Endo-Lysosomal Assembly Variations Among Human Leukocyte Antigen B (HLA-B) Allotypes

Abstract

The human leukocyte antigen class I (HLA-I) complexes comprise a highly polymorphic heavy chain, an invariant light chain, and a peptide. The heavy chain is encoded by three sets of genes: HLA-A, HLA-B, and HLA-C, with HLA-B being the most variable. HLA-I molecules function to present peptides derived from intracellular proteins to CD8+ T cells, which can recognize whether the peptides are self or non-self-derived. For assembly with peptide, HLA-I heavy and light chains associate with a group of chaperones called the peptide loading complex (PLC) in the endoplasmic reticulum (ER). Due to the importance of this pathway, the PLC is targeted by a variety of pathogens to evade the immune response. Additionally, cancer cells often have mutations or loss of PLC-encoding genes. HLA-I allotypes are known to differ in their dependencies on PLC components for their assembly and surface expression. Some allotypes that can assemble independently of PLC components display reduced cell surface stability and more rapid endocytosis. For such allotypes, the role of the endo-lysosomal system in the maintenance of constitutive HLA-I surface expression is unclear. Additionally, it is unclear whether such allotypes have advantages during presentation of exogenous (endocytosed) antigens. We hypothesized that a PLC-independent allotype such as B*35:01 is more efficient at endo-lysosomal assembly compared with a more canonical PLC-dependent allotype such as B*08:01. We addressed this hypothesis in the primary human antigen presenting cell (APC) subsets monocytes and monocyte-derived dendritic cells (moDCs), since APCs are generally efficient at binding exogenous antigen within endo-lysosomes. We found that the cell types differ in their HLA-B subcellular distributions and cell surface half-lives; moDCs had larger intracellular HLA-B pools and shorter surface half-lives compared to monocytes, particularly for B*35:01. moDCs displayed sub-optimal conformers of B*35:01 but not B*08:01. Monocyte HLA-B within the endo-lysosomal system was primarily localized to lysosomes, while moDCs contained HLA-B largely co-localized with Rab11+ recycling endosomes. Disruption of the endo-lysosomal pH gradient with bafilomycin rescued HLA-B surface expression in monocytes, but reduced B*35:01 expression and increased its co-localization with lysosomes in moDCs. B*08:01 was unaffected in moDCs. Together, these findings indicate that in moDCs, B*35:01 is more dependent on endo-lysosomal trafficking and assembly for its surface expression, while B*08:01 does not depend on endo-lysosomes for its cell surface expression. When monocytes and moDCs were used to compare efficiencies of the presentation of B*08:01 and B*35:01 epitopes from two exogenous antigens (cross-presentation efficiencies), we found that B*35:01 epitopes were cross-presented more efficiently, even when CTL sensitivities for peptide were matched across allotypes. We suggest that this difference arises from the ability of B*35:01 to utilize the endo-lysosomal pathway of cross-presentation more efficiently. The findings suggest that the compartmental flexibility afforded by the PLC-independent assembly modes of B*35:01 allows more efficient endo-lysosomal assembly and confers a cross-presentation advantage. Together, the findings of this study suggest a model wherein the ability of some HLA-B allotypes to undergo noncanonical assembly in the ER allows for supplemental assembly within the endo-lysosomes. This provides a mechanism for the HLA-I system to efficiently sample endo-lysosomal antigens. B*35:01, a relatively PLC-independent allotype, is dependent on endo-lysosomal trafficking for its constitutive expression in moDCs and can assemble with exogenous antigen within this pathway more effectively. We expect that these findings have important implications for understanding HLA-I associations with disease control.