AMOEBA PROTEUS - RU Press

CYTOPLASMIC FILAMENTS OF

AMOEBA PROTEUS

I. The Role of Filaments inConsistency Changes and Movement

THOMAS D. POLLARD and SUSUMU ITO

From the Department of Anatomy, Harvard Medical School, Boston, Massachusetts 02115.Dr. Pollard's present address is National Heart and Lung Institute, Section on CellularBiochemistry and Ultrastructure, National Institutes of Health, Bethesda, Maryland 20014

ABSTRACT

The role of filaments in consistency changes and movement in a motile cytoplasmic extractof Amoeba proteus was investigated by correlating light and electron microscopic observationswith viscosity measurements. The extract is prepared by the method of Thompson andWolpert (1963) . At 0 °C, this extract is nonmotile and similar in structure to ameba cyto-plasm, consisting of groundplasm, vesicles, mitochondria, and a few 160 A filaments . Theextract undergoes striking ATP-stimulated streaming when warmed to 22 °C. Two phasesof movement are distinguished . During the first phase, the apparent viscosity usuallyincreases and numerous 50-70 A filaments appear in samples of the extract prepared forelectron microscopy, suggesting that the increase in viscosity in caused, at least in part, bythe formation of these thin filaments. During this initial phase of ATP-stimulated movement,these thin filaments are not detectable by phase-contrast or polarization microscopy, butlater, in the second phase of movement, 70 A filaments aggregate to form birefringent micro-scopic fibrils . A preparation of pure groundplasm with no 160 A filaments or membranousorganelles exhibits little or no ATP-stimulated movement, but 50-70 A filaments form andaggregate into birefringent fibrils . This observation and the structural relationship of the70 A and the 160 A filaments in the motile extract suggest that both types of filaments maybe required for movement . These two types of filaments, 50-70 A and 160 A, are alsopresent in the cytoplasm of intact amebas . Fixed cells could not be used to study the distribu-tion of these filaments during natural ameboid movement because of difficulties in preservingthe normal structure of the ameba during preparation for electron microscopy .

INTRODUCTION

Although cytoplasmic streaming is a basic propertyof cells, the mechanism of the movement of sub-cellular particles is not clearly understood . Sincecytoplasm isolated from animal and plant cellsexhibits independent motility (Thompson andWolpert, 1963; Jarosch, 1956), the motive forcemust be generated in the cytoplasm . The structures

THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970,

thought most likely to produce movement are thetwo types of cytoplasmic filaments : (a) hollowfilamentous microtubules, and (b) solid filaments,which occur in several sizes . Microtubules serve acytoskeletal function in many cells and have beenimplicated in some types of motility, such as themovement of chromosomes in the mitotic spindle

pages 267-259

267

Dow

nloaded from http://rupress.org/jcb/article-pdf/46/2/267/1263396/267.pdf by guest on 07 April 2022

and of pigment granules in melanophores (Porter,1966) . As in the well studied contractile system ofmuscle, morphological evidence suggests that solidfilaments may be involved in cytoplasmic move-ments in many other cell types, including Physarum(Wohlfarth-Botterman, 1964), Nitella (Nagai andRebhun, 1966), ascidian epidermal cells (Cloney,1966), cultured fibroblasts (Buckley and Porter,1967), Difflugia (Wohlman and Allen, 1968) andsea urchin mesenchymal cells (Tilney and Gibbins,1969) . This view has been strengthened by twolines of evidence : (a) the isolation of actin fromPhysarum (Hatano and Oosawa, 1966), Acantha-moeba (Weihing and Korn, 1969), and sea urchineggs (Hatano, Kondo, and Miki-Noumura, 1969) ;and (b) the demonstration that muscle heavymeromyosin forms specific, ATP-dissociated, ar-rowhead-shaped complexes with thin filaments ina variety of chick embryo cells (Ishikawa, Bischoff,and Holtzer, 1969) and in Acanthamoeba (Pollard,Shelton, Weihing, and Korn, 1970) which arenearly identical with the complex between heavymeromyosin and actin filaments.

The closely related freshwater amebas, Amoebaproteus and Chaos carolinensis, have been used ex-tensively in studies of ameboid movement . Allenand his coworkers (Allen, 1961) determined thatthe axial endoplasm of Chaos has a lower con-sistency than the ectoplasm .' The higher consist-ency ectoplasm had been designated the GEL andthe lower consistency endoplasm the SOL . Thisterminology must be used cautiously, since Allen'sstudies have shown that the endoplasm is not astructureless SOL, but is a low consistency "gel ."Although light microscope studies have not directlydemonstrated gel structures in the ectoplasm, therestricted movement of ectoplasmic particles sug-gests that the ectoplasm is a two-phase system withthe organelles suspended in a loose gel network(Allen, 1961). Movement is dependent on gel

' The terms endoplasm and ectoplasm will be used asdefined by Allen (1961) . The endoplasm is the lowconsistency granular region in the central part of theameba. In moving monopodial specimens of Amoebaproteus, the endoplasm flows toward the advancing tipwhere it is everted in the fountain zone to form thehigher consistency, peripheral ectoplasm . The ecto-plasm is divided into granular ectoplasm and hyalineectoplasm . The hyaline ectoplasm, the clear area justinside the cell membrane, is sometimes referred to asthe ectoplasm, but this is confusing and will not beused. The endoplasm is separated from the ectoplasmby the shear zone .

268

THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970

structure (Allen, 1961), but no detailed studies onthe fine structure of the high and low consistencyregions of the cytoplasm have been reported .

Nachmias (1964) first reported fibrillar cyto-plasmic structures in electron micrographs ofChaos carolinensis . Long, thin filaments, about 75 Ain diameter, and thicker filaments, 150 A in di-ameter and up to 0 .5 p long, were seen in cellswhose plasma membranes were torn before fixationand in intact nonmotlie cells exposed to a pinocy-tosis-inducing solution .

In a related study, Simard-Duquesne andCouillard (1962) demonstrated that glycerinatedspecimens ofAmoeba proteus contract in the presenceof ATP and magnesium. Schäfer-Danneel (1967)examined the fine structure of such cells andshowed that a network of filaments 40-100 A indiameter associated with thicker filaments 160-220A in diameter becomes apparent during glycerinextraction . ATP-induced contraction of these ex-tracted amebas produced only a slight "condensa-tion" of the filament network . She suggested thatthe filaments play a role in the contraction of theglycerinated amebas and that filaments may alsobe important for the movement of the livingameba .

A major advance in our understanding of themechanism of ameboid movement was the dis-covery by Allen, Cooledge, and Hall (1960) thatparticles in cytoplasm freed from Chaos carolinensishad independent motility similar to the movementof particles in the intact cell . This approach wascarried further by Thompson and Wolpert (1963)who described a cytoplasmic fraction from homog-enized Amoeba proteus which exhibited vigorousmovement when ATP or ADP were added and itwas warmed from 4° to 20 °C. Electron microscopeexamination of aggregates formed at the end of thereaction revealed filaments which were 120 A indiameter and 5,000 A long . A more purified extractcontained filaments about 90 A in diameter(Wolpert, Thompson, and O'Neill, 1964) . Subse-quent reports (Morgan et al ., 1967) have shownmuch thinner filaments (20 A) in high-speed super-natant fractions of the extract . This high-speedsupernatant exhibited ATP-stimulated movementonly after the addition of the "vesicle fraction"which had previously been removed . Wolpert(1965) suggested that the thin filaments aggregateto form the thick filaments and that some interac-tion of these filaments results in contraction .

We have reexamined the motile cytoplasmic ex-

Dow


tract of Amoeba proteus described by Thompson andWolpert (1963) and have correlated light and elec-tron microscope observations with viscosity meas-urements to elucidate mechanisms of cytoplasmicconsistency changes and movement in this ameba .Our observations suggest that the "gelation" ofameba cytoplasm during movement is related tothe formation of labile 50-70 A filaments from pre-cursors in the groundplasm . We speculate thatcytoplasmic movement may depend on the interac-tion of these thin filaments with 160 A filamentsconstantly present in the groundplasm .

MATERIALS AND METHODS

Ameba Cultures

Amoeba proteus, strain PROT 1, was kindly suppliedby Dr. J . Griffin. The cells were grown in Prescottand Carrier medium (Prescott and Carrier, 1964)and fed Tetrahymena pyriformis according to themethod of Griffin (1960), modified for large scale(Griffin, personal communication) .

Preparation for Electron MicroscopySeveral thousand amebas were pipetted into a glass

vial, allowed to attach themselves to the surface andto undergo ameboid movement . A few milliliters offixative at 0 ° or 22 °C were gently added while thecells were observed with a phase-contrast or dissect-ing microscope. Fixatives used were based on theformaldehyde-glutaraldehyde fixative of Karnovsky(1965) modified according to Ito and Karnovsky(1968) by the addition of picric acid or related trini-trocompounds (trinitroresorcinol, trinitrocresol) atthe concentration of 0 .01%. 2% OS04 was added insome trials instead of the trinitrocompounds . Thefixatives were buffered with 0.1 M cacodylate, pH7.2. The fixatives were designated as follows : for-maldehyde-glutaraldehyde (FG), FG + picric acid(FGP), FG + trinitroresorcinol (FGR), FG +trinitrocresol (FGC), and FG + OS04 (FG-Os04) .After 10-90 min in one of the fixatives, the cells werewashed several times in 0.1 M cacodylate buffer, pH7 .2, and left in this solution overnight at 4 ° or 22 °Cbefore further fixation in 1 0 0 0304 buffered with0 .1 M cacodylate, pH 7 .2, at 22°C for 1 hr. Somesamples were treated with 0.5 or 3.0% uranylacetate in 0 .05 M Tris-maleate buffer pH 5.2 at22 °C for 1 hr after treatment with Os04 . Afterethanol dehydration, the cells were passed throughpropylene oxide and embedded in Araldite . Lightgold and silver sections were stained with saturatedor 3% aqueous uranyl acetate for 15 sec to 30 minfollowed by lead citrate (Venable and Coggeshall,1965) for 15 sec to 2 min.

Electron micrographs were taken on either anRCA EMU 3F or a Phillips 200 electron microscope .

Preparation of Cytoplasmic ExtractsThe motile fraction of ameba cytoplasm was pre-

pared by a modification of the method of Thompsonand Wolpert (1963) . Mass cultures of amebas werewashed in Chalkley's medium (Chalkley, 1930)containing 0 .5 MM MgCl2 and cooled to 4 °C for12-24 hr . The cooled cells were then concentratedby low-speed centrifugation, and 5 cc of the resultantslurry were centrifuged for 10 min at 18,000 rpm ina Type SW 39L Rotor (Spinco Division, BeckmanInstruments, Palo Alto, California) . This gave amaximum centrifugal force of 35,000 g at the tip ofthe tube and fragmented the cells into four distinctlayers similar to those described by Thompson andWolpert (1963) . The top clear layer was the sus-pending medium. The second layer was a viscidwhite material consisting of membrane-bounded bagsof cytoplasm. The third layer was brown and con-sisted of heavy fragments of the cells (food vacuoles,pieces of nuclei, mitochondria, and clusters of 150 Afilaments) . The fourth layer was white and consistedof crystals . The second layer (about 1-2 cc), con-sisting of membrane-bounded bags of cytoplasm, wastransferred to a teflon-glass homogenizer with anequal volume of glass-distilled water or Tris-maleatebuffer pH 7 .0 (10 or 20 mM) and homogenized withfive gentle strokes. Large fragments (mainly plasmamembrane) were removed by centrifugation at 1000g for 5 min . The supernatant was called Extract 1 .Further centrifugation of Extract 1 at 10,500 rpm ina Type 40 Rotor (Spinco) (maximum of 10,000 g)for 10 min removed clusters of 150 A filaments,mitochondria, rough endoplasmic reticulum, smoothvesicles, and plasma membrane fragments as a sub-stantial pellet . The supernatant was designatedExtract 2 . All steps in the preparative procedure werecarried out at 0-4 °C .

ATP (disodium salt-Sigma Chemical Co., St .Louis, Mo.) was freshly prepared for each experi-ment in glass-distilled water or 10 mm tris-maleatebuffer and brought to pH 7.0 with NaOH .

Light microscope observations and ciné-photo-micrographs were made with Zeiss phase-contrastand polarizing optics . Samples of extract were care-fully sealed under coverslips with petroleum jelly, sothat there was no movement of particles in controlsamples except random thermal (Brownian) motion .

Samples of the extract were prepared for electronmicroscopy by fixing with 3 0/0 glutaraldehyde in 0.2M cacodylate buffer, pH 7.0, FGC, or 1% 0304 in0 .1 M cacodylate buffer, pH 7 .2 for 2-15 hr at thetemperature of the sample (either 0 ° or 22 °C). Thefixed extract was concentrated into a pellet bycentrifugation at 140,000 g for 90 min. An alternate

T. D. POLLARD AND S. ITo Filaments of A . proteus . 1

269

Dow


method of fixation, which avoided concentrating theextract by centrifugation, was to solidify an enitresample of extract by adding a small amount of un-buffered 257 glutaraldehyde to a final concentrationof about 3%Jo for about 15 hr at 22 °C. The fixedsamples were washed with cacodylate buffer, osmi-cated, dehydrated, and embedded as described above .

Viscosity Measurement

An apparent viscosity of the extracts was measuredreproducibly with vertically mounted 0 .1 ml pipets .The pipet tips were heat constricted and were sub-merged in 0.2 ml of sample during viscosity measure-ments. The time for 0.07 cc of extract to flow fromthe pipet was compared with the time for an equalvolume of distilled water to flow out .

Dimension MeasurementThe dimensions of the various types of filaments

were measured precisely with a Nikon Profile Pro-jector (Nippon Kogaku K.K., Japan) directly fromoriginal negative electron micrographs taken atX 20,580. The microscope was calibrated with carbonreplica of a diffraction grating (E . F . Fullam, Inc .,Schenectady, N . Y .) .

OBSERVATIONS

The Effect of Fixation on Amoeba proteus

Adding any of the above fixatives to cultures ofactively moving amebas caused normal cytoplasmicstreaming to cease in less than 30 sec. WithFG-Os0 4 at 0 ° or 22°C or with any of the otherfixatives at 0 ° , there was no further visible cyto-plasmic movement . With the formaldehyde-glu-taraldehyde-based fixatives at 22 °C, the appear-ance of many of the amebas was distorted byviolent abnormal contractions and changes of shapefor 1-2 min following the addition of the fixative .

Substantial shrinkage of the amebas, up to one-half the original volume, occurred during fixationand embedding regardless of the technique used .High osmolarity fixatives caused more rapidshrinkage than low osmolarity fixatives . If theamebas were passed quickly through fixatives andbuffer washes for avoiding shrinkage there, theywould shrink during dehydration .Although the ectoplasm was readily distin-

guished from endoplasm in living amebas, a cleardistinction of these regions was no longer apparentin most fixed amebas .

270 THE JOURNAL OF CELL BIOLOGY • VOLUME 46,1070

Electron Microscopy of Intact Amebas

Both cytoplasmic matrix and membranous or-ganelles were preserved satisfactorily by Kar-novsky's formaldehyde-glutaraldehyde (FG) fixa-tive (Flickinger, 1968), but there was furtherimprovement in preservation when picric acid orrelated trinitro compounds were added to the FGfixative (Ito and Karnovsky, 1968) . The additionof trinitrocresol to FG (FGC) gave a more satis-factory fixation with fewer myelin figures thanFGR or FGP .

Washing fixed cells overnight in buffer solutionsbefore treatment with Os04 removed most of thesmall black cytoplasmic particles which puzzledprevious investigators. Flickinger (1968) obtainedsimilar results with an overnight wash of the FG-fixed amebas in distilled water .

Fig. 1 is a low power electron micrograph of asmall portion of an ameba showing the plasmamembrane, a cluster of thick filaments, mitochon-dria, two Golgi complexes, endoplasmic reticulum,and several types of vacuoles . The cytoplasmicmatrix (or groundplasm), which is preserved bythe improved fixation, contains free ribosomes andglycogen particles suspended by gray backgroundmaterial which may appear amorphous, granular,or reticular (Fig . 2) .

Two types of filaments but no microtubules wereobserved in the cytoplasm of randomly selectedamebas, confirming the observation of Nachmias(1968) on Chaos . Most sections of amebas had160 A wide, solid filaments up to 5000 A long(Figs . 1-3) . These filaments were most often foundin randomly oriented clusters near the plasmamembrane or groups of mitochondria. Occasion-ally, they were packed in a parallel array as if theywere in a stream of flow or under tension(Nachmias, 1964) . In some cells most of the 160 Afilaments were found singly, distributed through-out the cytoplasm.

The second type of filament observed was about70 A wide and indefinite in length (Figs . 3 and 4) .These thin filaments had to be searched for in thesectioned cells. They usually were found close togroups of 160 A filaments oriented in a parallel orreticular array .

In an effort to localize these filaments in anameba undergoing a well defined movement,single fountain-streaming amebas were fixed inFGP and embedded. Phase-contrast photomicro-

Dow


FIGURE 1 An electron micrograph of a small peripheral portion of Amoeba proteus showing the cyto-plasmic structures : thick filaments (TkF), plasma membrane (PM), pinocytotic vesicles (PV), foodvacuoles (FV), Golgi complexes (GC), mitochondria (Mt), and the cytoplasmic matrix, groundplasm .FGC fixation . X 17,000 .

graphs were taken at each step in the procedure .Shrinkage and abnormal movements during fixa-tion usually distorted the normal morphology . Onespecimen which appeared to be well preservedexcept for shrinkage and loss of distinction of theendoplasm from the ectoplasm was serial sectioned .Scattered 160 A filaments were found in both theendoplasm and the ectoplasm, without clear dif-ferences in any region . Few 70 A filamentsseen .

Cooling amebas to 4 °C for several hours andfixing at 4 °C did not reduce the apparent numberof 160 A filaments. There were too few 70 A fila-ments in fixed amebas for estimating the numberof these filaments at different temperatures .

were

Fine Structure of the Extracts

Electron microscope examination of thin sec-tions of Extract I revealed it to be a crude cyto-plasmic fraction of groundplasm, mitochondria,endoplasmic reticulum, smooth membranousvesicles, glycogen particles, and small fragmentsof plasma membrane (Fig. 5) . Occasionally,groups of 160 A filaments and, very rarely, a fewsmall bundles of 70 A filaments were seen in Ex-tract I kept at 0-4 °C (Table I) .

Centrifugation of Extract 1 for 10 min at10,000 g removed most of the membranous organ-elles and bundles of filaments . The supernatant,Extract 2, was almost pure groundplasm (Fig . 6)with very few small vesicles and microsomes . No

T. D . POLLARD AND S. ITo Filaments of A . proteus. 1

271

Dow


FIGURE 2 An electron micrograph showing the detailed structure of the cytoplasmic matrix or ground-plasm (GP), which consists of free ribosomes suspended in a matrix which is amorphous, granular, orreticular in different areas . A small group of thick filaments (TkF), plasma membrane, and a mitochondriaare also pictured . FGP fixation . X 39,000 .

filaments were observed in Extract 2 kept at 0-4 °C(Table I) .

In the phase-contrast microscope, the mitochon-dria and vesicles in the extracts appeared as darkparticles suspended in an amorphous gray back-ground (see the background in Figs. 7 and 8) .

Light Microscope Observation of the

Motile Extracts

Extract 1 underwent dramatic streaming andcontraction when warmed from 0 ° to 22 °C in thepresence of 2-5 mm ATP as described byThompson and Wolpert (1963) and demonstratedin a movie at the 1968 meeting of the AmericanSociety for Cell Biology (Pollard and Ito, 1968) .Movement was divided into two phases : Phase 1was the movement observed before the formationof fibrils visible in the phase-contrast microscope,and Phase 2 was the movement seen after theformation of visible fibrils. (Note : fibrils were ob-

272


served in the light microscope and filaments wereobserved in the electron microscope) .

In samples of extract sealed carefully undercoverslips, there was an even distribution of par-ticles surrounded by gray amorphous backgroundmaterial . In control samples, and initially inmotile samples, there was only Brownian move-ment of the particles. In Extract 1, the crude cyto-plasmic fraction, the initial phase of movement,Phase 1, began from 30 sec to 3 min after the slideswere transferred from 0 ° to 22 °C. First, there weresaltatory movements of individual particles . Grad-ually, larger groups of particles moved in unisonand appeared to be held in a semirigid structureinvisible in the phase-contrast and polarizing mi-croscopes. Large streams of particles formed andflowed in different directions. Two streams movingin opposite directions separated by an apparentlygelled area were often seen within one microscopefield . The effect was strikingly similar to the flow of

Dow


FIGURE 3 An electron micrograph showing the two types of cytoplasmic filaments : a large cluster ofrandomly oriented thick filaments (TkF) and a small parallel array of 50-70 A thin filaments (TnF) .FGC fixation. X 81,000.

axial endoplasm in a shell of ectoplasmic gel in anintact ameba and to the observations of streamingin broken amebas (Allen, Cooledge, and Hall,1960). The movement could be quite violent asareas of the extract contracted and retracted fromeach other; this resulted in the concentration ofparticles in the apparently gelled areas . This initialphase of movement lasted up to 10 min in Extract1, before movement gradually subsided .Control samples of extract without ATP con-

tinued to show only Brownian motion after warm-ing to 22 °C, and the particles remained evenlydistributed . Consequently, it was easy to tell whichsamples had moved, even after the movement hadceased, by the variation in concentration of par-ticles in the gelled areas and elsewhere . Convincingmovement was observed in eight out of elevenpreparations of Extract 1 prepared by the method

described above . Two of the failures occurred whendying ameba cultures were used to make theextract .

Phase 2 movement began after about 10 min at22 °C with the formation of very fine fibrils whichgradually increased in thickness as they were ob-served in the phase-contrast microscope . The fibrilsshowed strong, positive birefringence . Particleswere often attached to these fibrils and appearedto be pulled into clusters of other particles andfibrils as the fibril to which they were attachedshortened. As the fibrils shortened, there was nodetectable increase in their diameter ; their proxi-mal end appeared to vanish into the cluster ofparticles toward which they moved, giving one theimpression that the fibril was being "pulled into"the cluster . Circular arrays of fibrils hundreds ofmicrons in diameter sometimes formed and con-

T. D . POLLARD AND S . ITo Filaments of A . proteus. I 273

Dow


FIGURE 4 An electron micrograph of the periphery of an ameba showing an extended reticular arrayof 50-70 A thin filaments (TnF) between the plasma membrane and the nucleus (Ncl) . The thin fila-ments are organized into parallel arrays in some areas . The groundplasm is present in areas not oc-cupied by the thin filaments. The fibrous lamina (FL) is prominent in the nucleus . FGP fixa-tion. X 49,000 .

274

Dow


FiGuREs 5 and 6 Electron micrographs of thin sections of extracts of Amoeba proteus at 0°C .Fixation with 3% glutaraldehyde at 0 °C. Collection by centrifugation at 140,000 g .

FIGURE 5 Extract 1 at 0°C consists of mitochondria, vesicles, fragments of plasma membrane (PM),endoplasmic reticulum, and groundplasm (GP) . Occasional 160 A filaments and very rare 50-70 Afilaments are present in Extract 1 at 0°C-not shown. X 26,000.

FIGURE 6 Extract 2 at 0° C is essentially pure groundplasm . Rare vesicles, but no filaments are pres-sent in samples kept at 0-4 °C. X 45,000.

tracted toward their center, sweeping all of theparticles ahead of them into a large cluster visiblewith the naked eye . In both of these types of move-ment, the fibrils appeared to transmit the tensionnecessary to move the particles . Loose clusters ofparticles and fibrils contracted into tight clusters(Fig. 7) . This phase of movement lasted up to 90min and occurred in five out of eleven preparationsof Extract I .An alternate to this second phase of movement

was the formation of fibrils and spindle-shapedtactoids (Fig . 8) without visible movement or theformation of clusters of particles . The tactoids werealso birefringent (Fig . 9) . This was observed in fourout of eleven preparations of Extract 1 .

Phase 1 movement was observed in only two outof seven preparations of Extract 2, the purifiedcytoplasmic fraction, after warming to 22 °C withATP. When it did occur, it was transient and lastedless than 2 min. In six of seven preparations ofExtract 2, extensive networks of birefringent fibrilsand tactoids formed, starting about 10 min afterwarming with ATP. These fibrils and tactoids in-creased in diameter, but convincing movement wasnot observed . Consequently, little or no movementwas observed in either phase in preparations ofExtract 2 .

The fibrils and tactoids were cold labile . Theydisappeared in samples stored overnight at 4 °Cand returned in 10-20 min after rewarming to22°C. There was no movement during the reforma-tion of the fibrils .

Conditions for Movement and Fibril Formation

Warming of Extract 1 from 4 ° to 22 °C was re-quired for movement . Convincing Phase 1 move-ment was observed on one occasion in Extract 1without the addition of ATP, but ATP (2-5 mm)stimulated movement in multiple samples of eightout of eleven preparations of the extract . Extractsprepared with Tris-maleate buffer moved moreconsistently than those prepared with distilledwater. Extract stored at 4 °C lost the capacity forATP-stimulated movement after about 1-2 hr, asdescribed by Wolpert, Thompson, and O'Neill(1964) .For examining the calcium requirement for

movement, the calcium chelating agent EGTA(I , 2 -bis(2 -bicarboxymethylaminoethoxyethane) )at 0 .33 mm was added to Extract I with ATP . Themovement on warming to 22 °C was equivalent toPhase I and 2 movements stimulated by ATPalone. EGTA (0.33-1 .0 mm) alone did not stimu-late movement.

T. D . POLLAnn AND S. ITO Filaments of A . proteus . 1

275

Dow


276

TABLE I

Morphological Analysis of Extracts of Amoeba proteus

Fibril formation also required warming to 22 °C,but in contrast to movement, both ATP andEGTA stimulated fibril formation . EGTA pro-moted only a small amount of fibril formationcompared with ATP .

FIGURE 7 Phase-contrast photomicrographs of Extract 1 at 22° C with 4 Mm of ATP. One type of ATP-stimulated movement observed in Phase 2-the condensation of a cluster of particles and fibrils-is shown(Fig . 7 a) . The length of the cluster decreases about 35% in Fig . 7 b and about 607 in Fig . 7 c and d .This sequence was taken over a period of 10 min . The vesicles and mitochondria in the extract appearas small dark particles surrounded by amorphous gray material . X 250 .

TIIE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970

Viscosity MeasurementsSerial measurements of the apparent viscosity

were made on samples from five preparations ofExtract 1, while movement was monitored in twoother samples of the same preparation sealed under

Extract I Extract 2

Components at 0 °C Major : Groundplasm GroundplasmFree ribosomes Free ribosomesGlycogen particles Glycogen particlesRough endoplasmicreticulum

Smooth vesiclesPlasma membrane

vesiclesMitochondria

Minor : 160 A filaments Rough endoplasmicreticulum

70 A filaments (rare) Smooth vesiclesPlasma membrane

vesicles

New components 50-70 A filaments 50-70 A filamentsafter warming to (numerous) (numerous)22°C with ATP

ATP-stimulated Reliable Rare and uncon-motility vincing

Dow


FIGURE 8 A phase-contrast photomicrograph of fibrils and tactoids formed in Extract 1 after warmingto 22 °C with ATP . X 430 .

FIGURE 9 A polarizing photomicrograph of birefringent tactoids formed from Extract 1 with ATP at22°C. X 930.

coverslips . The apparent viscosity of Extract 1warmed to 22 °C with ATP increased to a maxi-mum, coinciding with the peak of Phase I move-ment, and then fell to control values as aggregatesof the extract precipitated out of solution in Phase2 (Fig. 10) . This increase in apparent viscosityoccurred only if there was active Phase 1 move-ment and never occurred if there was no move-ment, as in the controls without ATP .

The extent of the rise in viscosity during Phase 1appeared to be determined by the length of timebefore the precipitation of aggregates, after whichthe viscosity always decreased to control values .The preparation used for Fig. 10 demonstrated thegreatest increase in viscosity and no precipitateformed until after 8 min at 22 °C. In two prepara-tions, little or no rise in viscosity could be measureddespite good Phase I movement, but visible pre-cipitates formed in the viscometer after 2 or 3 min .In the remaining two preparations, in which inter-mediate viscosity increases were measured, pre-cipitates formed after 5 or 6 min . These precipitateswere not apparent in the samples sealed on micro-scope slides or in test tubes ; so it seemed likelythat the repeated flow of the extract into and outof the viscometer promoted aggregate formation .These aggregates consisted of clusters of particles,fibrils, and tactoids similar to those observed inPhase 2 .

Electron Microscopic Observations of theMotile Extracts

Samples of Extract I fixed during active move-ment (Figs. I1 and 12) had extensive regions inwhich the membrane fragments and organelleswere enmeshed in dense network of thin 50-70 Afilaments and clusters of thick 160 A filaments (seeTable I). The thick filaments were frequentlyfound at points at which groups of thin filamentsappeared to merge as in Fig. 11 . This arrangementof thick filaments at the center of radiating thinfilaments was also seen in moving samples ofExtract I prepared for electron microscopy with-out 140,000 g centrifugation . This gave someassurance that this organization existed in theextract and was not an artifact of the preparation .Some of the thin filaments appeared to abut onthe surface of the thick filaments, membranefragments, or organelles (Figs . I I and 12) .

In samples of Extract 1 fixed after the formationof birefringent fibrils, 70 A filaments were fre-quently arranged on long parallel arrays (Fig . 13) .The parallel arrays of thin filaments would be ex-pected to be birefringent, and the dimensions ofthese aggregates were the same as the dimensionsof the fibrils observed in the light microscope,strongly suggesting that the birefringent fibrilswere formed from 70 A filaments .

T. D . POLLARD AND S . ITo Filaments of A . proteus . 1

277

Dow


Extract 1 warmed to 22 °C without added ATPwas found to have a few small clusters of thin fila-ments more frequently than similar samples keptat 0-4 °C, but these filaments did not aggregateinto large fibrils visible in the light microscope .

The extensive arrays of birefringent fibrils whichform in Extract 2 after warming with ATP alsocomprised parallel arrays of 50-70 A filaments(Fig . 14), just as the fibrils in Extract 1 . No 160 Athick filaments were seen in Extract 2 (see TableI) .At the end stage of ATP-induced movement of

Extract 1, large pseudocrystalline aggregates ofthick or thin filaments formed (Figs . 15-17) . Thepseudocrystals of parallel thin filaments were thebirefringent tactoids seen in the light microscope(Figs . 8 and 9) . The parallel, tightly packed arrays

278

THE JOURNAL OF CELL BIOLOGY • VOLUME 46, 1970

MINUTES AT 22°C

FIGURE 10 The change of relative apparent viscosity of Extract 1 after warming to 22 °C f 4maI ATP. Relative apparent viscosity = s° rlo where t = time for 0.07 ml of Extract to flow frompipet viscometer, t„ = time for 0 .07 ml of water toJ flow from pipet viscometer, no = the apparentviscosity of water, taken to be 1 .00 here . • •, relative apparent viscosity of Extract 1 + 4

mm of ATP . 00, relative apparent viscosity of Extract 1 + volume of water equal to thevolume of added ATP . Two samples each of the Extract + ATP and Extract alone were observedin the phase-contrast microscope by an independent observer, while the viscosity was measured . Phase1 movement began at 2 min after warming to 22 °C. Fibrils appeared after 9 min, marking thebeginning of Phase 2 movement.

of thin 70 A filaments excluded all other structuralelements. The aggregates of thick filaments gen-erally excluded other structures, but there weresome amorphous matrix materials or a few thinfilaments between the thick filaments (Fig . 15) .

In cross-section, the thick filaments seen in Ex-tract 1 were solid with some fine lateral projections(Fig . 18) . In longitudinal section, the thick fila-ments measured up to 0 .5 µy long. The mean widthof the thick filaments (such as those in Fig. 18) was157 A (so = 22 A) . Some thick filaments haddistinct longitudinal striations, and others ap-peared to end in a spray of very fine threads .

The long, thin filaments in tactoids and fibrils(Figs. 13-16) were 72 A (sv = 10 A) wide andappeared slightly beaded at high magnification .A cross-section of a tactoid showed that the "70

Dow


FIGURE 11 An electron micrograph of Extract I undergoing Phase 1 movement 8 min after warmingto 22 °C with ATP . An extensive filamentous network surrounds and abuts on (arrow) the membranefragments and vesicles . Groups of 160 A thick filaments are located at points where groups of thin fila-ments intersect. Fixation with 3% glutaraldehyde at 22 °C. Collection at 140,000 g. X 21,000 .

279

Dow


FIGURE 12 An electron micrograph of Extract 1 undergoing Phase 1 movement 8 min after warmingto 22 °C with ATP. Thin filaments 50-70 A wide radiate from a cluster of 160 A thick filaments . Prepara-tion as in Fig. 11. X 80,000.

A" filaments were clearly separated from adjacentfilaments in an irregular array (Fig . 17) . Insamples fixed during Phase 1 and early Phase 2movement, there were 70 A filaments identicalwith those seen in the fibrils and tactoids, but therewere also substantial numbers of filaments whichappeared thinner (Figs . 12 and 13) . These thinnerfilaments were 54 A (so = 6 A) wide . These"50 A" thin filaments were frequently seen in con-tinuity with the 70 A thin filaments (Fig . 13) .There was no population of filaments intermediatein diameter between the 160 A thick and 70 A thinfilaments at any stage .

DISCUSSION

This study of ameboid movement was undertakento investigate the mechanisms of consistency

280


changes and movement in Amoeba proteus . The ob-servations reported above led us to conclude, asexplained in detail below, that increases in cyto-plasmic consistency are caused by the formation of50-70 A filaments from precursors in the ground-plasm. A second type of filament, about 160 A indiameter, was observed, and we postulate thatthese thicker filaments may interact with the thinfilaments to cause contraction of the cytoplasm .

Fixed Specimens of Amoeba proteus

Studies of living giant amebas (Allen, 1961) haveshown that the low consistency endoplasm of theameba appears to be reversibly converted to higherconsistency ectoplasm during movement . It washoped that it would be possible to fix an amebaduring movement and to examine the various

Dow


FIGURE 13 An electron micrograph of Extract 1 at 22 °C for 45 min with ATP. The parallel arraysof 70 A filaments are thought to be the birefringent fibrils observed to form during Phase 2 movement insamples warmed with ATP . The 70 A filaments in the upper parallel array (TnF) appear continuouswith thinner (about 50 A) filaments (arrow) bridging them to the cluster of 160 A filaments in the lowerright (TkF) . Preparation as in Fig . 11 . ER, endoplasmic reticulum . X 70,000.

regions of the cell in the electron microscope todirectly observe which components of the cell areresponsible for the difference in consistency be-tween the endoplasm and the ectoplasm . Suchstudies were previously attempted with Chaoscarolinensis (Komnick and Wohlfarth-Botterman,1965) and Saccamoeba sp .2 (Bhowmick, 1967), butthe results were not convincing because of thefollowing problems :

(a) One must observe the cell before and dur-ing preparation for electron microscopy to becertain that the cell being studied was movingnormally when fixed and to precisely identifyareas in electron micrographs as specific parts ofthe living cell. This has not previously beendone .

2 Originally stated to be Trichamoeba villosa butactually Saccamoeba sp ., strain F-13, Griffin (J .Griffin, personal communication) .

(b) Agonal contractions during fixation alterthe gross morphology and presumably the finestructure of the amebas. Certain structural fea-tures that are obvious in living amebas may dis-appear during fixation . For example, the cleardistinction between the endoplasm and the ecto-plasm in living amebas is usually lost duringfixation . This may explain the absence of ultra-structural distinction between the endoplasmand the ectoplasm in fixed amebas .

(c) Shrinkage of the fixed cells concentratesthe cytoplasm and may also disorganize someof the cytoplasmic structures . Schäfer-Danneel(1967) considered this problem but also wasunable to circumvent it .

(d) The cells should be moving-not sta-tionary, as when exposed to pinocytosis-inducingsolutions (Nachmias, 1964) or cold (Schäfer-Danneel, 1967) .

T. D. POLLARD AND S . ITO Filaments of A . proteus . 1

281

Dow


FIGURE 14 An electron micrograph of parallel arraysof 50-70 A filaments (TnF) in Extract 2 after warmingto 22°C for 8 min with ATP . These thin filaments arethrought to form the birefringent fibrils and tactoidswhich appear in Extract 2 warmed with ATP . Fixa-tion with FGC at 22 °C. X 65,000.

In an effort to avoid these problems, singlefountain-streaming amebas were fixed and em-bedded under close microscopic observation. Care-ful selection of specimens with minimal abnormalmovements during fixation and serial sectioningavoided the first problem, but shrinkage and im-perfect fixation, as observed in the phase-contrastmicroscope, limited this approach .

No fine structural difference could be detectedbetween regions of the cell known to be endoplasmand ectoplasm . This, of course, does not mean thatthere are no structural differences between theseregions in the living ameba . The most likely ex-planation for our inability to preserve this postu-lated differentiation is that the structures involved,presumably thin filaments (see below), are labile,and that the chemical and physical irritation of thecell by the fixative caused them to depolymerize .The loss of visual distinction of the ectoplasm fromthe endoplasm observed by phase-contrast micros-

282

THE JOURNAL OF CELL BioLOGy • VOLUME 46, 1970

FIGURE 15

An electron micrograph of Extract 145 min after warming to 22 °C with ATP. Pseudo-crystalline arrays of thick (TkF) and thin (TnF)filaments are homogeneous except for some amorphousmaterial between the thick filaments . The pseudo-crystals of 70 A filaments are thought to be the bire-fringent tactoids observed in the light microscope .Preparations as in Fig . 11. X 39,000.

copy during fixation, and the observation of Allenand Griffin (Allen, 1961) that the gel structure islabile and disrupted by agitation or other mildtreatments of the cell, are in agreement with thisexplanation.

Two Classes of Filaments in Amoeba proteus

Previous investigators working on Amoeba proteus

and Chaos carolinensis have observed cytoplasmicfilaments ranging in size from 20 to 220 A in diam-eter (Nachmias, 1964, 1968 ; Wolpert, Thompson,and O'Neil, 1964 ; Morgan, Fyfe and Wolpert,1967; Schäfer-Danneel, 1967) . All hypothesizedthat the thicker filaments were aggregations ofthinner filaments . The following observations arein favor of there actually being two classes of fila-ments in Amoeba proteus ; 160 A diameter rods about

Dow


FIGURES 16 and 17 Extract 1 after 45 min at 22°C with ATP . Preparation as in Fig. 11 .

FIGURE 16 A longitudinal section of a parallel array of thin 70 A filaments . The filaments have nodefinite substructure in this preparation, but they may be slightly beaded . X 145,000.

FIGURE 17 A cross-section of 70 A filaments in an irregular array . Individual filaments are clearlyseparated from each other. X 145,000 .

0 .5 µ long and 50-70 A thin filaments of indefinitelength

(a) The exact length of the thin filaments couldnot be determined in sectioned material, but theyappear to be quite long . The 160 A filaments areless than 0 .5,u long . It seems unlikely that the long70 A filaments would break up into 0.5 µ lengthsduring aggregation to form the 160 A filaments .

(b) There is no significant population of fila-ments of intermediate size in the cells or cyto-plasmic extracts . Careful measurement of thedistribution of filaments sizes supports this view.Assuming a random distribution of filament sizearound the mean wdith, 95% (mean f 2 sD) ofthe filaments would fall into these ranges : 160 Athick filaments, 113-201 A ; and 70 A thin fila-ments, 52-92 A . There is no overlap in thesedistributions . Overlap might be expected if the

70 A filaments aggregated to form the 160 Afilaments .

(c) Large numbers of 50-70 A filaments form inExtract 2 warmed to room temperature with ATP,but no 160 A filaments are found in Extract 2either before or after warming . Therefore, eitherthe 160 A filaments or factors necessary for theirformation are removed by centrifugation of Ex-tract I at 10,000 g for 10 min . The observation ofscattered bundles of 160 A filaments in the 10,000 gpellet favors the former explanation and makes itseem unlikely that the 70 A filaments aggregate toform the 160 A filaments.

(d) Pseudocrystals made up entirely of 160 Arods or 70 A filaments form after ATP-stimulatedmovement in Extract 1 . Mixed or heterogeneouscrystalline arrays are not observed . This may be an

T. D. POLLARD AND S. ITO Filaments of A . proteus. 1

283

Dow


FIGURE 18 160 A filaments in longitudinal and cross-section . Extract 1 8 min at 22 ° C with ATP . Thethick filaments are solid in cross-section with some fine lateral projections (circle) . There are distinctlongitudinal striations, and some filaments end in a spray of fine filaments (arrow). Preparation as inFig . 11 . X 145,000 .

indication that the two types of filaments havedifferent chemical compositions .

(e) The microscopic fibrils composed of 50-70 Afilaments are cold labile, while the 160 A rods arenot. The 50-70 A filaments, if present, are foundin very small numbers in extracts fixed at 0 °C. Onthe other hand, substantial numbers of 160 Afilaments are seen in whole cells and cell fractionsat 0-4 °C .

(f) In the intact ameba, the 50-70 A filamentsappear to be more labile and difficult to preservethan the 160 A filaments . Both types of filamentsare found in preparations of ameba extracts,where, as hypothesized above, the thin filaments'state of polymerization cannot be influenced bychanges in the cell membrane . Large numbers of160 A filaments are preserved in the cell and incytoplasmic extracts .

Clarification or reinterpretation of earlier ob-servations supports our hypothesis . Nachmias

284


(1964) described 150 and 75 A filaments in Chaoscarolinensis which are identical with the thick andthin filaments of Amoeba proteus . Wolpert,Thompson, and O'Neill (1964) demonstrated thatthe filaments formed from Extract 2 were thinner(80-90 vs . 120 A) than the filaments formed fromExtract 1 . They did not appreciate that they mayhave been dealing with two classes of filaments .With improved fixation techniques, the thinnerfilaments appear 50-70 A wide and are visualizedin both Extract 1 and Extract 2 . Their "spongy"material of Extract 1 is now resolved into networksof thin filaments . Their thicker filaments, some ofwhich measure up to 150 A wide in their Fig. 6,are identical with our 160 A filaments . Of specialinterest is their failure to find thick filaments inExtract 2, which supports our similar observation .Morgan, Fyfe, and Wolpert (1967) described20-30 A filaments in negatively stained extracts ofAmoeba proteus . These filaments appeared about

Dow


40 A wide in sectioned material and may be thesame as the thinnest filaments which we observedin sectioned material . Schäfer-Danneel (1967) alsoobserved thick and thin filaments in glycerinatedspecimens of Amoeba proteus .

Although the definite answer to the question ofhow many types of filaments are found in theseamebas will have to await their isolation andchemical characterization, these observations sug-gest that the thick and thin filaments are twochemically distinct types .

Thin Filaments

Long thin 50-70 A filaments are observed in in-tact cells and Extracts I and 2 . Very few thin fila-ments are seen in extracts at 0-4 °C, so the numer-ous filaments seen at 22 °C probably form fromprecursors in the groundplasm, the principalcomponent of these extracts .

At least two factors participate in the control ofthin filament formation : temperature and ATP .Warming the extracts to room temperature is anabsolute requirement for filament formation . Afew filaments form after warming, even in the ab-sence of ATP or EGTA, and neither of these com-pounds stimulates filament formation at 4°C. Inaddition, fibrils of thin filaments disappear whencooled to 4°C overnight . Although it is not knownwhether the filaments simply disaggregate orwhether they break down into subunits in the cold,these observations are consistent with filamentformation being a reversible endothermic reaction .Under appropriate conditions, the polymerizationof muscle actin has similar properties (Grant,1965) .

Addition of ATP greatly increases the number offilaments observed after warming the extracts toroom temperature. Because some filaments formwithout added ATP, it is not clear whether thepresence of ATP is an absolute requirement forpolymerization. If it is required, endogenous ATPmust be responsible for those filaments formedwithout its addition . Since the system probably isATP-poor after 24 hr at 4 °C, the number of fila-ments formed is limited.

Under suitable conditions, thin filaments ag-gregate laterally to form the birefringent fibrilsobserved in the light microscope. Morgan, Fyfe,and Wolpert (1967) reported that EGTA orEDTA will aggregate thin filaments from a high-speed supernatant of Amoeba proteus cytoplasm intofibrils, suggesting that the chelation of divalent

cations, especially calcium, is important in theaggregation process. We observed that fibril for-mation occurs in the crude cytoplasmic prepara-tions warmed with EGTA, ATP, or EGTA andATP. More fibrils formed in the presence of ATP,probably because more filaments formed, as dis-cussed above. Aggregation may be a separateprocess involving chelation of calcium by EGTA,EDTA, or ATP .The thin filaments vary in size between about

40 and 90 A. Three observations suggest that50 A filaments aggregate to form 70 A filaments :(a) There is considerable overlap in the distribu-tion of the sizes of the filaments . The 95% rangesare 42-66 A for the "50 A" filaments and 52-92 Afor the ` 70 A" filaments . (b) The 50 and 70 Afilaments are frequently observed in continuity .(c) The number of 70 A filaments in the extractsgradually increases as the proportion of 50 Afilaments decreases with time after warming to22°C . This concept of the relation of these twosizes of thin filaments agrees with the observationsof Morgan, Fyfe, and Wolpert (1967) on negativelystained filaments from Amoeba proteus .

These thin filaments are similar to muscle actinin size, shape, (Hanson and Lowy, 1963), forma-tion with the absorption of heat and proposedmolecular structure, including interaction withATP. Are they formed from an actin-like mole-cule? Evidence is rapidly accumulating that thincytoplasmic filaments from nonmuscle cells suchas these are formed from actin-like proteins .Hatano and Oosawa (1966) isolated actin fromslime mold, and Weihing and Korn (1969) puri-fied another from Acanthamoeba castellanii . Boththe purified ameba F-actin from Acanthamoebaand the thin filaments in glycerinated specimensof Acanthamoeba form "arrowhead" complexes withrabbit muscle heavy meromyosin (HMM) identi-cal with the complexes formed from muscle actinand HMM, demonstrating that the 60 A filamentsin that ameba are ameba F-actin (Pollard, Shel-ton, Weihing, and Korn, 1970) . In a subsequentreport (Pollard and Worn, 1970), we will showthat thin filaments in extracts of Amoeba proteusspecifically bind muscle heavy meromyosin toform arrowhead complexes which can be disso-ciated by Mg-ATP, strongly suggesting that thesefilaments are also actin .

To summarize what is known about the thinfilaments : they are actin-like 70 A filaments whichare cold labile, require both heat and ATP for

T. D . POLLARD AND S . ITo Filaments of A . proteus . 1

285

Dow


complete polymerization, and aggregate intomicroscopic birefringent fibrils in the presence ofcalcium chelators .

Thick Filaments

The 160 A wide, 0 .5 u long rodlike filamentsappear to be stable components of the ameba .They are present in cells and extracts at 0 °C aswell as 22 °C, in motile and nonmotile amebas andextracts, and in high and low viscosity extracts . Itis possible that more of these thick filaments formin Extract 1 after warming to 22 °C with ATP,but this was not certain from our observations .

Nachmias (1968) observed in negatively stained

preparations that the thick filaments of Chaos hada filamentous substructure similar to that whichwe have observed in sectioned material . She,therefore, postulated that the thick filamentsformed by the aggregation of 50-70 A thin fila-ments. As discussed above, it seems likely that thevery fine filaments which form these thick fila-ments may be distinct from the 50-70 A filaments .

Just as the thin filaments are similar to muscleactin, the thick filaments bear some superficialresemblance to myosin of striated muscle : bothare approximately 150 A rods with proposedfilamentous substructure and lateral projections .If these thick filaments are myosin, it is unusual .

Except for striated muscle, myosin is not aggre-gated into stable filaments under normal intra-cellular conditions . The myosin isolated fromslime mold (Hatano and Tazawa, 1968; Adelmanet al., 1968) is soluble at intracellular ionicstrengths. Smooth muscle myosin sometimes aggre-gates into filaments, but only under special condi-tions related to the state of contraction, pH, andconcentration of divalent cations (Kelly and Rice,1969; Schoenberg, 1969) .

Mechanism of Viscosity Changes

The technical difficulties discussed above makeit impossible to directly demonstrate in an intactfixed ameba any structural differences betweenthe endoplasm and ectoplasm which may accountfor their difference in consistency. Therefore, wehave had to rely on the study of the motile ex-tracts of ameba cytoplasm to formulate a hypothe-sis for the mechanism of viscosity changes andcontraction .The initial phase of ATP-stimulated movement

of Extract 1, Phase 1, closely resembles the cyto-

plasmic streaming described in intact and broken

286 TBE JOURNAL OF CELL BIOLOGY • VOLUME 46, 1970

amebas (Allen, Cooledge, and Hall, 1960) : nomicroscopic fibrils or strong birefringence develop,

and yet the extract moves vigorously and certainareas appear to have increased in consistency . Theapparent viscosity of the extract increases duringthis phase of movement (Fig . 10), and electronmicroscopy reveals that an extended network ofthin 50-70 A filaments has formed in certainregions of the extract (Fig . 11) .

The formation of thin filaments from precur-sors in the groundplasm is probably the cause ofthe increase in apparent viscosity of the extract .This is supported by the correlation of viscosityincrease with the appearance of large numbers ofthin filaments in the extract and is analogous tothe increase in viscosity associated with thepolymerization of other filamentous macromole-cules such as actin (Straub and Feuer, 1950) .

In addition, gel formation in amebas is knownto be an endothermic reaction (Marsland, 1964),so that the observation that heat is required for

the formation of the thin filaments in amebaextracts agrees with the concept that thin fila-ments account for the increased consistency of`gelled" cytoplasm .An alternate, but more complex, mechanism

for consistency changes is that the actin-like thinfilaments first polymerize and then interact witha myosin-like component to form an actomyosingel. If the thick filaments are the myosin-likecomponent, this mechanism is compatible withour observations, although it is not certain thatan actomyosin gel would form in the presence ofthe millimolar quantities of ATP in Extract 1 .

These observations suggest that the variationsin the consistency of different regions of theameba cytoplasm may be directly related tovariations in the proportion of thin filament pre-cursors polymerized into filaments . A network of

thin filaments could account for the limitedmovement of particles in the ectoplasm and theinvisible obstructions to the fall of crystals incentrifuged amebas (Allen, 1961) . In addition,thin filaments which became oriented under ten-sion could account for the increase in birefringenceobserved in specimens of Chaos carolinensis stretchedby the application of pressure and by sporadicreversals in the direction of streaming (Allen,

Francis, and Nakajima, 1965) . These filaments

must be present in low concentrations, and/or

they may be randomly oriented, because the bire-

fringence of ameba cytoplasm is low, with retarda-

Dow


tions on the order of 10`-10-5 (Allen, Francis,and Nakajima, 1965) .

A difficulty with this interpretation is that veryfew thin filaments were seen in fixed amebas, butas discussed above, the ectoplasmic gel may breakdown during fixation . It seems likely (Wolpertand Gingell, 1968) that changes in the cell mem-brane in response to external stimuli may controlthe state of gelation of the cytoplasm . The abnor-mal movements and changes in the ectoplasm ofthe amebas exposed to fixatives also may bemediated by the effects of the fixative on the cellmembrane. This defensive mechanism is absentin the cell-free extracts and may account for theease of fixing the thin filaments there .

Mechanism of Contraction

Contraction requires the development and trans-mission of tension . It is clear from our observa-tions in the light microscope and those earlier ofThompson and Wolpert (1963) that the fibrils,which form in Phase 2 movement of the cytoplas-mic extract, can transmit tension. We have shownthat these fibrils are birefringent and consist ofparallel arrays of 70 A filaments. Fibrils of anysize, from those just resolved in the phase-con-trast microscope to those more than 2 µ wide,appear to transmit tension. By analogy, one canargue that submicroscopic fibrils of thin filaments,or perhaps even single thin filaments, may also becapable of transmitting tension, such as thatnecessary for the saltatory movement of subcellu-lar extracts and in the cells.

The mechanism for the development of tensionis speculative, but a comparison of the motileExtract I with nonmotile Extract 2 gives somepreliminary data on this important problem (seeTable I) . Nonmotile Extract 2 contains the pre-cursors of thin filaments, which readily polymer-ized in the presence of ATP at 22 °C; therefore,the polymerization of thin filaments is alone in-sufficient to cause movement . This is consistentwith our light microscope observation that thereis no visible change in the shape of the fibrils ofthin filaments as they move toward aggregates ofparticles, suggesting that the fibrils are playing apassive role in the transmission of tension and arenot shortening or contracting themselves . Cen-trifugation of motile Extract 1 at 10,000 g tomake nonmotile Extract 2 removes thick fila-ments, vesicles, endoplasmic reticulum, andmitochondria . Any or all of these organelles may

be important for movement . This differs slightlyfrom the observations on the motility of extractsreported by Wolpert's group (Wolpert, Thomp-son, and O'Neill, 1964 ; Morgan, Fyfe, and Wol-pert, 1967). They found that the 9,000 g superna-tant was motile, but that 35,000 and 150,000 gsupernatants were nonmotile unless a "vesiclefraction," the "10-35,000 g pellet," was added tothese high-speed supernatants . They have notreported the components of this vesicle fraction,but this fraction may be similar to the 10,000 gpellet in the present study, as both have the lightmicroscope appearance of a vesicle fraction andboth are required for the movement of the super-natant .Which of the components of the 10,000 g

pellet are required for movement? This has notbeen proven, but speculation based on the ac-cepted model for muscle contraction (Hansonand Huxley, 1953) would suggest that the thickfilaments may be required to interact with thethin filaments to develop tension. The organiza-tion of the two types of filaments in the motileextract (Fig. 9), with thin filaments radiatingfrom clusters of thick filaments, is consistent withthe idea that the two types of filaments interactto produce the movement observed in theseextracts . At the end stage of the reaction, whenthe extract no longer moves, this organization islost, and the filaments aggregate into large pseu-docrystals of thin or thick filaments (Fig . 15) . Onthe other hand, the vesicles removed in the10,000 g pellet may be important for the controlof the development of tension as suggested byHoffman-Berling (1964), or the mitochondriamay be necessary to supply energy.

Wolpert, Thompson, and O'Neill (1964) foundthat Extract 1 moved only when warmed to roomtemperature with ATP or ADP, with ADPstimulating much less movement than ATP . Theyalso noted that the reaction was inhibited bymillimolar quantities of calcium. We noted thatmovement with ATP plus EGTA was equivalentto the movement stimulated by ATP alone . Theseobservations suggest that the contractile mecha-nism of Amoeba proteus is similar to that of musclein its requirement for ATP but dissimilar in itscontrol mechanism, because it is not calciumactivated like muscle (Weber, Herz, and Reiss,1964) .

T. D . POLLARD AND S . ITO Filaments of A . proteus . 1

287

Dow


CONCLUSION

This study confirms earlier observations that ex-tracts of Amoeba proteus move in a manner similarto the intact cell when warmed with ATP. Duringthis movement, there is an increase in apparent

viscosity associated with the formation of 50-70 A

filaments from precursors in the groundplasm . It

is postulated that interaction of these labile thin

Cell . J . Brachet and E. Mirsky, editors .Academic Press Inc., New York. 2:135 .

3. ALLEN, R. D ., J . W. COOLEDGE, and P . J . HALL .1960. Streaming in cytoplasm dissociated fromthe giant ameba, Chaos chaos . Nature (London) .187 :896 .

4. ALLEN,rR . D ., D. W. FRANCIS, and H . NAKAJIMA . *1965. Cyclic birefringence changes in pseudo-pods of Chaos carolinensis revealing the localiza-tion of the motive force in pseudopod extension .Proc . Nat . Acad . Sci. U. S. A . 54 :1153 .

5. BHOWMICK, D. K . 1967. Electron microscopy ofTrichamoeba villosa and amoeboid movement .Exp . Cell Res. 45 :570.

6. BUCKLEY, I . K., and K. R. PORTER. 1967 .Cytoplasmic fibrils in living cultured cells .Protoplasma . 64 :349 .

7. CHALKLEY, H. W. 1930. Stock cultures of amebaScience (Washington) . 71 :442.

8. CLONEY, R . A. 1966 . Cytoplasmic filaments andcell movements : Epidermal cells duringascidian metamorphosis . J. Ultrastruct . Res.14:300.

9. Flickinger, C . J . 1968. Mitochondrial poly-morphism in Amoeba proteus. Protoplasma . 66 :139 .

10. GRANT, R . J . 1965 . The reversibility of G-actin-ADP polymerization : physical and chemicalcharacterization . Doctoral Thesis, ColumbiaUniversity .

11 . GRIFFIN, J . L. 1960. An improved mass culturemethod for the large, free-living amebae .Exp. Cell Res. 21:170.

12. HANSON, J ., and HUXLEY, H . E . 1953 . Structuralbasis of the cross-striations in muscle . Nature

(London) . 172 :530.

13. Hanson, J ., and J. Lowy. 1963. The structure ofF-actin and of actin filaments isolated frommuscle . Nature (London) . 172:530.

filaments with 160 A filaments constantly presentin the groundplasm causes contraction of the cyto-plasm responsible for ameboid movement .

We thank Miss Jane Mueller for excellent assistanceand United States Public Health Training GrantNo. GM-406 for support .

Received for publication 22 December 1969, and in revisedform 19 February 1970 .

chim . Biophys. Acta . 127:488 .HATANO, S., and M. TAZAWA. 1968 . Isolation,

purification and characterization of myosin Bfrom myxomycete plasmodium. Biochim . Bio-phys. Acta . 154 :507 .

HOFFMANN-BERLING, H. 1964. Comment duringdiscussion . In Primitive Motile Systems in CellBiology. R. D . Allen and N. Kamiya, editors.Academic Press Inc., New York . 171 .

ISHIKAWA, H., R. BISCHOFF, and H. HOLTZER .1969. Formation of arrowhead complexes withheavy meromyosin in a variety of cell types.J. Cell Biol . 43 :312 .

ITO, S ., and M. J. KARNOVSKY . 1968. Formal-dehyde-glutaraldehyde fixative containingtrinitro compounds . J. Cell Biol . 39:168A.(Abstr. )

JARosCrs, R. 1956. Die Impulsrichtungsander-ungen bei der Induction der Induction derProtoplasmastromung . Protoplasma . 47 :478 .

KARNOVSKY, M. J. 1965. A formaldehyde-glutaraldehyde fixative of high osmolality foruse in electron microscopy . J. Cell Biol . 27 :137A (Abstr .)

KELLY, R. E., and R . V. RICE . 1969 . Ultra-

structural studies on the contractile mechanismof smooth muscle . J. Cell Biol. 42 :683 .

KOMNICK, H., and K. E. WOHLFARTH-BOTTER-

MAN. 1965 . Das Grundplasma und die Plasma-

filamente der Amoeba Chaos Chaos nachenzymatischer Behandlung der Zellmembran .

Z . Zellforsch . 66 :434 .

MARSLAND, D. 1964. Pressure-temperature studieson ameboid movement and related phenom-

ena . In Primitive Motile Systems in CellBiology, R . D. Allen and N . Kamiya, editors.Academic Press Inc., New York. 173 .

25. MORGAN, J ., D. FYFE, and L. WOLPERT . 1967 .

16 .

17 .

18 .

19 .

20 .

21 .

22 .

23 .

24.

288


REFERENCES

1 . ADELMAN, M. R., G. G. BORISY, M. L. SHELAN- 14 . HATANO, S ., H. KONDO, and T. MIKI-NOUMURA .Purification of sea urchin egg actin . Exp . CellSKI, R. C. WEISENBERG, and E . W. TAYLOR .

1968. Cytoplasmic filaments and tubules . Fed . Res . 55 :275 .Proc . 27:1186 . 15 . HATANO, S ., and F. OosAwA . 1966 . Isolation and

2 . ALLEN, R . D . 1961 . Amoeboid movement. In The characterization of plasmodium actin . Bio-

Dow


Isolation of microfilaments from Amoebaproteus . Exp . Cell Res . 48 :194 .

26 . NACHMIAS, V. T. 1964. Fibrillar structures inthe cytoplasm of Chaos chaos. J. Cell Biol .23 :183 .

27. NACHMIAS, V . T. 1968 . Further electron micro-scope studies on the fibrillar organization ofthe ground cytoplasm of Chaos chaos. J. CellBiol. 38 :40 .

28. NAGAI, R ., and L. T. REBHUN . 1966 . Cytoplasmicmicrofilaments in streaming Nitella cells. J.Ultrastruct . Res . 14 :571 .

29. POLLARD, T. D. and S. ITO. 1968. The role ofcytoplasmic filaments in viscosity changes andcontraction in extracts of Amoeba proteus. J.Cell Biol . 39:106A (Abstr .)

30. POLLARD, T . D., and E . D . KORN . 1970. Cyto-plasmic filaments of Amoeba proteus. II. Heavymeromyosin binding by thin filaments. Inpreparation .

31 . POLLARD, T . D ., E. SHELTON, R . WEIHING, andE. D . KORN . 1970 . Ultrastructural character-ization of F-actin isolated from Acanthamoebacastellanii and identification of cytoplasmicfilaments as F-actin by reaction with rabbitheavy meromyosin . J. Mol. Biol . 50. In press .

32. PORTER, K. R. 1966. Cytoplasmic microtubulesand their function. In Principles of Biomolecu-lar Organization . G . E. W. Wolstenholme andM. O'Connor, editors. Churchill Publishers,London .

33. PRESCOTT, D. M., and R. F. CARRIER. 1964 .Experimental procedures and cultural methodsfor Euplotes eurystomas and Amoeba proteus . InMethods in Cell Physiology . D. M. Prescott,editor . Academic Press Inc ., New York. 1.

34. SCHÄFER-DANNEEL, S . 1967 . Strukturelle andfunktionelle Voraussetzungen fur die Bewe-gung von Amoeba proteus . Z. Zellforsch . 78 :441 .

35. SCHOENBERG, C. 1969 . Electron microscope studyof the influence of divalent ions on myosinfilament formation in chicken gizzard extractsand homogenates . Tissue and Cell. 1 :83 .

36. SIMARD-DUQNESNE, N ., and P. COUILLARD. 1962 .Ameboid movement. Exp . Cell Res . 28 :85 .

37. STRAUB, F . B., and G . FEUER . 1950. Adenosine-triphosphate . The functional group of actin .Biochim . Biophys. Acta. 4 :455 .

38. THOMPSON, C. M., and L. WOLPERT. 1963 .Isolation of motile cytoplasm from Amoebaproteus. Exp. Cell Res . 32 :156 .

39. TILNEY, L., and J . R. GIBBINS. 1969 . Micro-tubules and filaments in the filopodia of thesecondary mesenchyme cells of Arbacia punc-tulata and Echinarachnius parna . J . Cell Sci . 5 :195 .

40. VENABLE, J . H., and R . COGGESHALL. 1965 . Asimplified lead citrate stain for use in electronmicroscopy . J . Cell Biol. 25 :407.

41 . WEBER, A., R. HERZ, and I . REISS. 1964. Theregulation of myofibrillar activity by calcium.Proc. Roy . Soc. London Ser. B . 1.60 :489 .

42. WEIHING, R., and E . D . KORN . 1969. Amebaactin : the presence of 3-methylhistidine .Biochem . Biophys. Res. Commun . 35:906 .

43. WOHLFARTH-BOTTERMANN, K. E . 1964 . Differ-entiations of the ground cytoplasm and theirsignificance for the generation of motive forceof amoeboid movement . In Primitive MotileSystems in Cell Biology . R. D. Allen and N.Kamiya, editors . Academic Press Inc., NewYork . 79.

44. WOHLMAN, A., and R . D. ALLEN. 1968 . Struc-tural organization associated with pseudopodextension and contraction during cell loco-motion in Difflugia . J. Cell Sci. 3 :105 .

45. WOLPERT, L . 1965. Cytoplasmic streaming andameboid movement . Symp . Soc. Gen . Microbiol.15 :270 .

46. WOLPERT, L., and D. GINGELL. 1968. Cellsurface membrane and ameboid movement .Symp. Soc . Exp. Biol . 22 :169 .

47. WOLPERT, L., C. M. THOMPSON and C. H .O'NEILL . 1964. Studies on the isolated mem-brane and cytoplasm of Amoeba proteus in rela-tion of amoeboid movement . In PrimitiveMotile Systems in Cell Biology . R. D. Allenand N. Kamiya, editors. Academic Press Inc.,New York. 143.

T. D . POLLARD AND S. ITO Filaments of A . proteus. 1

289

Dow


AMOEBA PROTEUS - RU Press

Documents

Transcript of AMOEBA PROTEUS - RU Press