Actin Cytoskeleton: A Nucleator Face-Off

Abstract

Actin assembly proteins initiate the formation of diverse cytoskeletal structures in a single cell. A new study shows that assembly factors compete for actin monomers, leading to homeostasis between different actin networks.

Competition fuels a dizzying array of processes in biology and life. It can be observed in organisms that battle for an ecosystem’s meager food supply, or when scientific researchers compete to publish in the pages of their favorite journal. Such battles for a limiting resource define exploitation competition, most commonly considered at the level of ecological communities. Do similar interactions also apply at the level of a single cell, where proteins might compete for a limiting substrate? In principle, the size and shape of subcellular structures might be confined by competition for a limiting factor, or alternatively by physicomechanical restraints or other mechanisms [1,2]. New research published in this issue of Current Biology by Burke et al. [3] reveals a striking homeostasis between distinct actin-based cytoskeletal networks in fission yeast cells. At the heart of this homeostatic relationship is competition for a limiting supply of actin, which forms these dynamic structures. From this new work, actin cytoskeletal networks join a growing list of organelles that are constrained by competition for a limited supply of building blocks.

Virtually every eukaryotic cell type contains multiple actin-based cytoskeletal structures dedicated to specific functions [4], but the quantitative relationship between these different structures has remained largely unclear. Actin itself is a globular monomeric protein, and these individual subunits can self-assemble into filamentous polymers. In cells, formation of actin filaments requires actin assembly factors, most notably formin proteins and the multi-subunit Arp2/3 complex. Formins and Arp2/3 complex are generally thought to assemble distinct, non-overlapping actin structures in cells. Such is the case in fission yeast cells, where Arp2/3 complex generates endocytic actin patches, while formins (named For3 and Cdc12 in this case) assemble polarized actin cables and the cytokinetic actin ring [5]. Both formins and Arp2/3 utilize actin monomers to form their respective structures, and numerous studies have investigated their regulated biochemical mechanisms [6]; however, the connection between these distinct actin structures has remained unclear. Burke et al. [3] demonstrate that inhibition of one assembly factor leads to enhanced activity by the other in cells. These and other experiments reveal that actin assembly factors compete for a limiting pool of actin monomers in cells, leading to a homeostatic relationship between distinct actin networks.

Burke et al. [3] employed a series of simple yet innovative techniques to identify and characterize this competition between actin structures. Cells treated with the Arp2/3 inhibitor CK-666 rapidly lost actin patches, consistent with the requirement for Arp2/3 in generating these dynamic structures [7,8]. Remarkably, loss of actin patches led to the rapid assembly of excess actin cables and rings. These ectopic structures required formins for their assembly and contained higher concentrations of actin filaments than endogenous formin-generated structures. Thus, formation of actin patches by Arp2/3 complex limits the assembly and size of actin structures by formin proteins.

As with any good competition, this phenomenon is not a one-way relationship. Previous work had shown that actin cables and rings disappear in the absence of formin proteins [9–11]. Burke et al. [3] found that loss of formin proteins also led to increased assembly of Arp2/3-dependent actin patches. Importantly, they showed that the overall number of actin patches increased, but the amount of actin per individual patch was unchanged. This means that formin-mediated actin assembly limits the initiation of new actin patches by Arp2/3 complex, but additional factors must limit the absolute size of each patch. In combination, these experiments reveal a previously unknown homeostasis between actin networks in cells (Figure 1). Though demonstrated in yeast cells, this principle likely applies to the balance of actin networks in a wide range of cell types in other organisms.

Figure 1. Homeostasis between different actin networks is fueled by competition.

A recent study [3] shows that formins and Arp2/3 complex compete for a limiting pool of actin monomers, leading to homeostatic levels of formin-generated actin cables and Arp2/3- assembled actin patches.

Homeostasis between actin networks could result from a simple competition for subunits between actin assembly factors. This model makes the simple prediction that the amount of both actin networks should scale with actin concentration. More actin would relieve the competition and build more of each network; less actin would make the limiting factor even scarcer. The answer turns out to be both complicated and fascinating. Increased actin levels drove the assembly of excess actin patches as predicted, but did not appear to have the same effect on cables [3]. Decreased actin levels inhibited actin patch assembly, but generated more cells with formin-assembled actin rings. This suggests that formin wins the competition when actin becomes scarce, although it remains possible that the increased number of cells with rings reflects slow dynamics of cytokinesis. Overall, these alterations to actin subunit concentration argue against a simple, linear competition between actin assembly factors. It seems likely that the many proteins studied as regulators of Arp2/3 and/or formins may also modulate the cellular competition between these assembly factors. In this sense, each nucleator likely serves as the captain of a larger team that works together in the hunt for cellular actin monomers. Of particular interest for team formin will be the actin-monomer binding protein profilin, which activates formins and shows genetic interactions consistent with a mediator of formin-versus-Arp2/3 competition [6,7,12,13].

In addition to uncovering homeostasis between actin networks, this work has revealed interesting differences between actin networks that coexist in the same cell. For example, excess formin-mediated actin assembly builds larger actin cables that contain higher actin concentration than normal. Thus, the size of a single cable is limited by the amount of actin available to formins. In contrast, increased actin assembly by Arp2/3 does not affect the size or dynamics of individual actin patches. Rather, a higher number of largely uniform actin patches appear in these cells. The uniform size of actin patches might reflect physical limits in patch size, or alternatively might mean that another component of patches serves as the limiting component to their size. This also suggests that the availability of actin to Arp2/3 limits the formation of new actin patches in cells. It will be interesting to see whether these network properties apply to actin networks assembled by formins versus Arp2/3 complex in other cell types, or alternatively whether these properties have been tailored to the specific needs of a yeast cell.

The exploitation competition that Burke et al. [3] have discovered in fission yeast likely applies to actin networks in diverse cell types. For example, the balance between exploratory filopodia and progressing lamellipodia might represent a similar formin-versus-Arp2/3 competition. In general, these findings can be extrapolated to other biological systems that are assembled by a limited supply of building blocks. The most obvious extensions are to other cytoskeletal networks formed by microtubules or intermediate filaments, but similar concepts might also determine the homeostatic size control of different membranous organelles that are assembled by limited lipid precursors. In this sense, future work on actin network homeostasis has the potential to reveal how additional players mediate competition between vast arrays of biological networks. These homeostatic control systems likely operate in a dynamic fashion as cells adjust to changes in their size, shape, and environment.