds), or the invE⁻ mutant (circles), respectively. Infected target cells were tested for lysis by clone SalTc 1.69.

FIG. 6.

Detection of Hsp 60 in Salmonella-infected cells. J774 cells were infected at an MOI of ∼100 followed by incubation in the presence of 0.1 mg of gentamicin/ml. At 0, 1, 2, and 3 h after infection, cells were harvested, resuspended with (+) or without (−) 30 μg of proteinase K (pk)/ml, and incubated for 15 min at 37°C. Following washing, cells were solubilized in SDS and the presence of bacterial Hsp 60 was detected by Western blotting with rabbit anti-Hsp 60. The arrow designates the Salmonella Hsp 60 species detected when cells were infected with strain SR11 (wild type, top panel) or invasion-deficient strain invA (middle panel) or invE (bottom panel). Extracts from uninfected cells (lane U) were analyzed in parallel. wt, wild type.

FIG. 7.

Chloroquine interferes with CTL recognition. Salmonella-infected P815 cells were treated with 50 μM chloroquine (circle) or were mock treated (square) prior to labeling with ⁵¹Cr and testing for lysis by bulk Salmonella-specific CTLs derived from the spleen of infected BALB/c mice (left panel) or the Salmonella-specific CTL clone SalTc 1.69 (right panel). Control uninfected cells are also shown (triangle).