Phagosome formation in Paramecium: effects of solid particles

AGNES K. FOK' , BENJAMIN C. SISON, Jr, MARILYNN S. UENO and RICHARD D. ALLEN

Pacific Biomedical Reseaich Center and Department of Microbiology, Snvder 306, 2S3S The Mall, L'nh'ersitv of Ilatvan, Honolulu,111 96S22, L'SA

•Author fur correspondence

Summary

Digestive vacuole (DV) formation in Parameciumcan be separated into four steps: sequestrationand recycling of the spent DV membrane, sweep-ing and concentrating of particles into the nascentDV, vacuole growth, and the release of the nas-cent DV. How the size, load and surface charge'ofsolid particles affected formation and sub-sequently the digestive processes in log-phasecells were investigated. Pulsing cells continuouslywith beads resulted in an initial linear increase,followed by a steady state, of labelled DVs. Abovea certain threshold concentration, the rate offormation and the size of the DVs formed (as wellas the steady state) all increased with increasingbead load, so that for a 16-fold increase in beadload, a corresponding fourfold increase in theincorporation of the recycled membrane into theDVs was observed. The threshold values, whichdepended on the sensitivity of the technique usedto score these DVs, were lowered as bead sizeincreased. The steady state of labelled DVs wasshown to correspond to a strict balance betweenthe formation and defecation rates as shown byefflux measurements and pulse-chase protocols

using two consecutive labels. The duration of thepulse required to reach these steady states wasinversely proportional to the logarithm of thebead number at low bead loads but remainedconstant at higher bead loads. The formationrates and the DV size were similar when cellswere pulsed with beads from 0-5 to 3/im, but DVsize increased using beads of 4-5 and 5-7 /im. Themaximal bead size that could be ingested was— 10 fun. Beads with a positive surface chargereduced the formation rate as well as the DV size.These results show that: (1) under normal con-ditions particle concentration can determine theformation rate and the DV size; (2) parameciacan form DVs continuously without any detect-able interruptions; and (3) the rate of defecationand thus the rate of recycling of spent DV mem-brane is dependent on the rate of DV formation.

Key words: Paramecium, phagosome formation,membrane recycling.

Introduction

A digestive lysosomal cycle in Paramecium can beseparated into four distinct processes: digestive vacuole(DV) formation, acidification—condensation, lyso-somal fusion-digestion, and defecation (Fok & Allen,1988). The first three processes occur in the processingperiod during which DVs are not lost (Foke/ al. 1982).The formation process can also be subdivided into foursteps. The first step involves the recycling of the spentDV membranes, which are transported in the form ofdiscoidal vesicles along the microtubular ribbons fromthe cytoproct back to the cytopharynx area (Allen &Fok, 1980). The second step involves the sequesteringand concentrating of particles into the nascent DV.

Journal of Cell Science 90, 517-524 (1988)Printed in Great Britain © The Company of Biologists Limited 1988

Ciliary membranelles in the oral region are thought topropel a large amount of water to the buccal cavity, andthe suspended particles carried along with the watercurrent are retained and concentrated in the nascentDV (Fenchel, 1980). The third step involves the fusionof membrane of the discoidal vesicles with the cyto-pharyngeal membrane to form the nascent DV (Allen,1974). The fourth step, the pinching off of the largenascent DV, is mediated by a mechanism sensitive tocytochalasin B (Cohen et al. 1984; Fok el al. 1985).Normally a DV can be formed and released within1 min of pulse, but in the presence of this drug it is notreleased for =?15 min. Cytochalasin B has no immediateinhibitor)' effect on the first three steps; thus, discoidal

517

vesicle membranes continue to be added to the cyto-pharynx resulting in the formation of an abnormallylarge nascent DV (Foke/ al. 1985).

The purpose of this study was to examine howparticle load, size and surface charge affected forma-tion, which in turn could affect the subsequent threeprocesses of the digestive cycle. The results showedthat particle concentration could influence the forma-tion rate as well as DV size and that the formationprocess in turn could regulate the movement andrecycling of the spent DV membrane as well as DVdefecation.

Materials and methods

Paraiiieciuni iiiiilliiiiicruiiuclealuin was maintained in axenicmedium (Fok & Allen, 1979; Fok el al. 1987) and studiedduring mid-log phase of growth. Polystyrene latex beads wereobtained from Sigma Chemical Co. (St Louis, MO) andSeradyn, Inc. (Indianapolis, IN), while fluorescent beadsand beads of different surface charges were obtained fromPolysciences, Inc. (Warrington, PA). Unless otherwise notedin the text, beads were used without prior washing.

Immediately before use, cells were washed with freshMillipore (0-22jim)-filtered axenic medium (phosphati-dylethanolamine (PE) was omitted to ensure the absence ofsmall emulsified particles) using a nylon cloth of 15 jfm mesh(Tetko Inc., Elmsford, NY). For each set of experiments theidentical cell number ml" ' was used, but between differentsets this number ranged between 2000 and 4000. To study theeffect of particle load on the formation rate and DV size, cellswere either pulsed for 3-5 min or fed continuously for up to120 min with various bead loads in the filtered medium. Tostudy the effect of bead size and surface charge, cells werepulsed for 3 min with beads ranging from 0 3 to 5-7^(m orbearing different surface charges. To determine the maximalparticle size that could be phagocytosed, cells were fed for45 min with beads up to 20;(m in diameter. At selected timepoints, cells were fixed with 5 % formalin in 005 M-phosphatebuffer, pH7-6. The mean number of labelled DV's/cell foreach time point was obtained by scoring DVs in 100 cellsusing either bright-field or fluorescence microscopy. The sizeof DVs was determined with a precalibrated Zeiss lightmicroscope. (In this study only labelled DVs were analysed,thus every mention of DVs refers only to labelled DVs.) Beadnumber was estimated according to the following equation(Polysciences Inc., data sheet no. 238):

Beads/ml = 6U' x 1012 x P" 1 x d'3 x jt~] ,

where \Y= grams of latex ml" ' , 6 = diameter in /.tm ofmonodisperse beads, and P = density of latex, 105gml~ ' .

Results

Formation rate

Bead concentration. Results, obtained after washingcells with Millipore-filtered axenic medium lacking PE

and pulsing continuously with varying bead concen-trations, showed that the accumulation of labelled DVscould invariably be divided into two periods: an initialperiod of linear DV formation followed by a secondperiod of no apparent formation (Fig. 1). During thefirst period formation rates increased rapidly withincreasing bead load and were proportional to thelogarithm of bead number from 107 to 10yml~'(Fig. 2). From 10y to 10" beads ml"1, the formationrates were still proportional to the logarithm of thebead numbers, although the increases in these rateswere less rapid, so that in the plot of the formation ratesagainst the logarithm of bead number, two straightlines intersecting at a bead number of about4 x l 0 8 m r ' could be drawn (Fig. 2). Below107 beads ml" ' , DVs contained so few beads that it wasdifficult to score them accurately (see section onthreshold value). This was compounded by a low rateof DV formation. Very similar results were obtained insix different sets of experiments, in which the cellnumber ml~' varied between 2000 and 4000 or whendifferent bead sizes were used.

Not only were the formation rates dependent on thebead load but the number of beads sequestered into avacuole increased dramatically when cells were pulsedwith increasing bead load. Fig. 3 shows a series of cellspulsed for 5 min over a 128-fold difference in beadconcentrations. At the three highest bead concen-trations, DVs were very large and seemed to be packedwith latex beads. But as the bead concentrations were

6-OxlO10

60Pulse time (min)

Fig. 1. Cells pulsed with varying concentrations offluorescent beads (0-26 j(m diameter) were fixed at varioustime points. After washing, counts of fluorescent DVs in3=100 cells were made for each time point. For cells pulsedwith the two lowest concentrations of beads, LuciferYellow was added to aid in the visualization of the labelledDVs.

518 A.K.Foketal.

0-8

0-6

0-4

I 0-2

—C»—O»—&--O-

50

40O

30-18.s

20 g

8 9 10Log bead number

11

Fig. 2. The formation rates and the duration of theprocessing period were obtained from two experiments inwhich fluorescent beads (0-26^m diameter) were used.The processing period, defined as the time betweenformation and the onset of defecation in a population ofcells (Fok el al. 1982), was estimated as the period betweenthe beginning of the pulse to the onset of a steady level.

diluted serially by a factor of 2, fewer beads weresequestered into the DVs, so that DVs in cells pulsedwith 1 75x10 beads ml~ or less had very indistinctboundaries in the light microscope (Fig. 3H).

During the second period, there was no net increasein DV formation as the mean number of labelled DVsreached a constant level for the remainder of the pulse(Fig. 1). The average number of DVs for the steadylevels increased with increasing bead concentration(data not shown), following a pattern identical to thatof the formation rates shown in Fig. 2. Also during thisperiod, a wide variation in the number of DVs inindividual cells was observed. At the lowest beadconcentration at which fluorescent DVs could still bescored accurately, >65 % of the cells had fewer than 10DVs, while in about 20% of the cells no labelled DVswere detected following the 90min pulse (Fig. 4A).With increasing bead concentration, a shift towardmore DVs and a decrease in the percentage of cellsdevoid of labelled DVs were seen (Fig. 4). At thehighest bead concentration used in this experiment,70 % of the cells had over 21 D Vs and very few cells hadless than 10 DVs (Fig. 4D).

Fig. 3. Cells were pulsed for 5 min with beads (0-8jUmdiameter) beginning with 2-13X 109 beads ml~' (A) anddecreasing in concentration stepwise by a factor of 2 to0017xl0'J beads ml"1 (B-H). These pictures were selectedto emphasize the differences in the intensities of thelabelled DVs in cells pulsed with different beadconcentrations and were not meant to represent the averageappearance of a cell at each bead concentration. X330.

The presence of a steady level indicated that citherparamecia were saturated with DVs during the initialperiod so that no new DV was formed and no labelledDV was egested, or the rates of formation and defeca-tion were equal. To distinguish between these twopossibilities, cells were first pulsed continuously for 1 h

Phagosome formation in Paramecium 519

with fluorescent beads. Then the cells were dividedinto two groups: one group remained in the initialpulse, while the second group was chased and fed againwith non-fluorescent beads. The results showed thatduring the first 42min, uptake of the first label wasrapid and linear (0-47 DVscell"1 min"1), reaching asteady level of 19-5DVscell~' (Fig. 5). In the cellsthat were chased and pulsed again with a second label,the non-fluorescent beads, the loss of the older fluor-escent DVs was linear and identical to their own

O *O O wi O(N oi Pi o ^

=: ^ s ^ a ^Vacuoles/cell

Fig. 4. The distribution of labelled DVs during the steadystate. Cells were pulsed with 6xlO7 (A), 2-4X108 (B),9-6X 108 (C) and 6x 10lu (D) fluorescent beads (0-26pmdiameter) per ml.

25r

20

15

•2 10E

.*, . * / « >

A•o-.-o

0 25 50 75 100 125Time (min)

Fig. 5. Cells were pulsed with fluorescent beads (0-06jUmdiameter) continuously ( • ) , or chased with axenic mediumafter 60 min of pulse followed by a second pulse with non-fluorescent beads ( 0 3 (im diameter). Filled and openarrows indicate the beginning of the chase and the secondpulse, respectively, and cells were then scored for bothfluorescent (O) and non-fluorescent DVs ( A ) .

formation rate. The formation rate of the non-fluor-escent DVs in these same cells was also linear(0-6 DVs cell"1 min"1), reaching a higher steady level30 min after the second pulse. These results showedthat cells were able to form DVs continuously and thatolder DVs were egested at a rate similar to theformation rate specific for each label. Thus the attain-ment of a steady level at each bead load was due to astrict balance between formation and defecation.

The above results also showed that the onset of thesteady state corresponded to the onset of the defecationof labelled DVs. Thus the time prior to steady state wasthe period previously designated as the processingperiod (Fok et al. 1982). When the length of theseprocessing periods was plotted against the logarithm ofthe bead concentrations, two straight lines interceptingat 2X108 beads ml"1 were obtained (Fig. 2). The dur-ation of the processing period, which was 48 min at thelowest bead concentration examined, was shortenedwhen the bead load was increased to 2 x l 0 s m l " ' .Further increase in bead number resulted in no furthershortening of this period (Fig. 2).

Bead size and surface charge. The formation rateswere similar when cells were pulsed with bead sizesranging between 0-5 and 4-5 jUm, but were loweroutside this size range (Fig. 6). With neutral or nega-tively charged beads of the same size and concen-tration, DV formation rates were identical (Table 1).However, with the positively charged beads, the forma-tion rate was reduced to 26%.

Threshold value. Paramecium appeared to require alarge number of solid particles before DV formationcould proceed rapidly. The threshold value belowwhich labelled DVs could not be scored accuratelyvaried with bead size and, more importantly, with thesensitivity of the methods used. When cells were pulsedwith non-fluorescent beads of 0-3, 0-6, 0-8 and 1-1 jttm,and the DVs were scored using bright-field mi-croscopy, the threshold values were 1263, 19-7, 8-3 and

0-8rO

a o0-6 -

n

iiii

" • >0 -

l i l t

2301

-21

2 3 4Bead diameter (fim

I19 I

C3

17 £OOCO

15 g

Fig. 6. Effects of bead size on the formation rates and thesize of DVs formed. Cells were pulsed for 3 min with beadsof varying diameter but all containing 0 0 6 % solid latex.Results were the means of two experiments.

520 A. K. Fok et al.

Table 1. Effects of the surface charge of latex beads on the foimation rates and the size of DVs fanned

Bead typeFormation rate

(DVscell"'nun ')Average DV

diameter (flm)DCV membrane

(/lm2 mm ')

Polystyrene (non-fluorescent)Polystyrene (fluorescent)HydroxylatcdCarboxylatcdAmmo

0-6S ±0-11 (10)0-68 ±0-13 (9)0-67 + 011 (10)0-60±0-12(10)0-17±0-04 (10)

17-7(1)18-0 (1)18-5(1)18-7(1)

15-0 ±0-7 (4)

640692720659120

Cells were pulsed with approximately 14 to 3-4x10" beads ml ' (0-2/im, diameter), which were washed once prior to the experiment.Values in parentheses represent the number of experiments. Total discoidal-vesicle (DCV) membrane area was determined by calculatingthe surface area of a DV of average diameter, since DV membrane comes from DCVs (Allen, 1974).

Table 2. Effects of particle number on DV size and the amount of discoidal vesicle membrane incoiporated

Particle number(10vmr')

Formation rate(DVsceir 'mirT1

Average DVdiameter (;<m)

Membrane utilized

4-262-131-070-530-270-13

111-051-010-920-720-73

25-521-117-817-015 615 5

2248 (11 678)1469 (7631)1006 (5 226)836 (4 343)551 (2862)551 (2862)

Cells were pulsed with latex beads of 0-8/im in diameter suspended in a Milliporc-filtered medium for 3 nun. Values in parenthesesrepresent the approximate number of discoidal vesicles incorporated into the DV membrane. The surface area of the membrane of adiscoidal vesicle (0-193/(ni ) was obtained by doubling the area of a disc having an average diameter of 0-35//in (Allen, 1974).

1-6 (XlO6) beads ml ', respectively, a difference of>800-fold. In terms of total solid latex these thresholdvalues differed by 16-fold. When cells were pulsed withfluorescent beads of 0-26fjm and scored using a fluor-escence microscope, the threshold was lowered by afactor of 20 when compared with that obtained using0-3 jum non-fluorescent beads. The addition of solubleLucifer Yellow, a fluorescent non-membrane-permeantsubstance (Swanson et al. 1985), to the fluorescentbeads resulted in an additional fourfold reduction inthe threshold value.

DV size and membrane surface area

Bead concentration. Not only were the formationrates increased with increasing bead load, but theaverage DV diameter also increased linearly (Table 2).Below 0-13 x l0 y beads ml"1, DV diameter could notbe measured as the boundary of most of these DVs wasdifficult to ascertain. At 0-13 X 109 beads ml""1, only15 % of the DVs were 20/im or larger (Fig. 7A). Withmore beads, DVs bigger than 20 fJ.m increased to almost70% (Fig. 7B). As the rate of formation and the size ofDVs increased with bead concentration, the amount ofdiscoidal vesicle membrane incorporated into theseDVs must also increase (Table 2). These results

Fig. 7. The size distribution of labelled DVs in cellspulsed for 3 min with 0-13 (A) and 4-26X 109 beads ml"1

(0-8 jtm diameter) (B).

showed that the rate of movement and utilization ofdiscoidal vesicles is tied to the formation process,which in turn is directly influenced by the bead load.

Bead size and surface charge. When cells werepulsed with beads ranging in size from 0 5 to 3/^m, theaverage DV diameter as well as the formation rateswere the same, so the amount of new membraneneeded for formation remained constant. However,when pulsed with 0-3 f.im beads, DV size and theformation rate both decreased, while DV size increasedwith 4-5 and 5-7jUm beads (Fig. 6). Fig. 8A shows a

50-

40-

30-

20-

£ 10-

1 o-| 50= 40-U

30-

20'

10

r\

A

1 1| |

B

7-10

10-1

5

15-2

0

| 1

20-2

5

25-3

0

30-4

0

DV diameter (;im)

Phagosome fonnation in Paramecium 521

Fig. 8. Cells pulsed for 3 min with 5-7/nn beads (A) andfor 45 min with 10/im beads (B). X27S.

cell pulsed with 5-7/«m beads, which formed a clusterthat filled the DV completely. To determine themaximal particle size that cells could phagocytose, cellswere pulsed for 45 min with 7-20 Jim beads. (The sizevariation of these beads was minimal, less than±0-1 jum, and prior to use they were washed to removethe surfactant, sodium dodecyl sulphate.) Cells took up7/tm beads readily. With the 10/<m beads, uptake wasextremely variable: some cells took in many, whileothers ingested only a few particles. Fig. 8B shows acell fed 10/im beads, two of which can be seen at oneend of the oral region (arrow). No labelled DVs wereformed with 15-20/im beads, indicating that themaximal particle size that can be engulfed by P.multimicroiuicleatiim would be about lOjUm.

The size of DVs formed was the same whether cellswere pulsed with neutral or negatively charged beads(0'2/(tn diameter); but when pulsed with positivelycharged beads of identical size and concentration, theaverage DV diameter was 15 fim. Combined with agreatly reduced formation rate, the amount of discoidalvesicle membrane needed to form these DVs was only18 % of that needed in cells pulsed with neutral ornegatively charged beads (Table 1).

Discussion

Studies on DV formation in Paramecium (MiiMeret al.1965) and Tetrahymena (Miiller et al. 1965; Rasmus-sen, 1976; Nilsson, 1979) have shown that solid par-ticles play an important role in stimulating this process.In this study we show that the formation rate and thesize of DVs formed are dependent on particle concen-tration. In addition, we believe this study providesseveral new findings. First, unlike Tetraiiymena,which has been reported to show some periodicity infeeding (Nilsson, 1972; Ricketts, 1971), paramecia canform DVs continuously without a break. Second, thefact that paramecia can form DVs continuouslysuggests that the discoidal vesicles, the DV-precursormembranes, must be readily available for fusion withthe cytopharyngeal membrane. This requires that the

membranes of the spent DVs at the cytoproct as well asthe retrieved membranes from the young condensingDVs must be transported efficiently and rapidly back tothe cytopharynx area. For example, when cells werepulsed with a 16-fold increase in bead concentration(0-8/im, Table 2), a fourfold increase in discoidalvesicle membranes is incorporated into the phagosomemembrane. This can be possible only if the rate ofrecycling of the spent DV membrane is similarlyincreased. These results suggest that under normalgrowth conditions, the rate of movement and utiliz-ation of the discoidal vesicles is tied to the formationprocess, which in turn is directly influenced by thebead load. This finding is strengthened by our thirdfinding, that the defecation rate is also directlyinfluenced by the formation rate. Thus, as the forma-tion rate is increased when cells are pulsed with higherbead concentrations, concomitant increases in the defe-cation rate take place, so that a steady level of labelledDVs is observed for each bead concentration (Fig. 1).Finally, as the formation rate is increased, the process-ing period is concomitantly shortened. All these resultssuggest that in paramecia the formation process, inaddition to being the first vital link in controlling thedigestive cycle, appears to influence the efficiency ofthe entire cycle.

The results obtained from the formation rate andDV size provide the first indication as to how fastmembranes are transported from the cytoproct to thecytopharynx region. In the experiment shown inTable 2, when cells were fed 4-3 X 10y beads ml"1, theformation rate was 1-1 DVcell"1 min"1 and DV diam-eter was 25'5/im, and the membrane needed everyminute would be equivalent to that contained in about12000 discoidal vesicles of 0-35 jum in diameter. Asthere are about 40 cytopharyngeal microtubular rib-bons (Allen, 1974), 300 vesicles min"1 would need tobe transported on each ribbon to provide this amountof membrane. Thus the membrane transport rate is100/im min"1, which is in line with known speeds ofvesicle movement along microtubules in other mem-brane transport systems (Vale, 1987). Thus it istheoretically possible for the cell to supply the amountof membrane required for the observed DV formationrate by recycling membranes of the spent DVs from thecytoproct along the 40 microtubular ribbons to thecytopharynx.

Paramecium under our experimental conditionsmight at first appear to be a relatively inefficient filterfeeder, as millions of particles ml" must be presentbefore sufficient particles are enclosed in a DV to makeit easily detectable. However, if the amount of fluid thecell needs to filter to form one DV is calculated, one isimpressed more by the efficiency rather than inef-ficiency in the cell's ability to obtain food from itssurrounding. In a milliliter of culture fluid containing

522 A. K. Fok et al.

2000 P. multiiniavnticleatum cells and 2-13 X 10y beads(0'8^tm, the concentration and size used routinely inour laboratory), only 0-04% of the volume is occupiedby the cell mass and 0-06 % by the beads. This is basedon a volume of 2-04xl0s/im3 for a Paramecium cell(200 fim long X 50 nm wide and approximating the cellvolume as twice that for a frustrum of a cone with alarge equatorial radius of 25, a small end radius of10Jim, and a height of 100jUm) and a volume of0-268(.im* per bead. Thus, even at the bead concen-tration used, 99-9 % of the culture is occupied by fluid.The amount of fluid a cell has to filter to obtain onebead will be 469jUm3. Assuming that about 50 % of theDV volume is occupied by the beads, a DV of 20fim indiameter will contain 7817 beads. Thus to collect thisnumber of beads at 50% trapping efficiency, the cellwill need to filter 7-3 X 106 finr1 of fluid. At a formationrate of 1 DVmin" ' , a cell will need to filter enoughfluid equivalent to 35 times its own volume. If thetrapping efficiency should drop to 25 %, the cell wouldneed to filter a volume equal to 70 times its bodyvolume. From these calculations it is apparent that thecell's ability to filter particles must be relatively ef-ficient.

Using the above calculations, we can determine theapproximate number of 0'8^im beads the cell willaccumulate at the threshold level (8-3xlOb), the levelat which individual labelled DVs contain too few beadsto be scored accurately. If we assume the same filteringrate of 7-3xlO('jttm3 per DV at this low bead concen-tration, and use a fluid volume of l-2x 10 ^im3 bead"1,we can determine that 30 beads are filtered at 50 %efficiency. This result serves to explain why DVscontain too few beads to enable them to be scoredaccurately.

Bead sizes between 0-5 and 4-5 im or beads with anegative surface charge have little negative effect on theformation rate or the DV diameter. On the other hand,beads with a positive charge reduce the formation ratedrastically and to a lesser extent reduce the size of theDVs formed. These combined effects will result in adramatic reduction in the amount of discoidal vesiclemembrane needed to form new DVs in these cells. Whya positive charge should have this effect is not known,although other studies in our laboratory have shownthat cationized but not native ferritin is toxic toparamecia (Westcot el al. 1985) and that DV acidifi-cation is inhibited in cells pulsed with positivelycharged beads (unpublished observation).

The maximum bead diameter that can be ingested inP. multimicronucleatum appears to be about lOjUm,which may be the maximum size to which the vestibu-lum and/or buccal cavity can be stretched to accommo-date food particles. Finally, while these results showclearly that solid particles stimulate DV formation, thequestion of whether paramecia can form DVs in the

absence of any particulate matter remains unansweredand will be the subject of a subsequent publication.

This work was supported in part by NSF grants DCB84-02881 and 85-02212 and NIH RCMI grant RRO3061.B.C.S. was supported by the MARC program funded withNIH grant GM 07684-06A1.

References

ALLEN, R. D. (1974). Food vacuole membrane growthwith microtubule-associated membrane transport inParamecium. J. Cell Biol. 63, 904-922.

ALLEN, R. D. & FOK, A. K. (1980). Membrane recyclingand endocytosis in Paramecium confirmed byhorseradish peroxidase pulse-chase studies. J. Cell Sci.45, 131-145.

COHEN, J. N., CARREAU DE LOUBRESSE, N. & BEISSON, J.

(1984). Actin microfilaments in Paramecium: localizationand role in intracellular movements. Cell Molil. 4,443-468.

FENCHEL, T. (1980). Suspension feeding in ciliatedprotozoa: structure and function of feeding organellcs.Arch. Pmtisletik. 123, 239-260.

FOK, A. K. & ALLEN, R. D. (1979). Axenic Parameciumcaudatum. 1. Mass culture and structure. J. Hrotozool.26, 463-470.

FOK, A. K. & ALLEN, R. D. (1988). The lysosome system.In Paramecium (ed. H. D. Gortz), pp. 301-324. Berlin:Springer-Verlag.

FOK, A. K., AZADA, E. A., UENO, M. S. & ALLEN, R. D.

(1987). Phagosomal acidification in Paramecium: effectson lysosomal fusion. Eur.J. Cell Biol. 43, 412-420.

FOK, A. K., LEE, Y. & ALLEN, R. D. (1982). Thecorrelation of digestive vacuole pH and size with thedigestive cycle in Paramecium caudatum. J. Protosool.29, 409-414.

FOK, A. K., LEUNG, S. S.-K., CHUN, D. P. & ALLEN, R.

D. (1985). Modulation of the digestive lysosomal systemin Paramecium caudatum. II. Physiological effects ofcytochalasin B, colchicine, trifluoperazine and compound48/80. Eur. J. Cell Biol. 37, 27-34.

MULLER, M., ROHLICH, P. & TORO, I. (1965). Studies onfeeding and digestion in protozoa. VII. Ingestion ofpolystyrene latex particles and its early effects on acidphosphatase in Paramecium mullimicmiiucleatum andTetrahymeua. J. Pivtozool. 12, 27-34.

NlLSSON, J. R. (1972). Further studies on vacuoleformation in Tetrahymeua pyrifonnis GL. ('. r. Trav.Lab. Carlsberg 39, 83-110.

NlLSSON, J. R. (1979). Phagotrophy in Tetrahymeua. InBiochemistiy and Physiology of Pmtozoa, vol. 2 (ed. M.Levandowsky & S. H. Hutner), pp. 339-379. New York,London: Academic Press.

RASMUSSEN, L. (1976). Nutrient uptake in Telraliymenapyrifonnis. Carlsberg Res. Commun. 41, 143-167.

RICKETTS, T. R. (1971). Periodicity of endocytosis in

Phagosome formation in Paramecium 523

Tetrahymena pyrifonnis. Protoplasma 73, 387-3%. microtubule-based motors. .4. Rev. Cell Biol. 3, 347-378.SWANSON, J. A., YIRINEC, B. D. & SILVERSTEIN, S. C. WESTCOT, D. M., FOK, A. K. & ALLEN, R. S. (1985).

(1985). Phorbol esters and horseradish peroxidase Flow of membrane markers from the coated pits instimulate pinocytosis and redirect the flow of pinocytosed Paramecium caudalum. J. Cell Biol. 101, 52a.fluid in macrophages. J. Cell Biol. 100, 851-859.

VALE, R. D. (1987). Intracellular transport using {Received 10 March 19SS - Accepted 12 April I9SS)

524 A. K. Fok el al.