Distribution of tubuiin and actin through the cell division cycle of the fission yeast Schizosaccharomyces japonicus var. versatilis: a comparison with Schizosaccharomyces pombe CAKOLINE E. ALFA* and JEREMY S. HYAMS Department of Biology, University College London, Gower Street, London WC1E 6BT, UK * Author for correspondence Summary Changes in the distribution of microtubules and Whereas in S. pombe cytoplasmic microtubules are F-actin through the cell division cycle of the fission reinitiated from a pair of microtubule organizing yeast Schizosaccharomyces japonicus var. versatilis centres (MTOCs) at the cell equator, in S. japonicus were investigated by fluorescence microscopy. The they arise by extensive microtubule growth from the fluorescence images obtained with SL japonicus were spindle poles. Western blots of cell extracts enriched markedly superior to those previously reported for S. for tubuiin by DEAE-Sephadex chromatography pombe and revealed new details of cytoskeletal or- showed that, like S. pombe, S. japonicus contains two ganization in this important genus. As in S. pombe, a-tubulins and a single /J-tubulin. Whilst the al- and /J- F-actin in S. japonicus was present as a concen- tubulins from the two species comigrated on one- tration of 'dots' at the growing poles of interphase dimensional polyacrylamide gels, the a2 species were cells and as a filamentous equatorial ring directing electrophoretically distinct Although fundamental the deposition of the cytokinetic septum. The tran- differences clearly exist between the two species, S. sition between these two states occurred at late japonicus could prove to be a useful tool in basic anaphase, in contrast to the situation in S. pombe studies of fission yeast cell biology. where the appearance of the equatorial actin ring is tightly coupled to the early events of mitosis. During the course of cytokinesis in S. japonicus the actin ring constricted and broadened, suggesting that it is Key words: tubuiin, actin, cell division cycle, contractile. Microtubule organization in S. japonicus Schizosaccharomyces japonicus var. versatilis, also revealed interesting differences from S. pombe. Schizosaccharomyces pombe. Introduction Adams, 1984), the dimorphic yeast Candida albicans (Anderson and Soil, 1986; Barton and Gull, 1988), the The yeast cytoskeleton is amenable to investigation by a apiculate yeast Saccharomycodes ludwigii (Dankova et al. combination of experimental approaches that is 1988) and the fission yeast Schizosaccharomyces pombe unmatched in any other group of organisms. These include (Marks and Hyams, 1985; Marks et al. 1986, 1987; Hagan the molecular cloning of the genes encoding the major and Hyams, 1988; Hagan et al. 1990). cytoskeletal proteins (Gallwitz and Sures, 1980; Ng and As with all these species, the effectiveness of immuno- Abelson, 1980; Hiraoka et al. 1984; Toda et al. 1984; fluorescence in S. pombe is limited by the size and shape of Thomas et al. 1985; Schatz et al. 1986; Mertins and the cell, as well as by the presence of a cell wall, which Gallwitz, 1987; Watts et al. 1987); the identification of must be at least partially removed to allow entry of interacting proteins by suppressor analysis (Novick et al. antibody probes. In an attempt to overcome the first of 1989; Adams and Botstein, 1989); the isolation and charac- these, we have turned our attention to another fission terization of cytoskeletal proteins by classic biochemical yeast, Schizosaccharomyces japonicus var. versatilis, methods (Kilmartin, 1981; Greer and Schekman, 1982; which is larger and significantly better for cytological Drubin et al. 1988; Liu and Bretscher, 1988) and the use of observation than S. pombe (Robinow, 1981; for a review of inhibitors of cytoskeletal function (Quinlan et al. 1980; fission yeast cytology, see Robinow and Hyams, 1989; for a Walker, 1982; Guthrie and Wickner, 1988; Jacobs et al. recent review of the taxonomy of the genus Schizosacchar- 1988). In addition, the structural rearrangements of tubu- omyces, see Sipiczki, 1989). Here we report that, whilst iin and F-actin that accompany progress through the cell there is considerable conservation of cytoskeletal organiz- division cycle of a number of yeasts have been extensively ation amongst the fission yeasts, significant differences characterized by fluorescence microscopy. These include also occur. In general, fluorescence images obtained with the budding yeasts Saccharomyces cerevisiae (Adams and S. japonicus were markedly superior to those with S. Pringle, 1984) and Saccharomyces uvarum (Kilmartin and pombe and revealed new details of cytoskeletal organiz- Journal of Cell Science 96, 71-77 (1990) Printed in Great Britain © The Company of Biologists Limited 1990 71 ation in this important genus. S. japonicus should prove Tris-HCl, pH7.4, 0.15 M NaCl) containing 0.1% (w/v) SDS and extremely useful in a range of studies in fission yeast cell 20% (v/v) methanol, using a Bio-Rad Trans Blot Cell and a biology. constant voltage of 40 V for 12-16 h. Following transfer, non- specific protein binding to the filters was blocked by incubation in TBS containing 0.05% (v/v) Nonidet P-40 and 2.5mgml"1 BSA (bovine serum albumin), for 30 min. Filters were subsequently Materials and methods incubated in primary antibody as follows: YOLl/34, 1/1000; KMX-1 (anti-^-tubulin; Birkett et al. 1985), 1/260-1/1000 in Strains and media TNP buffer (TBS containing 0.25 % (v/v) Nonidet P-40, 0.25 % The wild-type strain of S. pombe, 972h~, and S. japonicus strain (v/v) Tween 20), for 12-15 h. They were then washed four times in NCYC 419 (from the British National Collection of Yeast Cul- TNP buffer and incubated in a 1/250 dilution of the appropriate tures) were maintained on yeast extract medium (YE: Sgl"1 horseradish peroxidase-conjugated secondary antibody (Sigma), yeast extract (Difco), SOgl"1 glucose) solidified with 20gl~1 in TNP, for 3-6 h. After two washes in TNP and two washes in Bacto agar (Difco). Liquid cultures of S. japonicus were grown in TBS, filters were developed in TBS containing O^mgrnl""1 YE at 30°C to a density of 106 to 2xlO6 cells ml"1 on a reciprocal 4-chloro-l-napthol and 0.008% (v/v) hydrogen peroxide. shaker. S. pombe was grown in EMMp minimal medium (Nurse, 1976) to a density of 4xlO6 to 6xlO6 cells ml"1. In both cases this represents mid-exponential phase growth. Results Indirect immunofluorescence microscopy Fluorescence microscopy of F-actin and DNA Cells were processed for immunofluorescence microscopy follow- A population of S. japonicus cells stained with anti-actin ing fixation with either 4% formaldehyde (for actin) or 3% antibody plus DAPI to localize nuclear DNA is shown in formaldehyde plus 0.2 % glutaraldehyde (for tubulin) as described Fig. 1A and B, respectively. For comparison, an equival- by Hagan and Hyams (1988). The monoclonal anti-tubulin anti- ent population of S. pombe, prepared in the same manner, body YOL1/34 (Kilmartin et al. 1982) and the monoclonal anti- is shown in Fig. 1C and D. Whilst sharing the same actin antibody N350 (Amersham) were used at dilutions of 1/100. bacilliform shape, the S. japonicus cells are considerably larger (in the range 7 to 20 /an x 6 /mi) than those of S. Preparation of protein extracts pombe (7 to 14/imx3/xm). The two populations also Extracts of S. pombe and S. japonicus soluble proteins were showed striking differences in their staining with DAPI, prepared and enriched for tubulin using the method of Kilmartin the nuclei of S. japonicus appearing larger and' more (1981). Cultures of 1.61 were grown to densities of 2xlO7 to diffuse than those of S. pombe, which gave a more compact, 3 x 107 cells ml"1 (S. pombe) or 4 x 106 cells ml"1 (S. japonicus) and brighter DAPI image (compare Fig. IB and D). In both harvested for 5 min at 1000g at 4°C, in a Beckman TJ-6 refriger- species, actin staining showed concentrations of punctate ated centrifuge. The cells were washed once in ice-cold PEM F-actin at the poles of interphase cells and at the equators buffer (100 mM Na-Pipes (piperazLne-iV-iV'-bis-(2-ethanesul- phonic acid)), pH6.9, lmM EGTA (l,2-di(2-aminoethoxy)ethane- of cells undergoing cytokinesis, regions that in S. pombe NfltN'rN'-tetrei-eicetic acid), lmM MgSO<) to which was added are known to be the major zones of cell wall deposition lmM DTT (dithiothreitol) and lmM GTP (guanosine 5' triphos- (Mitchison and Nurse, 1985; Marks and Hyams, 1986). phate, lithium salt) plus the following protease inhibitors: 1 mM Although superficially similar, a more careful examin- PMSF (phenylmethylsulphonyl fluoride); 40//gml"1 leupeptin ation of the S. japonicus images revealed a number of and 3/igml"1 pepstatin A. The pellet was then resuapended in an significant differences in the temporal control of F-actin equal volume of the same buffer and the resulting slurry mixed redistribution through the cell division cycle from that with ice-cold 500/an glass beads (Sigma), in a ratio of 1:1. The described previously in S. pombe (Marks and Hyams, cells were disrupted in a Bead Beater (Life Science Laboratories), by 1 min bursts of homogenization with intervening 1 min periods 1985). To illustrate this, a sequence of individual S. on ice. The apparatus was cooled by means of an ice—water jacket. japonicus cells depicting one complete cell cycle was assembled (Fig. 2). As with S. pombe, actin was initially The homogenate was centrifuged at 18 000gfor 15 min (MSE 21 High Speed Centrifuge) followed by 100 000 g for 60 min (MSE monopolar (Fig. 2,2) but, in the majority of cells, switched Pegasus 65), both at 4°C. The high-speed supernatant was made to a bipolar distribution (Fig.