Uncovering fundamental mechanisms underlying "dynamic" mechanotransduction
Cells in our body are exposed to various mechanical forces from their neighbors and environment: for example, endothelial cells lining in the blood vessel are exposed to shear stress and pulsatile pressure from the blood flow. At the basal surface of the cell, however, cells are interfacing with something called extracellular matrix (ECM), which supports the cell not only chemically but also mechanically. In recent 20 years, it has been revealed that the rigidity of the extracellular matrix can greatly influence physiology and pathology of cells and tissues, including differentiation, survival, proliferation, altered drug response, and tumor progression. For example, in the case of tumor, the increase in the tissue stiffness without any changes in genetic information and chemical environment, can cause tumor progression. There is also an evidence showing that cancer-targeting drug does not work when cancer cells are highly contractile in very tensed environment.
Fig. 1. Dynamic and heterogeneous adhesion assembly and disassembly in the protrusion of Chinese Hamster Ovary epithelial cell, classified by Machine Learning
Mechanical sensing of the ECM by cells occurs through cell-matrix adhesions that link the ECM to the cytoskeleton. As a mechanical anchor, cell-matrix adhesions transmit mechanical forces from the actin cytoskeleton to the ECM. At the same time, they transduce the force they “felt” into biochemical signals, cellular response to which lead to changes in a wide variety of cell functions including structural reinforcement. While structure, function, and signaling by adhesions have been characterized at the level of large, mature focal adhesions, however, it has been elusive whether mechanosensing can occur before they reach their full maturation: i.e. in the state of nascent adhesions. Indeed these adhesions are very dynamic: they constantly form, turn over and mature (Fig.1).
Fig.2. (Left) vinculin of a CHO cell (green) and fluorescent beads on the gel (green) overlaid with traction vectors. (Right) Color coded of traction magnitude (blue to red, low to high) overlaid with traction vectors. Image credit: Alexia Bachir.
To study this dynamic mechanotransduction, 1) a tool to resolve small force from tiny adhesions and 2) a framework that allows linking mechanical force to highly heterogeneous molecular events are required. We address these issues experimentally and computationally. First we develop traction force microscopy (TFM) algorithm that is very sensitive to small forces from nascent adhesions as well as big forces from focal adhesions (Fig. 2.) while insensitive to noise in non-cell area, by introducing sparsity regularization into the inverse problem solution.
Second, we address the heterogeneity in the population of nascent adhesions with single particle tracking, machine learning and general image analysis of microscopic images (Fig. 3.). The measurements from these tools suggest evidence of independent force-transmission and mechano-sensitivity of a significant subset of nascent adhesions and highlight differential recruitment of early adhesion molecules (e.g. talin, vinculin, and paxillin).
We are expanding the framework of the TFM and adhesion-tracking-and-classification to 1) dynamic ECM rigidity sensing, 2) shear flow mechanotransduction and 3) cancer cell progression and metastasis.
Fig.3. Machine Learning of nascent and focal adhesions.
High-resolution Traction Force Microscopy Development
A critical component in MechanoBiology research is to be able to measure the traction force as accurately as possible. Based upon our recent development, we are refining the TFM tool with 1) better deformation tracking (for large deformation), 2) taking into account 3rd dimension, and 3) using magnetic force to calibrate force-deformation function in vitro.