J. Flaming1, R. Chandra1, L. Girard1, D. Ganguly1, J. Toombs1, J. Minna1, R. Brekken1 1University Of Texas Southwestern Medical Center,Hamon Center For Therapeutic Oncology Research,Dallas, TX, USA
Introduction:
Macrophages are key regulators of the immune landscape within the tumor microenvironment (TME) and the plasticity of macrophage phenotypes in the TME correlates with prognosis in Non-Small Cell Lung Cancer (NSCLC). Depending on their phenotype, macrophages in the TME can secrete pro-tumorigenic cytokines and chemokines, ultimately suppressing the function of other immune cells in the TME. The purpose of our study was to investigate how individual NSCLC cell lines alter macrophage phenotype in tumor cell, cancer-associated fibroblast (CAF) and macrophage co-cultures and to relate effects to the molecular characteristics of different NSCLCs. We hypothesize that immune suppression occurs through tumor-secreted signaling molecules that if identified and blocked, could alleviate macrophage-mediated immune suppression, resulting in improved anti-tumor immune responses
Methods:
We developed an in vitro organoid co-culture system (NSCLC tumor cells, human CAFs, and mouse macrophages) to interrogate cancer cell features causing heterogeneity of macrophage phenotypes across a panel of NSCLCs. We measured (with 4-7 replicates for each NSCLC cell line): mRNA expression in mouse macrophages with a panel of qPCR probes for important macrophage related genes (Arg, NOS2, IL-1β, IL-6, CHIL-3, SOCS3), and in selected cases, whole genome RNAseq; and protein expression using cytokine arrays measuring expression of 40 inflammatory cytokines. Positive controls were stimulation with LPS and IL-4.
Results:
Using our platform, we characterized over 70 NSCLC patient derived lines for their effects on mouse macrophage phenotype. We found: 1) the macrophage phenotypes induced by any one NSCLC were highly reproducible; 2) CAFs are important in the co-culture polarization of macrophages 3) three major clusters of cancer polarized macrophage phenotypes: high Arg (immune suppressive), high IL-1β (inflammatory) or high SOCS3 (JAK-STAT3 pathway) expression; 4) major oncogenotypes (KRAS, TP53, STK11, EGFR, BRAF) do not correlate with the induced macrophage phenotype. We selected 10 NSCLC “exemplars” lines representing each of these 3 clusters for RNA sequencing (mouse genes) and cytokine array protein (human) profiling. Across all clusters we found: 1) suppression of macrophage endocytosis pathways and activation of scavenger receptor A (SRA) signaling (M2 immunosuppressive phenotype); 2) increased expression of human IL6, IL8, and MCP1 proteins, which have been implicated in suppressing innate immune tumor sensing. Analyses of differences between the 3 clusters is ongoing.
Conclusion:
Patient derived NSCLC preclinical models have reproducible effects on macrophage phenotypes in co-cultures. Three major classes of NSCLC cell lines, which are not linked to oncogenotype, initiated reproducible macrophage alteration. Cytokines secreted by NSCLC cell lines appear responsible for these macrophage changes and this system provides an experimental platform to systematically test each as potential therapeutic targets.