Humoral determinants of checkpoint immunotherapy: anti-TL1A treatment has a synergistic effect with anti-PD-1 treatment

While the role of cellular immunity in checkpoint immunotherapy (CPI) for cancer is well established, the impact of antibody-mediated humoral immunity is comparably underexplored. Here we used rapid extracellular antigen profiling (REAP) to map the autoantibody reactome within a cohort of 374 cancer patients treated with CPIs and 131 healthy controls for autoantibodies against 6,172 extracellular and secreted proteins (the ‘exoproteome’). Globally, CPI-treated cancer patients harbored an exquisite diversity of autoreactivities that were elevated relative to controls, but changed minimally with treatment. Autoantibody signatures within CPI-treated patients strikingly distinguished them from controls. While associations of specific autoantibodies with immune-related adverse events (irAE) were sparse, we detected numerous individual autoantibodies that were associated with dramatically altered odds ratios (ORs) for response to therapy. These included autoantibodies against immunomodulatory proteins, such as cytokines, growth factors, and immunoreceptors, as well as tumor surface proteins. Functional evaluation of several autoantibody responses indicated that they neutralized the activity of their target proteins including type-I IFN (IFN-I), IL-6, OSM, TL1A, and BMPR1A/BMPR2. Modeling the effects of autoantibodies against IFN-I and TL1A in preclinical mouse tumor models resulted in enhanced CPI efficacy, consistent with their effects in patients. This single-cell RNA sequencing dataset evaluates the synergistic effect between anti-PD-1 and anti-TL1A treatments. Mice used for single RNA sequencing were engrafted with MC38b and treated with anti-mouse PD-1 and anti-mouse TL1A. Mice were euthanized 24 hours after the 2nd dose of saline, anti-mouse PD-1 and/ or anti-mouse TL1A and tumors were harvested for analysis. Tumor tissues were digested. Cell surface staining of single-cell suspensions from tumors was performed after two times wash with FACS buffer (PBS supplemented with 2% FBS). All samples were first stained with LIVE/DEAD™ Fixable Blue to exclude dead cells and incubated with Fc receptor blocking antibody (Biolegend, #101320). Then, cells were stained with anti-CD45 (30-F11) and anti-CD3 (17A2). Biological replicates for saline control (n=2) and treated group (n=4) were individually processed. Biological replicates were then pooled together at the single cell suspension stage before sorting with equivalent number of cells from each replicate. The following populations were sorted: P1: CD45+CD3+ (TIL cells), P2: CD45+CD3− (non-TIL immune cells), P3: CD45−CD3− (tumor and stromal cells). P1, P2, and P3 for each sample were then mixed back together at a 9:9:2 ratio, respectively. 20,000 cells from each treatment condition were loaded onto the 10x Genomics Chromium System. Library preparation was performed with 10x Genomics reagents according to the manufacturer’s instructions by the Yale Center for Genome Analysis (YCGA). Libraries were then sequenced with an Illumina NovaSeq 6000 at the YCGA.