aqueous medium it showed a rapid time-dependent decrease but it was significantly more ... They also indicate that the binding of S-I to F-actin is ... thought not only to act by sterically blocking the binding of ... activity is a general property displayed by S-1 covalently .... sodium thiosulfate, 0.5 mM 2-mercaptoethanol, 0.5 mM.
Eur. J. Biochem. 146, 391 -401 (1985) 0FEBS 1985
Further characterization of the structural and functional properties of the cross-linked complex between F-actin and myosin S-l Jean-Franqois ROUAYRENC, Raoul BERTRAND, Ridha KASSAB, Doris WALZTHONY, Martin BAHLER, and The0 WALLIMANN Centre de Recherche de Biochimie Macromolkculaire du Centre National de la Recherche Scientifique, Montpellier; and Eidgenossische Technische Hochschule, Institut fur Zellbiologie, Honggerberg, Zurich (Received June 26/September 25, 1984) - EJB 84 0687
Several structural and functional properties of the covalent complex, formed upon cross-linking of the myosin heads (S-1) to F-actin with l-ethyl-3-(3-dimethylaminopropyl)carbodiimide, were characterized. The elevated Mg2+-ATPase activity was measured during a 1-month storage of the complex under various conditions. In aqueous medium it showed a rapid time-dependent decrease but it was significantly more stable in the presence of 50% ethylene glycol at -20°C. The ATPase loss most likely reflects a progressive conformational change within the S-1 ATPase site resulting from its greater exposure to the medium, induced by the permanently bound F-actin. The covalent acto-S1 complex was submitted to depolymerization-repolymerizationexperiments using different depolymerizing agents (0.6 M KI; 4.7 M NH4Cl; low-ionic-strength solution). The depolymerization led to an immediate loss of the enhanced Mg2+-ATPase activity; this activity was almost entirely recovered upon repolymerization of the complex. The protein material formed upon depolymerization of the covalent acto-S1 was analyzed by gel chromatography, gel electrophoresis, analytical ultracentrifugation and electron microscopy. It comprised mainly small-sized actin oligomers associated with the covalently bound S-1 and only a limited amount of free G-actin. The results illustrate the relationships between the filamentous state of actin and its ability to stimulate the Mg2+-ATPase activity of S-1. They also indicate that the binding of S-I to F-actin is transmitted to several neighbouring actin subunits and strengthens the interactions between actin monomers. Acto-S1 cross-linked complexes were prepared in the presence of tropomyosin and the tropomyosin-troponin system. Under the conditions employed, the regulatory proteins were not cross-linked to actin or S-1 and did not affect the extent or the pattern of S-1 cross-linking to F-actin. Measurements of the elevated Mg2+-ATPase activity of the cross-linked preparations revealed that tropomyosin and the tropomyosin-troponin complex, in the absence of Ca2+,inhibit ATP hydrolysis; the extent of ATPase inhibition (up to 50%) was dependent on the amount of covalently bound S-I, being larger at low level of S-1 cross-linking; the addition of Ca2+ restored the ATPase activity to the control value. The data provide direct evidence that the regulatory proteins can modulate directly the kinetics of ATP hydrolysis by the covalent acto-S1 complex as has earlier been suggested for the reversible complex [Chalovich, J. M. and Eisenberg, E. (1982) J. Biol. Chern. 257, 2432-24371. Finally the selective limited digestion of tropomyosin with trypsin in the absence and presence of acto-S1 was employed as a probe of the binding of the regulatory protein to the covalent acto-S1. It indicated that the preliminary cross-linking of S-1 leads to an increase of the amount of tropomyosin bound to F-actin; however the extra binding of tropomyosin did not change its ability to regulate the activity of the covalent acto-S1 complex. The molecular mechanism of muscle contraction is based on the interaction of F-actin with the myosin heads which operate a cyclic displacement along the thin filaments to produce muscle shortening [I -31. The F-actin is also associated with tropomyosin and the troponin complex which account in vertebrate striated muscle for the regulation of the muscular motility and the sensitization of the contractile system to the calcium. Tropomyosin is thought not only to act by sterically blocking the binding of the myosin heads to actin (steric blocking model) [4, 51 but also to inhibit a kinetic step of the ATPase cycle [6]. Abbreviations. S-I, myosin subfragment 1 ;acto-Sl, covalent complex between F-actin and S-I ; carbodiimide, l-ethy1-3-(3-dimethylaminopropy1)carbodiimide ; EATP, 1,N6-etheno-adenosine 5'-triphosphate. Enzyme. ATPase, adenosine 5'-triphosphatase (EC 3.6.1.3).
For a better understanding of the myosin head- actin interaction, a model system in vitro which consists of the rigor complex between F-actin and myosin S-I covalently cross-linked by reaction with a zero-length cross-linker, 1ethyl-3-(3-dimethylaminopropyl)carbodiimide, was recently developed and investigated [7]. In this complex the S-1 heads are irreversibly bound to F-actin at the 20-kDa and 50-kDa regions of the heavy chain [7, 81. The isolated covalent complex exhibits a vastly enhanced Mg2+-ATPase activity which at least corresponds to the V of the actin-activated ATPase of the reversible acto-S1 complex at infinite actin concentration. The important increase of the Mg2+-ATPase activity is a general property displayed by S-1 covalently bound to F-actin as it is also observed when the two proteins are joined by the longer bis(imido esters) [9]. On the other hand, the ATP hydrolysis which occurs without dissociation of actin and S-1 is not only a feature of the chemically cross-
392 linked acto-S1 complex, but it seems also to occur during the ATPase reaction catalyzed by the reversible acto-S1 complex [lo]. Finally, kinetic studies on the covalently cross-linked acto-SI complex with ATP analogs suggest that neither the chemical mechanism of ATP hydrolysis nor the rate of the cleavage step is affected by the covalent attachment of actin [I 11. In this work we have extended our knowledge of the various properties related to the carbodiimide-produced covalent complex by first investigating its stability under various experimental conditions. We have also studied the relationship between the polymerization state of actin and the expression of the elevated Mg2+-ATPaseactivity of S-1. The data show a strong dependence between the enhancement of the Mg2+-ATPase activity and the extent of polymerization of the actin molecules in the covalent complex. Finally, we have investigated the influence of the covalent attachment of S-1 to actin on the binding of tropomyosin to F-actin and the capacity of the tropomyosin-troponin complex to regulate the elevated Mg2+-ATPase activity of the covalent acto-S1 complex, in the absence and presence of Ca 2+.The results provide direct evidence that the regulatory proteins can modulate the hydrolytic capacity of the covalent acto-S1 complex, as has previously been suggested for the reversible complex [6]. A preliminary report on some aspects of this work has appeared [12].
The Mg'+-dependent ATPase activities of the covalent acto-S1 complex were measured at 25 "C in a medium containing 50 mM Tris/HCl, 5 mM ATP, 30 mM KCI, 5 rnM MgCI2, pH = 7.5, in the presence of 0.5 mM CaClz or 1 mM EGTA. S-1 was added at 0.060 mg/ml. The actin-activated ATPase of the reversible complex was determined under similar conditions with 1 mg/ml of actin. Pi was measured colorimetrically by an automated phosphate system as previously described [8]. Cross-linking reactions
Cross-linking between F-actin and S-I was performed with 1-ethyl-3-(3-dimethylaminopropyl)carbodiimideas described by Mornet et al. [7]. F-actin (3 mg/ml) in 100 mM Mes, pH = 6.0, was incubated with 15 mM carbodiimide for 2 rnin at 20°C. The F-actin was then mixed with 5 vol. S-I solution in the same buffer (molar ratio actin:S-1 = 2). At various time intervals during the condensation, aliquots (0.5 ml) were withdrawn and the reaction was stopped by the addition of an equal volume of a dissociating solution containing 100 mM Hepes, pH = 8.5, 200mM KCl, 10mM MgCI2, 20mM sodium pyrophosphate. After centrifugation at 170000 x g, for 2 h, at 4"C, the pellet containing the covalent complex was gently suspended in 0.5ml of a 10mM Tris/HCl, pH = 8.0,30 mM KCl, 5 mM MgC12 solution. It was sonicated three times, each time for 1 min at 8 p m in an MSE 150-W sonicator. Aliquots (100 pi) were withdrawn from each fraction and assayed for Mg2+-ATPase activity. The supernatants MATERIALS AND METHODS were analyzed spectrophotometrically for their S-I content Reagents [7l. The cross-linking of S-1 to the F-actin- tropomyosin and Trypsin (treated with tosylphenylalanylchloromethane) was purchased from Worthington Biochemical Corp. 1-Ethyl- F-actin -tropomyosin - troponin complexes was carried out 3-(3-dimethylaminopropyl)carbodiimide was from Sigma. as follows. Tropomyosin or tropomyosin - troponin were first Sodium dodecyl sulfate was obtained from Serva (Heidelberg, associated to F-actin (3 mg/ml) at molar ratios of 1 :1:7 (tropomyosin: troponin: actin) in 10 mM Tris/HCl buffer, FRG). All other reagents were of analytical grade. pH = 8.0, 2 mM ATP, 30 mM KCI, 5 mM MgCi2, 1 mM EGTA. After incubation at 25°C for 30 min with continuous Protein preparations stirring, the mixed proteins were centrifuged at 170000 x g , The proteins used in this study were prepared from rabbit for 2 h, at 25°C. The F-actin pellets, containing at least 90% skeletal muscle. Myosin was isolated from rabbit back and of the added regulatory proteins, were suspended in 100 mM hind leg muscles according to Offer et al. [13]. S-1 was pre- Mes buffer pH = 6.0 and incubated for 2 min with 15 mM pared by digestion of myosin filaments with chyniotrypsin carbodiimide. S-I condensation was then carried out as above. [14, 151 and purified by gel filtration on Sephacryl S-200 Gel electrophoresis showed that the regulatory proteins rein 50 mM Tris/HCl buffer, pH = 7.5. F-actin was purified mained bound to F-actin under the cross-linking reaction according to Eisenberg and Kielley [16]. The protein pellet conditions; samples were removed at various times and prowas resuspended in 10mM KCl, 2 m M MgC12, 0.2mM cessed as for the covalent acto-S1 complex without regulatory CaC12 and 2 mM Tris/HCl, pH = 8.0, containing 1 mM proteins. Mg2 -ATPase activities were then measured either NaN, [17]. The muscle tropomyosin was prepared from the in the presence of 0.5 mM CaC1, or 1 mM EGTA. Binding of myofibrils washed with alcohol and ether according to tropomyosin to the acto-S1 cross-linked complex was carried Greaser and Gergely [I81 and purified by the procedure of out using 12'I-tropomyosin prepared according to Eaton [25]. Limited tryptic digestion of tropomyosin (1 mg/mi) was Hartshorne and Mueller [19]. The tropomyosin-troponin complex was extracted from the actomyosin powder by the carried out in the absence or presence of F-actin (molar ratio method of Graeser and Gergely [18]. Stock solutions of 1 : 7) in 50 mM Tris/HC1 buffer (pH = 8.0), 30 mM KCI and tropomyosin and tropomyosin-troponin complex were made 5 mM MgCI2, at 25"C, in time intervals of 0 - 30 min. A mass by dialysis against 63 mM sodium phosphate buffer, 0.37 M ratio of protease/tropomyosin of 1 :500 was employed. The KCl and 0.1 mM 2-mercaptoethanol, pH = 7.0 [20]. proteolytic reactions were stopped by adding 100 pi of the Protein concentration was determined spectrophotometri- digestion mixture to an equal volume of a solution containing cally with absorption coefficients of A;& n m = 7.5 cm-' for 1% sodium dodecyl sulfate, 5 mM 2-mercaptoethanol, S-1 [15], 11 cm-' for actin [21], 3.3 cm-' for tropomyosin 50 mM Tris/HCl (pH = 7.5); after 3 min at lOO"C, samples [22], and A;&,,m = 3.8 cm-' for the troponin-tropomyosin containing 0.030 - 0.050 mg tropomyosin were subjected to complex and 4.5 cm-' for troponin [16]. The following mol- gel electrophoresis. Sodium dodecyl sulfate/polyacrylamide ecular masses were used for calculating molar concentrations : gel electrophoresis was carried out in 5 - 18% or 10-23% 42 kDa for actin [23], 105 kDa for S-1 [8], 68 kDa for gradient slab gels containing 0.4% bisacrylamide. A 50 mM tropomyosin [24], and 80 kDa for troponin [16]. Tris/100 mM boric acid buffer (pH = 8.6) was used as migra+
393 tion buffer. Gels were stained with Coomassie brillant blue R-250 and destained according to Weber and Osborn [26]. Depolymerization of the covalent acto-SI complex
The covalent acto-S1 complex, formed during a 15-min condensation reaction, was isolated by centrifugation after treatment with Mg*+/pyrophosphate as specified above. The protein pellet (3 mg/ml) was homogenized in a Dounce homogenizer in 5 ml of an ice-cold depolymerizing solution containing 50 mM Tris/HCl (PH = 7.5), 0.6 M KI, 5 mM sodium thiosulfate, 0.5 mM 2-mercaptoethanol, 0.5 mM ATP, 0.5 mM CaCl,. At various times, aliquots (60 pl) were assayed for Mg2+-ATPase activity. A control of native S-1 was run under the same conditions. After incubation for 20 min, the solution was clarified by centrifugation at I70000 x g for 1 h at 4°C and submitted to chromatography on a Sephacryl S-300 column (150 x 1.9 cm) eluted with the same buffer and calibrated with the following protein markers : ovalbumin (43 kDa), bovine serum albumin (68 kDa), aldolase (158 kDa), catalase (232 kDa), xanthine oxidase (275 kDa), ferritin (450 kDa) and P-galactosidase (520 kDa). Native F-actin and carbodiimide-treated F-actin were also incubated in the KI solution and chromatographed on the same column as controls. Similar depolymerization experiments were carried out by using 4.7 M NH4C1 as depolymerizing agent in a buffer containing 50 mM imidazole/HCl, pH = 7.0, 0.5 mM dithioerythritol, 0.1 mM ATP, 0.5 mM CaCI,, 0.01% NaN3. To follow the repolymerization of the covalent acto-S1 and the recovery of its Mg2+-ATPase activity following the elimination of the depolymerizing agents, 3 ml of the acto-S1 solution treated with KI for 20min at 4°C were dialyzed at 4°C in an open dialysis bag against 2 m M Tris/HCl (pH = 8.0), 50 mM KC1, 5 mM MgC12, 0.2mM dithioerythritol, 0.1 mM CaCl,, 0.01 YO NaN3; aliquots (60 pl) were taken at various times and assayed for Mg2+ATPase activity. Analytical ultracentrifugation experiments carried out on the KI-depolymerized covalent acto-S 1 species were conducted in a MSE analytical ultracentrifuge supplied with a monochromator and a photoelectric scanner as previously described [27]. All experiments were performed at 20"C, in 0.6 M KI, with 0.7-0.8 mg/ml total protein concentration.
0.6 M KI, 1 mM NaN3, pH = 7.5), adsorbed for 60 s on the same kind of electron-microscopy-support films, washed first with six drops of depolymerization buffer and then with 1 mM sodium phosphate at pH 7.5 to prevent repolymerization of oligomers on the grids and subsequently negatively stained as described. The specimens were examined in a Jeol 100 C electron microscope equipped with an anti-contamination device at a beam voltage of 100 kV and a magnification of 50000 x . Pictures were taken on Agfa Gevaert scientific film and the negatives were enlarged three times to 150000 x as indicated in the legend of Fig. 4. The magnification was calibrated with catalase crystals. RESULTS Stability of the Mg2+-ATPaseactivity of the covalent acto-SI complex
Because the covalent acto-S1 complex is a useful tool for future studies of the kinetic properties of S-1 activation by actin, it seemed necessary to define the relative stability of its elevated Mg2 +-ATPase activity. We performed storage experiments of the acto-S1 preparation in different solutions and for various intervals of time. Just before storage, the acto-S1 complex was submitted to a sonication step which provided a practical means for homogenization of the protein suspension. Fig. 1 illustrates the influence of the storage conditions on the Mg2'-ATPase activity of the complex. This activity tended to decrease quickly, reaching 30% of the original value after 48 h at pH = 8.0 and 4°C. A reversible acto-S1 preparation (molar ratio actin: S-1 = 5) was also stored under similar conditions, as a control; the actin-activated ATPase was found after
t 30
I R
m
. ...
0, Lo
0
a
4: 4
0 15
-
Electron microscopy
Negative staining of performed acto-S1 complexes and carbodiimide-polymerized acto-S1 oligomers was performed essentially as described by Craig et al. [28]. F-actin-containing samples were diluted with thin filament buffer (2 mM Tris/HCl, 2 mM MgCl,, 0.1 mM CaC12,50 mM KCl, 1 mM NaN3, 0.2 mM 2-mercaptoethanol, pH = 7.58.0) to give a protein concentration of 50 pg/ml. One drop of diluted sample containing the decorated filaments was adsorbed for 60 s on either glow-discharged carbon-coated grids, silicium-monoxide-coated carbon grids [28] or carbon grids which had been made hydrophobic by holding them over hot molten dental wax for 15 s [29]. Subsequently the grids were washed with six drops of freshly prepared and filtered 1YOuranyl acetate (4°C) for 45 s and then blotted with a filter paper. The depolymerized oligomers were diluted to 10 pg/ml with the depolymerization buffer (2 mM Tris/HCl,
OO
10
20 Time (days)
30
Fig. 1. Influence of storage upon the Mg2+-ATPase activity of the covalent acto-Sl complex. After dissociation of uncross-linked S-I by Mg2+-PPi, the cross-linked acto-S1 complex (4.4 mg proteins/ml), isolated by centrifugation following 15-min reaction of S-I with carbodiimide-treated F-actin, was first submitted to sonication in 2 mM Tris/HCl, 50 mM KCl, 5 mM MgC12,0.2 mM dithioerythritol, 0.1 mM CaC12, 0.1 mM ATP, 0.01% NaN3, pH = 8.0. It was then stored in the presence of 50% ethylene glycol at -25°C ( O ) , 20% sucrose at 4°C (CI), and at 4°C without addition (A). All ATPase assays were performed on the same acto-S1 cross-linked preparation. Experiments were conducted on at least three different acto-S1 preparations with similar results
394 shows that the ATPase loss of the acto-S1 complex was fully reversible. Up to 90% of the elevated Mg2'-ATPase was progressively recovered after elimination of the depolymerizing reagent by a rapid dialysis. On centrifugation of the solution after a 2-h dialysis, all the ATPase activity was found associated with the protein pellet. NH4Cl (4.7 M) was also tested as depolymerizing agent and the repolymerization of the acto-S1 complex by fast dialysis under similar experimental conditions gave the same results, although the recovery of the Mg2+-ATPase activities was less extensive. - 0 15 0 60 120 180 Finally, depolymerization of the acto-S1 complex with 0 5 Time lminl concomitant loss of ATPase activity was also observed Fig. 2. Influence of the polymerization state of actin on the MgZf- without addition of any depolymerizing factor. It was brought ATPase activity of the covalent acto-Sl complex. (A) Suppression about by a 48-h dialysis against a low-ionic-strength solution of the elevated Mg2+-ATPase upon F-actin depolymerization. The containing 2 mM Tris/HCl, 0.2 mM ATP, 0.2 mM acto-S1 complex (4.4 mg/ml) was incubated in 50 mM Tris/HCl dithioerythritol, 0.1 mM CaC12, 0.1 mM NaN3, pH = 8.0. (pH = 7.5),0.6 M KI, 5 mM sodium thiosulfate, 0.5 mM 2-mercap- However, the extent of depolymerization was less extensive toethanol, 0.5 mM ATP, 0.5 mM CaC12, at 4°C. At various interval than in the presence of the depolymerizing agents. Thus, treattimes aliquots (60 pl) were assayed for Mg2+-ATPase activity as described under Materials and Methods. (B) Restoration of the elevat- ment of the covalent complex with KI proved to be the best ed MgZ+-ATPaseupon F-actin polymerization. The KI-treated acto- approach for studying the depolymerization-repolymerizaS1 preparation was dialyzed in an open bag at 4°C against 2 mM tion process of the cross-linked material. These experiments Tris/HCl (pH = 8.0), 50 mM KCI, 5 mM MgCI2, 0.2 mM demonstrate directly the dependency of the Mg2+-ATPase dithioerythritol, 0.1 mM CaCI2,0.01% NaN3. At the times indicated, activation on the polymerized state of F-actin.
t
-
A
samples (60 pl) were withdrawn and submitted to Mg2+-ATPase activity measurements
Influence of the cross-linked S-l on the structure of F-actin
To assess the effect of S-1 cross-linking on the structure of the actin filament, we have attempted to characterize the species formed under the depolymerizing conditions described above. For this purpose, we throughly investigated the KIproduced acto-S1 solutions. First the preparation was fractionated on Sephacryl S-300. The corresponding elution profiles are depicted in Fig. 3A. The F-actin control was totally converted into G-actin (Kav0.47). The carbodiimidetreated F-actin control was also extensively depolymerized giving rise to G-actin and, as expected, to some cross-linked actin oligomers. In contrast, the depolymerization of the covalent acto-S1 in 0.6 M KI led to only a small amount of G-actin with K,, 0.47; this limited amount of free G-actin suggests that only a few actin-actin bonds were broken within the actin filament cross-linked to S-3. Most of the actin material was eluted as a broad peak composed of actin oligomers containing the cross-linked S-1, as indicated by the Influence of the depolymerized-repolymerized states of actin electrophoretic analysis of the major fraction with Kay0.15 on the Mg2+-ATPaseactivity of the acto-Sl complex (Fig. 3C, lane a). This fraction contained a noticeable amount The activation of the Mg2+-ATPaseof S-1 is effected by of noncross-linked G-actin together with the cross-linked F-actin and not by G-actin [30]. In an attempt to obtain a actin - heavy-chain products with apparent masses of better insight into the relationship between the filamentous 180 kDa and 265 kDa [7]. For comparison, cross-linked acto-S1 treated with NH4C1 structure of actin and the great enhancement of the Mgz+ATPase activity in the cross-linked acto-S1, we investigated was also investigated (Fig. 3B). It gave the same the influence of actin depolymerization on the activity of the chromatographic profile as observed with KI except that covalent complex. Fig. 2A shows the time course of change oligomers with smaller size were formed as indicated by the in the Mg2+-ATPase upon treatment of the complex with elution volume of the major fraction with K,, = 0.27 and the 0.6 M KI. This enzymatic activity decreased quickly. Up to decreased amount of G-actin released from this fraction on 72% of the initial activity was lost after a 5-min incubation in the electrophoretic gels (Fig. 3C, lane b). In the G-actin posithe KI solution. After a 20-min reaction, the activity reached a tion of the chromotogram, a higher amount of free G-actin plateau corresponding to 20% of the original value. This was eluted together with dissociated light chains as judged by residual activity was related to a residual amount of acto-S1 gel electrophoresis of the protein peak (data not shown). In complex which could not be depolymerized. After centrifuga- contrast, the treatment with KI did not induce the dissociation tion a small pellet was discarded and the supernatant con- of the light chains. Although the depolymerization with tained about 80% of the starting protein material with no NH4CI was more efficient than in the presence of KI, these activated Mg2' -ATPase. The observed ATPase inhibition was overall data indicate that the permanent binding of S-1 marknot due to damage of S-I by the KI solution because the actin- edly increases the stability of the actin filaments. The KIactivated ATPase of the control S-1 incubated under similar treated acto-S1 solutions were further analyzed by analytical conditions was not appreciably affected. Moreover, Fig. 2 B centrifugation which provided the hydrodynamic parameters 7-days storage to have about 80% of the original value. The loss of ATPase activity of the covalent complex could be retarded by addition of 20% sucrose. After 10-days storage, 25% of the enzymatic activity was maintained in its presence. Because the decrease of the activity is fast in this storage medium, the enzymatic activities were no longer analyzed after the tenth day. In contrast, the stability of the complex was much higher when it was stored in 50% ethylene glycol at -25°C. After 7 days the activity remained constant and decreased thereafter at a slow rate reaching 50% of the original value after 30-days storage. Thus, cross-linked acto-S1 preparations still had a high Mg2I-ATPase activity when stored at sub-zero temperatures in the presence of a cryosolvent.
395 Table 1. Hydrodynamic properties of carbodiimide-treated F-actin and the whole covalent acto-SI preparation after depolymerization by 0.6 M KI The experimental conditions were as reported in Materials and Methods. f was an assumed value reported for G-actin [48] Parameters
0.5
0.25
t
L.7M NH,CI
"KO" 0.27
1
200 V, (rnll
a
/
B
1
300
b
Fig. 3. Chromatographic and electrophoretic characterization of the acto-Sl oligomers. 4 ml of the acto-S1 cross-linked preparation (4.4 mg/ml) in 50 mM Tris/HCl, pH = 7.5, 5 mM sodium thiosulfate, 0.5 mM 2-mercaptoethanol, 0.5 mM ATP, 0.5 mM CaCI2 containing 0.6 M KI (A) or 4 ml of the acto-S1 cross-linked preparation (4.4 mg/ ml) in 50 mM imidazole/HCl, pH = 7.0, 0.5 mM dithioerythritol, 0.1 mM ATP, 0.5 mM CaCI2, 0.01% NaN3 containing 4.7 M NH4CI (B) were passed on a column of Sephacryl S-300 (150 x 1.9 cm) equilibrated with the same buffer. (-.-.-) F-actin control; (.....) carbodiimide-treated F-actin control; (0-0) cross-linked actoS1. The peak fractions with K,, 0.15 and 0.27 were analyzed by sodium dodecyl sulfate/polyarylamide slab gel electrophoresis (5 18% gradient) (C) lane a, KI-produced fraction; lane b, NH,Clproduced fraction
shown in Table 1, The diffusion coefficient and the frictional ratio obtained for acto-S1 samples corresponded to a more elongated form as compared to those found for the carbodiimide-treated F-actin control. From these hydrodynamic data, an average molecular mass of 285 kDa was estimated for the oligomers. This value correlates with a
Sedimentation coefficient, s!j0,,, Anhydrous partial specific volume, V Molecular mass, M , derived from sedimentation-equilibrium Stokes radius: by ultracentrifugation by gel filtration Frictional ratio (derived from s, m and V )
Carbodiimidetreated F-actin
Acto-S1
3.1 S
11.7 S
0.732 ml . g-'
0.732 ml . g-'
42 kDa
285 kDa
3.16 nm 2.80 nm
5.75 nm 5.70 nm
1.37
1.32
Stokes radius of 5.7 nm which was independently determined by gel filtration chromatography for the major fraction eluting with K,, 0.27. Finally the KI-produced acto-S1 oligomers were examined by electron microscopy in comparison with the intact covalent acto-S1 complexes formed before and after the depolymerization-repolymerization cycle. The original actoS1 complex was constituted of a partially cross-linked meshwork of actin filaments that could only be processed for electron microscopy with great difficulty. However, the complex gave better pictures when it was repolymerized after the dissociation step of the residual S-1 by magnesium pyrophosphate (Fig. 4a). These complexes showed partial decoration as compared to the reversible acto-S1 (Fig. 4b) used as control, and the covalent binding of S-1 to actin seemed to be more or less random. Similar images were obtained with under-decorated reversible acto-S1 . When S-1 is lying on top of the actin cr-helix it seems to bind across the cr-helix at an angle of about 30". The repolymerization was not always complete since one can see some oligomers in the repolymerized preparations. Also the repolymerized filaments appeared relatively shorter. Electron microscopic studies (Fig. 4) carried out on four different protein fractions (F, F4) eluted from the column of Sephacryl-S-300 (Fig. 3A) showed a close correlation between the elution volume and the size of the oligomers; whithin each fraction, there seems to be some heterogeneity in size but the trend from larger to smaller-sized oligomers was clear in F1- F4 (Fig. 4, c - 0. In the smallest oligomers one cannot distinguish S-3 from the actin material. Modulation of the elevated Mg2 '-ATPase activity of the covalent acto-SI complex by tropomyosin and the tropomyosin-troponin system
We first studied the regulatory function of the tropomyosin alone upon the high hydrolytic activity of the irreversible acto-S1 complex. For this purpose the cross-linking reaction was allowed to proceed between carbodiimideactivated acto-tropomyosin and S-1 . Gel electrophoresis gave no evidence for any cross-linking between tropomyosin and actin or between tropomyosin and S-1. At the selected times, the reaction wa-s terminated by magnesium pyrophosphate,
396
Fig. 4a, b
the covalent acto-S1 formed was isolated and its MgzcATPase activity was assayed as illustrated in Fig. 5. A biphasic time course was observed. During the first period of crosslinking (8 min), the activity was approximately 20% lower in the presence of tropomyosin. Then the ATPase reached values similar to those found in the absence of tropomyosin. Since the observed extents of inhibition can proceed from a decrease in the amount of cross-linked S-I under the direct
influence of actin-bound tropomyosin during the cross-linking reaction, we measured concomitantly the amounts of cross-linked S-1 during the initial and last periods of the reaction after 8 min and 25 min, respectively. The data shown in Table 2 indicated no significant difference in the extent of S-I cross-linking to actin in the absence and presence of tropomyosin during the entire course of the reaction. Furthermore, the same effects of tropomyosin were obtained when it
397
Fig. 4. Electron microscopy of cross-linked acto-SI and acto-SI oligomers. Samples were negatively stained with 1% uranyl acetate. (a) Crosslinked acto-S1 complex after a KI depolymerization-repolymerizationcycle; (b) fully decorated reversible acto-S1 complex; (c - f) act031 oligomers corresponding to fractions F1 to F4 respectively, eluted from the Sephacryl S-300 column in the presence of0.6 M KI as described in Fig. 3 A ; ( g ) F-actin control. Magnification 150000 x . Bar represents 100 nm
was added to the ATPase assays after cross-linking of S-I to actin. Thus, tropomyosin seems to inhibit the rate of the elevated Mg2+-ATPase activity of the cross-linked S-I and this effect appears to be dependent on the extent of S-1 crosslinking; it was apparent only when up to 50% of the added S-1 was cross-linked (about 0.25 mol S-l/mol actin). At a
higher level of cross-linking the influence of tropomyosin was suppressed. A similar study was undertaken using S-I cross-linked to F-actin activated in the presence of tropomyosin and the troponin complex. As observed with tropomyosin alone, only the cross-linking between S-I and F-actin occurred. As
398 illustrated in Fig. 5, the Mg2+-ATPaseactivity of the covalent complex was markedly decreased over the entire course of the reaction, when it was assayed in the presence of EGTA; the inhibition reached maximally about 50% in the first period of the reaction and decreased to about 20% in the last period of cross-linking. However, when the Mg2+-ATPase assays were carried out in the presence of 0.5 mM CaC12, the activity was as high as that measured in the absence of regulatory proteins for the covalent acto-S1 complex used as control. Activation was also observed at 0.1 mM Ca2+. The Ca2+ concentrations employed have no apparent effect on the ATPase activity of the covalent complex lacking troponin ; also the EGTA concentration used (1 mM) has no impact on the Mg2+-ATPase activity of the cross-linked acto-S1. We conclude that the elevated Mg2+-ATPase activity of the crosslinked carbodiimide-produced acto-S1 can be modulated by the tropomyosin-troponin system in a Ca2+-dependent manner.
I
I
N
m
I
0' 0
20 Time ( m i d
10
30
LO
Interaction of tropomyosin with the covalent acto-S1 complex Because recent three-dimensional image reconstructions on the rigor F-actin -tropomyosin - S-1 complex indicate that the tropomyosin is bound close to the S-1 on the F-actin filament [31], we investigated the influence of the chemical cross-linking of S-1 to F-actin on the actin-tropomyosin interaction. The experimental approach we used in this study is based on the change in the tryptic susceptibility of tropomyosin upon its association with F-actin. As illustrated in Fig. 6A, the limited digestion of tropomyosin by trypsin resulted in characteristic peptides previously identified by Pato and Smillie [32]. The two main fragments formed, T1 and T2, with masses of 13 kDa and 17 kDa, correspond to the NH2- and COOH-terminal segments of tropomyosin (residues 1 - 133 and 134-284 respectively); T2 was further degraded at its NH2-terminal portion giving rise to a smaller peptide of 11 kDa (T4) [32]. The binding of F-actin to tropomyosin induced a dramatic change in the cleavage pattern and resulted in a total protection of tropomyosin against proteolysis (Fig. 6B). The presence of reversibly bound S-1 did not change the protective effect of actin on tropomyosin and conversely the tryptic cleavage pattern of acto-S1 was not affected by tropomyosin (Fig. 6B, lanes a, b); the S-1 heavy chain was cut only at the 27-kDa - 70-kDa connector segment as observed in the absence of tropomyosin [8]. Also, the cross-linking of S-I to the preformed F-actin tropomyosin complex did not alter the protective action of F-actin against the tryptic digestion of tropomyosin (Fig. 7A). In contrast, when tropomyosin was added to the isolated acto-S1 cross-linked preparation, two new features were noticed (Fig. 7B). First, the amount of tropomyosin bound to the complex was higher than expected from the stoichiometric ratio of 1 :7. Quantitative binding studies with '251-tropomyosin indicated that 2 - 3 mol tropomyosin were able to bind to 7 mol actin in the covalent acto-S1 complex; on the other hand, upon tryptic digestion only part of the bound tropomyosin was protected by F-actin whereas the remainder was degraded into the tropomyosin peptides. These results suggest that the primary covalent attachment of S-1 to F-actin modifies the subsequent binding of tropomyosin to F-actin.
Fig. 5. Ca2+-dependent regulation of the elevated Mg2+-ATPaseactivity of the covalent acto-Sl complex by the tropomyosin - troponin system. Actin was cross-linked to S-I in the absence and in the presence of tropomyosin and tropomyosin - troponin as indicated under Materials and Methods. At the times indicated the covalent actoS1 complex formed was isolated by dissociation with magnesium pyrophosphate and was assayed for Mg2+-ATPase activity. (m) Mg2+-ATPase of S-I cross-linked to F-actin- tropomyosin; ( 0 ) Mg2 -ATPase of S-I cross-linked to F-actin - tropomyosin - DISCUSSION troponin, determined in the presence of 1 mM EGTA; (0)M g 2 + The chemical cross-linking of the two major proteins inATPase of S-I cross-linked to F-actin - tropomyosin- troponin, the functional determined in the presence of 0.5 mM CaCI2. (A)MgZf-ATPaseof volved in muscle contraction, actin and part of the myosin molecule, offers the opportunity to study S-I cross-linked to F-actin +
s-I,
Table 2. Ejject of the binding of tropomyosin on the elevated Mg2+-ATPaseactivity of the covalent acto-S1 complexes and on the amount of cross-linked S-1 For acto-Tm-S1, the F-actin and S-I were cross-linked in the presence of tropomyosin, as indicated in Materials and Methods. For acto-S1Tm, tropomyosin was added to the preformed covalent acto-S1 complex (molar ratio F-atin/tropomyosin = 7 : 1). Mg2+-ATPasewas measured as pmol P i . min-' . (mg S-l)-' Complex
Acto-S1 Acto-Tm-S1 Acto-S1-Tm Acto-S1 Acto-Tm-S1 Acto-S1-Tm
MgZ -ATPase
Inhibition
S-I cross-linked to F-actin
min
pmol. min-' . mg-'
%
% initial
8 8 8 25 25 25
29 23 22 37 37 38
0 21 23 0 0 0
41 51 54 65 68 15
Time of crosslinking reaction
+
399
B
A
Fig. 6. Cleavage of'tropomyosin by trypsin. (A) Time course of the limited cleavage of tropomyosin by trypsin. Tropomyosin (I mg/ml) was incubated in 50 mM Tris/HCl, pH = 8.0, 30 mM KCI and 5 mM MgClz at 25"C, at a protease/substrate ratio of 1 : 500 (w/w). At the times indicated, protein aliquots were subjected to electrophoresis in 10-23% gradient polyacrylamide slab gel. (B) Time course of the tryptic digestion of tropomyosin - F-actin complex. Tropomyosin was treated with trypsin as in (A), in the presence of F-actin (molar ratio 1 :7). The tropomyosin-F-actin complex was also digested in the presence of S-I (2 mg/ml); the gel pattern of the ternary complex before (lane a) and after a 30-min reaction with trypsin (lane b) is shown. Electrophoresis was conducted as in (A). Tm = tropomyosin. Numbers represent molecular mass in kDa
A
6
Fig. 7 . Tryptic susceptibility of tropomyosin associated with the covalent acto-Sl complex. (A) The carbodiimide-treated F-actin - tropomyosin complex was reacted for 15 min with S-1 as indicated under Materials and Methods. The S-I - F-actin- tropomyosin cross-linked preparation was recovered in the pellet obtained after centrifugation of the reaction mixture in the presence of magnesium pyrophosphate. It was subjected to trypsin under the conditions reported in Fig. 6A. (B) The covalent acto-S1 complex (3 mg protein/ml) was incubated with tropomyosin for 1 h at 2 5 T , with continuous stirring in 10 mM Tris/HCl buffer, pH = 8.0, 2 mM ATP, 30 mM KCI, 5 mM MgCl2, 1 mM EGTA (Factin/tropomyosin molar ratio = 3). After centrifugation for 2 h at 170000 x g, the pellet was treated with trypsin as specified in Fig. 6A. Samples were taken at the times indicated and subjected to electrophoresis in 5- 18% gradient polyacrylamide slab gel. (P), Tryptic digest of tropomyosin; HC, heavy chain; numbers represent molecular mass in kDa
structure-function relationships within the rigor complex in vitro. The importance of the covalent actin-S1 complex lies in the permanent occupation of the actin site(s) on S-I and the expression by this complex of an enhanced Mg2+-ATPase. The data obtained in the present work indicated that in
aqueous medium the Mg2+-ATPase of the covalently bound S-I is rather labile; however it could be efficiently stabilized by the use of a cryosolvent such as ethylene glycol, and a low temperature. Since the Mg2+-ATPase loss is time-dependent and could be significantly retarded by ambiant factors, it is
400 unlikely that it was the result of the chemical reaction itself, but rather this feature seems to be an intrinsic property of the S-I covalently attached to F-actin. Indeed, it is an essential and well known property of F-actin to promote conformational changes in the S-I which open the ATPase site with concomitant acceleration of the rate of product release [33]. In contrast, in the absence of F-actin, the S-I ATPase site closes in upon itself and becomes less exposed to the environment [34]. It was shown that actin increases the rate of dissociation of EATP and the accessibility of the bound nucleotide within cross-linked acto-Sl [35]. Thus it is logical to assume that the ATPase site of the covalently attached S-I bears a permanently open structure which makes it more labile and more sensitive to protein denaturation processes which can be minimized by the use of cryosolvents. Therefore highly active act041 cross-linked preparations can be conveniently used for sub-zero kinetic investigations. The ability of the acto-S1 cross-linked complex to undergo a depolymerization-repolymerization cycle, as does actin alone, is in agreement with the brief report of Ikkai and Dreizen [36]. A striking feature of the covalent acto-S1 concerns the reversible loss of its elevated ATPase following F-actin depolymerization-repolymerization.These results directly illustrate the relationship between the extent of the Mg2+-ATPase activation and the filamentous structure of the actin. On the basis of the observations that neither G-actin nor G*-actin (or F-actin monomer) can activate the Mg2+ATPase activity [30] and that the acto-S1 oligomers resulting from the depolymerization of cross-linked acto-S1 were also inactive, it can be concluded that the F-actin configuration in the filament is a prerequisite for the stimulation of the Mg2+ATPase activity of the bound S-I. S-I is known to be an efficient polymerizing agent for actin 1371. The results of the depolymerization experiments on the cross-linked acto-Sl are in agreement with this idea as they indicate that its binding to actin strongly favors actin-actin interactions and gives rise to a large amount of actin oligomers under various depolymerizing conditions. The capacity of actin bound to S-1 to polymerize more readily than free actin explains the increase of the S-I :actin ratio within the cross-linked acto-S1 observed by Greene [38] when the complex was submitted to several cycles of depolymerization-repolymerization.The stabilizing influence of S-I on the F-actin structure is similar to that exhibited by other actin binding factors such as phalloidin which also protects actin against depolymerization by high concentrations of KI [39]. As S-I was cross-linked to F-actin in substoichiometric amounts, its impact on the actin structure obviously could not be uniform along the actin filament. Therefore, actin oligomers of different sizes, together with free G-actin, were produced. The depolymerization experiments were first initiated in an attempt to isolate, by gel chromatography under non-denaturing conditions, the major crosslinked species formed between S-I and one actin subunit [40]. However, the gel filtration patterns of the different depolymerized acto-S1 preparations indicated little if any of this species. If present, the latter should be eluted at the F4 position (KaV0.33). The major acto-S1-containing oligomer corresponding to F2 behaved as an entity with mass of 270 280 kDa. This value would be accounted for only by the presence of about four actin monomers and one S-I molecule. Thus the propagated influence of the bound S-I could reach at least four actin subunits within the F-actin filament. The contraction of vertebrate striated muscle is controlled by the Ca2+-sensitive tropomyosin-troponin system; the tropomyosin strand is thought to compete sterically with the
myosin heads for the myosin binding sites on F-actin [2]. However, recent studies on the influence of tropomyosin and the tropomyosin-troponin complex [6, 41, 421 on the kinetics of the actin-activated ATPase of skeletal muscle myosin have indicated that tropomyosin inhibits the rate of Mg2'-ATP hydrolysis. This inhibition does not result from a decrease of S-1 binding to regulated actin but rather from the blocking of a conformational transition in S-1 associated with the step of Pi release [6]. The effects of the regulatory proteins we observed on the elevated Mg2+-ATPase activity of the crosslinked acto-S1 are in agreement with these conclusions and directly illustrate that tropomyosin can exert its action by modifying the rate of ATPase activation of the covalently bound S-I by actin. Furthermore the extent of ATPase inhibition in the cross-linked acto-S1 is dependent on the amount of covalently bound S-I ; the ATPase is more inhibited at low than at high levels of S-1 cross-linking. This feature was first observed for the reversible S-I - actin- tropomyosin complex and is the manifestation of the S-l-induced activation of the Factin - tropomyosin complex reported by Lehrer and Morris [42]; during this activation it is thought that S-I binding shifts the tropomyosin position on the thin filament from a state associated with ATPase inhibition to a state associated with its activation; also, as expected from the findings of these investigators, the activity measured in the absence of Ca2+for the S-I cross-linked at low extent to the actin - tropomyosin troponin complex increased towards maximal values at the highest extent of S-I cross-linking. The influence of the concentration of the S-1 cross-linked to actin was even more striking in the case of the actin- tropomyosin complex since the inhibition of the elevated ATPase could be suppressed. Moreover Ca2+ acts as an activator of the ATPase of the cross-linked acto-S1 in the presence of the regulatory proteins. This indicates that the cross-linking of S-1 to actin does not abolish the conformational changes associated with F-actin regulation. During this study we used the limited tryptic digestion to probe the binding of tropomyosin to F-actin as compared to its binding to the reversible as well as the covalent F-actinS-I complexes. This approach was founded on the highly specific cleavage of tropomyosin by trypsin. Tropomyosin is also selectively cut by chymotrypsin [43]. The proteolytic susceptibility of tropomyosin can be related to the high flexibility of the molecule suggested by its crystal structure [44]; also other studies have indicated that tropomyosin, in solution, is in equilibrium between a rigid-chain closed state and a flexible-chain open state [45, 461. Its binding to F-actin leads to a total abolition of the tryptic sensitivity which, most probably, is caused by a change in the conformation of the tropomyosin towards a relatively more rigid structure. It could be also that the protection against tropomyosin cleavage was due to a simple steric blocking of the sensitive peptide bonds on combination with actin. It would be interesting to assess the proteolytic susceptibility of the smooth gizzard tropomyosin which was reported to be more rigid as compared to the skeletal tropomyosin [47]. The carbodiimidecatalyzed cross-linking of S-I to F-actin in the presence of tropomyosin can be done without any apparent change in the nature or the yield of the major cross-linked species between the S-I heavy chain and actin. On the other hand, tropomyosin influenced the Mg2+-ATPase activity to the same level whether it was added to F-actin before or after S-I cross-linking. The excess amount of tropomyosin that binds to actin in the latter case could be due to the formation of new interaction areas on F-actin resulting from its primary
40 1 cross-linking with S-I. As suggested by the data shown in Fig. 7 B, the tropomyosin molecules bound to these regions are probably loosely attached to F-actin and are more readily digested by trypsin in contrast to those specifically and tightly complexed to F-actin and which remain refractory to proteolysis. In conclusion, the cross-linked actoS1 can serve as a valuable tool not only for the study of the kinetic aspects of the ATPase cycle but also for investigating the dynamic conformational changes taking place in the thin filament during its regulation by the tropomyosin-troponin complex in the presence of S- 1. This research was supported by grants from the Centre National de la Recherche Scientifique, the Direction Gtnkrale de la Recherche et de la Technologie (convention 5-11834), the Institut National de la Santh et de la Recherche Mkdicale (C.R.E. 5-11850), the Fritz Hofmann-La Roche foundation zur Forderung wissenschaftlicher Arbeitsgemeinschaften, and an Eidgenossische Technische Hochschule training grant.
REFERENCES 1. Huxley, H. & Hanson, J. (1954) Nature (Lond.) 173, 973-976. 2. Huxley, H. E. (1969) Science (Wash. DC) 164, 1356-1366. 3. Sheetz, M. P. & Spudich, J. A. (1983) Nuture (Lond.) 303, 31 35. 4. Huxley, H. E. (1972) Cold Spring Harbor Symp. Quant. Biol. 37, 361 - 376. 5. Haselgrove, J . C. (1972) Cold Spring Harbor Symp. Quant. Biol. 37, 341 -352. 6. Chalovich, J. M. & Eisenberg, E. (1982) J . Biol. Chem. 257, 2432- 2437. 7. Mornet, D., Bertrand, R., Pantel, P., Audemard, E. & Kassab, R. (1981) Nature (Lond.) 292, 301 -306. 8. Mornet, D., Bertrand, R., Pantel, P., Audemard, E. & Kassab, R. (1981) Biochemistry 20, 2110-2120. 9. Labbe, J. P., Mornet, D., Roseau, G. & Kassab, R. (1982) Biochemistry 21, 6897-6902. 10. Shukla, K. K., Levy, H. M., Ramirez, F., Marecek, J. F., McKcever, B. & Margossian, S. S. (1983) Biochemistry 22, 4822 - 4830. 11. Webb, R. M. & Trentham, R. (1982) Fed. Proc. 41,6702. 12. Rouayrenc, J. F., Bertrand, R., Mornet, D., Pantel, P., Audemard, E. & Kassab, R. (1982) J . Muscle Res. Cell Moril. 3,464. 13. OfTcr, G., Moos, C. & Starr, R. (1973) f. Mol. Biol. 74, 653676. 14. Weeds, A. G. & Taylor, R. S. (1975) Nature (Lond.) 257, 5456. 15. Wagner, P. D. & Weeds, A. G. (1977) J . Mol. Biol. 109, 455473.
16. Eisenberg, E. & Kielley, W. W. (1974) J . Biol. Chem. 249,47424748. 17. Taylor, R. S. &Weeds, A. G . (1976) Biochem. J . 159,301 -315. 18. Greaser, M. L. & Gergely, J. (1971) J . Biol. Chon. 246, 42264233. 19. Hartshorne, D. J. & Mueller, H. (1969) Biochim. Biophy.7. Acta 175, 301 - 309. 20. Yang, Y. Z., Gordon, D. J., Korn, E. D. & Eisenberg, E. (1977) J . Biol. Chem. 252, 3374- 3378. 21. West, J. J., Nagy, B. & Gergely, J. (1967) J . Biol. Chem. 242, 1140- 1 145. 22. Cummins, P. & Perry, S. V. (1973) Biochem. J . 133, 765 - 777. 23. Korn, E. D. (1982) Physiol. Rev. 62, 672-7737, 24. Woods, E. F. (1967) J . Bid. Chem. 242, 2859-2871. 25. Eaton, B. L., Kominz, D. R. & Eisenberg, E. (1975) Biochemistry 14,2718-2725. 26. Weber, K. & Osborn, M. (1969) J . Bid. Chem. 244,4406-4412. 27. Roustan, C., Fattoum, A., Jeanneau, R. & Pradel, L. A. (1980) Biochemistry 19, 5168 - 5175. 28. Walzthony, D., Bahler, M., Wallimann, T., Eppenberger, H. M. & Moor, H. (1 983) Eur. J . Cell Biol. 30, 177- 181. 29. Craig, R., Szent-Gyorgyi, A. G., Beese, L., Flicker, P., Vibert, P. & Cohen, C. (1980) J. Mol. Biol. 140, 35-55. 30. Estes, E. J. & Gershman, L. C. (1978) Biochemistry 17, 24952499. 31. Taylor, K. A. & Amos, L. A. (1981) J . Mol. Bid. 147,297-324. 32. Pato, M . D. & Smillie, L. B. (1978) FEBS Letr. 87, 95-98, 32. 33. Lymn, R. W. & Taylor, E. W. (1971) Biochemistry 25, 46174623. 34. Biosca, J. A., Travcrs, F., Hillaire, D. & Barman, T. (1984) Biachemistry 23, 1947- 1955. 35. Rosenfeld, S. S. & Taylor, E. W. (1983) Biophys. f. 41, 301 a. 36. Ikkai, T. & Dreizen, P. (1984) Biophys. J . 45, 282a. 37. Yagi, K., Mase, R., Sakakibara, I. & Asai, H. (1965) J . Biol. Chem. 240,2448 - 2454. 38. Greene, L. E. (1984) Biophys. J. 45,222a. 39. Wieland, Th. & Faulstich, H. (1978) CRC Crit. Rev. Biochem. 5 , 185 - 260. 40. Sutoh, K. (1983) Biochemistry 22, 1579- 1585. 41. Sobieszek, A. (1982) J . Mol. Biol. 157, 275-286. 42. Lehrer, S. S. & Morris, E. P. (1982) J . Bid. Chem. 257, 80738080. 43. Pato, M. D., Mak, A. S. & Smillie, L. B. (1981) J . Bid. Chem. 256, 593 - 601. 44. Phillips, G. N . , Jr, Lattman, E. E., Cummins, P., Lee, K. Y. & Cohen, C. (1979) Nature (Lond.) 278,413-417. 45. Graceffa, P. & Lehrer, S. S. (1980) J . Biol. Chem. 255, 11 29611300. 46. Lehrer, S. S., Betteridge, D. & Graceffa, P. (1981) Ann. N. Y . Acad. Sci. 366, 285 -299. 47. Betteridge, D., Graceffa, P., Lehrer, S. S. & Seidel, J. (1983) Biophys. J . 41, 104a. 48. Mihashi, K. (1964) Arch. Biochem. Biophys. 107,441 -448.
J.-F. Rouayrenc, R. Bertrand, and R. Kassab, Centre de Recherche de Biochimie Macromolkculaire du Centre National de la Recherche Scientifique, Boite postale 5051, F-34033 Montpellier-Cedex, HCrault, France D. Walzthony, M. Bahler, and T. Wallimann, Institut fur Zellbiologie der Eidgenossischen Technischen Hochschule Zurich, Zurich-Honggerberg, CH-8093 Zurich, Switzerland
