MDSC differentiation arrest is a result of a combination of cell cycle and myeloid lineage alteration. A and B, Relative frequency of cells in each Seurat cluster from scRNA-seq data of MDSCs ± 1.5 µmol/L DAB. Relative levels were calculated on the basis of the standardized residuals from using Pearson χ2 overlaid on the UMAP plot of integrated samples (A) and grouped by functional annotation (B). Positive standardized residuals (yellow) for each cluster indicates increased frequency in the DAB-treated sample whereas a negative number (purple) indicates increased frequency in the DMSO control sample. A dotted line indicates standardized residual values of ±3. C, UMAP plot of the integrated scRNA-seq data from MDSCs ± 1.5 µmol/L DAB, showing the trajectory branches calculated by Monocle3 and represented over the 22 clusters defined by Seurat. Most relevant nodes at the end of the branches (black dots) and at intersections (red dots) are shown. Nodes used as starting (a) or end (b and c) points for regulon analysis across the branch are labeled. D, Heat map of the regulons significantly regulated by DAB calculated on AUCell (q < 0.01, |log2FC| >0.01) for each Seurat cluster. Color represents regulons significantly enriched in cells treated with DAB (red), those enriched in control samples (blue) and the ones that do not show significant changes (light yellow). E, Heat map of the regulons significantly regulated during differentiation in the direction from node a to node b in C, which spans from cluster 6 (immature granulocytes) to cluster 9 (intermediate granulocytes) in DMSO control cell cultures. Scale represents z-score of the normalized average AUCell score per regulon. F, Heat map of the regulons significantly regulated during differentiation in the direction from node a to node c in C, which spans from cluster 6 (immature granulocytes) to cluster 5 (mature PMN) in DMSO control cell cultures. Scale represents z-score of the normalized average AUCell score per regulon. DAB = dabrafenib.
ARTICLE ABSTRACT
The effect of targeted therapeutics on anticancer immune responses is poorly understood. The BRAF inhibitor dabrafenib has been reported to activate the integrated stress response (ISR) kinase GCN2, and the therapeutic effect has been partially attributed to GCN2 activation. Because ISR signaling is a key component of myeloid-derived suppressor cell (MDSC) development and function, we measured the effect of dabrafenib on MDSC differentiation and suppressive activity. Our data showed that dabrafenib attenuated MDSC ability to suppress T-cell activity, which was associated with a GCN2-dependent block of the transition from monocytic progenitor to polymorphonuclear (PMN)-MDSCs and proliferative arrest resulting in PMN-MDSC loss. Transcriptional profiling revealed that dabrafenib-driven GCN2 activation altered metabolic features in MDSCs enhancing oxidative respiration, and attenuated transcriptional programs required for PMN development. Moreover, we observed a broad downregulation of transcriptional networks associated with PMN developmental pathways, and increased activity of transcriptional regulons driven by Atf5, Mafg, and Zbtb7a. This transcriptional program alteration underlies the basis for PMN-MDSC developmental arrest, skewing immature MDSC development toward monocytic lineage cells. In vivo, we observed a pronounced reduction in PMN-MDSCs in dabrafenib-treated tumor-bearing mice suggesting that dabrafenib impacts MDSC populations systemically and locally, in the tumor immune infiltrate. Thus, our data reveal transcriptional networks that govern MDSC developmental programs, and the impact of GCN2 stress signaling on the innate immune landscape in tumors, providing novel insight into potentially beneficial off-target effects of dabrafenib.
An important, but poorly understood, aspect of targeted therapeutics for cancer is the effect on antitumor immune responses. This article shows that off-target effects of dabrafenib activating the kinase GCN2 impact MDSC development and function reducing PMN-MDSCs in vitro and in vivo. This has important implications for our understanding of how this BRAF inhibitor impacts tumor growth and provides novel therapeutic target and combination possibilities.