Article Information

• PDF (415 kb)
• Full Text with Large Figures
• Cited by in Scopus (4)

PubMed

• Articles by Andras Muhlrad
• Articles by Emil Reisler

Related Articles

• Cofilin and DNase I Affect the Conformation of the Small Dom...

  Cofilin and DNase I Affect the Conformation of the Small Domain of Actin Biophysical Journal, Volume 82, Issue 6, 1 June 2002, Pages 3134-3143 Irina V. Dedova, Vadim N. Dedov, Neil J. Nosworthy, Brett D. Hambly and Cris G. dos Remedios Abstract Cofilin binding induces an allosteric conformational change in subdomain 2 of actin, reducing the distance between probes attached to Gln-41 (subdomain 2) and Cys-374 (subdomain 1) from 34.4 to 31.4 Å (pH 6.8) as demonstrated by fluorescence energy transfer spectroscopy. This effect was slightly less pronounced at pH 8.0. In contrast, binding of DNase I increased this distance (35.5 Å), a change that was not pH-sensitive. Although DNase I-induced changes in the distance along the small domain of actin were modest, a significantly larger change (38.2 Å) was observed when the ternary complex of cofilin-actin-DNase I was formed. Saturation binding of cofilin prevents pyrene fluorescence enhancement normally associated with actin polymerization. Changes in the emission and excitation spectra of pyrene-F actin in the presence of cofilin indicate that subdomain 1 (near Cys-374) assumes a G-like conformation. Thus, the enhancement of pyrene fluorescence does not correspond to the extent of actin polymerization in the presence of cofilin. The structural changes in G and F actin induced by these actin-binding proteins may be important for understanding the mechanism regulating the G-actin pool in cells. Abstract | Full Text | PDF (241 kb)

• Control of Actin Reorganization by Slingshot, a Family of Ph...

  Control of Actin Reorganization by Slingshot, a Family of Phosphatases that Dephosphorylate ADF/Cofilin Cell, Volume 108, Issue 2, 25 January 2002, Pages 233-246 Ryusuke Niwa, Kyoko Nagata-Ohashi, Masatoshi Takeichi, Kensaku Mizuno and Tadashi Uemura Summary The ADF (actin-depolymerizing factor)/cofilin family is a stimulus-responsive mediator of actin dynamics. In contrast to the mechanisms of inactivation of ADF/cofilin by kinases such as LIM-kinase 1 (LIMK1), much less is known about its reactivation through dephosphorylation. Here we report Slingshot (SSH), a family of phosphatases that have the property of F actin binding. In , loss of function dramatically increased levels of both F actin and phospho-cofilin (P cofilin) and disorganized epidermal cell morphogenesis. In mammalian cells, human SSH homologs (hSSHs) suppressed LIMK1-induced actin reorganization. Furthermore, SSH and the hSSHs dephosphorylated P cofilin in cultured cells and in cell-free assays. Our results strongly suggest that the SSH family plays a pivotal role in actin dynamics by reactivating ADF/cofilin in vivo. Summary | Full Text | PDF (1135 kb)

• A New Internal Mode in F-Actin Helps Explain the Remarkable ...

  A New Internal Mode in F-Actin Helps Explain the Remarkable Evolutionary Conservation of Actin's Sequence and Structure Current Biology, Volume 12, Issue 7, 2 April 2002, Pages 570-575 Vitold E. Galkin, Margaret S. VanLoock, Albina Orlova and Edward H. Egelman Summary Actin is one of the most highly conserved eukaryotic proteins. There are no amino acid changes between the chicken and human skeletal muscle isoforms, and the most dissimilar actins still share more than 85% sequence identity . We suggest that large discrete internal modes of freedom within the actin filament may account for a significant component of this conservation, since each subunit must make multiple specific interactions with neighboring subunits. In support of this, we find that the same state of tilt of the actin subunit exists in both yeast and vertebrate striated muscle actin, and that in both the two domains undergo a “propeller rotation.” A similar movement of domains has also been seen in hexokinase , Hsc70 , and Arp2/3 , all structural homologs of actin , suggesting that such an interdomain hinge motion is common to proteins in this superfamily . Subunit-subunit interactions within the actin filament involve sequence insertions that are not present in MreB, a bacterial homolog of actin . Remarkably, we find that in the tilted state actin subunits make new contacts with neighboring subunits that also involve these inserts, suggesting a key role for these elements in F-actin polymorphism. Summary | Full Text | PDF (301 kb)

• …more

Copyright © 2006 The Biophysical Society. All rights reserved.
Biophysical Journal, Volume 91, Issue 12, 4490-4499, 15 December 2006

doi:10.1529/biophysj.106.087767

Muscle and Contractility

Antagonistic Effects of Cofilin, Beryllium Fluoride Complex, and Phalloidin on Subdomain 2 and Nucleotide-Binding Cleft in F-Actin

Andras Muhlrad^*, ,  , Israel Ringel^†, Dmitry Pavlov^‡, Y. Michael Peyser^* and Emil Reisler^‡

^* Institute of Dental Sciences, School of Dental Medicine, Hebrew University of Jerusalem, Jerusalem, Israel
^† Department of Pharmacology, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem, Israel
^‡ Department of Chemistry and Biochemistry and the Molecular Biology Institute, University of California, Los Angeles, California

Address reprint requests to Andras Muhlrad, Fax: 972-2-675-8561.

Actin has a central role in biological motility as an essential constituent of cytoskeleton and a partner of all myosin-based motor systems. In all eukaryotic cells actin exists in the rapidly interconverting monomer (G-actin) and polymer (filaments, F-actin) forms. The actin-based systems are highly dynamic and strongly regulated by a number of factors, including several actin-binding proteins. These factors can be subdivided into two antagonistic groups, which either stabilize or destabilize the structure of actin filaments.

Phalloidin, and inorganic phosphate (Pi) and its analogs—beryllium fluoride (BeFx) and aluminum fluoride (AlF₄)—belong to the group of F-actin stabilizing factors. These small molecules stabilize F-actin by reducing the critical concentration for polymerization and introducing conformational changes into the filament structure.

The complexes of beryllium and aluminum with fluoride were found to be good structural analogs of Pi 1 and are widely used in studying the activity of various nucleotide binding proteins, including G-proteins, Na⁺, K⁺-ATPase, tubulin, and others 2. Combeau and Carlier found that BeFx (BeFx stands for the and BeF₂(OH)⁻ complexes) and AlF₄ bind strongly to F-actin 3. BeFx and AlF₄ bind to F-actin with orders of magnitude higher affinity (K_d=2μM and 25μM, respectively) than Pi (K_d=1.5mM). BeFx competes with Pi for binding to the nucleotide-binding cleft of ADP-F-actin protomers at the place of the γ-phosphate of ATP 3. It stabilizes strongly F-actin by decreasing the rate of protomer dissociation (∼150-fold) and the critical concentration for polymerization (∼100-fold) 3. BeFx stabilizes in particular the structure of subdomain 2, as indicated by strong and cooperative inhibition of its cleavage by subtilisin (in the DNase I binding-loop (D-loop)) and trypsin (in the 60–69 loop) 4, and by electron microscopy studies 5. These effects of BeFx on F-actin are similar to those of Pi 3,4, but BeFx is more effective at much smaller concentrations. BeFx-induced changes in the C-terminus region were also detected by fluorescence 3 and proteolysis 4 methods.

Phalloidin has the strongest stabilizing effect on actin filaments. It decreases the critical concentration of actin polymerization, reduces the rate of monomer dissociation from both filament ends 6, and inhibits phosphate release from the nucleotide binding cleft after ATP hydrolysis. Phalloidin binds at the interface of three actin monomers 7,8,9 and stabilizes lateral interactions between the two filament strands.

Among actin destabilizing factors, the actin depolymerizing factor/cofilin (ADF)/cofilin) family of proteins, or AC proteins 10, have attracted much attention because of their important role in regulating actin dynamics in cells. These proteins change the twist of actin filaments 11, and destabilize, sever 12, and depolymerize 13 them by weakening longitudinal 14,15 and lateral 16,17 interprotomer contacts in F-actin. Extensive, cofilin-induced conformational changes in subdomain 2 of F-actin are readily monitored via quenching of the fluorescence of tetramethyl rhodamine cadaverine (TRC) (attached to Gln-41 on the D-loop 18), and a strong acceleration of subtilisin (between Met-47 and Gly-48) and tryptic cleavage (after Arg-62 and Lys-68) in subdomain 2 18.

The antagonistic structural effects of AC proteins and BeFx (and phalloidin) on F-actin appear consistent with their reciprocal inhibition of binding to F-actin 13 and the reported blocking of Acanthamoeba actophorin (AC protein) binding by BeFx 19 and human cofilin binding by phalloidin 20 to F-actin. On the other hand, we found that yeast cofilin removes rhodamine-phalloidin from F-actin 17, and obtained preliminary evidence 21 that BeFx inhibits the rate but not the extent of yeast cofilin binding to muscle F-actin. Following these observations, we used here the binding of yeast cofilin to BeFx-F-actin as a tool to study the structural effects of BeFx on actin filaments. Cofilin binding to F-actin was monitored via changes in the susceptibility of subdomain 2 to subtilisin and trypsin, and via changes in the fluorescence intensity of the TRC probe attached Gln-41 18 and of 1,N⁶-ethenoadenosine diphosphate (ɛ-ADP) bound in the nucleotide-binding cleft of F-actin. In addition, we also monitored the effect of cofilin on the release of BeFx from F-actin by ¹⁹F-NMR. We found that BeFx strongly inhibits the rate, but not the extent of cofilin binding, while cofilin greatly facilitates the dissociation of BeFx from F-actin, and affects the structure of subdomain 2 and the nucleotide binding cleft of F-actin. Our results indicate the existence of two types of BeFx-F-actin complexes, with different conformations and different rates of cofilin binding.

Tetramethyl rhodamine cadaverine (TRC) was obtained from Molecular Probes (Eugene, OR). ATP, 1,N⁶-ethenoadenosine triphosphate (ɛ-ATP) trypsin, soybean trypsin inhibitor, subtilisin (Carlsberg), phalloidin and phenylmethylsulfonyl fluoride (PMSF) were purchased from Sigma Chemical (St Louis, MO). Bacterial transglutaminase was a generous gift from Dr. K. Seguro (Ajimoto, Kawasaki, Japan).

G-actin was prepared from back and leg muscles of rabbit by the method of Spudich and Watt 22 and stored in G-buffer containing 5.0mM TrisHCl, 0.2mM CaCl₂, 0.2mM ATP, 0.5mM dithiotreitol, pH 8.0. F-actin was prepared from G-actin by polymerizing it with 2.0mM MgCl₂. Recombinant yeast cofilin was prepared as described before 23 with minor modifications 15. The concentrations of cofilin and unlabeled skeletal muscle α-actin were determined spectrophotometrically by using the extinction coefficients and , respectively. (The optical density of actin was measured in the presence of 0.5M NaOH, which shifts the maximum of absorbance from 280 to 290nm). Molecular masses were assumed to be 42 and 15.9 kDa for skeletal actin and yeast cofilin, respectively.

Labeled or unlabeled F-actin (10μM) was digested in the presence and absence of cofilin in pH 8.0 F-buffer (20.0mM TrisHCl, 2.0mM MgCl₂, 0.2mM ATP, 0.5mM dithiotreitol), and pH 6.5 F-buffer (20.0mM PIPES, 2.0mM MgCl₂, 0.2mM ATP, 0.5mM dithiotreitol), with 25μg/ml subtilisin or 800μg/ml trypsin, respectively. The products of digestion were run on SDS-PAGE. Protein bands on SDS gels were analyzed by densitometry.

Actin labeled with TRC at Gln-41 (TRC-actin) was prepared by incubating 50μM skeletal G-actin with 100μM TRC and 0.18mg/ml bacterial transglutaminase in G-buffer pH 8.0, at 22°C for 2h. Reagent excess was removed by filtering actin through a PD-10 column equilibrated with G-buffer. The extent of actin labeling for TRC was estimated using extinction coefficient of E₅₅₄=78,000cm⁻¹M⁻¹. The concentration of the labeled actin was measured by the Bradford protein assay 24, using native actin as a standard.

This was done essentially as described previously 25. Briefly, skeletal muscle G-actin was passed through a desalting column (Amersham, PD10, Piscataway, NJ) of Sephadex G-25 equilibrated with ATP-free G-buffer. The eluted actin was supplemented with a 20-fold molar excess of ɛ-ATP and was incubated for 1h on ice. Excess ɛ-ATP was removed from G-actin by passing it through another PD10 column. Actin was polymerized by addition of 2.0mM MgCl₂ and during the polymerization the actin-bound ɛ-ATP was hydrolyzed to ɛ-ADP.

Fluorescence measurements were carried out at 22°C with a PTI spectrofluorometer (Photon Technology Industries, South Brunswick, NJ) in pH 8.0 and pH 6.5 F-buffer. For TRC and ɛ-ADP, the excitation wavelength was set at 544 and 350nm and the emission at 580 and 412nm, respectively. Emission spectrum of ɛ-ADP-F-actin was recorded between 370 and 550nm wavelengths.

Spectra were obtained on the Varian Inova 500 instrument at 470.215MHz, in a 5-mm probe. Spectrum width was 50,000Hz, delay time 1s, and acquisition time 0.7s. Line broadening function was set to 20Hz. In all experiments temperature was kept at 20°C and the solutions were supplemented with 15% D₂O.

We showed before that labeling of Gln-41 on actin with tetramethyl rhodamine cadaverine (TRC) offers a convenient tool for monitoring the binding of cofilin to F-actin. The fluorescence intensity of TRC-F-actin is decreased by >70% upon cofilin binding, while the fluorescence intensity of G-actin changes little with the addition of cofilin 18. Here, we monitored the decrease in TRC fluorescence (Fig. 1) upon addition of 12μM cofilin to 10μM TRC-F-actin in the presence and absence of 60μM BeFx at pH 8.0 and 6.5 (BeFx alone did not affect the fluorescence of TRC-F-actin). In the absence of BeFx, the decrease in the fluorescence of TRC-F-actin by cofilin was essentially completed within the mixing time of the solutions, in agreement with our earlier observations 18. Yeast cofilin also binds to TRC-F-actin in the presence of BeFx at both pH-s, but at a much slower rate than in its absence. Fig. 1 shows that BeFx inhibits the rate, but not the extent of cofilin binding and that this binding is faster at pH 6.5 than at pH 8.0. The time course of fluorescence changes could be well fitted to a two-exponential expression, yielding the apparent first order fast (k_f) and slow (k_s) cofilin binding rates at pH 6.5 and 8.0, from which the second order association rate constants were calculated by taking into account their dependence on cofilin concentration (Fig. 2 and Table 1).

Display large version of this figure

Figure 1

Effect of cofilin on the fluorescence of BeFx-TRC-F-actin at pH 8.0 and 6.5. TRC-F-actin, 10μM, was incubated with 60μM BeCl₂ and 5mM NaF in a pH 8.0 or pH 6.5 F-buffer, overnight, on ice. Cofilin (12μM) was added to the sample (arrow) after the initial reading, and the time course of fluorescence intensity change was monitored at 22°C as described in Materials and Methods.

Display large version of this figure

Figure 2

Effect of cofilin concentration on the rate of cofilin binding to BeFx-TRC-F-actin. TRC-F-actin, 4μM, was incubated with 120μM BeCl₂ and 5mM NaF in a pH 8.0 F-buffer, overnight, on ice. Cofilin (6–40μM), as indicated on the figure, was added to the sample (arrow) after the initial reading, and the time course of fluorescence intensity change was monitored at 22°C as described in Materials and Methods. Cofilin concentration in μM is given on each curve.

To see whether cofilin binds to BeFx containing F-actin protomers or the binding is limited by the dissociation of BeFx from the protomers, we measured the effect of increasing cofilin concentration on the rate of binding (Fig. 2). We monitored the binding of 6–40μM cofilin to 4μM TRC-F-actin in the presence of 120μM BeFx. Relatively high concentration of BeFx was used to slow down the reaction. The binding rates were found to increase significantly with cofilin concentration, indicating that the binding is not limited by the dissociation of BeFx. All binding curves could be well fitted to a two-exponential expression. In the presence of 40μM cofilin and 4μM TRC-F-actin the k_f and k_s (apparent first-order rate constants) were 0.0515 and 0.0105s⁻¹, respectively. At the lowest cofilin concentration (6μM) the k_f and k_s rates were 0.0071 and 0.0012s⁻¹, respectively. The calculated second-order association rate constants were presented in Table 1. Because some AC proteins, such as ADF1 and actophorin, apparently do not bind to BeFx-F-actin 13,19, we have confirmed the binding of yeast cofilin to BeFx-TRC-F-actin also by their cosedimentation at pH 8.0 (Fig. 3). Similar results were obtained also with unlabeled actin, both at pH 6.5 and 8.0 (data not presented), showing that yeast cofilin binds to F-actin in the presence of BeFx.

Display large version of this figure

Figure 3

Cosedimentation of TRC-F-actin with cofilin in the presence and absence of BeFx at pH 8.0. TRC-F-actin, 10μM, was incubated with 60μM BeCl₂ and 5mM NaF in a pH 8.0 F-buffer, for 2h, on ice. Cofilin (12μM)was added to actin samples, which were incubated for 20min and then centrifuged at 80K at 20°C for 30min. Prespin (ps), supernatant (sup), and pellet (pell) samples of these solutions were analyzed by SDS-PAGE.

The binding of BeFx to the nucleotide binding cleft of F-actin stabilizes strongly the structure of subdomain 2, which is manifested in its resistance to subtilisin and trypsin cleavage 4. Cofilin binding has the opposite effect, as it increases dramatically the proteolysis of the D- and 60–69-loops by subtilisin and trypsin 18. We show in Figure 4A that cofilin increases the rate and extent of subtilisin cut in the D-loop of BeFx-F-actin at pH 6.5 and pH 8.0, with a bigger effect noted at the lower pH. This is consistent with the TRC-F-actin fluorescence results (Fig. 1). The rate and the extent of subtilisin cleavage increased with the time of cofilin incubation with BeFx-F-actin (Figure 4B). Similar results were obtained also for the tryptic cleavage of the 60–69 loop (after Arg-62 and Lys-68), which became faster and more extensive with cofilin incubation (Figure 4C). It should be noted, however, that even after 20min incubation with cofilin both the subtilisin and trypsin digestions were less extensive than in the presence of cofilin without BeFx. The results of proteolysis experiments confirm cofilin binding to BeFx-F-actin.

Display large version of this figure

Figure 4

Effect of cofilin on the limited proteolysis of BeFx-F-actin at pH 8.0 and 6.5. F-actin (33μM) was incubated with 100μM BeCl₂ and 5mM NaF in the pH 8.0 or pH 6.5 F-buffer, overnight, on ice. (A) After 1min incubation of 10μM F-actin or BeFx-F-actin with 12μM cofilin at 22°C, actin was digested by 25μg/ml subtilisin for 20, 50, and 80s. The samples were run on SDS-PAGE and analyzed by densitometry. (Solid symbols and solid line) pH 8.0; (open symbols and dotted line) pH 6.5; (■) BeFx only; (▴) BeFx and cofilin; (▾) no addition; (●) cofilin only. (B) After 0.5, 5, and 20min incubation of 10μM BeFx-F-actin with 10μM cofilin at 22°C, actin was digested with 18μg/ml subtilisin at pH 8.0 for 20, 50, and 80s. (○) BeFx only; (■) no BeFx and cofilin; (♦) 0.5min incubation with cofilin; (▾) 5min incubation with cofilin; (●) 20min incubation with cofilin; (▴) cofilin only, no BeFx. (C) After 0.5, 5, and 20min incubation of 10μM BeFx-F-actin with 10μM cofilin at 22°C it was digested by 0.8 mg/ml trypsin at pH 8.0 for 20, 50, and 80s. (■) BeFx only; (○) no addition; (▴) 0.5min incubation with cofilin; (▾) 5min incubation with cofilin; (●) 20min incubation with cofilin; (□) cofilin only, no BeFx.

To test whether cofilin removes the bound BeFx from the nucleotide-binding cleft of F-actin or binds to BeFx-actin without releasing this phosphate analog, we used ¹⁹F NMR. The free fluoride and beryllium were removed from BeFx-F-actin by extensive dialysis, after which the bound BeFx was released by denaturing actin with perchloric acid. Actin denaturation was needed because the actin bound fluoride does not give the ¹⁹F signal. The bound fluoride (0.6mol per mol actin) was calculated from the ¹⁹F NMR spectrum (Fig. 5). We added cofilin to another aliquot of the dialyzed F-actin and calculated the dissociation of the bound BeFx from actin from the recorded ¹⁹F NMR spectrum (Fig. 5). According to these measurements, ∼80% of the bound fluoride dissociates from F-actin upon 30min incubation with cofilin. These results indicate that the binding of cofilin induces conformational changes in the nucleotide-binding cleft of F-actin, and decreases BeFx affinity to actin.

Display large version of this figure

Figure 5

Removal of BeFx from BeFx-F-actin. BeCl₂ (150μM) and 5mM NaF were added to 168μM F-actin and incubated overnight, on ice. This was followed by dialysis (four changes) against 100-fold volume of F-buffer, pH 8.0, for 2 days. ¹⁹F NMR measurements were carried out at 20°C as described in Materials and Methods. (A) 0.5mM fluoride in F-buffer; (B) 0.5mM fluoride in 6% perchloric acid; (C) 104.0μM BeFx-F-actin in F-buffer; (D) 104.0μM BeFx-F-actin in 6% perchloric acid was centrifuged and the supernatant was measured; (E) 90.2μM cofilin was added to 104.0μM BeFx-F-actin in F-buffer.

In the light of the above findings we tested the effect of BeFx and cofilin on the nucleotide-binding cleft in F-actin by substituting the actin-bound ADP with its fluorescent ɛ-ADP analog. We found that upon addition of BeFx to ɛ-ADP-F-actin the fluorescence intensity of the bound ɛ-ADP increases by ∼11% (Figure 6A). The time course of the binding of BeFx to ɛ-ADP-F-actin (Figure 6B), which was fitted to a single exponential expression, is rather slow. The calculated second order association rate constant of this reaction is 3.67±0.57×10⁻⁵s⁻1μM⁻¹. BeFx does not affect the quenching of the fluorescence intensity of actin bound ɛ-ADP by nitromethane (Figure 6C), (K_SV values at pH 8.0 in the absence and presence of BeFx are 1.87±0.1M⁻¹ and 1.86±0.1M⁻¹, respectively). On the other hand cofilin significantly reduces the accessibility of nitromethane to ɛ-ADP in the nucleotide-binding cleft of F-actin (K_SV=0.926±0.04M⁻¹) (see also Muhlrad et al. 25). Addition of cofilin to ɛ-ADP-F-actin also increases the fluorescence intensity of the bound ɛ-ADP by ∼50% 25, which is significantly more than the BeFx induced increase (Figure 6A). The fluorescence change observed upon adding cofilin in the absence of BeFx to ɛ-ADP-F-actin is very fast while it is slow in the presence of BeFx (Figure 6B), which is consistent with TRC-F-actin fluorescence results (Fig. 1). When cofilin was added to BeFx-ɛ-ADP-F-actin the overall fluorescence increase was smaller (30%) than that with cofilin and BeFx-free ɛ-ADP-F-actin (50%) (Fig. 6).

Display large version of this figure

Figure 6

Effect of BeFx and cofilin on the fluorescence of ɛ-ADP-F-actin. To 8μM ɛ-ADP-F-actin in pH 8.0 F-buffer 0.1mM BeCl₂ and 5mM NaF was added, followed by the addition of 10μM cofilin. (A) Spectrum, (1) before addition of BeFx; (2) 35min after addition of BeFx; (3) 2h after addition of cofilin to BeFx-ɛ-ADP-F-actin; (4) cofilin added to ɛ-ADP-F-actin in the absence of BeFx. (B) Time course of fluorescence change following BeFx and cofilin additions. BeFx addition is shown by the first arrow; cofilin additions to BeFx-ɛ-ADP-F-actin and BeFx free ɛ-ADP-F-actin are indicated by the second (Cofilin A) and third arrow (Cofilin B), respectively. (C) Nitromethane was added to 8μM ɛ-ADP-F-actin in 10mM increments in the presence of 10μM cofilin (▾); in the presence of 0.1mM BeFx (■); and in the absence of cofilin or BeFx (●). The fluorescence intensity change was monitored at 22°C as described in Materials and Methods.

The time course of cofilin binding to BeFx-TRC-F-actin revealed two reaction steps, with fast and slow rate constants (Fig. 1 and Table 1). Because the binding of cofilin is accompanied by BeFx release from F-actin, as shown by NMR results, our data suggest the presence of at least two forms of BeFx-F-actin, tightly and weakly bound, which would account for the fast and slow cofilin binding. We tested this idea by varying the BeFx concentration (10–100μM) and its incubation time with F-actin (1 and 24h), and by using ∼1:2mol ratio of cofilin/actin. Predictably, both the relative extent and rate of the fast and slow fluorescence decrease upon cofilin binding to TRC-F-actin depended on BeFx concentration (up to 60μM), as did also the tryptic digestion of such actin (data not shown).

More revealing were the experiments in which we examined the effect of F-actin preincubation with BeFx on the binding of cofilin (Figure 7A). TRC-F-actin (10μM) was incubated with 60μM BeFx for 1 and 24h, respectively, and then mixed with cofilin (5.6μM). The initial, fast fluorescence decrease was faster and greater for the sample incubated with BeFx for 1h than for 24h. Similarly, the rate and the extent of subtilisin digestion of BeFx-F-actin in the presence of cofilin were greater after 1h than 24h incubation with BeFx (Figure 7B). These results indicate that the binding of cofilin to BeFx-F-actin depends on the incubation time of F-actin with this phosphate analog, i.e., most likely, on the ratio of the strongly to the weakly bound BeFx-F-actin complex, which increases with the time of incubation. Interestingly, in the absence of cofilin no difference was observed in the inhibition of subdomain 2 proteolysis between samples incubated for 1h and 24h with BeFx. In fact, full protection against proteolysis of F-actin by subtilisin appears already after 5min incubation with 60μM BeFx (Figure 7C). This result is consistent with the time course of the fluorescence intensity increase observed upon adding BeFx to ɛ-ADP-F-actin (Figure 6B). Similar results were obtained at pH 6.5 with subtilisin and at pH 8.0 with trypsin digestion (data not shown). Thus, although both the weakly and strongly bound BeFx-F-actin complexes are equally well protected against proteolysis in the absence of cofilin, they appear to present different binding environments to cofilin.

Display large version of this figure

Figure 7

Dependence of cofilin binding on the incubation time of TRC-F-actin with BeFx. TRC-F-actin, 10μM, was incubated with 60μM BeCl₂ and 5mM NaF in pH 8.0 F-buffer for 1 and 24h on ice. (A) 5.6μM cofilin was added to 10μM TRC-F-actin (arrow) and the time course of fluorescence intensity change was monitored at 22°C as described in Materials and Methods. (B) 10μM TRC-F-actin, which had been incubated with BeFx, was mixed with 12μM cofilin and digested by 25μg/ml subtilisin for 20, 50, and 80s at 22°C. The samples were run on SDS-PAGE and analyzed by densitometry. (▿) 1-h incubation with BeFx, no cofilin; (▾) 24-h incubation with BeFx, no cofilin; (○) 1-h incubation with BeFx, then cofilin added; (●) 24-h incubation with BeFx, then cofilin added; (■) no addition; (▴) cofilin only, no BeFx. (C) Effect of incubation time of BeFx with F-actin on its digestion with subtilisin. F-actin, 10μM, was incubated with 60μM BeFx or 5mM NaF for various time intervals, and then was digested with 50μg/ml subtilisin for 60, 120, and 240s at 22°C at pH 8.0. Digestion samples were run on SDS-PAGE and analyzed by densitometry. (■) no BeFx; (▴) 1-h incubation with 5mM NaF; (▾) 0.5-min incubation with BeFx; (□) 5-min incubation with BeFx; (♦) 1-h incubation with BeFx; (●) 24-h incubation with BeFx.

Because cofilin displaces BeFx from F-actin, we tested also the reverse case, i.e., cofilin displacement by BeFx. To this end, cofilin (4.0, 8.0, and 11.0μM) was added first to TRC-F-actin (10μM) at pH 6.5 or 8.0, and then mixed with 5mM NaF and 100μM BeCl₂. As shown in Fig. 8, there was an immediate, cofilin concentration-dependent drop in TRC-F-actin fluorescence (Table 2), reflecting the formation of a complex. This fluorescence decrease was reversed by BeFx slowly, and only to a small extent, at substoichiometric ratios of cofilin/actin (Table 2). The fluorescence intensity recovery decreased with increasing cofilin concentration. At pH 6.5 no fluorescence intensity recovery due to BeFx was observed at a saturating cofilin concentration (11μM; Table 2). Similar conclusion was reached from subtilisin digestion experiments; BeFx did not affect the digestion of F-actin saturated with cofilin (data not shown). These results indicate that BeFx reverses only partially the effect of substoichiometric cofilin on F-actin structure, but not when F-actin is saturated with cofilin. It appears that the binding of BeFx in the nucleotide binding cleft is inhibited in those actin protomers to which cofilin is attached.

Display large version of this figure

Figure 8

Reversal of cofilin effect on the fluorescence of TRC-F-actin by BeFx. To 10μM TRC-F-actin in pH 6.5 F-buffer 4.0–11.0μM cofilin was added (arrow) after the initial reading. This was followed by addition of 5mM NaF (second arrow) and 100μM BeCl₂ (third arrow) and the time course of fluorescence intensity change was monitored at 22°C, as described in Materials and Methods.

We found earlier that yeast cofilin, unlike some other AC proteins, binds to phalloidin–F-actin and dissociates phalloidin or rhodamine-phalloidin 17. Here we tested the effect of phalloidin on cofilin binding to F-actin by subtilisin digestion, taking advantage of the strong phalloidin inhibition of the D-loop cleavage 26. Phalloidin, 12μM, was added to 10μM F-actin containing 5 or 12μM cofilin and after 1.5 or 22h incubation at pH 8.0 the samples were digested with subtilisin. We also reversed the order of additions, adding phalloidin first and cofilin second. The results of such digestions in the presence of 12μM cofilin are presented in Fig. 9. The digestion pattern clearly shows that the addition of phalloidin to cofilin-F-actin decreases, whereas the addition of cofilin to phalloidin F-actin increases the rate and extent of subtilisin cleavage. However, the extent of actin cleavage after 90min incubation (Figure 9A) was greater when phalloidin was added to cofilin-F-actin than in the case of a reversed order of additions, when cofilin was added to phalloidin-F-actin. Such digestion differences disappeared after 22h incubation (Figure 9B), showing the slow equilibration of this system. These results indicate that phalloidin, unlike BeFx or inorganic phosphate 25, binds to F-actin also in the presence of cofilin.

Display large version of this figure

Figure 9

Phalloidin reverses the effect of cofilin on the subtilisin digestion of F-actin. F-actin (10μM) was incubated with cofilin (12μM) for 5min in F-buffer, at pH 8.0. Phalloidin (12μM) was added to that sample, and after 90min incubation at room temperature (A), or 22h incubation on ice (B), the samples were digested with 25μg/ml subtilisin for 20, 50, and 80s at 22°C, and then run on SDS-PAGE and analyzed by densitometry. In the reversed order experiment, 12μM phalloidin was added first to F-actin, and after 5min this was followed with the addition of 12μM cofilin. (○) No addition; (■) only phalloidin added; (▴) phalloidin added first then cofilin; (▵) cofilin added first then phalloidin; (●) only cofilin added.

The structural phosphate analog BeFx 2 binds strongly to F-actin, stabilizing its structure, and according to several reports prevents the binding of some ADF/cofilin proteins (Acanthamoeba actophorin 19, human cofilin 20, and plant ADF 13) to F-actin. We found in this study that in contrast to above-mentioned three members of the AC family, yeast cofilin binds to BeFx-F-actin, albeit at a much slower rate than to F-actin. This feature of yeast cofilin indicates that it has a higher affinity to F-actin than those AC family members that do not bind to F-actin in the presence of BeFx. The high affinity of yeast cofilin to F-actin may have physiological significance. We took advantage of this property of yeast cofilin and examined its binding to BeFx-actin to gain further insight into the changes caused in actin structure by this phosphate analog.

In agreement with Ressad et al. 27, we found that the binding of cofilin is faster at low than at high pH 25, although the depolymerizing effect is stronger at the high pH 28. In the presence of BeFx the binding rates at pH 8.0 and pH 6.5 are much closer (Table 1). It is possible that the nature of the beryllium fluoride complex that binds to F-actin contributes to the change in the rate difference at the two pH values. According to Combeau and Carlier 29, at low pH only binds, whereas at high pH both and BeF₂(OH)⁻ bind to F-actin (only ions with a single negative charge bind to the nucleotide-binding cleft of actin). may bind stronger to F-actin than BeF₂(OH)⁻ thereby, inhibits more effectively cofilin binding, which would explain the relatively stronger inhibition of cofilin binding at low pH. Similar conclusion was reached by studying the activation of transducin by the above two beryllofluoride complexes 30.

We monitored the binding of cofilin to F-actin by following the decrease in fluorescence intensity of the TRC group attached to Gln-41. The rate of cofilin binding was found to increase with increasing cofilin concentration without reaching a plateau even at 10:1 cofilin/actin molar ratio. This indicates that cofilin binds to BeFx containing F-actin protomers and the BeFx dissociation does not limit the rate of cofilin binding. This conclusion is supported by the results of Combeau and Carlier 3, who found by two methods (rapid dialysis and chasing out ⁷Be with unlabeled Be) that the rate of dissociation of BeFx from F-actin is extremely slow ∼10⁻⁶s⁻¹. We suggest that cofilin binding to BeFx containing protomers causes conformational changes at the BeFx binding site in the nucleotide-binding cleft. These changes induce dramatic decrease in the affinity of BeFx to F-actin, leading to its dissociation.

We studied the effects of BeFx and cofilin on the nucleotide-binding cleft of F-actin also by detecting changes in the fluorescence of ɛ-ADP bound to F-actin. The different effects of these two ligands on ɛ-ADP fluorescence and collisional quenching indicate that the structural changes induced by the two ligands in the cleft are different. The BeFx induced ɛ-ADP fluorescence intensity increase is consistent with the proposal of Combeau and Carlier 3,29 that BeFx is located at the place of γ-phosphate of ATP in the nucleotide-binding cleft of actin. The incomplete fluorescence intensity increase by cofilin in the presence of BeFx is presumably due to the residual bound BeFx inhibiting cofilin’s effect on the conformation of the nucleotide-binding cleft. This finding is consistent with the effect of cofilin on the extent of proteolysis of the subdomain 2 of F-actin in the presence and absence of BeFx.

The binding of cofilin to F-actin in the presence of BeFx was also monitored by the increase in proteolytic susceptibility of subdomain 2. Since cofilin increases while BeFx decreases proteolytic susceptibility, this method also measures BeFx dissociation. The results of proteolytic experiments show that the dissociation of BeFx is not complete even 20min after the addition of cofilin (Figure 4BC), when according to the fluorescence measurements F-actin is fully saturated with cofilin. This indicates that after 20min incubation with cofilin there are actin protomers to which BeFx and cofilin are simultaneously bound. The F-actin-bound residual BeFx inhibits considerably the proteolysis of subdomain 2 due to its strong cooperative effect on F-actin structure 4.

The binding of BeFx to F-actin is a relatively slow, two-step process 3. The first step is a fast equilibrium binding, which is followed by a slow isomerization step accompanied by a conformational change. We measured the rate of BeFx binding to F-actin by following the increase in the fluorescence intensity of ɛ-ADP-F-actin and found that it takes ∼5min for the fluorescence to reach the plateau after addition of BeFx (Figure 6B). About the same time is needed for F-actin to become fully resistant to proteolysis upon addition of BeFx (Figure 7C). However, cofilin binding experiments revealed that further changes occur in the structure of BeFx-F-actin, even after it became resistant to proteolysis. We showed that the binding of cofilin to BeFx-F-actin is also a two-step process consisting of a fast and a slow step. The amplitude of the fast step decreases with the time of incubation with BeFx; it is much smaller after 24h than 1h incubation with BeFx (Figure 7A). The ratio of the amplitudes of the fast and slow steps also depends on BeFx concentration; with the relative size of the first step decreasing with increasing BeFx concentration. We associate the fast and slow steps with the low and high affinity BeFx-F-actin complex, respectively. Since both complexes are resistant to proteolysis and their ɛ-ADP fluorescence is increased, we conclude that cofilin binding detects a second isomerization step in the BeFx-F-actin interaction, according to (Scheme 1):

(Scheme 1)

BeFx-F-actin is the fast equilibrium complex and BeFx-F-actin* and BeFx-F-actin** are the low- and high-affinity BeFx-F-actin complexes, respectively. A structural difference between the two complexes is suggested by the different degree of inhibition of cofilin binding. It appears that the same beryllium fluoride species binds to F-actin in the various BeFx-F-actin complexes, because the transformations between these complexes are not pH dependent.

It is interesting to compare the BeFx-F-actin complexes with the complexes of Pi-F-actin. According to Combeau and Carlier 3 there are two ADP-Pi-F actin complexes. Pi dissociates slowly from the ADP-Pi-F-actin* complex, which is a product of ATP hydrolysis accompanying actin polymerization, while it dissociates fast from the ADP-Pi-F actin complex, which is produced upon addition of Pi to ADP-F-actin. The two complexes are connected by an isomerization step, which limits the transformation of ADP-Pi-F-actin* to ADP-Pi-F actin according to (Scheme 2).

(Scheme 2)

The reversal of the isomerization step is extremely slow and thus, essentially only the ADP-Pi-F actin complex is obtained upon adding Pi to ADP-F-actin. We may assume that BeFx-F-actin** could be similar to the ATP-F-actin and BeFx-F-actin* to the ADP-Pi-F-actin* complex. This speculation is supported by the results of Combeau and Carlier 3 on the effect of BeFx on the fluorescence of pyrene-labeled F-actin. However, this hypothesis needs to be corroborated with further evidence.

Cofilin binding induces the dissociation of BeFx from F-actin, similarly to that of phalloidin 17. The removal of BeFx from actin indicates that cofilin causes allosteric conformational changes also in the nucleotide-binding cleft of F-actin where the BeFx is bound. Cofilin-induced conformational changes in the nucleotide cleft were indicated also by decreased phosphate affinity, changed fluorescence emission spectra, and decreased accessibility of F-actin-bound ɛ-ADP to collisional quenchers (Fig. 6 and Muhlrad et al. 25).

BeFx cannot bind to F-actin when the filaments are fully saturated with cofilin as shown by the inability of BeFx to reverse the cofilin-induced changes in the fluorescence intensity and proteolytic susceptibility. This indicates lower probability for the initial complex formation (BeFx-F-actin) and its reduced isomerization when the nucleotide-binding cleft of F-actin is allosterically changed by cofilin ((Scheme 1)). Cofilin probably inhibits the first isomerization step of BeFx, as the proteolytic susceptibility of the F-actin-cofilin complex remains high after the addition of BeFx. The effect of cofilin on BeFx binding is not cooperative; because the binding appears to be prevented only in those protomers of F-actin that are saturated with cofilin.

Phalloidin, unlike BeFx, can bind to cofilin-F-actin, which is fully saturated with cofilin. It competes with cofilin for F-actin and very slowly reverses the cofilin-induced disorder in the D-loop of subdomain 2 of F-actin. The different effect of BeFx and phalloidin on cofilin-F-actin is probably due to their binding to different sites on F-actin 3,7,8, and to the higher affinity of phalloidin to F-actin: (K_d=2.1 nM; 31) than BeFx (K_d=2μM; 3) to F-actin.

In contrast to Acanthamoeba actophorin, plant ADF, and human cofilin 19,13,20, yeast cofilin not only binds to F-actin in the presence of phalloidin or BeFx, but also facilitates their dissociation (Fig. 5; 17). The binding of cofilin to BeFx-F-actin revealed the existence of two types of BeFx-F-actin complexes, which transform to each other. The transformation of the weakly to the strongly bound complex is a very slow process, while the preceding step, of the initial complex formation, depends also on BeFx concentration. It is conceivable that the transformation of the complexes is accompanied by the movement of BeFx in the nucleotide binding cleft, which affects the conformation of the cleft and allosterically induces changes in the stability of the actin filament.

Overall, our studies resulted in several findings. By taking advantage of the high affinity of yeast cofilin to F-actin, which enables this cofilin to bind to F-actin also in the presence of BeFx or phalloidin, the antagonistic effects of the above factors were described quantitatively in this study. The existence of two BeFx-complexes with different stabilities was suggested by the pattern of cofilin binding to BeFx-F-actin. These complexes mimic different Pi-ADP-F-actin complexes, which exist during the hydrolysis of the F-actin bound ATP and upon the addition of Pi to ADP-F-actin. The study of the two BeFx-F-actin complexes may help to reveal the structure of functionally important Pi-F-actin adducts. Cofilin was found to affect the conformation of the nucleotide binding cleft of F-actin in addition to its influence on the structure of the DNase I binding loop and the 60–69 loop in subdomain 2. This effect of cofilin is manifested in the removal of BeFx from the nucleotide binding cleft, in the increase of F-actin bound ɛ-ADP fluorescence, a decrease in the accessibility of bound ɛ-ADP to collisional quencher, and the prevention of the tight BeFx binding. BeFx also increases the fluorescence intensity of the bound ɛ-ADP, which supports the earlier findings 3,29 that BeFx binds in the nucleotide cleft at the site of γ-phosphate of ATP. The detection of the unique effects of cofilin and the Pi analog, BeFx, on the actin structure will contribute to the understanding of the regulation of the cellular actin dynamics by AC-proteins and inorganic phosphate.