Download Profilin association with monomeric actin in

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Cytosol wikipedia , lookup

Cell growth wikipedia , lookup

Cell cycle wikipedia , lookup

Signal transduction wikipedia , lookup

Endomembrane system wikipedia , lookup

Tissue engineering wikipedia , lookup

Mitosis wikipedia , lookup

Extracellular matrix wikipedia , lookup

Cellular differentiation wikipedia , lookup

Cell culture wikipedia , lookup

Organ-on-a-chip wikipedia , lookup

Cell encapsulation wikipedia , lookup

JADE1 wikipedia , lookup

Cytoplasmic streaming wikipedia , lookup

Actin wikipedia , lookup

List of types of proteins wikipedia , lookup

Cytokinesis wikipedia , lookup

Amitosis wikipedia , lookup

Transcript
3779
Journal of Cell Science 112, 3779-3790 (1999)
Printed in Great Britain © The Company of Biologists Limited 1999
JCS0568
Profilin is predominantly associated with monomeric actin in Acanthamoeba
Donald A. Kaiser1, Valda K. Vinson2, Douglas B. Murphy2 and Thomas D. Pollard1,*
1Structural Biology Laboratory, Salk Institute for Biological
2Department of Cell Biology and Anatomy, Johns Hopkins
Studies, 10010 N. Torrey Pines Rd, La Jolla, CA 92037, USA
Medical School, 725 N. Wolfe Street, Baltimore, MD 21205-2196, USA
*Author for correspondence (e-mail: [email protected])
Accepted 4 August; published on WWW 18 October 1999
SUMMARY
We used biochemical fractionation, immunoassays and
microscopy of live and fixed Acanthamoeba to determine
how much profilin is bound to its known ligands: actin,
membrane PIP2, Arp2/3 complex and polyproline
sequences. Virtually all profilin is soluble after gentle
homogenization of cells. During gel filtration of extracts on
Sephadex G75, approximately 60% of profilin
chromatographs with monomeric actin, 40% is free and
none voids with Arp2/3 complex or other large particles.
Selective monoclonal antibodies confirm that most of
the profilin is bound to actin: 65% in extract
immunoadsorption assays and 74-89% by fluorescent
antibody staining. Other than monomeric actin, no major
profilin ligands are detected in crude extracts. Profilin-II
labeled with rhodamine on cysteine at position 58 retains
its affinity for actin, PIP2 and poly-L-proline. When
syringe-loaded into live cells, it distributes throughout
the cytoplasm, is excluded from membrane-bounded
organelles, and concentrates in lamellapodia and sites of
endocytosis but not directly on the plasma membrane.
Some profilin fluorescence appears punctate, but since no
particulate profilin is detected biochemically, these spots
may be soluble profilin between organelles that exclude
profilin. The distribution of profilin in fixed human A431
cells is similar to that in amoebas. Our results show that
the major pool of polymerizable actin monomers is
complexed with profilin and spread throughout the
cytoplasm.
INTRODUCTION
annexin-I (0.0015 second−1; Alvarez-Martinez et al., 1996) are
exceptions.
Profilin may be essential because it interacts directly with
actin monomers. In the absence of profilin, actin-dependent
processes fail in flies (Cooley et al., 1992; Verheyen and
Cooley, 1994), Dictyostelium (Haugwitz et al., 1994) and S.
pombe (Balasubramanian et al., 1994). Profilin binding to actin
is required for the N-WASp-induced microspike formation in
vertebrate cells (Suetsugu et al., 1998). Given the high
concentration of unpolymerized actin in non-muscle cells
(Bray and Thomas, 1976) and the high affinity of profilin for
actin monomers, a large fraction of profilin is expected to be
bound to actin monomers. This conclusion is supported by the
copurification of actin with profilin from cellular extracts by
poly-L-proline affinity chromatography (Tanaka and Shibata,
1985) and other methods (Tseng et al., 1984), but these studies
have not been quantitative.
Interactions of profilin with proline-rich sequences of
WASp family proteins, FH-domain proteins and VASP may
also be essential and may, in fact, contribute to the actin
defects in cells lacking profilin. The polyproline binding site
(Archer et al., 1994; Metzler et al., 1994; Mahoney et al.,
1997; Eads et al., 1998) is one of the most conserved features
of profilins, a family of proteins with highly variable
sequences. Since the binding sites for poly-L-proline and actin
(Schutt et al., 1993) are spatially separated, profilin can bind
poly-L-proline and actin simultaneously (Perelroizen et al.,
The small protein, profilin, is essential in yeast (Magdolen et
al., 1988; Balasubramanian et al., 1994), flies (Cooley et al.,
1992; Verheyen and Cooley, 1994) and mice (Witke et al.,
1993), but we do not know exactly why. The situation is
complicated by multiple profilin ligands: actin monomers
(Carlsson et al., 1977); polyphosphoinositides (Lassing and
Lindberg, 1985); the Arp2/3 complex (Machesky et al., 1994);
annexin-I (Alvarez-Martinez et al., 1996); and poly-L-proline
(Tanaka and Shibata, 1985) and proline-rich sequences of
proteins such as VASP (Reinhard et al., 1995), FH-domain
proteins (Frazier and Field, 1997) and N-WASp (Suetsugu et
al., 1998). Actin monomers (Perelroizen et al., 1994; Hajkova
et al., 1997; Vinson et al., 1998) bind profilin with the highest
affinity, followed by PIP2 (Machesky et al., 1990), Arp2/3
complex (Mullins et al., 1998b) and poly-L-proline
(Perelroizen et al., 1994; Petrella et al., 1996). GST profilin-I
and profilin-II bind proline-rich sequences of N-WASp with
submicromolar affinity (Suetsugu et al., 1998), but less is
known about the affinity of profilin for VASP and FH-domain
proteins. Most of these interactions are dynamic at steady state
due to high dissociation rate constants: 5 seconds−1 from actin
(Perelroizen et al., 1994; Vinson et al., 1998), >100 seconds−1
from poly-L-proline (Archer et al., 1994) and >50 seconds−1
from Arp2/3 complex (Mullins et al., 1998b). Slow
dissociation from PIP2 (0.1 second−1; Machesky, 1994) and
Key words: Antibody, Cell motility, Endocytosis, Pseudopod
3780 D. A. Kaiser and others
1994). Profilin binding to polyproline sequences in N-WASp
appears to be required for efficient N-WASp-induced
microspike elongation in vertebrate cells (Suetsugu et al.,
1998). A synthetic lethal interaction between the genes for
profilin (cdc3) and the FH-domain protein (cdc12) shows that
these proteins interact in vivo (Chang et al., 1997). FH-domain
proteins are required for S. pombe to form an actin filament
contractile ring (Chang et al., 1997) and for S. cerevisiae to
mate and establish an asymmetrical distribution of actin
patches (Evangelista et al., 1997). FH-domain proteins also
are required for cytokinesis in Drosophila (Castrillon and
Wasserman, 1994) and Aspergillus (Harris et al., 1997). On
the other hand, normal affinity for polyproline may not be
essential for profilin function, since a deletion mutant with
decreased affinity for polyproline can rescue profilin null
strains of S. cerevisiae (Haarer et al., 1993).
Localization of profilin in live and fixed cells has not
answered the question: which molecular partners bind profilin
in cells? After fixation, anti-profilin antibodies stain the
cytoplasm of amoebas (Tseng et al., 1984), plant cells (Vidali
and Hepler, 1997), vertebrate cells (Buss et al., 1992; FaivreSarrailh et al., 1993; Watanabe et al., 1997) as well as the
cleavage furrow of Tetrahymena (Edmatsu et al., 1992) and the
septum region of S. pombe (Balasubramanian et al., 1994;
Chang et al., 1997). Some profilin antibodies stain the nucleus
of fixed cells (Tseng et al., 1984; Mayboroda et al., 1997).
Fluorescent rat profilin-I, labeled on residue 41 and
microinjected into cultured cells, concentrates at cell-cell
contacts and in small spots rich in monomeric actin
(Tarachandani and Wang, 1996). Gold-labeled profilin
antibodies bind in the cytoplasm and to regions of plasma
membrane free of actin filaments in leukocytes and stimulated
platelets (Hartwig et al., 1989). Polyphosphoinositides may
mediate this binding to membranes (Lassing and Lindberg,
1985), but the fraction of profilin bound to membranes has not
been measured.
Using complementary biochemical and morphological
methods we find that a majority of Acanthamoeba profilin is
bound to actin monomers and is spread relatively uniformly
throughout the cytoplasm. In gel-filtration experiments no
profilin was associated with particles larger than monomeric
actin, so the other partners of profilin must be present in low
concentrations or have low affinity for profilin and actinprofilin complexes. The biologically significant conclusions of
this work are that most profilin in Acanthamoeba is bound to
actin monomers, profilin-actin complexes are the major pool
of polymerizable actin in the cell and Mg2+-ATP actin-profilin
complexes elongate the barbed ends of actin filaments at 70%
the rate of free actin monomers.
2 mM imidazole, pH 7.0, 1.0 mM EGTA, 0.2 mM ATP, 0.5 mM DTT,
0.2 mM MgCl2, 20 mM KCl, 1.0 mM NaN3.
Profilins
Profilin-I and profilin-II were purified from extracts of Acanthamoeba
by DEAE, poly-L-proline and CM-500 cation exchange
chromatography (Kaiser et al., 1989). Vinson et al. (1998) describe
the preparation of recombinant Acanthamoeba profilin-II cysteine
mutant rPIIN58C by PCR mutagenesis. Recombinant proteins were
expressed in E. coli, BL21 and purified according to the method of
Kaiser and Pollard (1996) but with 1-10 mM dithiothreitol in all
buffers. N58C-profilin-II was labeled specifically on its single
cysteine with tetramethylrhodamine maleimide (mixed isomers,
catalogue number T489, Molecular Probes) to stoichiometries of
0.2-0.4 dyes per profilin (Vinson et al., 1998). Rhodamine-N58Cprofilin-II was separated from free dye by gel filtration on Sephadex
G25 (fine), adsorption to poly-L-proline-agarose, elution with 8 M
urea, extensive dialysis against syringe loading buffer and
ultracentrifugation. Recombinant human platelet profilin was also
purified from BL21 DE3 bacterial lysates by poly-L-proline affinity
chromatography (Almo et al., 1994).
Actin polymerization
Monomeric actin was purified from Acanthamoeba (Pollard, 1984)
and stored in 2 mM Tris-HCl, pH 8.0, 0.1 mM CaCl2, 0.2 mM ATP,
0.5 mM DTT and 1 mM NaN3 at 4°C. For polymerization assays, actin
was supplemented with 5-8% pyrene labeled actin monomers
(Pollard, 1984) and fluorescence emission at 407 nM was monitored
after excitation at 365 nM using a PTI Alpha-scan spectrofluorometer
(Photon Technology International, Princeton, NJ). Ca2+ actin
monomers were converted to Mg2+ actin monomers by treatment with
200 µM K+-EGTA, pH 7.0, and 50 µM MgCl2 on ice prior to use.
MATERIALS AND METHODS
Monoclonal antibodies
We characterized mouse monoclonal antibodies to Acanthamoeba
profilin in detail (Kaiser et al., 1986, 1989; Vandekerckhove et al.,
1989; Kaiser and Pollard, 1996). Antibodies P1, P2, P3, P4, P5, P6
and P7 bind to both profilin-I and profilin-II. Antibodies P8 and P9
are specific for profilin-II. Antibodies P1, P2, P3, P8 and P9 do not
bind native profilin in solution but bind to profilin that has been
denatured or partially unfolded by immobilization on microtiter wells
or nitrocellulose. Antibody P4 immunoprecipitates free profilin and
profilin bound to actin, but antibody P5 immunoprecipitates only free
profilin. On immunoblots of whole Acanthamoeba, antibodies P1-P7
bind only to bands having the same mobility as profilin-I and
profilin-II (Kaiser et al., 1986, 1989; Fig. 1A). Antibodies P8 and P9
recognize a single band having the same mobility as profilin-II.
Monoclonal antibodies were purified from mouse ascites fluid by
ammonium sulfate fractionation and DEAE chromatography (Kiehart
et al., 1984). Monoclonal antibodies were labeled on lysines with
Cy3™ using the manufacturer’s protocol. Typically, 1.0 mg/ml of
antibody in coupling buffer was reacted with a molar excess of dye
for 30-60 minutes at 22°C. Labeling stoichiometries were controlled
by adjusting the buffer pH and the time of labeling. Monoclonal
antibodies were directly labeled with between 3.6 and 4.4 Cy3
dyes/IgG1, separated from free dye by gel filtration on Sephadex G-50
and dialyzed extensively.
Buffers
Dulbecco’s phosphate buffered saline (DPBS): 0.1 g/l CaCl2, 0.3 g/l
KCl, 0.2 g/l KH2PO4, 0.1 g/l MgCl2·6H20, 8.0 g/l NaCl, 2.16 g/l
Na2HPO4·7H2O; coupling buffer: 100 mM Na2CO3, pH ~8.5-9.5.
Syringe-loading buffer: 50 mM alanine, 50 mM proline, 24 mM
glutamine, 40 mM γ-aminobutyric acid, 1 mM EGTA, 2 mM MgCl2,
pH 6.5 (Sinard and Pollard, 1989). Tween-buffer: 150 mM NaCl,
0.1% (v/v) Tween-20 (polyoxyethylene-sorbitan monolaurate),
10 mM Tris-HCl, pH 7.5, 0.01% (w/v) Thimerosal. Extraction buffer:
Polyclonal antibodies
Purified recombinant human platelet profilin (Almo et al., 1994) was
conjugated to keyhole limpet hemocyanin using glutaraldehyde
(Reichlin, 1980). Rabbits were inoculated with a 1:1 mixture of
conjugate in Freund’s complete adjuvant and bled after several boosts
with conjugate in Freund’s incomplete adjuvant. Anti-profilin
antibodies were purified by affinity chromatography of serum on
recombinant human platelet profilin coupled to Sepharose CL4B. The
column was rinsed with DPBS, DPBS with 0.5 M KCl and antibodies
Profilin association with monomeric actin in Acanthamoeba 3781
were eluted with 1.0 M glycine-HCl, pH 2.5. Fractions were collected
in 0.1 M Tris base and dialyzed into DPBS. These antibodies bound
to a single band with the mobility of profilin on immunoblots of whole
A431 (Fig. 8A), HeLa and bovine aortic endothelial cells (data not
shown). Affinity purified polyclonal antibodies to recombinant human
platelet profilin-I were labeled with Cy3 dyes and purified like
monoclonal antibodies.
Fractionation of cell extracts
About 10 g of amoebas from 1 liter log-phase cultures in Neff’s
medium in shaken Fernbach flasks were washed with 50 mM KCl
at 22°C, mixed with 10 ml of ice-cold extraction buffer and lysed
with 18 strokes of a Dounce homogenizer. Separate 1.0 ml samples
of the crude cell homogenate were kept on ice (extract sample),
centrifuged at 16,000 g in an Eppendorf 5415C microfuge for
5 minutes at 4°C to yield a low speed supernatant (LSS) and
centrifuged at 200,000 g in a Beckman TL100 ultracentrifuge for
30 minutes at 4°C to yield a high speed supernatant (HSS). To
determine the concentration of profilin in the extracts, samples of
10-50 µl were diluted into boiling sample buffer for analysis by
SDS-PAGE and quantitative immunoblotting with purified profilin-I
and profilin-II standards. To analyze the pellets, 0.4 ml of the crude
cell homogenate was layered over a cushion of 0.8 ml extraction
buffer containing 20% sucrose before centrifugation at low or high
speed. Pellets were resuspended in 0.4 ml of sample buffer and run
on SDS-PAGE for immunoblotting. We also determined the
concentration of profilin in extracts using competitive binding
assays. Profilin standards and unknown extract samples were serially
diluted and tested for their ability to compete with purified profilin
adsorbed to microtiter wells for binding anti-profilin monoclonal
antibodies (Kaiser and Pollard, 1996).
Half milliliter samples of the LSS and HSS were fractionated on
a 1.5 × 25 cm column of Sephadex G75 (medium) in extraction
buffer at 4°C. Half milliliter fractions were analyzed separately or
in pools representing (a) large complexes (in the void volume,
calibrated with blue dextran), (b) profilin bound to monomeric actin
(calibrated with monomeric actin) and (c) free profilin (calibrated
with profilin). After SDS-PAGE (Laemmli, 1970) of samples
proportional in size to the volume of the pooled fractions, proteins
were transferred to BA-83 nitrocellulose membranes (Schleicher
and Schuell). The membranes were blocked and incubated with antiprofilin monoclonal antibodies, horseradish peroxidase-labeled
secondary antibodies and ECL chemiluminescence reagents
(Amersham). Luminograms of profilin standards of 1-200 ng and
unknowns were scanned with a ScanMaker III (MicroTek) and
quantitated with NIH-image software. Column fractions were also
adsorbed to microtiter wells (NUNC #439454) and incubated with
anti-profilin monoclonal antibodies, anti-actin monoclonal antibody
c4d6 (kindly provided by Dr James Lessard; Lessard, 1988) and
anti-Arp2 and Arp3 polyclonal antibodies (Kelleher et al., 1995).
Solid-phase ELISA was performed as described (Kaiser and Pollard,
1996).
Bead adsorption assays
Purified monoclonal antibodies P3, P4 and P5 were conjugated to
CNBr-activated Sepharose CL4B beads (Pharmacia) at a
concentration of 5 mg antibody per ml of packed beads. Poly-Lproline (Sigma, 40,000 MW) was conjugated to beads at 20 mg/ml
(Kaiser et al., 1989). For adsorption, 0.1-1.0 ml samples were mixed
with 0.1-0.5 ml packed beads in 1.5 ml microcentrifuge tubes and
incubated for 1 hour on ice with occasional mixing. Beads were
washed three times by centrifugation in 1.0 ml extraction buffer for 1
minute each and extracted with 6 M urea to release bound proteins
while retaining the covalently bound antibodies and poly-L-proline on
the beads. Samples were boiled in SDS-PAGE sample buffer and
analyzed by SDS-PAGE. Coomassie blue stained gels were scanned
and quantitated as above.
PIP2-binding
PIP2 (Boehringer Mannheim) was suspended at a concentration of
924 µM in 30 mM Hepes buffer, pH 7.3, 0.02% NaN3 and sonicated
for 1 minute at 22°C in a bath sonicator (Laboratory Supplies Co. Inc.,
Hicksville, NY). Mixtures of 185 µM PIP2 and 28.4 µM profilin were
chromatographed on a 0.75 × 47 cm column of Sephadex G-100
(Pharmacia) in 30 mM Hepes, pH 7.3, 0.02% NaN3. Fractions of
10 drops were collected and analyzed by SDS-PAGE. Coomassie blue
stained gels were scanned to determine the fraction of profilin bound
to PIP2 in the void volume, in trailing fractions and free profilin
fractions.
Syringe-loading
Cells were syringe-loaded as previously described (Clarke and
McNeil, 1992; Doberstein et al., 1993). Approximately 106
Acanthamoeba castellanii (Neff strain) from exponential cultures
were pelleted at 1000 g for one minute and resuspended in 300 µl of
labeled profilin-I or profilin-II in syringe-loading buffer, which
mimics intracellular conditions (Sinard and Pollard, 1989). This
suspension was forced through a 30 gauge needle by hand pressure
in a 1 ml tuberculin syringe and collected in a plastic microfuge tube.
The cells were redrawn through the needle and expelled twice more
making a total of five passes through the needle. Cells were rinsed by
centrifugation three times in 1.0 ml of Neff’s medium before
observation.
Cell fixation
Acanthamoeba from log-phase cultures were seeded onto glass
coverslips in Neff’s medium and grown for 2-96 hours at 22°C. The
favored method of fixation was to immerse coverslips in 3 ml of 20%
DPBS at pH 6.5 containing 1% glutaraldehyde at room temperature
for 15 minutes. After rinsing 3 times in 20% DPBS, the coverslips
were treated 10 times for 3-5 minutes each with a 3 ml solution of
0.1% sodium borohydride in 20% DPBS. Cells were permeabilized
with a solution of 0.2% Triton X-100 in 20% DPBS for 5-10 minutes.
Cells were incubated in 20% DPBS with 0.67% bovine serum albumin
(BSA) for 5-10 minutes before incubating with Cy3-labeled
monoclonal antibodies at concentrations between 1-100 nM in 20%
DPBS or 100% DPBS. Rhodamine phalloidin (Molecular Probes) was
diluted to 1-3 units per milliliter into DPBS with 0.67% bovine serum
albumin and used to stain fixed amoebas on coverslips. After 60
minutes incubation at room temperature, the coverslips were rinsed
five times in 20-100% DPBS and mounted on glass slides in 50%
glycerol. For determination of the state of profilin and its localization
in cells fixed in different ways, glutaraldehyde-fixed cells were
compared with cells fixed by immersion of coverslips in −20°C
methanol for 30 seconds or treatment with 3% trichloroacetic acid
(TCA) in either methanol at −20°C or 20% DPBS at room temperature
for 30 seconds. Cells were also fixed with 1.0% formaldehyde in
methanol at −20°C (Yonemura and Pollard, 1992). Human A431 cells
were fixed for 5 minutes in 4% formaldehyde in DPBS and
permeabilized with either 0.2% Triton X-100 or 100% methanol.
Antigen fixation experiments
Aliquots of 100 µl of 70 µM profilin-I or profilin-II were incubated
with 1 ml of several different fixatives: control 20% DPBS at room
temperature; 1.0% glutaraldehyde in 20% DPBS at room temperature;
or methanol ± 3% TCA at −20°C. After 15 minutes each profilinfixative solution was diluted 50-fold into 10 mM Tris-HCl, pH 8.0,
and 50 µl aliquots were dried down in microtiter wells for enzyme
linked immunosorbant assay (ELISA) using fourfold serial dilutions
of monoclonal antibodies to profilin in Tween buffer containing 1.0%
BSA (Kaiser et al., 1989).
Microscopy
We examined fluorescent cells with several light microscopes, all
equipped with epi-illuminators, 100 W mercury arc lamps and high-
3782 D. A. Kaiser and others
contrast, fluar-type oil immersion objectives. Live cells were observed
with a Zeiss Axiovert 135 fluorescence microscope or an Olympus
IMT2-NIC microscope. Electronic imaging of fluorescent specimens
was performed using a Photometrics cooled, charge-coupled device
(CCD) camera (PXL-1400) controlled with IPLab Spectrum 3.1P
software (Scanolytics, Fairfax, VA) on Macintosh computers or with
a Hamamatsu analog ICCD imaging system (C2400-98) and image
processor (Argus-10). A Bio-Rad MRC-600/Nikon Optiphot confocal
imaging system was used for confocal imaging of fixed cells. Phasecontrast images were acquired with a Hamamatsu analog CCD camera
(C2400-77). Differential interference contrast (DIC) microscopy was
performed with a ×100, 1.3 Neofluar lens. Illumination was provided
by a fiber optic cable attached to a mercury vapor arc. A Scion frame
grabber (LG-3) was used to transfer image frames from video tape to
a computer hard drive where they could be processed with NIH Image
and Adobe Photoshop software. Film-based (Plus-X 400 ISO film)
fluorescent and phase-contrast images of fixed cells were obtained
using a Leitz Orthoplan microscope. We quantitated the fluorescence
intensity of glutaraldehyde-fixed cells stained with Cy3 labeled
monoclonal antibodies P4 and P5 by scanning 35 mm negatives in a
Polaroid Sprint Scan 35 scanner using Adobe Photoshop v3.0
software. The fluorescence intensity of individual cells was measured
using NIH-Image v1.60 software. Non-specific background
fluorescence was determined in two ways with similar results: (1)
Regions of the images with no cells, having areas identical to each
individual cell, were measured and the intensity was subtracted from
that for each individual fluorescent cell. (2) Cells were treated in
methanol + 3% TCA which denatures the profilin in the cells,
destroying the native epitopes for antibodies P4 and P5. The
fluorescence of cells stained with the Cy3 antibodies was then
measured and subtracted from the specific staining intensities
obtained on the glutaraldehyde-fixed cells where the native structure
of profilin was preserved.
RESULTS
Profilin distribution during biochemical fractionation
To determine whether profilin is soluble or bound tightly to
large cellular components, we homogenized amoebas in an
equal volume of buffer, centrifuged and assayed for profilin by
immunoblotting with several different antibodies (Table 1). On
average, the concentrations of profilin in both the low speed
and high speed supernatants are the same as the whole
homogenate. Since 0.1-1.0% of the cells do not lyse, the small
amount of profilin that pellets through a sucrose cushion could
be in unlysed cells. This confirms less extensive experiments
with polyclonal antibodies (Tseng et al., 1984). Using both the
low and high speed supernatants, competitive ELISA with four
different monoclonal antibodies that react with both profilin-I
and profilin-II give similar values for the cellular profilin
concentration with an average of 104 µM. Quantitative
immunoblotting yields lower values in the range of 50 to
70 µM. We conclude that the amoeba contains about 100 µM
profilin, essentially all of which is soluble.
To determine if this soluble profilin is free or bound to other
proteins, we fractionated low and high speed supernatants by
gel filtration and assayed for profilin on immunoblots of pooled
fractions (Fig. 1A) or for profilin, actin and Arp3 by ELISA of
individual fractions (Fig. 1B). Conveniently, this column
separates the three fractions of interest: large particles (such as
Arp2/3 complex and vesicles) in the void volume; monomeric
actin; and free profilin. About 60% of the total profilin elutes
in the actin fractions and the balance is free (Table 2). No
profilin or actin is detected in the void volume with the Arp2/3
complex. Despite the fact that dilution and mass action must
dissociate profilin from actin during this experiment, a majority
of profilin is bound to actin monomers and no detectable
profilin or actin is associated with larger particles.
Beads carrying profilin ligands provide an independent
assay for partners of profilin in high speed supernatants (Fig.
2). Poly-L-proline binds profilin without interfering with actin
binding (Tanaka and Shibata, 1985; Kaiser et al., 1989;
Perelroizen et al., 1994). Immobilized poly-L-proline adsorbs
all of the profilin in the extracts along with some actin (molar
ratio actin:profilin = 1:7), but no other tightly associated
proteins detected by Coomassie blue staining. Control
monoclonal antibody P3, which binds denatured but not native
profilin, adsorbs no profilin, actin or other proteins from the
extract. Antibody P5, which binds native profilin but not
profilin-actin complexes, adsorbs profilin alone from the
extract (Fig. 2, lane P5). Antibody P4, which binds both free
profilin and profilin-actin complexes, adsorbs both profilin and
actin (molar ratio actin:profilin = 1:2.5) from the extracts
(Fig. 2, lane P4). Antibody P4 recovers more actin associated
with profilin than poly-L-proline, perhaps because the antibody
stabilizes the profilin-actin complex and/or actin is more easily
rinsed off the poly-L-proline beads. The total profilin adsorbed
by antibody P4 beads is equal to the free profilin adsorbed by
the antibody P5 beads plus a 1:1 ratio of profilin:actin. This
suggests that the only major ligand for profilin in cell extracts
Table 1. Recovery of profilin in cell fractions
Experiment
1
2
3
4
5
6
7
8
9
Mean
Lysis
temperature
Antibody used
for immunoblot
Low speed
supernatant
Low speed
pellet
0°C
0°C
0°C
0°C
0°C
0°C
22°C
0°C
0°C
P8 (PII)
P9 (PII)
P9 (PII)
P1-P7 (PI+PII)
P1-P7 (PI+PII)
P1-P7 (PI+PII)
P2 (PI+PII)
P2 (PI+PII)
P5 (PI+PII)
102%
100%
135%
89%
86%
120%
10%
9%
105%
6%
0%
High speed
supernatant
123%
88%
136%
89%
85%
117%
106%
81%
79%
100%
High speed
pellet
0%
10%
5%
5%
Cells were homogenized on ice or at room temperature in one volume of extraction buffer and centrifuged for 5 minutes at 16,000 g to yield a low speed
supernatant or for 30 minutes at 200,000 g to yield a high speed supernatant. Profilin was measured in the whole extract, supernatants and (in three experiments)
pellets after gel electrophoresis and quantitative immunoblotting with the monoclonal antibodies indicated. The values given are relative to profilin in the whole
homogenate.
Profilin association with monomeric actin in Acanthamoeba 3783
Table 2. Distribution of profilin in gel filtration
experiments
Experiment
Void
fraction
Actin monomer Free profilin
fraction
fraction
Immunoblots of pooled fractions
1. LSS (PI+PII)
2. HSS (PI+PII)
3. HSS (PI+PII)
4. LSS (PI+PII)
5. HSS (PI+PII)
Mean
0%
0%
0%
0%
0%
0%
55%
60%
63%
59%
68%
61%
45%
40%
37%
41%
32%
39%
ELISA of individual fractions
1. HSS, P8 (PII)
2. LSS, P1-P7 (PI+PII)
3. HSS, P1-P7 (PI+PII)
Mean
0%
0%
0%
0%
49%
54%
58%
54%
51%
46%
42%
46%
Low (LSS) or high speed supernatants (HSS) were fractionated by gel
filtration on Sephadex G-75 (Fig. 1) and assayed for profilin.
actin. Staining fixed cells with these antibodies supports this
conclusion (see Fig. 7, below).
Fig. 1. Distribution of profilin during fractionation of low-speed
supernatant (LSS) and high-speed supernatant (HSS) by gel filtration
chromatography. Bars labeled Void, Actin and Profilin show the
positions where the peaks of blue dextran, monomeric actin and
monomeric profilin run when chromatographed separately on the
column. Arrows indicate the salt volume. Conditions: 1.5 × 25 cm
column of Sephadex G-75 (medium) equilibrated with 2 mM
imidazole (pH 7.0), 1.0 mM EGTA, 0.2 mM ATP, 0.5 mM DTT,
0.2 mM MgCl2, 20 mM KCl, 1.0 mM NaN3. Fractions: 0.5 ml.
(A) Assay by immunoblots. Sample: 0.5 ml LSS (open circles) or
HSS (filled circles). Assays: Protein concentrations were estimated
with the Bradford assay (absorbance at 595 nm). Equivalent volumes
of fractions corresponding to the column sample (S), void volume
(V), monomeric actin fractions (A) and free profilin fractions (P)
were pooled and analyzed by immunoblotting with a pooled mixture
of monoclonal antibodies P1-P7 that detect both profilin-I and
profilin-II. In this experiment, 59% of the profilin in LSS and 68% of
the profilin in HSS chromatographed with the monomeric actin.
(B) Assay by ELISA. Sample: 0.5 ml HSS. Absorbance at 290 nm
(filled squares). For ELISA, a sample of each fraction was diluted
1:50 in 10 mM Tris and 50 µl were adsorbed to microtiter wells.
Antibody binding to column fractions is indicated by absorbance at
490 nm for anti-Arp3 (open squares), anti-profilin (open triangles)
and anti-actin (filled triangles).
is actin in the monomeric state. The yield of actin-profilin
complex obtained by both antibody P4 and poly-L-proline
beads is higher in concentrated than dilute extracts, as expected
from mass action. These assays show that as much as twothirds of the profilin in cell extracts is bound to monomeric
Effects of profilin and divalent cations on elongation
rates of actin
To determine the effect of physiological concentrations of
amoeba profilin on amoeba actin polymerization, we tested a
much wider range of profilin concentrations than previously
(Kaiser et al., 1986) on the elongation of Mg2+-ATP-actin from
actin filament seeds (Fig. 3). The elongation rate of 2.4 µM
amoeba actin declines with profilin concentration to about 6575% of control values and plateaus above 50 µM profilin. In
contrast, 75 µM profilin completely inhibits elongation of
Ca2+-ATP-actin. In similar experiments, Gutsche-Perelroizen
et al. (1999) found that much lower concentrations of bovine
profilin inhibited elongation of rabbit skeletal muscle actin
from spectrin-actin seeds using turbidity measurements: only
~30 µM profilin completely inhibited Ca2+-actin and only ~5
µM profilin inhibited Mg2+-actin to 60-70%, about the same
plateau we observe (Fig. 3).
Characterization of rhodamine-labeled profilin-II
We used profilin-II for experiments with live cells, since it
binds actin (Vinson et al., 1998) and poly-L-proline (Petrella
et al., 1996) equal to profilin-I and binds PIP2 better than
profilin-I (Machesky et al., 1990). Vinson et al. (1998)
document that rhodamine-S38C profilin-II and rhodamineN58C profilin-II bind actin and poly-L-proline. In small zone
gel filtration experiments Rho-N58C-PII binds PIP2 similar to
native profilin-II: 39-60% of native or recombinant profilin-II
and 50-70% of Rho-N58C-PII run in the void and trailing
fractions of G-100 columns ahead of free profilin.
Characterization of amoebas syringe-loaded with
fluorescent profilins
Syringe-loading lyses some cells, but fluorescent profilin fills
the cytoplasm of 10-40% of the surviving cells (Figs 4, 5).
Loaded cells are initially sluggish, but normal locomotion,
endocytosis and contractile vacuole function resume in all but
the most intensely labeled cells during 2 to 4 hours of
incubation in Neff’s medium. The presence of fluorescent
profilin in the cytoplasm facilitates observation of both
3784 D. A. Kaiser and others
Fig. 2. Adsorption assays for profilin and actinprofilin complexes in high-speed supernatants
of Acanthamoeba homogenates using beads
with immobilized poly-L-proline (PP) or
monoclonal antibodies P3, P4 or P5. Bound
proteins were eluted with 6 M urea, separated
by SDS-PAGE and stained with Coomassie
blue. Bands of actin and profilin (Pro) are
labeled.
endocytosis and exocytosis. Before and
after loading, contractile vacuoles cycle at
regular intervals, with a cycle to cycle
variation of less than 20%. In water the
average cycle times are 66 seconds for
unloaded cells (n=10) and 93 seconds for
loaded cells (n=4). In control and loaded
cells plasma membrane specializations for
macropinocytosis (called amoebastomes or crowns) form at
approximately the same frequency as contractile vacuole
contraction (Fig. 4A-C). Loaded cells survive for several days.
Although not observed, loaded cells may divide, since pairs of
fluorescent cells are common among many unloaded cells.
Distribution of fluorescent profilin in live amoebas
Rho-N58C-PII fluorescence fills the cytoplasm and is excluded
from organelles, including the nucleus, mitochondria, vacuoles
and large vesicles (Figs 4, 5). The cytoplasmic fluorescence
between these large organelles is generally homogeneous, but
Fig. 3. Effect of profilin on the initial rate of elongation of Ca2+ and
Mg2+ ATP-actin monomers. 2.4 µM Acanthamoeba actin monomers
(5% pyrenyl labeled) was polymerized from 5.0 µM actin filament
seeds in the presence of 0-380 µM Acanthamoeba profilin-I at 22°C.
Open circles, initial rates of elongation for Ca2+ actin monomers
polymerized from Ca2+ actin filament seeds in 1.0 mM CaCl2,
50 mM KCl, 10 mM imidazole, pH 7.0, 0.18 mM ATP, 0.45 mM
DTT and 0.8 mM NaN3. Filled circles, initial rates of elongation for
Mg2+ actin monomers polymerized from Mg2+ actin filament seeds
in 1.0 mM EGTA, 1.0 mM MgCl2, 50 mM KCl, 10 mM imidazole,
pH 7.0, 0.18 mM ATP, 0.45 mM DTT and 0.8 mM NaN3.
is finely granular in some areas (Fig. 4D). These small spots
of high fluorescence contrast are transient on a time scale of a
few seconds. The microscopic appearance and behavior of
Fig. 4. Time series (minute:seconds) of
cooled CCD fluorescence micrographs of
live amoebas syringe-loaded with
Rho-N58C-PII (A-D) or stained with the
membrane-binding vital dye FM4-64 (E).
(A,B) An amoeba loaded with
Rho-N58C-PII showing two cycles of
macropinocytosis (large arrows) and
contraction of the contractile vacuole
(small arrows). Lamellapodia extend first
towards the upper right and later to the
lower right (arrowheads are stationary).
(C) An amoeba with Rho-N58C-PII
fluorescence concentrating transiently near
the plasma membrane at a site of
macropinocytosis (dark arrows).
(D) Details of the amoeba seen in A and B
showing punctate accumulations of
fluorescent profilin and the low intensity of
fluorescence near the plasma membrane,
which is outlined (dotted lines) from
enhanced images. (E) A live amoeba
treated with the membrane-binding vital
dye FM4-64 which stains the plasma
membrane, contractile vacuole and other
organelles. Large membrane-bounded
organelles are excluded from the broad,
thin, lamellapodium as the cell extends a
pseudopod toward the upper left. Nu =
nucleus, CV = contractile vacuole. Quicktime movies of these cells are available at
http://perutz.salk.edu/picture_gallery/.
Profilin association with monomeric actin in Acanthamoeba 3785
Fig. 5. Cooled CCD fluorescence
micrographs of a time series
(minutes:seconds) taken during fixation
and permeabilization of an amoeba
syringe-loaded with Rho-N58C-PII and
rapidly migrating in Neff’s medium in a
flow chamber. At time zero 1%
glutaraldehyde began to flow through
the perfusion system, reaching the field
of view after 1 minute. Some cell
movements persist for 50 seconds
(arrow is stationary). After 19 minutes
in glutaraldehyde, strong
autofluorescence, obvious even in an
unlabeled cell seen below the labeled
cell at 20:00, 25:00, phase, obscures the
specific fluorescence. At 20:00, a
solution of 0.2% Triton X100 was
perfused through the flow chamber for 5 minutes. A Quick-time movie of this experiment is available at http://perutz.salk.edu/picture_gallery/.
these fluorescent spots does not distinguish whether they are
particles or cytosolic compartments between protein-excluding
organelles.
Rho-N58C-PII fluorescence is strong in the lamellapodia
of migrating cells (Figs 4A-B, 5), particularly considering the
short path length through relatively flat lamellapodia
compared with other parts of the cell. The profilin
fluorescence is deeper in the cortex than the plasma
membrane, which we stained with the vital fluorescent dye
FM4-64 (Fig. 4E). The zone with concentrated profilin has
few membrane-bounded organelles, which were also labeled
with the vital dye. We could not compare profilin and actin
directly in live cells because fluorescent actin and phalloidin
are toxic to the amoebas. In fixed cells, actin filaments
concentrate closer to the plasma membrane than profilin (Fig.
6B). Filopodia appear to be much richer in actin filaments
than profilin (Fig. 6).
Rho-N58C-PII
concentrates
transiently
around
amoebastomes during and immediately after internalization of
macropinocytic vesicles (Fig. 4A,C). In fixed cells, both
amoebastomes and macropinocytic vesicles stain brightly with
rhodamine-phalloidin (Fig. 6B). The long path length through
these large, three-dimensional cell surface specializations may
contribute to the higher intensity of profilin fluorescence in live
cells. On the other hand, fluorescent profilin does not
concentrate around contractile vacuoles either before or after
contraction (Fig. 4A,B) despite their large size (path length).
The high concentration of profilin-excluding membrane
vesicles around contractile vacuoles (Fig. 4E) might offset the
long path length.
Quantitation of free profilin and profilin bound to
actin by fluorescent antibody staining of fixed cells
Two monoclonal antibodies with native epitopes allowed us
to distinguish free profilin from profilin bound to actin in
fixed cells. Antibody P4 (Fig. 7), which binds free profilin
and profilin bound to actin (Fig. 2 and Kaiser and Pollard,
1996), stains glutaraldehyde-fixed cells much more intensely
than antibody P5 (Fig. 7), which binds free
profilin but not profilin-actin complexes (Fig. 2
and Kaiser and Pollard, 1996). In one experiment
(Fig. 7C, Expt. 1), the mean fluorescence
intensity, after correction for background, of
cells stained with P4 (profilin-actin + free
profilin) was 3.9 times that of cells stained with
P5 (free profilin). Thus 74% of the profilin in
these cells is bound to actin. In a separate
Fig. 6. Localization of total profilin and actin filaments
in glutaraldehyde-fixed amoebas. (A) Confocal
fluorescence (left) and phase-contrast (right)
micrographs of amoebas stained with 10 µg/ml Cy3labeled antibody P4 which binds profilin-actin
complexes as well as free profilin. The confocal image
is a superimposition of ten 1.0 µm confocal sections.
(B) Cooled CCD fluorescence (left) and DIC (right)
micrographs of amoebas stained with rhodaminephalloidin demonstrating localization of actin filaments
to an amoebastome (arrowheads), a newly-formed
endocytic vesicle (arrows), the cell cortex and
filopodia.
3786 D. A. Kaiser and others
Fig. 7. Discrimination of free profilin and total profilin
by monoclonal antibody staining of amoebas fixed with
glutaraldehyde. (A) Phase-contrast images.
(B) Conventional fluorescence micrographs. (Left) Cell
staining with Cy3 monoclonal antibody P5, which binds
only free profilin. (Right) Cell staining with Cy3
monoclonal antibody P4, which binds both free profilin
and profilin-actin complexes. Antibodies P4 and P5 have
similar affinities for profilin, are labeled with the same
number of Cy3 dyes, used at the same concentrations
(10 µg/ml) and photographed identically. Bar, 10 µm.
(C) Histograms showing the intensity distribution of
cells stained with P4 (open bars) and P5 (filled bars) for
two separate experiments: Experiment 1. The mean
fluorescence intensity of 37 cells stained with antibody
P4 is 3.9 times the mean fluorescence intensity of 39
cells stained with antibody P5. Experiment 2. The mean
fluorescent intensity of 16 cells stained with P4 is 8.9
times the mean fluorescence of 16 cells stained with P5.
experiment (Fig. 7C, Expt. 2), the P4-stained cells had 8.9
times the mean fluorescence intensity of the P5-stained cells,
so 89% of the profilin in these cells is associated with actin.
In control experiments, we used ELISA to test for possible
deleterious effects of glutaraldehyde treatment of profilin and
Cy3 labeling of the antibodies P4 and P5 on their binding to
profilin. P4 and P5 bind glutaraldehyde-treated profilin
slightly less well than native profilin, but binding was at least
75% the native profilin value. Cy3-labeling of P4 and P5
reduced their maximum binding to profilin but did not affect
the affinities (data not shown).
Since P4 and P5 have similar affinities for profilin, were
labeled to the same level with Cy3, were used at the same
concentrations and photographed identically, we conclude that
in glutaraldehyde-fixed cells 74-89% of the profilin is bound
to actin and the rest is not. Interpretation of this experiment
depends upon retention of profilin in its native conformation
during fixation and permeabilization, so we carried out
extensive controls to validate this assumption.
To choose methods for antibody staining, we fixed and
permeabilized cells loaded with fluorescent profilin (Fig. 5).
Cells continue to move for about 50 seconds in 1%
glutaraldehyde, but neither the distribution nor the intensity of
fluorescence changes appreciably (Fig. 5). Long term exposure
to glutaraldehyde produces strong autofluorescence in both
labeled and unlabeled cells and obscures the fluorescence in
the labeled cells (Fig. 5, 20 minutes). Treatment with 0.1%
sodium borohydride eliminates virtually all autofluorescence
(Figs 6A, 7). Permeabilization of cells with 0.2% Triton X-100
has no effect on the fluorescence (Fig. 5, 25 minutes) or
morphology of fixed cells (Fig. 5, phase). These results show
that little profilin is extracted from the amoebas
during fixation with glutaraldehyde and
permeabilization with Triton X-100.
Judging from phase contrast and DIC
microscopy, glutaraldehyde fixation preserves the
size, shape, pseudopods, organelles and general
appearance of live amoebas better than several other
fixatives including formaldehyde, formaldehydemethanol, methanol and methanol-TCA. As in live
cells,
the
anti-profilin
fluorescence
in
glutaraldehyde-fixed cells is spread throughout the
cytoplasm including lamellapodia, but is excluded from
membrane-bounded organelles, including the nucleus
(Figs 6A, 7). In contrast to live cells, some fixed cells have a
zone of low intensity anti-profilin staining between the cortex
and central part of the cell. Confocal micrographs (Fig. 6A)
confirm most of the impressions from conventional
fluorescence micrographs (Fig. 7), but emphasize the zone of
low fluorescence in the deep cortex.
Control experiments with both purified proteins and whole
cells show that glutaraldehyde fixation preserves the native
structure of profilin, while other fixatives tested do not. First,
when assayed by ELISA, glutaraldehyde-treated profilin-I or
profilin-II bind monoclonal antibodies to native epitopes (P4,
P5, P6 and P7). None of these antibodies to native epitopes
bound profilin treated with denaturing fixatives, such as
methanol containing 3% TCA. In contrast, antibodies with
denatured epitopes (P1, P2, P8, P9) bind best to profilins
treated with methanol containing 3% TCA. Second, these
conformation-sensitive antibodies allowed us to distinguish
between native and denatured profilin in fixed cells.
Antibodies with native epitopes stain cells fixed with
glutaraldehyde, but not cells fixed with methanol and 3%
TCA. Antibodies with denatured epitopes stain profilin in
cells treated with methanol and 3% TCA much more strongly
than cells fixed with glutaraldehyde. We assume that fixation
with methanol is responsible for profilin redistribution to
patches near the plasma membrane as observed in previous
micrographs from our laboratory (Machesky et al., 1994).
Thus fixatives containing methanol denature profilin and
redistribute profilin compared with live cells (Figs 4, 5) or
cells fixed in glutaraldehyde (Figs 6A, 7).
Profilin association with monomeric actin in Acanthamoeba 3787
through the poly-L-proline binding site, but these partners are
either sparse or bind weakly. Our results support the hypothesis
that the major function of profilin in Acanthamoeba is to
maintain a pool of actin monomers inhibited from spontaneous
self-nucleation or pointed end elongation but ready to interact
with Arp2/3 complex and activating proteins such as WASp or
Scar to form new actin filaments with free barbed ends
(Machesky et al., 1999) and to add explosively to these barbed
ends. In this way, profilin along with mechanisms controlling
the generation of barbed ends provides the cell with an
effective on/off switch for controlling actin polymerization.
Fig. 8. Localization of profilin and filamentous actin in human A431
cells. Human A431 cells were fixed in 4% formaldehyde,
permeabilized with 0.2% triton X100 and double-stained for profilin
and filamentous actin. (A) Staining with a Cy3-labeled affinity
purified polyclonal antibody to human platelet profilin-I. (A,
Inset) Anti-profilin immunoblot of A431 cells separated by 12.5%
SDS-PAGE, reacted with affinity-purified anti-profilin antibody and
detected by chemiluminescence. (B) Staining with bodipy
phallacidin to visualize actin filaments. (C) DIC image of cells
stained in A and B.
Localization of profilin and filamentous actin in
human A431 cells
We compared profilin localization in amoebas with
formaldehyde-fixed human A431 cells stained with an affinity
purified Cy3 labeled antibody (Fig. 8). Similar to fixed
(Figs 6B, 7) and live (Figs 4, 5) amoebas, profilin in A431 cells
distributes throughout the cytoplasm, is excluded from the
nucleus and concentrates at the leading edges of lamellapodia,
the only region where both profilin and filamentous actin
appear together in high concentrations (Fig. 8).
DISCUSSION
We used biochemical fractionation and localization
experiments with live and fixed cells to address two questions:
Which molecular partners bind profilin and where is profilin
located in live cells? Consistent results from a variety of
experiments show that most profilin in Acanthamoeba is bound
to actin monomers and spread throughout the cytoplasm. Some
profilin may be bound secondarily to other ligands, possibly
Cytoplasmic pools of profilin
Virtually all profilin-I and profilin-II is soluble after
homogenization of amoebas and most profilin is bound to actin
monomers. Little is bound tightly to membranes or large
particles. Even after gel filtration, which dilutes the sample and
separates free profilin from actin-profilin complexes (both
promoting dissociation by mass action), most profilin elutes in
the actin monomer fraction and none of the profilin or profilinactin complexes are associated with larger particles. Therefore
no profilin remains bound to molecules larger than actin and
no profilin-actin is bound strongly to a third partner. Some
potential partners like Arp2/3 complex are missed due to low
abundance (2 µM) and affinity (Kd = 7 µM) compared with
actin (about 100 µM, Kd = 0.1 µM). This does not discount the
physiological importance of potential partners such as Arp2/3
complex or FH-domain proteins, but indicates that their
concentrations or affinities for profilin are much lower than
those of monomeric actin. Immunoadsorption and staining
fixed cells with antibodies selective for free profilin and
profilin-actin complexes confirms the conclusion that a
minimum of about two-thirds of the cellular profilin is bound
to actin monomers. This accounts for an unpolymerized actin
pool of at least 60 µM (and possibly as high as 90 µM) in
Acanthamoeba.
Direct adsorption of profilin from cellular extracts with polyL-proline or antibodies immobilized on beads is useful for
confirming the abundance of profilin-actin complexes, but not
for ruling out other ligands that bind the polyproline site on
profilin. Bound poly-L-proline precludes binding other ligands
this site, and all of our antibodies that bind to native profilin
(P4, P5, P6 and P7) also interfere with poly-L-proline binding
(Kaiser and Pollard, 1996).
Localization of profilin in live and fixed cells
Experiments with live cells employed profilin-II labeled on a
single site where rhodamine does not interfere with binding to
known ligands. In contrast, labeling lysines, the method used
for plant profilin (Vidali and Hepler, 1997), compromised actin
and PIP2 binding by amoeba profilin. We used profilin-II,
because it binds well to all three classes of ligands: actin/Arp2,
poly-L-proline and PIP2. (It is not known if amoeba contain
annexin or if amoeba profilin binds annexin.) We calculate
from the results of Doberstein et al. (1993) that syringe loading
increased the cytoplasmic concentration of profilin less than
3%, not enough to alter the physiology. Syringe loading
allowed us to study numerous cells, which recovered function
better than microinjected cells (Sinard and Pollard, 1989).
Except for a 50% increase in the cycle time for contractile
vacuoles, loaded cells appeared normal.
3788 D. A. Kaiser and others
As expected from the biochemical results, the fluorescence
of the labeled profilin distributes relatively uniformly
throughout the cytoplasm of the amoeba. Fluorescent profilin
behaved similarly in live vertebrate cells (Tarachandani and
Wang, 1996) and plant cells (Vidali and Hepler, 1997). With
minor exceptions noted below, localization of profilin with
fluorescent antibodies in optimally fixed cells was the same as
fluorescent profilin in live cells.
Our observations on live and fixed amoebas differ from a
recent paper of Bubb et al. (1998), who reported that antibodies
to profilin-II concentrated on the plasma membrane of fixed
amoebas along with polyphosphoinositides. Neither our
fluorescent profilin-II nor our fluorescent profilin antibodies
associated strongly with membranes relative to their bright
cytoplasmic fluorescence or with the labeling of membrane
lipids with FM4-64. It is possible that labeling profilin
compromised lipid interactions required for membrane
binding, although the labeled profilin bound to PIP2 micelles
in vitro. Similarly, our monoclonal antibodies to profilin-II may
not recognize their epitope when profilin-II binds membranes.
Alternatively, profilin can concentrate unnaturally in the cell
cortex depending on the method of fixation. Our observations
do not rule out profilin-II being associated with membranes or
participating in the metabolism of polyphosphoinositides
(Goldschmidt-Clermont et al., 1991), but show that the
concentration of profilin on membranes does not greatly
exceed that in the cytoplasm in Acanthamoeba. Similarly,
fluorescent profilin did not appear to associate with membranes
in the live cell experiments of Tarachandani and Wang (1996).
Our observations of profilin in live and fixed amoebas and
fixed A431 cells differ in some aspects from reports on profilin
localization in fixed vertebrate cells. Buss et al. (1992) and
Mayboroda et al. (1997) reported profilin colocalized with
actin filaments in vertebrate cells, while amoeba profilin and
actin filaments are concentrated in different locations. In
contrast to our earlier work with polyclonal antibodies (Tseng
et al., 1984) and the observation of Mayboroda et al. (1997) on
vertebrate cells, we did not find any Rho N58C-PII in the
nucleus of live amoebas or any profilin in the nucleus of fixed
amoebas. As in neurons (Faivre-Sarrailh et al., 1993), amoeba
filopodia do not stain strongly with anti-profilin antibodies.
While loss of profilin and/or redistribution of this small soluble
protein during fixation of cells might be responsible for some
of these variable results, little profilin is lost during
glutaraldehyde fixation of amoebas (Fig. 5) or other cells
(Rothkegel et al., 1996; Schluter et al., 1997).
Comparison with biochemical properties
The present work agrees with biophysical studies showing that
profilins have a high affinity for actin (Perelroizen et al., 1994;
Vinson et al., 1998), such that most of the profilin is expected
to be bound to unpolymerized actin at the concentrations in
cells. A high concentration of actin-profilin complex spread
throughout the cytoplasm prepares the cell for explosive
growth of actin filaments wherever uncapped filament barbed
ends appear in the cytoplasm. Although profilin bound to actin
strongly suppresses spontaneous nucleation and prevents
elongation at the slow growing pointed end of actin filaments
(Pollard and Cooper, 1986; Pring et al., 1992), profilin-actin
complexes elongate the barbed end of actin filaments nearly as
fast as actin monomers (Fig. 3; Pollard and Cooper, 1986;
Kaiser et al., 1986; Pring et al., 1992; Pantaloni and Carlier,
1993; Gutsche-Perelroizen et al., 1999). As far as we know,
amoebas lack thymosin, but in cells with a high concentration
of thymosin, profilin can also shuttle actin subunits from the
non-polymerizable thymosin-actin complex to the barbed end
of actin filaments (Pantaloni and Carlier, 1993). The
micromolar concentration of capping protein in cells
terminates this rapid growth at the barbed ends of actin
filaments in about 1.5 seconds (Schafer et al., 1996).
Given the potential for rapid but transient barbed end growth
from the actin-profilin pool, the cell needs only to control the
formation of uncapped barbed filament ends. One possibility
is the dissociation of capping proteins from barbed ends by
membrane polyphosphoinositides (Janmey, 1994; Schafer et
al., 1996), but Eddy et al. (1997) suggest that new filaments
form by de novo nucleation rather than uncapping of barbed
ends. The Arp2/3 complex is the best candidate for this
nucleation activity (Kelleher et al., 1995; Welch et al., 1997;
Svitkina et al., 1997; Mullins et al., 1998a; Ma et al., 1998;
Mullins and Pollard, 1999). New evidence shows that a family
of proteins called Scar or WASp activate nucleation by Arp2/3
complex (Machesky et al., 1999; Rohatgi et al., 1999; Winter
et al., 1999), that profilin improves the fidelity of this on/off
switch (Machesky et al., 1999) and that preexisting actin
filaments promote nucleation by Scar and Arp2/3 complex
(Machesky et al., 1999).
Other pools of profilin
In live cells, some of the profilin in cytoplasm appears to be
concentrated in particles (Fig. 4D; Tarachandani and Wang,
1996; Vidali and Hepler, 1997). These profilin-rich structures
appear to change in shape and position with time. The particles
in NRK cells are rich in monomeric actin, judging from
colocalization with microinjected vitamin D-binding protein
(Tarachandani and Wang, 1996). Particles of similar size are
stained by anti-profilin antibodies in various fixed cells (Buss
et al., 1992; Mayboroda et al., 1997). However, during low and
high speed centrifugation, only a small fraction of profilin
pellets. This pelleted profilin could be associated with
membranes or submicrometer particles, or accounted for by the
few unlysed cells. Since so little profilin pellets at high speed,
these apparent particles in cells might contain lipids to make
them buoyant. However, the absence of profilin in the void
volume of the gel filtration columns is evidence against a
floating particulate fraction of profilin. Although we cannot
rule out an interesting particulate form of profilin, the particles
may simply be regions of cytoplasm surrounded by profilinexcluding membrane compartments.
Profilin concentrates in lamellapodia to a modest extent in
both live and fixed amoebas as well as live (Tarachandani and
Wang, 1996) and fixed (Buss et al., 1992; Mayboroda et al.,
1997) cultured vertebrate cells. In live motile amoebas the band
of fluorescent profilin (Figs 4, 5) often concentrates at the base
of the lamellapod (Fig. 4B), perhaps due to binding to prolinerich ligand(s). Because profilin can bind actin and poly-Lproline simultaneously, profilin immobilized by such ligands
may also carry actin monomers.
Several proteins in the endocytic pathway have been
identified as ligands for mouse brain profilins (Witke et al.,
1998). Clathrin appears to bind profilin-I (acidic isoform) and
dynamin-I appears to bind profilin-II (basic isoform) (Witke et
Profilin association with monomeric actin in Acanthamoeba 3789
al., 1998). Since fluorescent profilin-II accumulates around
some amoebastomes, the predominant endocytic structures in
amoebas (Fig. 4A,C), profilin may associate with similar
proteins in Acanthamoeba.
This work was supported by NIH research grants GM-26338 to
T.D.P. and GM-33171 to D.B.M. We are grateful to James Lessard for
providing anti-actin monoclonal antibody c4d6, Mas Sato and Sachiko
Karaki for help in preparing anti-profilin monoclonal antibodies,
Laurent Blanchoin for help with actin polymerization, Stephen M.
Mattessich for help with digital microscopy, Rodrigo I. Bustos,
Michael J. Delannoy and Pam Maupin for help with microscopy,
Susan D. Michaelis for providing FM4-64, Michael E. Ostap and
Enrique De La Cruz for helpful suggestions and Donna B. Kaiser for
proofreading the manuscript.
REFERENCES
Almo, S. C., Pollard, T. D., Way, M. and Lattman, E. E. (1994). Purification,
characterization and crystallization of Acanthamoeba profilin expressed in
Escherichia coli. J. Mol. Biol. 236, 950-952.
Alvarez-Martinez, M. T., Mani, J. C., Porte, F., Faivre-Sarrailh, C.,
Liautard, J. P. and Sri Widada, J. (1996). Characterization of the
interaction between annexin I and profilin. Eur. J. Biochem. 238, 777-784.
Archer, S. J., Vinson, V. K., Pollard, T. D. and Torchia, D. A. (1994).
Elucidation of the poly-L-proline binding site in Acanthamoeba profilin-I
by NMR spectroscopy. FEBS Lett. 337, 145-151.
Balasubramanian, M. K., Hirani, B. R., Burke, J. D. and Gould, K. L.
(1994). The Schizosaccharomyces pombe cdc3+ gene encodes a profilin
essential for cytokinesis. J. Cell Biol. 125, 1289-1301.
Bray, D. and Thomas, C. (1976). Unpolymerized actin in fibroblasts and
brain. J. Mol. Biol. 105, 527-544.
Bubb, M. R., Baines, I. C. and Korn, E. D. (1998). Localization of
actobindin, profilin I, profilin II and phosphatidylinositol-4, 5-bisphosphate
(PIP2) in Acanthamoeba castellanii. Cell Motil. Cytoskel. 39, 134-146.
Buss, F., Temm-Grove, C., Henning, S. and Jockusch, B. (1992).
Distribution of profilin in fibroblasts correlates with the presence of highly
dynamic actin filaments. Cell Motil. Cytoskel. 22, 51-61.
Carlsson, L., Nystrom, L. E., Sundkvisk, I., Markey, F. and Lindberg, U.
(1977). Actin polymerizability is influenced by profilin, a low molecular
weight protein in non-muscle cells. J. Mol. Biol. 115, 465-483.
Castrillon, D. and Wasserman, D. (1994). diaphanous is required for
cytokinesis in Drosophila and shares domains of similarity with the products
of the limb deformity gene. Development. 120, 3367-3377.
Chang, F., Drubin, D. and Nurse, P. (1997). cdc12p, a protein required for
cytokinesis in fission yeast, is a component of the cell division ring and
interacts with profilin. J. Cell Biol. 137, 169-182.
Clarke, M. S. and McNeil, P. L. (1992). Syringe loading introduces
macromolecules into living mammalian cell cytosol. J. Cell Sci. 102, 533541.
Cooley, L., Verheyen, E. and Ayers, K. (1992). Chickadee encodes a profilin
required for intercellular cytoplasmic transport during Drosophila
oogenesis. Cell 69, 173-184.
Doberstein, S., Baines, I. C., Weigand, G., Korn, E. D. and Pollard, T. D.
(1993). Inhibition of contractile vacuole function in-vivo by antibodies
against myosin-I. Nature 365, 841-843.
Eads, J. C., Mahoney, N. M., Vorobiev, S., Bresnick, A. R., Wen, K. K.,
Rubenstein, P. A., Haarer, B. K. and Almo, S. C. (1998). Structure
determination and characterization of Saccharomyces cerevisea profilin.
Biochemistry 37, 11171-11181.
Eddy, R. J., Han, J. and Condeelis, J. S. (1997). Capping protein terminates
but does not initiate chemoattractant-induced actin assembly in
Dictyostelium. J. Cell Biol. 139, 1243-1253.
Edmatsu, M., Hirono, M. and Watanabe, Y. (1992). Tetrahymena profilin is
localized in the division furrow. J. Biochem. 112, 637-642.
Evangelista, M., Blundell, K., Longtine, M., Chow, C., Adames, N.,
Pringle, J., Peter, M. and Boone. C. (1997). Bni1p, a yeast formin linking
cdc42p and the actin cytoskeleton during polarized morphogenesis. Science
276, 118-122.
Faivre-Sarrailh, C., Lena, J. Y., Had, L., Vignes, M. and Lindberg, U.
(1993). Location of profilin at presynaptic sites in the cerebellar cortex;
implication for the regulation of the actin-polymerization state during axonal
elongation and synaptogenesis. J. Neurocytol. 22, 1060-1072.
Frazier, J. A. and Field, C. M. (1997). Actin cytoskeleton: are FH proteins
local organizers? Curr. Biol. 7, R414-R417.
Goldschmidt-Clermont, P. J., Kim, J. W., Machesky, L. M., Rhee, S. G.
and Pollard, T. D. (1991). Regulation of phospholipase C-γ1 by profilin
and tyrosine phosphorylation. Science 251, 1231-1233.
Gutsche-Perelroizen, I., Lepault, J., Ott, A. and Carlier, M. -F. (1999).
Filament assembly from profilin-actin. J. Biol. Chem. 274, 6234-6243.
Haarer, B. K., Petzold, A. S. and Brown, S. S. (1993). Mutational analysis
of yeast profilin. Mol. Cell. Biol. 13, 7864-7873.
Hajkova, L., Bjorkegren Sjogren, C., Korenbaum, E., Nordberg, P. and
Karlsson, R. (1997). Characterization of a mutant profilin with reduced
actin-binding capacity: effects in vitro and in vivo. Exp. Cell Res. 234, 6677.
Harris, S., Hamer, L., Sharpless, K. and Hamer, J. (1997). The Aspergillus
nidulans sepA gene encodes an FH1/2 protein involved in cytokinesis and
the maintenance of cellular polarity. EMBO J. 16, 3474-3483.
Hartwig, J. H., Chambers, K. A., Hopcia, K. L. and Kwiatkowski, D. J.
(1989). Association of profilin with filament-free regions of human
leukocyte and platelet membranes and reversible membrane binding during
platelet activation. J. Cell. Biol. 109, 1571-1579.
Haugwitz, M., Noegel, A. A., Karakesisoglou, J. and Schleicher, M. (1994).
Dictyostelium amebas that lack G-actin-sequestering profilins show defects
in F-actin content, cytokinesis and development. Cell 79, 303-314.
Janmey, P. A. (1994). Phosphoinositides and calcium as regulators of cellular
actin assembly and disassembly. Annu. Rev. Physiol. 56, 169-191.
Kaiser, D. A., Sato, M., Ebert, R. and Pollard, T. D. (1986). Purification and
characterization of two isoforms of Acanthamoeba profilin. J. Cell Biol. 102,
221-226.
Kaiser, D. A., Goldschmidt-Clermont, P. J., Levine, B. A. and Pollard, T.
D. (1989). Characterization of renatured profilin purified by urea elution
from poly-L-proline agarose columns. Cell Motil. Cytoskel. 14, 251-262.
Kaiser, D. A. and Pollard, T. D. (1996). Characterization of actin and polyL-proline binding sites of Acanthamoeba profilin with monoclonal
antibodies and by mutagenesis. J. Mol. Biol. 255, 89-107.
Kelleher, J. F., Atkinson, S. J. and Pollard, T. D. (1995). Sequences,
structural models and cellular localization of the actin-related proteins Arp2
and Arp3 from Acanthamoeba. J. Cell Biol. 131, 385-397.
Kiehart, D. P., Kaiser, D. A. and Pollard, T. D. (1984). Monoclonal
antibodies demonstrate limited structural homology between myosin
isozymes from Acanthamoeba. J. Cell Biol. 99, 1002-1014.
Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly
of the head of bacteriophage T4. Nature 227, 680-685.
Lassing, I. and Lindberg, U. (1985). Specific interaction between
phosphatidylinositol 4,5-bisphosphate and profilactin. Nature 314, 472-474.
Lessard, J. L. (1988). 2 monoclonal antibodies to actin – one muscle selective
and one generally reactive. Cell Motil. Cytoskel. 10, 349-362.
Ma, L., Rohatgi, R. and Kirschner, M. W. (1998), The Arp2/3 complex
mediates actin polymerization induced by the small GTP-binding protein
Cdc42. Proc. Nat. Acad. Sci. USA 95, 15362-15367.
Machesky, L. M., Goldschmidt-Clermont, P. J. and Pollard, T. D. (1990).
The affinities of human platelet and Acanthamoeba profilin isoforms for
polyphosphoinositides account for their relative abilities to inhibit
phospholipase C. Cell Reg. 1, 937-950.
Machesky, L. (1994). Regulation of cytoskeletal dynamics and
phosphatidylinositide signaling by profilin. Johns Hopkins University
School of Medicine. Doctoral dissertation.
Machesky, L. M., Atkinson, S. J., Ampe, C., Vandekerckhove, J. and
Pollard, T. D. (1994). Purification of a cortical complex containing two
unconventional actins from Acanthamoeba by affinity chromatography on
profilin agarose. J. Cell Biol. 127, 107-115.
Machesky, L. M., Mullins, R. D., Higgs, H. N., Kaiser, D. A., Blanchoin,
L., May, R., Hall, M. and Pollard, T. D. (1999). Scar, a WASp-related
protein, activates nucleation of actin filaments by the Arp2/3 complex. Proc.
Nat. Acad. Sci. USA 96, 3739-3744.
Magdolen, V., Oechsner, U., Muller, G. and Bandlow, W. (1988). The introncontaining gene for yeast profilin (PFY) encodes a vital function. Mol. Cell
Biol. 8, 5108-5115.
Mahoney, N. M., Janmey, P. A. and Almo, S. C. (1997). Structure of the
profilin-poly-L-proline complex involved in morphogenesis and cytoskeletal
regulation. Nat. Struct. Biol. 4, 953-960.
Mayboroda, O., Schlüter, K. and Jockusch, B. (1997). Differential
3790 D. A. Kaiser and others
colocalization of profilin with microfilaments in PtK2 cells. Cell Motil.
Cytoskel. 37, 166-177.
Metzler, W. J., Bell, A. J., Ernst, E., Lavoie, T. B. and Mueller, L. (1994).
Identification of the poly-L-proline-binding site on human profilin. J. Biol.
Chem. 269, 4620-4625.
Mullins, R. D., Heuser, J. A. and Pollard, T. D. (1998a). The interaction of
Arp2/3 complex with actin: nucleation, high-affinity pointed end capping
and formation of branching networks of filaments. Proc. Nat. Acad. Sci. USA
95, 6181-6186.
Mullins, R. D., Kelleher, J. F., Xu, J. and Pollard, T. D. (1998b). Arp2/3
complex from Acanthamoeba binds profilin and crosslinks actin filaments.
Mol. Biol. Cell 9, 841-852.
Mullins, R. D. and Pollard, T. D. (1999). Structure and function of the Arp2/3
complex. Curr. Opin. Struct. Biol. 9, 244-249.
Pantaloni, D. and Carlier, M. -F. (1993). How profilin promotes actin
filament assembly in the presence of thymosin β4. Cell 75, 1007-1014.
Perelroizen, I., Marchand, J. B., Blanchoin, L., Didry, D. and Carlier,
M.-F. (1994). Interaction of profilin with G-actin and poly(L-proline).
Biochemistry 33, 8472-8478.
Petrella, E. C., Machesky, L. M., Kaiser, D. A. and Pollard, T. D. (1996).
Structural requirements and thermodynamics of the interaction of proline
peptides with profilin. Biochemistry 35, 16535-16543.
Pollard, T. D. (1984). Polymerization of ADP-actin. J. Cell Biol. 99, 769-777.
Pollard, T. D. and Cooper, J. A. (1984). Quantitative analysis of the effect
of Acanthamoeba profilin on actin filament nucleation and elongation.
Biochemistry 23, 6631-6641.
Pollard, T. D. and Cooper, J. A. (1986). Actin and actin-binding proteins. A
critical evaluation of mechanisms and functions. Annu. Rev. Biochem. 55,
987-1035.
Pring, M., Weber, A. and Bubb, M. (1992). Profilin-actin complexes
directly elongate actin filaments at the barbed end. Biochemistry 31, 18271836.
Reichlin, M. (1980). Use of glutaraldehyde as a coupling agent for proteins
and peptides. Meth. Enzymol. 70, 159-165.
Reinhard, M., Giehl, K., Abel, K., Haffner, C., Jarchau, T., Hoppe, V.,
Jockusch, B. M. and Walter, U. (1995). The proline-rich focal adhesion
and microfilament protein VASP is a ligand for profilins. EMBO J. 14, 15831589.
Rohatgi, R., Ma, L., Miki, H., Lopez, M., Kirchhausen, T., Takenawa, T.
and Kirschner, M. W. (1999). The interaction between N-WASP and the
Arp2/3 complex links Cdc42-dependent signals to actin assembly. Cell 97,
221-231.
Rothkegel, M., Mayboroda, O., Rohde, M., Wucherpfennig, C., Valenta,
R. and Jockusch, B. M. (1996). Plant and animal profilins are functionally
equivalent and stabilize microfilaments in living animal cells. J. Cell Sci.
109, 83-90.
Schafer, D. A., Jennings, P. B. and Cooper, J. A. (1996). Dynamics of
capping protein and actin assembly in vitro: uncapping barbed ends by
polyphosphoinositides. J. Cell Biol. 135, 169-179.
Schluter, K., Jockusch, B. M. and Rothkegel, M. (1997). Profilins as
regulators of actin dynamics. Biochim. Biophys. Acta 1359, 97-109.
Schutt, C., Myslik, J. C., Rozychi, M. D., Goonesekere, N. C. W. and
Lindberg, U. (1993). The structure of crystalline profilin-β-actin. Nature
365, 810-816.
Sinard, J. H. and Pollard, T. D. (1989). Microinjection into Acanthamoeba
castellanii of monoclonal antibodies to myosin-II slows but does not stop
cell locomotion. Cell Motil. Cytoskel. 12, 42-53.
Suetsugu, S., Miki, H. and Takenawa, T. (1998). The essential role of profilin
in the assembly of actin for microspike formation. EMBO J. 17, 6516-6526.
Svitkina, T. M., Verkhovsky, A. B., McQuade, K. M. and Borisy, G. G.
(1997). Analysis of the actin-myosin II system in fish epidermal keratocytes:
mechanism of cell body translocation. J. Cell Biol. 139, 397-415.
Tanaka, M. and Shibata, H. (1985). Poly (L-proline)-binding proteins from
chick embryos are a profilin and profilactin. Eur. J. Biochem. 151, 291-297.
Tarachandani, A. and Wang, Y. L. (1996). Site-directed mutagenesis enabled
preparation of a functional fluorescent analog of profilin: biochemical
characterization and localization in living cells. Cell Motil. Cytoskel. 34,
313-323.
Tseng, P. C. H., Runge, M. S., Cooper, J. A., Williams, Jr., R. C. and
Pollard, T. D. (1984). Physical, immunochemical and functional properties
of Acanthamoeba profilin. J. Cell Biol. 98, 214-221.
Vandekerckhove, J., Kaiser, D. A. and Pollard, T. D. (1989). Acanthamoeba
actin and profilin can be crosslinked between glutamic acid 364 of actin and
lysine 115 of profilin. J. Cell Biol. 109, 619-626.
Verheyen, E. M. and Cooley, L. (1994). Profilin mutations disrupt multiple
actin-dependent processes during Drosophila development. Development
120, 717-728.
Vidali, L. and Hepler, P. K. (1997). Characterization and localization of
profilin in pollen grains and tubes of Lilium longiflorum. Cell Motil.
Cytoskel. 36, 323-338.
Vinson, V. K., De La Cruz, E. M., Higgs, H. N. and Pollard, T. D. (1998).
Interactions of Acanthamoeba profilin with actin and nucleotides bound to
actin. Biochemistry 37, 10871-10880.
Watanabe, N., Madaule, P., Reid, T., Ishizaki, T., Watanabe, G., Kakizuka,
A., Saito, Y., Nakao, K., Jockusch, B. and Narumiya, S. (1997).
p140mDia, a mammalian homolog of Drosophila diaphanous, is a target
protein for Rho small GTPase and is a ligand for profilin. EMBO J. 16, 30443056.
Welch, M. D., DePace, A. H., Verma, S., Iwamatsu, A. and Mitchison,
T. J. (1997). The human Arp2/3 complex is composed of evolutionarily
conserved subunits and is localized to cellular regions of dynamic actin
filament assembly. J. Cell Biol. 138, 375-384.
Winter, D., Lechler, T. and Li, R. (1999). Activation of the yeast Arp2/3
complex by Bee1p, a WASP-family protein. Curr. Biol. 9, 501-504, S1.
Witke, W., Sharpe, A. H. and Kwiatkowski, D. J. (1993). Profilin deficient
mice are not viable. Mol. Biol. Cell 4, 149a.
Witke, W. A., Podtelejnikov, V., Di Nardo, A., Sutherland, J. D., Gurniak,
C. B., Dotti, C. and Mann, M. (1998). In mouse brain profilin I and profilin
II associate with regulators of the endocytic pathway and actin assembly.
EMBO J. 17, 967-976.
Yonemura, S. Y. and Pollard, T. D. (1992). Localization of myosin-I and
myosin-II in Acanthamoeba. J. Cell Sci. 102, 629-642.