Activation of human CD4+ T cells by targeting MHC class II epitopes to endosomal compartments using human CD1 tail sequences

Abstract

Distinct CD4(+) T-cell epitopes within the same protein can be optimally processed and loaded into major histocompatibility complex (MHC) class II molecules in disparate endosomal compartments. The CD1 protein isoforms traffic to these same endosomal compartments as directed by unique cytoplasmic tail sequences, therefore we reasoned that antigen/CD1 chimeras containing the different CD1 cytoplasmic tail sequences could optimally target antigens to the MHC class II antigen presentation pathway. Evaluation of trafficking patterns revealed that all four human CD1-derived targeting sequences delivered antigen to the MHC class II antigen presentation pathway, to early/recycling, early/sorting and late endosomes/lysosomes. There was a preferential requirement for different CD1 targeting sequences for the optimal presentation of an MHC class II epitope in the following hierarchy: CD1b > CD1d = CD1c > > > CD1a or untargeted antigen. Therefore, the substitution of the CD1 ectodomain with heterologous proteins results in their traffic to distinct intracellular locations that intersect with MHC class II and this differential distribution leads to specific functional outcomes with respect to MHC class II antigen presentation. These findings may have implications in designing DNA vaccines, providing a greater variety of tools to generate T-cell responses against microbial pathogens or tumours.

Figure 1

Diagram of DNA constructs created for this study. (a) Domain organization of the human CD1b cDNA including the leader peptide (black bar), the α1, α2 and α3 domains that constitute the extracellular/ectodomain (striped bar), the transmembrane (spotted bar), and the cytoplasmic tail domain (open bar). (b) Domain organization of the GFP-tCD1a–d constructs utilized in accompanying intracellular trafficking studies. These constructs differ from the parent CD1b sequence in that the CD1b ectodomain has been substituted with GFP sequence (green bar) and fused to all four CD1a–d cytoplasmic tail sequences (orange, open, pink and yellow bars). (c) Domain organization of the GroES-tCD1a–d constructs utilized in antigen presentation assays which are identical to the GFP-tCD1 constructs with the exception of the substitution of the CD1b ectodomain with GroES sequence (blue bar). (c) Domain organization of the GroES-tCD1a–d-GFP constructs utilized to quantify antigen expression and for use in T-cell assays. Constructs are organized in a manner similar to the GFP-tCD1 constructs described with two notable exceptions: (1) the M. leprae GroES cDNA replaces GFP sequence and (2) all constructs contain a vector-derived IRES-GFP sequence to facilitate quantification of gene expression.

Figure 2

Comparison of GFP-tCD1 fusion protein localization with wild-type CD1 counterpart. HeLa cells were cotransfected with the following pairs of plasmids: wild-type CD1a and GFP-tCD1a, wild-type CD1b and GFP-tCD1b, wild-type CD1c and GFP-tCD1c, and wild-type CD1d and GFP-tCD1d. The intracellular distribution of GFP (depicted in green) and the wild-type CD1 proteins (depicted in red) detected using isoform-specific antibodies and Alexa 565-labelled goat anti-mouse secondary antibody was analysed by confocal microscopy. The corresponding green and red confocal immunofluorescence images were then merged to detect compartments containing both proteins (in yellow); see right-hand column.

Figure 3

(a) Comparison of GFP-tCD1 fusion protein localization to endosomal markers. HeLa cells were transfected with GFP-tCD1a–d constructs (in green) and the intracellular distribution of GFP was compared to known endosomal markers (in red) for early/recycling (ARF6-T27N), early/sorting (CD71), and late endosomes/lysosomes/MIIC (CD63 and CD107a). Confocal immunofluorescence images in this figure result from the superimposition of the GFP and endosomal marker images for each transfection to detect colocalization (in yellow). (b) Intersection of GFP-tCD1 proteins and MHC class II+ vesicles. HeLa cells were transfected with GFP-tCD1a–d constructs and MHC class II expression induced by IFN-γ. The intracellular distribution of GFP is presented in green and that of HLA-DR in red. Images acquired for each GFP protein and HLA-DR were superimposed to determine the degree of colocalization between GFP-tCD1 and HLA-DR (yellow).

Figure 4

GroES-tCD1 fusion proteins activate CD4⁺ T cells. (a) Antigen-specificity and HLA restriction of the human GroES-reactive CD4⁺ T-cell line, 914B GroES. HeLa cells were transfected with constructs encoding HLA-DRB5*0101, HLA-DRB1*1504, or empty vector, as described, and used as antigen-presenting cells to stimulate GroES-reactive T cells in the presence of various antigens. IFN-γ ELISA was performed to determine the level of T-cell activation. Results are representative of two experiments. Error bars represent standard error of the mean (SEM). (b) Presentation of the GroES₂₈₋₃₉ epitope to GroES-reactive T cells requires endosomal acidification. HLA-DR-matched PBMCs incubated with concanamycin A (left panels) or chloroquine (right panels) before the addition of rGroES protein (0·01 μg/ml, upper panels) or synthetic GroES₂₈₋₃₉ peptide (0·1 μm, lower panels) were used as antigen-presenting cells to stimulate GroES-reactive T cells. Results in 4B are representative of two experiments. (c) HLA-DRB5*0101-expressing HeLa cells present GroES-tCD1 to T cells. HeLa cells were cotransfected with HLA-DRB5*0101-encoding plasmid and DNA and one of a panel of CD1-targeted GroES (or untargeted GroES) expression constructs. All transfectants were treated with rIFN-γ as described before use as antigen-presenting cells in culture with GroES-reactive T cells. GroES DNA concentrations utilized were 10, 33 and 100 ng with the exception of the GroES-tCD1a construct, which was also tested at 333 ng per transfection. Results shown are representative of two independent experiments, both performed with three independent preparations of plasmid DNA per construct tested. Mean fluorescence intensity values of GFP expression as a measure of antigen quantity was determined by flow cytometry and performed in triplicate tranasfections (mean ± SEM). The y-axis values shown represent IFN-γ production (mean ± SEM) from the T cells (determined in triplicate). (d) HLA transfected HeLa cells express functionally equivalent levels of HLA DRB5*0101. HeLa cells transfected with HLA DRB5*0101 and GroES plasmids were pulsed with GroES peptide (0·1 μm), then added to GroES-reactive T cells. IFN-γ was detected by ELISA. Values shown are the means of triplicates.