ANALYTICAL
BIOCHEMISTRY
78, 374-394 (1977)
The Isolation of Microquantities of Myosin from Amoeba proteus and Chaos carolinensis JOHN S. CONDEELIS' Department of Biological Sciences, State University of New York Albany, New York 12222 Received June 7, 1976; accepted November 23, 1976 Myosin has been isolated from the large carnivorous amoebae Chaos carolinensis and Amoeba proteus to 90% purity with the aid of microtechniques. These techniques for electrophoresis, dialysis, calorimetric protein, and ATPase assays and centrifugation permitted the isolation, purification, and characterization of microgram quantities of myosin from less than 1 ml of packed cells. The purification utilized sucrose extraction of the actomyosin, low ionic strength precipitation, and final separation of the actin and myosin by gel filtration in the presence of KI. The properties of carnivorous amoeba myosin are similar to those of several other cytoplasmic myosins. Its heavy chain molecular weight is greater than that of the heavy chain of rabbit skeletal muscle myosin. The ATPase activities of myosin from C. carolinensis and A. proteus were almost identical and demonstrated inhibition by EDTA and Mg++ and activation by Ca++ in the presence of 0.6 M KC1 and the absence of actin. Actin activated the Mg++ ATPase activity by an average factor of 6. No cofactor protein was required for this activity. These properties are similar to those of myosin isolated from Starfish eggs, Physarum, and Dictyostelium. Carnivorous amoeba myosin bound rabbit muscle F actin in the absence but not in the presence of Mg++ and ATP.
A satisfactory understanding of amoeboid movement cannot be obtained until an integration of information at the cellular and molecular levels is achieved. In turn, the details of movement at the molecular level cannot be understood until the properties of the molecules responsible for transducing chemical energy to mechanical work have been characterized. Because of their size and ease of maintenance in the laboratory,Amoeba proteus and Chaos carolinensis have been studied most intensively, and current theories on the physiology of motive-force production in amoeboid movement have been derived mainly from them (1,2). However, it is significant that actin and myosin have not been isolated from these amoebae. The major impediment to the critical step of isolating and characterizing the contractile proteins in these cells has been the problem of obtaining sufficiently large quantities of cytoplasm to make isolation by standard techniques feasible. For example, the isolation of conveniently ’ Present address: The Biological Cambridge, Massachusetts 02138.
Laboratoties,
Harvard University,
16 Divinity Ave.,
374 Copyright 0 1977 by Academic Press, Inc. All rights of reproduction in any form reserved.
ISSN OIM3-2697
ISOLATION
OF MYOSIN
FROM
AMOEBAE
375
handled amounts of myosin (l-3 mg) from Physarum requires about 100 g of plasmodium (3). This makes the isolation of contractile proteins from such organisms as A. proteus almost unthinkable, since only 0.5 g of amoebae can be grown over a similar time interval with the same culturing effort. To overcome this difficulty, several microtechniques have been developed to establish the properties of carnivorous amoeba actin and myosin and their probable physiological role. The methods are discussed in this and two separate publications (4,5). This study was undertaken to establish the existence of myosin in the large free-living amoebae Chaos carolinensis and Amoeba proteus by biochemical criteria and to demonstrate the suitability of this molecule as a putative mechanochemical transducer. METHODS Culture. A. proteus and C. carolinensis were obtained from Wards Biological Supply in wild culture. Amoebae were removed from wild culture with a Pasteur pipet and were placed in 6-in. plastic dishes (Falcon) containing Prescott and James medium (PJ) consisting of 6 mg of KCl, 4 mg of CaHPO,, and 4 mg of MgSO, in 1 liter of distilled water. Cultures were established by feeding the amoebae Tetrahymena pyriformis , allowing the amoebae to attach to the plastic substrate, and then pouring off the medium and replacing it with fresh PJ solution. This procedure was carried out several times each week to remove bacteria, algae, and residual ciliates. Amoebae were fed every 2 days in order to maintain the most rapid rate of growth possible. Tetrahymena were cultured under sterile conditions in 2% proteose peptone (Difco) which had previously been autoclaved in 60-ml aliquots and stored in tightly closed milk-dilution bottles. Tetrahymena were allowed to grow to high density (ca. 1 week) and were then harvested by pelleting and rinsing the cells several times with PJ to remove the proteose peptone. Amoebae were harvested by pooling the separate cultures in a large container at room temperature. Because of their large size, the amoebae settled rapidly when suspended in solution, whereas the other cells present remained suspended. The supernate was then decanted, and the amoebae were resuspended in distilled water. Three rinses were sufficient to remove the contaminating organisms as determined by cell counting. Care was taken to remove the PJ to avoid contamination of the initial ATPase assays with the phosphate contained in the culture solution and to avoid the possibility that the Ca++ in the PJ might result in contraction of the Ca++-sensitive cytoplasm (6) during extraction, causing it to pellet as a gel in the first centrifugation. Cells were harvested at such a density that an
376
JOHN S. CONDEELIS
average of 0.1 ml of loosely packed cells were obtained per dish of A. proteus and 0.25 ml per dish of C. carolinensis. Myosin Extraction. Between 0.1 and 1 ml of packed cells was suspended in an equal volume of extraction solution containing 60% sucrose, 20 mM Tris-HCl, O-2 mM EDTA, 0- 10 mM EGTA, 0.2 M NaCl, 5 mM DTT, pH 7.5, and the solution was homogenized in the cold (4°C) by grinding in a motor driven Teflon glass homogenizer (A. H. Thomas Co.). Homogenization was followed using Nomarski differential interference optics and was completed after 20 passes. The homogenizer was rinsed with an equal volume of half-diluted extraction solution, and both fractions were pooled and stirred for 15 min at 4°C. The pooled homogenate was centrifuged for 1 hr at 23,000g yielding a clear supernate (Sl) and a white pellet (Pl). The supernate was dialyzed overnight against 500 vol of 0.08 M NaCl, 20 mM imidazole-HCl, 1 mM CaCl,, 1 mM MgC&, and 1 mM DTT at pH 6.4 at 4°C. The precipitate was collected at 100,OOOg for 1 hr. The pellet was resuspended by homogenization in 0.1 M NaCl, a fraction was sampled for protein using the method of Hat-tree (7), and the remainder was mixed with an equal volume of 1.2 M KI or NaI, 10 m&r ATP, 10 mM DTT, 20 mM imidazole-HCl, and 2 mM MgClz at pH 7.0. The mixture was gently homogenized and clarified at 45,OOOg for 30 min. The supernate (S3) was chromatographed on a 0.6 x 27-cm column according to the method of Pollard et al. (8). To concentrate the myosin for the ATPase assays, Peak I was dialyzed against 50 mM KCl, 1 mM MgC&, and 10 mM imidazole-HCl at pH 7.0. The thick filaments of myosin were collected by centrifugation at 20,OOOg for 1 hr. Preparation of chromatography columns. The agarose chromatography gel was supplied by the manufacturer as a preswollen slurry (Bio-Gel A-15m, 200-400 mesh from Bio-Rad). The 0.6 x 27-cm column (Isco) was packed with gel that was pre-equilibrated with 0.6 M KCl, 0.02% sodium azide, and 1 mM EDTA at pH 7.0. Care was taken during preparation of the columns to insure that the void volume determined for each varied by less than 1% over a period of several weeks. The void volume was determined using the virus SPOl and the salt boundary, by using ATP. The Stokes radius of amoeba myosin was measured by the method of Ackers (9). Ackers’ column calibration parameter (r) was calculated using the distribution coefficient (Kd) determined for ferritin and catalase. Electron microscopy. Contractile proteins were prepared for electron microscopy by the negative staining method of Moore et al. (IO). Observations were made with a Philips EM300 using an accelerating voltage of 60 kV, a condenser aperture of 250 pm, and an objective aperture of 50 pm. Preparation ofproteins from muscle. Myosin was prepared according to the method of Kielley and Harrington (11) from rabbit back and leg muscles. Muscle actin was purified from an acetone powder according to
ISOLATION
OF MYOSIN
FROM AMOEBAE
377
Spudich and Watt (12). Proteolytic subfragments of muscle myosin were prepared according to the methods of Lowey and Cohen (HMM) (13) and Syent Gyorgy et al. (LMM) (14). Microcolorimetric assays. Protein was measured by the Hartree method (7), modified for l- to 20-~1 volumes by a l/50 reduction of all reagent volumes. This permitted routine determination of 0. I- 10 pg of protein. Bovine serum albumin was used for all standard curves (Fig. 1). ATPase assays. The following sequence was followed for the incubation of samples with ATP. (i) Assay buffer, 70 ~1 (a buffer containing KC1 or NaCl and imidazole, pH 7.0, to give the final concentration required), was added to a Kimax 6 x 50-mm tube followed by (ii) l-10 ~1 of stock solution of either CaCl,, MgC&, EGTA, or EDTA; (iii) l- 10 ~1 of protein solution; (iv) 5-10 ~1 of 20 mM ATP at pH 7.0. The final incubation solutions had the following compositions. (a) The buffer used for all myosin assays contained 0.6 M NaCl or KCl, 10 mM CaCl,, 10 mM MgCl, or 2 mM EDTA, 10 mM imidazole-HCl, and 2 mM ATP, pH 7.0. (b) The buffer used for actin activation consisted of 0.03 M KCl, O-2 mM MgCl*, 10 mM imidazole-HCl, O-5 mM EDTA, and 1 mM ATP at pH 7.0. (c) The buffer used for Ca++-sensitivity experiments was 0.03 M KCl, 1 mM MgClz, 0.1 mM CaCl, or 1 mM EGTA, 10 mM imidazole-HCl, and 1 mM ATP, pH 7.0. Modified sample
Hortree volume I - 20microlr
ters
1.61.4 1.2 I.0 5: r- .8a .6-
Micrograms
FIG. 1. The standard curve for the Hartree protein microassay in low and high salt buffers.
378
JOHN S. CONDEELIS
Two methods were used for the calorimetric determination of inorganic phosphate. The Martin and Doty method (15) was used according to the modification of Pollard and Kom (19), but was further modified for small volumes by a 100 reduction of all reagent volumes. The Koser method (16) was also used for the determination of inorganic phosphate after modification as follows. (i) The enzymatic reaction was stopped by the addition of 200 ,ul of a 1: 1 mixture of isobutanol-benzene followed by the addition of 200 ~1 of 1.5% ammonium molybdate dissolved in 0.35 N H,SO,. (ii) The tube was vortexed for 30 set, and the organic phase was allowed to separate. (iii) The organic phase, 100 ~1, was then removed and transferred to a clean 5 x 50 mm-tube. (iv) The color was developed by adding 10 ~1 of fresh SnCl solution prepared by diluting 1 part of stock solution (1 g of SnCl/lO ml of concentrated HCl) in 25 parts of 1 N H,SO,. The high sensitivity of the modified Koser technique (Fig. 2) made it useful for determination of ATPase activity in dilute protein solutions. However, at high protein concentrations (>5 mg/ml) the‘ photometric response of the modified Koser method became nonlinear. Hence, for Inorganic phosphorus ossoy ossoy volume IO - /OOmicroliters 2.0-
1.8 1.6 1.4 1.2 -
0.8 -
and Ooty
Mortm
0.018
0.002 0.006 0.01 Micromole
Pi
FIG. 2. The standard curves for the phosphate microassays comparing the sensitivity achieved with the modified Koser and modified Martin and Doty micromethods.
ISOLATION
OF MYOSIN FROM AMOEBAE
379
determination of inorganic phosphate in homogenates or solutions containing high concentrations of protein, the modified Martin and Doty micro-method described before was used (Fig. 2). Determination of optical density. The optical density of solution volumes down to 30 ~1 was measured using a Cary 15 spectrophotometer with specially made matched apertures inserted in both the sample and reference beams as shown in Fig. 3. Cuvettes with a IO-mm light path and 2-mm-wide sample space were purchased from Pyrocell. Pin-hole apertures, 1.O-mm diameter, restricted the light path to the small solution volume as shown in the insert to Fig. 3, thereby avoiding light scattering at the cuvette walls and bottom and the liquid miniscus. The optical density of the solution was determined at 750 nm for the protein assay, at 720 nm for the phosphate assays, and at 280,290, and 320 nm for ultraviolet protein determinations. Microdialysis. Dialysis was performed using the method shown in Fig. 4. Narrow tubes were made from stock glass tubing one end of which was closed in a bunsen flame. For volumes between 5 and 50 ~1,2.0 x 8.5-mm tubes were used, and, for volumes between 50 and 100 ~1, 4 x 8.5-mm tubes were used. Kimax 6 x 50-mm tubes were used for volumes between 100 and 250 ~1. The solution to be dialyzed was introduced into the tube, and %-in. thin-walled tubing (A. H. Thomas 3787-D20) was cut into small squares, spread across the open end, and fastened to the tube using a rubber band. In preparation for use, dialysis tubing was boiled for 30 min in 10 mM EDTA at pH 7.0, rinsed several times in distilled water, and stored at 4°C in
FIG. 3. The placement of apertures in the Cary 15 spectrophotometer light path between the beam splitter and cuvettes. Aperture size, relative to cuvette and sample size (30 PI), is shown in the insert which is drawn to scale.
380
JOHN S. CONDEELIS
Sample volume 5-250microliters
Dialysis
I
-
membrane
/‘Sample
FIG. 4. The method of dialysis of very small sample volumes.
10% ethanol and 10 mM EDTA at pH 7.0. Rubber bands were washed in 7 x detergent, rinsed, and soaked in distilled water until use. Mcrocentrifugation. The method used for centrifugation of volumes from 5 to 100 ~1 is shown in Fig. 5. A 50-ml polycarbonate tube (Sorval No. 780) was used as a mold into which 25 ml of bioelastomer (Dow Corning Silastic 328) was poured after thorough mixing with a polymerizing catalyst. Before polymerization was completed, a 4OOql plastic centrifuge tube (A. H. Thomas 2591-D15) was pushed into the elastomer at a60”angle to the tube walls, so that the small tube had a final angle of 90” to the rotor shaft. In addition, a pair of g-in. forceps was pushed into the plastic, thus forming two grab slots with which the whole mold could be pulled from the bottom of the polycarbonate tube without inverting the assembly. The plastic was polymerized overnight at 37°C. A Sorval No. 780 tube was chosen because this size enjoys interchangeable use with the medium-speed Sorval SS34 and IEC870 and high-speed IEC H192 rotors. For the removal of microscopic pellets, the 400~~1 soft plastic tube was removed from the mold, and the pointed tip was cut off with a razor blade just above the pellet. The pellet was then easily visible in the dissecting microscope and could be rapidly resuspended in buffer with minimal loss of material. Microefectrophoresis .-SDS electrophoresis was carried out in 1.5mmdiameter capillaries according to the method of Condeelis (4). The gels were stained with Coomassie brilliant blue and densitometered with a Gilford spectrophotometer. The molecular weight of the heavy chain of
ISOLATION
OF MYOSIN
FROM
AMOEBAE
381
40
35
IO
5
0 Radius,
cm
FIG. 5. The method of microcentrifugation and distribution of g forces drawn to scale. Orientation of the microtube 90” to the rotor shaft allowed collection of the pellet in the pointed tip. Sample volume shown is 10 ~1.
carnivorous amoeba myosin was estimated relative to the mobility of rabbit skeletal muscle myosin and human erythrocyte spectrin in SDS gels, using subunit molecular weights of 240,OOOand 220,OOOdaltons for spectrin (3 1). RESULTS The micro methods used in the preparation of myosin are summarized in Fig. 1-5. These methods were specifically designed for small-scale preparative biochemistry. The need for special equipment or complex handling procedures has been avoided, allowing the rapid use of these simple procedures on a routine basis. A summary of the extraction method is presented in Fig. 6. The extraction procedure used here was essentially that of Clark and Spudich (17) which was used to extract myosin from the cellular slime mold Dictyostelium discoideum. The extraction solution was modified by the addition of millimolar EGTA to remove free Ca++ and aid in homogenization (6). With the use of this solution, 40-60% of the Ca++ ATPase present in the crude homogenate was extracted into Sl. Sl became slightly turbid during dialysis. Decreasing the pH to 6.0
382
JOHN S. CONDEELIS 0.01 Add
-
I.0
equal.
cc of
volume
of
Homogenize Add
equal
volume
of Stir
SpinTr
Pl
l/2
packed
washed
6 extraction
cells solution
4, 50 passes 4 diluted
extraction
solution
h 15 minutes q
<23,000
S1 Dialyze of buffer:
overnight adjust
against pH to 6.0
500 with
volumes acetate
h Stir
15
minutes J
Spin
100,000
g 1 hour
4 52
Resuspend in 0.1 add equal volume sample buffer. Spin k! P3
4 45,000
g 30
M NaCl; of NaI
homogenize chromatography
and
minutes
I 53
Chromatography
on
4% agarose
FIG. 6. Outline of the procedure for purification of carnivorous amoeba myosin. PI and Sl are the pellet and supemate of the first spin, P2 and S2 are the pellet and supemate of the second spin, and P3 and S3 are the pellet and supemate of the third spin, respectively.
caused the supernate to become cloudy and resulted in a 30% increase in the amount of Ca++ ATPase recovered in P2. Gel electrophoresis indicated that a large amount of the protein present in P2 was myosin (Fig. 7). The amount of Ca++ ATPase extracted from P2 into S3 was quite variable, ranging from 20 to 90%. Rapid loss of the total Ca++ ATPase activity was characteristic of this step, suggesting that degradation of the myosin was occurring after P2 was redissolved. Gel electrophoresis supported this suggestion. Figure 7 (S3) shows the appearance of a band at 200,000, the intensity of which increased with the decrease in Ca++ ATPase and the intensity of the band at 225,000. To avoid degradation at this step, the time between resuspension of P2 and application of S3 to the chromatography column was made as short as possible. Gel filtration using the discontinuous buffer system completed purification of the myosin. The elution pattern obtained in one of the experiments is shown in Fig. 8. An average of 70% of the Ca++ ATPase activity applied to the column was recovered in the second peak (Peak I). Part of the lost ATPase was recovered in the void volume which was showr to contain a variable amount of myosin in filamentous form by electror microscopy and gel electrophoresis. The myosin obtained from the columr
ISOLATION
OF MYOSIN
383
FROM AMOEBAE
225Km-; 200K -
Sl
Pl
P2
S3
M
A
T
FIG. 7. Electrophoretic summary in SDS-polyacrylamide microgels of the extraction of amoeba myosin following the outline in Fig. 6. Electrophoresis was carried out in 25 mM Tris-glycine at pH 8.0. Standards: M, muscle myosin; A, muscle actin; T, brain tubulin.
could be stored without proteolytic degradation for at least 1 month as judged by gel electrophoresis. Figure 9 shows comparative electrophoresis patterns obtained in SDS with muscle and amoeba myosin. The heavy chain of both C. carofinensis (Fig. SC) and A. proteus (Fig. 9b) myosin has a molecular weight of approximately 225,000 as compared with the heavy chain of muscle myosin at 200,000 daltons (Fig. 9a). The heavy chain constituted over 90% of the protein in the amoeba gels, indicating a high degree of purification. The only other band on the amoebagels was a 135,000 molecular weight band of variable staining intensity relative to the heavy chain. This lighter band comigrated with one of the peptides constituting the HMM subfragment of muscfe myosin (Fig. 9d), suggesting that the variable band was partially degraded amoeba myosin. No light chains were observed on the amoeba gels.
384
JOHN S. CONDEELIS Kd 0.5
0A0
20 0
2.5
25 5.0 ml
30
35 7.5
40
45 10.0
FIG. 8. Typical result of gel filtration of S3 prepared from 0.5 ml of packed cells. Input to the 0.6 x 27-cm column of 4% agarose was 0.09 ml. Pi (A ,& refers to free phosphate eluting from the column as a background to the Ca++ ATPase activity. Kd is the column distribution coefficient.
The third peak eluted from the column contained (Fig. 1Oa) a small amount of myosin in addition to the major peptide which had a relative mobility approximately equal to that of LMM from rabbit muscle (Fig. lob) suggesting that this band (lOa) might be a proteolytic fragment of the amoeba myosin. These results indicate that a large amount of proteolytic degradation of the myosin occurred during the extraction. In an attempt to prevent proteolysis, the esterase inhibitor, phenylmethyl sulfonyl fluoride, was used (18). No substantial inhibition of the proteolysis was observed, even at the high concentrations demonstrated to have a substantial inhibitory effect (18). However, rapid execution of the critical steps resulted in a reduction of the loss of myosin due to proteolysis. The last peak eluted from the agarose column contained inorganic phosphate (Fig. 8)) presumably split from ATP during preparation of S3. In addition, this peak contained a variety of unidentified low molecular weight proteins, one of which comigrated with rabbit muscle actin on SDS microgels. The crude homogenates contained high ATPase activities which could not be explained by myosin alone as judged by SDS gel electrophoresis. In order to determine if any of the Ca ++ ATPase present in the crude homogenate was due to the mitochondrial Ca++ ATPase, enzyme assays were performed in the presence of 6.6 pg/ml of oligomycin. This concentration is sufficient to inhibit ATP-dependent Ca++ uptake by
ISOLATION
225Ke2ooK/--
OF MYOSIN
-
FROM AMOEBAE
-
385
-r*r” *
.
I
.:
a
b0
c
‘d
FIG. 9. SDS-polyacrylamide microgels containing (a) purified rabbit muscle myosin, (b) purified myosin from Amoeba profelts, (c) purified myosin from Chaos carolinensis, and (d) rabbit muscle HMM. The buffer was 25 mM Tris-glycine, pH 8.0.
mitochondria. Slightly over 10% of the Ca++ activity present in the crude homogenate was inhibited, whereas oligomycin had no effect on purified myosin. ATPase Activity The ATPase activities of A. proteus and C. carolinensis myosins were essentially identical. The average values of four experiments with each species of myosin are shown in Table 1. The enzymatic activity of amoeba myosin was very unstable with the loss of 20% activity per 24 hr even in the presence of 1 mM EDTA which has been found to stabilize the activity of muscle myosin. Amoeba myosin ATPase was depressed by K+ EDTA
386
JOHN S. CONDEELIS
FIG. 10. (a) Peak II from the 4% agarose chromatography column (Fig. 8). The major peptide in this gel has the same mobility as LMM from rabbit muscle myosin shown in gel b.
solutions unlike muscle and Acanthamoeba myosin (19). However, amoeba myosin was activated by Ca++ and inhibited by Mg++ in the absence of actin. The Ca++ activity ranged from 0.08 to 0.2 pmol of Pi/min/mg for both species of myosin and was insensitive to the salt concentration of the assay mixture when assayed at KC1 concentrations between 0.06 and 0.6 M. Znteraction of Amoeba Myosin with Actin
The physiologically significant enzymatic activity of myosin is believed to be the hydrolysis of ATP in the presence of F actin and Mg++ at physiological ionic strength. This activity is thought to represent the primary mechanochemical event in motive force production. In order to characterize this interaction, amoeba myosin was incubated with muscle actin in a myosin/actin weight ratio of 1: 1 in the buffer described in
ISOLATION
OF MYOSIN TABLE
SPECIFIC
ACTIVITY
AND ACTIN
ACTIVATION
387
FROM AMOEBAE 1 OF CARNIVOROUS
AMOEBA
MYOSIN
Specific activity (pm01 of piiminlmg at 22°C)
Amoeba proteus Chaos carolinensis
K+ EDTA
Na+ EDTA
0.01 0.01
0.01 0.01
Ca++
Mg++
0.14 0.10
0.02 0.03
Actin activation (pmol of piiminimg of myosin at 22°C) Myosiniactin ratio by weight
Mg++ ATPase
Amoeba proteus Chaos carolinensis
Rabbit muscle HMM
- Actin
+ Actin
0.02 0.02 0.002
0.1 0.12
0.1
1
1 2
Methods. Muscle HMM was incubated with muscle actin under similar conditions for direct comparison. Amoeba myosin was activated by an average factor of 6 from a specific activity of approximately 0.02 in the absence of actin to 0.12 pmol of Pi/min/mg of myosin in the presence of actin. The results varied from a five- to sevenfold activation (Table 1). The maximum effective actin concentration for myosin activation was not determined. Although no attempt was made to precisely determine the Mg++ requirement, l-2 mM MgCl, was sufficient for activation, whereas 2 mM EDTA efficiently inhibited the actomyosin activity. Caf+ was not required for activation of the Mg++ ATPase of amoeba myosin with muscle actin. A substantial Ca++-regulated contraction has been observed in freshly isolated cytoplasm from these large carnivorous amoebae (6). In view of this phenomenon, it was important to ask if crude preparations of amoeba actomyosin are Ca++-sensitive. Various fractions were tested for Ca++ sensitivity during amoeba myosin extraction in the presence of 0.1 mM CaCl, or 1 mM EGTA and 1 mM MgClz at physiological ionic strength (0.05), using the buffer system described in Methods. No Ca++ regulation could be detected in steps as early as S 1, with Ca++/EGTA ATPase activity equaling unity (Fig. 6). To determine if this loss was due to the loss of Ca++-regulatory light chains on the myosin in the presence of EDTA (20), fractions prepared in the absence of EDTA were assayed in the presence of highly purified muscle F actin ( 12) in a 1: 5 protein/actin ratio. No Ca++ sensitivity was detected in any of the fractions, even when EDTA or EGTA was removed from the extraction solution.
388
JOHN S. CONDEELIS
An important part of the primary mechanochamical event is believed to be the physical binding of myosin to actin. In order to investigate actin binding by amoeba myosin, muscle F actin was applied to a Formvar grid at a concentration of 0.5 mg/ml, and the grid was rinsed several times with a solution of amoeba myosin monomer at 0.2 mg/ml. The result of negatively staining such a preparation is shown in Fig. 11. The binding of amoeba myosin to actin was visualized as a 36-nm periodicity resulting from the interaction of myosin with the helical substructure of actin. Myosin was released from the actin by rinsing the grid with a buffer containing MgCl, and ATP (Fig. 12). DISCUSSION
The simple micromethods described in this paper permitted the isoiation and characterization of microgram quantities of myosin from less than 1 ml of packed cells. The use of such simple procedures should allow the characterization of contractile proteins from other organisms that have played an important role in our understanding of motile processes at the cellular level but have not been studied biochemically. The giant free-living amoebae are extremely active in phagocytosis. Their ability to engulf prey organisms that are large relative to their own cell volume indicates that the amoebae must contain a substantial digestive capacity. Apparently, it is this capacity that represents a major difficulty in isolating myosin from carnivorous amoebae. Several techniques for the isolation of amoeba myosin were initially attempted. The high salt extraction of actomyosin was performed by homogenizing the amoebae in a buffer containing 0.6 M KCI. Under these conditions, the myosin was rapidly degraded in the crude homogenate as observed by the disappearance of the 225,000 molecular weight band from SDS gels. Low ionic strength extraction of the soluble proteins was attempted by initially homogenizing the amoebae in a buffer containing 0.01 M KCl. The initial extraction of soluble proteins reduced the rate of degradation of the myosin once it was separated from the homogenate by centrifugation. However, once pelleted, the actomyosin could not be resuspended under any of the high ionic strength conditions usually employed. This irreversible aggregation of actomyosin at low ionic strength has also been observed by investigators working with Dictyostefium actomyosin (17). Sucrose has been used successfully to reduce the amount of myosin degradation occurring in cell homogenates (17,21). Extraction in isotonic sucrose was attempted by homogenizing cells in 0.15 M solution. The amount of degradation was substantially reduced, but the yield of myosin was less than 0.5% of the total Ca ++ ATPase activity present in the homogenate. Raising the sucrose concentration to 30% in the homogenate (17) increased the yield to between 1 and 3% of the initially available Ca++ ATPase .
ISOLATION
OF MYOSIN
FROM
AMOEBAE
389
FIG. 12. A control preparation of amoeba myosin monomer and muscle F-actin after being rinsed with Mg++ and ATP, demonstrating complete dissociation of the myosin and actin.
ISOLATION
OF MYOSIN
FROM
AMOEBAE
391
The time in the feeding cycle at which the cells were harvested was also a factor in controlling the amount of myosin degradation occurring during extraction. If the cells were fed and then harvested within 12 hr after returning to the extended streaming morphology, the amount of degradation was substantially lower as compared with that occurring in cells harvested after being starved for 48 hr. Perhaps, during starvation, the primary lysosomes accumulate in the cytoplasm, resulting in a higher hydrolytic capacity at the time of cell homogenization. The preparation of S3 from P2 included the steps during which most of the proteolysis of myosin occurred. Proteolysis could be minimized by rapidly redissolving P2 and clarifying and chromotographing S3. Prolongation of these steps resulted in a greater loss of myosin with separation of the resulting subfragments of myosin during the chromatography step. This suggests that the proteolytic subfragments of amoeba myosin might be prepared by incubating P2 or S3 in solution for various time intervals to optimize the yield of the subfragment of interest. EDTA-inhibited Ca++ ATPase activities in the crude homogenate were high, ranging from 0.058 to 0.156 pmol of Pi/min. Spectrophotometer scans of SDS microgels indicated that only one-tenth ot one-twentieth of these activities could be accounted for by myosin alone which usually comprises less than 5% of the total cell protein (22). There are several possibilities as to what this nonmyosin Ca++ ATPase in the homogenate might be. (i) The giant amoebae contain 200,000 mitochondria per cell (23). Mitochondria have been shown to exhibit a substantial Ca++-activated ATPase in the presence of ATP and millimolar Ca++, resulting in ATP-dependent uptake of Ca ++ (24). This suggestion is partially correct since oligomycin was shown to inhibit over 10% of the initial Ca++ ATPase in the crude homogenate. (ii) Contraction in isolated amoeba cytoplasm has been shown to exhibit Ca++ sensitivity (6). It is possible that a membrane fraction capable of Ca++ uptake might be present in amoeba cytoplasm to regulate the in vivo Ca++ concentration. Much of the Ca++ ATPase present in the crude homogenate may be associated with such a membrane fraction. (iii) The presence of nonspecific phosphatases might be responsible for the large ATPase activity present in the crude fractions. Such phosphatases split ATP in the presence of Mg++ and Ca++. but are inhibited by EDTA (25) as was amoeba myosin in the present study. To date, cells exhibiting amoeboid movement have been shown to contain filamentous myosin, similar in physical properties to that of muscle myosin with one exception: globular Acanthamoeba myosin (19). The following evidence indicates that carnivorous amoeba myosin is filamentous. (a) During extraction of carnivorous amoeba myosin, several peptides were observed on SDS gels that have a similar or identical mobility to the muscle subfragments HMM and LMM. These peptides copurify with
392
JOHN S. CONDEELIS
carnivorous amoeba myosin. (b) The binding of carnivorous amoeba myosin to actin results in the usual arrowhead pattern found with all other filamentous myosins, whereas globular myosin from Acanthamoeba does not produce an arrowhead structure upon binding to actin, but simply increases the diameter of the thin filament (26). (c) The heavy chain weight of carnivorous amoeba myosin on SDS gels is characteristic of the filamentous myosins found in Dictyostelium , marine eggs, and Physarum and is heavier than the heavy chain of vertebrate cytoplasmic myosins (Table 2). All of these myosins have heavy chains that are much larger than globular Acanthamoeba myosin. It should be noted that no light chains were observed on SDS gels containing carnivorous amoeba myosin. However, no attempt was made to establish the presence and characteristics of light chains associated with carnivorous amoeba myosin. Such a study is now in progress. (d) The equivalent hydrodynamic sphere of carnivorous amoeba myosin (Stokes radius) is approximately equal to that of filamentous Physarum myosin (27) (Table 2). (e) Purified carnivorous amoeba myosin is capable of forming thick filaments similar to those TABLE CHARACTERISTICS
OF MYOSINS Heavy
Refere”Ce
chain (daltons)
2
ISOLATEDFROM
Stokes radius (nm)
Actin activation 20-50
Cofactor requiremerit
DIFFERENT
SOURCES
ATPase" K+ EDTA
Ca++
NO
I.5
Light chain tdaltons)
Mg++
"C
0.19
0.002
2s
I ?O.Oao 2 18,ooo I 16,000
Striated muscle (rabbit)
22.30.32
200,OQo
19.2
Granulocyte
?I
200,000
19
3
NO
0.2
0. I
0.01
25
? 20,ooo
Platelet
33.34.35
200800
19
4
NO
0.55
0.44
0.02
37
2 19.000 2 16.000
Fibroblast
36
2oo.oa
?
9
NO
0.5
0.24
0.01
37
? ?O.oa
MacrophaSe
28.37
2cacaO
?
IO
Yes
0.56
0.56
0.05
37
2 20,cNJ 2 15,t?OO
Starfish
38
2lO.tn30
19.5
7
NO
0.3
0.45
25
I 20,Oca I 17,tNO
240,000
17.1
16
NO
0.03
1.0
0.03
22
2 21,ooo I 17.000
9
NO
0.01
0.09
0.01
25
I lE.OrN3 I 16.000
Y.?S
3.5
0.4
0.05
29
? 16,WLl ? 14.000
egg
Phwmm
3.27
?
(This Paper)
n Micromoles
of pi per minute
225,000
per milligram
5.5
50
17.4
6
NO
0.01
0.14
0.02
22
?
17.4
6
NO
0.01
0. I
0.03
22
?
of myosin
ISOLATION
OF MYOSIN FROM AMOEBAE
393
formed with nonmuscle filamentous myosins. Globular Acanthamoeba myosin does not form thick filaments (26). The filament-forming properties of carnivorous amoeba myosin are reported in a separate publication (5). A summary of the enzymatic properties of carnivorous amoeba myosin as compared with myosin isolated from several different organisms is presented in Table 2. Unlike myosin isolated from vertebrate cells and Acanthamoeba , carnivorous amoeba myosin is strongly inhibited by K+ EDTA, a property shared, so far, only by Physarum and Dictyostelium myosin. It is difficult to assess the importance of these differences since Ca++ and K+ EDTA ATPases have no discernable physiological significance. Muscle actin activated the Mg++ ATPase of amoeba myosin. Although the maximum effective actin concentration was not determined, the activation at a myosin/actin weight ratio of 1: 1 compares well with that of cytoplasmic myosin preparations not requiring a cofactor. Although no cofactor requirement was observed, it is possible that a cofactor might be necessary for actin activation of carnivorous amoeba myosin to the high levels observed in organisms requiring a cofactor (26,28). Further work is required to determine if such a cofactor exists in these amoebae. The results just described suggest that Ca++ regulation in the large carnivorous amoebae is not myosin linked. The possible absence of myosin-linked regulation suggests that the isolation of crude preparations of amoeba actin may permit the characterization of the components endowing Ca++ regulation on amoeba cytoplasm. Work is now in progress to isolate actin, Ca++-sensitized actin, and Ca++-sensitized actomyosin from these amoebae. At present, the results of this study demonstrate that myosin is present in the giant free-living amoebae, and that its mechanochemical properties are consistent with a sliding filament model (29) explaining the contractile nature of amoeboid movement. ACKNOWLEDGMENTS It is my great pleasure to thank Dr. Robert D. Allen for his advice, support, and encouragement from the inception to the completion of this study. I also wish to thank Dr. F. Walz, Dr. D. Edwards, and Dr. H. Ghiradella for advice and valuable discussions, Dr. D. Shub for supplying the virus SPOl, Mr. M. Koser for communicating his phosphate assay and Mr. D. Rice, Mr. R. Speck. and Mr. R. Loos for excellent technical assistance. This study was supported by research Grants GM-18854 and GM-22356 from the National Institute of General Medical Sciences administered by Dr. R. D. Allen.
REFERENCES 1. 2. 3. 4.
Allen, R. D. (1961)Exp. CellRes. 8, 17-31. Rinaldi, R., and Jahn. T. (1%3) J. Prutozool. 10, 344-357. Nachmias, V. T. (1974) J. Cell Bid. 62, 54-65. Condeelis. J. S. (1977) Anal. Biochem. 77. 195-207.
394
JOHN S. CONDEELIS
5. Condeelis, J. S. (1976) J. Cell Sci., in press. 6. Taylor, D. L., Condeelis, J. S., Moore, P. L., and Allen, R. D. (1973) J. Cell Biol. 59, 378-394. 7. Hartree, E. F. (1972) Anal. Biochem. 48, 422-435. 8. Pollard, T. D., Thomas, S. M., and Niederman, R. (1974)Anal. Biochem. 60,258-266. 9. Ackers, G. K. (1964) Biochemistry 3, 723-730. 10. Moore, P. L., Condeelis, J. S., Taylor, D. L., and Allen, R. D. (1973) Exp. Cell Res. 80, 493-49s. 11. Kielley, W. W., and Harrington, W. E. (1959) Biochim. Biophys. Acra 41, 401-422. 12. Spudich, J. A., and Watt, S. (1971) J. Biol. Chem. 246, 4866-4874. 13. Lowey, S., and Cohen, C. (1962) J. Mol. Biol. 4, 293-308. 14. Szent Gyorgyi, A. G., Cohen, C., and Philpott, D. E. (1960) J. Mol. Biol. 2, 133- 142. 15. Martin, J. B., and Doty, D. M. (1949) Anal. Chem. 21, 965-970. 16. Koser, M., unpublished method. 17. Clarke, M., and Spudich, J. A. (1974) J. Mol. Biol. 86, 209-222. 18. Fahmey, D. E., and Gold, A. M. (1963) J. Amer. Chem. Sot. 85, 997-1000. 19. Pollard, T. D., and Kom, E. D. (1973) J. Biol. Chem. 248, 4682-4690. 20. Lehman, W., Kendrick Jones, J. and Szent Gyorgyi, A. G. (1973) Cold Spring Harbor Symp. Quant. Biol. 37, 319-330. 21. Stossel, T. P., and Pollard, T. D. (1973) J. Biol. Chem. 248, 8288-8294. 22. Pollard, T. D., and Weihing, R. R. (1974) Crit. Rev. Biochem. 2, l-60. 23. Wolfe, S. C. (1972) Biology of the Cell, p. 103, Wadsworth, Belmont, California. 24. Spencer, T., and Bygrave, F. L. (1973) Bioenergetics 4, 347-362. 25. Bingham, E. W., and Zittle, C. A. (1%3) Arch. Biochem. Biophys. 101,471-482. 26. Pollard, T. D., and Kom, E. D. (1973) J. Biol. Chem. 248, 4691-4697. 27. Adelman, M. R., and Taylor, E. W. (1%9) Biochemistry 8, 4976-4988. 28. Stossel, T. P., and Hartwig, J. H. (1975) J. Biol. Chem. 250, 5706-5712. 29. Huxley, H. E. (1973) Nature (London) 243,445-450. 30. Eisenberg, E., and Moos, C. (1968) Biochemistry 7, 1486- 1498. 31. Kirkpatrick, F. (1976) Life Sci. 19, l-18. 32. Hayashi, Y., and Tonomura, Y. (1970) J. Biochem. 68, 665-674. 33. Adelstein, R., and Conti, M. (1972)ColdSpting HarborSymp. Quant. Biol. 37,599-605. 34. Adelstein, R., Pollard, T., and Kuehl, W. (1971) Proc. Nat. Acad. Sci. USA 68, 2703-2707. 35. Pollard, T., (1975) in Molecules and Cell Movement (Inout, S., and Stephens, R., eds.), Vol. 30, pp. 259-286. Raven Press, New York. 36. Adelstein, R., Conti, M., Johnson, G., Pastan, I., and Pollard, T. (1972) Proc. Nut. Acad. Sci. USA 69, 3693-3697. 37. Hartwig, J., and Stossel, T. (1975) J. Biol. Chem. 250, 56%-5705. 38. Mabuchi, I. (1976) J. Mol. Biol. 100, 569-582.