Jan 27, 1984 - VDepartment of Physiology, School of Dentistry, University of Pacific, San Francisco, .... scattering, LS,, and the initial rate of super-opalescence, VO, were ..... seen (for technical reasons, we could not photograph such.
Val. 260, No. 4, Issue of February 25, pp. 2321-2327.1985 Printed in U.S.A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 8 1985 by The American Society of Biological Chemists, Inc.
Skeletal Muscle Myosin Subfragment-1 Induces Bundle Formation by Actin Filaments* (Received for publication, January 27, 1984)
Toshio Ando$$ and Donald Scales7 From the $Cardiovascular Research Institute, School of Medicine, University of California, San Francisco, California 94143, the VDepartment of Physiology, School of Dentistry, University of Pacific, San Francisco, California 941 15, and the Department of Biological Chemistry, University of Maryland, Baltimore, Maryland 21201
As is well known,the light scattering intensity of Factin solutions increases immediately upon formation of the rigor complex with subfragment-1 (S-1). We have found that after the initial rise in scattering, there is a further gradual increase in scattering (we call it “super-opalescence”). Fluorescence and electron microscopic observations of acto-S-l solutions showed that super-opalescence results from formation of actin filament bundles once S-1 binds to F-actin. The actin bundles posses7ed transverse stripes witha periodicity of about 350 A, which suggested that in the bundles actin filaments are arranged in parallel register. The rate of the initial process of bundle formation (Le. sideby-side dimerization) could be approximately estimated by measuring the initial rate of super-opalescence (Vo).Vo had a maximum (V,”)at a molar ratio of regardless of the actin concentraS-1 to actin of %”%, tion, pH (6-8.5), Mg2* concentration (up to 5 mM), or ionic strength (up to 0.3 M KC1). Lower pH, higher Mg2+ concentration, and higher ionic strengthincreased V,”; Vo was proportional to the square of the actin concentration, regardless of the solution conditions.
headed subfragment of rabbit skeletal myosin,S-1,can assemble actin filaments into bundles under physiological conditions. The S-1-induced bundlewas characterized by measuring the initial rate of its formation undervarious conditions, observing the bundles by fluorescence and electron microscopy, and studying productswhich were formed by chemical cross-linking of acto-S-1 with a water-soluble carbodiimide under various conditions. MATERIALS ANDMETHODS
Protein Preparations-Myosin was prepared from rabbit skeletal back and leg muscles (6) and stored in 0.3 M KC1 and 50% glycerol at -20 “C. Chymotryptic S-1was obtained by the method of Weeds and Taylor (7). S-l(A1) and S-l(A2)were used together. Papain S-1 was prepared according to Margossian et al. (8). HMM was obtained by digesting myosin with a-chymotrypsin (9). Thesemyosin subfragments were stored a t -20 “C in thelyophilized form. The lyophilization was performed in 0.15 M sucrose, 5 mM ammonium acetate, and 0.2 mM dithiothreitol at pH7.0, and did not affect results below. Just before using S-1 and HMM, the lyophilized proteins were dialyzed overnight against a large volume of buffer containing 20 mM KC], 20 mM TES (pH 7.0), 2 mM MgSO,, and 0.2 mM dithiothreitol and then clarified by centrifuging for 90 min at 47,000 rpm using a 50-rotor. Acetone powder of rabbit skeletal muscle was prepared as described before (10). Actin was purified by the method of Spudich and Watt (11). For some purposes, actin was further purified by Sephacryl S200 gel chromatography; such actin gave the same results as those Actin is ubiquitous in eukaryotic cells and takes a variety obtained without S-200chromatography. A high concentration (-13 of forms ( i e . monomer, filament, bundles, and mesh-work mg/ml) of F-actin was stored in 0.3 mM ATP plus buffer A (20 mM structure) according to its functions in the cells. The forms KCI, 20 mM TES (pH 7.0), 2 mM MgSO,, 0.2 mM dithiothreitol, 1 are determined by the ionic conditions and by several types mM Na2HP04, and0.1 mM NaN3) a t 0 “C. Justbefore use, the stored of associated proteins. Actin filaments can be assembled into F-actin was centrifuged for 30-40 min a t 47,000 rpm using a 50-rotor, in order to remove free ATP. The pellet was suspended in buffer A bundles under nonphysiological conditions (at low pH (-4) and homogenized using a Teflon homogenizer. The centrifugation (l),or in the presence of high concentration of Mg2+ (2)), and homogenization were repeated once more. The molar concentrawithout the assistance of any proteins. Under physiological tion of chymotryptic S-1, papain S-1, and HMMwere determined by conditions,actinbundlesare formed by association with their absorption in 0.6 M NaCl using the El2 = 7.7, 8.3, and 7.0, specific proteins such as fascin, fimbrin, villin, filamin, or an respectively, and respective molecular weights of 1.15 X lo5, 1.33 X lo‘, and 3.5 X 10’. The molar concentration of F-actin was estimated actin-bindingprotein identified in lungmacrophages(see using E% = 6.7 and a molecular weight of 4.2 X lo4. Corrections for recent reviews, 3, 4). Trinick andOffer ( 5 ) have reported that turbidity were made as described before (12). the two-headed subfragment of skeletal muscle myosin, heavy Phalloidin-rhodamine-labeled F-actin(1p ~ was ) prepared accordmeromyosin (HMM’), cross-linksactinfilamentsto form ing to Yanagida et al. (13). 0.15 p~ S-1 was added to the labeled F“rafts,” by thebinding of eachheadto a differentactin actin in buffer containing 0.15 M KC], 20 mM TES (pH 7.0), 5 mM filament. In the present study, we report thateven the single- MgSO,, 0.2% (v/v) /3-mercaptoethanol, 1 mM Na2HP04, and0.1 mM NaN3, and the acto-S-1 solution was incubated for a few hours a t was diluted 5-10-fold in the same * This work was supported by National Science Foundation Grants 15 “C. Then, the acto-S-1 solution PCM 75-22698 and NHBLI-16883. The costs of publication of this buffer as above and put into a glass capillary 50 p~ in thickness. article were defrayed in part by the payment of page charges. This Phalloidin-rhodamine was purchased from Molecular Probes. Light Scattering-Light scattering intensity of acto-S-1 solutions article must therefore be hereby marked “aduertisement” in accordwas measured a t 400 nm by detecting light scattered in a direction ance with 18 U.S.C. Section 1734 solely to indicate this fact. perpendicular to that of the incident light propagation. A Hitachi P T o whom correspondence should be addressed. ’ T h e abbreviations used are:HMM, heavy meromyosin; S-1, Perkin-Elmer MPF-4 fluorometer was used for the measurements. subfragment-1; TES,N-tris[hydroxymethyl]methyl-2-amino-ethane- Since the units of light scattering intensity are arbitrary, the intensity sulfonic acid; MES, 2-(N-morpho1ino)ethanesulfonicacid; SDS, so- was normalized by assigninga value of 10 to the light scattering dium dodecyl sulfate; EDC, l-ethyl-3-(3-dimethylaminopropyl)- intensity of 5 p M F-actin in a solution containing 0.1 M KC1, 20 mM carbodiimide. TES (pH 7.01, 2 mM MgSO,, 1 mM Na2HP04,0.2 mM dithiothreitol,
2321
Myosin Subfragment-1 -induced Actin Bundles
2322
way). This seemed rather strange since the light scattering intensity of the rigor complex of acto-S-1 has been thought not to change with time, although it has been well-known that in super-precipitation, the turbidity of acto-myosin solutions increases with time, following the hydrolysisof ATP. So, we characterized super-opalescence by measuring it under various conditions. First, reversibility of super-opalescence was tested. Upon adding 10 ~ L ATP M t o an acto-S-1 solution which had been incubated for a while a n d had been showing super-opalescence, the light scattering intensity immediately dropped to a level equal t o that of F-actin alone (see Fig. 1). of After a while, the scattering rose sharply due to rebinding S-1 to F-actin and then againsuper-opalescenceappeared withalmostthesamerate as before.Therefore,superopalescence is reversible and reproducible. Second, the initial rate ( Vo)of super-opalescence was studiedas a function of S1concentration. As shown inFig. 2, the function was biphasic, with a maximum (VF)at a molar ratio, S-1/F-actin, of %-%, regardless of actin concentration. Although Vo seemed t o b e FIG. 1. Time course of change in light scattering intensity zero for the first few minutes when the S-l/actin ratio was of an F-actin solution.At arrow 1,0.75 p~ S-1was added to 5 p M above 0.5, V, never completely became zero. (When the light F-actin in solution I (0.15 M KC1,20 mM TES (pH7.0), 2 mM MgS04, scattering intensity of such acto-S-1 solutions was observed 1 mM Na2HP04, and0.2 mM dithiothreitol] at 15 “C. At arrow 2, 10 over 1 h, a definite increase in scattering was observed.) The p~ ATP was added. The light scattering was measured at 400 nm. initial rise of scattering due t o S-1binding to F-actin increased i n a slightly sigmoidal manner with increasing S-1 concentraand 0.1 mM NaN3 at 15 “C. A small volume (5-100 pl) of S-1 solution was added to 2 ml of F-actin solution in a thermostated fluorescence tion (Fig. 3). Similar nonlinearity of light scattering (or turcuvette and then immediately the light scatteringintensity was bidity) at low occupancy has been previously observed by recorded on time-scanning chart paper (Fig. 1).The initial rise of the others (15, 16). Such behavior might prevent us from conscattering, LS,, and the initial rate of super-opalescence, VO,were cluding that V z occurred at a ratio of bound S-l/actin = Yread from the chart paper. of thisconclusionrestsonlyon Electron Micrography-Droplets of solutions of 1 p~ F-actin or 1 %. However,thevalidity S-1 = bound S-1, and this seems assured by whether added p~ F-actin plus various amounts of S-1 were applied to freshly carbon-coated Formvar grids. After 20 s, the droplets were drawn off the high affinityof S-1 for actin (lo7M-’; Refs. 16 and 17);at with torn filter paper. A drop of 1% uranyl acetate was applied a ratio of ?k-%,more than 90% of the added S-1 should be immediately to the grid for another 20 s. This drop of negative stain bound. As shown inFig. 3, after incubating acto-S-1 solutions was then removed with filter paper. The grids were allowedto dry for several minutes before observing them at 80 kV in a Philips EM200 electron microscope. Before applying to grids, acto-S-1 solutions had been incubated for various length of time ( 2 min, 1 h, or 24 h) in solution I (0.15 M KCI,20 mM MES (pH 6.51, 2 mM MgS04, 1 mM Na,HP04, 0.2 mM dithiothreitol, and 0.1 mM NaNJ or solution 11 (10mM KC1,20 mM Tris-HCI (pH 8.0), 2 mM MgSO4,l mM Na2HP04, 0.2 mM dithiothreitol, and 0.1 mM NaNJ. FluorescenceMicroscopy-Fluorescent F-actin or acto-S-1 were observed with a Zeiss Photomicroscope I11 with epifluorescence optics, a Zeiss Neofluar x100 objective (oil immersion, NA 1.3), a 150watt mercury arc lamp, and a Zeiss rhodamine filter set (Aex -550 nm, X., >580 nm). The pictures of the fluorescent images were taken by a built-in camera using Kodak CF 1000 film. Since the exposure takes a long time and actin bundles moved quickly by diffusion, we focused on an inside surface of a glass capillary where some of the bundles were attached and immobile. Chemical Cross-linlzing-Acto-S-1 (20 p M F-actin plus 3 p M s-1) was chemically cross-linked by a water-soluble carbodiimide, l-ethyl3-(3-dimethylaminopropyl)carbodiimide,EDC, under various conditions, and the cross-linked products were studied by nongradient SDS-polyacrylamide gel electrophoresis (7.5% acrylamide). In order to see the ingredients of the cross-linked products, actin andS-1 were separately labeled with N-iodoacetyl-N’-(5-sulfo-l-naphthyl)ethylenediamine. Then, labeled S-1 plus F-actin or S-1 plus labeled Factin were cross-linked by EDC. The fluorescence bands after SDSpolyacrylamide gel electrophoresis were monitored by illuminating the gels by a UV lamp. The concentrations of labeled proteins were estimated by the method of Lowry et al. (14).
l2
RESULTS
As shown in Fig. 1, the light scattering intensityof F-actin S-1. It didnot,however,stay rose immediately on adding constant; the rise was followed by a further gradual increase in light scattering (the optical density behaved in the same
FIG. 2. Dependence of initial rate of super-opalescence on the molar ratioof S-l/actin. 0, 7 p M F-actin; 0, 5 p~ F-actin; A, 3 p~ F-actin. The solution condition is the same as that in Fig. 1.
2323
Myosin Subfragment-1 -induced Actin Bundles
/ " O
P
FIG. 3. Light scattering intensity of acto-Si-1solutions, as a function of S-1 concentration, at different times after adding immediately after adding S-1; A, 6 min later; 0, S-1 to F-actin. 0, 24 h later. The acto-S-1 (5 pM F-actin plus various amounts of s-1) in solution I was incubated at 15"C.
for 24 h at 15 "C, the light scattering rose significantly at every S-1 concentration. Third, the dependence of Vo on actin concentration was studied. The molar ratio of S-1 to actinwas kept constant at 0.15 (which gives V r ) or 0.35.As shown in Fig. 4, in both cases, Voincreased sigmoidally with increasing actin concentration. Plots of versus actin concentration gave straight lines at actin concentrations below 8 PM. That is, Vo was proportional to the square of the actin concentration. Since thisrelation held regardless of the S-1 to actin ratio, the initialstage of super-opalescence at differentratios (S-l/ actin) comes from the same molecular process. In otherwords, although super-opalescence probably results fromseveral consecutive processes, at the initial stageof super-opalescence a single process (( Vo [actin]') suggests side-by-side assembly of two actin filaments) occurs at both lower and higher ratios of S-l/actin. Fourth, the effect of varying pH, ionic strength, and Mg2+ concentration on theV r of super-opalescence was examined. As shown in Fig. 5, V? was highlysensitive to the ionic environment of acto-S-1, i.e. lower pH, higher ionic strength, and higher concentrations of M$+ increased Vr. This behavior suggests that electrostatic force is an inhibitory factor in the molecularprocesses of super-opalescence. It should be mentioned that the experiments in Fig. 5 were performed using a constant concentration of s-1 (i.e. 0.75 p M ) because theS-1concentration whichproduces the V r was little changed by the variations of H, ionic strength, and M$+ concentration. Linearity of versus actin concentration was again observed at these different ionic conditions (data not shown). Finally, all the foregoing experiments were repeated using papain S-1 and HMM, instead of chymotryptic S-1. Papain S-1 gave essentially the same results as those with chymotryptic S-1 (data not shown), except that value theof V r with papain S-1 was about one-third that with chymotryptic S-1
a
&
/
O
FIG. 4. Dependence of the initial rate of super-opalescence on the actin concentration.The molar ratio of S-l/actin was kept constant at 0.15 (0,0 or 0.35 (A, A). 0, A, Vo uersus F-actin concentration; 0 , A, Vouersus F-actin concentration. The solution condition is the same as that in Fig. 1.
2
under identical solution conditions.Although the "twoheaded myosin subfragment, HMM,produced super-opalescence, its V r was negligibly small. The V r under condition of pH 7.0,0.15 M KCl, and 2 mM MgS04 was only about 2% that with chymotryptic S-1 under the same condition. The optimum ratio of HMM heads to actinfor Vowas %-%. In order tovisualize the molecular events of super-opalescence, acto-S-1 was observed with an electronmicroscope. To follow the time course of the molecular events, acto-S-1was negatively stained at several times aftermixing chymotryptic S-1 with F-actin. Electronmicrographs of F-actin incubated for 24 h in solution I (pH 6.5, 0.15 M KC1; see "Materials and Methods")showed the usualsingle actinfilaments everywhere. In the case of a nonsaturated acto-S-1 sample (0.15 PM S-1 plus1p~ F-actin) which had been incubated for only 2 min in solution I before staining, the micrographs showed in places twofilaments clinging together ina crooked arrangement, in addition to stretchedsingle filaments (see Fig. 6A). When incubated for 1 h, long, straight bundles of 2-4 filaments appeared (Fig. 6B). On incubating for 24 h, thick, long bundles could be seen everywhere (see the low magnification electron micrograph Fig. 6C). A high magnification of these actin bundles is shown in Fig. 7.As marked with arrows, the bun.
\
0
!
C 1
a1
0
KC1 (M 1
I
0
1
2
a2
3
a3 4
5
MgSQ,(I”)
FIG.5 . Dependence of the maximum rate of super-opalescence on the solution condition. 0.75 p M S-1was added to 5 PM F-actin. 0, the pH was varied; A, the KC1 concentration was varied; 0, the MgSO, concentration was varied. Other conditions are the same as those of solution I.
4
diameter can be as great as 90 (20), negatively stained filaments have diameters of 60-80 A). In order to correlate further thesuper-opalescence of actoS-1 solutions with bundle formation seen in electron micrographs, acto-S-1 samples under conditions that give quite low rates of super-opalescence were observed with an electron microscope. First, a stoichiometric complex of acto-S-1 (1pM S-1 plus 1 p~ F-actin) incubated in solution I at 4 “C was observed. When the acto-S-1 hadbeen incubated for 1 h, the electron micrograph showed amixture of S-1-decorated single actin filaments and clinging and twisting filaments. (Note that in the case of acto-S-1 (0.15 p~ s-1plus 1 p~ F-actin) treated in the same way, quite organized bundles were already formed.) However, on incubating for 24 h, thick, long and organized bundles were formed (Fig. 8). Second, an acto-S-1 sample (0.15 p~ S-1plus 1 p~ F-actin) incubated in solution FIG.6. Electron micrographs of acto-S-1 at different time I1 (pH 8.0, 10 mM KCl; see “Materials and Methods”)at 4 “C stages after S-1was added to F-actin. 0.15 p~ S-1 was added to was observed. With 1-h incubation, the sample did not contain 1 PM F-actin in solution I. Then, at different times, the sample was any bundles, whereas with 24-h incubation, bundles identical stained with 1% uranyl acetate on carbon-coated copper grids. A, 2 min later; R, 1 h later; C, 24 h later. to those seen in Fig. 7 were formed. Since negative staining with 1% uranyl acetate (pH 4.0) was used for observing actin bundles by electron microscopy, seen (for technical reasons, we could not photograph such there may arise a question whether actin bundle formationis specimens). assisted by the uranyl acetate treatment. Since Yanagida et Summing up all the observations above, we can conclude al. (13)have recentlysucceeded in observingsingle actin that under conditions wherein a high rate of super-opalesfilaments by fluorescence microscopy using phalloidin-rho- cence was observed, long and well-organized actin bundles were formed in a short incubationof acto-S-1, whereas under damine-labeled F-actin, we employed their method in order to remove this question. As shown in Fig. 9, we observed conditions which give a quite low rate of super-opalescence, bright, long, thick fibers in an acto-S-1 sample which had bundles were not formed in a short incubation of acto-S-1. been incubated for several hours. However, such large fibers Therefore, it canbe concluded that the super-opalescence of were acto-S-1 solution resulted from formation of actin bundles. disappeared upon adding ATP, and short, thin filaments
Myosin Subfragment-I-induced Actin Bundles
2325
To study the nature of the linkage between actin and S-1 electrophoresis. As shown in Fig. 10,several productsapin actin bundles, acto-S-1 was chemically cross-linked by a peared upon cross-linking acto-S-1 (3 p~ chymotryptic S-1 zero-length cross-linker, EDC, under several conditions. The plus 20 p~ F-actin) by EDC. The main products, 180 and 170 to be one-to-one complexes cross-linked products were studied inSDS-polyacrylamide gel kDa, have been recently identified of actin plus S-1 heavy chain by Sutoh (21),Greene (22), and Heaphy and Tregear (23). There are several minor products above the 180 kDa. As shown in Fig. 11A, acto-S-1 in a low ionic strengthsolution producedmore of the 180-, 170-, 260-, and 270-kDa bands than it did ina high ionic strength solution. But, the 200-kDa was produced only in a high ionic strength solution. Furthermore, as shown in Fig. 11R, only the 200-kDa production depended on the incubation time of acto-S-1, i.e. the longer incubation increased the amount of 200-kDa production. After noting these characteristics, we examined whether the 200-kDa contains both actin and S-1 heavy chain. To do so, F-actin andS-1were separately labeled with N-iodoacetyl-N'-(5-sulfo-l-naphthyl)ethylenediamine which has been known to react specifically with cysteine 374 of actin (24) and SH, of the S-1 heavy chain (25) and then, labeled F-actin plus S-1or F-actinplus labeledS-1were crosslinked by EDC. As shown in Fig. 11C, the 200-kDa band was fluorescent in both cases, i.e. the 200-kDa contains both actin and S-1 heavy chain. DISCUSSION
Electron andfluorescence microscopic observations of actoS-1 showed that super-opalescence results from the formation of actin bundles. The initial rateof super-opalescence, Vo, is proportional to the squareof the actin concentration,regardless of S-l/actin and the ionic conditions. Therefore,the initial stageof super-opalescence seems tocome from a single process, i.e. side-by-side assembly of two actin filaments. So, Vo is proportional to the apparent rate of dimerization. We have also studied V, as a function of S-1 concentration. In this plot, Vo may not be a straightforward measure of the FIG. 7. A high magnification electron micrograph of actin filament bundles. Acto-S-1 (1 pM F-actin plus 0.15 p M s-1)was dimerization rate since the light scattering increment/dimer incuhated for 24 h in solution I a t 4 "C. Note that in places transverse formed may depend on theS-1concentration, However, qualitatively the biphasic nature of this plot clearly shows that stripes are clearlv seen. Some of them are marked witharrows. "
FIG. 8. A high magnification electron micrograph of actin filament bundles. Acto-S-1 (1 p~ F-actin plus 1 p~ S - 1 ) was incubated for 24 h in solution I a t 4 "C. Also, thissamplehas artransversestripesasmarkedwith rows.
Myosin Subfragment-1-induced Actin Bundles
2326
,180K -170K - 5 1 HC -
95K-
42K-
-H
PI(;. 9. A photograph of bundles of labeled actin filaments.
phalloidin-rhodamine-
L-1
C
0
A
"actin
--
actin
-
"
L-2
L S
FIG. 11. Cross-linking of acto-S-1 with EDC under several conditions. A, ionic strengthdependence of thechemicalcrosslinking; H , acto-S-1 (10 p~ F-actin plus 1.5 PM S-1) incubated for 1 h in 0.2 M NaCI, 2 mM MgSO4,20 mM MES (pH6.5).1 mM NazHPO,, and 0.1 mM NaN3 a t 20 "C was reacted with 2.5 mM EDC for 60 min; L,acto-S-1 (10 pM F-actin plus 1.5 p~ S-1) incubated for 1 h in 2 mM MgSO,, 20 mM MES (pH 6.5). 1 mM Na,HPO,, and 0.1 mM or 60 NaN3 a t 20 "C was reacted with 2.5 mM EDC for 15 min (L-1) min (L-2).B, dependence of chemical cross-linking on the incubation time of acto-S-1 solution. Acto-S-1 (20 p M F-actin plus 3 p M s-1) was incubated for 24 h (L)or 1 min (S) in 0.1 M NaCI, 2 mM MgSO4, 20 mM TES (pH 7.0), 1 mM Na,HPO,, and 0.1 mM NaN3 and then reactedwith 7.5 mM EDC for 10min a t 24 "C. C, chemical cross-linking of fluorescence-laheled acto-S-1. 20 p~ N-iodoacetyl-N'-(5-sulfo-l-naphthyl)ethylenediamine-laheled actin plus3 p~ S-1was incubated for 24 h in 0.15 M NaCI, 20 mM MES (pH 6.5),2 mM MgSO,, 1 mM NazHP04, and 0.1 mM NaN3, and then reacted with 2.5 mM EDC for 30 min. Right, 20 p M F-actin plus 3 p M N-iodoacetyl-N'-(5-sulfo-l-naphthyl)ethylenediamine-S-lwas treated in the same way as above. S-1 hc represents heavy chain of S-1.
Left.
half-periodicity of the actinhelix. And, S-1 molecules located on suchperiodic points connect two actin filaments. What do S-1molecules do which are not located on periodic the points? Before the bundlingof the two filaments hasbeen completed, some of the S-1 might have tried to cross-link the two filaments, but could do so only at the expense of bending and a c t in -la+.. twisting the filaments. So, cross-links by such ill-placed S-1 molecules will slow down the parallel assembly of actin filaFIG. 10. Electrophoretic pattern of acto-S-1 after chemi- ments. Therefore, a certain molar ratio (muchless than 1)of cally cross-linking with EDC. Acto-S-1 (3 p~ S-1 plus 20 p~ F- S-1to actinmaximizes the rateof bundling of actin filaments. actin) was incubated for 24 h in solution containing 0.15 M NaC1, 20 The present study was motivated by having observeda mM TES (pH 7.0), 2 mM MgSO,, 1 mM NazHP04, and0.1 mM NaN3 a t 4 "C, and then reacted with 3 mM EDC for 30 min. The reaction peculiar pyrophosphate effect on light scattering from actowas terminated by adding excess amount of 8-mercaptoethanol. S-1 S-1 (see Footnote2; a similar phenomenon hasbeen observed hc represents heavy chain of S-1. with acto-HMM by Fujime et al. (26)): on adding a small amount of PPi to a stoichiometric complex of acto-S-1, the the dimerization rate is biphasicbecause it is highly unlikely light scattering decreases and stays constant; further addition that the increment/dimer is biphasic. Nevertheless, it must of PPi decreases the light scattering more. This behavior is commonly observed (PPI has been known as a dissociating be conceded that the relations generating this plot are too reagent (17, 27)). However, such decrements are followed by complicated for quantitative deductions. ionic a gradual increase in light scattering. This PPi-induced superThe maximumrate, Vg, isquitesensitivetothe opalescence can now be explained by the dependence of Vo environment. Low pH, higherionic strength, and higher M$+ on S-l/actin (see Fig. 2). With a stoichiometric complex of concentration increase the V,". These are the same circumstances in which actin filaments are electrostatically neutral- acto-S-1, we would not see super-opalescence. But, once PPi ized. This is quite reasonablebecause it is easier to put actin reduces the fraction of bound S-1, the rate of super-opalesfilaments close to each other when less repulsiveforce is cence would become significant. Therefore, in the above observation, PPi was not per se essential for super-opalescence, exertedbetweenthefilaments.(Sincethelightscattering increment/dimer must be independent of the ionic condition, but just to produce an optimal ratio of bound S-l/actin for the comparison of V," values obtained under different ionic super-opalescence. Chemical cross-linking of acto-S-1 by EDC produced sevconditions is valid.) VT was achieved a t a molar ratio (S-l/actin) of %-%. Why eral products. Among these products, one with apparent M, of 200,000 had distinctive features. Its formation increased is such low ratio optimal for bundle formation? This seems longer beforeadding EDC. Since to be explained by considering that actin filaments in bundlesupon incubating the acto-S-1 the process of binding of S-1 to F-actin isvery fast, chemical are arranged in register. Because of this fact, points where two actin filaments are connected with each other in bundles T. Ando, unpublished results. must have a periodicity along the filaments identical to the
-e
Subfragment-1-induced Myosin cross-linking between actin and S-1 molecules shouldnot depend on the incubation time of acto-S-1, if we consider just ordinary binding of S-1 to actin. Actually, all cross-linked products except the 200-kDa product behaved in the expected way. Therefore, the200-kDa did not derive from the ordinary rigor complex of acto-S-1, but from an actin-S-1 binding involved in the bundlesof F-actin. This interpretation isalso supported by the fact that the 200-kDa product was not formed from acto-S-1 incubated for several hours in a low ionic solution. In such an acto-S-1 solution, bundles are not formed. As discussed in Refs. 4 and 28, proteins which are able to cross-link actin filaments are generally of two classes: one has a minimum of two actin-binding sites within onepolypeptide chain (e.g. villin (29) and gelsolin (30)), and one is a stable oligomer of polypeptides with single actin binding sites (e.g. filamin (31), a-actinin,caldesmon (32), andHMM (5)). Chymotryptic S-1 consistsof a heavy chain (95-kDa) and either of the alkali light chains, i.e. ALC-1 or ALC-2. The heavy chain has two domains, 20 and 50 kDa, available for actin binding (33). ALC-1can bind to actin in the acto-S-1 complex only at low ionic strength (34). Sinceformation of actin bundles by S-1 is inhibited by lowering the ionic strength, ALC-1 does not seem to be involved in the cross-linking of actin filaments by S-1. One of the possibilities is that S-1 cross-links two actin filaments in sucha way that the20-kDa domain binds to one filament and the 50-kDa domain binds to another filament. If this is the case, the 800-kDa product mentioned above must contain one S-1heavy chain and two actins. However, we could not determine the ratio of S-1/ actin in the200-kDa product because we could not obtain the 200-kDa component in sufficient amount. REFERENCES 1. Kawamura, M. & Maruyama, K. (1970) J . Biochem. (Tokyo) 6 8 , 855-899 2. Hanson, J. (1968) in Symposium on Muscle (Ernst, E. & Straub, F. B., eds) p. 999, Akademia, Kiado, Budapest 3. Weeds, A. (1982) Nature (Lond.) 296,811-816 4. Korn, E. D. (1982) Physiol. Reu. 6 2 , 672-737
Actin Bundles
2327
5. Trinick, J. & Offer, G. (1979) J . Mol. Biol. 133, 549-556 6. Tonomura, Y., Appel, P. & Morales, M. F. (1966) Biochemistry 5,515-521 7. Weeds, A. G. & Taylor, R.S. (1975) Nature (Lond.) 257, 54-56 8. Margossian, S. S., Lowey, S. & Barshop, B. (1975) Nature (Lond.) 258,163-166 9. Weeds, A. G. & Pope, B. (1977) J . Mol. Biol. 111 , 129-157 10. Ando, T. & Asai, H. (1979) J. Mol. Biol. 1 2 9 , 265-277 11. Spudich, J . A. &Watt, S.(1971) J. Biol. Chem. 2 4 6 , 4866-4871 12. Ando, T. (1984) Biochemistry 2 3 , 375-381 13. Yanagida, T., Nakase, M., Nishiyama, K. & Oosawa, F. (1984) Nature (Lond.) 3 0 7 , 58-60 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Bwl. Chem. 193,265-275 15. Chock, S. P., Chock, P. B. & Eisenberg, E. (1976) Biochemistry 15,3244-3253 16. White, H. D. & Taylor, E. W. (1976) Biochemistry 15,5818-5826 17. Greene, L. E. & Eisenberg, E. (1980) J. Biol.Chem. 2 5 5 , 543548 18. Hanson, J. & Lowy, J. (1963) J. Mol. Biol. 6, 46-60 19. Depue, R. H., Jr. & Rice, R. V. (1965) J. Mol. Biol. 1 2 , 302-303 20. Heuser, J. E. & Cooke, R. (1983) J. Mol. Biol. 169,97-122 21. Sutoh, K. (1983) Biochemistry 22, 1579-1585 22. Greene, L. E. (1984) J. Bid. Chem. 259, 7363-7366 23. Heaphy, S. & Tregear, R. (1984) Biochemistry 23, 2211-2214 24. Takashi, R. (1979) Biochemistry 1 8 , 5164-5169 25. Duie, J., Takashi, R., Ue, K. & Morales, M. F. (1976) Proc. Natl. Acud. Sci. U. S. A. 73, 302-306. 26. Fujime, S., Ishiwata, S. & Maeda, T. (1972) Biochim.Biophys. Acta 283,351-363 27. Granicher, D. & Portzehl, H. (1964) Biochim. Biophys. Acta 8 6 , 567-578 28. Sheterline, P. (1983) Mechanisms of Cell Motility: Molecular Aspects of Contractility, pp. 104-106, Academic Press, New York 29. Glenney, J. R., Jr., Geisler, N., Kaulfus, P. & Weber, K. (1981) J . Biol. Chem. 256,8156-8161 30. Bryan, J. & Kurth, M. C. (1984) J. Biol. Chem. 2 5 9 , 7480-7487 31. Davies, P. J. A., Wallach, D., Willingham, M., Pastan, I. & Lewis, M. S. (1980) Biochemistry 19, 1366-1372 32. Sobue, K., Muramoto, Y., Fujita, M. & Kakiuchi, S. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 5652-5655 33. Mornet, D., Bertrand, R., Pantel, P., Audemand, E. & Kassab, R. (1981) Biochemistry 20,2110-2120 34. Yamamoto, K. & Sekine, T. (1983) J. Biochem. (Tokyo) 94, 2075-2078