Discovery of Acanthamoeba Myosin-I, the First Unconventional Myosin

I discovered myosin-I when I was postdoctoral fellow in the laboratory of Edward Korn at the National Institutes of Health during the period July 1969 through June 1972 (Pollard & Korn 1973). At that point in time, little was known about myosins in non-muscle cells, but we expected that the soil amoeba, Acanthamoeba, would yield a myosin similar to those from muscles. Adelman and Taylor (Adelman & Taylor 1969) and Hatano and Tazawa (Hatano & Tazawa 1968) had found a myosin with conventional properties in the acellular slime mold, Physarum, and Bob Adelstein, Mike Kuehl and I (Adelstein, Pollard & Kuehl 1971) were characterizing a conventional myosin from human platelets at the same time. In those days before antibodies to myosin, SDS-gel electrophoresis and cloned DNA, the only way to assay for myosin was to use an enzyme assay for ATPase activity. Given lots of helpful advice from a fellow postdoc, Evan Eisenberg, I decided to assay for what were the most characteristic ATPase activities of muscle myosin: activity in EDTA and 0.5 M KCl (K-EDTA ATPase activity) and Mg-ATPase activity stimulated by actin filaments. The former is an unphysiological quirk of myosin. The latter is the physiologially relevant activity linked to motor activity.

High salt or pyrophosphate extracted lots of K-EDTA ATPase activity from homogenized cells. On gel filtration columns this K-EDTA ATPase activity chromatographed as a single peak, which also had modest actin-activated Mg-ATPase activity. The Stokes' radius of this peak was large, but not nearly as large as a conventional myosin. Nevertheless, I pushed on to purify the enzyme. After much trial and error over two years I obtained a preparation of the K-EDTA-ATPase, judged to be pure, since the specific activity was constant over the peak of protein and enzyme activity on the last of several columns. Subsequently we were shown how to run SDS-polyacrylamide gels, confirming the purity of the enzyme. This must have been one of the last proteins purified without the benefit of running a gel! But there were several hookers here.

First, the actin-activated Mg-ATPase activity disappeared during the purification on a hydroxylapatite column. Everyone in the lab assumed that the enzyme was partially denatured during purification, since muscle myosin readily lost actin activation due to sulfhydryl oxidation. However, on a visit to the University of Chicago, Ed Taylor suggested that I had separated an activator from the myosin. He was right. When I recombined fractions from the hydroxylapatite column without enzyme activity with the purified myosin, I got whopping actin-activated ATPase activity - one of the most exciting experiments I have ever done. I used this assay to purify a 95 kD cofactor protein. This was the first example of a protein that activates a myosin. Later Maruta and Korn (Maruta & Korn 1977) showed that the cofactor is a protein kinase that phosphorylates the heavy chain. Yet later, Brzeska (Brzeska et al 1996) found that the heavy chain kinase is related to the yeast Ste6 kinase and a member of the PAK (p21-activated) family of kinases. Thus PAK was purified in 1972 for the first time.

Second, this myosin had unusual physical properties. The heavy chain was only 140 kDa and the light chains were 16 and 14 kDa. The Stokes' radius showed that the protein was globular with only one copy of each subunit. This myosin was soluble at low and high ionic strength. I saw no evidence for bipolar filaments like those we made from platelet myosin. Although monomeric, it aggregated actin filaments into parallel bundles, suggesting that it had two actin binding sites. None of this was similar to muscle myosin or to Physarum or platelet myosin. The existence of a second actin-binding site was confirmed later by Lynch in Korn's lab (Lynch, Albanesi et al. 1986), when he isolated a proteolytic fragment from the C terminus of the tail and showed that it binds actin filaments in the presence of ATP.

Third, in the presence of cofactor, the Mg-ATPase activity depended on the concentration of actin in an unconventional way. Eisenberg and Moos (Eisenberg & Moos 1968) had shown that muscle myosin subfragment-1 or heavy meromyosin was activated with a hyperbolic dependence on the concentration of actin filaments, suggesting a binding isotherm. On the other hand, Acanthamoeba myosin was activated maximally by very low concentrations of actin filaments. Medium concentrations gave less activation, followed by activation again at high filament concentrations. This was quite mysterious to us. Later Albansesi, Fujisaki, Korn and Pantaloni (Albanesi, Fujisaki & Korn 1985; Pantaloni 1985) showed that this was due to the second, ATP-independent actin filament binding site on the tail of amoeba myosin.

I first presented this work at the 1972 Cold Spring Harbor Symposium on Quantitative Biology on muscle and contractile proteins. The muscle cognoscenti were polite but informed me in later years that they did not believe my story. They were certain that I had isolated a proteolytic fragment of a "real" myosin. Of course we were vindicated in 1986, when John Hammer (Hammer, Jung & Korn 1986) cloned and sequenced the genes for amoeba myosin-IB and -IC. He showed that the myosin-Is are encoded by different genes than myosin-II. The gene for the original myosin-I, myosin-IA, resisted cloning until Mike Ostap succeeded in my lab in 1996. Wei-Lih Lee finished characterizing the myosin-IA gene and the properties of its tail (Lee et al 1999).

Myosin-I got its name in 1977-78, when Korn's lab (Maruta & Korn 1977) and my lab (Pollard, Stafford & Porter 1978) independently purified a conventional myosin-II from Acanthamoeba. This myosin had two heads and a tail 85 nm long. Since it had two heads, we called it myosin-II. Thus the unconventional single-headed myosin became myosin-I. (We never dreamed that the myosin class numbers would today reach 15.) I missed amoeba myosin-II in my original studies because it has low K-EDTA ATPase activity (compared with Ca-ATPase activity) and very little actin-activation of its Mg-ATPase activity, unless regulatory phosphates are removed from the C-terminal tail piece (Collins & Korn 1980).

Contributed by Thomas D. Pollard The Salk Institute for Biological Studies March 2000

Additional Reference
Pollard, T. D. & E. D. Korn (1973). "The "contractile" proteins of Acanthamoeba castellanii. Cold Spring Harbor Symp. Quant. Biol. 37: 573-583.

View links to recent papers on myosin biochemistry and function from the PUBMED database

Return to the Myosin Motility & Biochemistry Page
Return to the Myosin Home Page

If you use material from the reviews at The Myosin Home Page, we ask that you cite either the Home Page and authors, or the appropriate source publication in your work.

Copyright 2000 All rights reserved