The metastatic spread of lung cancer to distant organ sites is a common complication of the disease; as many as 30-40% of patients present with metastases at diagnosis, at sites including the brain, bone, and liver (3). In the setting of brain metastasis specifically, 40% of lung cancer patients will go on to develop brain metastases, which confers significant decreases in patient quality of life and reduces overall survival to the order of months (1). Therefore, a more comprehensive understanding of the molecular mechanisms required for lung cancer metastasis, particularly to sites such as the brain, is necessary as it might lead to the design of new, more effective treatments.
Previous studies show that metastasizing cancer cells acquire new gene expression patterns upon arrival to distant organ sites (2). Changes in gene expression allow metastatic cancer cells to adapt to new environments, but may also expose these cells to unique vulnerabilities that could be targeted pharmacologically. In order to identify such targets, our lab recently generated a library of metastatic lung cancer cell lines with specific organ tropisms to distal metastatic sites including the brain, bone and liver. We then performed a genome-wide, 3D spheroid-based CRISPR/Cas9 genetic knockout screen to identify differences in genetic dependence between parental PC9 lung cancer cells and brain-metastatic PC9-BrM3 cells. Excitingly, our screen revealed that knockout of a select number of genes significantly impairs growth of brain-metastatic cells relative to parental PC9 cells—suggesting that brain-metastatic cells have developed a unique dependence on these genes for their survival.
The overall goal of this project is to characterize how one particular hit from this screen – encoded by the gene KIF2A – results in a genetic dependence unique to brain-metastatic lung cancer upon gene knockout. KIF2A encodes a kinesin protein, which localizes to the centrosome and is involved in the formation of mitotic spindles. Our recent research also has revealed that brain-metastatic lung cancer cells exhibit centrosomal amplification. Given this metastasis-specific dependence on KIF2A, metastasis-specific centrosomal amplification, and KIF2A’s relationship to the centrosome, we hypothesized that cell lines that are dependent on KIF2A will also exhibit centrosomal amplification. Interestingly, we also observed that PC9-BrM3 cells with sgKIF2A knockouts exhibited centrosome “declustering.” In other words, rather than a distinct, bright circle, their centrosomes appeared spread out and splattered. We thus hypothesized that KIF2A knockout induces centrosome declustering in PC9-BrM3 cells, but not PC9 parental cells.
To test both of these hypotheses, we performed immunofluorescent imaging experiments. First, to test the relationship between KIF2A dependence and centrosomal amplification, we cultured a panel of KIF2A-dependent and independent cell lines. We then seeded these cells for fixation and stained for pericentrin (a centrosome marker), and performed immunofluorescent imaging. We then analyzed centrosome content between conditions. We did not find a significant difference in centrosome content between KIF2A dependent and independent cell lines (Figure 1). Thus, we failed to support our hypothesis. It is possible that we would see a more meaningful relationship if we tested more cell lines.
To test our second hypothesis, we used CRISPR-Cas9 to transduce PC9 parental and PC9-BrM3 cells with sgKIF2A knockouts. We then again performed immunofluorescent imaging and compared centrosome “declustering” phenotypes between PC9 parental sgLacZ, PC9 parental sgKIF2A, PC9-BrM3 sgLacZ and PC9-BrM3 sgKIF2A conditions. Excitingly, we found that PC9-BrM3 sgKIF2A cells exhibit a centrosome “declustering” phenotype that PC9-BrM3 sgLacZ (Figure 2) and PC9 parental cells do not exhibit (Figure 3). This finding supports our hypothesis that KIF2A is involved in centrosomal maintenance in brain-metastatic cells, which may be related to the centrosomal amplification that brain-metastatic cells exhibit. In the future, we wish to explore this relationship with other centrosome-related genes that appeared as hits on our screen (such as NEK1). We also wish to explore how such knockouts may affect the metastatic potential of lung cancer cells. This data may reveal new avenues to target lung cancer metastasis.
References:
