By Jamshed Arslan, Pharm. D., PhD.
Cancers in the brain often come from tumors elsewhere in the body. Several adaptive mechanisms influence brain metastasis, such as blood brain barrier leakage that can be induced by stroma or disseminated tumor cells. The brain tumor microenvironment is complex, but the metastasizing cancer cells share common molecular features. To define such molecular characteristics of brain metastasis in vivo, a team led by Dr. Don Nguyen of Yale University School of Medicine developed an advanced RNA sequence-based approach. Their novel method identifies transcriptomic signatures shared by various metastasized tumors and distinguishes between tumor and stromal transcriptomes. Researchers transferred human tumor cells to immunosuppressed mice to generate a xenograft model of metastasis. Their innovative technique, called brain metastasis xenograft-RNA sequencing (BMX-seq), enabled them to appreciate both the tumor plasticity, e.g. altered expression of thousands of genes, and the stromal neuroinflammatory response, e.g. increased microglial-bound Tim3.
Transcriptomic features of brain metastasis are context-dependent
The team developed a xenograft model of metastasis by injecting human lung cancer cells with bioluminescent reporter into the immunosuppressed mice. As expected, the cancer metastasized to different parts of the brain. Their sophisticated RNA sequencing system for xenograft tissues revealed a distinctive gene expression of mouse cells and human cancer cells even at early stages of metastasis. The murine stroma had activated innate immune responses and downregulated ubiquitin pathways compared to xenograft brain metastasis. Transcriptomic analysis further revealed that tumor cells in the brain had activated pathways related to axonal guidance, epithelial-to-mesenchymal transition and calcium signaling relative to the same cells grown in culture. The team found a reduced expression of genes involved in hypoxia and angiogenesis in the metastasized cells in the brain relative to the same cells in subcutaneous tumors.
In other words, the BMX-seq technique showed that brain metastasis has transcriptomic features that are context-dependent. The next step was to study the response of brain metastatic stroma.
Tumor-bearing stroma exhibits neuroinflammatory response
In the human lung cancer cells that colonized mice brain, researchers found enhanced expression of genes abundant in the central nervous system, such as AKAP5, EFNB3, CHL1, L1CAM, and NCAM1. In response to such adaptations of metastasized cells to the brain microenvironment, brain stroma exhibited a neuroinflammatory response. For example, tumor-bearing stroma had increased expression of Tim-3 and other genes related to pro-inflammatory molecules (Il1A, Il1b), tumor-associated macrophages (C1qb), activated astrocytes (GFAP), macrophage phagocytosis (CD68), oxidative stress (Ncf1, Cyba), and tissue remodeling (Ccl8, Tgf beta1).
View Brain Markers
In summary, the in vivo transcriptomic signature identified by BMX-seq provides a detailed map of tumor-stromal interactions during the course of brain metastasis.
Significance for biomarker identification in brain metastasis
The transcriptomic hallmarks provided in this research may help identify novel biomarkers of brain cancers. The RNA-based approach can improve diagnosis, and the targets identified in this paper can lead to the treatment of advanced brain tumors.
Jamshed Arslan, Pharm D., PhD.
Previously at the University of Alabama at Birmingham, School of Medicine.
Dr. Arslan studies cell signaling in mitochondrial defects in
C. elegans and transgenic mice
Wingrove, Emily, et al. "Transcriptomic Hallmarks of Tumor Plasticity and Stromal Interactions in Brain Metastasis." Cell Reports, vol. 27, no. 4, 2019, pp. 1277–1292. https://doi.org/10.1016/j.celrep.2019.03.085