- Email: [email protected]
The myosin filament XIII. The sensitivity of LMM assembly to Mg·ATP
Biochimica et Biophysica Acta, 997 (1989) 182-187
182
Elsevier BBAPRO33451
The myosin filament XIII. The sensitivity of LMM assembly to Mg-ATP P r o k a s h K. C h o w r a s h i a n d F r a n k A. P e p e Department of Anatomy, School of Medicine, University of Pennsylvania, Philadelphia, PA (U.S.A.)
(Received 7 February1989)
Key words: Myosin;LMM; Mg. ATP; Assembly
An LMM fragment (M, 62 000) of myosin has been prepared which has aggregation properties that are sensitive to the presence of Mg- ATP. Aggregation of the LMM by reducing the ionic strength in the presence of 1 mM Mg- ATP produces non-periodic aggregates which gradually rearrange to paracrystals with a 43 nm axial repeat pattern. This fragment includes the C-terminal end of the myosin rod starting at residue 1376. Therefore, at least one of the Mg. ATP binding sites responsible for this effect is located somewhere along this region of the myosin rod. Although assembly of the rod fragment of myosin into paracrystais does not show sensitivity to Mg. ATP, assembly of intact myosin molecules to form filaments does show sensitivity to Mg- ATP. For myosin filaments, assembly initially gives a broad distribution around a mean length of 1.5 /zm, which sharpens around the mean length with time. The rearrangement of the ~ rods and intact myosin molecules both induced by the presence of Mg- ATP are probably related. These findings highlight the complexity of the cooperative interactions between different portions of the myosin molecule that are involved in determining the assembly properties of the intact molecule.
Introduction The assembly of myosin molecules in vitro has been shown to produce filaments which are structurally similar to native filaments in that they have a bare zone region of constant length in the middle of the filament and they have tapered ends [1]. However, the length of the filaments varies widely and this variation becomes greater with increasing length of the filaments [2]. Recently, a dilution procedure has been described which produces a length distribution around the native length of 1.5/tin with a standard deviation of _+0.2 pm [3,4]. The question of whether or not native filaments are themselves precisely determined in length or whether there is a narrow distribution of lengths cannot be answered unequivocally, since on separating the filaments for measurement, some loss of molecules from the ends may occur. Measurement of the A-band width may not correspond to filament length if there is some stagger between the filaments in the A-band lattice. The best measurements of native filament lengths that have been made are those of Morimoto and Harrington [5],
Correspondence: F.A. Pepe, Department of Anatomy, School of Medicine, Universityof Pennsylvania, Philadelphia, PA 19104-6058, U.S.A.
in which the distribution of lengths measured (mean of 1.53 pm with 80~ within ±0.17/tin) compares favorably with that obtained by in vitro assembly [3,4]. The assembly of myosin in vitro was found to be influenced by the presence of M g - A T E It was shown that the addition of 1 mM M g - A T P to a previously assembled filament suspension would lead to a redistribution in the lengths so that the distribution was sharpened around the native length of about 1.5 ttm [4]. This occurred even for variations of the assembly procedure which gave very wide distributions in the length of the assembled filaments. It was also found that if the assembly of myosin was done in the presence of 1 mM Mg-ATP a relatively broad distribution in length was obtained initially with redistribution on standing in the cold (0-4°C) to give a sharper distribution around the native length of 1.5 pm. Some experiments [4] have also shown that the minimum concentration of M g - A T P required to effect this redistribution of the lengths of the filaments was about 500/tM for a concentration of myosin of about 1 ~tM, therefore excluding the stoichiometric binding of the Mg. ATP to the myosin ATPase active site as being involved in the length-determining mechanism. Harrington and Himmelfarb [6] and Bartels et al. [7] have suggested that there are other binding sites for Mg. ATP on the myosin molecule. It is noteworthy that the concentration of M g - A T P in riving
0167-4838/89/$03.50 © 1989ElsevierSciencePublishersB.V.(BiomedicalDivision)
183 muscle cells [8] is within the range of that required for filament length redistribution of the in vitro assembled myosin filaments. Since a fragment of the myosin molecule is a part of the intact molecule, the assembly characteristics of the fragment must contribute to the d~termination of the assembly characteristics of the whole molecule. LMM assembles into paracrystals with the LMM rods arranged anti-parallel [9,10]. An anti-parallel packing arrangement is characteristic only of the bare zone region of the myosin filament, the rest of the assembly being by parallel alignment. However there must be a direct relationship between the anti-parallel and parallel assembly information which contributes to the filament assembly characteristics and which involves the transition from one to the other as the filament assembles beyond the bare zone region. In this study an LMM fragment has been identified (residues 1376-1939) with assembly properties sensitive to the presence of M g - A T E Its assembly has been compared with that of the myosin rod and with the assembly of the intact myosin molecule. Since the intact molecule has also been shown to be sensitive to the presence of Mg-ATP [4], it is likely that the identified peptide region of the myosin.rod of M r 62000 (LMM) used in this study contributes to that sensitivity.
Preparation of the rod fragment The rod fragment of myosin was prepared by the digestion of i~,soluble myosin with papain (Worthington-Cappel) e~entially as described previously [12]. The column-purified rabbit skeletal myosin, at a concentration of 12 mg/ml0 was dialyzed against 0.2 M ammonium acetate (pH 7.0). Under these conditions, the myosin is insoluble. To this suspension, papain was added to about 0.3 mg/ml or less. The reaction was allowed to proceed at room temperature for 15 min. Digestion was stopped by the addition of 0.1 M iodoacetic acid to a final concentration of 1 mM and the pH was adjusted to 6.0 to stop the reaction. The suspension was then centrifuged and the pellet was washed twice by re-suspension in 0.2 M ammonium acetate (pH 7.0). The washed pellet was dissolved in 0.6 M KC1/0.03 l,l phosphate buffer (pH 6.5). After solution, three volumes of ethyl alcohol were added and stirred at room temperature for 3 h. The precipitate was collected by centrifugation, dissolved in 0.6 M KCI/10 mM imidazole (pH 6.8) and dialyzed in the same buffer.
SDS-PA GE electrophoresis The procedure used for SDS-gel electrophoresis was the same as that previously described [13]. In these studies, 7.5% polyacrylamide gels were used.
Assembly of the LMM and rod fragments of myosin Materials and. Methods
Myosin was isolated from rabbit back and leg muscles and purified by column chromatography on DEAE Sephadex A50 as described previously [4]. The purified myosin was used for the preparation of the LMM and rod fragments of myosin as described below,
Preparation of the LMM fragment Column purified rabbit myosin was digested with ot-chymotrypsin (Sigma) as previously described [11]. The myosin at a concentration of 8 mg/ml in 0.5 M ammonium carbonate (pH 8.0) was digested with achymotrypsin in a weight ratio of myosin to a-chymotrypsin of 100:1 for 30 rain at room temperature. The reaction was stopped by adding phenylmethylsulfonyl fluoride (PMSF) to a final concentration of about 1 raM. The digest was dialyzed against 0.05 M KCI/5 mM phosphate buffer (pH 6.2) overnight. The precipitate was collected by centrifugation and dissolved in 0.6 M KCI/10 mM imidazole (pH 7.0). Three volumes of ethanol were added to this solution and this was stirred for 3 h at room temperature. The precipitate was collected and re-suspended in 0.6 M KCI/10 mM imidazole (pH 6.8) and dialyzed in the same buffer overnight. After dialysis, the solution was clarified by centrifugation.
Assembly of the LMM and rod fragments was carried out as previously described for myosin filaments [3,4]. The standard dilution procedure was used in all cases. The LMM or rod was at a starting concentration of 2 mg per ml in 0.6 M KCIp_0 mM imidazole (pH 6.8) either with or without the presence of 1 mM Mg ATP. The KCI concentration was reduced in two steps, from 0.6 to 0.3 M KCI and then from 0.3 to 0.15 M KCI. Starting with 0.5 ml of protein, 10 mM imidazole buffer (pH 6.8) was added in volumes of 35.3 /~1 at intervals of 5 s with 14 additions, bringing the KCI concentration to 0.3 M. Further reduction of the KCI concentration was by the addition of 16.7 lal of the buffer at 5 s intervals with 60 additions bringing the KCI concentration to 0.15 M. In those cases were I mM Mg-ATP was present, it was present both in the starting protein solution and in the dilution buffer. All dilutions were done at 0 °C while stirring in an ice-bath. On reaching 0.15 M KCI samples were taken immediately for electron microscopy. The suspension was allowed to stand in the celd room (0-4 ° C) for 24 h before samples were taken again.
Electron microscopy After assembly, the paracrystal suspension was placed on carbon-coated copper grids and washed with the same solutions in which the paracrystals were suspended. The wash was followed by negative staining
184
Mr --
110 Kd
--
62 Kd
C Fig. 1. LMM paracrystals assembled by the standard dilution procedure in the absence of Mg-ATP. (a) Paracrystals with a 43 nm axial repeat pattern were observed immediately on reaching 0.15 M KCI in the dilution procedure. Mag. 90000×. (b) Higher magnification to show the filamentous structures in the background. In some cases, these filamentous structures appear to be continuous with the structural elements of the paracrystal Mag. 240000)<. (c) SDS-PAGE gel of the LMM (62 kDa) and rod (110 kDa) fragments of myosin used in these studies.
with 1% uranyl acetate in water. Electron micrographs of the paracrystal in Fig. 1 were taken at a magnification of 20000 × on a Siemens Elmiskop I. Those in Figs. 3 and 4 were taken at 40 000 × on a Phillips 201. For measurements of the diameter of the filamentous structures associated with the paracrystals (Fig. lb), the electron micrographs were printed at a final magnification of 240000 x and a bit pad was used to measure the diameter of the filamentous structures. A print-out of the diameter distribution and the mean and standard deviation was obtained.
30 ¢.. (9
E == r-'-
:3 m
20
Amino acid sequencing The peptide was run on 6% SDS PAGE gds and electroblotted onto Polyvinylidene Difluoride (PVDF) membranes as described by Matsudaira [14]. The band was cut out on the PVDF membrane and sequenced by Edman degradation at the NH2-terminal end on an Applied Biosystems 470A Pulsed Liquid Sequencer in the laboratory of Dr. Ruth Hogue-Angeletti, Department of Pathology, University of Pennsylvania. Results
On SDS-PAGE, the LMM preparation used in this work had a chain weight of M r 62 000 and the rod had a chain weight of Mr 110000 (Fig. lc). Aggregates of the LMM preparation that were formed using the standard dilution procedure in the absence of Mg. ATP [3] are
0 L_ (l) ,.0
E :3 Z 10
I
I
I
I
I
I
I
I
1
2
3
4
5
6
7
8
10
Diameter (nm) Fig. 2. Diameter distribution of the filamentous structure associated with paracrystals assembled by the standard dilution procedure as shown in Fig. 1. The mean diameter is 6.4 nm with a standard deviation of + 1.3 nm (100 measurements).
185 shown in Fig. l a and b. The suspension of L M M aggiegates was sampled immediately after reaching a KCl concentration of 0.15 M by dilution. The aggregates did not change with time. Paraeryst-als with a d e a r ,axial repeat periodicity of 43 nm are observed. The axial repeat pattern consists of alternating dark and fight bands. In some eases the paraerystals appear to be very loosely assembled and the axial repeat pattern is not as well defined. The paracrystals appear to be made up of long thin strands which are about the same diameter as the thin filamentous material, which is evenly distributed in the background, and in some cases the background filamentous material appears to be continuous, with the long thin strands making up the loosely packed paracrystals (Fig. lb). The diameter of the thin filamentous material in the background of Fig. l b was measured and the distribution is shown in Fig. 2. These filamentous aggregates of L M M have a mean diameter of 6.4 nm with a standard deviation of _+1.3 nm (mean of 100). The aggregation of L M M by the standzrd dilution procedure in the presence of 1 m M M g - A T P [4] is shown in Fig. 3a as observed immediately on reaching 0.15 M KCI for the L M M preparation of M r 62000. In this case, assembly occurred to form spindle shaped non-periodic aggregates. Larger aggregates of the nonperiodic spindle shaped aggregates were also observed. In some cases, the spindle shaped aggregates appear to be attached end to end to form longer non-periodic aggregates. Also observed were larger non-periodic aggregates which seemed to be made up of the side-to-side aggregation of these long end-to-end aggregates. After 24 h at 0 - 4 ° C, the L M M aggregates were transformed into paracrystals with a well-defined 43 nm axial repeat pattern (Fig. 3b). No differences were observed between these aggregates and those obtained in the absence of Mg- ATP. The NH2-terminal amino acid sequence of the L M M fragment was obtained (Dr. Ruth Hogue-Angeletti, Department of Pathology, Central Sequencing Facility). Out of the first 14 residues of the rabbit L M M fragment (RLMM), 11 residues matched a portion of the published rat embryonic myosin heavy chain ( R E M H C ) sequence [15]. The sequence determined for the N H 2terminal end of the fragment and the corresponding R E M H C sequence are as follows: R-LMM, Glu-Thr-Asp-Ala-Ile-Gln- 9. -Thr-GluGlu-Leu-.Leu- ? -Ala; R E M H C , Glu-Thr-Asp-Ala-Ile-Gln-Arg-Thr-GluGlu-Leu-Glu-Glu-Ala. This sequence of residues beg~as at residue 1376. From here to the C-terminal end, which is residce 1939, is a chain weight of 61 198 assumh~g a mean residue weight of 108.7 [16]. This identifies the 62 k D a fragment as including the entire C-ternAnal end of the rod starting at residue 1376.
Fig. 3. LMM assembled by the standard dilution procedure in the presence of 1 mM Mg ATP. (a) Aggregatesobserved immediately on reaching 0.15 M KCI. The spindle-shaped non-periodic aggregates were observed in abundance. End to end and side to side aggregation of these spindle shaped aggregates was also seen. (b) After 24 h, paracrystals with a well-defined 43 nm axial repeat pattern were abundant. Some residual non-periodic aggregates were also observed in the background. Mag. 140000x.
Rod paracrystals assembled by the standard dilution procedure in the absence of 1 m M M g - A T P as was done for the L M M shown in Fig. l a and b gave paracrystals with a 14.3 nm axial repeat pattern as shown in Fig. 4. Rod paracrystals assembled in the presence of 1 m M Mg- ATP as was done for the L M M shown in Fig. 3 also gave paracrystals with a 14.3 nm
186
Fig. 4. Rod paracrystals assembled by the standard dilution procedure. The paracrystal shown was assembled in the presence of I mM Mg-ATP. In the absence of I mM Mg-ATP, rod paracrystals with the same 14.3 mn axial repeat pattern were observed.l In both cases, these were observed immediately after reaching 0.15 M KCI and 24 h later. Mag. 200000)<.
axial repeat pattern. In contrast to the LMM, no difference was observed in paracrystals assembled from the myosin rod either in the presence or in the absence of Mg. ATP. In the presence of Mg-ATP, the paracrystals showed no change in axial repeat pattern with time. Discussion
Interactions between myosin molecules leading to the assembly of the myosin filaments must involve the rod portion of the molecule, since this is the portion which is aggregated into the shaft of the filament with the $1 portions available on the surface to interact with the surrounding actin filaments. Presumably the bulky S1 heads are responsible for preventing the formation of the paracrystals characteristically obtained with the myosin rod or the LMM portion of the rod [17]. Whether or not some portion of the $1 is contributing to the interactions of the myosin rods in the backbone of the filament in addition to sterically controlling side to side assembly has not been determined. In this regard, we have found that removal of the LC2 myosin light chain leads to interference with the length and diameter determination of the filament and that the assembly of the LC2-deficient myosin is not affected by the presence or absence of M g - A T P [18]. These findings suggest an involvement of the $1, including its light chains, in the assembly process. In this study, it was found that conditions which produce a relatively narrow distribution of myosin filament lengths around the native length of 1.5/~m [3.4] also produce LMM paracrystals with a sharp 43 nm axial repeat pattern (Figs. 1 and 3b), and conditions which produce a broad distribution of myosin filament lengths produce LMM aggregates which are non-periodic (Fig. 3a). Even more significantly, conditions which lead to a sharpening of the length distribution of the myosin filaments, i.e., with time after standard dilution in the presence of Mg. ATP [4], concomitantly lead to a transformation of the non-periodic LMM aggregates (Fig. 3a) to LMM paracrystals with a well-defined 43
nm axial repeat pattern (Fig. 3b). These observations strongly suggest a contribution from the LMM portion of the myosin rod to the mechanism for the redistribution of filament lengths in the presence of Mg-ATP. The LMM portion of the myosin rod containing the region which confers sensitivity to M g - A T P has been identified as including amino acid residues 1376-1939, i.e., the entire C-terminal portion of the rod beginning with residue 1376. The minimum region of this portion of the rod that is responsible for the Mg- ATP sensitivity remains to be determined. This will be done by expressing the gene segment corresponding to this fragment in E. coli and then methodically decreasing the length of the fragment expressed until the M g - A T P sensitivity is lost. In this way, we may be able to define the Mg- ATP binding region. The myosin rod which is about double the length of the LMM includes the Mg-ATP-sensitive LMM portion of the rod (residues 1376-1939). However, assembly of the myosin rod under the same conditions as those used for the LMM fragment does not exhibit sensitivity to the Mg-ATP. Also, the rod formed paracrystals with a 14.3 nm periodicity in contrast to the 43 nm periodicity obtained with LMM. The interactions contributed by the NH2-terminal difference region in the rod essentially have masked the M g - A T P sensitivity of the LMM portion and have promoted a 14.3 nm stagger, while in the intact molecule, the M g - A T P sensitivity is restored. It is possible that the bulk of the S1 heads may sterically alter the interactions of the NH2-terminal portion of the rod with the backbone and make the contribution of the Mg-ATP-sensitive region of the rod more significant in the assembly characteristics of the whole molecule. This would suggest a complex involvement of the LMM, $2 and $1 portions of the myosin in the assembly process. We are presently investigating how different portions of the myosin molecule cooperate in the assembly process by starting with fragments of known unique assembly properties, and methodically studying how the addition of neighboring sequences affects the assembly characteristics. While non-periodic aggregates are formed initially in the presence of Mg-ATP, periodic aggregates are formed initially in the absence of Mg. ATP. With time, in the presence of Mg- ATP the non-periodic aggregates are transformed into periodic aggregates. Although from these experiments we cannot determine whether MgATP is affecting the kinetics [19] of the assembly reaction, we can conclude that it is effecting a redistribution of the packing arrangement in the paracrystals. This redistribution in packing arrangement in the paracrystal can be related to the redistribution in packing that must occur on sharpening of the length distribution of myosin filaments in the presence of Mg. ATP [4]. Harrington and Himmelfarb [6] and Chowrashi and Pepe [4] have shown that the presence of ATP will
187 promote dissociation of myosin molecules from myosin filaments in vitro. A decrease in length of myosin filaments from 1.6 to 0.8 p m was observed by the presence of 5 m M ATP [4]. However, under the same conditions using M g . ATP or M g 2+ alone it was found that at concentrations as high as 10 mM, filament lengths of 1.4-1.5 p m were obtained. With Mg 2+ alone or A T P alone, there was no substantial change in the length distribution of the filaments wi& time. Only with M g - A T P was there a sharpening of the length distribution around 1.5 or 1.6 /tin with time. We have also shown previously that the effect of M g - ATP on myosin fileanent length distribution is not the result of phosphorylation, since the same effect is obtained using M g - A M P P N P and Mg pyrophosphate [4]. Filamentous material was associated with the paracrystals, (Fig. lb), and was found to have a mean diameter of 6.4 n m with a standard deviation of +_+_1.3 n m (mean of 100). This filamentous material may be related to the 6.5 nm dimension of the unit cell in L M M paracrystals studied by Yagi and Offer [10] by x-ray diffraction and electron microscopy. Subfilaments have been observed in transverse sections of vertebrate skeletal myosin filaments with a center to center spacing of about 3 or 4 nm. These are arranged with three pairs of subfilaments on the surface of the filament and with three additional subfilaments centrally located [20]. Maw and Rowe [21] have reported the splaying of myosin filaments into three parallel units which would each correspond to a pair of surface subfilaments and one central subfilament in the observations made by Pepe et al. [20]. It is conceivable that either this group of three subfilaments or pairs of subfilaments could correspond to the approx. 6 n m diameter filamentous material observed in relation to the paracrystals in Fig. lb. Two or three subfilaments close packed with a center spacing of about 4 nm could easily be measured as about 6 n m because of the fall off in density at the edges. Other proteins have been suggested as playing a role in myosin filament assembly, such as C-protein [22] or titin [23]. In the work reported here the L M M and rod fragments used were obtained from column-purified myosin and in previous work with myosin assembly [4], column-purified myosin was also used. Therefore, the assembly properties observed in these studies are related to the myosin molecules themselves. The effect of other proteins on the assembly would probably be limited to that of fine tuning rather than to any significant involvement in the length determining process [3]. The two basic findings of this study are that at least one Mg-ATP-sensitive region of the myosin rod is located somewhere along the C-terminal residues 1376-1939 and that the inclusion of the NH2-terminal residues to give the myosin rod results in a loss of this
sensitivity. It is likely that there is a relationship between the M g . ATP-induced rearrangements leading to a redistribution of myosin filament lengths, and the similarly induced rearrangem~nt form non-periodic aggregates to paracrystals of L M M (residues 1376-1939), since they are both triggered by the presence of M g . ATP. Both of these rearrangements involve a change in the assembly of the rod portion of the myosin molecule. We have therefore identified one region of the myosin rod, sensitive to the presence of M g - A T P , which contributes to the assembly characteristics observed for the intact molecule. Acknowledgements We are pleased to acknowledge the assistance of Beth Maguire and Valerie La Fave in this work, and Dr. Ruth Hogue-Angeletti, Department of Pathology, for the amino acid sequencing. This work was supported by USPHS Grant H L 15835 to the Pennsylvania Muscle Institute and in part by the Muscular Dystrophy Association. References
1 Huxley, H.E. (1963) J. Mol. Biol. 7, 281-308. 2 Pepe, F.A. (1982) (Stracher, A., ed.), pp. 105-149, Academic Press, New York. 3 Pepe, F.A., Drucker, 1:¢.and Chowrashi, P.K. (1986) Prep. Biochem. 16, 99-132. 4 Chowrashi, P.K. and Pepe, F.A. (1986) J. Muse. Res. Cell Motility, 7, 413-420. 5 Morimoto, K. and Harrington, W.F. (1974) J. Mol. Biol. 83, 83-97. 6 Harrington, W.F. and Himmelfarb, S. (1972) Biochemistry 11, 2945-2952. 7 Bartels, E.M., Cooke, P.H., Elliott, G.F. and Hughes, R.A. (1984) Proc. Physical. Soc. 358, 80P. 8 Infante, A.A. and Davies, R.E. (1965) J. Biol. Chem. 240, 3996-4001. 9 Bennett, P. (1981) J. Mol. Biol. 146, 201-221. 10 Yagi, N. and Offer, G.W. (1981) J. Mol. Biol. 151,467-490. 11 Safer, D. and Pepe, F.A. (1980) J. Mol. Biol. 136, 343-358. 12 Lowey, S., Slayter, H.S., Weeds, A.G. and Baker, H. (1969) J. Mol. Biol. 42, 1-29. 13 Chowrashi, P.K. and Pepe, F.A. (1982) J. Cell Biol. 94, 564-573. 14 Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038. 15 Strehler, E.E., Strehler-Page, M., Perriard, J., Periasamy, M. and Nadal-Ginard, B. (1986) J. Mol. Biol. 190, 291-317. 16 Schulz, G.E. and Schirmer, R.H. (1979) Principles of Protein Structure, Springer-Verlag, New York, p. 2. 17 Chowrashi, P.K. and Pepe, F.A. (1977) J. Cell Biol. 74. 136-152. 18 Chowrasbi, P.K., Pemrick, S.M. and Pepe, F.A. (1989) Biochim. Biophys. Acta 990, 216-223. 19 Davis, J. (1988) Annu. Rev. Biophys. Chem. 17, 217-239. 20 Pepe, F.A., Ashton, F.T., Street, C. and Weisel, J. (1986) Tissue Cell 18, 499-508. 21 Maw, M.C. and Rowe, A.J. (1980) Nature 286, 412-414. 22 Moos, C. (1972) Cold Spring Harbor Symposium on Quant. Biol. 37, 93-95. 23 Wang, K. and Wright, J. (1987) J. Cell Biol. 105, 27a.
182
Elsevier BBAPRO33451
The myosin filament XIII. The sensitivity of LMM assembly to Mg-ATP P r o k a s h K. C h o w r a s h i a n d F r a n k A. P e p e Department of Anatomy, School of Medicine, University of Pennsylvania, Philadelphia, PA (U.S.A.)
(Received 7 February1989)
Key words: Myosin;LMM; Mg. ATP; Assembly
An LMM fragment (M, 62 000) of myosin has been prepared which has aggregation properties that are sensitive to the presence of Mg- ATP. Aggregation of the LMM by reducing the ionic strength in the presence of 1 mM Mg- ATP produces non-periodic aggregates which gradually rearrange to paracrystals with a 43 nm axial repeat pattern. This fragment includes the C-terminal end of the myosin rod starting at residue 1376. Therefore, at least one of the Mg. ATP binding sites responsible for this effect is located somewhere along this region of the myosin rod. Although assembly of the rod fragment of myosin into paracrystais does not show sensitivity to Mg. ATP, assembly of intact myosin molecules to form filaments does show sensitivity to Mg- ATP. For myosin filaments, assembly initially gives a broad distribution around a mean length of 1.5 /zm, which sharpens around the mean length with time. The rearrangement of the ~ rods and intact myosin molecules both induced by the presence of Mg- ATP are probably related. These findings highlight the complexity of the cooperative interactions between different portions of the myosin molecule that are involved in determining the assembly properties of the intact molecule.
Introduction The assembly of myosin molecules in vitro has been shown to produce filaments which are structurally similar to native filaments in that they have a bare zone region of constant length in the middle of the filament and they have tapered ends [1]. However, the length of the filaments varies widely and this variation becomes greater with increasing length of the filaments [2]. Recently, a dilution procedure has been described which produces a length distribution around the native length of 1.5/tin with a standard deviation of _+0.2 pm [3,4]. The question of whether or not native filaments are themselves precisely determined in length or whether there is a narrow distribution of lengths cannot be answered unequivocally, since on separating the filaments for measurement, some loss of molecules from the ends may occur. Measurement of the A-band width may not correspond to filament length if there is some stagger between the filaments in the A-band lattice. The best measurements of native filament lengths that have been made are those of Morimoto and Harrington [5],
Correspondence: F.A. Pepe, Department of Anatomy, School of Medicine, Universityof Pennsylvania, Philadelphia, PA 19104-6058, U.S.A.
in which the distribution of lengths measured (mean of 1.53 pm with 80~ within ±0.17/tin) compares favorably with that obtained by in vitro assembly [3,4]. The assembly of myosin in vitro was found to be influenced by the presence of M g - A T E It was shown that the addition of 1 mM M g - A T P to a previously assembled filament suspension would lead to a redistribution in the lengths so that the distribution was sharpened around the native length of about 1.5 ttm [4]. This occurred even for variations of the assembly procedure which gave very wide distributions in the length of the assembled filaments. It was also found that if the assembly of myosin was done in the presence of 1 mM Mg-ATP a relatively broad distribution in length was obtained initially with redistribution on standing in the cold (0-4°C) to give a sharper distribution around the native length of 1.5 pm. Some experiments [4] have also shown that the minimum concentration of M g - A T P required to effect this redistribution of the lengths of the filaments was about 500/tM for a concentration of myosin of about 1 ~tM, therefore excluding the stoichiometric binding of the Mg. ATP to the myosin ATPase active site as being involved in the length-determining mechanism. Harrington and Himmelfarb [6] and Bartels et al. [7] have suggested that there are other binding sites for Mg. ATP on the myosin molecule. It is noteworthy that the concentration of M g - A T P in riving
0167-4838/89/$03.50 © 1989ElsevierSciencePublishersB.V.(BiomedicalDivision)
183 muscle cells [8] is within the range of that required for filament length redistribution of the in vitro assembled myosin filaments. Since a fragment of the myosin molecule is a part of the intact molecule, the assembly characteristics of the fragment must contribute to the d~termination of the assembly characteristics of the whole molecule. LMM assembles into paracrystals with the LMM rods arranged anti-parallel [9,10]. An anti-parallel packing arrangement is characteristic only of the bare zone region of the myosin filament, the rest of the assembly being by parallel alignment. However there must be a direct relationship between the anti-parallel and parallel assembly information which contributes to the filament assembly characteristics and which involves the transition from one to the other as the filament assembles beyond the bare zone region. In this study an LMM fragment has been identified (residues 1376-1939) with assembly properties sensitive to the presence of M g - A T E Its assembly has been compared with that of the myosin rod and with the assembly of the intact myosin molecule. Since the intact molecule has also been shown to be sensitive to the presence of Mg-ATP [4], it is likely that the identified peptide region of the myosin.rod of M r 62000 (LMM) used in this study contributes to that sensitivity.
Preparation of the rod fragment The rod fragment of myosin was prepared by the digestion of i~,soluble myosin with papain (Worthington-Cappel) e~entially as described previously [12]. The column-purified rabbit skeletal myosin, at a concentration of 12 mg/ml0 was dialyzed against 0.2 M ammonium acetate (pH 7.0). Under these conditions, the myosin is insoluble. To this suspension, papain was added to about 0.3 mg/ml or less. The reaction was allowed to proceed at room temperature for 15 min. Digestion was stopped by the addition of 0.1 M iodoacetic acid to a final concentration of 1 mM and the pH was adjusted to 6.0 to stop the reaction. The suspension was then centrifuged and the pellet was washed twice by re-suspension in 0.2 M ammonium acetate (pH 7.0). The washed pellet was dissolved in 0.6 M KC1/0.03 l,l phosphate buffer (pH 6.5). After solution, three volumes of ethyl alcohol were added and stirred at room temperature for 3 h. The precipitate was collected by centrifugation, dissolved in 0.6 M KCI/10 mM imidazole (pH 6.8) and dialyzed in the same buffer.
SDS-PA GE electrophoresis The procedure used for SDS-gel electrophoresis was the same as that previously described [13]. In these studies, 7.5% polyacrylamide gels were used.
Assembly of the LMM and rod fragments of myosin Materials and. Methods
Myosin was isolated from rabbit back and leg muscles and purified by column chromatography on DEAE Sephadex A50 as described previously [4]. The purified myosin was used for the preparation of the LMM and rod fragments of myosin as described below,
Preparation of the LMM fragment Column purified rabbit myosin was digested with ot-chymotrypsin (Sigma) as previously described [11]. The myosin at a concentration of 8 mg/ml in 0.5 M ammonium carbonate (pH 8.0) was digested with achymotrypsin in a weight ratio of myosin to a-chymotrypsin of 100:1 for 30 rain at room temperature. The reaction was stopped by adding phenylmethylsulfonyl fluoride (PMSF) to a final concentration of about 1 raM. The digest was dialyzed against 0.05 M KCI/5 mM phosphate buffer (pH 6.2) overnight. The precipitate was collected by centrifugation and dissolved in 0.6 M KCI/10 mM imidazole (pH 7.0). Three volumes of ethanol were added to this solution and this was stirred for 3 h at room temperature. The precipitate was collected and re-suspended in 0.6 M KCI/10 mM imidazole (pH 6.8) and dialyzed in the same buffer overnight. After dialysis, the solution was clarified by centrifugation.
Assembly of the LMM and rod fragments was carried out as previously described for myosin filaments [3,4]. The standard dilution procedure was used in all cases. The LMM or rod was at a starting concentration of 2 mg per ml in 0.6 M KCIp_0 mM imidazole (pH 6.8) either with or without the presence of 1 mM Mg ATP. The KCI concentration was reduced in two steps, from 0.6 to 0.3 M KCI and then from 0.3 to 0.15 M KCI. Starting with 0.5 ml of protein, 10 mM imidazole buffer (pH 6.8) was added in volumes of 35.3 /~1 at intervals of 5 s with 14 additions, bringing the KCI concentration to 0.3 M. Further reduction of the KCI concentration was by the addition of 16.7 lal of the buffer at 5 s intervals with 60 additions bringing the KCI concentration to 0.15 M. In those cases were I mM Mg-ATP was present, it was present both in the starting protein solution and in the dilution buffer. All dilutions were done at 0 °C while stirring in an ice-bath. On reaching 0.15 M KCI samples were taken immediately for electron microscopy. The suspension was allowed to stand in the celd room (0-4 ° C) for 24 h before samples were taken again.
Electron microscopy After assembly, the paracrystal suspension was placed on carbon-coated copper grids and washed with the same solutions in which the paracrystals were suspended. The wash was followed by negative staining
184
Mr --
110 Kd
--
62 Kd
C Fig. 1. LMM paracrystals assembled by the standard dilution procedure in the absence of Mg-ATP. (a) Paracrystals with a 43 nm axial repeat pattern were observed immediately on reaching 0.15 M KCI in the dilution procedure. Mag. 90000×. (b) Higher magnification to show the filamentous structures in the background. In some cases, these filamentous structures appear to be continuous with the structural elements of the paracrystal Mag. 240000)<. (c) SDS-PAGE gel of the LMM (62 kDa) and rod (110 kDa) fragments of myosin used in these studies.
with 1% uranyl acetate in water. Electron micrographs of the paracrystal in Fig. 1 were taken at a magnification of 20000 × on a Siemens Elmiskop I. Those in Figs. 3 and 4 were taken at 40 000 × on a Phillips 201. For measurements of the diameter of the filamentous structures associated with the paracrystals (Fig. lb), the electron micrographs were printed at a final magnification of 240000 x and a bit pad was used to measure the diameter of the filamentous structures. A print-out of the diameter distribution and the mean and standard deviation was obtained.
30 ¢.. (9
E == r-'-
:3 m
20
Amino acid sequencing The peptide was run on 6% SDS PAGE gds and electroblotted onto Polyvinylidene Difluoride (PVDF) membranes as described by Matsudaira [14]. The band was cut out on the PVDF membrane and sequenced by Edman degradation at the NH2-terminal end on an Applied Biosystems 470A Pulsed Liquid Sequencer in the laboratory of Dr. Ruth Hogue-Angeletti, Department of Pathology, University of Pennsylvania. Results
On SDS-PAGE, the LMM preparation used in this work had a chain weight of M r 62 000 and the rod had a chain weight of Mr 110000 (Fig. lc). Aggregates of the LMM preparation that were formed using the standard dilution procedure in the absence of Mg. ATP [3] are
0 L_ (l) ,.0
E :3 Z 10
I
I
I
I
I
I
I
I
1
2
3
4
5
6
7
8
10
Diameter (nm) Fig. 2. Diameter distribution of the filamentous structure associated with paracrystals assembled by the standard dilution procedure as shown in Fig. 1. The mean diameter is 6.4 nm with a standard deviation of + 1.3 nm (100 measurements).
185 shown in Fig. l a and b. The suspension of L M M aggiegates was sampled immediately after reaching a KCl concentration of 0.15 M by dilution. The aggregates did not change with time. Paraeryst-als with a d e a r ,axial repeat periodicity of 43 nm are observed. The axial repeat pattern consists of alternating dark and fight bands. In some eases the paraerystals appear to be very loosely assembled and the axial repeat pattern is not as well defined. The paracrystals appear to be made up of long thin strands which are about the same diameter as the thin filamentous material, which is evenly distributed in the background, and in some cases the background filamentous material appears to be continuous, with the long thin strands making up the loosely packed paracrystals (Fig. lb). The diameter of the thin filamentous material in the background of Fig. l b was measured and the distribution is shown in Fig. 2. These filamentous aggregates of L M M have a mean diameter of 6.4 nm with a standard deviation of _+1.3 nm (mean of 100). The aggregation of L M M by the standzrd dilution procedure in the presence of 1 m M M g - A T P [4] is shown in Fig. 3a as observed immediately on reaching 0.15 M KCI for the L M M preparation of M r 62000. In this case, assembly occurred to form spindle shaped non-periodic aggregates. Larger aggregates of the nonperiodic spindle shaped aggregates were also observed. In some cases, the spindle shaped aggregates appear to be attached end to end to form longer non-periodic aggregates. Also observed were larger non-periodic aggregates which seemed to be made up of the side-to-side aggregation of these long end-to-end aggregates. After 24 h at 0 - 4 ° C, the L M M aggregates were transformed into paracrystals with a well-defined 43 nm axial repeat pattern (Fig. 3b). No differences were observed between these aggregates and those obtained in the absence of Mg- ATP. The NH2-terminal amino acid sequence of the L M M fragment was obtained (Dr. Ruth Hogue-Angeletti, Department of Pathology, Central Sequencing Facility). Out of the first 14 residues of the rabbit L M M fragment (RLMM), 11 residues matched a portion of the published rat embryonic myosin heavy chain ( R E M H C ) sequence [15]. The sequence determined for the N H 2terminal end of the fragment and the corresponding R E M H C sequence are as follows: R-LMM, Glu-Thr-Asp-Ala-Ile-Gln- 9. -Thr-GluGlu-Leu-.Leu- ? -Ala; R E M H C , Glu-Thr-Asp-Ala-Ile-Gln-Arg-Thr-GluGlu-Leu-Glu-Glu-Ala. This sequence of residues beg~as at residue 1376. From here to the C-terminal end, which is residce 1939, is a chain weight of 61 198 assumh~g a mean residue weight of 108.7 [16]. This identifies the 62 k D a fragment as including the entire C-ternAnal end of the rod starting at residue 1376.
Fig. 3. LMM assembled by the standard dilution procedure in the presence of 1 mM Mg ATP. (a) Aggregatesobserved immediately on reaching 0.15 M KCI. The spindle-shaped non-periodic aggregates were observed in abundance. End to end and side to side aggregation of these spindle shaped aggregates was also seen. (b) After 24 h, paracrystals with a well-defined 43 nm axial repeat pattern were abundant. Some residual non-periodic aggregates were also observed in the background. Mag. 140000x.
Rod paracrystals assembled by the standard dilution procedure in the absence of 1 m M M g - A T P as was done for the L M M shown in Fig. l a and b gave paracrystals with a 14.3 nm axial repeat pattern as shown in Fig. 4. Rod paracrystals assembled in the presence of 1 m M Mg- ATP as was done for the L M M shown in Fig. 3 also gave paracrystals with a 14.3 nm
186
Fig. 4. Rod paracrystals assembled by the standard dilution procedure. The paracrystal shown was assembled in the presence of I mM Mg-ATP. In the absence of I mM Mg-ATP, rod paracrystals with the same 14.3 mn axial repeat pattern were observed.l In both cases, these were observed immediately after reaching 0.15 M KCI and 24 h later. Mag. 200000)<.
axial repeat pattern. In contrast to the LMM, no difference was observed in paracrystals assembled from the myosin rod either in the presence or in the absence of Mg. ATP. In the presence of Mg-ATP, the paracrystals showed no change in axial repeat pattern with time. Discussion
Interactions between myosin molecules leading to the assembly of the myosin filaments must involve the rod portion of the molecule, since this is the portion which is aggregated into the shaft of the filament with the $1 portions available on the surface to interact with the surrounding actin filaments. Presumably the bulky S1 heads are responsible for preventing the formation of the paracrystals characteristically obtained with the myosin rod or the LMM portion of the rod [17]. Whether or not some portion of the $1 is contributing to the interactions of the myosin rods in the backbone of the filament in addition to sterically controlling side to side assembly has not been determined. In this regard, we have found that removal of the LC2 myosin light chain leads to interference with the length and diameter determination of the filament and that the assembly of the LC2-deficient myosin is not affected by the presence or absence of M g - A T P [18]. These findings suggest an involvement of the $1, including its light chains, in the assembly process. In this study, it was found that conditions which produce a relatively narrow distribution of myosin filament lengths around the native length of 1.5/~m [3.4] also produce LMM paracrystals with a sharp 43 nm axial repeat pattern (Figs. 1 and 3b), and conditions which produce a broad distribution of myosin filament lengths produce LMM aggregates which are non-periodic (Fig. 3a). Even more significantly, conditions which lead to a sharpening of the length distribution of the myosin filaments, i.e., with time after standard dilution in the presence of Mg. ATP [4], concomitantly lead to a transformation of the non-periodic LMM aggregates (Fig. 3a) to LMM paracrystals with a well-defined 43
nm axial repeat pattern (Fig. 3b). These observations strongly suggest a contribution from the LMM portion of the myosin rod to the mechanism for the redistribution of filament lengths in the presence of Mg-ATP. The LMM portion of the myosin rod containing the region which confers sensitivity to M g - A T P has been identified as including amino acid residues 1376-1939, i.e., the entire C-terminal portion of the rod beginning with residue 1376. The minimum region of this portion of the rod that is responsible for the Mg- ATP sensitivity remains to be determined. This will be done by expressing the gene segment corresponding to this fragment in E. coli and then methodically decreasing the length of the fragment expressed until the M g - A T P sensitivity is lost. In this way, we may be able to define the Mg- ATP binding region. The myosin rod which is about double the length of the LMM includes the Mg-ATP-sensitive LMM portion of the rod (residues 1376-1939). However, assembly of the myosin rod under the same conditions as those used for the LMM fragment does not exhibit sensitivity to the Mg-ATP. Also, the rod formed paracrystals with a 14.3 nm periodicity in contrast to the 43 nm periodicity obtained with LMM. The interactions contributed by the NH2-terminal difference region in the rod essentially have masked the M g - A T P sensitivity of the LMM portion and have promoted a 14.3 nm stagger, while in the intact molecule, the M g - A T P sensitivity is restored. It is possible that the bulk of the S1 heads may sterically alter the interactions of the NH2-terminal portion of the rod with the backbone and make the contribution of the Mg-ATP-sensitive region of the rod more significant in the assembly characteristics of the whole molecule. This would suggest a complex involvement of the LMM, $2 and $1 portions of the myosin in the assembly process. We are presently investigating how different portions of the myosin molecule cooperate in the assembly process by starting with fragments of known unique assembly properties, and methodically studying how the addition of neighboring sequences affects the assembly characteristics. While non-periodic aggregates are formed initially in the presence of Mg-ATP, periodic aggregates are formed initially in the absence of Mg. ATP. With time, in the presence of Mg- ATP the non-periodic aggregates are transformed into periodic aggregates. Although from these experiments we cannot determine whether MgATP is affecting the kinetics [19] of the assembly reaction, we can conclude that it is effecting a redistribution of the packing arrangement in the paracrystals. This redistribution in packing arrangement in the paracrystal can be related to the redistribution in packing that must occur on sharpening of the length distribution of myosin filaments in the presence of Mg. ATP [4]. Harrington and Himmelfarb [6] and Chowrashi and Pepe [4] have shown that the presence of ATP will
187 promote dissociation of myosin molecules from myosin filaments in vitro. A decrease in length of myosin filaments from 1.6 to 0.8 p m was observed by the presence of 5 m M ATP [4]. However, under the same conditions using M g . ATP or M g 2+ alone it was found that at concentrations as high as 10 mM, filament lengths of 1.4-1.5 p m were obtained. With Mg 2+ alone or A T P alone, there was no substantial change in the length distribution of the filaments wi& time. Only with M g - A T P was there a sharpening of the length distribution around 1.5 or 1.6 /tin with time. We have also shown previously that the effect of M g - ATP on myosin fileanent length distribution is not the result of phosphorylation, since the same effect is obtained using M g - A M P P N P and Mg pyrophosphate [4]. Filamentous material was associated with the paracrystals, (Fig. lb), and was found to have a mean diameter of 6.4 n m with a standard deviation of +_+_1.3 n m (mean of 100). This filamentous material may be related to the 6.5 nm dimension of the unit cell in L M M paracrystals studied by Yagi and Offer [10] by x-ray diffraction and electron microscopy. Subfilaments have been observed in transverse sections of vertebrate skeletal myosin filaments with a center to center spacing of about 3 or 4 nm. These are arranged with three pairs of subfilaments on the surface of the filament and with three additional subfilaments centrally located [20]. Maw and Rowe [21] have reported the splaying of myosin filaments into three parallel units which would each correspond to a pair of surface subfilaments and one central subfilament in the observations made by Pepe et al. [20]. It is conceivable that either this group of three subfilaments or pairs of subfilaments could correspond to the approx. 6 n m diameter filamentous material observed in relation to the paracrystals in Fig. lb. Two or three subfilaments close packed with a center spacing of about 4 nm could easily be measured as about 6 n m because of the fall off in density at the edges. Other proteins have been suggested as playing a role in myosin filament assembly, such as C-protein [22] or titin [23]. In the work reported here the L M M and rod fragments used were obtained from column-purified myosin and in previous work with myosin assembly [4], column-purified myosin was also used. Therefore, the assembly properties observed in these studies are related to the myosin molecules themselves. The effect of other proteins on the assembly would probably be limited to that of fine tuning rather than to any significant involvement in the length determining process [3]. The two basic findings of this study are that at least one Mg-ATP-sensitive region of the myosin rod is located somewhere along the C-terminal residues 1376-1939 and that the inclusion of the NH2-terminal residues to give the myosin rod results in a loss of this
sensitivity. It is likely that there is a relationship between the M g . ATP-induced rearrangements leading to a redistribution of myosin filament lengths, and the similarly induced rearrangem~nt form non-periodic aggregates to paracrystals of L M M (residues 1376-1939), since they are both triggered by the presence of M g . ATP. Both of these rearrangements involve a change in the assembly of the rod portion of the myosin molecule. We have therefore identified one region of the myosin rod, sensitive to the presence of M g - A T P , which contributes to the assembly characteristics observed for the intact molecule. Acknowledgements We are pleased to acknowledge the assistance of Beth Maguire and Valerie La Fave in this work, and Dr. Ruth Hogue-Angeletti, Department of Pathology, for the amino acid sequencing. This work was supported by USPHS Grant H L 15835 to the Pennsylvania Muscle Institute and in part by the Muscular Dystrophy Association. References
1 Huxley, H.E. (1963) J. Mol. Biol. 7, 281-308. 2 Pepe, F.A. (1982) (Stracher, A., ed.), pp. 105-149, Academic Press, New York. 3 Pepe, F.A., Drucker, 1:¢.and Chowrashi, P.K. (1986) Prep. Biochem. 16, 99-132. 4 Chowrashi, P.K. and Pepe, F.A. (1986) J. Muse. Res. Cell Motility, 7, 413-420. 5 Morimoto, K. and Harrington, W.F. (1974) J. Mol. Biol. 83, 83-97. 6 Harrington, W.F. and Himmelfarb, S. (1972) Biochemistry 11, 2945-2952. 7 Bartels, E.M., Cooke, P.H., Elliott, G.F. and Hughes, R.A. (1984) Proc. Physical. Soc. 358, 80P. 8 Infante, A.A. and Davies, R.E. (1965) J. Biol. Chem. 240, 3996-4001. 9 Bennett, P. (1981) J. Mol. Biol. 146, 201-221. 10 Yagi, N. and Offer, G.W. (1981) J. Mol. Biol. 151,467-490. 11 Safer, D. and Pepe, F.A. (1980) J. Mol. Biol. 136, 343-358. 12 Lowey, S., Slayter, H.S., Weeds, A.G. and Baker, H. (1969) J. Mol. Biol. 42, 1-29. 13 Chowrashi, P.K. and Pepe, F.A. (1982) J. Cell Biol. 94, 564-573. 14 Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038. 15 Strehler, E.E., Strehler-Page, M., Perriard, J., Periasamy, M. and Nadal-Ginard, B. (1986) J. Mol. Biol. 190, 291-317. 16 Schulz, G.E. and Schirmer, R.H. (1979) Principles of Protein Structure, Springer-Verlag, New York, p. 2. 17 Chowrashi, P.K. and Pepe, F.A. (1977) J. Cell Biol. 74. 136-152. 18 Chowrasbi, P.K., Pemrick, S.M. and Pepe, F.A. (1989) Biochim. Biophys. Acta 990, 216-223. 19 Davis, J. (1988) Annu. Rev. Biophys. Chem. 17, 217-239. 20 Pepe, F.A., Ashton, F.T., Street, C. and Weisel, J. (1986) Tissue Cell 18, 499-508. 21 Maw, M.C. and Rowe, A.J. (1980) Nature 286, 412-414. 22 Moos, C. (1972) Cold Spring Harbor Symposium on Quant. Biol. 37, 93-95. 23 Wang, K. and Wright, J. (1987) J. Cell Biol. 105, 27a.