HLA reduction of human T cells facilitates generation of immunologically multi-compatible cellular products

Abstract

Adoptive cellular therapies have shown enormous potential, but are complicated by personalization. Because of HLA mismatch, rejection of transferred T cells frequently occurs, compromising the T-cell graft's functionality. This obstacle has led to the development of human leukocyte antigen (HLA) knock-out (KO) T cells as universal donor cells. Whether such editing directly affects T-cell functionality remains poorly understood. In addition, HLA KO T cells are susceptible to missing-self recognition through NK cells and lack of canonical HLA class I expression may represent a safety hazard. Engineering of non-canonical HLA molecules could counteract NK cell recognition, but further complicates the generation of cell products. We here show that HLA KO does not alter T-cell functionality in vitro and in vivo. While HLA KO abrogates allogeneic T-cell responses, it elicits NK-cell recognition. To circumvent this problem, we demonstrate that selective editing of individual HLA class I molecules in primary human T cells is possible. Such "HLA reduction" not only inhibits T-cell alloreactivity and NK-cell recognition simultaneously, but also preserves the T-cell graft's canonical HLA class I expression. In the presence of allogeneic T cells and NK cells, T cells with remaining expression of a single, matched HLA class I allele show improved functionality in vivo in comparison to conventional allogeneic T cells. Since reduction to only a few, most frequent HLA haplotypes would already be compatible with large shares of patient populations, this approach significantly extends the toolbox to generate broadly applicable cellular products.