Peptide maps of the myosin isoenzymes of Acanthamoeba castellanii.

Extracts of Acanthamoeba castellanii contain four myosin-like ATPases (Maruta, H., Gadasi, H., Collins, J.H., and Korn, E.D. (1979) J. Biol. Chem. 254, 3624-3630): double-headed Acanthamoeba myosin II and single-headed Acanthamoeba myosins IA, IB, and IC, which have heavy chains of 170,000, 130,000, 125,000, and 130,000 daltons, respectively, as well as different light chains. In the accompanying paper, evidence is presented that suggests that Acanthamoeba myosin IC is the same molecule as Acanthamoeba myosin IA plus a regulatory 20,000-dalton peptide. This conclusion is confirmed by the identity of the peptide maps obtained by limited proteolysis of the heavy chains of Acanthamoeba myosins IA and IC by Staphylococcus aureus V8 protease. However, peptide maps of the heavy chains of Acanthamoeba myosins IA, IB, and II obtained by limited proteolysis by the Staphylococcus protease and chymotrypsin and by chemical cleavage by cyanogen bromide and cyanylation have few, if any, peptides in common. From this evidence, and the enzymatic and subunit data in the accompanying paper, it is concluded that the three Acanthamoeba myosin isoenzymes, IA (IC), IB, and II, are products of different genes.


genes.
In the accompanying (1) and previous (2)(3)(4)(5)(6)(7) papers, we have shown that extracts of Acanthamoeba castedzanii contain four myosin-Iike ATPases. Acanthamoeba myosin II is a doubleheaded enzyme with a pair of heavy chains of about 170,000 daltons and two pairs of light chains of about 17,500 and 17,000 daltons (5,8). Three single-headed enzymes are also present in Acanthamoeba (1). Acanthamoeba myosin IA contains a single heavy chain of about 130,000 daltons and single light chains of about 17,000 and 14,000 daltons; Acanthamoeba myosin IB contains a heavy chain of about 125,000 dakons and light chains of about 27,000 and 14,000 daltons; Acanthamoeba myosin IC contains a heavy chain of about 130,000 dakons and light chains of about 20,000, 17,000, and 14,000 daltons. In addition to its different native molecular weight and subunit composition, Acanthamoeba myosin II differs dramatically in its enzymatic properties from Acanthamoeba myosins IA, IB, and IC which are enzymatically rather similar to each other (1).
The question naturally arises whether any of these four enzymes have a common origin. We presented evidence in the previous paper (1) that Acanthamoeba myosin IC consists of Acanthamoeba IA plus a loosely bound 20,000-dalton regulatory peptide that inhibits some of the ATPase activities of the enzyme. On the basis of their subunit compositions and l The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. enzymatic properties, however, we speculated in the accompanying paper (1) that Acanthamoeba myosins IA( IB, and II might have independent origins. This supposition can be directly tested in at least two ways. First, attempts can be made to convert Acanthamoeba myosin II, by controlled proteolytic digestion, to products resembling Acanthamoeba myosins IA, IB, and IC in subunit composition and enzymatic properties. Second, and without regard to the added complexity imposed by their different light chain compositions, the heavy chains of the several myosins can be compared by peptide mapping. If, as we supposed (l), Acanthamoeba myosin IC is Acanthamoeba myosin IA plus a 20,000-dalton peptide, then peptide maps of the 130,000-dahon heavy chains of the two enzymes should be identical. If the 125,000-dalton heavy chain of Acanthamoeba myosin IB were derived from the 130,000dalton heavy chain of IA, or if they have a common origin, then essentially ah of the peptides in digests of the heavy chain of IB should also be present in digests of IA and approximately 95% of the peptides in digests of the heavy chain of IA should occur in digests of the heavy chain of IB. Similarly, if the heavy chains of Acanthameoba myosins IA, IB, and IC are derived from the 170,000-dakon heavy chains of Acanthamoeba myosin II, almost ah of the peptides in digests of the heavy chains of IA, IB, and IC should be present in digests of the heavy chain of II and about 75% of the peptides in digests of the heavy chain Acanthamoeba myosin II should occur in the digests of the heavy chains of Acanthamoeba myosins IA, IB, and IC.
Finally, consider the possibility that Acanthamoeba myosins IA, IB, IC, and II might ah be degradation products of a common precursor molecule. The largest heavy chain found in a myosin from any source is about 225,000 daltons (9). Even if the 170,000-dalton chain of Acanthamoeba myosin II and the 130,000-dalton chain of Acanthamoeba myosin IA were derived from opposite ends of a 225,000-dahon precursor, the minimum overlap would be 75,000 dakons (( 170,000 + 130,000) -225,000). Therefore, about 60% of the peptides in digests of Acanthamoeba myosin I heavy chain would also be present in digests of Acanthamoeba myosin II heavy chain, while 44% of the peptides in digests of the heavy chain of II would occur in digests of the heavy chain of I. In fact, the overlapping sequence would almost certainly be greater than the arithmetical minimum because all the myosin heavy chains would have to include that portion of the precursor molecule that contained the ATP-binding and ATP hydrolytic sites and the actin-binding site (7). The data in this paper show that tryptic digestion can convert Acanthamoeba myosin II to a single-headed molecule. However, the enzymatic activities of the tryptic digestion products are identical to those of the original enzyme and their subunit compositions also have no similarity to Acanthamoeba myosin IA, IB, or IC. Peptide maps of the heavy chains of Acanthamoeba myosins IA, IB, IC, and II after digestion with either Staphylococcus aureus V8 protease, chymotrypsin, cyanogen bromide, or cyanylation confirm the suspected identity of the heavy chains of Acanthamoeba myosins IA and IC, but, more importantly, show that they and the heavy chains of Acanthamoeba myosin IB and Acanthamoeba myosin II are the products of three different genes.

EXPERIMENTAL PROCEDURES
Materials-Acanthamoeba myosins IA, IB, and IC and muscle actin and myosin were prepared as described in the accompanying paper (1). Acanthamoeba myosin II was prepared by the procedure of Pollard et al. (8). Partially purified Acanthamoeba myosin I heavy chain kinase was prepared as described previously (6). ATP, cyanogen bromide, imidazole, and Tris were purchased from Sigma Chemical Co.; dithiothreitol and 5,5'-dithiobis(2-nitrobenzoic acid) from Calbiochem Corp.; ultrapure guanidine from Schwarz/Mann; [14C]KCN and N-ethyl [2,3-'4C]maleimide from Amersham Corp.; [Y-~'P]ATP from New England Nuclear; tosylphenylalanyl chloromethyl ketonetreated trypsin, soybean trypsin inhibitor, and'chymotrypsin from Worthington Biochemicals; Staphylococcus aureus V8 protease from Miles Laboratories; the chemicals for polyacrylamide gel electrophoresis and Bio-Gel A-15m from Bio-Bad; and Sephadex G-25 (fine) from Pharmacia Fine Chemicals. All other chemicals were reagent grade and deionized water was used throughout.
Methods-Protein concentrations were estimated by the procedure of Lowry et al. (10) with bovine serum albumin (Armour Laboratories) as standard. ATPase activities were measured by the rate of release of E3*P]P, from [y-32P]ATP as described (1). The ability of Acanthamoeba myosin heavy chain kinase to stimulate actin-activated myosin Mg*+-ATPase activity was assayed as described by Maruta and Korn (6). Incorporation of 32P into protein was measured by the filter paper assay (11).
S-(N-Ethylsuccinimido)-myosins were prepared by incubating 1 to 2 mg of enzyme with 2.2 mM N-ethyl [2,3-'4C]maleimide in 1 ml of 0.5 M KCl, 10 mM phosphate buffer, pH 7.0, for 45 min at 0°C. The reaction was terminated by addition of dithiothreitol to a final concentration of 25 mM (12). Incorporation of radioactivity was determined by separating the polypeptides by dodecyl sulfate-polyacrylamide gel electrophoresis using gels prepared with N,N'-diallyltartardiimide:acrylamide (molar ratio, 1O:l). Gels were scanned at 550 nm and 1.6-mm sections were dissolved in 2% NaI04 at room temperature (13). The samples were added to 10 ml of Aquasol (New England Nuclear) and radioactivity was determined in a scintillation counter.
Acanthamoeba myosin II (2 mg/ml) was digested with trypsin (40 &ml) in 0.5 M KCl, 1 mM dithiothreitol, and 10 mM imidazole chloride, pH 7.0, at 22'C for 10 min. The digestion was stopped by addition of soybean inhibitor at a S-fold weight excess relative to trypsin. The digestion products were separated by chromatography on Bio-Gel A-15m (1.5 x 30 cm) into two peaks of Ca*+-ATPase activity with molecular weights of about 310,000 (major) and 125,000 (minor). The major fraction was reincubated with one-tenth its mass of trypsin under conditions identical to the original digestion and the products were separated by agarose gel chromatography into two peaks of Ca*+-ATPase activity with approximate molecular weights of 275,000 and 105,000.
Dodecyl sulfate-polyacrylamide gel electrophoresis was routinely carried out in a slab gel apparatus using the discontinuous buffer system described by Laemmli (14). Gels were stained with Coomassie blue according to Fairbanks et al. (15). Molecular weight standards included muscle myosin and actin, phosphorylase A, and bovine serum albumin.
Peptide mapping by limited proteolysis in dodecyl sulfate-polyacrylamide gels was carried out essentially as described by Cleveland et al. (16). Acanthamoeba myosins were initially separated into heavy and light chains by dodecyl sulfate-polyacrylamide gel electrophoresis and the gels were briefly stained with Coomassie blue and rapidly destained. The regions containing the heavy chains were cut out and the slices were placed immediately in the sample wells of a second gel to which protease (Staphylococcus aureus V8 or chymotrypsin) had previously been added in amounts given in the figure legends. Electrophoresis on the second gel was begun immediately.
The cyanylation reaction with [14C]KCN was carried out according to the two-step procedure of MacLeod et al. (17). Acanthamoeba myosins (0.2 to 1 mg) that had been dialyzed against deionized water and lyophilized were suspended in 0.5 to 1 ml of a solution containing  (105) were analyzed on 7.5% acrylamide gels. This sample of intact Acanthamoeba myosin II had partially broken down during prolonged storage in sodium dodecyl sulfate. The minor bands of less than 170,000 were not present in the material subjected to trypsin digestion. For details see "Experimental Procedures." 6 M guanidine hydrochloride, 10 mM dithiothreitol, 1 mu EDTA, and 0.1 M Tris-chloride, pH 8.2. The reaction was allowed to proceed at room temperature for 4 h. Dithionitrobenzoic acid was added to 50 mM, the pH was adjusted to 8, and the mixture was allowed to stand for 4 h at room temperature. Excess reagent was removed by exbaustive dialysis against water and the modified protein was recovered by lyophilization. The pellet was resuspended in 0.5 to 1 ml of 6 M guanidine, 1 mM EDTA, 0.1 M Tris-chloride, pH 8.2, containing 50 &i of ['Y?]KCN and the solution was left at room temperature for 15 min. Nonradioactive KCN was added to 50 mM and, after an additional 15 min at room temperature, the protein was desalted by passage through Sephadex G-25. The protein peak was pooled and incubated for 18 h at 37°C to allow the cleavage reaction to occur. The high molecular weight reaction products were recovered from Sephadex G-25, dialyzed against water, and lyophilized. The yield of the cleavage reaction was calculated from the per cent of the radioactive protein that was recovered.
The cyanogen bromide cleavage was generally carried out by reacting proteins in 80% formic acid at room temperature for 67 h with a 500-fold excess of cyanogen bromide relative to methionine residues (17). Most of the solvent and reagent were removed by drying under reduced pressure and residual formic acid was removed by adding water and lyophilizing.

RESULTS
Tvptic Digestion ofAcanthamoeba Myosin II-There was almost no change in either Ca'+-ATPase activity (600 and 622 nmol . min-' . mg-', before and after) or (K',EDTA)-ATPase activity (158 and 114 nmols min-' e mg-', before and after) when Acanthamoeba myosin II (400,000 daltons) was subjected to limited digestion with trypsin as, described under "Experimental Procedures." The two major reaction products (native molecular weights about 310,000 and 125,000) sepa-II IA ..a TOOK IA rated by gel filtration both contained a major polypeptide of about 100,000 daltons and three polypeptides between approximately 70,000 and 75,000 daltons by dodecyl sulfatepolyacrylamide gel electrophoresis (Fig. 1). These results suggest that the larger product was a two-headed enzyme (analogous to muscle heavy meromyosin) and the smaller product a single-headed enzyme (analogous to subfragment-1), both with enzymatic properties similar to those of Acanthamoeba myosin II. Complete enzymatic activity was also recovered after further tryptic digestion of the isolated 310,000-dalton product product (Ca2+-ATPase, 593 mnol . min-' -mg-'; (K+,EDTA)-ATPase, 105 mnol~min-'srng-').
The major products of this second digestion had apparent molecular weights of about 275,000 and 105,000, by agarose gel chromatography, and both major peptide components of about 70,000 daltons and 55,000 daltons (Fig. 1). None of the products of tryptic digestion was a substrate for Acanthamoeba myosin I heavy chain kinase (6). Thus, although tryptic digestion can convert Acanthamoeba myosin II into products with native and subunit molecular weights less than those of Acanthamoeba myosin IA, IB, or IC, the products retain the enzymatic properties characteristic of the Acanthamoeba myosin II from which they were derived. Acanthamoeba myosin II is also not converted into enzymes resembling Acanthamoeba myosin I when incubated with papain' or with whole cell extracts (5,8).
Protease Peptide Maps -The products formed by Staphylococcus aureus V8 protease digestion of the 170,000-dalton heavy chain of Acanthumoeba myosin II, the 130,000-dalton heavy chain of Acanthamoeba myosin IA, and the lOO,OOOdalton products of tryptic digestion ofAcanthamoeba myosin II (see Fig. 1) are compared in Fig. 2 (left). There appear to be no common peptides in the maps derived from the heavy chains of Acanthamoeba myosins II and IA. As expected, ' H. Gadasi (Fig. 2, right) shows that the peptide maps of the 130,000-dalton heavy chains of IA and IC are identical to each other but are almost totally dissimilar to the peptide map of the 125,000-dalton heavy chain of IB. Certainly, there are many fewer, if any, peptides of identical electrophoretic 120mobility in the maps of Acanthamoeba myosins IB and IA than would be expected if the 125,000-d&on chain of IB were derived from the 130,000-dalton chain of IA. These results are extended and confiied by comparison of the peptide maps produced by limited proteolysis by chymotrypsin (Fig. 3). The heavy chains of Acanthamoeba myosins 11, IA, and IB produce maps with essentially no common peptides.
Cyanylation Peptide Maps-Chemical cleavage at cysteine residues by the cyanylation reaction produced different peptide maps for Acanthamoeba myosins IA, IB, and II (Fig. 4). By Coomassie blue staining, essentially one peptide of about 95,000 daltons was obtained from Acanthamoeba myosin II, while cyanylation of both IA and IB produced many different peptides mostly of lower molecular weight. At least five of the six major peptides derived from Acanthumoeba myosin IB had different electrophoretic mobilities than any of the peptides derived from IA and there were no peptides derived from IB corresponding to the higher molecular weight peptides derived from IA. Almost all of the peptides derived from Acanthamoeba myosins IA and IB were larger than their light chains and, therefore, must have been derived from their heavy chains even though the intact molecule was used in the cyanylation reaction. Acanthamoeba myosin II contains about 36 cysteine residues (8). We assume that there were multiple cleavages in the cyanylation reaction and that many small peptides were removed by chromatography on Sephadex G-25 or were unresolved at the bottom of the electropho-.' 42. FIG. 4. Comparison of the cyanylation products of Acanthamoeba myosins IA, IB, and II. The high molecular weight cyanylation cleavage products were recovered from Sephadex G-25 and analyzed on dodecyl sulfate-polyacrylamide gels (7.5%). Each of the three pairs of gels shows the intact myosin (left) and its cleavage products (right). retie gel. Cyanylation of the lOO,OOO-dalton tryptic peptide derived from Acanthamoeba myosin II (Fig. 1) gave the same pattern (not shown) as obtained for the native enzyme.
The results with Acanthamoeba myosin IA are incompatible with the absence of cysteine residues in the amino acid analyses reported by Pollard and Kom (2). Primarily for that reason, we repeated the cyanylation procedure using ['"Cl-KCN. Acanthamoeba myosins IA, IB, and II were all labeled with radioactivity and, therefore, each must have contained reactive cysteine residues. Between 48 and 56% of the radioactivity was recovered in the high molecular weight cleavage products recovered from each of the three myosins. Autoradiography of peptide maps obtained from these samples revealed labeled peptides corresponding to those shown in Fig.  4 except that the material at approximately 95,000 daltons derived from Acanthamoeba myosin II was resolved into two bands, presumably because of incomplete cleavage. The presence of cysteine residues in Acanthamoeba myosin IA was also confirmed by reaction with N-ethy1 [2,3-14C]maleimide under conditions that activate the Ca'+-ATPase activity and almost completely inhibit the (K+,EDTA)-ATPase activity of skeletal muscle myosin (12). Identical changes in enzymatic activities were obtained for Acanthamoeba myosin IA and dodecyl sulfate-polyacrylamide gel electrophoresis of the labeled enzyme showed the radioactivity to be localized almost entirely in the 130,000-d&on heavy chain.

Cyanogen Bromide
Peptide Maps-Similarly, the peptide maps of Acanthamoeba myosins IA, IB, and II obtained after chemical cleavage at methionine residues by cyanogen bromide show very few, if any, common bands (Fig. 5) and the maps, as a whole, are very different.
Cleavage of Acanthamoeba myosin II heavy chain was clearly incomplete since the sum of the molecular weights of the products (Fig. 5) exceeds the molecular weight of the heavy chain (170,000). Reaction in 2% dodecyl sulfate, 7% formic acid, conditions which others (18) have found to give more complete cleavage of muscle myosin than occurs in 80% formic acid, did not result in greater cleavage of Acanthamoeba myosin II (Fig. 5). There were too many peptides of too low molecular weight generated from Acanthamoeba myosins IA and IB to assess whether their cleavage was complete.

DISCUSSION
The identity of the peptide maps of the 130,000-dalton heavy chains of Acanthamoeba myosins IA and IC obtained by limited proteolysis by Staphylococcus aureus V8 protease confinms the conclusion tentatively reached in the accompanying paper from enzymatic data and subunit composition: Acanthamoeba myosin IC is Acanthamoeba myosin IA plus a 20,000-dalton regulatory peptide. In contrast, limited cleavage of the heavy chains of Acanthamoeba myosins IA, IB, and II by proteases, cyanylation, and cyanogen bromide produced sets of peptide maps that have very few, if any, peptides in common. The results are certainly incompatible with the predicted 95% identity of peptides if Acanthamoeba myosin IB were derived from IA and almost 75% identity if either were derived from Acanthamoeba myosin II. Although any one procedure might have given a misleading result, the fact that four different procedures gave the same answer strongly indicates that the amino acid sequences of the heavy chains of Acanthamoeba myosins IA( IB, and II are different and, therefore, that they are the products of different genes. The apparent light chains (1) and several other properties (2)(3)(4)(5)(6)(7) of these enzymes are also different, adding further strength to the conclusion that they are true isoenzymes. However, the relatively low molecular weights of these myosins, and especially the fact that Acanthamoeba myosins IA, IB, and IC are single-headed enzymes, makes it necessary to consider whether the isolated enzymes are degradation products of unidentified native molecules. No experimental evidence in support of this possibility has been found (1,4,5,8). Moreover, whether or not the isolated Acanthamoeba myosins are the functional forms in uiuo, there still must be at least three myosin isoenzymes in this organism because the peptide maps of the heavy chains of the isolated enzymes are too different for them to have been derived from a common precursor.
The ratio of Acanthamoeba myosin I to Acanthamoeba myosin II has remained the same in cultures maintained in this laboratory for almost 10 years (2) and Acanthamoeba myosins IA, IB, and IC occur in the same proportions in cultures that have been maintained separately for 5 years in different laboratories.
We have recently found that the myosin isoenzymes also occur in the same ratio in a strain of Acanthamoeba (I-D-4) cloned by Dr. R. J. Neff, Vanderbilt University, in 1976 and selected for its encystment efficiency. This strain also has different nutritional requirements than that which we used in most of our experiments.
We believe it is unlikely that alI these cultures would still contain the same mixed population of cells but absolute proof that the myosin isoenzymes are in a single cell will require cloning the amoebae' or reacting them with specific myosin antibodies.
Myosin isoenzymes have been shown to occur in rabbit skeletal muscle (19-23), nematode muscle (24), and chicken tissues (25), but not yet in any other non-muscle cell. Possibly, the complexity of motile activities in free living, exponentially growing amoeba requires the presence of myosins not required, at least in substantial amounts, in differentiated cells. Processes such as ameboid motility, cell division, saltatory motion, endocytosis, and exocytosis, although all dependent on actomyosin, may utilize significantly different molecular mechanisms involving different myosin isoenzymes. It is generally assumed that motile events in non-muscle cells involve a sliding filament mechanism analogous to that of muscle, with the actin filaments attached to the structures to be moved and the myosin organized into bipolar filaments that pull the actin filaments. Acanthamoeba myosin II (8), but not Acanthamoeba myosin I (2), has been shown to make bipolar filaments in uitro. More speculatively, actin filaments might be made to slide relative to one another by mechanisms requiring myosin molecules but not myosin filaments, for example, in the way that the tubulin-dynein system seems to function in ciliary axonemes (26,27). Moreover, movement of intracellular organelles has been envisaged as possibly occurring by the interaction of actin filaments with myosin attached to the organelles (28, 29) and such a mechanism also might not require filaments. Therefore, although we do not rule out the possibility that the single-headed Acanthamoeba myosins I may be degradation artifacts of two-headed myosins or, alternatively, that they may be able to form bipolar filaments, these myosins might function in Acanthamoeba by mechanisms that do not require myosin filaments.