Actin's Other Life

How cell movement and migration are managed by internal actin structures.

We are familiar with actin in muscles, where actin filaments occur in almost crystalline arrays, interdigitated by structural proteins, myosin cables, and nerve / ionic regulators, causing the macroscopic flexing we rely on to get around. But how to individual cells get around? They use actin too, but very differently.

Most cells have actin and myosin inside, but they have many roles, forming a "cytoskeleton" that helps shape and move the cell, but also forming avenues of transport, where myosin attaches to actin on its motor end, and to other things like organelles or vesicles of proteins to be secreted on its other end. For cell movement, actin extends at the leading part of cells through its own polymerization, not by being pushed or pulled by myosin. Then, once the leading edge finds a place outside it that likes and adheres to, myosin acts on the actin network behind to drag the rest of the cell along.

A neuronal growth cone, with actin stained in red, microtubules in green. Actin polymerization is the motive force at the leading edge where cells figure out where they want to go.

Thus much of the story of cell movement, especially at the leading edge, as in neurons trying to find their way around to the right destination in the developing brain, comes down to the selective polymerization of actin, which is, naturally, a carefully regulated process. A recent paper discussed the role of actin-binding proteins in this regulation. It has long been known that cells have some proteins that encourage actin polymerization (profilin, Arp2, thymosin beta 4, formins), and others that inhibit polymerization, or even cleave existing filaments (CAPz, cofilin, gelsolin, severin), or medate crosslinking and branching (fascin, filamin). These proteins are all generally regulated by phosphorylation, so they can be quickly and reversibly controlled by the various signaling cascades that receive signals from the local cell surface and result in a variety of protein kinase (phosphorylation) activities. Thus the main research question is filling in the details of how actin management at the leading edge of cells is orchestrated. And this has been quite difficult, since the scales are very small in time and space, and the tuning quite subtle. Delete a gene, and it may have paradoxical effects, this being such a crude manipulation.

The current authors decided to look into a relatively simple question- how is actin polymerization controlled by profilin, its most typical binding partner in the monomeric state, by its own ATP hydrolysis, and by formins, which are the main actors in speeding up actin polymerization? Actin all by itself polymerizes quite enthusiastically. Thus the cell keeps non-filamentous actin in dimers with various controlling partners such as profilin and thymosin beta 4, to keep a lid on excess polymerization. Profilin is the major partner, and binds actin very tightly. It strictly restricts actin addition to + ends of filamentous actin, not to the - ends, and not other actin monomer commplexes. It also impedes polymerization to a slight degree, even on + ends of actin filaments, due to its sticking a small finger into the interface where the next actin molecule would bind.

Structures of actin (teal and green) and profilin (pink) at the + end of a polymerizing filament (bottom). Profilin sticks a small structure into the cleft where the next actin monomer would add (zoom), slowing polymerization slightly.

The authors create a set of mutant profilin proteins, with altered binding strength to actin. The only useful ones were very subtle alterations. Binding too well inhibited actin polymerization completely, while poor binding rendered profilin entirely useless. But changes of about 5-fold in binding strength were telling in their effect.

The growth of actin filaments is strongly affected by the binding strength of profilin to actin. Profilin mutants that are weaker binders (pink) accelerate polymerization, while a stronger binder (purple) slows it down drastically.

The next step was to add in the effect of formin, a protein that binds to the actin-profilin dimer at the growing end of actin filaments in such a way as to encourage profilin to leave, and also encourage the next actin-profilin dimer to add on. Formin also induces a helical shape change in the actin filament that makes it more stable, resistant to the action of, for example, cofilin, which breaks it down. Thus addition of formins had dramatic effects, speeding up actin polymerization by several-fold, depending on which of a variety of formin versions was used. Not only that, but this system of formin + profilin managed actin monomers rendered the polymerization reaction almost entirely insensitive to the concentration of profilin / actin dimers and actin in general, as long as there was enough profilin to soak up all the free actin.

The growth of actin filaments is strongly accelerated by formin, a protein that binds to the profilin-actin complex at the growing ends of actin filaments, dissociating them to allow more units to add. DAAM1 is a weak formin, while mDia2 and mDia1 are progressively stronger formins, which show progressively faster acceleration of filament growth, in these direct microscopic assays. The arrow marks the growing actin filament end, and at bottom, graphs of actin growth, where the two dimensions above are reduced to one to give a time vs elongation graph.

This leads to the basic finding of the paper, which is that, assuming that free actin is immediately taken up into complexes of various kinds, either in new filaments or with profilin and other monomer binding proteins, the cell does not regulate actin growth by making more of it or altering the bulk amount of the actin:profilin dimers. There certainly isn't the time. Cells use the various specialized accessory proteins to orchestrate actin activities, and then regulate those in turn by the signal-driven phosphorylation events, which constitute an enormous field we won't get into here.

When formins and profilins are present at physiological concentrations vs actin, polymerization rates are insensitive to actin monomer concentrations, and instead depend on the identity of the formin, and presumably its post-translational regulation.

Yet one oddity remains- that actin is itself an ATP-ase. What is the role of this ornate property? The authors demonstrate quite conclusively that disabling the ATP-ase of actin does not alter the assembly and growth characteristics they are studying. Profilin still binds, heterodimers still polymerize, and formin still accelerates that polymerization. This is contrary to at least some old models, which rely on profilin being known to be an exchange factor for actin monomers, promoting the exchange of ADP for fresh ATP. Which was then thought to be essential for actin polymerization. Not at all - ATP does something else as part of actin, which seems to be to create a two-state system where actin with ATP is competent to for some functions, such as profilin binding, and stabilizing existing filaments, while actin with ADP is competent for other things, like release of actin from the other (-) ends of filaments. But the actual hydrolysis is immaterial to actin polymerization. All this was, sadly, already known over a decade ago, so it is not entirely clear why these observations were made in this article. At any rate, it is good that groups are still working on actin and applying ever more modern methods to quantitative studies of its function.

"Importantly, we found that ATPase-deficient actin was able to elongate actin filaments with nearly the same rates as wildtype actin at saturating profilin-actin concentrations. This clearly demonstrates that profilin release from the barbed end does not require cleavage of the β-γ phosphodiester bond of ATP in actin. More generally, the lack of assembly-related defects for ATPase deficient actin is consistent with the notion that ATP hydrolysis serves an essential function unrelated to filament assembly."

• When Anthony Fauci is muzzled, we have a problem.
• If we didn't feed them, we wouldn't have rats.
• Yes, we can make the drugs we need.
• Total abdication of judicial and constitutional responsibility in the McGann decision.
• What happens when Texas runs an election.
• Maybe Biden should keep a low profile.