ABSTRACT
The effect of dimethylsulphoxide (DMSO) on the organization of the microtubular system of the mouse oocyte has been examined. Exposure to DMSO causes the immediate appearance of multiple, cold-resistant microtubular asters associated with the foci of peri-centriolar material (PCM) normally present in the oocyte. More prolonged exposure to DMSO leads to progressive disassembly of the spindle, and as a result dispersal of the chromosomes and polar PCM foci occurs, and tubulin polymerization becomes confined to PCM-organized asters. Those astral microtubules located between the PCM foci and the cortex of the oocyte appear to be particularly stable, resulting in the development of lengthening radial bundles of microtubules between the PCM and the surface and the progressive movement of the PCM′foci towards the centre of the cell. In contrast, after activation of the oocyte the microtubules generated in the presence of DMSO remain located in a cortical mesh. The effects of DMSO do not appear to be fully reversible in most oocytes. We discuss the implications of these results both for the cytoplasmic organization of the oocyte and zygote, and for the attempts at cryopreservation of human oocytes for therapeutic use in infertility programmes.
Introduction
The ovulated oocyte of the mouse is arrested in second meiotic metaphase and its microtubules are located exclusively in the spindle, which lies eccentrically in the cell, adjacent and parallel to the surface. The spindle is atypical in that it is barrel shaped, anastral and lacks organized centrioles (Fig. 1A,B; Szollosi, Calarco & Donahue, 1972; Wassarman & Fujiwara, 1978; Calarco-Gillam, Siebert, Hubble, Mitchison & Kirschner, 1983; reviewed Karsenti & Maro, 1986). Whilst no centriolar structures are evident in the oocyte, the peri-centriolar material (PCM), which in most other cells associates with the centriole to form the centrosome, is present in two beaded rings which form the ‘rims’ at each end of the barrel-shaped spindle (Fig. 1C); additional PCM foci are dispersed throughout the cytocortex (Calarco-Gillam et al. 1983; Maro, Howlett & Webb, 1985; Maro, Johnson, Webb & Flach, 1986; Schatten, Schatten, Mazia, Balczon & Simerly, 1986). Although in the unfertilized oocyte some of the PCM is located at the spindle poles, it cannot alone support the polymerization of microtubules, thus accounting for the absence of both polar asters and cytocortical asters. Rather, in the oocyte it is the meiotic chromo-somes which modify the local conditions to favour microtubule polymerization. Thus, if endogenous chromosomes are dispersed within the oocyte (Maro et al. 1986; Pickering & Johnson, 1987) or exogenous chromosomes introduced into the oocyte (Van Blerkom & Bell, 1986), each group of chromosomes favours the polymerization of microtubules locally. This chromosome-microtubule complex recruits adjacent PCM, which appears to have a role in the spatial ordering of the spindle and in the parallel alignment of the microtubules (Karsenti & Maro, 1986; Maro et al. 1986; Pickering & Johnson, 1987). However, the termination of meiosis and the progress of the activated oocyte to interphase, is associated with a change in the internal milieu of the cell that favours the polymerization of microtubules nucleated on the PCM located around the cell cortex and the formation of a cortical mesh of microtubules (Maro et al. 1985; Schatten, Simerly & Schatten, 1985; Schatten et al. 1986). This change, the nature of which is unclear, amounts to a decrease in the threshold or critical concentration of free tubulin required for polymerization.
Oocytes examined for the distribution of chromatin (B,F,H,J,L), tubulin (A,D,E,G,I) or PCM (C,K). (A) Tubulin distribution in unfertilized mouse oocyte; note the barrel-shaped anastral spindle with the chromosomes on the metaphase plate (B). (C) The distribution of polar PCM at each end of the spindle and of cytocortical foci of PCM viewed here en face and arrowed. (D-F) An oocyte exposed to 1·5M-DMSO at 4°C and then fixed within 10 min; note that a barrel-shaped spindle is retained, but that it is now astral, particularly so in E, and that multiple asters have also formed around the cortex. (G,H) An oocyte taken in 0·25 M steps to 1·5M-DMSO at 4 °C and then fixed; note the more central position of the mesh of microtubule asters within the cell compared with D and E. (I,J) An oocyte taken directly to 1·5M-DMSO at 37°C and then fixed; during preparation for microscopy, the oocyte lysed releasing the microtubular mesh from the cell as a single unit. (K,L) PCM distribution in an oocyte incubated directly in DMSO at 37°C; note the relatively central location of the PCM, the halos of microtubules that surround each PCM focus and the dispersing chromosomes. (A,B,D-J, X400; C,K,L, x630.)
Oocytes examined for the distribution of chromatin (B,F,H,J,L), tubulin (A,D,E,G,I) or PCM (C,K). (A) Tubulin distribution in unfertilized mouse oocyte; note the barrel-shaped anastral spindle with the chromosomes on the metaphase plate (B). (C) The distribution of polar PCM at each end of the spindle and of cytocortical foci of PCM viewed here en face and arrowed. (D-F) An oocyte exposed to 1·5M-DMSO at 4°C and then fixed within 10 min; note that a barrel-shaped spindle is retained, but that it is now astral, particularly so in E, and that multiple asters have also formed around the cortex. (G,H) An oocyte taken in 0·25 M steps to 1·5M-DMSO at 4 °C and then fixed; note the more central position of the mesh of microtubule asters within the cell compared with D and E. (I,J) An oocyte taken directly to 1·5M-DMSO at 37°C and then fixed; during preparation for microscopy, the oocyte lysed releasing the microtubular mesh from the cell as a single unit. (K,L) PCM distribution in an oocyte incubated directly in DMSO at 37°C; note the relatively central location of the PCM, the halos of microtubules that surround each PCM focus and the dispersing chromosomes. (A,B,D-J, X400; C,K,L, x630.)
It has proved possible to induce atypical polymerization of tubulin around PCM foci in unfertilized mouse oocytes in three ways. First, addition of taxol, a drug which stabilizes microtubules, induces aster formation in association with the PCM at the poles of the spindle and with the foci of PCM in the cytocortex (Maro et al. 1985, 1986). Second, if the spindle is dismantled partially by cooling to room temperature, it seems that the rise in free tubulin levels so generated exceeds the threshold for polymerization of PCM-nucleated microtubules, and asters form cytocortically as well as at the poles of the disas-sembling spindle (Pickering & Johnson, 1987). Third, the spindle may be dismantled partially or completely by exposure to nocodazole (which binds to and thereby sequesters free tubulin so reducing the effective free tubulin concentration); on removal of the drug, there is a transiently elevated level of free tubulin and PCM-nucleated asters form until the preferential polymerization around the chromosomes has again resequestered the tubulin within the spindle (Pickering & Johnson, 1987). In this paper, we report on a fourth way of inducing asters in unfertilized oocytes. Dimethylsulphoxide (DMSO) is used extensively as a cryoprotectant for many cells including oocytes and zygotes (Schneider, 1986; Wilmut, 1986). We find that it also has profound effects on the microtubular system of the oocyte. Our results there-fore have implications not only for an understanding of how the organization of oocyte microtubules is regulated, but may also have relevance to the practice of oocyte cryopreservation.
Materials and methods
(A) Egg recovery
Oocytes were recovered from 3-to 5-week-old (C57B1.10× CBA)F1 or MF1 mice after superovulation with 5i.u. pregnant mares’ serum gonadotrophin (Folligon, Intervet) followed 48 h later by 5 i.u. human chorionic gonadotrophin (hCG; Chorulon, Intervet). The females were killed 12·5 h post-hCG and the oviducts placed in warm phosphate-buffered saline (PBS). Oocytes were released into drops of warm Medium 2 containing 4mgml−1 bovine serum albumin (M2+BSA; Fulton & Whittingham, 1978). Cumulus cells were removed by brief exposure to 0·1 M-hyaluronidase (Sigma) and stored in drops of M2+BSA at 37°C for a maximum of 45 min prior to use. For in vitro fertilization, oocytes were recovered in the same way and then inseminated as described in detail in Maro, Johnson, Pickering & Flach (1984).
(B) Incubation conditions
Oocytes were incubated in M2+BSA in glass cavity blocks containing the appropriate DMSO concentration and pre-equilibrated at the appropriate temperature in a water bath. Nocodazole (10 mM stock in DMSO at —20°C; Lot T-4–112, NIH) was used at 10 μM.
(C) Immunocytochemical labelling
Oocytes were retained at the appropriate temperature, at the appropriate DMSO concentration and/or in the appropriate drugs throughout processing to fixation. Removal of the zona pellucida was accomplished by brief exposure to acid Tyrodes’ solution (Nicolson, Yanagimachi & Yanagimachi, 1975). Oocytes were carried through processing in the specially designed chambers as illustrated and described in Maro et al. (1984). For microtubule demonstration, cells were fixed at 37°C for 30–45 min in 2·0% formaldehyde (BDH) in PBS, washed in PBS, extracted for 10 min in 0·25 % Triton-X 100 (Sigma) and finally washed in PBS (Maro et al. 1985). The numbers of cells examined are indicated in Tables 1·5. An additional 445 oocytes were examined for MTOC distribution under different conditions; cells were extracted for 2 min at 30°C in HPEM buffer (Maro et al. 1985) containing 0·1% Triton X-100, washed in HPEM buffer and fixed for 30 min with 2·0% formaldehyde in HPEM buffer. Tubulin was visualized with a monoclonal anti-alpha-tubulin antibody (YL]/2; Kilmartin, Wright & Milstein, 1982) followed by FTTC anti-rat IgG. PCM was visualized by a human anti-PCM serum (Calarco-Gillam et al. 1983) followed by FITC anti-human IgG. Demonstration of chromosomes was achieved by incubating fixed cells in Hoechst dye 33258 (5μgml−1 in PBS) for 20 min at 20°C.
(D) Microscopy
Samples were mounted in Citifluor (City University, London) and viewed on a Leitz Ortholux II with filter sets L2 for fluorescein-labelled reagents and A for Hoechst dye. Photographs were taken on Kodak Tri-X film using a Leitz Vario-Orthomatic photographic system. Due to the large size of the oocytes optical sections are illustrated in the Figures.
Results
DMSO is used as a cryoprotectant at a concentration of around 1·5m. On addition to cells, DMSO acts osmotically to cause transient shrinkage until equilibration of intracellular and extracellular levels of DMSO is achieved. The maximum degree of shrinkage can be reduced by adding the DMSO in 0·25 M increments of 10min each. Both direct and incremental addition of DMSO were used in these experiments. Potentially more serious is the osmotic expansion of cells that accompanies removal of DMSO, since this can lead to cell lysis. This danger can be eliminated by incremental removal of DMSO in 0·25 M steps of 10 min each, a procedure undertaken in all experiments reported here.
(A) Early effects of DMSO addition
Direct addition of oocytes to 1·5M-DMSO at 37°C, followed by fixation for analysis within 10–20min, yielded three striking changes (Table 1, compare lines 1 and 2; Fig. 1D-J). First, the barrel shape of the spindle was retained, but the poles became astral indicating nucleating activity by the polar foci of PCM (most clearly seen in Fig. ID in which the astral poles are relatively poorly developed compared with Fig. 1E,G,I). Second, multiple cytoplasmic microtubular asters formed at other PCM foci not associated with the spindle poles (Fig. 1D,E,G,I). Third, in most oocytes these multiple asters were concentrated in the centre of the oocyte (Fig. 1G) and cosegregated there with the normally cytocortical foci of PCM (Fig. 1K,L). The asters (both polar and cytoplasmic) formed the basis of a network of micro-tubules which, in some cells that lysed during processing, was extruded from the cell as an intact single unit (Fig. 11,J). An almost identical picture was observed in oocytes that were added to DMSO at 37 °C and then immediately transferred to 4°C for lh of incubation (Table 1, line 3), and in oocytes transferred directly to DMSO at 4°C followed by analysis within 10-20 min (Table 1, line 4). The only difference between direct addition of oocytes to DMSO at 4°C and direct addition at 37 °C (with or without subsequent cooling) was that at 4°C the mesh of asters was less well developed in about 70 % of the oocytes, and in these oocytes was located cortically not centrally (compare Fig. IE,F at 4°C with Fig. 1G,H at 37°C). In contrast to these results, exposure of oocytes to 4 °C in the absence of DMSO resulted in rapid net depolymerization and loss of all or most microtubules (Table 1, line 5; Pickering & Johnson, 1987). We conclude that 1·5M-DMSO favours the polymerization of tubulin nucleated by PCM and that this polymerization can even proceed at low temperatures.
(B) More prolonged exposure to DMSO
More prolonged exposure to DMSO was achieved in several ways, but in all cases very similar results were seen. Thus, incremental addition of DMSO (0-25 M steps of 10 min each totalling 60 min of increasing DMSO exposure) at either 37°C (Table 2, line 1) or 4°C (Table 2, line 2), incremental addition to 1·5 m-DMSO at 37°C followed by an incubation of 1h at either 37 °C (Table 2, line 3) or 4 °C (Table 2, line 4), or direct addition at 37°C followed by lh at 37°C (Table 2, line 5) all resulted in a different picture from that observed following immediate analysis. Either no spindle, or just a spindle remnant, was evident, the remnant being detected most readily by its residual poles and their relationship to the chromosomal cluster. Usually, these poles were incorporated at one side of a tightly knit central cluster of asters associated with PCM (Fig. 2A,B). From this central cluster, bundles of microtubules radiated out to the surface of the cell (Fig. 2A,C). In those oocytes exposed to DMSO at 37 °C during some part of the protocol, but not in those exposed at 4°C throughout (Table 2, line 2), loss of an identifiable spindle was often associated with the dispersal of chromosomes from the equatorial plate (Fig. 1L). Usually the dispersal was not widespread, but rather the chromosomes appeared to have been scattered locally in both the cytocortex and the more central regions of the oocyte. The chromosomes were not associated obviously with microtubules, tending only to be bypassed by the radiating microtubules between the central asters and the surface.
Oocytes examined for the distribution of chromatin (B,E,G), tubulin (A,C,D) and PCM (F,H). (A-C) The distribution of microtubules and chromosomes in oocytes incubated in DMSO for 1 h at 37°C; note the absence of a spindle and the central position of the asters, which are connected to the surface by long strands of microtubule bundles. (D,E) Oocytes incubated in DMSO plus nocodazole; note that the DMSO has provided some stabilization of microtubules, but that the spindle is shorter than in control oocytes; however, the nocodazole has prevented the nucleation of asters by the DMSO. (F-H) Similar to D and E but showing distribution of PCM; the polar PCM remains but is closer to the equator due to the shortening of the spindle, whilst the cortical PCM (viewed here en face) remains cortical and has not been driven towards the centre of the oocyte. (A,B,D-G, ×400; C,H, ×630).
Oocytes examined for the distribution of chromatin (B,E,G), tubulin (A,C,D) and PCM (F,H). (A-C) The distribution of microtubules and chromosomes in oocytes incubated in DMSO for 1 h at 37°C; note the absence of a spindle and the central position of the asters, which are connected to the surface by long strands of microtubule bundles. (D,E) Oocytes incubated in DMSO plus nocodazole; note that the DMSO has provided some stabilization of microtubules, but that the spindle is shorter than in control oocytes; however, the nocodazole has prevented the nucleation of asters by the DMSO. (F-H) Similar to D and E but showing distribution of PCM; the polar PCM remains but is closer to the equator due to the shortening of the spindle, whilst the cortical PCM (viewed here en face) remains cortical and has not been driven towards the centre of the oocyte. (A,B,D-G, ×400; C,H, ×630).
To determine whether the central displacement of the microtubule-organizing activity of the oocyte was some nonspecific result of the addition of DMSO or was a consequence of the polymerization of micro-tubules induced by DMSO, we examined the effects of DMSO on microtubule organization in the presence of 10 μM-nocodazole, a drug that binds to tubulin monomers and prevents their polymerization. Nocodazole added to control oocytes caused loss of polymerized tubulin, and hence of the spindle, and the scattering of both the chromosomes and the polar PCM around the cortex (Table 3, line 1 - see also Maro et al. 1986 and Pickering & Johnson, 1987). In contrast, if oocytes were added to medium containing both nocodazole and DMSO, the chromosomes remained in a tight cluster associated with a residual shortened spindle and polar PCM (Fig. 2D-H). However, no asters formed in association with the PCM (polar or cytoplasmic; Fig. 2D) and no central migration of cytoplasmic PCM occurred (Table 3, line 3; Fig. 2F,H). To determine whether the inward migration of PCM also required polymerized micro-filaments, oocytes were exposed to DMSO in the presence of cytochalasin D, a drug which destabilizes microfilaments (Maro et al. 1984). Under such conditions, the oocyte microtubule pattern resembled that of oocytes in DMSO alone (Table 3, line 4).
Influence of nocodazole and cytochalasin D on the DMSO-induced effects on the organization of microtubules in the oocyte

We therefore conclude that prolonged exposure to DMSO causes the principal organizing focus for microtubules in the oocyte to be transferred from the chromosomes to the PCM and that the tubulin polymerization that results drives the PCM towards the centre of the cell.
(C) Effect of DMSO concentration
DMSO appears to be influencing the stability of microtubules with respect both to cooling and to patterns of polymerization. We next examined whether there was a reduced concentration of DMSO to which oocytes could be exposed at 4 °C for 30 min and which resulted in stabilization of the existing spindle but no net increase in polymerization associated with PCM. The results are presented in Table 4, and reveal that as the concentration of DMSO falls below 0·75 M, the barrel-shaped spindle structure is no longer preserved at low temperature. Loss of PCM-mediated polymerization occurred at DMSO concentrations below 0·5 M. Thus, there does not appear to be a DMSO concentration at which a normal spindle is preserved in the absence of PCM-mediated polymerization.
(D) Reversibility of DMSO action
DMSO was removed at either 4°C or 37°C. Removal after Ih at 4 °C followed by immediate analysis revealed rudimentary elements of a spindle and few or no PCM-organized microtubules were detected (Table 5, lines 1 and 2; Fig. 3A,B). Exposure of oocytes to 37 °C during or after removal from DMSO (regardless of the duration of exposure) resulted in the restoration of a barrel-shaped spindle in some oocytes (Table 5, lines 3–6; Fig. 3C,D). However, in these cells, the chromosomes were not always all located on the equatorial plate, one or more stray chromosomes being evident (Fig. 3E-H). Of those oocytes lacking barrel-shaped spindles, a few lacked spindles altogether, some had partial spindles and others had abnormal spindles with multiple poles or errant bundles of microtubules (Fig. 3I-N). Some of these latter oocytes retained polar and cortical asters, although usually much reduced in extent and always located cortically rather than centrally. In some oocytes lacking a properly reconstituted spindle, chromosomal dispersal was evident.
Oocytes examined for distribution of chromatin (B,D,F,H,J,L,N,P), tubulin (A,C,E,G,I,K) or PCM (M,O). (A,B) Oocyte placed in DMSO at 4 °C for 1 h and then returned to control medium at 4 °C and fixed immediately; note that the spindle and asters evident in DMSO at 4 °C (see Fig. 1D-H) have disappeared almost completely. (C,D) Oocyte treated as in A and B, but then warmed to 37°C; note restoration of apparently normal shape to the spindle. (E-H) Oocytes exposed to DMSO at 37°C and then returned to control medium at 37°C; note irregularities in spindle organization and chromosomes displaced from the equatorial plate (arrowed). (I,J) Oocyte exposed to DMSO at 4 °C and then returned to control medium at 37°C; note in J that the chromosomes are split into two clumps (one clump out of focal plane) and in I that microtubules have formed into disorganized bundles (first polar body arrowed). (K,L) Oocyte exposed to DMSO at 37°C and returned to control medium at 37°C - note detached chromosome (arrowed) and abnormal spindle shape. (M-P) Oocytes treated as in K and L but examined for PCM distribution; note that all the PCM is clustered around the spindle, in M and N in three clusters at the poles (arrowed) of a tripolar spindle, in O and P in a semicircle around the spindle. (A-L, ×360; M-P, ×570.)
Oocytes examined for distribution of chromatin (B,D,F,H,J,L,N,P), tubulin (A,C,E,G,I,K) or PCM (M,O). (A,B) Oocyte placed in DMSO at 4 °C for 1 h and then returned to control medium at 4 °C and fixed immediately; note that the spindle and asters evident in DMSO at 4 °C (see Fig. 1D-H) have disappeared almost completely. (C,D) Oocyte treated as in A and B, but then warmed to 37°C; note restoration of apparently normal shape to the spindle. (E-H) Oocytes exposed to DMSO at 37°C and then returned to control medium at 37°C; note irregularities in spindle organization and chromosomes displaced from the equatorial plate (arrowed). (I,J) Oocyte exposed to DMSO at 4 °C and then returned to control medium at 37°C; note in J that the chromosomes are split into two clumps (one clump out of focal plane) and in I that microtubules have formed into disorganized bundles (first polar body arrowed). (K,L) Oocyte exposed to DMSO at 37°C and returned to control medium at 37°C - note detached chromosome (arrowed) and abnormal spindle shape. (M-P) Oocytes treated as in K and L but examined for PCM distribution; note that all the PCM is clustered around the spindle, in M and N in three clusters at the poles (arrowed) of a tripolar spindle, in O and P in a semicircle around the spindle. (A-L, ×360; M-P, ×570.)
In oocytes in which a barrel-shaped spindle reformed, PCM was located at the spindle poles.
However, unlike the situation prior to DMSO exposure, in 81 % of 114 such oocytes examined all the observable PCM was associated closely with the spindle, either at its poles or lateral to it, rather than being dispersed round the cytocortex (Fig. 3M-P).
(E) DMSO-induced polymerization in activated oocytes
During the recovery and handling of oocytes, a few became activated parthenogenetically and had reached anaphase or telophase of the second meiotic division by the time of fixation. We observed that in such oocytes, addition of DMSO also produced a polymerization of microtubules, but that the pattern of distribution differed from that in nonactivated oocytes. We therefore inseminated oocytes with spermatozoa in vitro, removed sample oocytes at 30 min intervals over a period of 2h, placed them in 1·5 M-DMSO at 37 °C for 30 to 90 min and analysed the patterns of microtubule distribution. Unfertilized oocytes and oocytes in the very earliest stages of anaphase at the time of fixation (as assessed from their chromatin organization) showed a microtubule pattern in DMSO similar to that described earlier. However, oocytes in late anaphase or telophase differed in that the network of microtubules that formed was organized in a more delicate meshwork and in all 235 oocytes examined this network was localized in the cortex of the oocyte (Fig. 4A–G) rather than having moved towards the centre. No radially organized tubules as described above (Fig. 2A-C) were evident. This pattern was identical to that seen in oocytes activated parthenogenetically and did not therefore require a spermatozoal component. Addition of DMSO suppressed spindle rotation, and thus polar body extrusion, in some (Fig. 4D,E) but not all (Fig. 4F,G) oocytes. Removal of the DMSO followed by incubation for 1h at 37 °C resulted in a reduction of the cortical mesh to the level seen in controls (Maro et al. 1986). However, when removal of DMSO was at 4°C, microtubules were lost throughout the cell except in the region between the two sets of anaphase chromosomes (Fig. 4H,I), suggesting that these (presumptive midbody) microtubules are more resistant to cold disruption.
Fertilized oocytes exposed to DMSO and examined for the distribution of chromatin (A,E,G,I) or tubulin (B,C,D,F,H). (A-C) An anaphase oocyte in which the spindle remnant is seen end on with one set of chromosomes in focus (A); an optical section in (B) shows the cortical mesh of microtubules, whilst (C) shows an en face view. (D,E) Telophase oocyte in which polar body extrusion has not occurred; note the cortical mesh. (F,G) En face view of a telophase oocyte in which polar body extrusion has occurred; note cortical mesh of microtubules. (H,I) Late anaphase oocyte exposed to DMSO at 4°C and then washed free of DMSO at 4 °C for 1 h; note stability of the microtubules linking the two chromatin masses. (×400.)
Fertilized oocytes exposed to DMSO and examined for the distribution of chromatin (A,E,G,I) or tubulin (B,C,D,F,H). (A-C) An anaphase oocyte in which the spindle remnant is seen end on with one set of chromosomes in focus (A); an optical section in (B) shows the cortical mesh of microtubules, whilst (C) shows an en face view. (D,E) Telophase oocyte in which polar body extrusion has not occurred; note the cortical mesh. (F,G) En face view of a telophase oocyte in which polar body extrusion has occurred; note cortical mesh of microtubules. (H,I) Late anaphase oocyte exposed to DMSO at 4°C and then washed free of DMSO at 4 °C for 1 h; note stability of the microtubules linking the two chromatin masses. (×400.)
Discussion
Appropriate organization of the second meiotic spindle of the oocyte is essential if a euploid zygote is to be formed at fertilization (Edwards, 1958; Webb, Howlett & Maro, 1986). In this paper, we demonstrate that DMSO, a cryoprotectant used widely in the preservation of frozen oocytes and embryos, exerts profound effects on microtubular organization and, in most oocytes, these effects are not fully reversible.
DMSO has been shown to exert stabilizing effects on the assembly of purified tubulin monomers in vitro (Himes, Burton, Kersey & Pierson, 1976; Himes, Burton & Gaito, 1977; Burton & Himes, 1978). The DMSO is thought to act by reducing the effective water concentration around the tubulin, thereby increasing the concentration of monomer and driving the reaction towards net assembly. However, these in vitro microtubules were not stable at 4°C unless microtubule-associated proteins were also present.
The observations on microtubule formation in situ reported here are consistent with such an explanation. Thus, a rise in the effective concentration of tubulin monomer would explain the appearance in the oocyte of multiple asters associated with the PCM, both at the spindle poles and cytocortically. It is also possible that DMSO causes other changes in the oocyte, for example, a reduction in the threshold level of free tubulin at which polymerization occurs in association with PCM. Presumably, once microtubules have been assembled due to the action of DMSO, associated proteins then combine with them to protect against cold disruption (Magistrini & Szoilosi, 1980; Pickering & Johnson, 1987). An analogous change has been reported for sea-urchin eggs treated with hexyleneglycol, which thus appears to function like DMSO (Endo, Toriyama & Sakai, 1983).
Once the DMSO-driven polymerization of PCM-associated microtubules is achieved, the continuing turnover of tubulin in both asters and spindle leads to the gradual unravelling of the latter, whilst the former proliferate. With loss of the spindle, dispersal of both the polar PCM and the chromosomes occurs. Conversely, if the concentration of free tubulin is reduced concurrent with the addition of DMSO (as achieved here by addition of nocodazole to sequester monomer chemically; Table 3), chromosomal control over polymerization is retained, a modified spindle structure persists and PCM-nucleated asters do not form. These observations support the notion that during M-phase the critical concentration for polymerization of tubulin is lower in the vicinity of the chromosomes than elsewhere in the oocyte (Maro et al. 1985; Karsenti & Maro, 1986).
The proliferation of astral microtubules appears to be oriented spatially in that those radial arrays of microtubules between the PCM foci and the surface are particularly prominent. Presumably, micro-tubules formed on the superficial side of the initially cortical PCM foci are stabilized preferentially compared with those formed in other orientations. Such a spatial ordering of microtubule stability could then drive the PCM centrally. That the central movement of PCM is indeed related to tubulin polymerization is shown by the action of nocodazole in inhibiting central migration in the presence of DMSO. In contrast, inhibition of microfilaments by addition of cytochalasin D concurrent with the DMSO did not block the central migration. The microtubule-stabilizing activity associated with the cytocortex that is implied by these results appears to be lost at, or shortly after, the reactivation of meiosis. At this time a fine cortical mesh forms naturally and its formation appears to be exaggerated in the presence of DMSO. However, in no case did the polymerization induced by DMSO in activated oocytes lead to stable radial microtubules or to a central migration of PCM. It is of interest that several hours later in the first cell cycle a central migration of both the pronuclei and the PCM does occur, and that this migration is dependent upon microtubule polymerization and on the formation of radiating astral arrays that are not too dissimilar to those induced by DMSO (Màro et al. 1985; Schatten et al. 1985, 1986). This observation raises the possibility that there is a cell-cycle-associated change in the properties of the zygote cytocortex involving the stabilization of microtubules and/or the way in which the PCM is linked to the surface. The possibility that such changes in the activity of the cortex might play a role in the spatial organization within the cell has, of course, wider developmental implications (Stephens,1986; Kirschner & Mitchison, 1986). There is evidence that localized modulation of microtubule stability and organization arising from changes in the cytocortex may be an important step in intracellular reorganization and thereby cell diversification in the cleavage-stage mouse embryo (Houliston, Pickering & Maro, 1987; Johnson, Chisholm, Fleming & Houliston, 1986; Johnson & Maro, 1986), as well as in the orientation of migration of Dictyostelium cells (Yumara & Fukui, 1983), differential cleavage in Caenorhabditis (J. White, personal communication), the ordering of nuclei in the early Drosophila embryo (Foe & Alberts, 1983), and the determination of axis formation in Xenopus (Danilchick & Gerhart, 1986).
Since the effects of DMSO on the microtubules, the PCM and the chromosomes of the oocyte are so profound, and since DMSO is used widely as a cryoprotectant for human oocytes during infertility therapy, we examined whether its influence was fully reversible. The results were not entirely reassuring. Whilst a ‘normal’ looking spindle reformed in some oocytes, a high incidence of abnormal spindles and dispersed chromosomes was evident. Moreover, even those spindles that appeared to be normal were associated with an atypical distribution of PCM, in most oocytes all of the PCM being associated closely with the spindle rather than dispersed through the cytocortex. Retention of the PCM around the spindle could lead to abnormal segregation of both the chromosomes and of the PCM at activation. Loss of much or most PCM to the second polar body could be deleterious for early cleavage divisions, which depend exclusively upon maternally-derived PCM for their organization (Maro et al. 1985; Szollosi et al. 1972). Moreover, since glycerol, the other major cryoprotectant used in oocyte-storage programmes, is also reported to have stabilizing effects on microtubules (Lee & Timasheff, 1975), it may not be possible to avoid the problematic results obtained with DMSO. Neither will reduction of the DMSO concentration be useful, since its use at 1M concentration, below which its cryoprotectant action is not effective, did not reduce the deleterious effects (Table 4). We are currently examining the consequences of exposure of oocytes to DMSO for early development of the mouse oocyte, as well as the effects of DMSO on the cytoskeletal organization of human oocytes. It seems clear already, however, that the difficulties experienced in the freezing of oocytes (Glenister, Wood, Kirby & Whittingham, 1987) may result as much or more from the deleterious effects of cryoprotectant as from the freezing process itself.
ACKNOWLEDGEMENTS
We wish to thank Martin George and Brendan Doe for their technical help, Evelyn Houliston and Bernard Maro for valuable advice and discussion, and Drs J. Kilmartin and M. Kirschner for reagents. The work was supported by grants from the Medical Research Council and the Cancer Research Campaign.