Inorganic phosphate regulates the binding of cofilin to actin filaments Andras Muhlrad 1 , Dmitry Pavlov 2 , Y. Michael Peyser 1 and Emil Reisler 2 1 Institute of Dental Sciences, School of Dental Medicine, Hebrew University of Jerusalem, Israel 2 Department of Chemistry and Biochemistry and the Molecular Biology Institute, University of California, USA Actin dynamics, the polymerization and depolymeriza- tion of actin filaments and formation of ordered actin assemblies, is critical to many events of cell motility, including the movement of whole cells, cell division, vesicular transport and exo- and endocytosis. The essential processes of actin dynamics are closely regula- ted in the cell by a large number of actin binding pro- teins and small molecules. The large family of actin depolymerizing factor (ADF) ⁄ cofilin (AC) proteins [1] has a central role in regulating actin dynamics. These small proteins, which are ubiquitous in all eukaryotic cells, increase the depolymerization and nucleation of actin filaments and accelerate their treadmilling [2]. AC proteins accelerate the turnover of actin filaments by severing them [3–6], thereby increasing the number of the free pointed and barbed ends, or by increasing monomer dissociation from the pointed end of fila- ments [7], and ⁄ or by both processes [8–10]. The action of these proteins on actin is promoted in most cases by increasing pH [4], and is regulated by phosphatidyl inositides [11,12] and their phosphorylation (except for yeast cofilin) by kinases [13,14]. The binding of ADF ⁄ cofilin to F-actin induces large conformational changes in the structure of actin fila- ments, including changes in their mean twist [15], and the weakening of longitudinal contacts between proto- mers along t he lon g -pitch he lix [16,17] and lateral contacts between the two strands [18,19]. According to electron microscopy evidence, the weakening of longitudinal contacts in F-actin is due to conformational changes in subdomain 2 of actin, including the disordering of the DNase I binding loop (D-loop) [20]. Solution studies also showed cofilin induced conformational changes in actin, particularly in subdomain 2 and its D-loop. The fluorescence intensity of probes [tetramethyl rhodamine cadaverine (TRC) and dansyl diethylamine] attached to Gln41 on the D-loop decreased dramatically upon bind- ing of cofilin to F-actin [17,21]. Moreover, subtilisin Keywords actin; cofilin; collisional quenching; fluorescence; limited proteolysis Correspondence A. Muhlrad, Institute of Dental Sciences, School of Dental Medicine, Hebrew University of Jerusalem, PO Box 12272, Jerusalem 91120, Israel Fax: +972 2 675 8561 Tel: +972 2 675 7587 E-mail: muhlrad@cc.huji.ac.il (Received 3 January 2006, accepted 8 February 2006) doi:10.1111/j.1742-4658.2006.05169.x Inorganic phosphate (Pi) and cofilin ⁄ actin depolymerizing factor proteins have opposite effects on actin filament structure and dynamics. Pi stabilizes the subdomain 2 in F-actin and decreases the critical concentration for actin polymerization. Conversely, cofilin enhances disorder in subdomain 2, increases the critical concentration, and accelerates actin treadmilling. Here, we report that Pi inhibits the rate, but not the extent of cofilin binding to actin filaments. This inhibition is also significant at physiological concen- trations of Pi, and more pronounced at low pH. Cofilin prevents conforma- tional changes in F-actin induced by Pi, even at high Pi concentrations, probably because allosteric changes in the nucleotide cleft decrease the affinity of Pi to F-actin. Cofilin induced allosteric changes in the nucleotide cleft of F-actin are also indicated by an increase in fluorescence emission and a decrease in the accessibility of etheno-ADP to collisional quenchers. These changes transform the nucleotide cleft of F-actin to G-actin-like. Pi regulation of cofilin binding and the cofilin regulation of Pi binding to F-actin can be important aspects of actin based cell motility. Abbreviations AC, ADF ⁄ cofilin; ADF, actin depolymerizing factor; D-loop, DNase I binding loop; K sv ,K sv ¼ [(F 0 ⁄ F))1]Æquencher M )1 , where F 0 and F are fluorescence values in the presence and absence of quencher respectively; Pi, inorganic phosphate; TRC, tetramethyl rhodamine cadaverine. 1488 FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS cleavage of the D-loop between Met47 and Gly48 and the tryptic cut after Arg62 and Lys68 in the 60–69 loop of subdomain 2 became accelerated greatly upon addi- tion of cofilin [21], confirming significant changes in the subdomain 2 structure. The structure and dynamics of F-actin are also affected by the presence of phosphate (Pi) in the nuc- leotide binding cleft [22,23]. In ATP-containing solu- tions, under physiological conditions, G-actin contains tightly bound ATP in the nucleotide binding cleft. The bound ATP hydrolyzes to ADP and Pi subsequent to the polymerization of G-actin. The formed Pi slowly dissociates from the nucleotide cleft of the F-actin pro- tomers. The rate of dissociation is limited by an isome- rization step, which is accompanied by conformational changes. The structure of actin filament is highly dynamic, as there is continuous net association of monomers at the barbed end and a net dissociation at the pointed end, resulting in filament treadmilling. When ATP is present in the medium, there is always ATP or ADP–Pi in the cleft of actin protomers located in the vicinity of the barbed end of the filament, pro- ducing an ATP or ADP–Pi cap at this end. Because of the presence of this cap at the barbed end the critical concentration for polymerization of F-actin is low. F-ADP-actin can also be polymerized from ADP– G-actin in the absence of ATP. In this case all the proto- mers contain ADP and there is no cap at the barbed end. ADP–F-actin is less stable and has a high critical concentration for polymerization (reviewed in [24]). Pi from the medium can bind stoichiometrically to the nucleotide cleft of ADP–F-actin protomers and ADP– G-actin if the Pi concentration in the solution is high enough. The K d for Pi in ADP–F-actin protomers is 1.5 mm at pH 7.0 [23], while in G-ADP-actin the K d for Pi is an order of magnitude higher [25]. Because K d increases with increasing pH [23], it was concluded that the actin bound Pi species is H 2 PO 4 – [23]. Pi lowers sig- nificantly the critical concentration for polymerization of ADP-actin [26] by decreasing the rate of protomer dissociation at both filament ends, and in particular at the barbed end [23]. On the other hand, Pi has a negli- gible effect on the critical concentration for polymeriza- tion of F-actin in the presence of ATP, because of the protective ATP or ADP–Pi cap at the barbed end of the filament. However, Pi stabilizes F-actin structure both in the absence (ADP–F-actin) and presence of ATP (F-actin with ATP or ADP–Pi cap at the barbed end) in the medium. This is manifested in the ‘straightening’ of the actin filament [22], the stabilization of longitudinal bonds between protomers [27–29], the prevention of breakage and destruction of filaments upon exposure to SDS and potassium iodide (KI) [30], and the decreased tryptic susceptibility of the 60–69 loop in subdomain 2 [31]. Phosphate and its BeFx analog {BeFx represents BeF 3 – ÆH 2 O or BeF 2 (OH) – Æ(H 2 O) beryllium fluoride complexes [32]}, like other actin filament stabilizing factors (e.g., phalloidin, tropomyosin, etc.), are antag- onistic to ADF ⁄ cofilin, which destabilizes filaments. Carlier et al. [7] reported that plant ADF does not bind to BeFx containing F-actin and hypothesized that the binding of ADF to ATP or ADP–Pi containing F-actin protomers is also prevented. According to sev- eral reports [3,33,34], Pi inhibits the severing of actin filaments by the AC-protein Acanthamoeba actophorin, and slows, whilst BeFx completely prevents, the bind- ing of actophorin to actin filaments. In addition, acto- phorin was reported to promote the dissociation of Pi from freshly polymerized F-actin [34]. However, no attempt was made to study the effect of physiological phosphate concentrations on cofilin binding to F-actin at various pH values relevant to the regulation of AC proteins. In view of the important and antagonistic effects of ADF ⁄ cofilin and phosphate on actin dynamics, we examined here the binding of cofilin to F-actin and its effect on F-actin structure in the presence of various concentrations of Pi at pH 8.0 and 6.5. We also studied the effect of cofilin on the nucleotide cleft of F-actin. We used yeast cofilin because in its presence the critical con- centration for actin polymerization is relatively low (0.7 lm) even at pH 7.8 [7]. Thus, at high enough F-actin concentrations, its depolymerization does not significantly influence data analysis. To monitor cofilin binding, we used fluorescence and proteolysis methods. TRC-labeled F-actin shows a large fluorescence inten- sity decrease upon cofilin binding [21], while subdo- main 2 cleavage by subtilisin and trypsin is greatly accelerated by cofilin, independently of the increased fil- ament treadmilling [21]. We found here that the rate of cofilin binding is strongly inhibited by Pi even at physio- logical Pi concentrations. Cofilin greatly reduces the affinity of Pi to the nucleotide cleft because of conform- ational changes, which render the cleft of F-actin pro- tomers G-actin like. The mutual regulation of cofilin and Pi binding to F-actin is pertinent to the regulation of actin dynamics in the cell. Results Fluorescence measurements of the effect of Pi on the rate of cofilin binding to TRC–F-actin A convenient way to study the binding of cofilin to F-actin is using actin in which Gln41 is labeled with A. Muhlrad et al. Pi regulates cofilin binding to F-actin FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS 1489 TRC. The fluorescence intensity of TRC–F-actin is decreased > 70% upon binding of cofilin, while the fluorescence intensity of G-actin changes little upon addition of cofilin [21]. The binding of 5 lm cofilin to 4 lm TRC–F-actin in the presence of 2 mm and 30 mm Pi at pH 8.0 and 6.5 was measured by monitor- ing fluorescence changes in a stopped-flow fluorometer (Fig. 1). To saturate F-actin, we used 30 mm Pi because of its low affinity to F-actin [23,25]. To check the effect of physiological concentrations of Pi [35,36], we also assayed cofilin binding in the presence of 2mm Pi. The ionic strength of all reaction mixtures was adjusted with KCl. In the absence of Pi, the TRC fluorescence intensity decreased fast upon addition of cofilin (Fig. 1). The rate of cofilin binding depended on pH in the absence of Pi; it was about four-fold fas- ter at pH 6.5 than at pH 8.0 (Table 1). Both 30 and 2mm Pi significantly inhibited the rate but not the extent of cofilin binding, however, the inhibition was greater at the higher Pi concentration (Fig. 1). In the presence of Pi, unlike in its absence, the rate of cofilin binding was pH independent between pH 6.5 and 8.0 (Table 1). We also checked the effect of phosphate on the fluorescence emission spectrum of TRC–F-actin (data not shown). We found that except for a very small red shift, Pi essentially did not affect the spec- trum of the TRC moiety on the D-loop of F-actin. The incubation time (1–24 h) of TRC–F-actin with Pi had no effect on cofilin binding (data not shown). Effect of cofilin on Pi binding to F-actin and on the conformation of the nucleotide binding cleft Although Pi inhibited the rate of cofilin binding to F-actin, we could not detect any displacement of cofilin by even 30 mm Pi, using either TRC–F-actin fluores- cence or subtilisin digestion assays (data not shown). This indicates very low affinity of Pi to cofilin-occupied F-actin. It is plausible that cofilin induced changes in the nucleotide binding cleft are responsible for a reduction of Pi affinity to F-actin. To test whether such changes indeed occur in F-actin, we examined the fluorescence emission of etheno-ADP (e-ADP) on actin, with and without the bound cofilin. Figure 2A shows a 54% fluorescence increase upon binding of cofilin to F-actin, confirming nucleotide cleft perturbation. Addition of cofilin to G-actin increased slightly (6.3%) the fluores- 0.1 0.2 0.3 0.4 0.5 01020304050 Time (sec) Fluorescence (A.U.) A 30 mM Pi 2 m M Pi 0 m M Pi 01020304050 Time (sec) Fluorescence (A.U.) B 0 0.1 0.2 0.3 0.4 0.5 30 mM Pi 2 m M Pi 0 m M Pi Fig. 1. Stopped-flow fluorescence measurements of the effect of 30 m M Pi on the binding of cofilin to TRC–F-actin at pH 8.0 and 6.5. TRC–F-actin, 4 l M, was incubated with 30 mM NaPi or KPi in pH 8.0 or pH 6.5 F-buffer, respectively, for 1 h on ice. The ionic strength was equalized in all solutions with NaCl or KCl at pH 8.0 and 6.5, respectively. Cofilin (5.0 l M) was added and the time course of fluorescence intensity change was monitored in a stopped-flow instrument at 20 °C, as described in Experimental procedures. (A), pH 8.0; (B), pH 6.5. Table 1. Effect of Pi on the rate constants of cofilin binding to TRC–F-actin at pH 8.0 and 6.5. Data were obtained by fitting the binding curves in Fig. 1 to mono-exponential expression. All rates were normalized to the rate determined in the absence of Pi at the same pH. pH Pi concentration (m M) Rate constants of cofilin binding (s )1 ) Normalized rates (%) 8.0 0 0.1924 100 2 0.1261 65.5 30 0.0444 23.0 6.5 0 0.8676 100 2 0.1283 14.8 30 0.0438 5.0 Pi regulates cofilin binding to F-actin A. Muhlrad et al. 1490 FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS cence of e-ATP in the nucleotide cleft (data not shown). We also studied the effect of cofilin on the nitromethane quenching of e-ADP in F-actin (Fig. 2B). The K SV values obtained in the absence and presence of coflilin were 3.32 and 1.41, respectively, which indicates that the cofilin induced perturbation in the nucleotide bind- ing cleft decreases the accessibility of the F-actin bound nucleotide to collisional quenchers. Monitoring the Pi stabilization of the D-loop in subfragment 2 by subtilisin digestion It has been shown that BeFx Pi analogs protect sub- domain 2 of F-actin from digestion by trypsin and subtilisin and Pi inhibits the tryptic cleavage of the 60–69 loop of subdomain 2 [31]. Here, we show that Pi also inhibits strongly the subtilisin cleavage of the D-loop in F-actin at pH 8.0 (Fig. 3) and 6.5 (data not shown). This inhibition decreased with decreasing Pi concentration, but could be observed even at 2 mm Pi. The inhibitory effect of Pi on the subtilisin digestion indicates stabilization of the D-loop. Monitoring the binding of cofilin to F-actin by limited proteolysis The effect of Pi on the binding of cofilin to F-actin was also studied by subtilisin cleavage of the D-loop in subdomain 2. We showed previously that the digestion of subdomain 2 of F-actin by trypsin and subtilisin was greatly accelerated by cofilin [21]. We used the effects of cofilin and Pi on actin proteolysis to assay cofilin binding to F-actin in the presence of 30 mm Pi at pH 6.5 and 8.0. To this end, TRC– F-actin (10 lm) was digested with subtilisin in the presence and absence of 12 lm cofilin and 30 mm Pi (Fig. 4). The digestion was started 30 s after the 0 100000 200000 300000 A 380 420 460 500 Wavelength (nm) Fluorescence (A.U.) 8 µ M cofilin no cofilin B 01020304050 0.95 1.00 1.05 1.10 1.15 1.20 Nitromethane (m M) F o / F Fig. 2. Effect of cofilin on the fluorescence emission spectrum and quenching of e-ADP in the nucleotide cleft of F-actin. The fluores- cence spectrum (A) and quenching by nitromethane (B) of 8 l M e-ADP–F-actin in 2 mM MgCl 2 ,20mM PIPES, pH 6.5, were moni- tored as described in Experimental procedures. n, no cofilin; m, 8 l M cofilin. Fig. 3. Inhibition of subtilisin digestion of F-actin by Pi at pH 8.0. TRC–F-actin (10 l M) was incubated with 0–30 mM Pi at pH 8.0 for 10 min, and then digested with 100 lgÆmL )1 subtilisin for 60, 120 and 180 s. The ionic strength was equalized in all samples with KCl. The samples were run on SDS ⁄ PAGE and analyzed by densi- tometry. n,noPi;m,2m M Pi; .,5mM Pi; r,10mM Pi; d,20mM Pi; h,30mM Pi. A. Muhlrad et al. Pi regulates cofilin binding to F-actin FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS 1491 addition of cofilin. The ionic strength was equalized in all samples by adding KCl or NaCl. Cofilin greatly accelerated the rate of subtilisin cleavage both in the presence and absence of Pi at pH 6.5 and pH 8.0. At pH 8.0 the digestion in the presence of cofilin was equally fast with and without Pi, while at pH 6.5 the acceleration of the cleavage was lar- ger in the absence of Pi than in its presence. We also studied the effect of incubation time of Pi–F-actin with cofilin on the rate of subtilisin diges- tion at pH 8.0 and 6.5 (Fig. 5). At pH 8.0, after 3 min incubation with cofilin, the cleavage rate of Pi–F-actin became as fast as that of F-actin without Pi, and 20 s incubation with cofilin was enough to almost completely activate the subtilisin digestion (Fig. 5A). Similar results were obtained by trypsin digestion of Pi–F-actin in the presence of cofilin at pH 8.0 (data not shown). On the other hand, at pH 6.5 the subtilisin cleavage of Pi–F-actin was still inhibited after 20 s, and even 180 s, incubation with cofilin (Fig. 5B). The rate of cofilin-activated subtilisin cut of Pi–F-actin is faster at pH 8.0 than at pH 6.5, while the binding rates of cofilin at these pH values, as monitored by TRC fluorescence, are equal. Discussion Cofilin and phosphate are important regulators of actin dynamics in the cell. Their effects on the struc- ture of actin filaments are antagonistic to each other. Cofilin, and AC proteins in general, disorder the Fig. 4. Effect of cofilin on the subtilisin digestion of Pi-TRC–F-actin at pH 8.0 and 6.5. TRC–F-actin (10 l M) was incubated with 30 mM NaPi or KPi in pH 8.0 or pH 6.5 F-buffer, respectively, for 1 h on ice. The ionic strength was equalized in all solutions with NaCl or KCl at pH 8.0 and 6.5, respectively. After 30 s incubation with 12 l M cofilin, the actin was digested with 25 lgÆmL )1 subtilisin for 20, 50, 80 s at 22 °C. The samples were run on SDS ⁄ PAGE and analyzed by densitometry. Closed symbols, pH 6.5; open symbols, pH 8.0; circle, Pi only, no cofilin; downward triangle, no addition; diamond, Pi and cofilin; sqaure, cofilin only, no Pi. Fig. 5. Effect of incubation time with cofilin on the rate of subtilisin digestion of Pi–F-actin. TRC–F-actin (10 l M) was incubated with 30 m M NaPi or KPi in a pH 8.0 or pH 6.5 F-buffer, respectively, for 1 h on ice. The ionic strength was equalized in all solution with NaCl or KCl at pH 8.0 and 6.5, respectively. After 20 or 180 s incu- bation with 12 l M cofilin, actin was digested with 25 lgÆmL )1 sub- tilisin for 20, 50, 80 s at 22 °C in the pH 8.0 or pH 6.5 F-buffer, respectively. The samples were run on SDS ⁄ PAGE and analyzed by densitometry. (A), pH 8.0; (B) pH 6.5. ., Pi only, no cofilin; n,no addition; r, Pi and 20 s incubation with cofilin; d,Piand3min incubation with cofilin; m, 20 s incubation with cofilin only, no Pi. Pi regulates cofilin binding to F-actin A. Muhlrad et al. 1492 FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS filament structure by changing the conformation of subdomain 2, while phosphate has a stabilizing effect on F-actin structure. The antagonistic effects of Pi and AC proteins on F-actin are also indicated by increased dissociation of Pi from actin filaments in the presence of Acanthamoeba actophorin and inhibited binding of this AC protein at high, nonphysiological Pi concen- tration [34]. Here, we found that the rate but not the extent of yeast cofilin binding to F-actin is inhibited even at physiological concentrations of Pi at pH values 8.0 and 6.5, which promote and inhibit AC protein induced actin depolymerization, respectively [4]. In agreement with Ressad et al. [37], we observed that in the absence of Pi the binding of cofilin is faster at low pH than at high pH. This is in contrast to cofi- lin’s depolymerizing effect, which is stronger at high pH. However, according to our observations, the rate of cofilin binding to F-actin in the presence of 2–30 mm Pi is the same at pH 6.5 and 8.0. The influ- ence of Pi on the relative rates of cofilin binding at these two pH values can be explained by the finding that the Pi species that binds to F-actin is H 2 PO 4 – [23]. At pH 8.0 the main species of Pi is HPO 4 2– , while at pH 6.5 the H 2 PO 4 – species is dominant. Thus, at equal concentrations more Pi is bound to F-actin at pH 6.5 than at pH 8.0, yielding apparently greater inhibition of cofilin binding at pH 6.5 than at pH 8.0 (relative to the rate observed in the absence of Pi). However, another explanation of the stronger inhibition of cofi- lin binding at lower pH can be also suggested. Accord- ing to the molecular dynamics modeling of Wriggers & Schulten [38], Pi leaves the nucleotide cleft of F-actin through a back door mechanism. In this mechanism, His73, which is methylated with pKa ¼ 6.56, has a central role, because its positively charged, protonated form inhibits the release of the negatively charged Pi. Because the protonation of histidine decreases with increasing pH, it follows that the affinity of Pi to F-actin is lower at high pH. The pH dependence of the Pi inhibition of cofilin binding may contribute to the pH regulation of the actin depolymerizing activity of cofilin, which is important in view of the localized pH fluctuations in the cell. Subtilisin digestions of F-actin revealed that after 30 s incubation with cofilin at pH 8.0 the rate and the extent of actin cleavage was the same in the presence and absence of 30 mm Pi. On the other hand, at pH 6.5 the rate of F-actin cleavage was inhibited by Pi even after 3 min incubation with cofi- lin, suggesting incomplete displacement of Pi from actin by cofilin. The presence of residual Pi is not detected via TRC fluorescence data, because these are unchanged by Pi (data not shown) and monitor only cofilin binding to F-actin. The difference in subtilisin digestion of F-actin at the two pH values probably derives from the stronger binding of Pi at pH 6.5 than at pH 8.0 (see above). It is also possible that the effect of Pi (in the nucleotide cleft) on F-actin structure is cooperative, as is the case for the BeFx analog of Pi [31], and that the Pi coopera- tivity is higher at pH 6.5 than at pH 8.0. It is more difficult to detect any significant effect of Pi on the preformed F-actin–cofilin complexes. High concentra- tions of Pi (30 mm) cannot displace cofilin from F-actin, probably because of further weakening of the intrinsically weak binding of Pi to F-actin. This may be due to cofilin induced allosteric changes in the nucleotide binding cleft of actin. The cofilin induced increase in the fluorescence and a decrease in quenching of e-ADP (Fig. 2) also indicate nucleo- tide cleft perturbation in F-actin. The latter indicates that in the presence of cofilin the bound nucleotide in F-actin becomes less accessible and probably more buried in the F-actin structure. Notably, cofilin bind- ing has an opposite effect on ADP and Pi in the nucleotide binding cleft, i.e., it buries ADP and faci- litates Pi release from the cleft. We may speculate that the Pi release is promoted by opening of the ‘back door’ [38], while ADP becomes less accessible through the closing of the ‘main door’ on the top of the protomer or by some other mechanism. Similarly in G-actin, cofilin decreases the accessibility of the bound nucleotide, and inhibits nucleotide exchange [40] perhaps by closing the ‘main door’ of the nuc- leotide binding cleft. Taken together, cofilin binding transforms the nucleotide binding cleft of F-actin protomers to a more G-actin-like state, as indicated by the higher fluorescence and less accessibility to collisional quenchers of the bound e-ADP, which is characteristic of G-actin [40]. Moreover, the affinity of Pi to G-actin is an order of magnitude lower than to F-actin [25], similarly to cofilin containing F-actin. Although these results per se do not show weaker Pi binding to F-actin in the presence of cofi- lin, they document cofilin induced changes in the nucleotide cleft, which may be linked to lower Pi affinity to F-actin. The lowering of Pi affinity by cofilin has a physiological significance in the forma- tion of lamellipodia of moving cells because it facili- tates the binding and branching activity of Arp2 ⁄ 3, which can bind only to ADP–F-actin protomers on F-actin [41]. Pi, which stabilizes F-actin structure and decreases the critical concentration for actin polymerization, and cofilin, which has the opposite effect, are antagonistic to each other. This may be physiologically significant A. Muhlrad et al. Pi regulates cofilin binding to F-actin FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS 1493 because Pi concentration in the cell can approach 2mm [35,36], which particularly at low pH decreases the rate of cofilin binding to F-actin. On the other hand, cofilin accelerates Pi dissociation from F-actin [34] and prevents conformational changes that accom- pany Pi binding. Thus, Pi, cofilin, and pH can intri- cately regulate actin dynamics with a profound impact on the actin based cell motility. Experimental procedures Reagents TRC and e-ATP were obtained from Molecular Probes (Eugene, OR). ATP, trypsin, soybean trypsin inhibitor, sub- tilisin (Carlsberg), phenylmethylsulfonyl fluoride were pur- chased from Sigma Chemical Co. (St Louis, MO). Bacterial transglutaminase was a generous gift from K Seguro (Ajimoto Co., Inc., Kawasaki, Japan). Proteins G-actin was prepared from back and leg muscles of rabbit by the method of Spudich & Watt [42] and stored in G-buf- fer containing 5.0 mm Tris ⁄ HCl, 0.2 mm CaCl 2 , 0.2 mm ATP, 0.5 mm dithiotreitol, pH 8.0. F-actin was prepared from G-actin by polymerizing it with 2.0 mm MgCl 2 . Yeast cofilin was prepared as described previously [17]. The con- centrations of cofilin and unlabeled skeletal muscle a-actin were determined spectrophotometrically by using the extinc- tion coefficients E 1% 280 ¼ 9.2 and E 1% 290 ¼ 11.5 cm )1 , respectively. (The optical density of actin was measured in the presence of 0.5 m NaOH, which shifts the maximum of absorbance from 280 nm to 290 nm). Molecular masses were assumed to be 42 and 15.9 kDa for skeletal actin and yeast cofilin, respectively. Proteolysis Labeled or unlabeled F-actin (10 lm) was digested in the presence and absence of cofilin at pH 8.0 (in 20.0 mm Tris- HCl, 2.0 mm MgCl 2 , 0.2 mm ATP, 0.5 mm dithiotreitol) and pH 6.5 (in 20.0 mm PIPES, 2.0 mm MgCl 2 , 0.2 mm ATP, 0.5 mm dithiotreitol) by 25 lgÆmL )1 subtilisin or 800 lgÆmL )1 trypsin, respectively. The products of diges- tions were analyzed by SDS ⁄ PAGE. Protein bands on SDS gels were analyzed by densitometry. Chemical modification Actin labeled with TRC at Gln41 (TRC–actin) was pre- pared by incubating 50 lm skeletal G-actin with 100 lm TRC and 0.18 mgÆmL )1 bacterial transglutaminase in G-buffer pH 8.0, at 22 °C for 2 h. Reagent excess was removed on PD-10 filtration column (Amersham Pharmacia Biotech Inc., Piscataway, NJ) equilibrated with G-buffer. Preparation of e-ADP–F-actin ATP in skeletal muscle G-actin was substituted with e-ATP as follows. G-actin was passed through a desalting column (Amersham, PD10) of Sephadex G-25 equilibrated with ATP-free G-buffer. The eluted actin was supplemen- ted with 20-fold molar excess of e-ATP and was incubated for 1 h on ice. Excess e-ATP was removed from G-actin by passing it through another PD10 column. Actin was polymerized by addition of 2.0 mm MgCl 2 and during the polymerization the actin-bound e-ATP was hydrolyzed to e-ADP. Fluorescence measurements Fluorescence emission spectra were recorded in a PTI spectrofluorometer (Photon Technology Industries, South Brunswick, NJ), in G-buffer for G-actin, and in G-buffer containing 2.0 mm MgCl 2 for F-actin. The excitation wave- length for TRC and e-ADP was set at 544 and 350 nm, respectively. For quenching of e-ADP and time course of TRC fluorescence change the emission wavelength was set at 420 and 583 nm, respectively. The time course of TRC fluorescence change was monitored in an Applied Photo- physics (Leatherhead, Surrey, UK) SX-18 MV stopped-flow apparatus supplied with excitation and emission monochro- mators. Acknowledgements This work was supported by USPHS grant AR 20231 and NSF grant MCB 0316269 (to E.R.). References 1 Bamburg JR (1999) Proteins of the ADF ⁄ cofilin family: Essential regulators of actin dynamics. Annu Rev Cell Dev Biol 15, 185–230. 2 Lappalainen P & Drubin DG (1997) Cofilin promotes rapid actin filament turnover in vivo. Nature (London) 388, 78–82. 3 Maciver SK, Zot HG & Pollard TD (1991) Characteri- zation of actin filaments severing by actophorin from Acanthamoeba castellanii. J Cell Biol 115, 611–620. 4 Hawkins M, Pope B, Maciver SV, Brauweiler A & Weeds AG (1993) Human actin depolymerizing factor mediates a pH sensitive destruction of actin filaments. Biochemistry 32, 9985–9993. 5 Du J & Frieden C (1998) Kinetic studies on the effect of yeast cofilin on yeast actin polymerization. Biochem- istry 37, 13276–13284. Pi regulates cofilin binding to F-actin A. Muhlrad et al. 1494 FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS 6 Theriot JA (1997) Accelerating on a treadmill: ADF ⁄ cofilin promotes rapid actin filament turnover in the dynamic cytoskeleton. J Cell Biol 136, 1165–1168. 7 Carlier M-F, Laurent V, Santolini J, Melki R, Didry D, Xia G-X, Hong Y, Chua NH & Pantaloni D (1997) Actin depolymerizing factor (ADF ⁄ cofilin) enhances the rate of filament turnover: implication in actin-based motility. J Cell Biol 136, 1307–1323. 8 Moriyama K & Yahara I (1999) Two activities of cofilin, severing and accelerating directional depoly- merization of actin filaments, are affected differentially by mutations around the actin binding helix. EMBO J 18, 6752–6761. 9 Pope BJ, Gonsior SM, Yeoh S, McGough A & Weeds AG (2000) Uncoupling actin filament fragmentation by cofilin from increased subunit turnover. J Mol Biol 298, 649–661. 10 Yeoh S, Pope B, Mannherz HG & Weeds AG (2002) Determining the differences in actin binding by human ADF and cofilin. J Mol Biol 315, 911–925. 11 Yonezawa N, Nishida E, Iida K, Yahara I & Sakai H (1990) Inhibition of the interactions of cofilin, destrin, and deoxyribonuclease I with actin by phosphoinosi- tides. J Biol Chem 265, 8382–8386. 12 Kusano K, Abe H & Obinata T (1999) Detection of a sequence involved in actin-binding and phosphoinosi- tide-binding in the N-terminal side of cofilin. Mol Cell Biochem 190, 133–141. 13 Bamburg JR, Khatib FA & Bernstein BW (1984) Speci- ficity and regulation of brain actin depolymerizing fac- tor. J Cell Biochem 8b, 115. 14 Yang N, Higuchi O, Ohashi K, Nagata K, Wada A, Kangawa K, Nishida E & Mizuno K (1998) Cofilin phosphorylation by LIM-kinase 1 and its role in Rac-mediated actin reorganization. Nature 393, 809–812. 15 McGough A, Pope B, Chiu W & Weeds AG (1997) Cofilin changes the twist of F-actin: implication for actin filament dynamics and cellular function. J Cell Biol 138, 771–781. 16 Galkin VE, Orlova A, Lukoyanova N, Wriggers W & Egelman EH (2001) Actin depolymerizing factor stabi- lizes an existing state of F-actin and change the tilt of F-actin subunits. J Cell Biol 153, 75–86. 17 Bobkov AA, Muhlrad A, Kokabi K, Vorobiev S, Almo SC & Reisler E (2002) Structural effects of cofilin on the longitudinal contacts in F-actin. J Mol Biol 323, 739–750. 18 McGough A & Chiu W (1999) ADF ⁄ cofilin weakens lateral contacts in the actin filament. J Mol Biol 291, 513–519. 19 Bobkov AA, Muhlrad A, Shvetsov A, Benchaar S, Scoville D, Almo SC & Reisler E (2004) Cofilin (ADF) affects lateral contacts in F-actin. J Mol Biol 337 , 93–104. 20 Galkin VE, Orlova A, Van Loock M, Shvetsov A, Reisler E & Egelman EH (2003) ADF ⁄ cofilin use an intrinsic mode of F-actin instability to disrupt actin filaments. J Cell Biol 163, 1057–1066. 21 Muhlrad A, Kudryashov D, Peyser YM, Bobkov AA, Almo SC & Reisler E (2004) Cofilin Induced Conforma- tional Changes in F-actin Expose Subdomain 2 to Proteolysis. J Mol Biol 342, 1559–1567. 22 Nonomura Y, Katayama E & Ebashi S (1975) Effect of phosphate on the structure of actin filament. J Biochem (Tokyo) 78, 1101–1104. 23 Carlier M-F & Pantaloni D (1988) Binding of phos- phate to F-ADP-actin and role of F-ADP–Pi-actin in ATP-actin polymerization. J Biol Chem 263, 817–825. 24 Korn ED, Carlier M-F & Pantaloni D (1987) Actin polymerization and ATP hydrolysis. Science 238, 638–644. 25 Wanger M & Wegner A (1987) Binding of phosphate ions to actin. Biochim Biophys Acta 914, 105–113. 26 Rickard JE & Sheterline P (1986) Cytoplasmic concen- trations of inorganic phosphate affect the critical concentration for assembly of actin in the presence of cytochalasin D or ADP. J Mol Biol 191, 273–280. 27 Orlova A & Egelman EH (1992) Structural basis for the destabilization of F-actin by phosphate release following ATP hydrolysis. J Mol Biol 227 , 1043–1053. 28 Belmont LD, Orlova A, Drubin DG & Egelman EH (1999) A change in actin conformation associated with filament instability after Pi release. Proc Natl Acad Sci USA 96, 29–34. 29 Pollard TD, Goldberg I & Schwarz WH (1992) Nucleo- tide exchange, structure, and mechanical properties of filaments assembled from ATP-actin and ADP-actin. J Biol Chem 267, 20339–20345. 30 Dancker P & Fischer S (1989) Stabilization of actin filaments by ATP and inorganic phosphate. Z Natur- forsch [C] 44, 698–704. 31 Muhlrad A, Cheung P, Phan BC, Miller C & Reisler E (1994) Dynamic properties of actin: structural changes induced by beryllium fluoride. J Biol Chem 269, 11852– 11858. 32 Combeau C & Carlier M-F (1989) Characterization of the aluminum and beryllium fluoride species bound to F-actin and microtubules at the site of the c-phosphate of the nucleotide. J Biol Chem 264, 19017–19021. 33 Maciver SK & Pollard TD (1994) Actophorin preferen- tially binds monomeric ADP-actin over ATP-bound actin: consequences for cell locomotion. FEBS Lett 347, 251–256. 34 Blanchoin L & Pollard TD (1999) Mechanism of inter- action of Acanthamoeba actophorin (ADF ⁄ cofilin) with actin filaments. J Biol Chem 274, 15538–15546. 35 Burt CT, Glonek T & Barany M (1977) Analysis of living tissue by phosphorus-31 magnetic resonance. Science 195, 145–149. A. Muhlrad et al. Pi regulates cofilin binding to F-actin FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS 1495 36 Gillies RJ, Ogino T, Shulman RG & Ward DC (1982) 31P nuclear magnetic resonance evidence for the regula- tion of intracellular pH by Ehrlich ascites tumor cells. J Cell Biol 95, 24–28. 37 Ressad F, Didry D, Xia G-X, Chua N-H, Pantaloni D & Carlier M-F (1998) Kinetic analysis of the interaction of actin-depolymerizing factor (ADF) ⁄ cofilin with G- and F-actins. J Biol Chem 273, 20894–20902. 38 Wriggers W & Schulten K (1999) Investigating a back door mechanism of actin phosphate release by steered molecular dynamics. Proteins: Structure, Function Genet 35, 262–273. 39 Nishida E (1985) Opposite effects of cofilin and profilin from porcine brain on rate of exchange of actin-bound adenosine 5¢-triphosphate. Biochemistry 24, 1160–1164. 40 Root DD & Reisler E (1992) The accessibility of etheno-nucleotides to collisional quenchers and the nucleotide cleft in G- and F-actin. Protein Sci 1, 1014– 1022. 41 Pollard TD & Borisy GG (2003) Cellular motility driven by assembly and disassembly of actin filaments. Cell 112, 453–465. 42 Spudich JA & Watt S (1971) Regulation of skeletal muscle contraction. I. Biochemical studies of the inter- action of the tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. J Biol Chem 246, 4866–4876. Pi regulates cofilin binding to F-actin A. Muhlrad et al. 1496 FEBS Journal 273 (2006) 1488–1496 ª 2006 The Authors Journal compilation ª 2006 FEBS . measurements of the effect of Pi on the rate of cofilin binding to TRC–F -actin A convenient way to study the binding of cofilin to F -actin is using actin in. affinity of Pi to F -actin is lower at high pH. The pH dependence of the Pi inhibition of cofilin binding may contribute to the pH regulation of the actin depolymerizing