Depletion of SUN1 disrupts mechanosensitive collective cell migration via nuclear and cytoskeletal remodeling

Introduction: Cell migration is a crucial physiological process involved in embryogenesis, wound healing, regeneration and even in metastasis. Numerous previous studies have investigated biochemical and mechanical events associated with both single and collective cell migration. However, there are gaps in knowledge about how nuclear-cytoskeletal communication and subnuclear components, either synergistically or independently, regulate collective cell migration. In our current study, we use LINC (LInker of Nucleoskeleton and Cytoskeleton) complex disruption to understand how nuclear and cytoskeletal changes cause macroscale disruption in collective cell migration. The LINC complex is made of SUN domain proteins on Inner nuclear membrane, which have physical connections with chromatin, and KASH domain proteins on the outer nuclear membrane, which have contact with cytoskeletal elements. Better understanding of how depleted functioning of LINC, specifically SUN1, alters nuclear, cytoskeletal, and multicellular mechanosensing could provide novel migration-related insights into disease states like cancer and laminopathies.

Materials and

Methods: We use human mammary epithelial cells, MCF-10As and their LINC disrupted counterparts (hereafter stated as SUN1L cells; courtesy of Professor Tanmay Lele). SUN1L cells express dominant negative construct for truncated version of SUN domain proteins; this results in sequestration of LINC complex away from the nucleus, resulting in LINC disruption. We make polyacrylamide hydrogels of varying stiffnesses and coat them with collagen type 1 (0.05 mg/ml). We seed the cells as monolayer for 30hrs, and subsequently perform live or fixed imaging. Any applicable drug treatments are performed after 30hrs of culture. We use PIVlab and custom MATLAB codes for analyzing timelapse data, whereas Fiji is used for image analysis.

Results, Conclusions, and Discussions: We observed that SUN1L cells (with LINC disruption) have significantly lower migration speed as compared to MCF 10A controls (Fig B,D). To further delineate this difference, we defined stiffness sensitivity as the difference in average velocity magnitudes between stiff (5kPa) and soft (0.1kPa) gel conditions. Similarly, we defined LINC sensitivity as the difference between average velocity magnitude of control and LINC disrupted cells in a particular stiffness, which indicates the relative importance of SUN1 depletion mechanosensing in any given matrix stiffness. We found that SUN1L cells have reduced stiffness sensitivity than MCF 10A controls (Fig F). We also found that, LINC sensitivity reduces for softer gels (0.1kPa), compared to the stiffer gels (5kPa and 50kPa)(Fig E). Overall, these results demonstrate broad scale disruption to mechanosensitive response of epithelial collectives.
Next, we sought to understand whether LINC disruption affects nuclear and cytoskeletal states within individual cells that form the epithelial collectives. First, we imaged for heterochromatin marker, H3K27me3, and found it was slightly higher in SUN1L cells (Fig I), indicating chromatin compaction after SUN1 depletion. To understand whether SUN1-independent chromatin disruption causes migration changes, we performed SAHA and Anacardic acid (AA) treatments to open or close chromatin, respectively, both of which reduced stiffness sensitivity for 10A controls. Interestingly, for SUN1L cells, stiffness sensitivity increased after closing (AA) but reduced after opening (SAHA) chromatin (Fig F).
To assess cytoskeletal changes, we imaged F-actin and pMLC. While the actin structure appeared unchanged after SUN1 depletion, the pMLC levels were slightly higher in SUN1L cells (Fig G, H), which indicates rise in contractility after LINC disruption. We introduced several cytoskeletal perturbations which reduced the stiffness sensitivity of 10A controls whereas the SUN1L cells were less affected and, in some case, gained stiffness sensing (Fig F). These results suggest that actomyosin mediated contractility may not be the key drivers of mechanosensitive migration here.
We show LINC disruption alters the traditional understanding of cytoskeletal-dependent mechanosensitive cell migration. As such, this could potentially impact our interpretation of cell migration responses to stiffening tumor microenvironments, especially in the context of laminopathies and other mutations targeting nuclear-cytoskeletal communication.

Acknowledgements (Optional):