TAP- and Proteasome-Dependent Endogenous Antigen Loading of HLA Class II in Leukemic Blasts Introduces a Promising New Target for Generating Leukemia-Specific CD4+ T Cells

Abstract

According to the classical HLA class II antigen presentation pathway, exogenous antigens are processed in the endosomal/lysosomal pathway and associate with HLA class II after exchange with the class II-associated invariant chain peptide (CLIP). For this reason, the relative amount of CLIP presented by HLA-DR (DR) molecules (CLIP/DR amount) can be considered as an indicator for HLA class II antigen loading. Previously, we showed that Invariant Chain (Ii) down-modulation in the Kasumi-1 and THP-1 AML cell lines led to marked declines in CLIP/DR amount [Van Luijn et al., Haematologica 2008; 93(s1), Abstract 0029]. In addition, the total amount of cell surface-expressed DR was reduced on Kasumi-1 blasts, in line with the need of Ii for the transport of newly synthesized HLA class II molecules into the endosomal/lysosomal pathway. Surprisingly, in THP-1 blasts, Ii down-modulation hardly affected DR expression at the cell surface. In the present study, we further explored the Ii-independent pathway of HLA class II antigen presentation in leukemic blasts. Not only in the THP-1, but also in another AML cell line, the KG-1, Ii down-modulation had no effect on DR expression levels, as determined by flow cytometry. Since DR expression does require peptide binding, Ii-independency in these AML blasts may be achieved by endogenous antigen loading in the endoplasmic reticulum (ER). To test this hypothesis, supply of endogenously derived peptides into the ER was blocked by viral proteins interfering with the function of the transporter associated with antigen processing (TAP). As expected, TAP inhibition in KG-1 blasts by the viral UL49.5 protein (which changes the conformation of TAP and mediates its degradation) resulted in a strong down-regulation (7.7-fold) of HLA class I. Strikingly, TAP inhibition also induced a clear DR− KG-1 blast population (52.3% of total; MFI=1.4) next to the original DR+ KG-1 blast population (36.5% of total; MFI=288.9), demonstrating that DR expression is partly TAP-dependent in KG-1 blasts. Upon sorting of both populations, TAP−DR− blasts had a decreased expression of intracellular Ii as compared to TAP−DR+ blasts (3.9-fold) and wild type blasts (4.7-fold). Additionally, confocal microscopy revealed that in TAP−DR+ blasts, DR localised to the cell surface, indicating that Ii is able to rescue cell surface expression of DR. Indeed, Ii down-modulation in TAP−DR+ blasts caused a 2.3-fold decline in DR expression. The observed differences in TAP, Ii and DR expression between these KG-1 variants were confirmed by Western blot analysis. Furthermore, blocking of proteasome function by the specific inhibitors MG-132 and Bortezomib also caused a marked decrease of DR expression on KG-1 blasts (MFI declined from 289.2 to respectively 76.4 and 114.5). Accompanying reduction in HLA class I levels ascertained specific proteasome inhibition. This confirmed that at least part of the antigens presented by DR on KG-1 blasts was derived from endogenous sources. Similar results were obtained with THP-1 blasts, as both TAP and proteasome inhibition clearly affected DR expression. In line with our observations that in Kasumi-1 blasts, DR expression is Ii-dependent, addition of the TAP and proteasome inhibitors to Kasumi-1 blasts did not affect cell surface expression of DR. In conclusion, our data reveal an alternative Ii-independent, but TAP- and proteasome-dependent cross-presentation pathway in different AML cell lines, which involves HLA class II loading of endogenous antigens in the ER. Therefore, this alternative pathway may serve as a potent immunomodulatory target in leukemic blasts to activate CD4+ T cells specific for a broad range of leukemia-associated antigens.