Growth-factor-induced, persistent fibroblast migration is mediated by mechanical insulation of cell front and back

Cell migration is a crucial process during development, the immune response and wound healing.
As a consequence, aberrant cell migration can lead to tumor metastasis or autoimmune disorders. In order to migrate directionally, cells have to orchestrate a complex machinery of cytoskeletal, adhesion and signaling components to reach their place of destination. During directional migration, cells maintain a polarized state, which means that cell front and back have to be co-ordinated in a robust way. This complex process is not yet fully understood and gaining mechanistic information in model cell migration systems might help to explain regulation of cell migration in vivo.
An important question is how local phenomena such as fine leading edge dynamics, Rho-GTPase signaling or cytoskeletal organization impact on establishment and maintenance of cell polarization to co-ordinate front and back activities during directional cell migration. Current models of global front and back co-ordination mostly originate from highly polarized, rapidly moving cells such as neutrophils or Dictiostelium[OP1]. However, due to the small size of these cells, it is not possible to study local phenomena such as fine leading edge dynamics, happening within time scales of seconds and spatial scales of single micrometers. In order to address questions concerning these fine dynamics, fibroblasts are a widely used model system due to their large and flat morphology. The limitation of this system is that fibroblasts are not highly polarized, precluding the study of front/back co-ordination. An integrated, multi-scale view of directional cell migration view is therefore missing.
To overcome this limitation we engineered a system, which enabled us to study polarized and persistent cell migration on multiple time and length scales. In this experimental set up we allowed rat embryonic fibroblasts (REF52) to migrate on fibronectin coated line substrates. Unstimulated cells (referred to as hapto cells) or platelet derived growth factor (PDGF) treated cells (referred to as chemo cells) were studied in a variety of static or live cell imaging experiments using different spatio-temporal resolution. Hapto cells were found to undergo transient episodes of polarization and therefore characterized by low migration persistence as well as low migration velocity. In contrast, chemo cells showed a drastic increase of migration persistence, enabling them to migrate in one specific direction for hours. At the same time migration velocity was elevated by more than five times. This provides an excellent model system to study polarized cell migration.
In order to explain these drastic changes in migration persistence and velocity as global migration parameters, we examined cytoskeletal, adhesion and signaling dynamics at high spatio-temporal resolution. We found that hapto cells displayed classic features previously observed during mesenchymal cell migration. A protrusive lamellipodium led to the formation of initial adhesions, called focal complexes. Directly behind the lamellipodium, these adhesions then matured in mechanosensitive adhesions, called focal adhesions, through interaction with the contractile lamella. Front adhesions connected to back focal adhesions through stress fibers. Thus, as previously proposed, this front/back linkage coupled with stress fiber tension allows the front to pull the back, leading to tail retraction.
Surprisingly, we observed different actin and adhesion dynamics in persistently migrating chemo cells. These cells remodeled their cytoskeleton and developed two distinct front and back functional modules which are mechanically uncoupled. Specifically, a non-contractile front module, containing a constant sized zone of podosome-like structures (PLSs) replaced the lamella and precluded contractile, retrograde F-actin flow at this subcellular localization. This allowed the PLS zone to function as mechanical insulator, leading to loss of maturation of focal complexes to focal adhesions. Tail retraction was then mediated by a 2nd contractile module that consists of a myosin cluster positioned directly at the back of the PLS zone. Thus, a front module pushes in direction of cell migration, and a contractile back module directly follows the PLS zone and pulls the cell back allowing for tail retraction.
By evaluating front/back motion co-ordination, and using drug perturbations, we formally showed, that the protrusive front is mechanically uncoupled from the contractile back module. From a signaling point of view, we found that the PLS zone acts by locally inhibiting RhoA mediated contractility at the leading edge, allowing uncoupling of cell front and back.
We propose, that mechanical uncoupling of cell front and back by establishment of the PLS zone enables highly efficient and persistent fibroblast cell migration during exposure to a uniform concentration growth factor. The finding that cells do not necessarily require a chemokine gradient to migrate uni-directionally, but can polarize efficiently by simple exposure of an uniform concentration of growth factor in combination with topological confinement of the ECM, might have significance in vivo too. For example, neural crest cells can migrate directionally during collective cell migration in absence of a gradient. During cancer metastasis, a macrophage-tumor cell paracrine loop allows for collective cell streaming in one specific direction on collagen fibrils. While this was suggested to involve chemotaxis, it is conceivable that chemokinesis might therefore be sufficient to induce directional cell migration on the collagen fibril.
Our finding of mechanical uncoupling of cell front and back during chemokinesis on line substrates might provide a mean for generation of directional cell migration.
[OP1]Took away time and length scales