Metastasis is the most lethal aspect of cancer, yet current therapeutic strategies do not target its key rate-limiting steps. We have previously shown that the entry of cancer cells into the blood stream, or intravasation, is highly dependent upon in vivo cancer cell motility, making it an attractive therapeutic target. To systemically identify genes required for tumor cell motility in an in vivo tumor microenvironment, we established a novel quantitative in vivo screening platform based on intravital imaging of human cancer metastasis in ex ovo avian embryos. Utilizing this platform to screen a genome-wide shRNA library, we identified a panel of novel genes whose function is required for productive cancer cell motility in vivo, and whose expression is closely associated with metastatic risk in human cancers. The RNAi-mediated inhibition of these gene targets resulted in a nearly total (>99.5%) block of spontaneous cancer metastasis in vivo.

Tumor cell extravasation is a key step during cancer metastasis, yet the precise mechanisms that regulate this dynamic process are unclear. We utilized a high-resolution time-lapse intravital imaging approach to visualize the dynamics of cancer cell extravasation in vivo. During intravascular migration, cancer cells form protrusive structures identified as invadopodia by their enrichment of MT1-MMP, cortactin, Tks4, and importantly Tks5, which localizes exclusively to invadopodia. Cancer cells extend invadopodia through the endothelium into the extravascular stroma prior to their extravasation at endothelial junctions. Genetic or pharmacological inhibition of invadopodia initiation (cortactin), maturation (Tks5), or function (Tks4) resulted in an abrogation of cancer cell extravasation and metastatic colony formation in an experimental mouse lung metastasis model. This provides direct evidence of a functional role for invadopodia during cancer cell extravasation and distant metastasis and reveals an opportunity for therapeutic intervention in this clinically important process.

Planar cell polarity (PCP) signaling has been shown in different studies to either promote or inhibit the malignancy of breast cancer. Using the 21T cell lines, which were derived from an individual patient and represent distinct stages of progression, we show that the prototypical PCP ligand, WNT5A, is expressed highest in 21MT-1 cells (invasive mammary carcinoma) and lowest in 21PT (atypical ductal hyperplasia) and 21NT (ductal carcinoma in situ) cells. Overexpression of WNT5A decreased spherical colony formation and increased invasion and in vivo extravasation only in 21NT cells; whereas overexpression increased migration of both 21PT and 21NT cells. WNT5A overexpression also increased RHOA expression of both cell lines and subsequent RHOA knockdown blocked WNT5A-induced migration, but only partially blocked WNT5A-induced invasion of 21NT cells. PCP can signal through VANGL1 to modulate AP-1 target genes (e.g. MMP3) and induce invasion. VANGL1 knockdown inhibited WNT5A-induced invasion of 21NT cells, but had no effect on WNT5A-induced migration of either 21PT or 21NT cells. WNT5A-induced MMP3 expression was seen only in 21NT cells, an effect that was VANGL1 dependent, but independent of AP-1. We thus provide evidence that PCP signaling can act in a context dependent manner to promote breast cancer progression.

Abstract The formation of invasive, rapidly growing metastatic lesions is a critical step in cancer metastasis, the cause of more than 90% of cancer deaths. Development of novel therapeutic approaches that block the invasion step of metastasis is one of the highest priorities for clinical cancer research. For this reason we completed the first genome-wide in vivo shRNA screen for genes that directly contribute to invasive metastatic lesion formation. Using state of the art intravital imaging, we identified over fifty novel regulators of invasive metastatic colony formation in vivo. Interactome analysis links these genes to key cellular processes including: transcriptional regulation of gene expression, mRNA processing and cytoskeletal remodeling. The target list was then prioritized based on clinical gene expression profiles that negatively correlated with key cancer endpoints including metastasis, cancer-specific and overall survival. Pharmacological and shRNA-mediated knockdown of the high priority targets in human cancer cell lines such as prostate cancer and melanoma specifically blocked cancer cell migration and invasion in vitro and in vivo. Moreover, shRNA-mediated knockdown of these genes blocked human cancer cell metastasis in the avian embryo and mouse preclinical models of metastasis. Finally, immunohistochemical analysis on clinical prostate cancer and melanoma tissue samples showed strong correlation with disease progression and metastasis. In summary, we have identified numerous novel genes that that are functionally involved in cancer invasion and metastasis that may serve as predictive markers for disease aggressiveness and represent exciting new pharmacological targets to block cancer invasion and metastasis. Citation Format: Konstantin Stoletov, David Bond, Hon Sing Leong, Emma Woolner, Srijan Raha, Amy Robertson, Francis Wong, Andries Zijlstra, John D. Lewis. In vivo whole genome shRNA screen reveals novel targets to block cancer metastasis. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4972. doi:10.1158/1538-7445.AM2014-4972

Abstract During the metastatic process, cancer cells must undergo trans-endothelial migration from the vessel lumen into underlying tissue in a process known as extravasation. Little is known about the dynamic mechanical aspects of this process. To address this, we performed real-time, sub-cellular resolution intravital imaging of human cancer cells during arrest, intravascular migration and extravasation using a shell-less avian embryo xenograft model. We find that extravasation occurs at endothelial junctions and that the majority of extravasation events occurred during the 12 hours subsequent to the intravenous injection of cancer cells. In the majority of extravasation events, long cytoplasmic extensions identified as invadopodia were observed prior to extravasation which breached underlying endothelium. We observed the release of microparticles during these invadopodial extension events prior to extravasation, and this resulted in a ∼40% reduction in cell volume post-extravasation. Based on this, we hypothesized that invadopodia are required for extravasation, and that cancer cell extravasation could be abrogated by inhibiting factors required for invadopodia function. We found that treatment with Src kinase inhibitors significantly reduced invadopodia formation in vivo and resulted in a ∼60% decrease in extravasation events when compared to vehicle treated controls. Furthermore, a higher proportion of cells in the Src kinase inhibitor-treated animals remained intravascular compared to vehicle control. In conclusion, we determined that 1) extravasation occurs at junctions between adjacent endothelial cells, 2) cancer cells form invadopodia that breach the endothelial layer prior to successful extravasation and 3) cancer cells undergoing extravasation release cancer microparticles into both the vessel lumen and tissue interstitium with a corresponding decrease in cell volume. Pharmacological inhibition of invadopodia by Src kinase inhibitors (Bosutinib, Dasatinib) reduces cancer cell extravasation, revealing a novel and potentially important mechanism of action against metastatic cancers. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 968. doi:10.1158/1538-7445.AM2011-968

Abstract Metastasis contributes to over 90% of cancer mortalities. Metastatic disease is a dynamic, multi-step process that requires invasion and migration of cancer cells from the primary tumour site and leads to colonization at a distant location. Migration events mediated by the family of Rho GTPases are facilitated by acto-myosin contraction at the cell rear. In vitro studies have demonstrated that knockdown of RhoA causes insufficient actomyosin contraction and decreased cell detachment. Migration of cancer cells in vivo does not only require detachment from the primary tumor, but also relies on invasion of the cell into the surrounding stromal tissue. Formation of functional invadopodia drives invasion and a key component of invadopodia is cortactin. The objective of this study is to combine inhibitory short hairpin RNA (shRNA) technology with an in vivo metastasis model to identify mechanisms of cancer cell migration that have not previously been studied in an animal model. I hypothesize that inhibition of mediators required for cellular rear detachment and invadopodia formation, such as RhoA and cortactin respectively, will prevent migration of human fibrosarcoma (HT1080) and human epidermoid carcinoma (Hep3) cell lines in vivo. Inhibition of RhoA and cortactin was performed by lentiviral infection of Hep3 and HT1080 cell lines with individual pLOK.1-puro vectors containing a 21-mir sequence to target and decrease mRNA expression. The migratory ability of both cell lines following RNA inhibition was analyzed in vitro using a scratch wound assay and in vivo using a migration and metastasis model in the shell-less chicken embryo, whereby transduced HT1080 and Hep3 cells are injected intravenously and subsequently arrest in the chorioallantoic membrane (CAM). We used a microscope mounted incubator and spinning disc confocal microscopy to visualize cells in real time and quantify their migration parameters. Results obtained from in vitro migration assays indicate that inhibition of cortactin does not affect the migration of Hep3 and HT1080 cell lines across a rigid substratum. Diminished levels of cortactin also had no significant effect on the migration of these cells in vivo. Importantly, inhibition of RhoA caused a significant decrease in cell migration in vitro for both cell lines and in vivo for Hep3 cells (p<0.05). This data suggests that RhoA, a key mediator of cellular rear detachment, is required for in vivo migration of human epidermoid carcinoma cells while cortactin, implicated in invadopodia structures, is not required. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2367. doi:10.1158/1538-7445.AM2011-2367