Compartmentalized MHC class I antigen processing enhances immunosurveillance by circumventing the law of mass action

Abstract

MHC class I molecules function to display peptides generated from cellular and pathogen gene products for immune surveillance by CD8(+) T cells. Cells typically express approximately 100,000 class I molecules, or approximately 1 per 30,000 cellular proteins. Given "one protein, one peptide" representation, immunosurveillance would be heavily biased toward the most abundant cell proteins. Cells use several mechanisms to prevent this, including the predominant use of defective ribosomal products (DRiPs) to generate peptides from nascent proteins and, as we show here, compartmentalization of DRiP peptide generation to prevent competition from abundant cytosolic peptides. This provides an explanation for the exquisite ability of T cells to recognize peptides generated from otherwise undetected gene products.

Fig. 1.

Saturation of the class I pathway by Venus-Ub-peptide expression. L-K^b cells were infected by V-U-SIIN (A and E). L-D^b cells were infected by V-U-ASN (B and F), V-U-SSL (C and G), and V-U-LSL (D and H). Flow cytometry was used to quantitate the time postinfection versus expression of Venus (Lower) and cell surface CPC expression (Upper), as determined by the binding of the indicated TCR-like antibodies specific for the cognate peptide complex. The MOI (0.1–10) and percentage of infected cells as determined by Venus expression at the last time point measured are indicated for each infection. Similar results were obtained in an additional experiment.

Fig. 2.

RGY competes with cytosolic or ER-targeted SIIN peptide. L-K^b cells were coinfected at an MOI of 1–10, with the SIIN-expressing rVV indicated along with the rVV expressing the peptide indicated as a mCherry-Ub-peptide gene product. K^b-SIIN cell surface expression was then measured by 25-D1.16 staining at the times indicated postinfection. (A) V-U-SIIN. (B) M-U-SIIN. (C) EGFP-U-SIIN. (D) Mg-OVA. (E) ES-OVA. (F) Similarly, L-D^b cells were coinfected with V-U-LSL and the panel of rVVs expressing other peptides. D^b-LSL levels were determined by 2E1 staining at the specific times postinfection. Similar results were obtained in three additional experiments.

Fig. 3.

Overall cell surface class I expression as a measure of peptide loading. L-K^b and L-D^b cells were infected with rVVs expressing Venus-Ub-peptide gene products. D^b and K^b surface levels were determined using specific antibodies at predetermined times postinfection.

Fig. 4.

Lack of competition between RGY and peptides from standard gene products. L-K^b cells were coinfected at an MOI of 1:10, with the SIIN expressing rVV indicated, along with the indicated rVV expressing the peptide indicated as a mCherry-Ub-peptide gene product. Similar results were obtained in six or more additional experiments.

Fig. 5.

Cytosolic RGY or SIIN do not compete with natural VV-encoded K^b-restricted peptides. (A and B) L-K^b cells were infected at an MOI of 10 with rVVs UV irradiated for the times indicated (in parentheses) and incubated with or without zLLL to block proteasome-dependent generation of peptides from VV and cellular gene products. Cell surface CPC expression was determined at the times indicated postinfection by staining with 25-D1.16 (A) or ASN-E10 (B). Similar results were obtained in a repeat experiment. (C) Low molecular weight, TCA-soluble peptides recovered from L-K^b cells infected for 6 h with the rVVs indicated were HPLC-fractionated. Fractions were tested for antigenic activity using in vitro–expanded polyclonal rVV-specific CD8⁺ T cells, as determined by the fraction of cells synthesizing IFN-γ on exposure to the corresponding fraction. The eluting times for the synthetic peptides corresponding to defined immunodominant K^b-restricted peptides B8R, A47L, and A19L are indicated by black arrows; other major antigenic activities are indicated by red arrows. The antigenic activity present in fraction 122 derived from RGY-expressing cells is a contaminant likely emanating from the B8R fractions; it was not present in a repeat experiment that yielded otherwise similar results. NP₃₆₆-expressing virus demonstrates greater antigenic activity in many fractions (particularly fractions 143–145, containing an undefined peptide), most likely due to higher levels of vaccinia virus gene expression.