1. Cytoskeletal and membrane regulation during leader bleb-based migration in non-adhesive confinement. Leader bleb-based migration (LBBM) that occurs after MAT under high confinement and low adhesion is a very recently discovered and uncharacterized mode of migration. It is also very scary, as it is fast and persistent in the absence of external cues. Although I do not think that metastasis could be mediated by cue-deficient migration, LBBM could promote the ability of cells to continue migrating between small regions of tissue that lack the appropriate cues or ligands, giving rise to the plasticity in tumor cell behavior. LBBM-like morphology has been observed in migrating cells in vivo during development and can be seen in Erik Sahais movies of melanoma tumors in mice. Thus, understanding the mechanism of this weird and newly discovered migration mode is definitely worthwhile. The cytoskeletal dynamics driving LBBM have been reasonably well-characterized, however, the basic cell anatomy, i.e. the localization and organization of cytoplasmic organelles that allow the cell to migrate and survive the tight confinement and unusual morphology of LBBM, have not be studied. In addition, how actin regulatory proteins mediate this unusual morphology and motility is not elucidated. Greg Adams from Xuebai Yaos lab arrived a year or so ago and has performed a survey of the localization and effects of overexpression of fluorescent-tagged markers of all major cytoskeletal and membrane-bounded organelle systems as well as an array of actin binding and regulatory proteins in human A375 metastatic melanoma cells undergoing LBBM under confinement. It should be noted that if gene over-expression promotes MAT or LBBM, it may be a candidate metastasis-promoter and thus a possible drug target. We developed image analysis algorithms to determine the fraction of the organelle in the cell body versus the leader bleb and its spatial distribution within these compartments, as well as the effect of overexpression on the frequency of MAT and cell motility parameters. Preliminary results indicate that most organelles are localized in the cell body; however, microtubules, endoplasmic reticulum, Golgi, and some mitochondria reside in the leader bleb. Greg has preliminary data suggesting that overexpression of the actin crosslinking proteins -actinin and filamin, but not any other actin regulatory proteins, promote MAT and faster LBBM. As these are well-known effectors of cell cortical mechanical properties whose molecular regulation are reasonably well characterized, we are excited to dissect the molecular regulation of these proteins in mediating cortical contractility that drives LBBM, and to determine their role in metastasis. I also have a new post doc, Ankita Jha, from Tomas Lecuits lab that is starting in March, who, in collaboration with Jennifer Lippincott-Schwartz wants to work on the role of membrane-based signaling in regulating polarization of the cytoskeleton in LBBM and confined collective cell migration. 2. Nuclear Integrity during metastatic melanoma cell migration in tumors. Michelle Baird, the infamous cloning guru from Mike Davidsons lab who was willed to me as a grad student from his death-bed, has started an exciting project in which we will use in vitro confined migration assays to look for proteins that mediate nuclear protection, and then take that knowledge in vivo and image the consequences during melanoma metastasis in mice. To hone in on which proteins could be playing a role in this process and are relevant to melanoma metastasis, we are taking advantage of our Bioinformatics core facility. We are using publicly available RNA-Seq data from patient melanoma tumor samples in various stages of disease progression and applying differential expression and pathway analysis tools to identify significantly upregulated cytoskeletal or nuclear/nucleoskeletal genes in invasive melanoma compared to dysplastic nevi. We have identified a few dozen high priority targets, and will perform siRNA in a well-characterized invasive human melanoma cell line 1205lu that we were pointed to by our melanoma collaborator in NCI, Glenn Merlino. We will use tumor spheroids in collagen gels to examine nuclear shape and dynamics, migration through confined micro-channels to assay nuclear integrity, and FISH to assay genomic instability. We will then do our typical Waterman lab imaging-based mechanistic cell biology to determine how our best target mediates nuclear protection via the cytoskeleton. We will then characterize the role that altering nuclear shape has on metastatic behavior and intratumor dynamics using intravital microscopy in a dorsal skin window of primary melanoma in a nude mouse. I am excited to gear up for in vivo imaging, which will be done in collaboration with our Imaging Core facility director, Chris Combs, who has this as his primary expertise. 3. ECM fibril curvature control of cell polarization and migration. In many cellular environments in vivo in both normal and disease states, collagen and other ECM components are organized into wavy, sinusoidal bundles rather than purely linear fibers. In breast ductal carcinoma, the more aligned and straight the collagen fibers around breast ducts the poorer the prognosis, suggesting that straight bundles promote cancer outgrowth. In contrast, in ovarian stromal tumors, higher curvature radii of wavy collagen fibers are predictive of tumor grade. Thus, it may be important to understand how and why cells respond differently to the topological cue of ECM fiber curvature. To investigate how cells polarize and migrate on wavy ECM fibrils, Bob Fischer is collaborating with John Fourkas and Wolfgang Losert in Physics at U Maryland using nanofabrication to create substrates with sinusoidal ridges of approximate cross-section of large collagen fibrils, but with varying wavelength or amplitude in ranges similar to what we can measure from published images of tumor ECMs from the Keely, Fischbach, and Campagnolo labs, along with unpublished data from Claudia Fischbach (Physics, Cornell). Bob is analyzing the effect of amplitude and wavelength on cell shape polarization and migration. Preliminary live-cell imaging shows that each convex curve of the waveform (1/2 wavelength) sets up a leading edge, with actin polymerization and assembly of myosin II filaments that flow rearward and coalesce into contractile arcs that span 1/2 wavelength, such that each full wave within a cell gives rise to two opposing leading edges. This suggests that fibril curvature sets up polarity by dictating the position of leading edge polymerization and arc formation. We thus predict that wavelengths >2x the cell diameter will produce one leading edge oriented towards the convex curve and cell polarization and migration perpendicular to the direction of wave propagation, while wavelengths <2x the cell diameter will set up multiple opposing leading edges and arcs that coalesce into stress fibers that span multiple waves parallel to the direction of wave propagation, thus polarizing cell migration along this axis. We will analyze localization of polarity markers of the cytoskeleton and signaling (FAs, Arp2/3, free barbed ends, PIP3), and will test the role of integrin engagement, PI3K activity and myosin II activity. We predict that detection and response to fibril wavelength will require myosin II-dependent arc