Actin filaments lie in a wide range of orientations in the leading edge of a migrating cell, according to images from Stefan Koestler, Victor Small (Austrian Academy of Sciences, Vienna, Austria), and colleagues. What looks like disorder might help organize both protrusion at the front and retraction behind.

The work casts doubt on the textbook view of the network in the very front of a migrating cell, known as the lamellipod. In this view, the actin filaments branch off one another at consistent 70° angles. Small's group has been questioning this model for years, worried that fixation techniques might introduce branch-like artifacts by causing the filaments to collapse.

In the new work, the authors tried to avoid these problems by using a stronger fixative and drying the samples directly in a negative stain. Live cells were first viewed using light microscopy to determine whether lamellipodia were protruding, pausing, or retracting. They were then immediately fixed and examined by EM.

Protruding lamellipodia, the images revealed, contained actin filaments that hit the plasma membrane at a wide range of angles, varying from 15° to 90°. During pauses, more filaments aligned more nearly parallel to the membrane. Small thinks the rearrangement comes naturally as the cell front slows its forward movement. “As cells slow, some actin filaments stop growing, but others are still polymerizing. Those filaments have to change orientation,” he says, since their growing plus ends are tracking along the membrane.

The change might create an organization of filaments that maintains the retracting cell edge by lying parallel to it. It might also help construct the cytoskeleton just behind the lamellipod, in the lamella, where contractile bundles are built from myosin and antiparallel actin arrays. In an upcoming JCB paper, Small and colleagues show that actin bundles from filopodia also contribute to contractile bundles in the lamella.

“What we're saying,” Small explains, “is that lamellipodia and filopodia are filament factories not only for protrusion but also for constructing the cytoskeleton behind. One way to get antiparallel filaments in the lamella is by reorienting those at the front.”

Although Small hopes to change how people envision the lamellipodial actin array, he knows more work is yet to be done. “We haven't disproven branching yet,” he says. “We'll need 3D imaging to put the nail in that coffin. But what we've seen makes the branching model unlikely.”


Koestler, S.A., et al.
Nat. Cell Biol.