Distinct functional constraints driving conservation of the cofilin N-terminal regulatory tail

Sexton, Joel A.; Potchernikov, Tony; Bibeau, Jeffrey P.; Casanova-Sepúlveda, Gabriela; Cao, Wenxiang; Lou, Hua Jane; Boggon, Titus J.; De La Cruz, Enrique M.; Turk, Benjamin E.

doi:10.1038/s41467-024-45878-9

Download PDF

Article
Open access
Published: 16 February 2024

Distinct functional constraints driving conservation of the cofilin N-terminal regulatory tail

Nature Communications volume 15, Article number: 1426 (2024) Cite this article

581 Accesses
3 Altmetric
Metrics details

Subjects

Abstract

Cofilin family proteins have essential roles in remodeling the cytoskeleton through filamentous actin depolymerization and severing. The short, unstructured N-terminal region of cofilin is critical for actin binding and harbors the major site of inhibitory phosphorylation. Atypically for a disordered sequence, the N-terminal region is highly conserved, but specific aspects driving this conservation are unclear. Here, we screen a library of 16,000 human cofilin N-terminal sequence variants for their capacity to support growth in S. cerevisiae in the presence or absence of the upstream regulator LIM kinase. Results from the screen and biochemical analysis of individual variants reveal distinct sequence requirements for actin binding and regulation by LIM kinase. LIM kinase recognition only partly explains sequence constraints on phosphoregulation, which are instead driven to a large extent by the capacity for phosphorylation to inactivate cofilin. We find loose sequence requirements for actin binding and phosphoinhibition, but collectively they restrict the N-terminus to sequences found in natural cofilins. Our results illustrate how a phosphorylation site can balance potentially competing sequence requirements for function and regulation.

Emergence of fractal geometries in the evolution of a metabolic enzyme

Article Open access 10 April 2024

Pooled multicolour tagging for visualizing subcellular protein dynamics

Article Open access 19 April 2024

Spatially organized cellular communities form the developing human heart

Article Open access 13 March 2024

Introduction

Phosphorylation is perhaps the most common reversible post-translational protein modification and has widespread roles in regulating cell behavior. Sites of phosphorylation evolve rapidly, in part because kinase recognition motifs can be introduced by a small number of mutations in flanking residues^1,2. However, regulatory sites of phosphorylation are necessarily under additional sequence constraints beyond those required for kinase recognition. For example, intracellular environments harbor many protein kinases—up to ~500 for human cells³. Accordingly, just as the sequence surrounding the site must be tuned to its cognate kinase, there is also selective pressure to evade phosphorylation by other kinases^4,5,6. Furthermore, regulatory phosphorylation often occurs in functionally important regions, so that variability in the local sequence may be limited to maintain basic function^7,8. Finally, the specific sequence context of a phosphorylation site may dictate how phosphorylation modifies protein behavior, for example, by generating a binding site for a phosphopeptide interaction module⁹. These factors are potentially in competition with one another, but there is currently limited insight into how they are balanced in any given phosphorylation site.

One example of a phosphoregulatory event occurring at a highly conserved site involves proteins in the cofilin/ADF (actin depolymerizing factor) family (hereafter referred to collectively as cofilin). Cofilin is a small actin-binding protein important for dynamic regulation of the actin cytoskeleton in eukaryotes and essential for multiple actin-dependent processes, including cytokinesis, cell polarity, migration, neurite extension, and mechanosensing^10,11,12. Cofilin binds cooperatively to actin filaments, promoting severing at junctions between bare and decorated segments, as well as depolymerization, to produce new filament ends and a pool of actin monomers for reassembly^{13,14,15,16,17,18,19,20,21,22}.

Due to its major impact on the actin cytoskeleton, cofilin activity is tightly regulated by several mechanisms. Principal among these is phosphorylation at Ser3 within its disordered N-terminal region, catalyzed in most animals and some protists by members of the LIM kinase family (LIMK1/2 and TESK1/2)²³. The N-terminal tail is an essential part of the actin-binding interface, and Ser3 phosphorylation inactivates cofilin by dramatically decreasing its affinity for actin²⁴. The phosphomimetic cofilin^S3D mutant binds to filamentous actin weakly but with higher cooperativity than the wild-type (WT) protein, and it severs filaments poorly, even when bound^24,25,26.

The N-terminus of animal cofilins contains a conserved acetyl-Ala2-Ser3-Gly4-Val5 sequence (by convention, we use residue numbering based on the unprocessed precursor)¹¹. Orthologs from other kingdoms vary in the sequence upstream of Ser3 and may have other β-branched or aliphatic residues in place of Val5, but the phosphoacceptor residue and Gly4 are invariant. Factors driving the conservation of this sequence are currently unclear. Aside from full N-terminal deletion mutants that are non-functional, only the phosphomimetic S3D and phosphorylation-resistant S3A variants have been examined with respect to their effects on actin in vitro or in cells^{24,25,26,27,28}.

Recognition of cofilin by LIMK occurs through a docking interaction between the cofilin globular domain and a site adjacent to the kinase catalytic cleft on the kinase C-terminal lobe^29,30. Consequently, the N-terminal region projects into the LIMK active site in a non-canonical orientation that lacks complementary binding interactions typically observed in other protein kinase-substrate complexes. Limited mutational analyses have suggested loose sequence requirements surrounding the phosphorylation site, save for an unusually strict exclusion of Thr as a phosphoacceptor residue³⁰. Attempts to identify a LIMK phosphorylation site sequence motif using peptide libraries have been unsuccessful, presumably because the docking interaction is essential for efficient phosphorylation^29,31.

A platform for the evaluation of large numbers of N-terminal sequence variants would facilitate the discovery of those features important for the various aspects of cofilin functionality. Here we exploit the widespread requirement for cofilin to establish a budding yeast system for high-throughput evaluation of sequence variants. S. cerevisiae lacking their sole cofilin ortholog, Cof1, are inviable^32,33, but ectopic overexpression of mammalian cofilin-1 can rescue growth^24,29,34. In this context, we have reported that inducible expression of human LIMK1 suppresses growth in a manner dependent on Ser3 phosphorylation²⁹. With this system, we can evaluate in parallel the impact of a mutation on either core cofilin function or its regulation by LIMK based on its ability to support growth in the absence or presence of LIMK induction, respectively. Here, we have applied the system to screen a comprehensive collection of N-terminal sequence variants and subsequently conducted biochemical analysis of a panel of mutants. We found unexpectedly loose sequence requirements for each of the various aspects of cofilin function and regulation, with each dominated by its own distinct position within the N-terminal sequence. Overall conservation of the N-terminal region thus appears to be constrained by the collective requirements imposed by the full complement of cofilin functionality. These observations may have general implications for how regulatory phosphorylation sites emerge and are maintained through evolution.

Results

A dual-functionality cofilin library yeast competitive growth screen

To define cofilin N-terminal sequence requirements for actin severing and phosphoregulation, we leveraged the capacity of mammalian cofilin to substitute for yeast Cof1 in supporting growth²⁴. We used a yeast strain in which the endogenous COF1 gene was placed under the control of a tetracycline-repressible promoter (TeTO₇-COF1)³⁵. This strain grew poorly in the presence of doxycycline (DOX) due to loss of endogenous Cof1 expression, but ectopic expression of either yeast Cof1 or human cofilin-1 rescued growth (Fig. 1a). Our previous work with a temperature-sensitive cof1 mutant strain demonstrated that ectopic expression of the LIMK1 catalytic domain (LIMK1^CAT) completely blocked growth in the context of human cofilin-1²⁹. Likewise, we found that inducible expression of LIMK1^CAT inhibited the growth of TeTO₇-COF1 yeast expressing human cofilin-1 in a manner dependent on its phosphorylation at Ser3 (Supplementary Fig. 1). Expression of full-length (FL) LIMK1, which has lower cofilin kinase activity than LIMK1^CAT, caused only partial growth suppression (Fig. 1b). We chose to use FL LIMK1 for our screens as lower activity should allow us to better distinguish cofilin mutants that impact LIMK regulation to varying degrees. We next constructed a saturation mutagenesis library of the human cofilin-1 N-terminal region in which Ala2, Gly4, and Val5 were combinatorially randomized as all twenty potential canonical amino acids (Fig. 1c). Ser3 was maintained as either a serine or threonine residue to preserve a potential phosphorylation site, producing a library with 16,000 (20³ × 2) unique sequences. The library was introduced into a constitutive yeast expression plasmid with an N-terminal His₆ tag included to prevent potential sequence-dependent post-translational modifications (e.g., initiator methionine removal and acetylation) and to help normalize levels of expression. The His₆ tag did not prevent cofilin from rescuing yeast growth, severing actin, or being phosphorylated by LIMK1 (Fig. 1a, Supplementary Fig. 2). TeTO₇-COF1 yeast harboring the inducible LIMK1 expression plasmid were transformed with the cofilin library and expanded in liquid media with a portion of the culture reserved as a starting point sample. The remaining cells were grown in raffinose to derepress the LIMK1 expression construct and treated with DOX to repress Cof1 expression. The culture was then split and propagated in either glucose or galactose to repress or induce LIMK1 expression, respectively (Fig. 1d). At various times, the cultures were sampled for plasmid DNA extraction. The variable library sequence within the recovered plasmids was PCR-amplified and subjected to next-generation sequencing. This analysis indicated how the relative abundance of each component of the library changed over time (Fig. 1e, Supplementary Data 1 and 2). In the context of competitive growth, DOX + glucose conditions should enrich for viable cells expressing functional library variants. Growth in DOX + galactose should deplete the culture of functional cofilin variants that could be phosphorylated and inhibited by LIMK1. We first analyzed our data from the DOX + glucose condition to identify core determinants of cofilin function.

**Fig. 1: A dual-functionality screen of a cofilin-1 N-terminal mutagenesis library.**

Functional constraints on the cofilin N-terminal sequence in supporting growth

Yeast grown in glucose were analyzed for enrichment of cofilin sequences that could presumably bind and sever actin filaments. We compared sequence representation between the beginning of culture in glucose and ~6 population doublings later (Figs. 1e and 2a), beyond which changes were less pronounced. We ranked each sequence by its average log₂ fold change in the normalized read count across three independently performed replicate screens (Fig. 2a). When analyzed this way, 4099 of the 16,000 library sequences were enriched in the population, suggesting that the majority of N-terminal substitutions were deleterious to cofilin function. As anticipated, native cofilin sequences performed well, including those from yeast (Arg-Ser-Gly-Val, rank 27), human (Ala-Ser-Gly-Val, rank 87), and Arabidopsis (Ala-Ser-Gly-Met, rank 262), though the most enriched sequence was non-native (Phe-Thr-Gly-Leu) (Fig. 2a, Supplementary Data 2). Examination of the 400 most enriched sequences overall identified a Leu-X-Gly-Met consensus motif. (Supplementary Fig. 3a). Within the N-terminal sequence, the fifth position was the most stringently selective, with aliphatic or β-branched amino acids (73% of enriched sequences) being apparently required for function (Fig. 2b–d) and Met being most commonly enriched (17% of enriched sequences). By comparison, other positions within the sequence were less selective. Despite being highly conserved in plants and animals, the second residue (Ala2 in humans) was effectively indiscriminate, with modest enrichment for non-native hydrophobic over charged residues (Fig. 2d). Likewise, enriched sequences had roughly equal representation of Ser (45%) and Thr (55%) at the phosphoacceptor (third) position. Finally, the selection at the fourth residue was intermediate. As found in all WT cofilin sequences, Gly was overrepresented at this position in enriched sequences (8% overall). Despite being absent from all native cofilin sequences, hydrophobic residues were generally selected at this position, with charged residues and most small residues being depleted.

**Fig. 2: Cofilin library screen results for yeast growth rescue.**

Because our combinatorial library includes all possible combinations of residues at random positions, it has the potential to reveal cases where there is functional cooperativity between specific pairs of residues. In such cases, two residues would be co-enriched more frequently than expected if they behaved independently. We found many instances where residues appeared to co-vary in this manner. For example, Leu and Gly residues are represented similarly at position 4 within the 400 most enriched sequences. However, only Gly4 sequences were significantly enriched for a Val residue at position 5 (Supplementary Fig. 3a–c). Likewise, sequences with Val5 more prominently included Gly4 compared to those sequences containing Met5 (Supplementary Fig. 3d, e). The Gly4-Val5 pair, which is found invariably in animals and fungi, was uniquely over-represented within these most highly enriched sequences (Supplementary Fig. 3f) and may provide a fitness advantage over sequences combining independently favorable residues. The residue at position 4 likewise appeared to influence selectivity at position 2 in unexpected ways. For example, in the context of Gly4 but not Leu4, Glu2 was consistently preferred over a similar acidic Asp2 residue (Supplementary Fig. 3g). We furthermore noted distinct sequence preferences in the context of the two phosphoacceptor residues (Supplementary Fig. 4). Enriched Thr3 sequences in which the otherwise identical Ser3 sequence was depleted, for example, had a preponderance of charged residues at position 4 (Supplementary Fig. 4a, b). This phenomenon may reflect the capacity for Thr3 to better promote cofilin function, allowing it to compensate for the presence of these otherwise unfavorable residues. We also observed strong covariation of Thr3 with Thr5, as reflected in a large difference in mean log₂ fold change for Thr3/Thr5 variants compared with Ser3/Thr5 variants (1.7 vs. −0.28) (Supplementary Fig. 4c–e). While the basis for this and other examples of covariation are not immediately obvious, they are likely to promote either specific conformations or dynamics that enhance actin binding affinity or the capacity to sever actin once bound.

To confirm the results from the screen, we selected a number of mutants with varying degrees of enrichment for further analysis (Table 1). We initially assayed these mutants for their ability to support yeast growth on solid media, a format distinct from the competitive growth scheme used for the screen. Human cofilin-1 single or double-point mutants were expressed in TeTO₇-COF1 yeast and grown with or without DOX to control endogenous Cof1 expression. In keeping with their enrichment in the screen, cofilin^V5M supported growth similarly to WT human cofilin (Fig. 3a), while yeast expressing cofilin^V5E failed to grow. In contrast, Ala2 mutations had little effect on yeast growth, with both cofilin^A2L and cofilin^A2D supporting growth in the presence of DOX. Growth of yeast expressing the double mutant cofilin^A2L,V5M, which incorporates residues most highly enriched at each position, was indistinguishable from that of yeast expressing cofilin^WT or cofilin^V5M. Substitutions to Gly4 displayed a range of growth phenotypes that generally correlated with behavior in the screen (Fig. 3b). Yeast expressing cofilin^G4L grew similarly to yeast with cofilin^WT. Cofilin^G4A, which was slightly depleted in the screen, and cofilin^G4F, which was in the bottom half of enriched sequences, each supported growth albeit at slower rates than for cofilin^WT. The more substantially depleted variants, G4P, G4E, and G4K, provided, at best, a marginal amount of growth. Overall, we found that for most variants enriched in the context of competitive growth in liquid culture, apparent differences in fitness did not translate into overt differences in growth rate on solid media. Furthermore, mutants that can support growth on solid media but at lower rates than cofilin^WT were likely depleted in the screen due to being outcompeted by higher fitness variants. We note that the extent of growth rescue did not generally correlate with the level of cofilin expression (Supplementary Fig. 5). However, one of the most damaging mutations, cofilin^V5E, expressed at very low levels suggesting that it and perhaps other non-functional Val5 substitutions may destabilize the protein.

Table 1 Properties of cofilin variants chosen for individual evaluation

Full size table

**Fig. 3: Impact of cofilin N-terminal mutations on yeast growth and actin binding.**

We next examined how the ability of a cofilin variant to support growth correlated with its biochemical behavior in vitro. We expressed and purified cofilin mutants using a system that produces untagged protein with the native N-terminal sequence (initiating with Ala2 or the desired substitution). We initially evaluated the panel of mutants for binding to actin filaments measured in a fluorescence quenching assay and further analyzed a subset of these mutants for their capacity to sever actin filaments in a TIRF microscopy-based assay. Phosphorylated cofilin^WT and phospho-mimetic cofilin^S3D were used as weak binding controls^25,28,36. We used the cofilin concentration producing half-maximal occupancy (K_0.5) as a proxy for relative overall binding affinity³⁷. Cofilin^WT bound to actin filaments with a K_0.5 value of approximately 0.6 µM (Fig. 3c–e, Table 1), whereas cofilin^S3D required ~8-fold higher concentrations to bind actin filaments. As anticipated, phosphorylated cofilin^WT displayed almost no detectable binding at the concentrations tested. Almost all mutants that had supported growth on solid media, including the G4F, A2D, V5M, and the double mutant A2L/V5M, displayed actin binding comparable to cofilin^WT (Fig. 3), and where tested, these mutants likewise severed actin filaments similarly to WT (Supplementary Figs. 6 and 7). Cofilin^G4A, which supported growth at a lower rate than cofilin^WT, showed somewhat weaker actin binding (K_0.5 = 3.2 μM), as did the G4K variant that failed to support growth (K_0.5 = 3.5 μM). Of all variants tested, only cofilin^G4E showed essentially no binding to actin filaments (Fig. 3e), and accordingly, it had no detectable severing activity (Supplementary Fig. 6). Consistent with its low level of expression in yeast, cofilin^V5E formed soluble aggregates when expressed in bacteria, and we were unable to isolate monomeric protein. Overall, we found a general correlation between yeast growth rate and actin binding, with a threshold K_0.5 of ~2 μM being required to support maximal growth.

Identification of cofilin variants inhibited by LIMK1

To identify cofilin variants susceptible or resistant to LIMK1, we examined the relative abundance of functional variants in our screen following induction of the kinase by growth in galactose. We noted that depletion of the cofilin^WT sequence from the culture lagged 1–2 population doublings behind the induction of LIMK1. We, therefore, followed variant representation from this time point onward over the course of 10 population doublings in the presence of either glucose or galactose (Fig. 1e), and we ranked sequences by the log₂ fold change in abundance. We found that approximately 1/3 of functional cofilin sequences (those enriched in glucose) became depleted in the presence of galactose, suggesting that they were phosphorylated and inhibited by LIMK1, with the remaining 2/3 persisting in the presence of galactose and thus presumably resistant to LIMK1 expression (Fig. 4a, Supplementary Fig. 8, and Supplementary Data 2). Functional sequences depleted in galactose overwhelmingly had Ser3 as the phosphoacceptor residue and were dominated by small residues at the Gly4 position, while conversely, the majority of LIMK1-resistant sequences contained Thr3 and large hydrophobic residues (primarily Leu and Phe) in position 4 (Fig. 4b, c and Supplementary Fig. 8).

**Fig. 4: Cofilin library screen results for yeast growth rescue with and without LIMK1.**

We initially examined the impact of phosphorylation site residue substitutions LIMK1-induced phosphorylation and growth inhibition in yeast (Fig. 5a). In contrast to yeast expressing cofilin^WT (with Ser3), those expressing cofilin^S3T were unaffected by LIMK1 induction and grew similarly to yeast harboring cofilin^S3A. Because our cofilin library did not include Tyr3 variants, we also examined the growth of yeast expressing cofilin^S3Y. While cofilin^S3Y only partially rescued growth in the presence of DOX, growth was further inhibited by galactose-induced expression of LIMK1. Matching these observations, both cofilin^WT and cofilin^S3Y became phosphorylated upon LIMK1 induction (Fig. 5b) as judged by mobility shift on Phos-tag SDS-PAGE, while only a trace amount of cofilin^S3T phosphorylation was observed. In similar experiments performed with yeast expressing the hyperactive LIMK1^CAT, cofilin^S3T was still phosphorylated to only a small extent and fully supported growth (Supplementary Fig. 9). Consistent with these results and a recent report³⁰, LIMK1 phosphorylation of cofilin in vitro was blocked by either S3A or S3T mutation and was reduced but not eliminated when Ser3 was substituted with Tyr (Fig. 5c). LIMK2 was similarly unable to phosphorylate cofilin^S3T (Supplementary Fig. 9d), indicating that this strong phosphoacceptor residue preference is conserved across isoforms. Collectively, these results confirm that while LIMKs are dual-specificity kinases in the sense that they can phosphorylate both Ser and Tyr, they are almost entirely devoid of activity on Thr residues.

**Fig. 5: Effects of cofilin Ser3 substitutions on phosphorylation by LIMK1.**

Cofilin/ADF proteins are members of a larger family of actin-binding proteins having an ADF fold. Theoretically, other ADF domain proteins could be regulated by phosphorylation at sites analogous to Ser3. For example, the N-terminal region of twinfilin proteins, which consist of two ADF domains arranged in tandem, has a potential phosphoacceptor residue at position 5, analogous to cofilin Ser3 (Fig. 5d). Notably, this residue is Thr in mammalian twinfilins, but is a Ser in the single ortholog from budding yeast, which lack LIMKs. Furthermore, key features of the LIMK docking region of cofilin centered on helix α5 are conserved to the N-terminal ADF domain of twinfilin (Supplementary Fig. 10)²⁹. We therefore wondered whether the identity of the phosphoacceptor residue might serve to insulate twinfilin from phosphoregulation by LIMKs. We examined if LIMK1 could phosphorylate mouse or budding yeast twinfilin in vitro and whether phosphorylation was influenced by the identity of the putative phosphoacceptor. We found that LIMK1 robustly phosphorylated budding yeast twinfilin at ~40% of its cofilin phosphorylation rate yet had barely detectable activity on mouse twinfilin-1 (Fig. 5e). As anticipated, substituting Ser5 of yeast twinfilin with Thr abolished phosphorylation, confirming Ser5 to be the site of LIMK1 phosphorylation and suggesting flexibility in the identity of that residue in an organism lacking LIMKs. However, LIMK1 did not substantially phosphorylate mouse twinfilin-1^T5S, indicating a Ser phosphoacceptor residue is necessary but not sufficient for LIMK phosphorylation of ADF fold proteins. Collectively these results suggest that insulating twinfilin from inhibition by LIMK1 involves both the phosphoacceptor and additional structural features that prevent phosphorylation.

Identification of phosphoregulatory constraints on the cofilin N-terminal sequence

In our screen, LIMK1-resistant variants were dominated by those having a Thr phosphoacceptor residue. However, Ser3 sequences having Leu, Phe, or Trp at position 4 also persisted despite LIMK1 expression, while Ser3 sequences depleted upon LIMK1 induction were enriched for Gly and other small residues at position 4 (Fig. 4b). We confirmed in growth assays on solid media that cofilin^G4L and cofilin^G4F, but not cofilin^G4A, protected yeast from LIMK1-induced growth inhibition (Fig. 6a, Supplementary Fig. 11). The most straightforward explanation for these results would be if variants with bulky hydrophobic residues at position 4 were poor substrates of LIMK1. However, we found instead that cofilin^G4L and cofilin^G4F became highly phosphorylated upon expression of LIMK1 in yeast, as did cofilin^G4A (Fig. 6b, c). Furthermore, LIMK1 phosphorylated cofilin^G4A, cofilin^G4L, and cofilin^G4F at rates faster than cofilin^WT in vitro, while cofilin^G4P was not detectably phosphorylated (Fig. 6d). A comparison of steady-state kinetic parameters confirmed that cofilin^G4F (k_cat = 0.55 ± 0.05 s⁻¹, K_M = 12 ± 3 µM) was a more efficient LIMK1 substrate than cofilin^WT (k_cat = 0.20 ± 0.03 s⁻¹, K_M = 9 ± 3 µM) over the full range of substrate concentrations (Fig. 6f, g). Collectively these observations suggest that the lack of cofilin^G4F and cofilin^G4L growth phenotypes in yeast expressing LIMK1 were not due to reduced phosphorylation.

**Fig. 6: Effects of cofilin Glycine 4 substitutions on LIMK1 substrate suitability.**

As neither cofilin^G4F nor cofilin^G4L was impaired as a LIMK1 substrate, an alternative explanation for their ability to support growth in the presence of LIMK1 is that those mutations alleviate the inhibitory effect of Ser3 phosphorylation. To investigate this possibility, we used cofilin variants harboring a phosphomimetic S3D mutation to model constitutive phosphorylation. Consistent with its inability to sever actin, cofilin^S3D, as previously shown²⁴ was unable to support yeast growth (Fig. 7a). In contrast, yeast expressing cofilin^S3D,G4F or cofilin^S3D,G4L grew at equivalent rates to those expressing their respective Gly4 single mutant. When examined in vitro, both cofilin^S3D,G4F and cofilin^S3D,G4L bound actin similarly to their non-phosphomimetic counterparts, effectively correcting the S3D substitution (Fig. 7b). Together, these results suggest that the presence of a large hydrophobic residue immediately downstream of the phosphoacceptor residue uncouples cofilin inhibition from phosphorylation. Accordingly, the absolute conservation of a Gly residue at position 4 may be due at least in part because it is required for phosphoregulation of actin-binding and severing.

**Fig. 7: Cofilin Gly4 mutations counter the inhibitory effects of phosphomimetic S3D substitution.**

Discussion

Cofilin regulation of actin depolymerization is essential for normal cytoskeletal function^10,11,12. Much of the functional regulation of cofilin converges on its N-terminal site of phosphorylation, yet how this short four amino acid stretch simultaneously impacts its relationship with its regulators, the LIM kinases, and its effector, actin, has not been explored. We used a yeast-based dual-functionality screen of a cofilin N-terminal sequence library and subsequent functional analysis to simultaneously determine the fitness of all possible combinations of N-terminal residues and to assess the importance of each of these amino acids. We identified features required for the maintenance of cofilin core functionality, targeting of cofilin phosphorylation by its kinases, evasion of non-cognate kinase phosphorylation, and phosphoregulation of its ability to bind and sever actin. We found that each aspect of cofilin functionality is mediated by distinct sequence features of its N-terminal sequence, which collectively restrain sequence variation across the entire tail (Fig. 7c).

This system allowed us to simultaneously determine the fitness of all possible combinations of N-terminal residues. Previous mutagenesis library screens have similarly used yeast growth as a readout for kinase activity and docking interactions^38,39,40. Other approaches, such as phage display or droplet-based screening, could theoretically allow high-throughput analysis of actin binding, with a potential advantage of consistent levels of expression and a single readout⁴¹. However, it would be difficult to conduct large-scale in vitro screens examining actin severing or phosphorylation by LIMK. Some potential advantages of our approach are that all interactions occur in a eukaryotic cellular environment, that selection requires actin severing (and not simply binding) activity, and that we are able to simultaneously screen for cofilin function and regulation by LIMK.

Given the high degree of conservation of the cofilin N-terminal sequence, we were surprised to find relatively loose constraints on the sequence for maintaining sufficient activity to support growth. The deleterious effect of most substitutions was largely due to selectivity for aliphatic and β-branched amino acids at the fifth position. This range of residues is reflected in various cofilin orthologs, including those from mammals (Val), Arabidopsis (Met), maize (Leu), Acanthamoeba (Ile), and starfish (Thr)^24,42,43. In all three-dimensional structures of cofilin solved to date, the sidechain of this residue contacts the cofilin globular domain, suggesting that it may stabilize the protein (Supplementary Fig. 12). Indeed, cofilin^V5E aggregated readily during expression and purification. The suitability of variants at other positions was not readily predicted from known structures. For example, despite allowing interaction with actin, the presence of bulky residues in place of Gly4 is incompatible with the actin-binding mode observed by cryo-EM and would thus require an alternative conformation. By contrast, a more conservative G4A substitution was deleterious. Substitutions to Ala2 were most tolerated overall and had little effect on cofilin severing activity. This residue is absent from Acanthamoeba actophorin and is thus dispensable for actin binding⁴⁴. We conjecture that the conservation of small residues at this position may be important for maintaining N-terminal processing that would be dispensable in the context of our screen. We note that in our viability-based screen, variants promoting the fastest growth would not necessarily be those that bind actin with the highest affinity and that a hyperactive cofilin could be deleterious for growth. However, when examined individually, mutations that imparted fitness defects consistently reduced actin binding, and there was a general correlation between rank in the screen and affinity. Finally, we note that many variants capable of supporting growth did appear to have fitness defects evident in a pooled competitive growth format, but a large number were indistinguishable from the WT sequence.

Our finding that almost all cofilin variants with a Thr3 phosphoacceptor residue were resistant to LIMK1 expression and our subsequent biochemical analysis agree with a recent report that cofilin^S3T could not be phosphorylated by LIMK³⁰. The few Thr3 cofilin sequences selectively depleted by LIMK1 were strongly enriched for Tyr residues at the adjacent position (~40% of sequences, Supplementary Fig. 13). As LIMK1 can efficiently phosphorylate cofilin^S3Y, it would seem likely that these variants are phosphorylated on Tyr4 rather than Thr3³⁰. Ser/Thr kinase phosphoacceptor specificity is canonically determined by the identity of the residue immediately downstream of the DFG motif in the kinase activation loop (DFG + 1), which appears to promote a conformation facilitating phosphate transfer⁴⁵. However, due to the non-canonical nature of the interaction, cofilin Ser3 is too distant from the LIMK DFG + 1 residue for it to impart substrate specificity²⁹. Furthermore, the relative openness of the LIMK active site and the flexibility of the cofilin N-terminus suggest that steric clashes with the Thr3 sidechain are unlikely to preclude phosphorylation. Alternatively, LIMK and cofilin^S3T may form a non-productive complex that sequesters the phosphoacceptor from the catalytic center. The majority of Ser/Thr kinases across species prefer Ser over Thr, potentially explaining why Ser3 is conserved even in organisms lacking LIMK⁴⁶.

In addition to optimizing a substrate for its cognate kinase, a phosphorylation site sequence may also be tuned to avoid the large number of other kinases present within the cell. For example, autophosphorylation sites in the human epidermal growth factor receptor appear to be suboptimal as a means to avoid phosphorylation by multiple non-receptor Tyr kinases⁵. As most kinases make extended interactions on either side of the phosphosite, the location of Ser3 so close to the cofilin N-terminus may facilitate evasion of non-cognate kinases⁴⁷. By contrast, LIMK-cofilin binding does not require these canonical interactions, as demonstrated by its ability to phosphorylate the N-terminal Ser1 of actophorin⁴³. Notably, the extended N-terminal sequences of yeast and plant cofilin orthologs may facilitate a more canonical mode of interaction in organisms lacking LIMKs. By a similar token, because it lacks selective pressure to evade LIMKs, yeast twinfilin can be an efficient LIMK substrate, which we could attribute to its Ser phosphoacceptor. Because LIMK-induced growth suppression appears entirely reversed in the context of a cofilin S3A mutation, twinfilin phosphorylation does not appear to confer a growth defect, but its potential phosphoregulation under other conditions warrants further investigation. Though conservation of Thr at the N-terminus of mammalian twinfilin does not appear necessary to evade phosphorylation by LIMK, it may contribute to its insulation from other pathways.

Save for the exclusion of Pro residues, LIMK1 was promiscuous with respect to the sequence surrounding the cofilin phosphorylation site, consistent with a prior report examining a small panel of mutants⁴⁸. Indeed, LIMK-resistant sequences having bulky hydrophobic residues in place of Gly4 were phosphorylated faster by LIMK1. Our observation that G4F and G4L substitution can correct the phosphomimetic S3D mutation in vitro and in cells suggests that Gly4 is critical for cofilin to be regulated by phosphorylation. We postulate that this phenomenon underlies the widespread conservation of Gly4 in all cofilins. How might Gly4 optimize phosphoregulation? In molecular dynamics (MD) simulations, Ser3-phosphorylation or S3D mutation increases the dynamics of the N-terminal region and compromises its docking to actin²⁵. A flexible Gly4 residue could be required for phosphorylation to increase dynamics, which may contribute substantially to the undocking of the N-terminus. Alternatively, if the hydrophobic binding site for the N-terminal tail repels a charged phospho-Ser3, the sidechain of Leu or Phe could recapitulate the N-terminal binding interaction if bound in a distinct conformation. It has also been suggested that phospho-Ser3 participates in intramolecular interactions with one or more basic patches within the cofilin globular domain to sequester the N-terminus away from actin^49,50. In this scenario, reducing the flexibility of the N-terminal sequence by Gly4 substitution might preclude these intramolecular interactions, leaving the N-terminus free to interact with actin through the canonical binding mode. Further research will be required to distinguish between these models. Overall, our studies reveal specific determinants of cofilin regulation and how it interacts with and severs actin filaments.

Methods

Plasmids

The high-copy constitutive yeast expression vector (pRS423-GPD) for His₆-tagged human cofilin-1, the bacterial expression vector for N-terminally His₆-SUMO-tagged human cofilin-1 (in pEBDuet28a), and the galactose-inducible yeast expression vectors (pRS415-GAL1) for N-terminally FLAG-tagged LIMK1^CAT and full length LIMK1 were previously described²⁹. The cofilin-1 library and all cofilin mutants expressed in yeast using the pRS423-GPD vector were His₆-tagged (MASHHHHHHGAGA) N-terminal to the initiator methionine. The bacterial expression vector for N-terminally His₆-tagged human cofilin-1 was generated by excising the SUMO coding sequence from pEBDuet28a-His₆-SUMO-cofilin-1 by Gibson assembly. The high-copy yeast expression vector for yeast Cof1 was generated by shuttling the genomic DNA fragment from pRS316-COF1 (produced by Mark Hochstrasser’s laboratory)³⁴ into the EcoRI site of pRS423. Bacterial expression vectors for GST-tagged mouse twinfilin-1 and yeast twinfilin were a gift from the Bruce Goode laboratory^51,52. Point mutations in all plasmids were made by either QuikChange site directed mutagenesis or by Gibson assembly following amplification with mutagenic primers. Oligonucleotides used in this study for cloning and mutagenesis are provided in Supplementary Data 3.

To generate the cofilin-1 N-terminal mutagenesis library, an oligonucleotide pool spanning the region was synthesized with phosphoramidite trimer mixtures encoding all 20 amino acids at the sites encoding Ala2, Gly4, and Val5, and an ACC/AGC mixture (encoding a mix of Ser and Thr) in place of Ser3 (core DNA sequences encoding the four residues are provided in Supplementary Data 1, and the full sequence is provided in Supplementary Data 3). Oligonucleotides were PCR amplified and inserted into pRS423-GPD-His₆-cofilin-1 plasmid using the 5’ BamHI restriction site upstream of the start codon and a 3’ SalI restriction site that was introduced by silent mutagenesis within the cofilin-1 coding sequence. Electrocompetent DH10β E. coli (Invitrogen ElectroMAX) were transformed with ligation products, providing at least 1000 transformants per library variant. Plasmid library DNA was recovered from the pool of transformed bacterial colonies, PCR amplified, and sequenced on an Illumina NovaSeq instrument to confirm the full representation of all variants.

Cofilin library screening

The S. cerevisiae TeTO₇-COF1 strain³⁵ (Horizon Discovery TH_5610) was transformed sequentially with pRS415-GAL1-FLAG-LIMK1 and the cofilin library plasmid pool and frozen in aliquots to conduct replicate screens. For each screen, transformed yeast was pooled and grown at 30 °C with shaking in liquid culture with selective media (SC-His-Leu) containing 2% glucose at a starting OD₆₀₀ of 0.1. After 4–5 population doublings, a portion was diluted to an OD₆₀₀ of 0.1 with SC-His-Leu media containing 2% raffinose and 10 mg/L DOX. After the culture reached an OD₆₀₀ of 1–1.5, a portion was reserved for plasmid recovery (T₀ sample), and the remaining culture was split and diluted into SC-His-Leu liquid media containing 10 mg/L DOX and either 2% glucose or 2% raffinose/1% galactose to OD₆₀₀ = 0.1. Cultures were subjected to three growth and dilution cycles in which they were grown to an OD₆₀₀ of 1–2 and diluted back to an OD₆₀₀ of 0.1, saving a portion for plasmid recovery at each step (T₁–T₃ samples). After the screen was complete, plasmids were recovered from cell pellets using a QIAGEN Spin Miniprep kit. The N-terminal variable region of the plasmid pool was PCR amplified to attach sequencing adaptors and add barcodes, and sequenced on an Illumina NovaSeq instrument.

The relative representation of each cofilin variant at each timepoint was calculated by dividing the number of reads that variant by the total read count. The enrichment or depletion of a cofilin-1 library sequence in the absence of LIMK1 induction was calculated as the ratio of its relative representation at T₀ and T₂ in glucose. Enrichment or depletion following LIMK1 induction was calculated from timepoints T₁ and T₃ in glucose or galactose. Sequences were ranked by the average log fold changes in representation across three replicate screens. Data were processed using Microsoft Excel 16.75. Values in the heat map in Fig. 2 are the mean fold change of all library sequences containing the indicated residue at the position shown. Probability logos were generated using pLogo⁵³ with the indicated sequences as the foreground and the full library as the background. Logos indicate the probability that the frequency of a given residue at each position is different from the background frequency of all sequences in the library.

Immunoblotting

Yeast lysates were fractionated by either conventional SDS-PAGE or Mn²⁺ Phos-tag SDS-PAGE before transfer to polyvinylidene difluoride (PVDF) membranes. Phos-tag acrylamide gels were prepared with 12% acrylamide, 100 µM MnCl₂, and 50 µM Phos-tag Acrylamide (Fujifilm AAL-107). Membranes were blocked with 5% powdered milk in Tris-buffered saline solution with 0.05% Tween-20 (TBST) for 2 h at room temperature and incubated with primary antibodies diluted 1:2000 in TBST containing 5% milk overnight at 4 °C. Primary antibodies included: anti-Penta-His (QIAGEN, #34650), anti-FLAG M2 (Sigma-Aldrich, #F3165), and anti-Kss1 (Santa Cruz, #sc-6775-R). Membranes were washed with TBST before incubation with fluorophore-labeled anti-mouse IgG (LI-COR Biosciences, #D10603-05) or anti-rabbit IgG (Invitrogen, #A21109) secondary antibodies at 1:20,000 dilution in TBST containing 5% bovine serum albumin (BSA) for 1 h at room temperature. Membranes were washed with TBST and imaged with an Odyssey CLx imager (LI-COR Biosciences). Fluorescence signals were quantified using Image Studio 5.2.5 software.

Yeast growth assays

To evaluate the ability of cofilin variants to support growth, overnight cultures of TeTO₇-COF1 yeast transformed with the indicated expression vectors were diluted to OD₆₀₀ of 0.1 in selective media and grown in a shaking incubator at 30 °C until 3–4 doublings had occurred. Portions of each culture were reserved for immunoblotting, and a series of 5-fold dilutions (OD₆₀₀ range from 0.5 to 0.0008) were spotted onto SC-His agar plates with or without 10 mg/L DOX. Plates were incubated at 30 °C for 3–4 days before imaging. Lysates for immunoblotting were prepared by pelleting cells at 4700×g for 2 min and vortexing the cell pellet with glass beads in TCA extraction buffer (10 mM Tris, pH 8.0, 10% TCA, 25 mM NH₄OAc, 1 mM EDTA). Proteins were pelleted by centrifugation and resuspended in 100 mM Tris (pH 11.0) with 3% SDS before boiling for 5 min. Protein concentrations were determined by BCA assay, and equal amounts were separated by SDS-PAGE and subjected to immunoblotting as described above.

The effect of LIMK1 expression on cofilin phosphorylation and yeast growth was evaluated using TeTO₇-COF1 yeast co-transformed expression vectors for the indicated cofilin variants and pRS415-GAL1-FLAG-LIMK1, pRS415-GAL1-FLAG-LIMK1^CAT or empty pRS415-GAL1 vector as a control. Yeast was grown overnight in SC-His-Leu containing 2% raffinose to derepress the GAL promoter, diluted to an OD₆₀₀ of 0.1, and allowed to grow through 3–4 population doublings at 30 °C. A portion of the cells were spotted and grown as described above on SC-His-Leu agar plates containing 2% glucose, 2% glucose with 10 mg/L DOX, or 2% raffinose/1% galactose with 10 mg/L DOX. The remaining culture was split, and LIMK1 expression was either induced or repressed by adding galactose to 1% or glucose to 2%. After 12 h, equal OD₆₀₀ units of each culture were pelleted, and lysates were prepared by TCA extraction as described above. Lysates were fractionated by both conventional SDS-PAGE and Mn²⁺ Phos-tag SDS-PAGE before immunoblotting.

Protein purification

Human cofilin-1 variants were expressed in E. coli Rosetta(DE3) cells overnight at 16 °C by induction with 0.5 mM IPTG. Cells were pelleted, snap-frozen on dry ice/EtOH, thawed on ice, and resuspended in bacterial lysis buffer (20 mM Tris, pH 7.5, 140 mM NaCl, 1 mM DTT, 0.4% Igepal CA-630, 1 mM MgCl₂, 200 µg/ml lysozyme, 10 µg/ml pepstatin A, Roche cOmplete protease inhibitor cocktail, and 30 U/ml DNAse 1). Cells were lysed by sonication and pelleted by centrifugation (13,870×g for 30 min at 4 °C). The supernatant was extracted and incubated with TALON affinity resin (Takara Bio) while rotating at 4 °C for 1–2 h. The resin was transferred to a gravity flow column, drained, and washed once with 0.5% Igepal CA-630 in PBS and once in 10 mM imidazole, 20 mM Tris, pH 8.0, 140 mM NaCl. His₆-SUMO-tagged cofilin variants were washed an additional time with cofilin storage buffer (10 mM HEPES, 100 mM NaCl, 1 mM DTT, 10% glycerol) and eluted from the resin by overnight digestion with SUMO protease at 4 °C. N-terminally His₆-tagged cofilin variants were eluted with imidazole (20 mM Tris, pH 7.5, 250 mM imidazole, 100 mM NaCl, 1 mM DTT) and dialyzed into cofilin storage buffer overnight at 4 °C. Cofilin used for actin binding and severing assays was further purified by size exclusion chromatography on a Superdex200 10/300 GL column in 25 mM Tris, pH 8.0, 0.5 mM DTT, 0.25 mM EGTA. Cofilin eluted as a single peak across multiple fractions, which were pooled and concentrated using Amicon Ultra centrifugal filter concentrators (3 kDa cutoff). Protein concentration was determined by A₂₈₀ absorbance using a Thermo Scientific 2000 nanodrop instrument.

GST-tagged yeast and mouse twinfilin variants were expressed in E. coli, and cells were lysed as described above. The clarified cell lysate was incubated with glutathione Sepharose resin for 2 h at 4 °C. The resin was washed once with 0.5% Igepal CA-630 in PBS and once with cofilin storage buffer. Proteins were eluted by overnight digestion with GST-tagged HRV 3C protease at 4 °C. Twinfilin concentration was determined by comparison to BSA standards following SDS-PAGE and Coomassie-staining.

FLAG-tagged human LIMK1^CAT for in vitro kinase assays was purified using a yeast expression system. TeTO₇-COF1 yeast transformed with pRS423-GPD-His₆-cofilin-1^S3A (to support the growth of yeast during LIMK induction) and pRS415-GAL1-FLAG-LIMK1^CAT were grown in selective liquid media (SC-His-Leu) containing 2% raffinose at 30 °C with shaking to an OD₆₀₀ of 2. LIMK1 expression was induced by the addition of galactose and nutrient-rich yeast extract peptone (YP) broth, bringing the final concentrations of media to 1×YP media and 1% galactose. Cells were induced for 12 hr. at 30 °C before cells were harvested, pelleted, washed once with water, and snap frozen. The cell pellet was thawed on ice and resuspended in yeast lysis buffer (50 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, 10% glycerol, 0.5 mM DTT, 10 µg/ml leupeptin, 2 µg/ml pepstatin A, 2.5 mM NaPP_i, 10 µg/ml aprotinin, 1 mM β-glycerophosphate and 1 mM Na₃VO₄). Cells were lysed by vortexing with glass beads at 4 °C. Cell lysates were pelleted, and the recovered supernatant was incubated with M2 anti-FLAG affinity resin in batch for 1–2 h at 4 °C. The resin was washed twice with yeast lysis buffer and twice with FLAG wash buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 1 mM DTT, 0.01% Igepal CA-630, 10% glycerol, 1 mM β-glycerophosphate, and 100 µM Na₃VO₄) before the protein was eluted with FLAG wash buffer containing 0.5 mg/mL 3×FLAG peptide (Sigma). LIMK1 concentration was determined by running alongside BSA standards on SDS-PAGE with Coomassie staining.

To generate Ser3-phosphorylated cofilin, the FLAG-LIMK1 catalytic domain was expressed in yeast and purified as described above but not eluted from the resin. Purified cofilin^WT was incubated with resin-bound LIMK1 catalytic domain overnight at 4 °C in kinase reaction buffer (50 mM Tris, pH 7.5, 100 mM NaCl, 5 mM MgCl₂, 5 mM MnCl₂, 1 mM DTT) with 1 mM ATP. The resin was then pelleted, and the supernatant dialyzed overnight at 4 °C into 50 mM Tris, pH 6.8 with 1 mM DTT. Phosphorylated and unphosphorylated cofilin species were separated by cation-exchange chromatography using a MonoS 5/50 GL column. Cofilin phosphorylation was assessed by Mn²⁺ Phos-tag SDS-PAGE.

Actin filament binding assays

The binding affinity of cofilin variants for actin filaments was compared using in vitro pyrene quenching assays as previously reported^25,37,54. Serial dilutions of purified cofilin variants were incubated with polymerized pyrene-labeled actin (1 µM, ~83% labeled) in 1× KMI buffer (20 mM imidazole, pH 7.0, 50 mM KCl, 2 mM MgCl₂, 2 mM DTT, 200 µM ATP, 1 mM NaN₃) for 1 hr at 25 °C. Pyrene quenching was measured in a plate reader (excitation 366 nm, emission scan 409 nm). Data were normalized to the signal for unquenched pyrene-actin alone and saturated cofilactin filaments, and they were fitted to a model-independent sigmoidal dose-response curve using GraphPad Prism 9. The cofilin concentration at half-maximal binding was defined as K_0.5.

Actin severing assays

The severing activity of cofilin variants was assessed using TIRF microscopy as previously described²⁵. Alexa-647 labeled actin filaments (5% labeled) were polymerized for 1 h in 1× KMI buffer. Serial dilutions of purified cofilin variants were incubated with the polymerized actin for 1 h before placement on poly-L-lysine coated glass slips and imaged directly by TIRF microscopy. The recorded images were skeletonized using ImageJ FIJI and analyzed with Persistence Software version 4.2.3 to quantify the actin filament lengths⁵⁵. Persistence Software is available to download for free at delacruzlab.yale.edu/persistence-software.

Radiolabel kinase assays

Purified cofilin or twinfilin (2 µM) was incubated with LIMK1 catalytic domain (2 nM) and ATP (12.5 µM with 0.1 µCi/µL [γ-³²P]ATP) in kinase reaction buffer (defined above) at 30 °C. At 5 min and 10 min incubation, aliquots were removed and quenched with 4x SDS-PAGE loading buffer. Samples were boiled for 5 min and fractionated by 15% acrylamide SDS-PAGE. Gels were Coomassie stained and dried before exposure to a phosphor screen. Phosphor imaging and quantification of radiolabel incorporation were conducted using a BioRad Molecular Imager FX Pro Plus with Image Lab software. Images shown are from the 10 min timepoint. Phosphorylation rates were calculated from the entire time course and normalized to that of wild-type cofilin-1. Michaelis-Menten parameters for cofilin^WT and cofilin^G4F were determined from reactions run at seven substrate concentrations ranging from 50 to 0.5 μM. To calculate absolute rates, a [γ-³²P]ATP standard curve (2.5, 5, and 10 nCi) was imaged alongside the Coomassie-stained gel. Reaction kinetics were calculated using Microsoft Excel 16.75 and GraphPad Prism 9.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Next-generation sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) repository under accession code GSE242403. The processed data are available in the Supplemental Data 1 and 2 files. Yeast strains and plasmids described in this study are available upon request from the corresponding author. Source data are provided in this paper.

References

Studer, R. A. et al. Evolution of protein phosphorylation across 18 fungal species. Science 354, 229–232 (2016).
Article CAS PubMed ADS Google Scholar
Landry, C. R., Freschi, L., Zarin, T. & Moses, A. M. Turnover of protein phosphorylation evolving under stabilizing selection. Front. Genet. 5, 245 (2014).
Article PubMed PubMed Central Google Scholar
Manning, G., Whyte, D. B., Martinez, R., Hunter, T. & Sudarsanam, S. The protein kinase complement of the human genome. Science 298, 1912–1934 (2002).
Article CAS PubMed ADS Google Scholar
Shah, N. H., Lobel, M., Weiss, A. & Kuriyan, J. Fine-tuning of substrate preferences of the Src-family kinase Lck revealed through a high-throughput specificity screen. eLife 7, e35190 (2018).
Article PubMed PubMed Central Google Scholar
Cantor, A. J., Shah, N. H. & Kuriyan, J. Deep mutational analysis reveals functional trade-offs in the sequences of EGFR autophosphorylation sites. Proc. Natl Acad. Sci. USA 115, E7303–E7312 (2018).
Article CAS PubMed PubMed Central ADS Google Scholar
Johnson, J. L. et al. An atlas of substrate specificities for the human serine/threonine kinome. Nature 613, 759–766 (2023).
Article CAS PubMed PubMed Central ADS Google Scholar
Strumillo, M. J. et al. Conserved phosphorylation hotspots in eukaryotic protein domain families. Nat. Commun. 10, 1977 (2019).
Article PubMed PubMed Central ADS Google Scholar
Ochoa, D. et al. The functional landscape of the human phosphoproteome. Nat. Biotechnol. 38, 365–373 (2020).
Article CAS PubMed Google Scholar
Seet, B. T., Dikic, I., Zhou, M. M. & Pawson, T. Reading protein modifications with interaction domains. Nat. Rev. Mol. Cell. Biol. 7, 473–483 (2006).
Article CAS PubMed Google Scholar
Mizuno, K. Signaling mechanisms and functional roles of cofilin phosphorylation and dephosphorylation. Cell. Signal. 25, 457–469 (2013).
Article CAS PubMed Google Scholar
Maciver, S. K. & Hussey, P. J. The ADF/cofilin family: actin-remodeling proteins. Genome Biol. 3, reviews3007 (2002).
Article PubMed PubMed Central Google Scholar
Kanellos, G. & Frame, M. C. Cellular functions of the ADF/cofilin family at a glance. J. Cell Sci. 129, 3211–3218 (2016).
CAS PubMed Google Scholar
Bibeau, J. P., Gray, S. & De La Cruz, E. M. Clusters of a few bound cofilins sever actin filaments. J. Mol. Biol. 433, 166833 (2021).
Article CAS PubMed PubMed Central Google Scholar
Elam, W. A., Kang, H. & De La Cruz, E. M. Biophysics of actin filament severing by cofilin. FEBS Lett. 587, 1215–1219 (2013).
Article CAS PubMed PubMed Central Google Scholar
Suarez, C. et al. Cofilin tunes the nucleotide state of actin filaments and severs at bare and decorated segment boundaries. Curr. Biol. 21, 862–868 (2011).
Article CAS PubMed PubMed Central Google Scholar
De La Cruz, E. M. How cofilin severs an actin filament. Biophys. Rev. 1, 51–59 (2009).
Article PubMed Google Scholar
Maciver, S. K. & Weeds, A. G. Actophorin preferentially binds monomeric ADP-actin over ATP-bound actin: consequences for cell locomotion. FEBS Lett. 347, 251–256 (1994).
Article CAS PubMed Google Scholar
Wioland, H., Jegou, A. & Romet-Lemonne, G. Quantitative variations with pH of actin depolymerizing factor/cofilin’s multiple actions on actin filaments. Biochemistry 58, 40–47 (2019).
Article CAS PubMed Google Scholar
Wioland, H. et al. ADF/Cofilin accelerates actin dynamics by severing filaments and promoting their depolymerization at both ends. Curr. Biol. 27, 1956–1967.e1957 (2017).
Article CAS PubMed PubMed Central Google Scholar
Hocky, G. M., Sindelar, C. V., Cao, W., Voth, G. A. & De La Cruz, E. M. Structural basis of fast- and slow-severing actin-cofilactin boundaries. J. Biol. Chem. 296, 100337 (2021).
Article CAS PubMed PubMed Central Google Scholar
Schramm, A. C., Hocky, G. M., Voth, G. A., Martiel, J. L. & De La Cruz, E. M. Plastic deformation and fragmentation of strained actin filaments. Biophys. J. 117, 453–463 (2019).
Article CAS PubMed PubMed Central ADS Google Scholar
Schramm, A. C. et al. Actin filament strain promotes severing and cofilin dissociation. Biophys. J. 112, 2624–2633 (2017).
Article CAS PubMed PubMed Central ADS Google Scholar
Ribba, A. S., Fraboulet, S., Sadoul, K. & Lafanechere, L. The role of LIM kinases during development: a lens to get a glimpse of their implication in pathologies. Cells 11, 403 (2022).
Article CAS PubMed PubMed Central Google Scholar
Moriyama, K., Iida, K. & Yahara, I. Phosphorylation of Ser-3 of cofilin regulates its essential function on actin. Genes Cells 1, 73–86 (1996).
Article CAS PubMed Google Scholar
Elam, W. A. et al. Phosphomimetic S3D cofilin binds but only weakly severs actin filaments. J. Biol. Chem. 292, 19565–19579 (2017).
Article CAS PubMed PubMed Central Google Scholar
Nagaoka, R., Abe, H. & Obinata, T. Site-directed mutagenesis of the phosphorylation site of cofilin: its role in cofilin-actin interaction and cytoplasmic localization. Cell Motil. Cytoskeleton 35, 200–209 (1996).
Article CAS PubMed Google Scholar
Lappalainen, P., Fedorov, E. V., Fedorov, A. A., Almo, S. C. & Drubin, D. G. Essential functions and actin-binding surfaces of yeast cofilin revealed by systematic mutagenesis. EMBO J. 16, 5520–5530 (1997).
Article CAS PubMed PubMed Central Google Scholar
Huehn, A. R. et al. Structures of cofilin-induced structural changes reveal local and asymmetric perturbations of actin filaments. Proc. Natl Acad. Sci. USA 117, 1478–1484 (2020).
Article CAS PubMed PubMed Central ADS Google Scholar
Hamill, S., Lou, H. J., Turk, B. E. & Boggon, T. J. Structural basis for noncanonical substrate recognition of cofilin/ADF proteins by LIM kinases. Mol. Cell 62, 397–408 (2016).
Article CAS PubMed PubMed Central Google Scholar
Salah, E. et al. Lessons from LIMK1 enzymology and their impact on inhibitor design. Biochem. J. 476, 3197–3209 (2019).
Article CAS PubMed Google Scholar
Mezna, M. et al. Development of a high-throughput screening method for LIM kinase 1 using a luciferase-based assay of ATP consumption. J. Biomol. Screen. 17, 460–468 (2012).
Article CAS PubMed Google Scholar
Lappalainen, P. & Drubin, D. G. Cofilin promotes rapid actin filament turnover in vivo. Nature 388, 78–82 (1997).
Article CAS PubMed ADS Google Scholar
Nakano, K., Satoh, K., Morimatsu, A., Ohnuma, M. & Mabuchi, I. Interactions among a fimbrin, a capping protein, and an actin-depolymerizing factor in organization of the fission yeast actin cytoskeleton. Mol. Biol. Cell. 12, 3515–3526 (2001).
Article CAS PubMed PubMed Central Google Scholar
Kang, H. et al. Site-specific cation release drives actin filament severing by vertebrate cofilin. Proc. Natl Acad. Sci. USA 111, 17821–17826 (2014).
Article CAS PubMed PubMed Central ADS Google Scholar
Mnaimneh, S. et al. Exploration of essential gene functions via titratable promoter alleles. Cell 118, 31–44 (2004).
Article CAS PubMed Google Scholar
Moriyama, K. & Yahara, I. The actin-severing activity of cofilin is exerted by the interplay of three distinct sites on cofilin and essential for cell viability. Biochem. J. 365, 147–155 (2002).
Article CAS PubMed PubMed Central Google Scholar
De La Cruz, E. M. Cofilin binding to muscle and non-muscle actin filaments: Isoform-dependent cooperative interactions. J. Mol. Biol. 346, 557–564 (2005).
Article PubMed Google Scholar
Ahler, E. et al. A combined approach reveals a regulatory mechanism coupling Src’s kinase activity, localization, and phosphotransferase-independent functions. Mol. Cell 74, 393–408 (2019).
Article CAS PubMed PubMed Central Google Scholar
Shi, G. et al. Proteome-wide screening for mitogen-activated protein kinase docking motifs and interactors. Sci. Signal. 16, eabm5518 (2023).
Article CAS PubMed PubMed Central Google Scholar
Bhaduri, S. & Pryciak, P. M. Cyclin-specific docking motifs promote phosphorylation of yeast signaling proteins by G1/S Cdk complexes. Curr. Biol. 21, 1615–1623 (2011).
Article CAS PubMed PubMed Central Google Scholar
Rossenu, S. et al. A phage display-based method for determination of relative affinities of mutants. Application of the actin-binding motifs in thymosin beta 4 and the villin headpiece. J. Biol. Chem. 278, 16642–16650 (2003).
Article CAS PubMed Google Scholar
Bowman, G. D. et al. A comparative structural analysis of the ADF/cofilin family. Proteins 41, 374–384 (2000).
Article CAS PubMed Google Scholar
Blanchoin, L., Robinson, R. C., Choe, S. & Pollard, T. D. Phosphorylation of Acanthamoeba actophorin (ADF/cofilin) blocks interaction with actin without a change in atomic structure. J. Mol. Biol. 295, 203–211 (2000).
Article CAS PubMed Google Scholar
Quirk, S. et al. Primary structure of and studies on Acanthamoeba actophorin. Biochemistry 32, 8525–8533 (1993).
Article CAS PubMed Google Scholar
Chen, C. et al. Identification of a major determinant for serine-threonine kinase phosphoacceptor specificity. Mol. Cell 53, 140–147 (2014).
Article PubMed PubMed Central Google Scholar
Mok, J. et al. Deciphering protein kinase specificity through large-scale analysis of yeast phosphorylation site motifs. Sci. Signal. 3, ra12 (2010).
Article PubMed PubMed Central Google Scholar
Miller, C. J. & Turk, B. E. Homing in: Mechanisms of substrate targeting by protein kinases. Trends Biochem. Sci. 43, 380–394 (2018).
Article CAS PubMed PubMed Central Google Scholar
Amano, T., Tanabe, K., Eto, T., Narumiya, S. & Mizuno, K. LIM-kinase 2 induces formation of stress fibres, focal adhesions and membrane blebs, dependent on its activation by Rho-associated kinase-catalysed phosphorylation at threonine-505. Biochem. J. 354, 149–159 (2001).
Article CAS PubMed PubMed Central Google Scholar
Narita, A. ADF/cofilin regulation from a structural viewpoint. J. Muscle Res. Cell Motil. 41, 141–151 (2020).
Article CAS PubMed Google Scholar
Pope, B. J., Zierler-Gould, K. M., Kuhne, R., Weeds, A. G. & Ball, L. J. Solution structure of human cofilin: actin binding, pH sensitivity, and relationship to actin-depolymerizing factor. J. Biol. Chem. 279, 4840–4848 (2004).
Article CAS PubMed Google Scholar
Johnston, A. B., Collins, A. & Goode, B. L. High-speed depolymerization at actin filament ends jointly catalysed by Twinfilin and Srv2/CAP. Nat. Cell Biol. 17, 1504–1511 (2015).
Article CAS PubMed PubMed Central Google Scholar
Hilton, D. M., Aguilar, R. M., Johnston, A. B. & Goode, B. L. Species-specific functions of twinfilin in actin filament depolymerization. J. Mol. Biol. 430, 3323–3336 (2018).
Article CAS PubMed PubMed Central Google Scholar
O’Shea, J. P. et al. pLogo: a probabilistic approach to visualizing sequence motifs. Nat. Methods 10, 1211–1212 (2013).
Article PubMed Google Scholar
Cao, W., Goodarzi, J. P. & De La Cruz, E. M. Energetics and kinetics of cooperative cofilin-actin filament interactions. J. Mol. Biol. 361, 257–267 (2006).
Article CAS PubMed Google Scholar
Graham, J. S. et al. Multi-platform compatible software for analysis of polymer bending mechanics. PLoS ONE 9, e94766 (2014).
Article PubMed PubMed Central ADS Google Scholar

Download references

Acknowledgements

We thank Cameron Schmitz for technical assistance in generating cofilin mutants. This research was supported by the National Institute of General Medical Sciences of the National Institutes of Health under awards R01 GM102262 (to B.E.T. and T.J.B.) and R35 GM136656 (to EMDLC). J.S. and G.C.S. were supported by a T32 training grant from the National Institute of General Medical Sciences, and G.C.S. was supported by American Heart Association grants 835293 and 23DIVSUP1058562. TP was supported by the Yale College Dean’s Office and the Hahn Scholars Program.

Author information

Authors and Affiliations

Department of Pharmacology, Yale School of Medicine, New Haven, CT, 06520, USA
Joel A. Sexton, Hua Jane Lou, Titus J. Boggon & Benjamin E. Turk
Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, 06520, USA
Tony Potchernikov, Jeffrey P. Bibeau, Gabriela Casanova-Sepúlveda, Wenxiang Cao, Titus J. Boggon & Enrique M. De La Cruz

Authors

Joel A. Sexton
View author publications
You can also search for this author in PubMed Google Scholar
Tony Potchernikov
View author publications
You can also search for this author in PubMed Google Scholar
Jeffrey P. Bibeau
View author publications
You can also search for this author in PubMed Google Scholar
Gabriela Casanova-Sepúlveda
View author publications
You can also search for this author in PubMed Google Scholar
Wenxiang Cao
View author publications
You can also search for this author in PubMed Google Scholar
Hua Jane Lou
View author publications
You can also search for this author in PubMed Google Scholar
Titus J. Boggon
View author publications
You can also search for this author in PubMed Google Scholar
Enrique M. De La Cruz
View author publications
You can also search for this author in PubMed Google Scholar
Benjamin E. Turk
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

J.A.S. and B.E.T. designed the study, and J.A.S, T.P, J.P.B., G.C.S, W.C., and H.J.L. performed experiments. J.A.S., H.J.L., and G.C.S. generated new reagents used in this study. J.A.S., T.P., J.P.B., G.C.S, W.C., T.J.B., E.M.D., and B.E.T. designed experiments, analyzed data, curated data, and prepared figures. J.A.S. and B.E.T. wrote the initial draft, with additional editing provided by T.P., W.C., T.J.B., and E.M.D.

Corresponding author

Correspondence to Benjamin E. Turk.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Dominic Ng and the other anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Reporting Summary

Description of Additional Supplementary Files

Peer Review File

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Sexton, J.A., Potchernikov, T., Bibeau, J.P. et al. Distinct functional constraints driving conservation of the cofilin N-terminal regulatory tail. Nat Commun 15, 1426 (2024). https://doi.org/10.1038/s41467-024-45878-9

Download citation

Received: 31 July 2023
Accepted: 06 February 2024
Published: 16 February 2024
DOI: https://doi.org/10.1038/s41467-024-45878-9

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.