Quantitative assay of human antibody-dependent cell-mediated cytotoxicity (ADCC) using peripheral blood mononuclear cells (PBMCs)

Abstract

Abstract ADCC can be essential for anti-viral immunity and supports many tumor immunotherapies. It is a challenge to quantify and one wants a fast assay that uses few lymphocytes when patient blood is limited. Three considerations affected our new assay design: 1) normal B cells within PBMCs introduce ‘cold target’ competition of radiolabeled tumor B cells if anti-B cell antibodies are present in the ADCC assays; 2) type 1 antibodies to CD20 B cell epitopes are internalized and cleared from the cell surface while type 2 anti-CD20 antibodies remain external; and 3) target cells with MHC I initiate KIR-regulation of lysis that will vary among donors. To address these issues, we used Daudi B leukemia cells that lack MHC I and pretreated the Daudi cells with the humanized nonfucosylated/glycoengineered monoclonal type 2 anti-CD20 antibody obinutuzumab (GA101 G2). Target cell pretreatment prevented ‘cold target’ competition. We used 4 hr 51Cr-release for its sensitivity and varied the PBMC to target (E:T) cell ratios to obtain lytic units. Flow cytometric counts of CD16A Fc-receptor positive cells with TrueCountR beads determined the exact number of ADCC effectors within the unfractionated PBMCs. There was low NK activity to Daudi cells, which required subtraction of simultaneous controls without antibody. We plotted ADCC activity as a log function of the E:T ratios and observed that ADCC was dependent upon interactions of multiple CD16A-positive effectors per target cell, which was expressed as lytic units per 106 CD16A cells, that varied 50 fold among donors. In addition, the linear slopes of CD16A cell lytic cooperativity varied 8 fold. In summary, we have established a fast quantitative assay for comparison of human ADCC that uses as few as 8 million PBMCs.