THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 264, No. 32, Issue of November 15, pp. 19333-19339,1989 Printed in U. S. A.

Purification and Characterization of a Third Isoform of Myosin I from Acanthamoeba castellanii*

(Received for publication, July 3, 1989)

Thomas J. Lynch, Hanna Brzeska, Hidetake Miyata, and Edward D. Korn From the Laboratory of Cell Biology, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892

A third isoform of myosin I has been isolated from Acanthamoeba and designated myosin IC. Peptide maps and immunoassays indicate that myosin IC is not a modified form of myosin IA, IB, or 11. However, myosin IC has most of the distinctive properties of a myosin I. It is a globular protein of native M, -162,000, apparently composed of a single 130-kDa heavy chain and a pair of 14-kDa light chains. It is soluble in MgATP at low ionic strength, conditions favoring filament assembly by myosin 11. Myosin IC has high Ca2+- and (K+,EDTA)-ATPase activities. Its low Mg2+-ATPase activity is stimulated to a maximum rate of 20 s-l by the addition of F-actin if its heavy chain has been phosphorylated by myosin I heavy chain kinase. The dependence of the Mg2+-ATPase activity of myosin IC on F-actin concentration is triphasic; and, at fixed concentrations of F-actin, this activity in- creases cooperatively as the concentration of myosin IC is increased. These unusual kinetics were first dem- onstrated for myosins IA and IB and shown to be due to the presence of two actin-binding sites on each heavy chain which enable those myosins I to cross-link actin filaments. Myosin IC is also capable of cross-linking F- actin, which, together with the kinetics of its actin- activated Mg2+-ATPase activity, suggests that it, like myosins IA and IB, possesses two independent actin- binding domains.

Myosin I was first isolated from Acanthamoeba castellanii by Pollard and Korn (1) and subsequently shown to exist as two isoforms termed myosins IA and IB (2). Peptide mapping showed that their heavy chains (140 kDa for myosin IA and 125 kDa for myosin IB) differed in primary structure (3). Their single light chains (17 kDa for myosin IA and 27 kDa for myosin IB) have not been studied in great detail since they are not essential for the ATPase activities of myosin I (4). A third form’ was also identified, but it appeared to be myosin IA with an additional, weakly associated 20-kDa poly- peptide (2).

Under physiological conditions, myosins IA and IB are globular, monomeric proteins, lacking the extended tail of other myosins and showing no tendency to self-associate into ordered structures (1, 5). Based on its functional properties (6, 7) and primary structure (8, 9), the NHZ-terminal 80 kDa of the heavy chain of each myosin I is analogous to the

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This modified form of myosin IA had been referred to as myosin IC, a term which now refers to the third form of myosin I described in this paper.

globular head (subfragment 1) of muscle myosin. The remain- ing COOH-terminal portion of the myosin I heavy chain possesses binding sites for membranes (10) and for F-actin (11) which presumably anchor myosin I to these structures, while the subfragment 1 domain generates force against the actin filament with which it associates.

In the absence of actin, both myosins IA and IB are highly active Ca2+- and (K+,EDTA)-ATPases (1, 2). The actin-acti- vated M$+-ATPase activities of both myosins depend on the phosphorylation of a single site on their heavy chains, located near the center of their subfragment 1-like domains (12-15). The Me-ATPase activity of both myosins I is a complex function of the concentration of F-actin (14, 16) and, at fixed concentrations of actin, exhibits positive cooperativity with respect to myosin concentration (17). Both properties arise from the ability of myosin I to cross-link actin filaments (11).

In this paper, we report the existence of a third isoform of myosin I in Acanthamoeba which we refer to as myosin IC. Its physical and enzymatic properties resemble those of my- osins IA and IB, but its heavy and light chains are unique. In the accompanying paper (15), we report the sequences of the regulatory phosphorylation sites of myosins IA, IB, and IC. In the course of this work, it became evident that the myosin IC heavy chain is the product of the gene previously identified (18) as coding for the myosin IB heavy chain and that the myosin IB heavy chain is the product of the gene previously identified (9) as coding for a myosin IL heavy chain.

MATERIALS AND METHODS AND RESULTS~

DISCUSSION

Myosin IC is the third isozyme of myosin I to have been isolated from Acanthamoeba. Myosin IC is clearly a myosin by virtue of its ATPase activities, including its actin-activated M$’-ATPase activity. However, its size and native molecular weight are those of a globular, monomeric (single headed) protein similar to myosins IA and IB and different from the larger and highly asymmetric two-headed myosin 11. In par- ticular, the small frictional ratios of all three myosins I indicate the absence of extended rod-like tails. Consequently, myosin IC is freely soluble at low ionic strength in the presence of MgATP, conditions that induce nearly quantita- tive assembly of myosin I1 into bipolar filaments (19-21). It

Portions of this paper (including “Materials and Methods,” “Re- sults,” Figs. 1-10, and Tables 1-111) are presented in miniprint at the end of this paper. The abbreviations used are: SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; EGTA; [ethylene- bis(oxyethylenenitrilo)]tetraacetic acid TLCK, N“-p-tosyl-L-lysine chloromethyl ketone; TPCK, N-tosyl-L-phenylalanine chloromethyl ketone. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

19333

19334 Acanthamoeba Myosin IC

should be noted that, in the absence of nucleotide, myosin IC does precipitate if the concentration of KC1 is reduced much below 100 mM; this nucleotide-dependent solubility is shared by myosins IA and IB.

Myosin IC is also immunologically distinct from myosin I1 as it is recognized by an antiserum raised against myosin IB that does not detectably react with myosin 11. Finally, myosin IC is a substrate for Acanthamoeba myosin I heavy chain kinase that does not phosphorylate myosin 11. Phosphoryla- tion of what appears to be a single site on the myosin IC heavy chain increases its actin-activated Me-ATPase activ- ity about 14-fold, which is low compared to the 40-80-fold stimulation of myosins IA and IB. Since this difference is primarily due to the relatively high actin-activated M F - ATPase activity of myosin IC before phosphorylation in vitro and since the maximal incorporation of phosphate in vitro is less than 0.9 mol/mol of myosin IC, we think it likely that the heavy chain of myosin IC, as isolated, is phosphorylated at its regulatory site to a greater extent than is myosin IA or IB .

At moderate to high concentrations of actin, the M e - ATPase activity of phosphorylated myosin IC is an apparent first-order function of actin concentration. However, this relationship does not hold at low concentrations of actin or if the ratio of actin to myosin is decreased by increasing the concentration of myosin IC in the assay. Specifically, the actin-activated Mg?'-ATPase activity of myosin IC displays myosin-dependent, positive cooperativity that is characteris- tic of myosins IA and IB (17). It has been shown that this cooperativity is due to the ability of myosin I to cross-link actin filaments (6,11,22,23). We infer the same relationship for myosin IC since it also cross-links F-actin. Furthermore, there is direct evidence for two actin-binding domains in the myosin IA heavy chain which would permit it to bind simul- taneously to two actin filaments (11). Although we have not addressed this point, the overall similarity between myosins IA and IC in their physical properties, enzymatic activities, and interactions with actin, suggests that myosin IC also possesses two independent actin-binding domains.

Despite the similarities among the three myosins I, their heavy chains differ sufficiently in primary structure to be distinguishable from each other by limited peptide mapping. Hence, the three isozymes are not derived from each other or from a common precursor. Furthermore, partial sequence data obtained directly from all three myosins I (11, 15) and the complete sequences of two of them deduced from their ge- nomic DNA (8,9) show conclusively that the three myosin I heavy chains are products of different genes. Data in the accompanying paper (15) indicate that the heavy chain of myosin IC is the product of the gene originally identified as coding for the heavy chain of myosin IB (18). This misiden- tification occurred because the existence of myosin IC was not then known; the heavy chains of myosins IB and IC have similar mobilities on sodium dodecyl sulfate-polyacrylamide gel electrophoresis; and, if it behaved like the anti-myosin IB serum used in this study, the anti-myosin IB serum used previously to identify the i n vitro translation products would have cross-reacted with the myosin IC heavy chain. The myosin IB heavy chain in fact appears to be encoded by the recently sequenced gene (9) designated myosin IL (see accom- panying paper (15)). Whereas the misidentification of the genes does not alter any previous conclusion or interpretation, future experiments with the proteins that rely on their de- duced sequences require their correct identification.

In early studies (1,2), myosin IA was reported to contain a single heavy chain of 130-140 kDa and two light chains of 17

and 14 kDa; and myosin IB, a single heavy chain of 125 kDa and two light chains of 27 and 14 kDa. The amounts of the 14-kDa light chain associated with each isoform varied from preparation to preparation, and its stoichiometry was always less than 1:l with respect to the heavy chain and the other light chain. In more recent preparations of myosins IA and IB (6, 7, 11, 14, 24), the 14-kDa light chain has been absent altogether; and the native molecular weights of these two proteins were consistent with each possessing a single light chain (5). We now think it very likely that the early prepa- rations of both myosins IA and IB included variable amounts of myosin IC.

The question naturally arises why Acanthumoeba possesses three isoforms of myosin I, which in turn is subsumed under the broader question of the in situ function of myosin I. Recently, two physiological roles for myosin I have been suggested propelling intracellular organelles along actin fila- ments (10, 25) and generating force in the cortex of the cell to drive processes such as the advance of the leading edge of a migrating cell or particle capture at the onset of phagocytosis (26). Myosin I may be associated with the actin filament matrix immediately beneath the plasma membrane or directly associated with the plasma membrane or with membranous organelles (10, 25, 27, 28). Insofar as their interactions with F-actin have been characterized i n vitro, all three myosin I isozymes would behave similarly in their associations with actin filaments unless the interactions of each myosin I iso- zyme with actin were differentially modulated by other factors i n situ. There is as yet no evidence for such a mechanism. Membrane-bound myosin I may be distributed uniformly on the cytoplasmic face of the plasma membrane or membrane- bound organelles and activated, presumably by heavy chain phosphorylation, only where needed. Since all three myosin I isozymes appear to be equivalent substrates for the myosin I heavy chain kinase, this would not account for the diversity of myosin I isoforms. Alternatively, myosin 1 may only inter- act with the membranes at specific sites. This would imply the existence of membrane receptors for myosin I which could discriminate between myosin I isoforms. Presumably, myosin I would associate with membranes through its COOH-termi- nal domain(s) in order to allow the NHP-terminal, catalyti- cally active 80-kDa segment to interact with actin (10, 25). Interestingly, it is in their COOH termini that the greatest divergence in the sequences of the myosin 1B and IC heavy chains occurs (8,9); the myosin IA heavy chain has not been sequenced. If these segments of the myosin I heavy chains were capable of interacting with other cellular components, such as a putative set of membrane receptors, the differences in primary structure could specify a discrete set of interactions for each myosin I isoform.

1.

2.

3.

4.

5.

6.

7.

8.

REFERENCES Pollard, T. D., and Korn, E. D. (1973) J. Bid. Chem. 248,4682-

4690 Maruta, H., Gadasi, H., Collins, J. H., and Korn, E. D. (1979) J.

Biol. Chem. 254,3624-3630 Gadasi, H., Maruta, H., Collins, J. H., and Korn, E. D. (1979) J.

Bioi. Chem. 254,3631-3636 Maruta, H., Gadasi, H., Collins, J. H., and Korn, E. D. (1978) J .

Biol. Chem. 253,6297-6300 Albanesi, J. P., Fujisaki, H., Hammer, J. A., 111, Korn, E. D.,

Jones, R., and Sheetz, M. P. (1985) J. Bid. Chem. 260,8649- 8652

Brzeska, H., Lynch, T. J., and Korn, E. D. (1988) J. Bioi. Chem.

Brzeska, H., Lynch, T. J., and Korn, E. D. (1989) J. Bwl. Chem.

Jung, G., Korn, E. D., and Hammer, J. A., I11 (1987) Proc. Natl.

263,427-435

264,10243-10250

Acad. Sci. U. S. A. 84,6720-6724

Acanthamoeba Myosin IC 19335 9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21. 22.

23.

Jung, G., Schmidt, C. J., and Hammer, J. A., I11 (1989) Gene

Adams, R. J., and Pollard, T. D. (1989) Nature 340,565-568 Lynch, T. J., Albanesi, J. P., Korn, E. D., Robinson, E. A., Bowers,

B., and Fujisaki, H. (1986) J. Biol. Chem. 261, 17156-17162 Maruta, H., and Korn, E. D. (1977) J. Biol. Chem. 2 5 2 , 8329-

8332 Hammer, J. A., 111, Albanesi, J. P., and Korn, E. D. (1983) J.

Biol. Chem. 258,10168-10175 Albanesi, J. P., Hammer, J. A., 111, and Korn, E. D. (1983) J.

Biol. Chem. 2 5 8 , 10176-10181 Brzeska, H., Lynch, T. J., Martin, B., and Korn, E. D. (1989) J.

Biol. Chem. 264, 19340-19348 Pollard, T. D., and Korn, E. D. (1973) J. Biol. Chem. 248,4691-

4697 Albanesi, J. P., Fujisaki, H., and Korn, E. D. (1985) J. Biol. Chem.

Hammer, J. A., 111, Jung, G., and Korn, E. D. (1986) Proc. Natl.

Pollard, T. D., Stafford, W. F., 111, and Porter, M. E. (1978) J.

Collins, J. H., Kuznicki, J., Bowers, B., and Korn, E. D. (1982)

Pollard, T. D. (1982) J . Cell Biol. 9 5 , 816-825 Albanesi, J. P., Coue, M., Fujisaki, H., and Korn, E. D. (1985) J.

Albanesi, J. P., Lynch, T. J., Fujisaki, H., and Korn, E. D. (1986)

(Amst.), in press

260,11174-11179

Acad. Sci. U. S. A. 8 3 , 4655-4659

Biol. Chem. 253,4798-4808

Biochemistry 21,6910-6915

Biol. Chem. 2 6 0 , 13276-13280

J. Biol. Chem. 261 , 10445-10449

24.

25. 26.

27. 28.

29.

30.

31.

32. 33. 34.

35.

36. 37. 38.

39.

40. 41.

Lynch, T. J., Brzeska, H., and Korn, E. D. (1987) J . Biol. Chem.

Adams, R. J., and Pollard, T. D. (1986) Nature 322 , 754-756 Fukui, Y., Lynch, T. J., Brzeska, H., and Korn, E. D. (1989)

Gadasi, H., and Korn, E. D. (1980) Nature 286,452-456 Miyata, H., Bowers, B., and Korn, E. D. (1989) J. Cell B i d , in

Spudich, J. A., and Watt, S. (1971) J. Biol. Chem. 2 4 6 , 4865-

Collins, J. H., and Korn, E. D. (1981) J. Biol. Chem. 2 5 6 , 2586-

C&, G. P., Robinson, E. A., Appella, E., and Korn, E. D. (1984)

Bradford, M. M. (1976) Anal. Bwchem. 7 2 , 248-254 Laemmli, U. K. (1970) Nature 227, 680-685 Otter, T., King, S. M., and Witman, G. B. (1987) Anal Biochem.

262,13842-13849

Nature 341 , 328-331

press

4871

2595

J . Biol. Chem. 259,12781-12787

162,370-377 Hawkes, R., Niday, E., and Gordon, J. (1982) Anal. Biochem.

119,142-147 Laskey, R. A. (1980) Methods Enzymol. 65, 363-371 Maruta, H., and Korn, E. D. (1981) J. Biol. Chem. 256,499-502 MacLean-Fletcher, S. D., and Pollard, T. D. (1980) J. Cell Biol.

Fujisaki, H., Albanesi, J. P., and Korn, E. D. (1985) J. Biol. Chem.

Korn, E. D. (1982) Methods Cell Biol. 25B, 313-333 Albanesi, J. P., Fujisaki, H., and Korn, E. D. (1984) J. Bid. Chem.

85,414-428

2 6 0 , 11183-11189

259, 14184-14189

IUTERIALS AND UETKODS

Analvtical centrifuqatiog - Analyses were performed as descrlbed by Albanesi et al. ( 5 ) . Samples Of myosln IC vere dlalyned lnto 10 Ipl( TrLS.HC1, pH 7.5, 100

Measurements were made I n an AN-D rotor fitted Wlth a double sector cell, run EO! KC1, 1 EO! dithiothreltol, at a final concentratlon of 0.16 mg/ml.

sedlmentatlon equlllbrium Ya6 determined after Centrlfuqatlon at 11.000 rpm, at I" a Beckman model E centrlfuqe equlpped vlth a photoelectric scanner. The

3 ' C and 4.5 'C, far 16 h. SedLmentation velocity was measured at 23 'C for 45 mln after reaching 48.000 TPm.

IC was facllrtated by several revlslona of the previously described procedure Purification of mvoain I - The purification of adequate quantities of myosin

r l 7 l i 1 0 ) . The oresence of mvosln I durlnm the Durlfication was monitored

The cells, about 1 kg, were harvested by centrlfuqation at 1000 x q far 5 .in The axenic qrovth of Rcanthalnoeba csstellanj2 has been fully descrlbed (40).

and washed t w i c e with about 3 Val of 10 mJ4 imldazole.HC1. DH 7 . 0 , 7 5 mJ4 NaCl. at 4 .C. All subsequent Steps Were at 4 .C . The cell pellets Yere resuspended ln 2 vol of 3 0 EO! laidasole.HC1, pH 7.5, 75 mJ4 KC1, 12 EO! rodiun pyrophosphate,

pepstatin A, 2-5 mq/llter leupeptln, 0.5 m phenylmethylsulfonyl fluoride, and 5 m dlthrothreitol, 20 aq/llter soybean trypsln Inhibitor. 10 mq/liter

honoqenlzed ~n 100 lo1 Dounce homogenizers (type B pestles, 15 strokes, an ice). The homogenate Was centrlfuqed for 1 h In a Sorvall GSA rotor a t 12.000 rpa. The turbid Supernatant was brought to 0.5 m diisop~~pylfluorophosphate and

was decanted carefully from the pellet and turbid layer at the bottomb of the centrlfuqed for 6 h rn a B e c b a n type 19 rotor at 14,000 rpm. The supernatant

centrlfuqe bottles. The vol~me of the recovered supernatant was approximately equal to the volume of extraction buffer added rnitially. The supernatant. which was usually about pH 6.6. Vas titrated to about pH 8 . 0 by the addltion Of

~ ~ ~~

19336 Acanthamoeba Myosin IC

*e 43 'L P-11 ' ADP-Ag

&o*/ A B C A B C

31- .L' P

- "" u

)LC -

-200

-116

-66

.45

.3 1

,21.5 44.4 ,6.5

Acanthamoeba Myosin IC 19337

r

P - l l IC -./ IA

0 4 0 8 2 0

1 2 1 2 1 2

II I6 IC

anti-MI anti-Mll ".- v:"

II IA I6 IC It IA IB IC

'.O t

- ;

0.2 1 / I I IA IC IB II IA IC IB

I I I I I I O 20 40 60

[Aclm]. pM

80 100 1

Acanthamoeba Myosin IC

I I I I

IMYOSINI. pM

I I I I 0.1 0.2 0.3 0.4

[MYOSIN]. pM

Y

-26

HI

e7

*S 4

3

Coomassle

IA IC PC__

c. c- "3 "- ""

-&e-

"

4. ""

3- 8-

8- ""-A" -

-70.

-42 -38

'H-UTP-Labeled

IA "- IC

HC- L. 3 -HC

Acanthamoeba Myosin IC

MYOSIN IC TABLE I1

PHYSICAL PROPXRTIES OP LCANTHIllOBBA MYOSIN I ISOZYUES

Myosln IAa uyosm Myosin IC

4 Molecular Welqht

0 38 66 108 130

MYOSIN IA

OR

4

NH?-

0 38 64 91 112 140

k n ' H - U T P - I A B E L E D B 3 ' P - I A B E L E D

rig. 10. schematie_unr.q.ntation of the oriqin of the tnmtic Dentides Of

apparent molecular weights Of the tryptic peptides estimated by SDS-PACE; those B Y O ~ I ~ S I A am3 I C . Numbers are in kDa: those within the bars refer to the

under the bars refer to the distances from the NH terminus Of the sites Of cleavac~e. The Dositions of the major trVDtiC Sites $; myosin IA. when digested under ion-denatbring conditions, wire p;dviously determcnned ( 2 4 ) . The proposed map for myosin IC is based on the data from Figs. 8 and 9 and ~n the assumption that the phosphorylatlon site and UTP-labeling site are located along the heavy chain in Dositions similar to those Of myosin IA. The central 70-kDa fragment of myosln'Ic is further digested primaril'y at one sxte (yieldmg the 28- and 42-

at two major tryptic sites. ThlS map is confirmed in the accompanying paper kDa peptides) whereas the central 74-kDa fragment of myosin IA can be cleaved

( 1 5 ) which reports the NH teminal sequence of the 70-kDa fragment and the phosphorylation site locat!ei wlthln it. When calculated from the amino acld sequence deduced from Its genomlc DNA ( a ) , the lndlcated tryptlc sltes of myosln

welght of its heavy chain 1s 121 koa. IC are located 35- , 60.. and 104-kDa from the NH2-termlnus and the m o l e c v l a r

TABLE I

PORIPICATION OP ACANTHIllOEBA MYOSIN I ISOBYMKS

iK*.EDTAI-ATPase

Fraction Protei" Total Specific Activlty Yleld

mq paol/mln Pmol/mln/mg %

Honogenate 4 9 , 8 2 0 9 4 0 0 . 0 2 1 0 0

Extract 2 4 , 6 1 5 6 8 6 0 . 0 1 7 3

OE-52 5,116 1 5 6 0 . 0 7 38

216 184 119

100 7 9 29

0 . 4 6 0 . 4 2 0 . 2 4

22

A ADP-A9 B

c

A Mono-Q B

C

2 4 . 3 4 . 5 4 . 3

6 . 7 4.1 1 4 . 4 2 . 0 1 2 . 7

6 7 . 8 2 . 8 3 5 . 2 1 4 . 4

1 . 8 12 3 . 3

59.8 8.9 8 . 4 6.3

11

Native

SDS-PAGE

1 5 9 , 0 0 0 1 5 0 , 0 0 0 1 6 2 . 0 0 0

t17.000 140,000 1 2 5 , 0 0 0

t 2 7 . 0 0 0 t 1 4 . 0 0 0 1 3 0 , 0 0 0

Stakes Radius (nm) 6 . 2 5 . 9 6 . 3

Sedimentation Coefficients 6 . 6 5 6 . 5 5 6.14

Frictional Ratio 1 . 7 1 . 7 1 . 7

TABLE I11

ATPASE ACTIVITIES OP ACANTHMOEBA IYOBIW I ISOZYMBB

Condltmns Myosin IA Myosin IB Myosin IC

5"

Kt, EDTA 2 2 . 2 2 1 . 4 1 5 . 3

ca2+ 2 . 1 4.4 3 . 1

Mgz+ 0 . 3 0.3a 0 . 2

Mg" + F-Actin 0.6 0 . 5 a 1 . 5

Mq" + F-Act1n +

Heavy Chaln Phosphorylated 1 8 . 1 1 7 . 4 ' 2 0 . 1

'Data from Albanesl et al. (14. 1 7 ) .