Kinetics of actin depolymerization: influence of ions, temperature, age of F-actin, cytochalasin B and phalloidin

Biochimica et Biophysica Acta 873 (1986) 387-396 Elsevier

387

BBA 32630

Kinetics of actin depolymerization: influence o f ions, temperature, age o f F-actin, cytochalasin B and phalloidin Hanns Wendel and Peter Dancker lnstitut flir Zoologie, Technische Hochschule Darmstadt, Darmstadt (F.R.G.) (Received 13 June 1986)

Key words: Actin depolymerization; Cytochalasin B; Phalloidin; Depolymerization kinetics; (Rabbit skeletal muscle)

Actin, labelled with the fluorescent dye N-(3-pyrenyl)maleimide, was diluted below its critical concentration and depolymerization was followed by measuring the declining fluorescence intensity. The time courses of depolymerization were fitted to a sum of three exponentials. In most eases there was a fast initial phase followed by one or three slower ones. Increasing MgCI 2 concentration slowed down depolymerization velocity, as did substitution of Tris-maleate buffer by phosphate buffer. Older F-actin preparations depolymerized more slowly than younger ones. Phalloidin strongly decreased depolymerization velocity even after sonication. In the presence of eytochalasin B depolymerization was more uniformly exponential than in the absence of cytochalasin B; overall depolymerization velocity was decreased by cytochalasin B. The results are discussed on the assumption that depolymerization kinetics reflect the length distribution of aetin filaments during depolymerization.

Introduction The protein actin is not only the main part of thin filaments of muscle but also is a constituent of the contractile and cytoskeletal apparatus of all eucaryotic cells. For the biophysicist actin provides a simple model for the self-assembly process. It is the ability of monomeric G-actin to polymerize to filaments of F-actin and to depolymerize from F-actin to G-actin that makes it very suitable for its contractile and skeletal functions in the cell (for a comprehensive review see Ref. 1). The polymeric state of actin depends on the ionic conditions. The conditions in the cell are such that pure actin should remain permanently in the polymerized state. When (as is the case in the nonCorrespondence address: Dr. Peter Dancker, Institut fiir Zoologic, Technische Hochschule Darmstadt, Schnittspahnstrasse 10, D-6100 Darmstadt, F.R.G.

muscle eucaryotic cell) different physiological states of the cell are associated with either polymerized or depolymerized actin, one must assume that the state of actin is under the control of regulatory proteins. Many such proteins have been described in recent years [1-4]. One prerequisite for understanding the action of these proteins is to understand the mechanism of polymerization and depolymerization of pure actin. Whereas many studies have been conducted on the polymerization kinetics of actin (for review see Refs. 1, 5, 6), relatively little is known about the kinetics and mechanism of actin depolymerization. The reason for this may be that only very recently has the introduction of fluorescent pyrene dyes [7] provided a convenient tool for studying actin depolymerization. The strategy for measuring depolymerization is to dilute polymerized actin below the 'critical concentration' (that is the concentration below which actin exists only as mono-

0167-4838/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

388 mers; see Oosawa and Kasai [8] for a classical review). The fluorescent signal (which is much higher in polymerized actin than in depolymerized) is still strong enough to be measurable in strongly diluted solutions. Depolymerization is followed by the decay of the fluorescent signal in the depolymerizing actin. In the present work we studied actin depolymerization under various ionic conditions and we show that the kinetics are rather complex. The interpretation of our results is based on the hypothesis that the complex kinetics reflect the length distribution of filaments. This theoretical approach has been introduced for the depolymerization kinetics of microtubules by Johnson and Borisy [9], and has been extended by Karr et al. [10] and Kristofferson et al. [11]. Methods P r e p a r a t i o n o f actin. The preparation of actin acetone powder (from mixed skeletal muscle of rabbit) and the extraction of the acetone powder were performed essentially as described by Spudich and Watt [12] and Pardee and Spudich [13]. Actin was extracted with 2 mM Tris-HC1, p H 8.0, 0.1 mM ATP, 0.2 mM CaCI 2. After polymerization the extract was dialyzed (for depolymerization) against the extraction buffer and gel-filtered through Sephadex G-150 [14]. P r e p a r a t i o n o f p y r e n e - l a b e l l e d actin. The procedure was similar to that described by Kouyama and Mihashi [7]. As a pyrene dye we used N-(3pyrene)maleimide. Actin acetone powder was extracted with 2 mM Tris-HC1 p H 8.0, 0.1 mM ATP, 0.1 mM CaC12, 1.5 mM NaN 3. The extract was diluted to 1 m g / m l actin and polymerized for 2 h at room temperature by the addition of 2 mM MgC12. Then the appropriate volume of pyrene maleimide stock solution (10 mM in acetone) was added to give a final concentration of 2 . 1 0 - 5 M. After 30 rain at room temperature the assay was stored overnight in the refrigerator. The next morning the polymerized pyrene-actin was centrifuged for 2 h in a preparative ultracentrifuge (Beckman 50 Ti Rotor at 40000 rpm), homogenized (with a Teflon homogenizer) and dialyzed for 2-3 days against the extraction buffer. After depolymerization pyrene-actin was ultra-

centrifuged once more as described to remove polymerized actin and contaminations. A G-actin stock solution (1 m g / m l ) in extraction buffer (approximately 20% pyrene-actin and 80% unlabelled actin) was divided into two parts: one part was polymerized in 2 mM MgC12, the other part was left as unpolymerized G-actin and served as reference in the experiments. The experiments were performed in a Shimadzu R F 520 dual-beam fluorescence spectrophotometer. A small amount (between 10 and 60 #g) of F-actin was put in the measuring cuvette and by adding 2 ml depolymerizing buffer (essentially extraction buffer, variations are indicated in the legends), the whole solution (with an actin concentration of 0.12-0.72/~M) was mixed. The reference cuvette contained the same concentration of unpolymerized G-actin. Registration was started when the depolymerizing buffer was added to the measuring cuvette. Care was taken to perform the dilutions as identically as possible. All experiments were conducted several times and only for those sets of measurements which gave reproducible results are representative examples given in the figures and tables. The instrument measures the fluorescence differences between the two cuvettes. It can easily be shown that this difference is proportional to the concentration of polymerized actin (F-actin) in the measuring cuvette. Let A be the concentration of total actin in each cuvette. In the reference cuvette all the actin is G-actin. In the measuring cuvette the sum G t + F t always equals A ( G t and F~ are the concentrations of G- and F-actin at time t). a is a factor by which G-actin fluoresces less than F-actin. Thus the fluorescence signal ft recorded by the instrument is, at every time t: L = ~ + aG, - aA

with F~ = A - G , and otA = fluorescence of the reference cuvette. ft = A - G~ + aGt - aA f,~A(l-a)-G,(l-a) I,-- ( I - ( , ) ( A - C , )

/,= ( 1 - ~ ) < Since (1 -- a) is constant, f, is proportional to the F-actin concentration.

389

The temperature of the cuvettes was adjusted to 20°C by thermally regulated circulating water; excitation and emission wavelengths wore 365 nm and 410 nm, respectively; the slit in the excitation beam was minimal (5 nm) to avoid photobleaching, and the slit in the emission beam was 10 nm. Protein concentration was determined f r o m optical absorption at 290 nm using an absorption coefficient of 26 600 M - ] [15]. Analysis of the curves. The time course of depolymerization was recorded with a chart recorder connected to the fluorometer. In each experiment the depolymerization kinetics under different conditions were compared. Each figure is representative of the results of at least five experiments. At first sight the curves resemble an exponential decay but closer inspection reveals that they could not generally be described by only one exponential; however, a fit to the sum of three exponentials gave a sufficiently exact description of the curves (see Fig. 1). The data of some of the presented experiments were treated in the following way: the curves from the recorder were digitized (with a H I P A D TM Digitizer, Houston Instruments). The values were then fitted (by the 'least-squares method') to a sum of three exponentials: ft = C1e-*'t + C2e-*:t + C3e-k3'

ft is the fluorescence intensity at time t, kl, k2

and k 3 are descriptive constants with the dimensions of a rate constant and C 1, C2, C3 are also constants. The sum of C1, C2 and C3 corresponds to the initial fluorescence at t = 0. Slight deviations from the actually recorded fluorescence intensity are due to the fitting procedure. As an example of the fitting procedure digitized data from the fluorescence traces of Fig. 3 are compared with the function derived by the fitting procedure (Fig. 1). When the decay of fluorescence intensity corresponded to a single exponential, the last two terms of the above expression disappeared. In most instances the recorder curves could not adequately be described by only one exponential. The significance of the second derivative (including the meaning of C') of the above formula is given in the Discussion. Results

Depolymerization in different ionic mifieus Fig. 2 shows the depolymerization of actin after dilution to 5 # g / m l ( = 12 #M) in three concentrations of MgCI 2. In all cases there was a fast initial phase followed by a slow one. In order to discriminate better between the single phases of the presented experiments the numerical values describing the fit of particular experiments are tabulated (see Table I for the experiment of Fig. 2). In the absence of Mg 2+, the curve can be fitted by two exponentials, in other cases a reasonable fit is

90-

90.

>,

g~ c

60-

6o.

~ c

gc ~,. ~g ao-

~

g

-~

30.

u.

ff

÷

o

io

2'o Time

3'0

[rain)

Fig. 1. Comparison of the digitized fluorescence data (symbols) of Fig. 3 with the exponential functions derived from the fitting procedure (drawn line). Upper curve: depolymerization of actin in the presence of phosphate. Lower curve: depolymerization in the presence of Tris-maleate. For further details see Fig. 3.

Time (min}

Fig. 2. Depolymerization of actin in three different MgCI 2 concentrations. Depolymerization was initiated by diluting Factin to 5 # g / m l = 0.12 # M into depolymerizing buffer. The fluorescence signal is proportional to the F-actin concentration in the sample cuvette.

390 TABLE I PARAMETERS OF THE DEPOLYMERIZATION CURVE OF Fig. 2 A N D ITS S E C O N D D E R I V A T I V E ( W H I C H GIVES T H E L E N G T H D I S T R I B U T I O N OF F I L A M E N T S ; SEE DISCUSSION) The values of C or C ' indicate the extent to which the three exponentials contribute to the overall form of the depolymerization kinetics (C) or to the length distribution (C'). C is measured in the same arbitrary units of fluorescence intensity as in Fig. 2. The units of C' are also arbitrary and derived from two-fold differentiation. The dimension of k is m i n - L 0 m M MgCI 2 C1 C2 C3

59.4 36.4 0

1 m M MgCI 2

2 m M MgCI 2

34.0 11.5 41.2

34.3 10.5 49.2

kI k2 k3

1.52 0.07 0

1.14 0.27 0.02

0.71 0.11 0

C{ C~

137 0.18

44 0.84

17 0.13

c;

o

0.02

zation of actin: in the presence of 10 mM phosphate, 0.36 FM actin depolymerized much more slowly than in the presence of 10 mM Tris-maleate. This is consistent with the observation of Nonomura et al. [16] that in the presence of inorganic phosphate actin filaments are much more regular and straight than in the presence of Tris buffer. Table II shows that in the presence of phosphate the kinetics deviate less from a single exponential (only two C values) than in the presence of Trismaleate. This may be due to the specific molecular structure of the phosphate anion, but could equally well be due to the higher ionic strength effect of phosphate as compared to Tris.

Influence of the age of F-actin preparation

o

possible only with three exponentials. The time required for a half-maximal fluorescence decay increased with increasing MgC12 concentration. Qualitatively similar pictures can be obtained (not shown) when depolymerization is measured at various concentrations of KCI. The experiment of Fig. 3 shows that not only cations but also anions influence the depolymeri-

From the experiment in Fig. 4 it is obvious that the age of the F-actin preparation profoundly influences the depolymerization behaviour. The older the preparation, the slower the depolymerization and the less pronounced the difference between the fast and the slow phase, so that eventually this difference tends to disappear. In the 6-day-old actin (see Table III) the curve-fitting procedure no longer distinguished between C~ and C2 and took them both together to a new value of C~. C2 in Table III with a k 2 of zero shows that after an exponential decay of two-thirds of the 6-day-old polymerized actin there was one-third left which decayed only very slowly (see Discussion).

90-

T A B L E II

~.

~3

60 -

30-

PARAMETERS OF THE DEPOLYMERIZATION CURVE O F Fig. 3 A N D T H E L E N G T H D I S T R I B U T I O N D E R I V E D F R O M IT

phate

See legend to Table I for further details.

C1 C2 C3

Tris- m a l e a t e

0

1'o

2'0

3'o

T i m e (rain)

Fig. 3. Comparison of actin dcpolymerization in 10 m M phosphate buffer, p H 7.0, and 10 m M Tris-maleate buffer, p H 7.0. Dcpolymerization was initiated by diluting F-actin to 15 /~g/ml = 0.36/~M into depolymerizing buffer which instead of T r i s - H O contained either phosphate or Tris-maleate buffer.

kI k2 k3

10 m M phosphate

10 Tris-maleate

43.5 39.5 0

37.5 36.0 22.0

0.17 0.01 0

2.89 0.48 0.03

C{

1.26

3.13

c~ c;

0 0

8.3 0.03

391 90-

900.4 }JM

PHD

60-

m ~

-_='~

m D m m

3d

u ~3 3 0 o 0)

2(I ld

o

;o

~'o Time

do

(rain}

0

Fig. 4. Depolymerization of F-actin of various ages. G-actin was polymerized in 2 m M MgCI 2 for 30 rain at room temperature and was then put into the refrigerator. After the times (days) indicated 50 Fg actin were withdrawn and diluted to 25 ttg/ml = 0.6/~M into depolymerizing buffer.

1'0

210 Time

30

(rain)

Fig. 5. Depolymerization of actin in the presence of phailoidin (PHI)). The experiment was initiated by diluting F-actin to a final concentration of 15/~g/ml = 0.36/~M into depolymerizing buffer, containing the PHD concentration indicated. F-actin was polym¢fized without PHD. At the time indicated by the arrowhead this particular sample was sonicated for 30 s (60 W with the mierotip of a Branson Sonifier B-12).

Influence of actin-binding drugs Cytochalasin B (from the mold Helminthosporium dematoideum) and phalloidin (from the mushroom Amanita phalloides) are both known to strongly influence the polymerization of actin. Phalloidin stabilizes the polymeric form of actin (F-actin) [17-19]. Low concentrations of cytochalasin B retard, higher concentrations of cytochalasin B accelerate actin polymerization, suggesting that cytochalasin B stimulates the nucleation phase of actin polymerization [20] but inhibits the elongation phase, most probably by binding to the filament ends [21,22]. The experiments with phalloidin are presented in Fig. 5. It can be seen that very low phalloidin concentrations (4 nM; actin concentration 360 nM) dramatically retarded the depolymerization of actin. It is remarkable that even in this case a

TABLE III PARAMETERS OF THE DEPOLYMERIZATION CURVE OF Fig. 4 AND OF ITS SECOND DERIVATIVE For further details see Table I. Age of aefin

(71

C2

k]

k2

C~

C~

1 2 3 6

65.3 55.0 44.5 63.5

21.8 29.5 37.8 27.0

1.05 0.67 0.29 0.14

0.03 0.03 0.02 0

72.0 24.7 3.74 1.24

0.02 0.07 0.015 0

fast initial phase could be observed; however, the contribution of this fast phase to the whole kinetics was only very small at inhibiting phalloidin concentrations. It is known that treatment with ultrasound (sonication) enhances actin depolymerization (for a recent report see Ref. 23). In order to see whether this is also true in the presence of phalloidin, the following experiment was conducted. In the presence of 0.4 /~M phalloidin 0.36 #M actin did not depolymerize. After 30 s of sonication, however, a slow depolymerization began to occur which did not stop after the cessation of sonication, indicating that the alteration of actin produced by sonication (most probably fragmentation) 15ersisted (Fig. 5). The effect of cytochalasin B is shown in Fig. 6. The most remarkable fact is that the fast phase became slower and the slow phase became faster in the presence of cytochalasin B. As a consequence of the acceleration of the slow phase complete depolymerization was reached earlier in the presence of cytochalasin B than in the absence of cytochalasin B. The curve became more uniform. This is reflected in the semilogarithmic plot of Fig. 6B, where the values measured in the presence of 375 and 15 #M cytochalasin B declined more or less regularly, indicating uniform exponential de-

392 90~

90"

~ E -= = >.

oo>~ ~ 30-

1'0

2'0

]

1vO

! 30

2g0 Ti me (rain}

2I"

0

3'6 ' 3'8 ' 4.0

0

.E oo 3-

S 2 I.L _5 =

%

0.8

,,"~1 30

4-

°°

E 1.2

o ~ ~ 30" #-

,..:- a 0 5-

~ 1.6

10

20

3O

Fig. 7. Depolymerization of actin at different temperatures. The experiments were initiated by diluting F-actin to a final concentration of 20 /~g/ml =0.4 /~M into depolymerizing buffer. The temperatures of the cuvettes were adjusted to the desired values. The main part of the figure shows three depolymerization curves obtained at the temperatures indicated, the inset shows an Arrhenius plot of all measurements performed.

Time {min)

Fig. 6. Depolymerization of actin in the presence of cytochalasin B (CB). The experiments were initiated by diluting F-actin to a final concentration of 30 /~g/ml = 0.72 /LM into depolymerizing buffer. One experiment was performed without CB, in the other experiment 15/~M CB was in the depolymerizing buffer. A shows the depolymerization curves; in B the natural logarithm of the fluorescence signal is plotted against time.

p o l y m e r i z a t i o n kinetics in the presence of cytoc h a l a s i n B. (The significance of the small lag p h a s e in the presence of c y t o c h a l a s i n B r e m a i n e d unclear.)

Depolymerization at different temperatures F - a c t i n d e p o l y m e r i z a t i o n was m e a s u r e d at various temperatures. T h r e e o f these m e a s u r e m e n t s a r e p r e s e n t e d in Fig. 7. I n T a b l e IV the results of all m e a s u r e m e n t s are shown. T h e curves in Fig. 7 c o u l d r e a s o n a b l y well b e fitted b y two e x p o n e n tials. A s c a n be seen, the p r o p o r t i o n of the two e x p o n e n t i a l s (the C values) d i d n o t change, o n l y the rate c o n s t a n t s i n c r e a s e d with increasing temperature. I n the inset in Fig. 7 the l o g a r i t h m i c values of k 1 are p l o t t e d a g a i n s t the reciprocal of the a b s o l u t e t e m p e r a t u r e ( A r r h e n i u s plot). O n e c a n calculate j u s t in a f o r m a l w a y a n activation energy of a b o u t 27 k J . M - i ; however, its physical m e a n i n g is difficult to evaluate b e c a u s e k m e a sured in this e x p e r i m e n t is a f u n c t i o n of b o t h f i l a m e n t length d i s t r i b u t i o n a n d the r a t e c o n s t a n t

TABLE IV PARAMETERS OF THE DEPOLYMERIZATION CURVE OF Fig. 7 For further details see Table I. t (°C)

cI

c2

k]

k2

1

82.1 77.1 77.5 83.0 85.0 80.0 95.5 89.0 68.8

4.4 10.8 6.2 9.2 7.2 5.2 5.2 8.4 5.7

0.14 0.12 0.20 0.26 0.29 0.36 0.47 0.74 0.98

0

5 10 15 20 25 30 35 40

0.02 0.013 0.02 0.01 0.02 0.006 0.08 0.06

g o v e r n i n g the release of m o n o m e r i c units of each filament. Discussion

P r i n c i p a l l y there are two ways in which actiL f i l a m e n t s c o u l d d e p o l y m e r i z e . T h e y c o u l d disass e m b l e explosively, releasing all their subunits at the same time, so that a long f i l a m e n t m a y d i s a p p e a r with the s a m e velocity as a short one, or they m a y release their subunits form their ends in a sequential fashion. Since all the evidence available

393 suggests that during polymerization actin filaments grow by adding subunits onto the filament ends, it is quite reasonable to assume that during depolymerization actin filaments decay by losing subunits from their ends. If so - as is assumed for the following discussion - the velocity of depolymerization should be proportional to the number concentration of filaments and hence of filament ends. Every change in depolymerization rate should reflect a change in the concentration of filaments. In other words, every time the depolymerization velocity changes to a lower value, a group of filaments with a certain length has disappeared. This interpretation is valid only if we disregard (as has also been done by Johnson and Borisy [9], Karr et al. [10] and Kristofferson et al. [11]) the back-reaction (reassociation of monomers to filaments). This is justified when the actin concentrations investigated are significantly lower than the critical concentration under the respective conditions. This holds true for our experiments, since actin was diluted into depolymerizing buffer (which normally is used to store G-actin, i.e., in which the critical concentration is in the range of several mg per ml), so that our actin concentrations were well below critical concentration. The small addition of MgCI 2 due to transference of Mg 2+-polymerized F-actin to the cuvette (maximally 60 # M MgC12) did not significantly lower the critical concentration. Even in 1 mM MgCl 2 the critical concentration was under our conditions not lower than 25 # g / m l (data not shown). That reassociation can be disregarded is further corroborated by the experiment shown in Fig. 8, in which it is shown that DNAase I (which strongly binds to actin monomers (thus inhibiting a possible back-reaction, cf. Hitchcock et al. [24]) did not influence the depolymerization kinetics. Our interpretation also disregards the possibility that there might be breakage or annealing of filaments, but we do not think that filament breakage (which is certainly possible during the mixing procedure before the measurements start) played a significant role during the depolymerization process itself and, if so, this should have led to an increase (rather than to a decrease) in depolymerization velocity. Therefore, we suppose that there are as many length classes of filaments as there are depolymerization velocities. The larger

60.

o

1"0

Time (min)

A

a'0

Fig. 8. Depolymerizationof actin in the presenceof DNAase I. 60 #g F-actin, the highest concentration used in this paper, w e r e diluted into 2 ml of depolymerizingbuffer. - - , no DNAase I; . . . . . , molar ration ofDNAase I to actin 1:1; . . . . . . , molar ratio of DNAase I to actin 2.5 : 1. 31000 was assumed as the molecularweight of DNAase I.

the change in velocity, the more frequent must have been the length class of filaments that has just disappeared. Since the shorter filaments disappear before the longer ones, the time axis of the velocity distribution curve reflects the length axis of the length distribution curve. Therefore, one can obtain the length distribution of filaments from the first derivative of the depolymerization rate with respect to time, that is the second derivative of the original depolymerization curve (Johnson and Borisy [9]). In a graph of the second derivative of the depolymerization curve the units of the abscissa depend on the rate constant of the depolymerization reaction at the single filament end and on the length difference between the various classes of filaments. The earlier the change in depolymerization velocity occurs the larger is the rate constant a n d / o r the smaller the length difference between the various classes. Since we are able to fit our depolymerization curves to a sum of two or three exponentials (only in a few cases could the time course be described by a single exponential; see below and Ref. 29), we can describe the length distribution also by a sum of exponentials (since an exponential function remains exponential after differentiation). The length distribution derived from a depolymerization curve of the form y=C~e-k~ ~ +C2e-k~+C3e-k3~

394

will be y " = C~e-ka x + C~e - k 2 x + C~e-k~ ~

with C " = k 2 C m and m =1,2 and 3

In the depolymerization curve x has the meaning of time, in the second derivative x has the meaning of Al/k* with A 1 being the length difference between the various length classes of filaments and k* being the rate constant (which is the sum of the rate constants of the two filament ends) which determines the length change of the single filament per time. It is not possible to extract the values of k* from our data since A l is not known. The values of k (which are not k*!) describe the shape of the curves and are related to the change in filament concentration with time (or length). Since the mechanistic meaning of k is difficult to assess, the numerical values of k are of minor significance. Different values of k (k 1, k 2, k3) in one experiment only show that a particular time course can be described by more than one exponential. Since the values of k appear as squares in the constant term of the second derivative (C'), it follows that the exponential term with a lower k contributes less to the length distribution than it contributes to the depolymerization kinetics. Conversely, a small fraction of longer filaments (a small value of C') contributes more to the form of depolymerization kinetics than to the length distribution. In other words, a slight deviation from a single exponential in the length distribution results in a noticeable deviation from a single exponential in the depolymerization kinetics. This means that the length distribution of filaments prevailing in the measuring cuvette at the beginning of the experiment is more uniformly exponential than the depolymerization curve would suggest. The experiments of the present work can be roughly divided into two groups. In the one group the changing of one parameter (e.g. changing MgCI 2concentrations or substitution of phosphate buffer for Tris buffer) results in a marked change in the ratio of the values of C of the depolymerization curve (Tables I-III). This is, according to the hypothesis of this work, due to a change in length

distribution of the actin filaments. The depolymerization kinetics reflect the length distribution in the measuring cuvette. In the case of various MgC12 concentrations or in the comparison between phosphate and Tris buffer, F-actin was always taken from the same stock solution so that the change in length distribution could only occur during the addition of F-actin to the depolymerization assay. Therefore, the mechanical stress of blowing actin into the cuvette influenced the length distribution in different ways depending on the depolymerizing medium. Blowing actin into phosphate buffer or into 1-2 mM MgCl 2 obviously retained an initial deviation in length distribution from a single exponential, whereas in the absence of MgC12 or in the presence of Tris-maleate buffer F-actin was obviously fragmented in such a way that a more or less exponential length distribution ensued. Such a fragmentation (yielding more filament ends) can also explain the higher values of k which are observed in low MgC12 or in "Iris buffer. A change in the values of k can also be found in the data of Table III (referring to actin preparations of different ages). In this experiment there was a tendency for the C1 values to become lower and the C2 values to become higher. This suggests that in the older preparations the length distribution had shifted to longer filaments. The filaments might even be so long that they comprise, e.g. after 6 days, about one-third of the polymerized actin, as suggested by Fig. 4. An increase in filament length is necessarily connected to a reduction in filament number, thus explaining the reduced values of k 1 in Table III. The second group of experiments referred to above is represented by the experiment with varying temperatures. In this experiment the proportions of the C values did not significantly change, indicating that increasing temperatures did not primarily change the length distribution but increased the rate constant of subunit release at the filament ends. The experiment with the drugs cytochalasin B and phaUoidin indicates that both drugs retard subunit release at the filament ends. This is particularly pronounced in the case of phalloidin (consistent with the older experiments of Estes et al. [18] and Coluccio and Tilney [19]) where sonication could induce only a small increase in de-

395

polymerization velocity. An alternative explanation (instead of assuming retardation of subunit release at the single filament ends) for the phalloidin effect would be possible if during depolymerization in the presence of phalloidin the number concentration of filaments was lower than in the absence of phalloidin. This, however, is unlikely, since actin was pipetted (where filament breakage or its inhibition by phalloidin could have occurred) into the cuvette in the absence of phalloidin so that an altered length distribution by phalloidin at the beginning of the depolymerization process cannot be expected. Reannealing of broken filaments in the presence of phaUoidin is also unlikely, since the acceleration of depolymerization by sonication persisted after the stopping of sonication (Fig. 5). In the case of cytochalasin B two effects appear to be superimposed: the retardation of the first depolymerization phase reflects the slowing down of subunit release at the filaments ends and confirms the observation of Nishida et al. [25] and Bonder and Mooseker [26]. The acceleration of the second phase suggests a fragmentation of actin filaments in the presence of cytochalasin B, as is frequently postulated by various authors (cf. Refs. 27, 28). This latter result is reminiscent of the observation of Walsh et al. [29], who have found that the protein villin (which blocks the 'barbed' end of actin filaments and fragments filaments) also converts biphasic depolymerization kinetics into exponential. The difference with our observation is that in the presence of villin the slow phase was not accelerated but rather extended over the whole range of depolymerization, but the similarity remains that in both cases a 'capping' by a fragmenting substance makes depolymerization kinetics more uniform. Together with the experiment at a low MgC12 concentration (see Fig. 2 and Table I) the cytochalasin experiment suggests that an exponential length distribution might frequently result from filament fragmentation (although Inoue et al. [30] deduced from electron micrographs that during the initial stage of polymerization actin in the presence of cytochalasin D has a Poisson-type length distribution, as compared to an exponential length distribution in the absence of cytochalasin D).

In conclusion

This work has shown that the measuring of actin depolymerization is a sensitive tool for investigating various factors which influence actin filament structure. The present work illustrates that the depolymerization kinetics reflect both the length distribution of actin filaments and the rate constant governing subunit release at the filament ends.

Acknowledgement We thank the Deutsche Forschungsgemeinschaft for financial support.

References 1 2 3 4

Kom, E.D. (1982) Physiol. Rex,. 62, 672-737 Weeds, A. (1982) Nature 296, 812-816 Schliwa, M.C. (1982) Cell 25, 587-590 Stossei, T.P., Chaponnier, C., Ezzell, R.M., Hartwig, J.H., Janmey, P.A., Kwiatkowski, D.J., Lind, S.E., Smith, D.B., Southwick, F.S., Yin, H.L. and Zaner, K.S. (1985) Annu. Rev. Cell Biol. 1, 353-402 5 Neuhaus, J.M., Wanger, M., Keiser, T. and Wegner, A. (1983) J. Muscle Res. Cell Motility 4, 507-527 6 Frieden, C. (1985) Annu. Rev. Biophys. Biophys. Chem. 14, 189-210 7 Kouyama, T. and Mihashi, K. (1981) Eur. J. Biochem. 114, 33-38 80osawa, F. and Kasai, M. (1971) in Subunits in Biological Systems (Timasheff, S.N. and Fasman, G.D., eds.), Marcel Dekker, New York 9 Johnson, K.A, and Borisy, G.G. (1977) J. Mol. Biol. 117, 1-31 10 Karr, T.L., Kristofferson, D. and Purich, D.L. (1980) J. Biol. Chem. 255, 8560-8566 11 Kristofferson, D., Karr, T.L. and Purich, D.L. (1980) J. Biol. Chem. 255, 8567-8572 12 Spudich, J.A. and Watt, S. (1971) J. Biol. Chem. 246, 4866-4871 13 Pardee, J.D. and Spudich, J.A. (1982) Methods Enzymol. 85 Part B, 164-171 14 MacLean-Fletcher, S. and Pollard, T. (1980) Biochem. Biophys. Res. Commun. 96, 18-27 15 Gordon, D.J., Yang, Y.-Z. and Korn, E.D. (1976) J. Biol. Chem. 51, 7474-7479 16 Nonomura, Y., Katayama, E. and Ebashi, S. (1975) J. Biochem. 78, 1101-1104 17 Dancker, P., LiSw, I., Hasselbach, W. and Wieland, T. (1975) Biochim. Biophys. Acta 400, 407-414 18 Estes, J.E., Selden, L.A. and Gershman, C. (1981) Biochemistry 20, 708-712 19 Coluccio, L.M. and Tilney, L.G. (1984) J. Cell Biol. 99, 529-535

396 20 Dancker, P. and L/Sw, I. (1979) Z. Naturforsch. 34c, 555-557 21 Brown, S.S. and Spudich, J.A. (1979) J. Cell Biol. 83, 657-662 22 Flanagan, M.D. and Lin, S. (1980) J. Biol. Chem. 255, 835-838 23 Carlier, M.F., Pantaloni, D. and Korn, E.D. (1984) J. Biol. Chem. 259, 9987-9991 24 Hitchcock, S.E., Carlsson, L. and Lindberg, U. (1976) Cell 7, 531-542 25 Nishida, E., Ohta, Y. and Sakai, H.C. (1983) J. Biochem. 94, 1671-1683

26 Bonder, E.M. and Mooseker, M.S. (1986) J. Cell Biol. 102, 282-288 27 Maruyama, K., Hartwig, J.H. and Stossel, T.P. (1980) Biochim. Biophys. Acta 626, 494-500 28 Hartwig, J.H. and Stossel, T.P. (1979) J. Mol. Biol. 145, 563-581 29 Walsh, T.P., Weber, A., Higgins, J., Bonder, E.M. and Mooseker, M.S. (1984) Biochemistry 23, 2013-2021 30 Inoue, Y., Tashira, A. and Maruyama, K. (1985) Biomed. Res. 6, 343-346