interaction of actin with phalloidin:: polymerization and stabilization of f-actin

8
Biochimica et Biophysica Acta, 400 (1975) 407~t14 © Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands BBA 37124 INTERACTION OF ACTIN WITH PHALLOIDIN: POLYMERIZATION AND STABILIZATION OF F-ACTIN P. DANCKER', I. LOWb, W. HASSELBACH" and TH. WIELAND b'* aMax-Planck-Institut fiir Medizinische Forschung, Abteilung Physiologie und bAbteilung Naturstoff- Chemie, D 69 Heidelberg, Jahnstr. 29 (G.F.R.) (Received February 26th, 1975) SUMMARY The cyclic peptide phalloidin, one of the toxic components of Amanita phal- loides prevented the drop of viscosity of F-actin solutions after the addition of 0.6 M KI and inhibited the ATP splitting of F-actin during sonic vibration. The data con- cerning ATP splitting are consistent with the assumption (a) that only 1 out of every 3 actin units of the filaments needs to be combined with phalloidin in order to suppress the contribution of these 3 actins to the ATPase activity of the filament and (b) that all actin units of the filaments can combine with phalloidin with a very high affinity. Phalloidin did not only stabilize the actin-actin bonds in the F-actin structure but it also increased the rate of polymerization of G-actin to F-actin. The ability of F-actin to activate myosin ATPase was not affected by phalloidin. The tropomyosin-troponin complex did not prevent the stabilizing effect of phalloidin on the F-actin structure. INTRODUCTION Recent studies from these laboratories [1-4] have shown that the intoxication of rats with the cyclic peptide phalloidin, a toxic component of the green deathcap toadstool Amanita phalloides causes the formation of thin filaments in liver cells which lie in the vicinity of the cell membrane. These filaments are actin-like micro- filaments as judged from their electronmicroscopic appearance and their ability to form with muscle heavy meromyosin the well-known arrowhead structure. These filaments were not disrupted by 0.6 M KI. Obviously phalloidin preserves the struc- ture of these filaments. This prompted the search for similar effects on isolated muscle actin polymers. First results of these studies which show that phall0idin inhibits the depolymerization of F-actin by KI [5] have already been published [4, 6]. The present study analyzes the way in which phalloidin interacts with rabbit skeletal muscle actin. It will be shown that phalloidin not only stabilizes the F-actin structure but also enhances the rate of polymerization of monomeric globular G-actin to polymeric filamenteous F-actin. * To whom correspondence should be addressed.

Upload: p-dancker

Post on 28-Aug-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Biochimica et Biophysica Acta, 400 (1975) 407~t14 © Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands

BBA 37124

INTERACTION OF ACTIN WITH PHALLOIDIN:

POLYMERIZATION AND STABILIZATION OF F-ACTIN

P. DANCKER', I. LOW b, W. HASSELBACH" and TH. WIELAND b'* aMax-Planck-Institut fiir Medizinische Forschung, Abteilung Physiologie und bAbteilung Naturstoff- Chemie, D 69 Heidelberg, Jahnstr. 29 (G.F.R.)

(Received February 26th, 1975)

SUMMARY

The cyclic peptide phalloidin, one of the toxic components of Amanita phal- loides prevented the drop of viscosity of F-actin solutions after the addition of 0.6 M KI and inhibited the ATP splitting of F-actin during sonic vibration. The data con- cerning ATP splitting are consistent with the assumption (a) that only 1 out of every 3 actin units of the filaments needs to be combined with phalloidin in order to suppress the contribution of these 3 actins to the ATPase activity of the filament and (b) that all actin units of the filaments can combine with phalloidin with a very high affinity. Phalloidin did not only stabilize the actin-actin bonds in the F-actin structure but it also increased the rate of polymerization of G-actin to F-actin. The ability of F-actin to activate myosin ATPase was not affected by phalloidin. The tropomyosin-troponin complex did not prevent the stabilizing effect of phalloidin on the F-actin structure.

INTRODUCTION

Recent studies from these laboratories [1-4] have shown that the intoxication of rats with the cyclic peptide phalloidin, a toxic component of the green deathcap toadstool Amanita phalloides causes the formation of thin filaments in liver cells which lie in the vicinity of the cell membrane. These filaments are actin-like micro- filaments as judged from their electronmicroscopic appearance and their ability to form with muscle heavy meromyosin the well-known arrowhead structure. These filaments were not disrupted by 0.6 M KI. Obviously phalloidin preserves the struc- ture of these filaments. This prompted the search for similar effects on isolated muscle actin polymers. First results of these studies which show that phall0idin inhibits the depolymerization of F-actin by KI [5] have already been published [4, 6]. The present study analyzes the way in which phalloidin interacts with rabbit skeletal muscle actin. It will be shown that phalloidin not only stabilizes the F-actin structure but also enhances the rate of polymerization of monomeric globular G-actin to polymeric filamenteous F-actin.

* To whom correspondence should be addressed.

408

METHODS

Preparation of material Actin has been prepared from rabbit skeletal muscle. The actin used for the

measurements of viscosity and ATP hydrolysis has been prepared according to [7] and was free of the regulatory proteins tropomyosin and troponin. For the determi- nation of the polymerization rates the water extract of the acetone actin powder (extracted for 20 min at 0 °C) was used immediately after extraction in order to ensure that the experiments started with true G-actin. ATP was added to the extract to a final concentration of 0.1 raM. Some of these actin solutions contained minor traces of regulatory proteins as seen by sodium dodecyl sulfate gel electrophoresis. Phalloidin was a gift of Dr A. Buku, who prepared it in our laboratory.

Measuremen ts Viscosity (2 ml samples) has been measured at 20-22 °C in a spiral capillary

viscosimeter. The flow time for water was 40 s. ~/Spec. has been defined as usual as (t/to) -- 1 with t ---- flow time of the sample and to -- flow time of the corresponding protein free medium. Sonic vibration was performed at 20 °C with a Branson sonifier B 12 using a microtip for 4.5 ml samples at 20 kc/s and 50 W. The inorganic phosphate produced during sonic vibration has been measured according to Marsh [8]. Acto- myosin ATPase was measured according to Dancker and Hoffmann [7].

The rate of polymerization of G-actin to F-actin has been observed by meas- uring (at 25 °C) the increase of light scattering intensity of the actin solution at 400 nm in a Hitachi-Perkin-Elmer Fluorescence Spectrophotometer MPF 2 A. The scattered light has been measured perpendicular to the incident light. The values of light scattering intensity indicated at the ordinates of the figures are deflections of the registration device and are hence arbitrary values depending on apparatus parameters like gain, slit width etc. Besides the constituents indicated in the figures the actin solutions used for the measurements of the polymerization rate always contained 0.1 mM ATP and 5 mM Tris. HC1 buffer (pH 8.0). Polymerization was always started by the addition of MgClz. In the absence of MgCIz no polymerization occurred during the time of measurement.

Protein concentrations were determined by a modified biuret method or by the Lowry method [9].

Molecular weights The following molecular weights have been used for the calculations in this

paper: 45 000 for actin and 800 for phalloidin.

R ES U LTS

The effect of phalloidin on the viscosity of F-actin (in the absence and presence of KI) and on the rate of polymerization of actin

Table l shows the viscosity of actin that has been incubated overnight in the different polymerizing media indicated. Only in the presence of Mg 2+ alone (without additional components) was the viscosity lower, but in the other cases (addition of nucleotides, phalloidin, or KC1) equal values of viscosity were reached in these

409

T A B L E I

Specific viscosity of F-actin (1.5 mg/ml) polymerized for 12 h in the media indicated (buffer concen- tration: 5 mM Tris-HC1, pH 8.0). The viscosity was measured either in the absence or in the presence of 0.6 M KI.

- -KI + K I

0.7 mM Mg 2÷ 0.70 0.03 0.7 mM Mg z+ 0.1 mM ADP 1.60 0.03 0.7 mM Mg 2÷ 0.1 mM ATP 1.62 0.03 0.7 mM Mg 2÷ 50/~M phalloidin 1.59 1.59 0.7 mM Mg 2+ 0.1 M KCI 1.62 0.03

12i B 10-

"C ~8-

g . j i

_c 2-

~d ~ rm

e..r.e~e-e"e"e"e

J t(m[n)

Fig. 1. Acceleration of polymerization of G-actin to F-actin by phalloidin. Polymerization was started immediately after the addition of phalloidin by adding MgCI~. The solutions (2 ml) con- tained 2.6 mg actin (58 nmol), 1 mM MgCI2, 0.1 mM ATP and the following concentrations of phalloidin: C)--©, no phalloidin; 0 - - -0 , 5 nmol phaUoidin; I - - I , 10 nmol phalloidin; A- -A, 60 nmol phalloidin; 0 - - 0 , 100 nmol phalloidin.

experiments. In order to guarantee that maximal viscosity is attained in every case it is necessary to add sufficient ADP or ATP (as has been done in some samples of the table) because sometimes in the presence of phalloidin or KC1 without ADP or ATP the viscosity did not reach its maximal value. Addition of 0.6 M KI reduced the high viscosity of F-actin nearly completely. In the presence of phalloidin, however, 0.6 M KI reduced the viscosity only very little (e.g. about 10~o in the experiment of Figure 2A) or even not at all (Table I).

410

1,2-

1.0-

0.~-

0.6- d

0.4-

0 . 2 -

A

X) 100-

~ 8 0 -

>~ 60"

4 0 ~

20-

1.0 0:2 0:4 o~ 0:8 I'~ 0:2 0'.4 0'.0 0:8 Amount o f added phalloidin (mol /mol ac t i n ) Amount of added phal lo id in(mol /mol act in)

Fig. 2. A. The preservation of high viscosity of F-actin in 0.6 M KI by increasing amounts of phal- loidin. F-actin pellets (partially polymerized in 0.7 mM MgCI2) were homogenized in 1 mM Tris- HCI buffer (pH 7.4) so that the actin concentration was 1.2 mg/ml. After 3 h ATP and MgClz were added to final concentrations of 0.15 mM and 0.7 mM, respectively. The amounts of phalloidin indicated in the figure were added and the whole solution was kept overnight at 4 °C. O, specific viscosity prior to the addition of KI; O, specific viscosity after the addition of 0.6 M KI. B. Inhibition by phalloidin of ATP splitting during sonic vibration of F-actin. Prior to the measurements the sam- ples were treated like those of Fig. 2A with the exception that they have stood overnight in 0.1 M KCI and 0.8 mM ATP. Phalloidin was added immediately before sonic vibration started. After 30 rain of sonic vibration an aliquot was deproteinized with 3.5 ~ trichloroacetic acid and the inorganic phosphate (P0 determined. The rate of ATP splitting was calculated from the increase of P~ content during the time of vibration. The rate is expressed as ~ of the rate measured in the absence of phal- loidin. The figure includes the results of two experiments with the following rates of ATP hydrolysis : 100 % (tool Pi/mol actin/30 rain) O, 6.2; II, 4.6. Drawn line represents the following function: A = 100 (1 p)3 with A -- rate of ATP hydrolysis (~) and p -- amount of added phalloidin (tool/tool actin). This line represents the rate of ATP hydrolysis which would prevail if this rate would be pro- portional to any group of 3 actin units free of phalloidin and if the total added phalloidin would be bound in a purely statistical manner to the actin filaments (each actin unit representing one binding site).

These results imply that in the presence of phalloidin the same degree of polymerizat ion has eventual ly been reached that could be reached by other means too (MgCI~ plus sufficient amounts of nucleotides or KCI). How far the velocity of polymerizat ion was influenced by phal loidin is demonstra ted in Fig. 1, which shows the increase of light scattering of the polymerizing actin solution. One can see that gradual increase of phal loidin gave rise to a gradual increase of the rate of polymer- ization. The maximal rate was reached not before phalloidin and actin were present in equimolar amounts . Addi t ion of further phalloidin still increased the rate, al though

only to a small extent. A similar quant i ta t ive relat ion between actin and phalloidin holds true for

the protect ion of F-act in by phal loidin against depolymerizat ion by KI. F rom Fig. 2A it can be seen that F-act in was completely protected against KI only when the phal- loidin and actin subuni ts were present in equimolar amounts , a l though a high degree

411

of protection was already reached when the molar ratio of phalloidin to actin was very small.

Since the interpretation of the viscosity of solutions of such large filaments like F-actin is difficult, it is more convenient to analyze quantitatively the interaction between actin and phalloidin by means of another indicator for changes of the F-actin structure. This will be described in the following section.

The effect of phalloidin on the ATP spfitting of F-actin during sonic vibration The structure of F-actin cannot only be disturbed by anions like iodide but

by mechanical influences too. Asakura [10] was the first to show that during sonic vibration F-actin in an ATP-containing solution splits ATP. Further studies [11-13] have shown that during ultrasonication the cations and nucleotides, which are bound to F-actin could be easily exchanged with the respective ions of the medium. This is remarkable since normally these ligands are "buried" in the F-actin structure, so that they rarely exchange with the medium [14, 15]. Obviously, sonic vibration gives rise to local loosenings of the F-actin structure, which allowed the exchange of bound ADP with ATP of the medium. The subsequent "healing" of the local perturbation induces the splitting of the terminal phosphate bond of ATP in a way which is similar to the ATP splitting during polymerization of G-actin. This loosening-healing cycle continues during the time of sonication.

Fig. 2B shows that the ATPase activity of F-actin during sonic vibration can be completely inhibited by phalloidin. The events which are involved in ATP splitting during sonic vibration are independent of divalent cations: ATP splitting also occurred in the presence of EDTA or ethylene-bis(oxyethylenenitrilo)-tetracetic acid (EGTA) (this has already been observed by B~ir~iny and Finkelman [11]). Likewise, phalloidin also inhibited ATP splitting in the presence of EDTA or EGTA.

Fig. 2B shows that the rate of ATP hydrolysis was diminished considerably when only a small proportion of the actin units could have been combined with phalloidin. The line drawn in Fig. 2B is derived from simple probabilistic consider- ations [16] assuming (a) that the probability for each actin unit of the F-actin fila- ments to become combined with phalloidin is equal to the molar ratio of added phalloidin to actin and (b) that the binding of phalloidin to one actin unit inhibits the contribution of three units to the ATPase activity of the filament.

The dependence of the polymerizing effect of phalloidin on the ionic conditions The results discussed so far have shown that phalloidin prevented the loosening

of the F-actin structure and accelerated the polymerization of G-actin to F-actin. The extent of acceleration was greatly dependent on the other constituents of the medium. This is shown in Fig. 3. As can .be seen from Fig. 3A there is practically no polymerization in the absence of Mg 2+ during the period of observation. Addition of Mg 2+ enhanced the velocity considerably. Increase of Mg z+ to 3 mM had no further effect although it did have an effect in the absence of phalloidin (not shown in the figure).

Whereas the velocity of polymerization in the presence of phalloidin was increased by Mg 2+ it was decreased by KI. Fig. 3B demonstrates that in the presence of phalloidin the rate of polymerization was considerably reduced by increasing concentrations of KI.

412

6-

1

n o M g 2 +

1 2 3 t ( m i n )

12"

c~10 -

6-

4 - c

2 -

0 . 1 M KI

• O , 2 M K

0 . 5 M I

i I i , , 1 2 3 4 5

t ( r a i n )

Fig. 3. A. Increase of polymerization rate of actin by MgCI2 in the presence of phalloidin. Phalloidin was added immediately before polymerization was started by the addition of MgClz. The solutions (2 ml) contained besides MgCI2 0.1 mM ATP, 5 mM Tris-HCl buffer (pH 8.0), 2.6 mg actin (58 nmol) and 60 nmol phalloidin. B. Decrease of polymerization rate by KI. KI has been added prior to the addition of both phalloidin and MgCI2. The light scattering intensity was already elevated in 0.1 M KI because polymerization had already begun to occur before the measurements have been started by the addition of MgCI2. At higher concentrations of KI no polymerization could be seen before the addition of MgCI2. Besides KI the solutions (2 ml) contained 2.6 mg actin (58 nmol), 125 nmol phalloidin, 1 mM MgClz, 0.1 mM ATP and 5 mM Tris .HC1 buffer (pH 8.0).

This an t agon i sm between K I and pha l lo id in can be shown in the reverse way too : in the presence o f 0.3 M K I there was no po lymer i za t ion at all when pha l lo id in was absent . Add i t i on o f phal lo id in , however, induced a marked polymer iza t ion .

F-actin-phalloidin interaction in the presence of tropomyosin-troponin In ver tebra te s t r ia ted muscle the thin f i laments are not only composed o f

F-ac t in but of the regula tory pro te ins t r opomyos in and t ropon in too [17-19]. These pro te ins are able to influence the mechanica l p roper t ies o f the thin f i laments [20]. Fig. 4 demons t ra t e s that F-ac t in in the F -ac t i n - t ropomyos in - t ropon in complex was equal ly well p ro tec ted by pha l lo id in agains t depo lymer iza t ion by K I as was pure actin. Only the increase of viscosi ty induced by the add i t ion of the regula tory pro te ins was abol i shed ( indicat ing the d issocia t ion o f t r o p o m y o s i n - t r o p o n i n f rom the actin fi laments).

Ne i the r the ac t iva t ion o f myosin ATPase by actin nor the calcium regula t ion o f the ATPase o f ac tomyos in which conta ined the regula tory pro te ins was affected by pha l lo id in .

DISCUSSION

The descr ibed results have shown that (a) pha l lo id in stabil izes F-ac t in against influences which tend to loosen the s t ructure o f the actin f i lament (KI , sonic v ibra t ion)

413

5 1 2 ~

O.4

1.0 210 TM-TP (rng/ml)

Fig. 4. Stabilization by pha|loidin of actin containing the regulatory proteins tropomyosin and troponin (TM-TP) against the depolymerization by 0.6 M KI. Solutions containing 1.3 mg/ml actin (29 ffM) 0.1 M KC1 and the concentrations of tropomyosin-troponin indicated at the abscissa have been prepared before phalloidin has been added to a concentration of 60 ffM. For measuring viscosity the solutions were diluted by adding 0.1 M KCI to the twofold volume. Tropomyosin-troponin has been prepared as the entire complex according to Dancker [23]. Full symbols, viscosity measured before the addition of KI; open symbols, viscosity measured after the addition of 0.6 M KI. O, ©, actin without phalloidin; II, D, actin in a solution containing phalloidin.

and (b) that phalloidin accelerates the rate of polymerization of G-actin to F-actin. Both results are obviously consequences of the binding of phalloidin to actin. That phalloidin can be bound by actin has been shown [21]. The relationship between the rate of ATP hydrolysis during sonic vibration of F-actin and the amount of added phalloidin (Fig. 2B) shows that it is reasonable to assume that the added phalloidin molecules are totally bound and are statistically distributed over the entire actin filaments. This interpretation of Fig. 2B (other interpretations may be possible too) implies that phalloidin possesses a very high affinity to actin and that it is equally well bound by each actin unit. From Fig. 2B it can be further concluded that at least three adjacent actin units must be able to respond to sonic vibration in order to catalyze ATP hydrolysis.

Also for stabilizing the actin filament against KI it was not necessary that each actin unit contained one phalloidin molecule (the curve of Fig. 2A is still more non- linear than that of Fig. 2B) but it seems to be sufficient that the filaments are stabilized only at some points. Again, the fact that maximal stabilization was reached only when phalloidin was present in an amount equimolar to that of actin suggests that phalloidin is statistically distributed over the actin filaments. The strong non-linear relation between added phalloidin and the viscosity in 0.6 M KI may further imply that it is sufficient to stabilize only some filaments by phalloidin in order to preserve the high viscosity of the entire solution.

The following aspect of the phalloidin action deserves particular consideration : KI had no (or only a small) effect on the viscosity of F-actin in the presence of phal- loidin but KI did have a retarding effect on the velocity of polymerization even in the presence of phalloidin, in other words, there was no complete "protection" by

414

pha l lo id in agains t K I dur ing po lymer iza t ion . One explana t ion for this hehaviour is to assume that , on the one hand, the b inding o f pha l lo id in increases the p robab i l i t y for fo rming an ac t in-ac t in bond in the growing actin f i lament and that , on the o ther hand, K I decreases this p robab i l i t y so tha t the ac tual veloci ty which can be observed results f rom both an tagonis t ic influences. One mus t fur ther assume that once such a bond has been fo rmed this bond in the presence o f pha l lo id in canno t be d is turbed by K1 or sonic vibra t ion. This implies tha t pha l lo id in not only faci l i tates the forming o f an act in-act in b o n d bu t also stabil izes the formed bonds.

In conclus ion: pha l lo id in influences the mutua l in teract ion of the single actin units dur ing fo rma t ion and preserva t ion of the actin po lyme r in qui te a specific way. U p to now noth ing can be said as to the molecu la r mechanism of the pha l lo id in act ion. Since pha l lo id in is a pept ide, one may speculate tha t the pha l lo id in-ac t in in terac t ion mimics a specific "a l los ter ic" p ro te in -p ro te in in terac t ion which may be re levant for the state o f non-muscu la r actin which may undergo changes between po lymer ized and unpo lymer ized states (in con t ras t to muscle actin which remains po lymer ized dur ing the whole life t ime o f the muscle cell, cf. also [22]). The inter- ac t ion o f pha l lo id in with act in differs f rom the hypothet ica l p ro te in -pro te in inter- ac t ion in tha t it fixes act in in the po lymer ized state thus d is turb ing a poss ible po lymer - i za t ion -depo lymer i za t ion cycle o f cell actin which may be vital at least for liver cells (cf. [4]).

REFERENCES

1 Govindan, V. M., Faulstich, H., Wieland, Th., Agostini, B. and Hasselbach, W. (1972) Natur- wissenschaften 59, 521-522

2 Govindan, V. M., Rohr, G., Wieland, Th. and Agostini, B. (1973) Hoppe-Seyer's Z. Physiol. Chem. 354, 1159-1161

3 Lrw, I., Lengsfeld, A. M. and Wieland, Th. (1974) Histochemistry 38, 253-258 4 Lengsfeld, A. M., Lrw, I., Wieland, Th., Dancker, P. and Hasselbach, W. (1974) Proc. Natl.

Acad. Sci. U.S. 71, 2803-2807 5 Guba, F. (1950) Nature 165, 439 6 Lrw, I. and Wieland, Th. (1974) FEBS Lett. 44, 340-343 7 Dancker, P. and Hoffmann, M. (1973) Z. Naturforsch. 28c, 401-421 8 Marsh, B. B. (1959) Biochim. Biophys. Acta 32, 357-361 9 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265

10 Asakura, S. (1961) Biochim. Biophys. Acta 52, 65-75 11 Bfir~iny, M. and Finkelman, F. (1962) Biochim. Biophys. Acta 63, 98-105 12 Asakura, S., Tanigushi, M. and Oosawa, F. (1963) Biochim. Biophys. Acta 74, 140-142 13 Kakol, I. and Weber, H. H. (1965) Z. Naturforsch. 20b, 977-989 14 Martonosi, A., Gouvea, M. and Gergely, J. (1960) J. Biol. Chem. 235, 1700-1703 15 Grubhofer, N. and Weber, H. H. (1961) Z. Naturforsch. 16b, 435 444 16 Asai, H. and Morales, M. F. (1965) Biochemistry 4, 830-838 17 Hanson, J. and Lowy, J. (1963) J. Mol. Biol. 6, 46-60 18 Ebashi, S. and Endo, M. (1968) Prog. Biophys. Mol. Biol. 18, 123-183 19 Hanson, J., Lednev, V., O'Brien, E. J. and Bennett, P. M. (1973) Cold Spring Harbor Syrup.

Quant. Biol. 37, 311-318 20 Oosawa, F., Fujima, S., Ishiwata, S. and Mihashi, K. (1973) Cold Spring Harbor Syrup. Quant.

Biol. 37, 277-285 21 Wieland, Th. and Govindan, V. M. (1974) FEBS Lett. 46, 351-353 22 Allison, A. C. (1973) Ciba Foundation Symp. 14 (new series) Associated Scientific Publishers,

Amsterdam, 109-143 23 Dancker, P. (1974) Z. Naturforsch. 29c, 496-505