J Biol Chem, Vol. 274, Issue 49, 34637-34645, December 3, 1999

Tropomodulin Increases the Critical Concentration of Barbed End-capped Actin Filaments by Converting ADP·P_i-actin to ADP-actin at All Pointed Filament Ends^*

Annemarie Weber, Cynthia R. Pennise, and Velia M. Fowler^§¶

From the  Department of Biochemistry and Biophysics, University of Pennsylvania, Philadelphia, Pennsylvania 19104 and the ^§ Department of Cell Biology, The Scripps Research Institute, La Jolla, California 92037

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES

The pointed end capping protein, tropomodulin, increases the critical concentration of barbed end capped actin, i.e. it lowers the apparent affinity of pointed ends for actin monomers. We show here that this is due to the conversion of pointed end ADP·P_i-actin (low critical concentration) to ADP-actin (high critical concentration) when 70-98% of the ends are capped by tropomodulin. We propose that this is due to the low affinity of tropomodulin for pointed ends (K_d ~ 0.3 µM), which allows tropomodulin to rapidly exchange binding sites and transiently block access of actin monomers to all pointed ends. This leaves time for ATP hydrolysis and phosphate release to go to completion between successive monomer additions to the pointed end. When the affinity of tropomodulin for pointed ends was increased about 1000-fold by the presence of tropomyosin (K_d < 0.05 nM), capping of 95% of the ends by tropomodulin did not alter the critical concentration. However, the critical concentration did increase when the tropomodulin concentration was raised to the high values effective in the absence of tropomyosin. This may reflect transient tropomodulin binding to tropomyosin-free actin molecules at the pointed ends of the tropomyosin-actin filaments without a high affinity tropomodulin cap, i.e. the ends that determine the value of the actin critical concentration.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES

Tropomodulin is a ~40-kDa actin- and tropomyosin-binding protein that caps the pointed ends of actin filaments (1). Tropomodulin isoforms are associated with the actin cytoskeleton in a variety of post-mitotic, differentiated cell types in vertebrates, including striated muscle, erythrocytes, lens fiber cells, and neurons (2, 3). Recently, tropomodulin homologs have also been identified in flies (4) and in worms (2, 3) but not in yeast or fungi. In vertebrate striated muscle, tropomodulin is tightly bound to both tropomyosin and actin at the pointed ends of the thin filaments where it is believed to function to maintain thin filament length, sarcomere organization, and contractile function (1, 5). Although tropomodulin function has generally been considered in the context of its tight association with the stable, tropomyosin-actin filaments in striated muscle and erythrocytes, it is an open question as to whether tropomodulin could also regulate the assembly of more dynamic actin filaments in other contexts (6).

This possibility is suggested by the observation that tropomodulin, in addition to inhibiting elongation, increases the steady state monomer concentration of barbed end-capped actin 2-fold, close to the value for ADP-actin (7). This is not the result of monomer sequestration by tropomodulin, because tropomodulin binds exclusively to the pointed filament ends and not to actin monomers or alongside actin filaments (7, 8). Thus, tropomodulin must have increased the pointed end critical concentration (the critical concentration of barbed end-capped actin filaments).

The critical concentration of barbed end capped actin filaments represents the monomer concentration at steady state with the actin filament pointed ends and is the same after polymerization of actin monomers to filaments or after partial depolymerization of filaments (F-actin) to monomers (G-actin). The value of the pointed end critical concentration is determined by the affinity of the filament ends for actin monomers and is independent of the number of pointed filament ends or the total amount of F-actin (for a detailed treatment see Ref. 9). This means that the effect of tropomodulin on the pointed end critical concentration (7) must have been due to a decrease in the affinity of pointed ends for actin monomers and not to a decrease in the number of free pointed ends because of capping by tropomodulin. The aim of the present study was to determine how tropomodulin changes the affinity of pointed ends for monomers and increases the pointed end critical concentration.

The apparent affinity of filament ends for actin monomers depends on the nucleotide content of the actin molecules at the filament ends. As the result of ATP hydrolysis, newly incorporated ATP-actin molecules very rapidly become ADP·P_i-actin and then change to ADP-actin after the slow release of phosphate (for reviews see Refs. 10 and 11). The affinity of the filament ends for actin monomers is higher before than after phosphate release, thus the critical concentration is lower for ADP·P_i ends than for ADP ends. Here, we present data indicating that the tropomodulin-induced increase in the critical concentration depends on the conversion of pointed end ADP·P_i-actin to ADP-actin. To influence the critical concentration, this conversion must take place at the free pointed ends because tropomodulin-blocked filament ends are not at steady state with actin monomers. We propose that the relatively low affinity of tropomodulin for pointed ends allows tropomodulin to move rapidly from one pointed end to another and thus to bind transiently to all ends. This would have the effect of slowing down monomer addition at all ends, permitting ATP hydrolysis and phosphate release between monomer additions.

Consistent with this mechanism, we find that the critical concentration is not increased by high affinity binding of tropomodulin to tropomyosin-actin filament pointed ends. However, the steady state monomer concentration at the end point of polymerization is increased at high tropomodulin concentrations, the same concentrations that increase the critical concentration in the absence of tropomyosin. We attribute this to transient tropomodulin binding to tropomyosin-free actin molecules at the pointed ends of those tropomyosin-actin filaments that are not blocked by a high affinity tropomodulin cap and therefore are at steady state with G-actin. We explain the reasons why we expect most pointed ends of tropomyosin-actin filaments to terminate with one or more actin molecules that are not bound to tropomyosin.

In cells, tropomodulin might increase the critical concentration in regions of the cell where all barbed ends are tightly capped and the extent of tropomodulin-capping of the tropomyosin-free actin molecules at the pointed ends is between 70 and 98%. An increase in the critical concentration would be expected to increase the size of the sequestered actin monomer reservoir because the free actin monomer concentration is in equilibrium with monomers bound to sequestering proteins (12). This could play a role in the regulation of actin-based motility and/or the reorganization of the actin cytoskeleton during cell differentiation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES

Proteins-- Rabbit skeletal muscle actin was prepared from rabbit muscle acetone powder as described previously (13). Pyrenyl labeling of muscle actin was carried out according to Kouyama and Mihashi (14) with the modifications described previously (15). Actin was stored in liquid nitrogen and defrosted as described previously (13). ADP-actin was prepared in 50% sucrose by incubating 25 µM ATP-G-actin with 1.0 mM glucose and 0.05 mg/ml hexokinase and 50 µM P¹,P⁵-di(adenosine-5'-pentaphosphate) (myokinase inhibitor) and 2.0 mM ADP overnight. When the stock solution is diluted into the assays (about 10-fold), it is essential to have myokinase inhibitor present, otherwise there is some reconversion to ATP-actin. The ADP-actin had the same critical concentration in the absence and presence of a barbed end capper (gelsolin), indicating complete conversion of ATP-actin to ADP-actin. We prepared and stored ADP-actin in 50% sucrose because, as recently confirmed (16), Oosawa and co-workers (17) had shown that even nucleotide-free actin remained polymerizable in sucrose. The ADP-actin is quite stable in 50% sucrose with little loss in activity over several days. Interestingly, in the assays that contain 100 mM KCl and 2 mM MgCl₂ without sucrose, ADP-G-actin is stable overnight, although in low salt without sucrose ADP-G-actin deteriorates within hours (18). (Because of that, the conversion period from calcium-actin to magnesium-actin in low salt was shortened to 2 min for ADP-actin). The critical concentration of our ADP-actin varied between 1.2 and 1.5 µM. Recombinant chicken skeletal muscle tropomodulin was expressed in Escherichia coli and purified to homogeneity as described (19). Gelsolin, a generous gift from J. Bryan, was prepared as described previously (20). Tropomyosin was prepared according to Ref. 21 and stored as the lyophilized powder. Protein concentrations were determined for actin, gelsolin, tropomyosin, and tropomodulin by light absorption, using E₂₉₀ = 24.9 mM¹ cm¹, E₂₈₀ = 150 mM¹ cm¹, E₂₇₆ = 24 mM¹ cm¹, and E₂₈₀ = 14.7 mM¹ cm¹, respectively. Gelsolin concentrations were also followed by tryptophane fluorescence (290 nm) calibrated against the specific extinction.

Measurements of Elongation Rates-- Measurements of elongation rates were carried out as described previously (13), using pyrenyl-labeled actin (extent of labeling is indicated in the legends) and gelsolin-capped actin filaments as nuclei for polymerization (in the absence of tropomyosin, actin:gelsolin = 10-20:1 and in the presence of tropomyosin, actin:gelsolin = 150:1). Actin polymerization was followed continuously for 5-7 h and measured again the next morning. The fluorescence changes (excitation, 366.5; emission, 407 nm) were standardized against a Raman excitation peak and measured in a photon counting fluorimeter (Photon Technology International, Princeton, NJ). All experiments were carried out at 20 °C with Mg²⁺-actin (converted from Ca²⁺-actin as described previously (13)) in a medium containing 10 mM imidazole buffer, pH 7.0, 0.1 M KCl, 2 mM MgCl₂, 1 mM azide, 1 mM dithiothreitol, 0.5 mM ATP, and 0.1 mM CaCl₂ (polymerizing medium). Gelsolin-capped actin filaments serving as nuclei for polymerization were obtained by copolymerizing actin (usually 10 µM) with gelsolin in the presence of calcium. Average sizes for filaments are given in the legends as the ratios of actin:gelsolin. Because the accuracy of the rate constants depends on the accuracy of the nuclei concentration we checked the biological activity of our gelsolin preparation in low salt and the presence of calcium, by titrating gelsolin against increasing concentrations of pyrenyl-actin until the fluorescence reached a plateau, when gelsolin-actin dimer formation was complete. This usually occurred at the stoichiometry of two actin molecules per gelsolin, indicating that all of the gelsolin was active. For the preparation of tropomyosin-actin filaments, tropomyosin in excess over that necessary for filament saturation (excess of about 1.0 µM) was mixed with G-actin and, when present, tropomodulin (excess of 0.1-2.5 µM), before copolymerization was started by the addition of salt. In addition, 1.0 µM tropomyosin was added directly to the assay medium during elongation and depolymerization to ensure that actin filaments were always saturated with tropomyosin.

Measurements of the Critical Concentration-- The effect of tropomodulin on the critical concentration was measured by calculating the G-actin concentration at the end point of polymerization (see below for calculations). End points of polymerization were measured either at the end of the day or after overnight incubation to ensure completion of polymerization. (The 24-h end points could not always be used: sometimes the pointed ends were not stable and some of the F-actin depolymerized again overnight.) In some experiments, to minimize errors at very low critical concentrations or at the very low polymerization rates associated with a high extent of capping, the critical concentration was measured instead with the nullpoint method. For nullpoint measurements, actin filaments were mixed with G-actin in concentrations near the anticipated critical concentration to ascertain whether the G-actin polymerized ([G-actin] > critical), or the filaments depolymerized ([G-actin] < critical), or there was no change ([G-actin] = critical). For each tropomodulin concentration (and controls) about three to five measurements with different actin concentrations were made.

Calculations-- The concentrations of F-actin and G-actin were calculated according to Equation 1,

(Eq. 1)

where a, fluorescence of 1 µM G-actin and b, fluorescence of 1 µM F-actin, and [G-actin] = [total actin]  [F-actin]. The extent of capping, 1  (p_free/p_total), was calculated from the inhibited elongation rate, (c  c) k₊[p_free]/(c  c) k₊[p_total], where p_total is the total concentration of pointed ends (equivalent to the gelsolin concentration), p_free is the concentration of uncapped pointed ends, and c  c is the polymerizable G-actin concentration, i.e., [total actin]  c. In the absence of tropomyosin, we obtained the value for k₊[p_free], the pseudo first order rate constant (k'₊) for the rate of formation of F-actin, from the half-time for reaching the end point of polymerization, k'₊ = 0.7/t_1/2. Assuming that k₊ remains constant, the ratio of the pseudo first order rate constants is (0.7/t_1/2 control/0.7/t_1/2 tropomodulin) = p_free/p_total. At the low elongation rates associated with higher tropomodulin concentrations, when we could not be certain that the end point of elongation was reached on the same day, we calculated F-actin at the end point of polymerization according to ([F-actin] = [total actin]  c), using a value for c that was determined in separate assays by the null point method. If this is not done (as in our previous experiments (7)), the polymerizable G-actin concentration (c  c) is underestimated leading to a disproportionally high value for k'₊ at higher tropomodulin concentrations, compared with the value for p_free/p_total that can be calculated for these tropomodulin concentrations from the K_d value determined at low tropomodulin. This leads to the false conclusion of a leaky cap (7). In the presence of tropomyosin, the extent of capping was calculated by dividing the elongation rate in the presence of tropomodulin by the control rate measured over the identical range of c  c. The K_d of tropomodulin for the pointed ends was calculated according to K_d = [p_free] [tropomodulin_free]/[p-tropomodulin], where tropomodulin_free is the unbound tropomodulin, and p-tropomodulin is the pointed end-bound tropomodulin.

There are several caveats as regards our calculations. First, the assumption that the monomer on-rate at the pointed end (k₊) is constant during elongation is not quite correct in so far as k₊ depends on the nucleotide content of the pointed end actin. Because with increasing tropomodulin and decreasing rate of G-actin binding the fraction of pointed end ADP-actin increases and the fraction of ADP·P_i-actin decreases, the ratio k₊/k_+ control is probably a little lower than 1.0 at high tropomodulin concentrations, resulting in a slight overestimation of the extent of capping. Second, there is also an unavoidable small inaccuracy in the normalization of c  c. This is due to the fact that the value for the critical concentration that applies during net elongation is lower than the critical concentration at the end point of polymerization because the critical concentration also depends on the nucleotide content and decreases with increasing ADP·P_i-actin at the pointed ends. Thus, the ratio (c  c)/(c  c)_control that applies during elongation may not be identical to the ratio calculated using the critical concentration values at the end point of polymerization. However, the difference is probably within error of our measurements.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES

Pointed End Capping by Tropomodulin Is Associated with an Increase in the Critical Concentration of Tropomyosin-free Actin-- We showed previously that 4 µM tropomodulin increases the critical concentration of gelsolin-capped actin about 2-fold (7). The critical concentration of gelsolin-capped actin is entirely determined by the apparent K_d of the pointed ends for actin monomers because gelsolin completely blocks all barbed filament ends from interacting with the free actin monomers (the K_d of gelsolin for the barbed filament ends is in the picomolar range). We show now that the critical concentration of the gelsolin-capped actin filaments increases with increasing tropomodulin concentrations until it reaches a plateau at about 1.0-1.2 µM (Fig. 1A), a value close to the critical concentration of ADP-actin. The observation that a plateau is reached is additional confirmation of our previous data (7), showing that tropomodulin increases the steady state monomer concentration by increasing the concentration of free, unsequestered actin monomers (i.e. the critical concentration). By contrast, if tropomodulin had increased the steady state monomer pool by monomer sequestration, the monomer concentration would have continued to go up with increasing tropomodulin until all F-actin had been depolymerized.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effect of increasing tropomodulin concentrations in the presence of ATP on the critical concentration of gelsolin-capped actin (A) and the extent of pointed end capping (B). A, short actin filaments (actin:gelsolin = 10; 7 nM gelsolin) were mixed with 20% pyrenyl-ATP-G-actin (1.3-1.6 µM G-actin total) under polymerizing conditions in the presence of increasing tropomodulin concentrations as indicated. ATP-actin was prepared from ADP-actin by overnight incubation with 3 mM creatine phosphate and 0.1 mg/ml creatine phosphokinase. The critical concentration was calculated from the end point of polymerization measured after 24 h (closed circles) or 72 h (closed squares) (see "Experimental Procedures"). B, the ratio of free pointed ends to total pointed ends (pfree/ptotal) was obtained as described under "Experimental Procedures." In some of the experiments (open circles), the critical concentrations and the elongation rates were determined in separate assays and in others (closed circles), the critical concentrations were calculated from the end point of polymerization as for A. Closed circles, 1.3 µM (20% pyrenyl-actin) G-actin was added to 70 nM F-actin (actin:gelsolin 10:1); open circles, 2.5 µM (10% pyrenyl-actin) G-actin was added to 100 nM F-actin (actin:gelsolin 10:1). Tropomodulin was added directly to the assay medium.

Tropomodulin Capping Completely Blocks the Interactions between Pointed Ends and Free G-actin-- At a calculated value of 97% tropomodulin capping of pointed ends (based on the K_d value), we measured a 95-97% inhibition of the elongation rate (Fig. 1B). This is different from our earlier results when, limited by the amount of tropomodulin available to us, we used lower tropomodulin concentrations and obtained a value of 75-80% for maximal inhibition of the elongation rate by extrapolation to infinite tropomodulin concentration (7). We therefore concluded at that time that tropomodulin was a leaky cap, allowing interactions between fully capped pointed ends and free monomers to continue at a residual rate of 20-25% of the control. However, this conclusion was in error because we had underestimated the concentration of polymerizable actin (c  c) at higher tropomodulin concentrations, which led to a positive value for the elongation rate on extrapolation to 100% capping, rather than zero as expected for a blocking capper (for more details see "Calculations" under "Experimental Procedures"). In the more recent experiments shown here, complete blocking of the pointed ends by tropomodulin is strongly suggested by the fact that the observed inhibition of 95-97% of the elongation rate was within 2% of that expected from the calculated extent of capping.

The Critical Concentration Increase Requires the Presence of ATP-- A possible explanation for the tropomodulin-induced decrease in the affinity of the pointed ends for actin monomers (the basis for the increase in the critical concentration) would be a change in the bound nucleotide of the pointed end actin molecules from ADP·P_i to ADP (see "Introduction"). Therefore, we investigated the effect of tropomodulin on the critical concentration of actin filaments whose pointed end actin molecules had been converted to ADP-actin prior to the addition of tropomodulin. Tropomodulin did not increase the high critical concentration of ADP-actin (about 1.2 µM) any further (Fig. 2A), although ADP-actin pointed ends were capped to a similar extent as the pointed ends in the presence of ATP (compare Figs. 1B and 2B). The elimination of the critical concentration effect was completely reversible; replacing ADP with ATP (by incubation of the ADP-actin with creatine phosphate + creatine phosphokinase) restored the tropomodulin-induced increase in the critical concentration. This is shown by the experiment of Fig. 1A, which was done with actin that had been reconverted from ADP-actin to ATP-actin. These data support the idea that the tropomodulin-induced increase in the critical concentration is caused by the conversion of the pointed end actin molecules from ADP·P_i-actin to ADP-actin.

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Effect of increasing tropomodulin concentration on the critical concentration of gelsolin-capped ADP-actin (A) and the extent of pointed end capping of ADP-actin (B). A, the experiment and the calculations were carried out as in Fig. 1A, except that the filament concentration was higher (12 nM gelsolin; actin:gelsolin = 10; 20% pyrenyl-actin) and the ADP-G-actin concentrations varied between 1.3 and 2 µM. B, 2 µM ADP-G-actin (20% pyrenyl-actin) was added to 120 nM F-actin in ADP (actin:gelsolin 10:1).

The Critical Concentration Increase Requires the Release of Phosphate, the Second Step of ATP Hydrolysis-- If the tropomodulin-induced increase in the critical concentration is caused by the conversion of the pointed end actin molecules from ADP·P_i-actin to ADP-actin, then we might expect that inhibition of phosphate release would prevent the tropomodulin effect. Indeed, tropomodulin lost its ability to increase the critical concentration of ATP-actin by the addition of 100 mM inorganic phosphate (Table I), which prevents phosphate release after the hydrolysis step (22). Because the critical concentration in the presence of 100 mM inorganic phosphate is very low (Fig. 3), we measured the effect of tropomodulin on the critical concentration by a null point method (see "Experimental Procedures"). The pointed end critical concentration was not altered by 10 µM tropomodulin (Table I), which inhibited elongation by 85% under these conditions (data not shown). Note that in the absence of phosphate, this extent of inhibition is sufficiently high for a near maximal increase in the critical concentration (Fig. 1, compare A and B). This experiment suggests that tropomodulin can no longer exert its effect on the critical concentration when the shift from ADP·P_i-actin to ADP-actin at the pointed ends is blocked. Elimination of the tropomodulin-induced critical concentration increase by inhibiting phosphate release after ATP hydrolysis also indicates that the tropomodulin effect depends on phosphate release (ADP·P_i  ADP) and not on the hydrolytic step itself (ATP  ADP·P_i).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   A, effect of increasing phosphate concentrations on the barbed end critical concentration (uncapped filaments) and the pointed end critical concentration (gelsolin-capped filaments). B, extrapolation of the data in A to higher phosphate concentrations where the two critical concentrations converge to the same value. The critical concentrations were determined by the null point method using 50% pyrenyl actin. The ionic strength was maintained constant by the addition of decreasing concentrations of sulfate, pH 7.0.

The interpretation of these experiments is based on the consensus view that phosphate shifts the equilibrium from the ADP state to the ADP·P_i state by entering the nucleotide pocket of polymerized ADP-actin (10). This is probably true, but it has never been proved. For instance, it has not been shown that at high phosphate levels the critical concentration is the same for both filament ends, as it should be if the bound nucleotide is the same at both filament ends. In Carlier and Pantaloni's experiments (22), the critical concentration for the barbed ends was still about five times higher than that for the pointed ends at the inorganic phosphate concentrations used (75 mM, pH 7.4).

Therefore, we compared the critical concentrations of uncapped actin (close to the barbed end critical concentration) and gelsolin-capped actin (pointed end critical concentration) in 100 mM inorganic phosphate (Fig. 3). We found that the two critical concentrations were very close at 100 mM inorganic phosphate (Fig. 3A) and that the linear extrapolation of the last part of the curves converged to 0.018 µM actin in the presence of about 200 mM inorganic phosphate (Fig. 3B). These data strongly support the assumption that inorganic phosphate enters the nucleotide pocket of actin.

The data in Fig. 3B lead to another conclusion not related to pointed end capping. Because the decrease in the critical concentration of uncapped filaments on the addition of high inorganic phosphate reflects a decrease in the barbed end critical concentration from about 0.1 µM to about 0.018 µM, these data also show that the barbed ends do not entirely consist of ADP·P_i-actin (and possibly some ATP-actin) but that the barbed ends also contain some ADP-actin that can be converted to ADP·P_i-actin at saturating inorganic phosphate concentrations.

The Effect of Tropomodulin on the Critical Concentration of Tropomyosin-Actin Filaments-- Tropomodulin had a very similar effect on the pointed end critical concentration in the presence and absence of tropomyosin when the effects were compared at the same tropomodulin concentrations. Thus, concentrations of tropomodulin below 0.1 µM had no significant effect on the critical concentration in the presence (Table II) or absence (Fig. 1A) of tropomyosin, whereas 3 µM tropomodulin increased the critical concentration of tropomyosin-actin from 0.35 to 0.8 µM and from 0.5 to 0.9 µM in two independent experiments (data not shown). This was comparable with the effect of 3 µM tropomodulin on the pointed end critical concentration in the absence of tropomyosin (Fig. 1A).

However, the effect of tropomodulin on the critical concentration was quite dissimilar in the presence and absence of tropomyosin when the critical concentrations were compared as a function of tropomodulin saturation of the pointed ends. This is due to the large difference in the affinity of tropomodulin for tropomyosin-containing and tropomyosin-free pointed ends. In the absence of tropomyosin, the K_d of tropomodulin for the capping of pointed ends was ~0.2-0.3 µM (Fig. 1B), as shown previously (7). In contrast, we have now determined that the K_d of tropomodulin for tropomyosin-containing pointed ends is 0.05 nM or less; elongation of tropomyosin-actin filaments was maximally inhibited (~98% in Fig. 4) by 0.5 nM free tropomodulin, i.e. tropomodulin in excess over the concentration of pointed ends (taken to be equal to the gelsolin concentration). Increasing the free tropomodulin concentration to 9 nM had no further effect (Fig. 4). Thus, in contrast to pure actin filaments (Fig. 1), tropomodulin concentrations that maximally inhibit elongation of tropomyosin-actin filaments have no effect on the pointed end critical concentration (Fig. 4).

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Tropomodulin does not increase the pointed end critical concentration of tropomyosin-actin filaments at stoichiometric ratios of tropomodulin to pointed filament ends. Tropomodulin-tropomyosin-actin filaments (stock solution: 10 µM G-actin, 19% pyrenyl-actin, 2.0 µM tropomyosin, 72 nM tropomodulin, and 66 nM gelsolin) (actin:gelsolin = 150:1) were diluted 11-fold into 1.5 µM ATP-G-actin, containing 1.0 µM tropomyosin (final tropomodulin concentration, 6.6 nM (large open circles). Additional tropomodulin to bring the total concentrations to 7 nM (open triangles), 8 nM (open squares), and 15 nM (small open circles) was added directly to the assay. Closed circles, control for tropomyosin-actin filaments in the absence of tropomodulin. The elongation rate was not different than pure actin filaments (not shown). The numbers on the figure indicate the nanomolar concentrations of free tropomodulin (in excess over gelsolin, i.e. pointed ends). After about 75 min of incubation, inhibition of the elongation rate varied between 92 and >97% for all tropomodulin concentrations. Note that our previous experiments (7), which showed that tropomyosin inhibited the elongation rate at the pointed end, were due to a small amount of tropomodulin contamination in the tropomyosin preparations.

These experiments led to an additional conclusion not directly related to the effect of tropomodulin on the critical concentration. Because capping was almost complete at concentrations of tropomodulin stoichiometric to the concentration of pointed ends, the experiment in Fig. 4 also shows that only one tropomodulin molecule is necessary to completely block elongation from tropomyosin-actin pointed ends.

The large increase in the tropomodulin affinity for pointed filament ends on the addition of tropomyosin means first that at tropomodulin concentrations that were too low to produce an increase in the critical concentration (e.g. 60 nM tropomodulin; Table II), 95-98% of the pointed ends were actually tropomodulin-capped in the presence of tropomyosin, whereas they were virtually uncapped in the absence of tropomyosin. Second, although the increase in the critical concentration at micromolar tropomodulin in the absence of tropomyosin was associated with an increasing extent of pointed end capping (to 90-97%) (Fig. 1B), this could not have been the case in the presence of tropomyosin because capping was already 95-98% saturated at 9 nM tropomodulin (Fig. 4). Thus, it appears that the increase in the critical concentration observed at high tropomodulin concentrations in the presence of tropomyosin is not due to the tight capping of the tropomyosin-containing pointed ends by tropomodulin. This suggests the existence of a population of tropomyosin-actin filaments whose pointed ends have a low affinity for tropomodulin.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES

Tropomodulin Decreases the Monomer Affinity of the Pointed Ends by Converting All Ends to ADP-Actin-- The data presented in this study show that tropomodulin increases the pointed end critical concentration to a value very close to that of ADP-actin by converting ADP·P_i-actin at the pointed filament ends to the lower affinity ADP-actin. Thus, when added to ADP-actin, tropomodulin does not increase the critical concentration any further. Tropomodulin also has no effect on the critical concentration when the release of phosphate from ADP·P_i-actin is prevented by the inclusion of 100 mM inorganic phosphate. Furthermore, conversion of ADP·P_i-actin to ADP-actin is the natural consequence of pointed end capping because this slows down monomer binding and therefore allows phosphate release from ADP·P_i-actin to go to completion between successive monomer additions. Theoretically, a similar effect could also be achieved without inhibition of the rate of elongation by a binding protein capable of increasing the rate of phosphate release. However, modelling of our data (see "Appendix" and Fig. 7) suggests that tropomodulin does not affect the rate constant of phosphate release.

Tropomodulin Changes the Nucleotide Content of the Actin at the Uncapped Pointed Ends by Binding Transiently to All Pointed Ends and Slowing Down Monomer Addition-- To increase the critical concentration, tropomodulin must cause a change in the nucleotide content of actin molecules on pointed ends that are at steady state with G-actin. One possible mechanism would be leaky capping by tropomodulin, as we had proposed earlier (7), which would allow continuing monomer interaction with capped pointed ends that consist entirely of ADP-actin. However, the data presented here strongly suggest that tropomodulin forms a blocking cap at the pointed end. Therefore, tropomodulin must cause a change in the nucleotide content of actin molecules on pointed ends that are not capped, because only these and not the tropomodulin-blocked ends are at steady state with G-actin.

A mechanism that would explain the effect of tropomodulin on the uncapped pointed ends is suggested by the relatively low affinity of tropomodulin for the pointed ends (K_d ~ 0.3 µM). Low affinity implies that tropomodulin has a fast off-rate constant from pointed ends (k = K_d × k₊), assuming the on-rate constant (k₊) is high, e.g. diffusion limited. As a result, the dwelling time of tropomodulin on any individual pointed end is expected to be short, and it can move rapidly from one pointed end to the other, blocking each pointed end, but only for a short period of time. In other words, although a high affinity capper blocks some of the ends all of the time, a low affinity capper blocks all of the pointed ends some of the time. Thus, all pointed ends can bind actin monomers during the intervals after the departure of one tropomodulin molecule and before the arrival of another. With increasing tropomodulin saturation, the capping periods will become longer, giving each pointed end ADP·P_i-actin molecule more time to release its phosphate before another actin molecule has the chance to bind. The critical concentration is increased in the presence of tropomodulin because the only pointed ends that are available for binding actin monomers are pointed ends that have been converted from ADP·P_i-actin to ADP-actin.

In summary, we propose that capping by tropomodulin is equivalent to a decrease in the monomer on-rate constant for binding to the pointed ends, which decreases in proportion to the decreasing fraction of uncapped pointed ends. This view is supported by modelling of our data (see "Appendix" and Fig. 7). Between 70 and 97% saturation of pointed ends with tropomodulin, the observed increase in the critical concentration with increasing capping fits rather well a modelled curve plotting the critical concentration as a function of decreasing on-rate constants for the pointed ends (Fig. 7). Below 70% capping the data points start to fall below the modelled curve. This is understandable when one considers that transient tropomodulin capping is equivalent to a decrease in the rate constants of actin binding to pointed ends only if each pointed end has been capped at least once before the next actin monomer has a chance to bind. To achieve this at 70% capping, only half of the bound tropomodulin molecules need to dissociate and rebind elsewhere, whereas at 50 or 30% capping each tropomodulin molecule needs to dissociate and rebind elsewhere once or twice, respectively. When the saturation of the pointed ends with tropomodulin falls below 70%, the modelling of Fig. 7 indicates that a significant number of ends bind an actin monomer without first having been capped by a tropomodulin molecule. Thus at low extents of capping, the decrease in monomer on-rate constant (and increase in critical concentration) is not directly proportional to the extent of capping.

Tight Binding of Tropomodulin to Tropomyosin-Actin Filament Pointed Ends Does Not Lead to an Increase in the Critical Concentration-- Tropomodulin binds very tightly to tropomyosin-actin pointed ends because the interaction with the N-terminal end of tropomyosin (8, 19, 23) adds to the overall binding strength of tropomodulin for pointed ends, increasing the affinity more than 1000-fold over that for pointed ends without tropomyosin (K_d of less than 0.05 nM as compared with about 0.3 µM) (Fig. 4 and Ref. 7). Because each muscle tropomyosin rod spans seven actin molecules along the filament, tropomodulin can bind tightly to the filament end only when a tropomyosin rod extends to the tip of the pointed end so that tropomodulin fits precisely into its binding sites on both actin and tropomyosin (1, 5).

The high affinity of tropomodulin for tropomyosin-actin pointed ends can explain why 95-98% saturation of the pointed ends with tropomodulin had no effect on the critical concentration. Tightly bound tropomodulin remains bound to the same pointed end most of the time and cannot bind transiently and rapidly to all pointed ends. As a result, the ADP·P_i-actin monomers at the residual uncapped pointed ends (the pointed ends that are at steady state with G-actin) are not converted to ADP-actin prior to addition of another actin monomer, because actin monomers have access to the uncapped pointed ends at all times.

Considering the high affinity of tropomodulin for tropomyosin-actin pointed ends, it is surprising that a significant fraction of the pointed ends of tropomyosin-actin filaments remained uncapped (2-5% according to the residual elongation rate) even in the presence of 3 µM tropomodulin. This concentration of tropomodulin would have been expected to saturate more than 99.99% of the ends, based on the measured K_d of about 0.05 nM for tropomodulin binding to tropomyosin-actin filaments. It is conceivable that the uncapped ends are generated by spontaneous filament breakage or spontaneous actin nucleation. However, because both processes generate equal amounts of barbed and pointed ends, this would not explain why the concentration of free barbed ends was negligible in our experiments (indicated by a steady state monomer concentration of about 0.5-0.6 µM and by the sensitivity of the steady state G-actin concentration to tropomodulin); also see Ref. 7.

Another possibility is that a minor population of gelsolin-capped actin filaments that are less than seven subunits long (i.e. contain less than fourteen actin subunits) may co-exist with longer tropomyosin-actin filaments. These short filaments would be too short to accommodate a tropomyosin molecule and therefore would not be capped tightly by tropomodulin. Although the steady state length distribution of pure actin filaments is exponential (5, 6, 9), the length distribution of barbed end capped, tropomyosin-actin filaments was reported by Broschat (24) to be bimodal and somewhat shifted to longer lengths. Although Broschat did observe a significant fraction of tropomyosin-actin filaments ~20 monomers long, her technique was not designed to detect filaments less than 8-10 monomers in length, and thus the presence of very short filaments cannot be ruled out (24).

The Effect of High Concentrations of Tropomodulin on the Critical Concentration in the Presence of Tropomyosin Can Be Explained by the Mechanism of Tropomyosin-Actin Filament Polymerization-- We observed that high concentrations of tropomodulin (~3 µM) did lead to an increase in the pointed end critical concentration in the presence of tropomyosin, despite capping of 95-98% of the tropomyosin-actin pointed ends. We propose that this is due to tropomodulin binding to tropomyosin-free actin molecule(s) present at the tip of the uncapped pointed ends remaining at steady state after polymerization of tropomyosin-actin filaments (Fig. 5). These tropomyosin-free actin molecule(s) are a natural consequence of polymerization of tropomyosin-actin filaments in which seven actin molecules/strand must be assembled prior to binding of a tropomyosin molecule (Fig. 5) (24, 25). This means that net polymerization of tropomyosin-actin filaments is limited by the extent of net actin assembly, which ceases at the critical concentration for tropomyosin-free actin. Thus, at steady state, the pointed ends of tropomyosin-actin filaments terminate in tropomyosin-free actin molecules (Fig. 5, bottom filament). Transient binding of tropomodulin to these tropomyosin-free actin molecules would be expected to result in the conversion of their bound ADP·P_i to ADP as described above for pure actin, thus leading to an increase in the critical concentration.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Two-step elongation mechanism for pointed ends of tropomyosin-actin filaments. For simplicity, only one actin and tropomyosin strand is depicted. In the first step, actin monomers elongate from the pointed end to form a filament with a stretch of seven tropomyosin-free actin monomers. In the second step, a tropomyosin molecule binds to this bare stretch of seven actins. Filaments continue to elongate with this two-step process until the G-actin concentration falls to the critical concentration for pure actin. The top filament is an example of a filament in which the tropomyosin molecule extends to the tip, providing a tight binding site for tropomodulin. The second filament is an example of a filament ending in a bare actin stretch too short to bind tropomyosin. Bottom filament depicts the steady state configuration of the pointed ends with one tropomyosin-free actin monomer at steady state with the G-actin critical concentration.

This two-step mechanism for polymerization of tropomyosin-actin filaments (Fig. 5) can explain why it took such a long time for tropomodulin capping to reach its maximal extent during elongation (Fig. 4 and Ref. 7). Because at the low tropomodulin concentrations (nM) used in the elongation experiments of Fig. 4, tropomodulin binding is likely to be much slower (k₊ [tropomodulin] = 0.01-0.1 s¹) than is actin binding (k₊ [G-actin] = 0.5 -2 s¹), the rate of formation of these tight bonds is expected to be very slow. In fact, we have observed that the maximum extent of capping by tropomodulin was achieved sooner when the tropomodulin concentration was increased (7).¹

Conclusions-- First, every agent that lowers the on-rate of monomer binding to pointed ends (or barbed ends) without simultaneously lowering the rate of phosphate release promotes the conversion of pointed end ADP·P_i-actin to ADP-actin. Conversely, any agent that does the opposite and increases the on-rate of monomer binding to filament ends, e.g. profilin at the barbed ends, will maintain the actin molecules at the filament ends in the high affinity ADP·P_i state (26).

Second, conversion of ADP·P_i-actin to ADP-actin leads to an increase in the critical concentration only if the converted pointed ends continue to interact with the pool of free monomers. This interaction can occur either when the cap is leaky or when a blocking capper binds transiently because of its low affinity for pointed ends, as does tropomodulin. It will be interesting to know whether the very complex pointed end capper, Arp2/3 (27) increases the critical concentration by one of these mechanisms or whether it acts in still another way.

Third, in living cells, one would definitely expect a tropomodulin effect on the steady state monomer concentration in regions where the free tropomodulin concentration is 3 µM or higher and where all the actin filaments are tightly capped at their barbed ends and contain tropomyosin-free actin molecules at their pointed ends. These can be either actin filaments without any tropomyosin or bare actin subunits at the end of tropomyosin-actin filaments. The additional presence of tightly tropomodulin-capped tropomyosin-actin filaments without bare actin tails should have no effect on the steady state G-actin concentration.

	ACKNOWLEDGEMENTS

In the early phases of this work we profited from many valuable discussions with Martin Pring. We are also very grateful to all the colleagues who were of immense help with putting this paper together, discussing and clarifying concepts and giving assistance in improving the writing. We especially thank warmly Enrique De La Cruz, Vivianne Nachmias, Michael Ostap, Sally Zigmond, and Ryan Littlefield. We also acknowledge the expert technical assistance of Jeannette Moyer for purifying the recombinant tropomodulin.

	FOOTNOTES

^* This work was supported by National Institute of Health Research Grants GM53029 and HL15835 (to A. W. and the Pennsylvania Muscle Institute) and GM34225 (to V. M. F.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Cell Biology, MB24, Scripps Research Inst., 10550 N. Torrey Pines Rd., La Jolla, CA 92037. Tel.: 858-784-8277; Fax: 858-784-8753; E-mail: velia@ scripps.edu.

¹ A. Weber and V. M. Fowler, unpublished data.

	APPENDIX

What Is the Proportion of ADP·P_i-Actin and ADP-Actin at the Pointed Ends of Barbed End-capped Actin Filaments?-- Although we have discussed so far only ADP·P_i-actin and ADP-actin pointed ends as if actin formed a linear polymer rather than a double-stranded helix, there are actually four pointed end actin configurations that can be distinguished from each other by their content of ADP·P_i-actin and ADP-actin (Fig. 6). It is likely that these four types of pointed ends differ from each other in their rate constants of monomer binding and release because they offer four different binding sites for the associating as well as the dissociating actin molecules (indicated by the direction of the arrows in Fig. 6). When we refer to the conversion from ADP·P_i-actin to ADP-actin we have in mind the conversion of the different ADP·P_i-actin containing pointed ends (P₁ to P₃ in Fig. 6) to a uniform population of all-ADP-actin pointed ends (D in Fig. 6).

View larger version (37K): [in this window] [in a new window]   Fig. 6.   Diagram of four possible kinds of pointed ends with different combinations of ADP·Pi-actin and ADP-actin (D to P3). The nucleotide content of the three terminal actin molecules determines the affinity of pointed ends for monomers. The actin molecules poised for dissociation (hatched circles, outgoing arrows) are held in identical binding sites in D and P1, but in two different sites in P2 and P3. The binding sites for the incoming monomer (incoming arrows) are identical in P2 and P3 but different for each P1 and D. The nucleotide for the shaded actin monomers does not affect the monomer affinity of the pointed ends (bound ATP-monomers are not shown because their lifetime is too short).

The steady state distribution between these four kinds of ends depends on the relationship between the rate of monomer binding and the rate of phosphate release. The steady state distribution can be calculated from the observed pointed end critical concentration (c = 0.5-0.6 µM), the different monomer on-rate constants (k₊) and the rate constant of phosphate release (k_cat) according to the following equations.

(Eq. 2)

(Eq. 3)

(Eq. 4)

The correctness of the calculated distribution is checked by substituting the fraction of the total pointed ends (p_total) occupied by each of the four configurations (i.e. p₃/p_total, p₂/p_total etc) into the equation that defines the critical concentration in terms of the sum of the on-rates and off-rates of the pointed ends.

(Eq. 5)

The calculated distribution between the four types of pointed ends is verified as correct if the calculated critical concentration is equal to the value of the observed critical concentration, c_, used to calculate this distribution.

Using our measured values for some of the on-rate constants at pointed ends and some estimated values for others (Table III) together with a published value for the rate constant for phosphate release (k_cat = 0.005 s¹) (28), 98% of the pointed ends were calculated to consist entirely of ADP·P_i-actin with a corresponding critical concentration that was 30 times lower (0.02 µM) than the observed value (0.6 µM). This means that either the on-rate constants used for the calculation of the distribution were far too high or the rate constant for phosphate release far too low. The on-rate constants cannot be far too high because they have been measured directly for our actin preparations a number of times, using calibrated concentrations of gelsolin-capped filaments (see "Experimental Procedures"). Consequently, the rate constant of phosphate release at the pointed ends must be much higher than the rate constant of phosphate release at the barbed filament ends (29).

View this table: [in this window] [in a new window]   Table III Rate constants used to calculate critical concentrations and the distribution of ADP·Pi-actin and ADP-actin On-rate constants were calculated according to the expression k+ = k/c in those cases where a series of values for k and c showed less scatter than the measured on-rate constants. Constants without affix are measured values; the values for k+p3 and kp3 were measured in 100 mM inorganic phosphate.

Therefore, we adjusted the rate constant for phosphate release (arriving at a final value of 0.26 s¹) until the calculated value for the critical concentration matched the observed critical concentration (0.6 µM). According to these calculations 74% of the pointed ends consist of only ADP-actin, 15% contain one ADP·P_i-actin molecule, and 11% have 2 and 3 ADP·P_i-actin molecules.

Modelling the Critical Concentration Change as a Function of Decreasing Actin Monomer On-rate Constants at All Pointed Ends (Fig. 7)-- As discussed above, transient capping is expected to shift the distribution of the ends toward the all ADP state by lowering the effective monomer on-rate constants for all pointed ends. To test whether our data conform to this model, we calculated the effect of lowering the on-rate constants for actin binding on the nucleotide content of the pointed ends and on the associated critical concentration. With increasing capping {1  (p_free/p_total)}, the on-rate constants were decreased according to k₊/k_+ control = p_free/p_total. These values were entered into Equations 1-3 to obtain the distribution of the nucleotide configurations of the pointed end actin molecules. This then allowed the calculation of the critical concentration according to Equation 4. The data points for three different experiments fit the modelled curve fairly well at tropomodulin saturation of pointed ends above 70%. The reasons why the data points do not fit well at lower extents of tropomodulin saturation are given under "Discussion."

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Modelling the critical concentration assuming that tropomodulin acts by lowering the effective on-rate constants of actin monomers for all pointed ends. The solid line represents the modelled curve showing the increase in the critical concentration when the on-rate constants are decreased in inverse proportion to the observed extent of capping (upper abscissa) by a factor = k+/k+ control = pfree/ptotal. The relationship between decreasing rate constants and increasing extent of capping is indicated by the relationship between the upper and lower abscissa. The solid circles represent data points from three different experiments with different protein preparations, showing the critical concentration as a function of increasing capping.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION APPENDIX REFERENCES