Context-restricted PD-(L)1 checkpoint agonism by CTLA4-Ig therapies inhibits T cell activity

Summary: T cell surface CTLA4 sequesters the costimulatory ligands CD80 and CD86 on antigen-presenting cells (APCs) to prevent autoimmunity. Therapeutic immunosuppression by recombinant CTLA4-immunoglobulin (Ig) fusion proteins, including abatacept, is also attributed to CD80/CD86 blockade. Recent studies show that CTLA4-Ig binding to APC surface cis-CD80:PD-L1 complexes can release the inhibitory ligand PD-L1, but whether this contributes to T cell inhibition remains unclear. Here, we show that PD-L1 liberation by CTLA4-Ig is strictly limited, both in extent and context, relative to PD-L1-competing anti-CD80 antibodies. At APC surface CD80:PD-L1 ratios exceeding 2:1, CTLA4-Ig therapies fail to release PD-L1 regardless of their CD80 affinity. Additionally, introducing flexibility into CTLA4-Ig by modifying its rigid homodimer interface produces biologics that retain bivalent CD80 binding without dissociating cis-bound PD-L1. These findings demonstrate that CTLA4-Ig therapies liberate PD-L1 through a CD80 reorientation mechanism that imposes a strict context dependence to their PD-1 checkpoint agonism and resultant T cell inhibition.