Regulation of Cell Shape in Euglena Gracilis. Iii

y. Cell Set. 74, 219-237 (1985) 219 Printed in Great Britain © Company of Biologists Limited 1985

REGULATION OF CELL SHAPE IN EUGLENA GRACILIS. III. INVOLVEMENT OF STABLE MICROTUBULES

CAROLE L. LACHNEY AND THOMAS A. LONERGAN* Department of Biological Sciences, University of New Orleans, Lakefnmt, Netu Orleans, LA 70148, U.SA.

SUMMARY The role of cytoplasmic microtubules in a recently reported biological clock-controlled rhythm in cell shape of the alga Euglena gracilis (strain Z) was examined using indirect immunofluorescence microscopy. The resulting fluorescent patterns indicated that, unlike many other cell systems, Euglena cells apparently change from round to long to round cell shape without associated cytoplasmic microtubule assembly and disassembly. Instead, the different cell shapes were correlated with microtubule patterns, which suggested the movement of stable microtubules to accomplish cell shape changes. In live intact cells, these microtubules were demonstrated by immunofluorescence to be stable to lowered temperature and elevated intracellular Ca2"1" levels, treatments that are commonly used to depolymerize microtubules. In cells extracted in detergent at low temperature or in the presence of elevated Ca2+ levels, the fluorescent image of the microtubules was disrupted. Transmission electron microscopy confirmed the loss of one subset of pellicle microtubules. The difference in microtubule stability to these agents between live intact cells and cells extracted in detergent suggested the presence of a microtubule-stabilizing factor in live cells, which is released from the cell by extraction with detergent, thereby permitting microtubule depolymerization by Ca2+ or lowered temperature. The calmodulin antagonist trifluoperazine prevented the Ca2+- induced disruption of the fluorescent microtubule pattern in cells extracted in detergent. These results implied the involvement of calmodulin in the sensitivity to Ca2+ of the microtubules of cells extracted in detergent.

INTRODUCTION The determination of cell shape is thought to involve the specific arrangement of the microtubule component of the cytoskeleton. Changes in cell shape are thought to result from a reorganization of the microtubule cytoskeleton. Many studies have supported the conclusion that an increase in cellular asymmetry (anisometry) is accompanied by microtubule polymerization and a decrease in asymmetry, or cell rounding, is accompanied by microtubule depolymerization (Porter, 1966,1980). These studies typically employ microtubule-disrupting agents to remove the microtubule component of the cytoskeleton. Cell shape changes, such as the retraction of heliozoan axopods (Schliwa, 1976) or neuroblastoma neurites (Olmsted, 1981) in response to low temperature, or the rounding of normally pyriform Ochromonas cells (Brown & Bouck, 1973) as a result of exposure to colchicine or increased pressure,

•Author for correspondence. Key words: microtubules, cell shape, Euglena. 220 C. L. Lachney and T. A. Lonergan accompany the observed disruption of microtubules. The conclusion that the loss of microtubules is related to the loss of cellular asymmetry is strengthened by the observation that removal of the microtubule-disrupting agents permits the recovery of normal cell morphology as well as repolymerization of the microtubules. Only a few studies have been performed with naturally occurring shape changes, i.e. those not chemically or physically induced. The transition from round to flattened mouse 3T3 fibroblasts will occur if cells are permitted to settle onto a solid substrate. The microtubule cytoskeleton, which closely matches the changing cell shape, is observed to polymerize during the cell spreading event (Osborn & Weber, 1976). Mouse neuroblastoma cells begin to extend neurites when permitted to settle onto a solid substrate, and these neurites contain bundles of polymerizing microtubules (Olmsted, 1981; Solomon &Magendantz, 1981). In the colonial alga Volvox, daughter colonies invert, a process that is characterized by elongation of individual cells in the colony. Within the inverting cells are microtubules that are apparently synthesized as cell elongation takes place and lie parallel to the axis of elongation (Viamontes, Focht- mann & Kirk, 1979). A new system has been reported in which Euglena gracilis cells change shape between two extremes, round and long cells, every 24 h (Lonergan, 1983). The time scale (h) of the cell shape change makes it amenable to ultrastructural study. Colchicine inhibits the cell shape changes inEuglena (Lonergan, 1983), imply- ing the involvement of microtubules, but without microscopic examination of the microtubules this involvement was a matter of speculation. The present study was undertaken to examine the role of microtubules in these shape changes, which are at least superficially similar to the shape changes of metaboly (Guttman & Ziegler, 1974; Hofmann & Bouck, 1976), but slower. This article reports our characterization of the pellicle microtubules of E. gracilis by indirect immunofluorescence. We have found a correlation between cell shape and microtubule arrangement that, unlike the above systems, apparently did not involve the polymerization and depolymerization of microtubules. Examination of microtubule arrangements of different cell shapes suggested instead the movement of a stable population of microtubules by intracellular contractile forces. We further characterized the stability of the pellicle microtubules to commonly used microtubule-disrupting agents. In addition to stability to colchicine, previously reported by Silverman & Hikida (1976), we found Euglena microtubules to be stable to low temperature and elevated intracellular Ca2+ levels in live cells. Exposure of detergent- extracted cells to elevated Ca2+ levels or low temperature resulted in loss of microtubules, suggesting the existence of an extraction-labile stabilizing factor. Experiments with the calmodulin antagonist trifluoperazine have implicated calmodulin in the Ca2+ - induced depolymerization of microtubules in Euglena cells extracted in detergent.

MATERIALS AND METHODS Cell cultures Axenic cultures of Euglena gracilis (strain Z) were grown from samples of liquid stocks in 250 ml flasks containing 175 ml Cramer—Myers liquid culture medium (Cramer & Myers, 1952). Cultures Stable microtubules in Euglena 221 were aerated with filter-sterilized air at an approximate rate of 0-51 min"1, magnetically stirred, and maintained at 25°C. The illumination consisted of a 10-h light/14-h dark cycle at fluorescent light intensity of 300 /iEinsteins m~2 s"1.

Fluorescent antibody staining Samples for antibody staining consisted of 20 ml portions at a density of 0-75 X 10s to 1-0 X 105 cells/ml (late log phase), removed at the times of the extremes of the cell shape rhythm, 0900 standard time (ST, 9:00am, round cells, corresponding to lights-on of the illumination cycle), 1400 ST (2:00 pm, long cells, 5 h into the light period) or at times between the extremes of shape. In order to prevent cell shape changes, control cells or cells exposed to various experimental conditions were fixed in suspension at room temperature by slowly adding equal volumes of 6% (w/v) paraformaldehyde to the cell suspension while gently mixing on a vortex mixer, resulting in a final 3% (w/v) paraformaldehyde solution. Cells were collected by centrifugation in a Damon IEC clinical centrifuge at maximum setting for 30 s and gently resuspended in a second fixative consisting of 3% (w/v) paraformaldehyde in a microtubule-stabilizing buffer consisting of lOOmM-PIPES (pH 6-9), 1 mM-MgS04, 2mM-EGTA, 330 mM-sorbitol (Lloyd etal. 1979; Lloyd, Slabas, Powell & Lowe, 1980), for 10min. Cells were centrifuged and resuspended in microtubule-stabilizing buffer without sorbitol, containing 2-5% Triton X-100 (v/v), and extracted for 15min while agitating in a gyratory shaker (175 rev./min). All subsequent centrifugations were at 1600 rev./min (500 g), for 10 min in a Beckman TJ-6 tabletop centrifuge. After extraction, cells were centrifuged and resuspended in 5 ml 80% (v/v) acetone for 5 min while agitating in a gyratory shaker to remove chlorophyll, which causes an interfering red fluorescence. Cells were then centrifuged and resuspended twice for 5 min, while agitating in 5 ml microtubule buffer without sorbitol (to hasten extraction of soluble cytoplasmic contents), followed by three additional resuspensions (for 5 min) in microtubule buffer containing 0-lmM-glycine (pH 7-4) to reduce excess aldehydes. Cell pellets were then rinsed twice in microtubule buffer and centrifuged in a Beem capsule to obtain a 1 - 2/il pellet. These cells were incubated in 50 fA sheep anti-bovine brain tubulin (Caabco, Inc., Houston, TX, 225^g/ml antibody in 125 mM-borate and 75mM-NaCl, pH 8-4) for 45 min at 25°C. Cells were rinsed three times in 10 drops of microtubule buffer and resuspended in 50fil FITC- conjugated rabbit anti-sheep IgG (Miles-Yeda, Rehovot, Israel, 100-200 ^g/ml specific antibody in lOmM-phosphate and 75mM-NaCl, pH 7-4) 45 min at 25°C. Cells were rinsed three times in 10 drops of microtubule buffer, and resuspended for storage in 1 drop of microtubule buffer containing 1% (w/v) NaN3. Cells were stored refrigerated in covered Beem capsules. Cells were viewed by making wet mounts of approximately 1 /A cell suspension and 1 (A mounting medium consisting of 1% (w/v) />-phenylenediamine in glycerol at pH 8. Cells were viewed and photographed with an Olympus Model BHT microscope with BH-RFL-W epifluorescence illumination, using a 100X oil- immersion objective lens. Cells were photographed on Kodak Plus-X Pan film with 90 s manual film exposures using an Olympus OM-2 camera.

Lysed cell system The order of the fixation and extraction steps was reversed to expose extracted cells to the same concentrations and durations of microtubule-disrupting agents as administered in vivo. Samples (20 ml) of cell cultures were centrifuged and resuspended in the microrubule-stabilizing buffer without sorbitol but containing 2-5% Triton X-100. Various agents were added to this buffer: 180/m-Ca2+ (and no EGTA); or trifluoperazine plus 180/m-Ca2+ (and no EGTA). Cells were agitated for 15 min in a gyratory shaker, centrifuged and resuspended in buffer plus agent (no detergent) for 2-25 h. For low-temperature experiments, the extraction was at 1°C in buffer containing detergent plus EGTA for 15 min, then resuspension in buffer minus detergent at 1°C for 2-25 h. After each of the above procedures, cells were centrifuged and fixed in 3% paraformaldehyde/ microtubule buffer for 10 min, followed by 5 min in 80% acetone, after which they were processed as above for either indirect immunofluorescence or electron microscopy.

Transmission electron microscopy Specimens for electron microscopy were prepared as if for antibody staining, including all rinses, but antibodies were omitted. Control (prepared for antibody staining) or experimental (low 222 C. L. Lachney and T. A. Lonergan temperature, lysed cells, etc.) cells were resuspended in calf serum, and centrifuged, and the pellet was fixed in 4% glutaraldehyde in 0-1 M-cacodylate buffer (pH 7-4). Samples were post-fixed in 1% OsO4, dehydrated through an acetone/propylene oxide series, and embedded in an Epon- Araldite mixture. Silver sections were stained with uranyl acetate and lead citrate. Specimens were viewed and photographed at 80 kV with a Philips 200 transmission electron microscope. As a control, cells were removed directly from the culture and processed for electron microscopy as described above.

Analysis of experimental results The extent of microtubule disruption was estimated by systematic scanning of a microscopic field and scoring each cell as having intact fluorescent striations (microtubules), no visible striations, or partially intact striations. At least 400 cells were scored and the results of each experiment expressed as percentage striated, percentage non-striated, or percentage partially striated. The decision to declare Englena microtubules stable or labile to a particular microtubule-disrupting agent was based on the results of a 2 X 2 G-test for a significant difference ata= 0.05 between percentage striated cells of the particular treatment, and percentage striated cells of the appropriate control, as previously described (Lonergan, 1983).

Reagents All reagents were purchased from Sigma. Trifluoperazine was kindly supplied by Smith Kline and French Laboratories.

Fig. 1. Indirect immunofluorescence micrographs of Euglena cells showing antibody- labelled pellicle microtubules. Cells fixed in formaldehyde were extracted with Triton X- 100 and acetone, stained with sheep anti-bovine brain tubulin, followed by FITC- conjugated rabbit anti-sheep IgG. The microtubules of the submembrane pellicle, seen as bright fluorescent striations on the cell surface, wind helically around the cell. The flagellum at the cell anterior stains intensely but is out of the plane of focus in these micrographs. Pellicle microtubules are present in both round and long cell-shape extremes, as well as intermediate shapes, A. Round cell, typical of approximately 70% of cells in culture at dawn of the light/dark cycle; cell length 22fjm. B. Intermediate-shaped cell; length 37/im. c. Long cell, typical of approximately 70% of the cell in culture at midlight portion of the light/dark cycle; length, 48/im. X 1600; bar, 5 fim. Stable microtubules in Euglena 223

RESULTS Fluorescent antibody staining of the pellicle microtubules Indirect immunofluorescent staining of cells fixed in formaldehyde and extracted in detergent using FITC-labelled rabbit anti-sheep immunoglobulin G (IgG) direc- ted against sheep anti-tubulin resulted in regularly spaced fluorescent striations that represent antibody-labelled pellicle microtubules (Fig. 1). These striations were seen on both the upper and lower surfaces of the cell by focusing through the cell at x 1000 magnification. The flagellum was also intensely stained by the antibodies. Staining of Euglena cells at the times of the two extremes of shape of the cell shape rhythm (dawn and mid-light of the light—dark cycle) demonstrated the presence of the pellicle microtubules at both extremes of shape as well as all intermediate shapes. Fig. 1A shows a round cell, typical of most cells in the culture in the early light portion of the illumination cycle; c shows a long cell typical of the majority of cells in the middle of the light portion of the cycle; and B shows a cell shape typical of the populations intermediate between these two extremes (e.g. after 2h of light). The fluorescence pattern indicates that the arrangements of the pellicle microtubules in a general sense was similar for the full range of cell shapes, with the microtubules regularly spaced and winding helically around the surface of the cell. The specificity of the primary and secondary antibodies for E. gracilis microtubules was demonstrated, with controls in which each antibody was replaced by buffers or non-specific antibody (data not shown). Cells that had been fixed and extracted were

Fig. 2. Transmission electron micrographs of the pellicle ridges of Euglena. Thin sections through the ridged surface of the alga show groups of microtubules associated with each ridge, A. Section through a cell fixed in formaldehyde shows four microtubules associated with each ridge (1-4). B. Section through cell fixed and extracted as if for fluorescent- antibody staining shows loss of microtubules 2 and 3, and the persistence of microtubules 1 and 4.1-4, microtubules; er, endoplasmic reticulumjpm, plasma membrane;/), pellicle material. X 9500; bar 0-1 /im. 224 C. L. Lachney and T. A. Lonergan incubated m 50 [A borate-buffered saline (solvent for primary antibody) instead of primary and secondary antibodies. The resulting dim yellow fluorescence represents autofluorescence of cellular remains after extraction. In a second control, the primary antibody was replaced by borate-buffered saline, followed by fluorescent second antibody. There is an increase in fluorescence, representing non-specific staining of the second antibody, compared to the autofluorescence seen in the buffer-only control. For a third control, the primary antibody was replaced by 200^g/ml sheep IgG for the first antibody incubation, followed by the fluorescent second antibody. The resulting increase in fluorescence was not comparable to the specific staining pattern seen with anti-tubulin followed by second antibody. Transmission electron microscopy shows that the pellicle microtubules are associated with the characteristic ridges of the intracellular pellicle of E. gracilis. Thin sections through these ridges (Fig. 2A) revealed the presence of four microtubules associated with the protein pellicle material at the times of both extremes of cell shape. After preparation for antibody staining, two microtubules remained in each ridge, those that are labelled / and 4 in Fig. 2A and 2B. The other microtubules, 2 and J, were apparently removed during the extraction procedure.

A B

A3" 54 86V- >v V \ \h 84*7 f ~\ 26°\N J 1 35 45 Cell length (urn)

Fig. 3. Analysis of angles of microtubules in different cell shapes. The micrographs in Fig. 1 and 29 other micrographs of a series of cell lengths were traced and the angles of the microtubules to the anterior to posterior axis of the cell, defined by the position of the flagellum, were measured. A. Angles of microtubules of the round cell in Fig. 1A. Anterior to posterior axis is shown as a vertical line. Cell anteriors are uppermost in the diagram; cell length, 22/un; average of three angles, 83 °. B. Angles of microtubules of the intermediate-shaped cell in Fig. 1B; cell length, 37/un; average angle, 56°. c. Angles of microtubules of the long cell shown in Fig. lc; cell length, 48^an; average angle, 37°. D. Correlation of angles of microtubules to cell length. Three angle measurements per cell, similar to those shown in A through c, were averaged and plotted against cell length. As cell length increases, the average angle of the microtubules to the anterior to posterior axis of the cell decreases from nearly perpendicular in round cells to more acute angles in longer cells. Coefficient of correlation, r=-0-878. Stable microtubules in Euglena 225 These pellicle microtubules were seen as the specifically stained fluorescent striations in Fig. 1. Large samples of cells viewed at the times of the two extremes of shape contained the full range of cell shapes, with a larger proportion of round cells in the morning, and a larger proportion of long cells in the mid-afternoon. Intermediate-shaped cells were always present. The fact that the microtubules were present in virtually all cells at all times implies that the round-to-long and long-to-round shape transitions are accomplished without the complete disassembly and reassembly of the microtubules. Inspection of stained Euglena cells of different shapes revealed a pattern of microtubule organization that corresponded to cell shape: the angles of the microtubules in relation to the anterior to posterior axis of the cell changed with increasing cell elongation. Fig. 3 shows diagrammatically how the angles of the microtubules were measured from photomicrographs such as Fig. 1. The trend can be summarized as follows: in round cells the microtubules were nearly perpendicular (approx. 80-85°) to the anterior to posterior axis of the cell, in cells of intermediate shape the angles of the microtubules were intermediate (approx. 60-70°) between the two extremes, and in the longest cells the angles were much more acute (approx. 40°). A plot of average angle of the microtubules versus cell length is shown in Fig. 3D. The good correlation (r = —0*878) of angle with cell length reflects the consistent pattern we saw when inspecting cells of varying shapes.

Fig. 4. Immunofluorescent criteria for microtubule stability to various microtubule- disrupting treatments. Cells from various treatments were placed into one of the three following categories, A. Striated cells, with intact, intense fluorescent striations, presumably representing intact pellicle microtubules. B. Partially striated cells, in which the fluorescent striations are incomplete over half of the cell. Partial striations are interpreted as partially disrupted microtubules. c. Non-striated cell, showing no evidence of fluorescent striations on any surface of the cell. The absence of striations is interpreted as a loss of pellicle microtubules. X 1800; bar, 5/im. 226 C. L. Lachney and T. A. Lonergan

Stability of Euglena microtubules in live cells The effects of three microtubule-depolymerizing agents on the pellicle microtubules were examined. Fig. 4 illustrates the criteria for classifying cells as striated, partially striated, or non-striated, the measure by which we determined the stability or lability of the microtubules to various agents. We interpreted intense, intact fluorescent striations (Fig. 4a) as intact microtubules. Partial striations are shown in Fig. 4B. Close examination of micrographs of such cells shows that these interrupted striations are more or less continuous with intact striations. The larger, globular fluorescent spots, which tend to obscure the interrupted striations, were present in the non-specific IgG control and represent non-specific staining. There were always cells with partial striations in all treatments, possibly representing a loss of microtubules in the fixation and extraction for antibody staining. Cells showing no sign of fluorescent striations (Fig. 4c) were interpreted as having lost their microtubules. Live cells were exposed to 2*5 mM-colchicine, beginning at dawn of the light/dark cycle (the time of the greatest proportion of round cells) for 2-5 h, at the mid-light

Control

Fig. 5. Immunofluorescence and electron micrographs of colchicine-treated cells, A. Cell incubated in 2-5 mM-colchicine 21 h before preparation for antibody staining. Striations indicate intact microtubules (see Fig. 1, untreated cell). X 1500; bar, 5/wn. B. Stability of microtubules to colchicine treatment is quantitated by comparison of proportions of striated cells (representing intact microtubules) in colchicine-treated samples to untreated samples. For clarification of the descriptive classes see Fig. 4. (D) Striated; (•) partial; (•) non-striated, c. Thin sections through cells incubated in 2-5 mM-colchicine for 2-5 h show that microtubules I and 4 persist through the extraction procedure (compare with Fig 2B). X 95 000; bar, 01 urn. Stable microtubules in Euglena 227

Control

>•

Fig. 6. Immunofluorescence and electron micrographs of cells incubated at 1 °C. Live cells were maintained in growth medium in an ice bath at 1 °C for 2-5 h with normal illumination, aeration and stirring, before preparation for antibody staining. Lowered temperature results in rounding of virtually all cells, A. Fluorescent striations indicate stability of microtubules to lowered temperature (see Fig. 1, untreated cells). This cell is viewed from the anterior, where the pellicle strips and microtubules wind into the canal leading to the water reservoir. The flagellum, mostly out of focus, is partially visible as the rodlike structure in the centre of the microtubule whorl. X 1800; bar, 5 pm. B. Proportions of cells in the three descriptive classes defined in Fig. 4 in a typical low-temperature experiment. Symbols as in Fig. 5B. C. Section through cell incubated at low temperature, fixed, and extracted for antibody staining, shows microtubules / and 4 present, as in colchicine-treated (Fig. 5c) and untreated cells (Fig. 2B). X 95 000; bar, 0-1 fun. portion of the cycle (the time of the greatest proportion of long cells) for 2-5 h and for 21 h beginning at the middle of the light cycle. Fig. 5A shows that even after 21 h of exposure to colchicine the microtubules were present, as evidenced by the striated fluorescence pattern. Neither the arrangement, in terms of angles characteristic of cell length, the intensity of staining, nor the completeness of the striations, was affected by exposure of cells to colchicine. Fig. 5B quantitates the proportion of striated cells observed in a sample from a typical colchicine experiment and demonstrates that it was comparable to that of untreated cells: 93% of colchicine-treated cells are fully striated, compared to 82% in untreated cells. For the sample size this does not represent a statistically significant (P<0-05) difference. Electron microscopy confirmed that the two microtubules, J and 4, which persisted through the fluorescent- antibody procedure, were present after colchicine treatment (Fig. 5c). The effect of cold treatment was examined by maintaining Euglena cultures at 1°C for 2*5 h before preparation for antibody staining. Although virtually all cells became 228 C. L. Lachney and T. A. Lonergan spherical under these conditions, Fig. 6A shows that the microtubules remained intact, as shown by fluorescent striations of the same intensity and completeness as those in untreated cells. After cold treatment, 89% of the cells were striated, compared to 82% in non-treated cells (Fig. 6B). Electron microscopy confirmed that microtubules 1 and 4 were present (Fig. 6c), as they were in both colchicine-treated and untreated cells (compare with Figs 2, 5). Therefore, cell rounding induced by lowered temperature is apparently not the result of microtubule depolymerization in E. gracilis, unlike fibroblasts, heliozoans and numerous other cell types. The effect of raising Ca2+ concentration on microtubule stability was examined by exposing cells to Cramer-Myers growth medium (180/iM-Ca2+ containing the calcium ionophore A23187) for 2.5 h before fluorescent-antibody staining. Fig. 7A shows intact, intense fluorescent striations, indicating that the pellicle microtubules were stable to this level of intracellular Ca2+. A high proportion (66%) of the cells observed were striated (Fig. 6B). This does not represent a statistically significant (P< 0-05) difference compared to the control. The effects of alteration of intra- and extracellular levels of Ca2+ on cell shape in Euglena are the subject of another paper (Lonergan, 1984).

Lability of Euglena microtubules in lysed cells Fig. 2A shows that cells removed from the culture had four microtubules in each pellicle ridge. The procedure used for preparing the cells for antibody staining resulted in the loss of two microtubules, but two microtubules were consistently present (Fig. 2B). The same two microtubules apparently resisted depolymerization by lowered temperature (Fig. 6) and elevated intracellular Ca2+ levels (Fig. 7). To

Ca2+ Control 100 *=401 82% n = 380 75 66% -

•= 50 - 32% 17% If 2% /%j 1%

Fig. 7. Immunofluorescence micrograph of cell exposed to elevated intracellular Ca2+ levels. Live cells were incubated in 40/M-A23187 ionophore and 0.1 % (v/v) dimethyl- sulphoxide in culture medium containing 180 /iM-Ca2+ for 2-5 h with normal illumination, aeration and stirring before preparation for antibody staining, A. Intense, intact striations indicate stability of pellicle microtubules to these conditions (compare with Fig. 1, untreated cell). X 1600; bar, 5JOTI. B. Quantitation of the effects of elevated Ca2+ on microtubules, as determined by the presence of fluorescent striations. Symbols as in Fig. 5B. Stable microtubules in Euglena 229

Ca2+ lysed Ca2+ live n=455 n = 401 81 %| 66%

32% 15% 4%l 2% JZ] I MM

Fig. 8. Immunofluorescence and electron micrographs of cells extracted in the presence of 180/iM-Ca2+ before fixation, A. No vfluorescent striations are visible on any surface of the cell. The diffuse background fluorescence is of similar intensity as the autofluorescence (no antibody) controls. The surface fluorescence appears as spots that represent non- specific staining. At the right is the stump of the flagellum that has been apparently removed by this treatment. X 1900; bar, 5/zm. B. Comparison of the effects of 180[M- Ca2+ on the presence of fluorescent striations in intact (live) and detergent-extracted (lysed) cells exposed to 180 /Ai-Ca 2+. c. Electron microscopy shows that only microtubule 1 remains in most ridges (compare with Figs 2B, 5C, 6C; striated cells with microtubules 1 and 4 present). Microtubule 4 is usually absent or, if present, appears to be somewhat disrupted. Loss of fluorescent striations is associated with loss of microtubule 4. X 95 000; bar, 0-1 i&n. Symbols as in Fig. 5B. examine whether these two microtubules were stable to these agents when most of the cytoplasmic contents were removed, cells extracted in detergent were exposed to elevated levels of Ca2+ or lowered temperature for 2-5 h before fixation. The reversal of fixation and extraction steps without the addition of a microtubule-disrupting agent (lysed-cell control) resulted in 73% striated cells, compared to 82% striated cells in untreated, intact cells. This does not represent a statistically significant (/)<005) difference. Neither the angles of the striations, relative to cell shape, the intactness, nor the intensity of the striations was altered by reversal of these steps in these cells. When cells were extracted in 180/iM-Ca2+, fixed and stained, the striated fluorescence pattern of the microtubules was lost in a large proportion of the cells (Fig. 8). Fig. 8B compares the effect of this treatment on striations with the effect of elevated Ca2+ levels on intact cells. Only a small proportion (15%) remained striated, while the majority (81%) showed no evidence of striations. The difference in the percentage of striated cells between these two treatments represents a significant 230 C. L. Lachney and T. A. Lonergan

1°C lysed 3 live n=3% 89% n=455 o 74% _ 5 75- a- u * 50- >u

I 25- 14%12%^H ' 10% 1 ^^H

Fig. 9. Immunofluorescence and electron micrographs of cells extracted at 1 °C. A. The surface of this cell displays a diffuse surface fluorescence similar to that in Fig. 8A (cells extracted in elevated Ca*+ levels), instead of fluorescent striations. The flagella are not affected by this treatment. X 1 900; bar, 5 (Jm. B. Comparison of the effects of low temperature on the microtubules of intact (live) and detergent-extracted (lysed) cells shows microtubule sensitivity to low temperature in extracted but not intact cells. Symbols as in Fig. 5B. C. Electron microscopy shows that microtubule 4 is usually absent, while microtubule 1 is always present. X 95 000; bar, 0-1 /an.

(P<005) difference. The background fluorescence characteristic of the autofluorescence controls was present, along with a diffuse surface fluorescence where the striations would be, as well as the non-specific fluorescent spots. The flagellaappeare d to be removed by this treatment, and virtually all cells had the oval shape seen in Fig. 8A. Electron microscopy demonstrated that only microtubule 1 remained after the inverted fixation/extraction treatment, while microtubule 4 was severely disrupted or absent (Fig. 8c), as compared to the two microtubules, 1 and 4, seen in striated cells (compare Fig 2B, 5C and 6c). Extraction of cells at 1°C before fixation resulted in the same degree of microtubule disruption as with elevated Ca2+ levels, as can be seen in Fig 9. The surface striations were present in only 14% of the cells, compared to the 89% of cells that remained striated when intact cells were exposed to low temperature for the same length of time. The difference in percentage of striated cells between these two treatments represents a significant (P<005) difference. Microtubule 4 was lost, and microtubule 1 remained (Fig. 9c), as it did in cells extracted in elevated Caz+ levels (Fig. 8c). Unlike cells extracted in 180^M-Ca2+, the flagella remained attached to the cells, and elon- gated shapes were commonly seen (compare with Fig. 8A). Stable microtubules in Euglena 231

Stabilization of microtubules in lysed cells by trifiuoperazine The calmodulin inhibitor trifluoperazine (TFP) was included in extraction buffer containing 180^M-Ca2+ in an initial attempt to examine the possible involvement of calmodulin in the Ca2+-induced microtubule depolymerization observed in lysed cells. Fig. 10E shows that although 10 and 100 ^JM-TFP did not cause an appreciable change in the small proportion of striated cells, 500/iM-TFP stabilized the microtubules against the effect of Ca2+, as shown by the presence of fluorescent striations in 87% of the cells, which is significantly different (P<0-05) from the cells extracted in Ca2+ alone. The photographic series of Fig. 10A through c was prepared with strict uniformity of exposure to show a trend with increasing levels of TFP: as the concentration of TFP was increased from 0 to 10 to 100 to 500 jUM, an increase in the level of surface fluorescence was seen in the form of a hazy brightness of the cell surface. We interpreted these results as fluorescent-labelling of tubulin that is dispersed upon

180/iM-Ca24 Ca2+/i0/iM-TFP Ca27l00/m-TFP Ca2+/500^M-TFP

Fig. 10. Fluorescent images of cells extracted in increasing levels of the calmodulin antagonisttrifluoperazineinthepresenceofl80/*i-Ca2+. A. Cell extracted in 180/*M-Ca2+ showing no striations. B. Cell extracted in 180/iM-Ca2+ plus 10^M-TFP showing no striations but increased surface fluorescence, c. Cell extracted in 180/iM-Ca2+ plus lOO/M-TFP shows no striations but there is still more surface fluorescence, D. Cell extracted in 180/iM-Ca2+ plus 500/iM-TFP is striated. The micrographs A, B and c were printed with exactly the same exposure, both on the film recording and in printing, to demonstrate the dramatic increase in fluorescence seen in this series, x 1900; bar, 5 fan. E. Quantitation of the effects of raising the concentration of TFP in the Ca2"1" extraction buffer on the presence of fluorescent striations. 232 C. L. Lachney and T. A. Lonergan microtubule depolymerization. As the TFP concentration is increased, tubulin becomes more organized, until at 500 jUM TFP, the microtubules are stabilized. These results implicate calmodulin as a possible mediator of the Ca2+-induced microtubule depolymerization in the detergent-extracted (lysed) cells, since the sup- posed depolymerization is inhibitied by a calmodulin antagonist. Without specific visualization of calmodulin by immunofluorescence or isolation of calmodulin from Euglena cells, this evidence must be considered preliminary. Work is currently in progress in our laboratory to localize calmodulin by indirect immunofluorescence.

DISCUSSION Three types of shape change in the euglenoids have been reported. The shape change most widely described, metaboly or euglenoid motion, is a relatively rapid process (seconds), and has been characterized by light (Murray, 1981) and scanning electron (Guttman & Ziegler, 1974) microscopy. Possibly, a more rapid shape change is the shock response, where virtually all cells in a sample will quickly round up in response to a variety of physical and chemical stimuli (Bovee, 1982; Murray, 1981). The rhythmic transitions in cell shape that we have examined require several hours (Lonergan, 1983), making this the slowest of the three types of change. All three types of shape change have the same two extremes in cell shape, round and long, as well as similar shapes transitional between the two extremes. With these observations made, we propose that it is likely that the changes in microtubule patterns seen in the biological clock-controlled shape changes are similar or identical to those of both metaboly and the shock response. There are two possible schemes for the production of these three types of shape change. First, there may be one contractile mechanism operating on different time scales to produce, on the one hand, the very rapid shock response and the relatively rapid changes of metaboly and, on the other hand, the rhythmic distribution in cell shapes (lengths). This possibility is supported by the observation that the same cell shapes occur in the three shape changes. The second possibility, that two or more different contractile mechanisms produce the three shape changes, is suggested by the difference in time scale involved. The calcium-binding protein spasmin, involved in the contraction of many protozoa (Routledge & Amos, 1977), is characterized by a very rapid contraction, compared to that of actomyosin of skeletal muscle. It is possible that a rapidly contracting spasmin or spasmin-like system produces the shock response while an actomyosin system produces the metaboly and biological clock- controlled shape changes of Euglena. Both these systems would apparently have to exert force on the same structural elements, i.e. microtubules and pellicle strips, since the shapes produced in the three responses can be described in terms of the orientation of either of these elements. We are currently examining via immunofluorescence the distribution of actin and myosin in E. gracilis. Whatever forces are responsible for these shape changes in Euglena, certain generalizations can be made. The microtubules remain intact, apparently stable to Ca2+ and cold shock, unlike shape changes induced by chemical or physical treat- Stable microtubules in Euglena 233 merits in many other cell types, or naturally occurring shape changes involving new synthesis of microtubules, such as in Volvox (Viamontes et al. 1979) or 3T3 cells (Osborn & Weber, 1976). The contractile forces themselves are apparently sensitive to Caz+ and the demonstration by Murray (1981) that the cisternae of the endoplasmic reticulum in the pellicle strips contain Ca2+ suggests the likelihood that some or all of these shape changes are initiated by Ca2+ flux that is controlled by Ca2+-sequestering membranes. A very appealing hypothesis is that microtubules act as a stable, rigid cytoskeleton in a literal sense, with at least one cytomuscular system operating upon it to produce changes in cell shape. Such a scheme is consistent with the suggestions by Arnott & Smith (1969) and Guttman & Ziegler (1974) that the surface array of stable microtubules functions to return the cell to a non-random shape. We have further characterized the stability of Euglena pellicle microtubules by comparing the responses of intact and lysed cells to conventional microtubule- disrupting treatments. There are many examples of the range in stability of microtubules to various agents thought to cause depolymerization. These include the difference in stability to cold and colchicine of the two kinds of spindle microtubules (Brinkley & Cartwright, 1975), the cold- and colchicine-stable flagellar microtubules (Behnke & Forer, 1967), a subclass of neurotubules demonstrated by cycles of assembly—disassembly in vitro to be cold-stable but Ca2+-labile (Job, Fischer & Margolis, 1981; Webb & Wilson, 1980), and heliozoan axopodal microtubules, which depolymerize in response to drugs, pressure, low temperature and elevated Ca2+ levels (Schliwa, 1976). The pellicle microtubules of E. gracilis have previously been characterized as resistant to depolymerization by colchicine (Silverman & Hikida, 1976) and increased pressure (Hofmann & Bouck, 1976). We report that the microtubules are also resistant to depolymerization by lowered temperature and elevated intracellular Ca2+ levels, when live cells are treated before fixation or extraction. Several arguments can be offered in favour of a microtubule-stabilizing factor in intact Euglena cells. One line of reasoning suggesting a microtubule-stabilizing factor is the striking difference in stability of Euglena microtubules in intact versus detergent-extracted cells. The microtubules are apparently stable in live, intact cells to low temperature (Fig. 6) and elevated Ca2+ (Fig. 7), but are lost when cells are extracted in the presence of Ca2+ (Fig. 8) or low temperature. What apparently happens is the removal or disruption of a factor or circumstance in the intact cell, permitting microtubule depolymerization. In other cell types, various factors have been proposed to confer microtubule stability. In lysed kidney epithelial cells, microtubules depolymerize upon exposure to Ca2+ in the micromolar range. In these cells, microtubules are also stabilized to elevated Ca2+ levels by addition of exogenous microtubule-associated proteins or by alterations in the extraction protocol, such as the inclusion of polyethylene glycol, or the use of the detergent Brij-58 instead of Triton X-100 to stabilize microtubule- associated proteins (Schliwa, Euteneuer, Bulinski & Izant, 1981). A low molecular weight compound in the cold- but not Ca2+-stable subset of neurotubules has been implicated in tubule stability by Webb & Wilson (1980). The low molecular weight compound taxol has been found to promote microtubule assembly both in vitro 234 C. L. Lachney and T. A. Lonergan (Schiff, Fant & Horwitz, 1979) and in vivo (De Brabander et al. 1982), and to stabilize microtubules to the depolymerizing effects of low temperature (Schiff et al. 1979; Schiff & Horwitz, 1980) and elevated Ca2+ levels (Schiff etal. 1979). Although the usefulness of such a compound in the study of the behaviour of microtubules is considerable, we think that a question that is especially pertinent to our study is whether the microtubules of Taxus, the plant from which taxol is isolated (Wani et al. 1971), are stable to microtubule-disrupting treatments. The possibility exists that stable microtubules, and microtubule-stabilizing factors, are more the rule than the exception in eukaryotic cells. It seems reasonable that an over-riding microtubule-stabilizing factor could account for stability to drugs, elevated pressure, low temperature and elevated Ca2+ levels, agents that presumably operate by quite different mechanisms to depolymerize microtubules (Schliwaef al. 1981). The alternative explanation would be the existence of separate circumstances that stabilize the microtubules to each of these agents. Electron-microscopic examination of the intact versus lysed cells reveals another unusual observation. Within the Euglena pellicle there exists a range of microtubule stability. Microtubules 2 and 3 were lost sometime during the extraction procedure, preventing the drawing of firm conclusions about their stability to low temperature or elevated Caz+ levels (Fig. 2B). Silverman & Hikida (1976) demonstrated the stability of all the microtubules of the pellicle ridge to colchicine as well as Triton X- 100, the detergent used in this study for permeabilizing the cells. We have noted that the notch region of each ridge, where microtubules 2 and 3 are found, appears to be especially disorganized by the extraction procedure, possibly representing a weaker surface integrity in this region than elsewhere in the pellicle. In the two remaining microtubules, 1 and 4, we see another difference in stability. Both are stable to cold and Ca2+ in intact cells, but microtubule 4 is labile to both agents in lysed cells, whereas microtubule 1 is not (Figs 8c, 9c). Hofmann & Bouck (1976) indicated that microtubule 4 in all likelihood becomes microtubule 1 by intussusception in daughter pellicle strips during cell division. We think that this change in relative position is coincident with a change in supramolecular stability: when or shortly after microtubule 4 becomes microtubule 7 in a new strip, it acquires characteristics (stabilizing factors) that render it stable to cold and Ca2+ even in lysed cells. Why microtubule 1 does not appear as a fluorescent striation is yet another enigma (Figs 8c, 9c). Either the antibodies do not recognize this microtubule, or the microtubule is not accessible to the antibodies. Comparison of these results demonstrates that the indirect immunofluorescence technique is capable of resolving single microtubules. Microtubule I in each ridge is apparently not visible as a fluorescent striation (Figs 8c, 9c), while the presence of microtubules / and 4 is seen as a striation (Figs 2B, 5C, 6C). Therefore, the single microtubule 4 is apparently responsible for each fluorescent striation. An alternative interpretation of these results is that at least two fluorescently labelled microtubules in close proximity are required for the labelling to become visible. Since Osborn, Webster & Weber (1978) have elegantly and directly demonstrated that individual Stable microtubules in Euglena 235 microtubules are resolvable by indirect immunofluorescence microscopy, we chose the former interpretation as the simpler of the two alternatives. The reversal of the Ca2+-induced loss of striations in lysed cells by trifluoperazine implies the presence and involvement of calmodulin in Euglena microtubule stability. Likewise, the increase in surface fluorescence with increase in trifluoperazine suggests a dose-dependent dispersion of tubulin. The reversal of what is presumed to be Ca2+-induced microtubule depolymerization by the calmodulin antagonist trifluoperazine is similar to that reported for lysed kidney epithelial cells (Schliwa, 1980) .The reported existence of calmodulin in Euglena, as reviewed by Klee, Crouch & Richman (1980), strengthens the preliminary conclusion that calmodulin mediates the observed microtubule depolymerization. A role for the proposed stabilizing factor might be suggested in preventing the interaction of Ca2+and calmodulin. What is the significance of microtubule stability? Our results suggested to us the distinction between stability of the microtubules to applied agents (colchicine, Ca2+, etc.) and temporal stability of the cytoplasmic microtubule complex. We propose that whether a cell type possesses a stable cytoplasmic microtubule complex determining its cell shape can be partially predicted from an examination of the cell's normal environment, but that the occurrence of stability to applied agents is not yet explic- able. Microtubule arrays in mammalian fibroblasts (Osborn & Weber, 1976) and amoebae of the slime mouldDictyostelium (White, Tolbert & Katz, 1983) are remark- ably similar in appearance, as revealed in particular by immunofluorescence microscopy. Both types of cell creep along a solid substrate. We assume this creeping motility demands a labile cell shape. The cytoplasmic microtubule complex of these cells is apparently labile in response to these cell shape demands, as evidenced by the fact that the microtubule arrays closely correspond to cell shape, which is observably labile. In contrast, the cytoplasmic microtubule complexes of Euglena (present study), Ochromonas (Bouck & Brown, 1973), various trypanosomes (Angelopoulos, 1970), and nucleated erythrocytes (Fawcett, 1981) are apparently stable, at least throughout interphase. All these cell types inhabit fluid environments. We assume that motility in fluids relies on maintenance of a specific, appropriate cell shape, and we think it is also safe to assume that the maintenance of these shapes is provided at least in part by microtubules. While some cells, seemingly predictably, possess cytoplasmic microtubule complexes that are stable during cell shape changes, there is no predictable relationship as to which organisms would possess microtubules stable to the typical depolymerizing agents. For example, Ochromonas microtubules, which are apparently present throughout interphase, are sensitive to colchicine and increased pressure (Brown & Bouck, 1973). Dictyostelium microtubules, on the other hand, polymerize and depolymerize during crawling, but are somewhat stable to cold (White et al. 1983). We propose that temporal stability of the cytoplasmic microtubule complex is more significant to cytoskeletal functioning and the determination of cell shape than chemical or mechanical stability of the microtubules to treatments never encoun- tered in.the day-to-day existence of these cells. The stability of microtubules to such 236 C. L. Lachney and T. A. Lonergan depolymerizing treatments may well be biologically incidental to the functioning of the cell.

We thank Dr Sherwood Githens III for helpful criticisms during the preparation of this manuscript. This work represents a portion of a thesis submitted by C.L.L. to the Graduate School of the University of New Orleans in partial fulfilment of the requirements for the degree of Master of Science. Portions of this work were presented at the N.I.H. Fogarty International Conference, 'The Cytoplasmic Matrix and the Integration of Cellular Function', Bethesda, MD, October, 1983.

REFERENCES ANGELOPOULOS, E. (1970). Pellicular microtubules in the Family Trypanosomatidae. J. Protozool. 17, 39-51. ARNOTT, H. J. & SMITH, H. E. (1969). Analysis of microtubule structure in Euglena granulata. J. Phycol. 5, 68-75. BEHNKE, 0. & FORER, A. (1967). Evidence for four classes of microtubules in individual cells. J. Cell Sci. 2, 169-192. BOUCK, G. & BROWN, D. (1973). Microtubule biogenesis and cell shape in Ochromonas. I. The distribution of cytoplasmic and mitotic microtubulea. J. Cell Biol. 5, 340—359. BOVEE, E. C. (1982). Movement and locomotion in Euglena. In The Biology of Euglena, vol. Ill, Physiology (ed. D. E. Buetow,), pp. 143-168. New York: Academic Press. BRINKLEY, B. R. & CARTWRIGHT, J. JR (1975). Cold-labile and cold-stable microtubules in the mitotic spindle of mammalian cells. Ann. New York Acad. Sci. 253, 428—439. BROWN, D. L. & BOUCK, G. B. (1973). Microtubule biogenesis and cell shape in Ochromonas. II. The role of nucleating sites in shape development. J. Cell Biol. 56, 360-378. CRAMER, H. & MYERS, J. (1952). Growth and photosynthetic characteristics of Euglena gracilis. Arch. Mikrobiol. 17. 384-402. DE BRABANDER, M., GEUENS, G., NUYDENS, R., WILLEBRORDS, R. & DE MEY, J. (1982). Microtubule stability and assembly in living cells: The influence of metabolic inhibitors, taxol, and pH. Cold Spring Harbor Symp. quant. Biol. 66, 227-240. FAWCETT, D. (1981). The Cell, pp. 768-771. Philadelphia: W. B. Saunders. GUTTMAN, H. N. & ZIEGLER, H. (1974). Clarification of structures related to function in Euglena gracilis. Cytobiologie 9, 10-22. HOFMANN, C. & BOUCK, G. B. (1976). Immunological and structural evidence for patterned intussusceptive surface growth in a unicellular organism. .7. Cell Biol. 69, 693-715. JOB, D., FISCHER, E. H. &MARGOLIS, R. L. (1981). Rapid disassembly of cold-stable microtubules by calmodulin. Proc. natn. Acad. Sci. U.SA. 78, 4679-4682. KLEE, C, CROUCH, T. & RICHMAN, P. (1980.) Calmodulin. A. Rev. Biochem. 49, 489-515. LLOYD, C. W., SLABAS, A. R., POWELL, A. J. & LOWE, S. B. (1980). Microtubules, protoplasts and plant cell shape. Planta 147, 500-506. LLOYD, C. W., SLABAS, A. R., POWELL, A. J., MACDONALD, G. & BADLEY, R. A. (1979). Cytoplasmic microtubules of higher plant cells visualised with antitubulin antibodies. Nature, Lond. 279,239-241. LONERGAN, T. A. (1983). Regulation of cell shape in Euglena gracilis. I. Involvement of the biological clock, respiration, photosynthesis, and cytoskeleton. PI. Physiol. 71, 719-730. LONERGAN, T. A. (1984). Regulation of cell shape in Euglena gracilis. II. The effects of altered extra- and intracellular Ca2+ concentrations and the effect of calmodulin antagonists. J. Cell Sci. 71, 37-50. MURRAY, J. M. (1981). Control of cell shape by calcium in the Euglenophyceae.J. Cell Sci. 49, 99-117 OLMSTED, J. B. (1981). Tubulin pools in differentiating neuroblastoma cells. J. Cell Biol. 89, 418-423. OSBORN, M. & WEBER, K. (1976). Tubulin-specific antibody and the expression of microtubules in 3T3 cells after attachment to a substratum. Expl Cell Res. 103, 331-340. Stable microtubules in Euglena 237 OSBORN, M., WEBSTER, R. & WEBER, K. (1978). Individual microtubules viewed by immunofluorescence and electron microscopy in the same PtK.2 cell. J. Cell Biol. 77, R27-34. PORTER, K. R. (1966). Cytoplasmic microtubules and their functions. In Principles oj' Biomolecular Organization (ed. G. E. W. Wolstenholme&M. O'Connor), pp. 308-356. Boston: Little, Brown and Co. PORTER, K. R. (1980). Some notes on the early characterization of microtubules. In Microtubules and Microtubule Inhibitors (ed. M. De Brabander & J. De Mey), pp. 555-568. Amsterdam: Elsevier/North Holland Biomedical Press. ROUTLEDGE, L. M. & AMOS, W. B. (1977). Calcium-binding contractile proteins in protozoa. In Calcium-Binding Proteins and Calcium Function (ed. R. H. Wasserman et al.), pp. 439—453. Amsterdam: Elsevier/North Holland. SCHIFF, P., FANT, J. & HORWWTZ, S. (1979). Promotion of microtubule assembly in vitro by taxol. Nature, Land. 277, 665-667. SCHIFF, P. & HORWITZ, S. (1980). Taxol stabilizes microtubules in mouse fibroblast cells. Proc. natn.Acad. Sci. U.SA. 77, 1561-1565. SCHLIWA, M. (1976). The role of divalent cations in the regulation of microtubule assembly. J. Cell Biol. 70, 527-540. SCHLIWA, M. (1980). Pharmacological evidence for an involvement of calmodulin in calcium- induced microtubule disassembly in lysed tissue culture cells. In Microtubules and Microtubule Inhibitors (ed. M. De Brabander& J. DeMey), pp. 57-70. Amsterdam: Elsevier/North-Holland Biomedical Press. SCHLIWA, M., EUTENEUER, U., BULINSKI, J. C. & IZANT, J. G. (1981). Calcium lability of cytoplasmic microtubules and its modulation by microtubule-associated proteins. Proc. natn. Acad. Sci. U.SA. 78, 1037-1041. SILVERMAN, H. & HIKIDA, R. S. (1976). Pellicle complex of Euglena gracilis: Characterization by disruptive treatments. Protoplasma 87, 237-252. SOLOMON, F. & MAGENDANTZ, M. (1981). Cytochalasin separates microtubule disassembly from loss of asymmetric morphology. J. Cell Biol. 89, 157-161. VIAMONTES, G. I., FOCHTMANN, L. J. & KIRK D. L. (1979). Morphogenesis in Volvox. Analysis of critical variables. Cell 17, 537-550. WANI, M., TAYLOR, H., WALL, M., COGGON, P. & MCPHAIL, A. (1971). Plant antitumor agents. VI. The isolation and structure of taxol, a novel antileukemic and antitumor agent from Taxus brevifolia.J. Am. Chetn. Soc. 93, 2325-2327. WEBB, B. C. & WILSON L. (1980). Cold-stable microtubules from brain. Biochemistry 19, 1993-2001. WHITE, E., TOLBERT, E. & KATZ, E. (1983). Identification of tubulin inDictyostelium discoideum: Characterization of some unique properties. ,7. Cell Biol. 97, 1011—1019.

{Received 29 June 1984 -Accepted, in revised form, 11 October 1984)