A revised model of nuclear actin import: Importin 9 competes with cofilin, profilin, and RanGTP for actin binding

While predominantly studied in cytoplasmic contexts, actin plays critical roles in the nucleus, regulating genome accessibility, transcription, and DNA repair. Cell-based studies have contributed to a widely accepted model in which the import factor importin 9 (IPO9) acts in concert with the actin filament-severing protein cofilin to transport actin into the nucleus. The classical nuclear localization signal on cofilin is thought to anchor IPO9 to cofilin-bound actin monomers, driving the formation of an import-competent tripartite actin-cofilin-IPO9 complex. Contrary to this established model of actin import, we demonstrate that IPO9 directly binds to monomeric actin with midnanomolar affinity and, rather than promoting IPO9-actin complex formation, cofilin competitively inhibits the binding of IPO9 to actin. To define the mechanism of IPO9 binding, we subject monomeric actin to competitive IPO9 binding in the presence of well-established actin-binding molecules and find proteins that engage either the barbed face, profilin, or pointed face, DNase I, competitively limit IPO9-actin complex formation, whereas sterically less demanding binding partners, thymosin beta-4 and latrunculin B, do not. Consistent with these findings, we demonstrate that IPO9 modestly decreases the rate of actin filament assembly, a process that requires both actin faces, and that IPO9 exhibits minimal binding to actin filaments. Finally, we identify unexpected affinity between the nuclear import release factor RanGTP and monomeric actin; however, a tripartite IPO9-actin-RanGTP complex does not form. The competitive interactions observed between IPO9 and cytoplasmic actin-binding proteins suggest that dynamically coupled equilibria mediate the nuclear transport of actin monomers.