Salmonella polarises peptide-MHC-II presentation towards an unconventional Type B CD4+ T-cell response

Abstract

Distinct peptide-MHC-II complexes, recognised by Type A and B CD4(+) T-cell subsets, are generated when antigen is loaded in different intracellular compartments. Conventional Type A T cells recognize their peptide epitope regardless of the route of processing, whereas unconventional Type B T cells only recognise exogenously supplied peptide. Type B T cells are implicated in autoimmune conditions and may break tolerance by escaping negative selection. Here we show that Salmonella differentially influences presentation of antigen to Type A and B T cells. Infection of bone marrow-derived dendritic cells (BMDCs) with Salmonella enterica serovar Typhimurium (S. Typhimurium) reduced presentation of antigen to Type A T cells but enhanced presentation of exogenous peptide to Type B T cells. Exposure to S. Typhimurium was sufficient to enhance Type B T-cell activation. Salmonella Typhimurium infection reduced surface expression of MHC-II, by an invariant chain-independent trafficking mechanism, resulting in accumulation of MHC-II in multi-vesicular bodies. Reduced MHC-II surface expression in S. Typhimurium-infected BMDCs correlated with reduced antigen presentation to Type A T cells. Salmonella infection is implicated in reactive arthritis. Therefore, polarisation of antigen presentation towards a Type B response by Salmonella may be a predisposing factor in autoimmune conditions such as reactive arthritis.

Figure 1

MHC-II accumulates in MVBs in Salmonella-infected cells. MelJuSo were infected for 20 min with invasive GFP-S. Typhimurium (MOI 50). Cell surface MHC-II was labelled (L243) at 12 h post-infection and then cells were fixed (A) or further incubated until 20 h post-infection before fixation (B, C, E and F). Cell sections were processed for cryo-immunoelectron microscopy and HLA-DR localisation was visualised with Protein A-gold (10 nm). (D) Graph represents average amount of gold (HLA-DR)/MVB in each cell analysed. Average amount of gold/MVB was calculated for at least 15 cells per condition and comparison of distributions was assessed by unpaired two-tailed t-test. Boxed areas from (B) and (C) are magnified twofold in (E) and (F), respectively. Histograms show surface HLA-DR measured by flow cytometry in infected and uninfected MelJuSo at time points indicated. Refer to Supporting Information Fig. 1A for gating strategy. Data are representative of two independent experiments.

Figure 2

MHC-II down-regulation by Salmonella requires clathrin but not invariant chain-directed trafficking. (A) HeLa cells stably expressing HLA-DR WT (DRα,β) and cytoplasmic tail mutants were generated. HLA-DR surface expression was assessed by flow cytometry at 20 h post-infection with invasive GFP-S. Typhimurium and compared with HeLa-CIITA (Ii positive) cells. Refer to Supporting Information Fig. 1A and B for gating strategy and representative flow cytometry data. Graph shows percent of normal HLA-DR surface expression in uninfected (GFP-negative) cells combined from at least four independent experiments. (B) HeLa cells stably expressing HLA-DR WT (DRα,β)(Ii negative) were transfected with AP-2, clathrin or control siRNAs. Cells were infected with invasive GFP-S. Typhimurium after 5 days of AP-2 or clathrin depletion and surface HLA-DR was assessed as described in (A). Western blot shows AP-2 and clathrin depletion from representative cell lysates after 5 days of siRNA treatment. The loading control is β-actin. Graph shows percent of normal surface HLA-DR expression in uninfected (GFP negative) cells combined from four independent experiments. Comparison of distributions was performed by unpaired (A) or paired (B) two-tailed t-tests.

Figure 3

Salmonella downregulates I-A and I-E surface expression and presentation of antigen to CD4⁺ T cells. (A) BMDCs were infected with opsonised GFP-S. Typhimurium (MOI 10) then I-A^k (OX6) and I-E^k (14.4.4s) surface expression was compared in infected (GFP positive) and uninfected (GFP negative) CD11c/CD11b⁺ BMDCs by flow cytometry. Refer to Supporting Information Fig. 1A for gating strategy. Histograms (upper panels) show I-E^k surface expression in infected and uninfected BMDCs from a representative of at least four independent experiments. Graphs (lower panels) show percent of normal (GFP negative) I-A^k or I-E^k surface expression combined from four independent preparations of BMDCs infected with WT or SPI2-deficient (ΔssaV) S. Typhimurium. (B) BMDCs (in triplicate) were uninfected or infected with opsonised WT, HKWT or ΔssaV S. Typhimurium (MOI 10). From 20 h post-infection, cells were incubated with HEL protein and Type A CD4⁺ T hybridoma cells (3A9) at a ratio of 5 T cells: 1 BMDC. After 24 h, culture supernatants were harvested and T-cell activation was quantified by IL-2 ELISA. Graph shows percent of normal mean (uninfected) I-A^k-dependent HEL presentation to Type A T cells combined from at least four independent experiments. Antigen presentation in uninfected BMDCs is shown as a dashed line. Comparison of distributions was performed by paired two-tailed t-tests.

Figure 4

Salmonella infection enhances presentation of exogenous peptide to Type B T cells. BMDCs (in triplicate) were infected with opsonised WT (A and B), SPI2-deficient (ΔssaV) (B) or HKWT (B) GFP-S. Typhimurium (MOI 10). From 20 h post-infection, cells were incubated with HEL protein or HEL_46–61 peptide and 3A9 (Type A) or 11A10 (Type B) T hybridoma cells at a ratio of 5 T cells: 1 BMDC. After 24 h, culture supernatants were harvested and T-cell activation was quantified by IL-2 ELISA. (A) Graphs show mean IL-2 concentration from a representative of at least four independent experiments. Error bars represent SD. (B) Graphs show percent of normal (uninfected) I-A^k-dependent HEL_46–61 presentation to Type A or B T cells combined from at least three independent experiments. Antigen presentation in uninfected BMDCs is shown as a dashed line. Comparison of distributions was performed by paired two-tailed t-tests.

Figure 5

Exposure to Salmonella is sufficient to enhance presentation of exogenous peptide to Type B T cells. BMDCs (in triplicate) were infected with opsonised WT, SPI2-deficient (ΔssaV) or HKWT GFP-S. Typhimurium (MOI 10, unless specified (C)). For antigen presentation, BMDCs were incubated with HEL_46–61 peptide and 11A10 (Type B) T hybridoma cells at a ratio of 5 T cells: 1 BMDC. After 24 h, culture supernatants were harvested and T-cell activation was quantified by IL-2 ELISA. (A) At 20 h post-infection, culture supernatant was harvested and incubated with fresh BMDCs, HEL_46–61 peptide and 11A10 (Type B) T cells. Where indicated, culture supernatant was filtered (0.45 μm) prior to incubation with fresh BMDCs. (B) At 20 h post-infection, GFP-S. Typhimurium-infected BMDCs were sorted from the exposed but uninfected population. Refer to Supporting Information Fig. 1A for representative gating strategy. Presentation of 0.5 μM HEL_46–61 peptide to 11A10 (Type B) T cells was compared for BMDCs that were unexposed, exposed but uninfected, or infected with S. Typhimurium. Comparison of distributions was performed by unpaired two-tailed t-tests. (A–C) Data shown are the mean + SD and are representative of one out of at least four independent experiments.