Mechanistic Principles Underlying Regulation of the Actin Cytoskeleton

Mechanistic Principles Underlying Regulation of the Actin Cytoskeleton

Mechanistic principles underlying regulation of the PNAS PLUS actin cytoskeleton by phosphoinositides Yosuke Senjua, Maria Kalimerib, Essi V. Koskelaa,1, Pentti Somerharjuc, Hongxia Zhaoa, Ilpo Vattulainenb,d, and Pekka Lappalainena,2 aInstitute of Biotechnology, University of Helsinki, FI-00014 Helsinki, Finland; bDepartment of Physics, Tampere University of Technology, FI-33101 Tampere, Finland; cFaculty of Medicine, University of Helsinki, FI-00014 Helsinki, Finland; and dDepartment of Physics, University of Helsinki, FI-00014 Helsinki, Finland Edited by Kai Simons, Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany, and approved September 14, 2017 (received for review March 29, 2017) The actin cytoskeleton powers membrane deformation during many (ADF)/cofilins, gelsolin, and twinfilin] or prevent the assembly of cellular processes, such as migration, morphogenesis, and endocyto- new actin monomers into filament ends (e.g., heterodimeric cap- – sis. Membrane phosphoinositides, especially phosphatidylinositol ping protein) (7 10). Conversely, PI(4,5)P2 activates proteins that 4,5-bisphosphate [PI(4,5)P2], regulate the activities of many actin- promote actin filament assembly (e.g., N-WASP) or function as binding proteins (ABPs), including profilin, cofilin, Dia2, N-WASP, linkers between actin filaments and the plasma membrane (e.g., – ezrin, and moesin, but the underlying molecular mechanisms have ezrin, moesin, radixin, and talin) (11 15). As a result, an increase remained elusive. Moreover, because of a lack of available method- in the plasma membrane PI(4,5)P2 induces actin filament assem- ology, the dynamics of membrane interactions have not been exper- bly beneath the membrane, whereas decreasing the levels or imentally determined for any ABP. Here, we applied a combination availability of PI(4,5)P2 at the plasma membrane diminishes actin filament assembly and enhances filament disassembly (16–19). of biochemical assays, photobleaching/activation approaches, and Cell biological studies have also provided evidence that membrane atomistic molecular dynamics simulations to uncover the molecular interactions are critical for the proper in vivo functions of many principles by which ABPs interact with phosphoinositide-rich mem- ABPs, including cofilin, vinculin, formins, and N-WASP (20–25). branes. We show that, despite using different domains for lipid bind- Despite the central biological roles of ABPs, the molecular ing, these proteins associate with membranes through similar mechanisms by which they interact with cellular membranes remain multivalent electrostatic interactions, without specific binding pock- largely unknown. So far, atomistic details, as derived by a combi- ets or penetration into the lipid bilayer. Strikingly, our experiments nation of mutagenesis and molecular dynamics simulation ap- reveal that these proteins display enormous differences in the dy- proaches, have been reported only for membrane interactions of namics of membrane interactions and in the ranges of phosphoino- the heterodimeric capping protein (26, 27), whereas, for other sitide densities that they sense. Profilin and cofilin display transient, central ABPs, the principles of membrane interactions have low-affinity interactions with phosphoinositide-rich membranes, remained elusive. Moreover, whether distinct ABPs interact with whereas F-actin assembly factors Dia2 and N-WASP reside on membranes through similar or different affinities has not been phosphoinositide-rich membranes for longer periods to perform their reported. Most importantly, the dynamics of membrane interactions functions. Ezrin and moesin, which link the actin cytoskeleton to the have not been experimentally determined for any ABP. The kinetics plasma membrane, bind membranes with very high affinity and slow of membrane interactions have fundamental consequences for dissociation dynamics. Unlike profilin, cofilin, Dia2, and N-WASP, they the cellular functions of proteins. This is because the membrane do not require high “stimulus-responsive” phosphoinositide density for membrane binding. Moreover, ezrin can limit the lateral diffusion Significance of PI(4,5)P2 along the lipid bilayer. Together, these findings demon- strate that membrane-interaction mechanisms of ABPs evolved to Membrane phosphoinositides have emerged as key regulators of precisely fulfill their specific functions in cytoskeletal dynamics. the actin cytoskeleton in cell migration, morphogenesis, cytokine- BIOCHEMISTRY sis, and endocytosis. However, the molecular mechanisms by which actin cytoskeleton | phosphoinositides | protein–lipid interactions | actin-binding proteins (ABPs) interact with phosphoinositide-rich signal transduction | molecular dynamics simulations membranes remain remarkably poorly understood. By applying a combination of biochemical, biophysical, and atomistic molecular olymerization of actin filament networks against membranes dynamics simulation approaches on six central ABPs, we discovered Pprovides forces for many vital cellular processes, including that they employ multivalent electrostatic interactions for mem- generation of plasma membrane protrusions in cell migration and brane binding. Strikingly, our experiments revealed that these morphogenesis, as well as the formation of plasma membrane in- proteins display enormous differences in their membrane in- vaginations in endocytosis (1, 2). The dynamics and the 3D orga- teraction dynamics and in the ranges of phosphoinositide densities nization of actin filament arrays in these processes are precisely that they sense. These differences precisely correlate with the controlled by a large array of actin-binding proteins (ABPs), whose specific functions of these proteins in cytoskeletal dynamics. These findings uncover molecular principles by which membrane phos- activities are in turn regulated by various signaling pathways (3–5). phoinositides regulate dynamics and architecture of the actin In addition to kinase/phosphatase cascades, which can activate or cytoskeleton in cells. inhibit central actin-regulatory proteins, membrane phospholipids, especially phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2]and Author contributions: Y.S., M.K., H.Z., I.V., and P.L. designed research; Y.S., M.K., E.V.K., and P.S. phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P3], have emerged performed research; Y.S. and M.K. analyzed data; and Y.S., M.K., I.V., and P.L. wrote the paper. as important regulators of actin dynamics. PI(3,4,5)P3 contributes to The authors declare no conflict of interest. actin dynamics mainly by regulating the Rho family small GTPases, This article is a PNAS Direct Submission. whereas PI(4,5)P2 controls cytoskeletal dynamics more directly by Published under the PNAS license. regulating ABPs (6). 1Present address: Department of Bioproducts and Biosystems, Aalto University, 00076 PI(4,5)P2, which is the most abundant phosphorylated de- Aalto, Finland. rivative of phosphatidylinositol at the plasma membrane, interacts 2To whom correspondence should be addressed. Email: [email protected]. with multiple ABPs. PI(4,5)P2 typically inhibits proteins that cat- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. alyze actin filament disassembly [e.g., actin-depolymerizing factor 1073/pnas.1705032114/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1705032114 PNAS | Published online October 9, 2017 | E8977–E8986 Downloaded by guest on September 25, 2021 association and dissociation rates, as well as the lateral mobility of N-WASP and cofilin-1 were previously shown to function as proteins along the membrane plane, determine their subcellular phosphoinositide density sensors (36, 38). Therefore, we examined localization, their accessibility to other interaction partners, and the whether a local increase in the PI(4,5)P2 density on the membrane effects of “stimulus-responsive” PI(4,5)P2 synthesis/hydrolysis on affects the membrane binding of other ABPs as well. Cosedimenta- their functions. For example, cell biological studies suggested that tion assays revealed that all six proteins exhibited an increase in cofilin stably binds to the plasma membrane in carcinoma cells and membrane binding with respect to PI(4,5)P2 concentration, suggest- is released only following epidermal growth factor-induced PI(4,5) ing that they can sense the local density of PI(4,5)P2 on the mem- P2 hydrolysis (21, 28, 29). Moreover, some animal, plant, and slime brane (Fig. S3). The binding curves displayed sigmoidal functions, at mold formins, which promote actin filament nucleation and elon- least for profilin-1, cofilin-1, Dia2, and N-WASP, and Hill coeffi- gation, were proposed to require the N-terminal phosphoinositide- cients (nH) were greater than 1 [profilin-1 nH = 3.7; cofilin-1 nH = binding region for proper anchoring to the plasma membrane (20, 4.2; Dia2 basic domain (BD) nH = 3.0; N-WASP polybasic motif and 30–33). However, such functions would require stable, high-affinity GTPase binding domain (B-GBD) nH = 2.3], indicating cooperative binding of these proteins to the PI(4,5)P2-rich membranes, which membrane binding with respect to the PI(4,5)P2 density in the has not been experimentally demonstrated. membranes. Importantly, whereas binding of the FERM domains of ∼ ∼ Here, we applied a combination of biochemical, biophysical, ezrin and moesin saturated at 2 mol%, PI(4,5)P2, 10 mol% PI and atomistic molecular dynamics simulation approaches to reveal (4,5)P2

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    10 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us