Project Summary
Resistance to anti-CTLA-4, PD-1/PD-L1 (PD) T cell-based immune checkpoint blockade (ICB) therapy in large
subsets of cancer patients necessitates the search for other tumor-associated T cell immune checkpoints that
critically affect tumor immunity. Lymphocyte activating gene 3 (LAG-3) is a T-cell inhibitory receptor and
represents a promising target for cancer immunotherapy, as monoclonal antibodies (mAb) blocking LAG-3 with
its known ligand, MHC-II, in combination with anti-PD-1 show modest but promising efficacy in recent clinical
trials. The FDA recently accepted for priority review the biologics license application for LAG-3 blocking mAb
relatlimab and anti-PD-1 nivolumab fixed-dose combination for treating melanoma. However, very few single-
agent activities of these MHC-II blocking anti-LAG-3 mAbs were observed, and no useful biomarkers for patient
stratification have been suggested thus far. Motivated by numerous studies suggesting that the
immunosuppressive effect of LAG-3 is largely independent of MHC-II, we identified for the first time that
fibrinogen-like protein 1 (FGL1) is a high affinity and major functional ligand for LAG-3 whose interaction interface is
distinct from MHC-II. FGL1 is a hepatocyte-secreted protein that can be detected in circulation. We found that
soluble FGL1 inhibits antigen-specific T cell activation in a LAG-3 dependent manner. FGL1 genetic ablation or
antibody blockade in mice promotes anti-tumor immunity in established mouse tumors, even in FGL1 negative
tumors, suggesting a functional role of host-derived FGL1. In contrast to its minimal expression in normal tissues
except the liver, FGL1 is upregulated by cancer cells in multiple cancers, such as non-small cell lung cancer
(NSCLC) and melanoma; however, the role of tumor-associated FGL1 in the control of tumor immunity is still
unclear. The major anti-LAG-3 clinical programs currently focus on those blocking only the MHC-II/LAG-3
interaction, which do not elicit any single-agent activities. Thus, it is critical to mechanistically dissect the
immunological function of the FGL1/LAG-3 axis in the tumor microenvironment to design better LAG-3 targeted
approaches for clinical application. In this project and through two complementary aims, we will leverage our
extensive experience and expertise in cancer immunology and immunotherapy to 1) determine the functional role
of tumor-associated FGL1 in modulating tumor immunity, its cancer induction mechanism by oncogenic Kras
mutations and functional biomarker value in predicting responses to immunotherapies; 2) develop FGL1-guided,
next-generation anti-LAG-3 cancer immunotherapy. We will compare the anti-tumor efficacy of our in-house anti-
human or mouse LAG-3 antibodies that differentially block either or both FGL1/MHC-II and map the LAG-3
binding interfaces of these antibodies via CryoEM and X-ray crystallographic studies. In sum, our proposed
studies will significantly deepen our understanding of the newly defined FGL1/LAG-3 axis and foster new
immunotherapy designs for treating human cancers.
