Water and DMSO membrane permeability characteristics of in-vivo- and in-vitro-derived and cultured murine oocytes and embryos

Transcription

1 Molecular Human Reproduction vol.4 no.1 pp , 1998 Water and DMSO membrane permeability characteristics of in-vivo- and in-vitro-derived and cultured murine oocytes and embryos R.T.Pfaff 1,2, J.Liu 1, D.Gao 1, A.T.Peter 2, T.-K.Li 3 and J.K.Critser 1,2,4,5,6 1 Cryobiology Research Institute at Methodist Hospital of Indiana, Indianapolis, IN 46202, 2 Department of Veterinary Clinical Sciences, School of Veterinary Medicine, Purdue University, West Lafayette, IN 47907, 3 Department of Biochemistry, Indiana School of Medicine; Indianapolis, IN 46202, 4 Departments of Obstetrics and Gynecology, Physiology and Biophysics and Pediatrics, Indiana School of Medicine, Indianapolis, IN 46202, and 5 Advanced Fertility Institute, Methodist Medical Plaza North, Indianapolis, IN 46280, USA 6 To whom correspondence should be addressed at: Cryobiology Research Institute, Cancer Research Building, Wells Research Center, Room 431B, 1044 W. Walnut Street, Indianapolis, IN 46202, USA Although embryo cryopreservation is routine for many mammalian species, it is important to know how the fundamental cryobiology of these cells changes with development. Progressive cleavage divisions result in a reduction in the blastomere surface area available for water and cryoprotectant mass transport. Therefore, the membrane permeability of murine oocytes, zygotes, 2-cell, 4-cell, and 8-cell embryos to water (L p ), and dimethylsulphoxide (P DMSO ), and the reflection coefficient, sigma (σ) were determined. Oocytes or zygotes were recovered, cumulus cells removed, then cultured until use. Oocytes and embryos were immobilized and perfused with treatment solutions at 24 C. Osmotically induced cell volume changes over time were videotaped followed by image analysis. The L p values in the presence of dimethylsulphoxide (DMSO) were 0.77, 0.81, 0.94, 0.86, and 1.10 µm/min / atm, and the P DMSO values were 1.85, 2.04, 2.41, 1.95, and cm/min for oocytes, zygotes, 2, 4, and 8-cell embryos respectively. The L p values in the presence of DMSO were significantly (P 0.05) higher than those in the absence of DMSO. Treating the whole embryo as a single osmotic entity leads to significantly (P 0.05) elevated P DMSO estimates relative to those based upon measurements of individual blastomeres. These data indicate that both L p and P DMSO estimates are lower when predicted on an individual blastomere basis. The data also show that neither L p nor P DMSO differ among oocytes, zygotes, 2-cell and 4-cell embryos. However, the significantly higher L p and lower P DMSO of the 8-cell stage support the hypothesis that fundamental cryobiological differences may require developmental stage-specific embryo cryopreservation protocols. Key words: cleavage stages/dmso/membrane permeability Introduction Since the first successful cryopreservation of mouse embryos (Whittingham et al., 1972), embryos of many mammalian and a few non-mammalian species have been successfully cryopreserved (Leibo, 1986). Embryos from cattle, mice and humans are routinely cryopreserved. However, the developmental stage at which cryopreservation is usually performed differs among these different species. This is determined, in part, by the practicality of recovery techniques, culture methods, and by embryo survival rates after cryopreservation. Murine embryos can be readily recovered from the oviductal ampulla as one-cell embryos, but are routinely cryopreserved at the 8-cell stage because higher survival rates have been obtained at this stage (Miyamoto and Ishibashi, 1977; Takeda et al., 1987). Bovine embryos do not usually enter the uterus before the morula stage, but can be recovered relatively easily and non-surgically from the uterus as morulae or blastocysts. They are routinely cryopreserved at these developmental stages with subsequent conception rates similar to those following non-frozen embryo transfer or artificial insemination (Willadsen et al., 1978; Elsden et al., 1982). The demand for efficient and efficacious embryo cryopreservation protocols is enhanced by the rapidly increasing utilization of in-vitro maturation and fertilization and the subsequent in-vitro culture of resulting embryos. These techniques create a greater need to preserve embryos for later use. For example, in assisted reproduction clinics most human embryos are derived from in-vitro fertilization and are often cryopreserved either at the pronuclear stage (zygotes) or as early cleavage stage embryos when supernumerary embryos are present. Membrane water and cryoprotectant permeability coefficients have been estimated predominantly for oocytes and zygotes. There is very little information available for cleavage stage embryos. The water permeability (hydraulic conductivity) for oocytes has been determined for mice (Leibo, 1980; Benson and Critser, 1994), cattle (Myers et al., 1987), and hamsters (Benson and Critser, 1994), and estimates of some cryoprotectant permeabilities for oocytes have been determined for mice (Jackowski et al., 1980, using glycerol; McGrath et al., 1992a, using DMSO or 1,2 propanediol), and humans (Fuller et al., 1992, for 1,2 propanediol). The hydraulic conductivity for zygotes has been estimated for mice (Leibo, 1980), hamsters European Society for Human Reproduction and Embryology 51

2 R.T.Pfaff et al. (Shabana and McGrath, 1988), goats (Le Gal et al., 1994), cattle (Ruffing et al., 1993) and humans (Hunter et al., 1992), and the cryoprotectants permeability for zygotes has been estimated for mice (Jackowski et al., 1980 using glycerol; McGrath et al., 1992a using DMSO and 1,2 propanediol). There are very few reports of experiments regarding the permeability of early embryo stages other than zygotes. Schneider and Mazur (1984) reported permeability coefficients of bovine early blastocysts using glycerol and DMSO. Mazur and Schneider (1986) also carried out measurements on entire early embryos; they found that the membrane permeability increased from the zygote stage to the blastocyst stage. The ratio of surface area to volume, however, is not constant among cleavage stages, but increases as the volumes of individual cells decrease. Post-cryopreservation survival rates tend to be higher for embryos at later developmental stages with higher cell numbers. This phenomenon could be due to: (i) the decreased likelihood of embryonic death if one or few cells are damaged during the cryopreservation process, i.e. the loss of even a few cells of an 8-cell embryo might not hinder further development, but the loss of one cell of a 1-cell embryo is obviously fatal; (ii) the sensitivity to chilling of oocytes (Leibo et al., 1995) and some in-vitro derived embryos (Pollard and Leibo, 1994); (iii) different sensitivity to cryoprotectant toxicity (Fahy, 1984; Arakawa et al., 1990); (iv) different available surface area to volume ratios (Mazur, 1963), and (v) different membrane permeabilities at different developmental stages to water and cryoprotectant (Jackowski and Leibo, 1976; Jackowski et al., 1980). The objective of this experiment was to determine membrane permeabilities of mouse oocytes, zygotes and cleavage stage embryos to water and DMSO based upon measurements made at the level of the individual cell, and to compare these values to those made on the entire embryo. Materials and methods Source of embryos Ovulation was stimulated in female B6C3F1 mice using i.p. injections of 7.5 IU equine chorionic gonadotrophin (ecg; Diosynth Inc, Chicago, IL, USA) and 5 IU human chorionic gonadotrophin (HCG; Sigma Chemical Co., St Louis, MO, USA) 48 h after the ecg injection. To obtain zygotes, B6C3F1 female mice were mated 1:1 with ICR males after injection with HCG. To recover Metaphase-II oocytes, mating was omitted after injection with HCG. Oocytes and zygotes were recovered h after injection with HCG. Oocytes and zygotes were placed in phosphate-buffered saline (PBS; Life Technologies Inc, Grand Island, NY, USA) supplemented with 1 mg/ml hyaluronidase (Sigma Chemical Co) and 4 mg/ml bovine serum albumin (BSA; Sigma Chemical Co) for ~5 min to remove the cumulus cells. Oocytes and zygotes were then washed by transfer through three 40 µl culture drops of serum-free, gas-pre-equilibrated Ham s F10 medium under oil, and cultured in vitro in humidified 5% O 2 /5% CO 2 /90% N 2. Oocytes and zygotes were subjected to osmotic experiments within h after collection. Cleavage stage embryos (2-, 4-, and 8-cell stage) were obtained by in-vitro culture of recovered zygotes to the respective experimental stage. The solutions used for the determination of osmotically inactive 52 volumes (V b ) and hydraulic conductivity (L p ) consisted of PBS as base medium. The osmolality of the solutions was adjusted by addition of calculated amounts of reagent grade water or sodium chloride (NaCl; Sigma Chemical Co.) to obtain solutions of 235, 285, 400, 570 and 790 mosm. The osmolalities of the solutions were then verified via osmometer. The solutions were supplemented with 0.1% (w/v) BSA. For determination of osmotically inactive volumes, embryos were perfused for 5 10 min. For determination of the hydraulic conductivity, embryos were equilibrated in 285 mosm solution, then perfused for 10 min with a 600 mosm solution, followed by perfusion for 10 min with a 285 mosm solution. To estimate all three membrane permeability parameters (L p,p DMSO, and σ), embryos were perfused with 1.5 M DMSO in PBS supplemented with 0.1% (w/v) BSA for min. Viability assay To verify viability of the embryos upon completion of osmotic experiments, individual embryos were washed through five drops of Ham s F10, placed into culture drops and cultured to the blastocyst stage. Viability of oocytes was verified using fluorescein diacetate (FDA; Rotman and Papermaster, 1965). Only embryos that developed to the blastocyst stage, and oocytes with a positive FDA stain were included in the data analysis. Data collection and image analysis To obtain optimal spatial resolution, a microperfusion chamber was built to continuously perfuse an oocyte or embryo, which was immobilized by a holding pipette. This technique was modified from that previously described by Gao et al. (1994) and yields optimal resolution combined with maximal magnification. The volume changes of the cells were videotaped using 40 magnification on an inverted Nikon (Fryer Co, Carpentersville, IL, USA) compound microscope. Video images were captured at predetermined time intervals using the software Vidcap 1.00 (Microsoft Corporation). Every captured frame (Figure 1) was analysed utilizing a customized image analysis program. The following two measurement methods were applied. Method I: measurements of individual cells Due to their spherical shape, the cell volume (V) and surface area (A) data for oocytes and zygotes can be readily calculated from the measurement of the radius (r) using: 4π V 3 r 3, A 4π r 2. [1] Data acquisition from blastomeres of 2-, 4-, and 8-cell stage embryos was more complex. The shape of individual blastomeres of the cleavage stage embryo deviates from spherical due to their impingeing upon one another. For blastomeres of the 2-cell stage embryo, the formula for an oblate spheroid was used to calculate the volume and surface area. Then, the values of volume and surface area were adjusted by subtracting those of the impinging zone segment as shown in the schematic drawing of Figure 2. The assumption was made that this area does not participate significantly in the membrane mass transport, since part of it is not available for direct contact with the extracellular medium. The volume and surface area of the 2-cell are derived from: and, πa 2 b V 6 V zone [2]

3 Permeability characteristics of oocytes and early cleavage stage embryos Figure 1. Images of (a) an oocyte and cleavage stage (b) 2-cell, (c) 4-cell, and (d) 8-cell mouse embryos captured from a videotape. During the experiments, the oocytes and embryos were immobilized by a holding pipette inside a microperfusion chamber. a 2 π b 2 π (1 ε) A ln 2r ab h 2 π, [3] 2 4ε (1 ε) where the volume of the zone segment V zone (π/3) h 2 2 (3r ab h 2 ), h 2 is the height of the zone (the subscript of h 2 refers to a 2-cell embryo), r ab (a b)/4, and ε 1 (b/a) 2 (see Figure 2 for definitions of a and b). For blastomeres of 4- and 8-cell stage embryos, the formula for the sphere was applied in the calculations. Because of the deviation from a perfect sphere, half of the average of the maximum and minimum diameter was used as the radius (r) of the sphere to calculate the volumes and surface areas. Corrections similar to those for blastomeres of 2-cell stage embryos were performed by subtracting the zones of blastomere-blastomere contact. Thus, the values for volume and surface area of one blastomere of the 4- and 8-cell stage embryos are given by: and 4π V 3 r 3 n s V zone [4] A 4π r 2 2n s rh 4 π, [5] where V zone (π/3) h 2 4 (3r h 4 ) and, h 4 r r 2 d 2 /4. The term n s refers to the number of contact surfaces (see Figure 2 for definition of d). It was assumed that there are three contact surfaces (n s 3) per blastomere in a 4-cell embryo, and four contact surfaces (n s 4) per blastomeres in an 8-cell embryo. Method II: measurements of the entire embryo Two diameters (minimum and maximum) were obtained by measuring across the whole embryo; a quarter of the sum of the two diameters was then used as the radius to calculate the volume and surface area for the entire embryo. This was done to test whether measurements across the entire embryo would lead to similar parameter estimates as obtained in method I. Determination of the permeability coefficient of cryoprotectant agents (CPA) Kedem and Katchalsky (1958) formulated a theoretical model for the permeability of membranes to non-electrolytes based on the theory of irreversible thermodynamics. For a ternary solution consisting of a permeable solute (in this scenario, the cryoprotectant DMSO), an impermeable solute (NaCl), and water, the total transmembrane volume flux (J v ) and permeable solute flux (J s ) are presented as: 1 dv Jv L p [(P e P i ) RT(M e n M i n) σrt(m e s M i s)] [6] A and, dt 1 dn i s Js (1 σ)m s J v P s (a e s a i s), [7] A dt where V is cell volume and N i s is the number of moles of CPA inside the cell, P is the hydrostatic pressure, M is osmolality and a is activity, m e s m i s m s ln m e s ln m i s (where m is the molal concentration). The superscripts i and e refer to intra- or extracellular cell compartment respectively. L p, P s and σ are the hydraulic conductivity of water, the permeability coefficient of the CPA, and the reflection coefficient. The temperature and universal gas constant are given by T and R (see Table I for explanations of descriptions, units and values of the symbols). Mazur and Miller (1976) calculated the osmolality (M j g) and activity 53

4 R.T.Pfaff et al. (a j g) for glycerol by using quadratic functions of molality to account for non-ideal solution effects: and M j g φ j g m j g υ g [8] a j g γ j g m j g, [9] where φ j g 1 p o m j g, and γ j g 1 g o m j g. The superscripts j i or e refer to intra- or extracellular region. In the present studies, however, the osmotic coefficient (φ s ) and activity coefficient (γ s ) of the CPA are set to be one. Therefore, this yields: a j s m j g, M j s m j s. [10] The above approximation is based on the following: (i) The values of coefficients (p o and 2g o ) are constant and can be ignored since a single concentration 1.5 M DMSO was used, and (ii) the volume change is determined by the gradients of osmolality and activity differences across the membrane, not by their absolute values. For the impermeable solute (NaCl), the osmotic coefficient (φ n ) was assumed to be constant. Then the intracellular osmolality of NaCl is given by: V (o) V b V(o) s M i (o) n M n, [11] V V b V s where V b is the osmotically inactive cell volume and V s N i s V s (V s is the partial molar volume of the CPA). The superscript (o) represents the values at t 0. The relation between N i s and m i s is given by: N i s (V V b V s )m i s, [12] where (V V b V s ) denotes the volume of intracellular water. Combining equations [6], [7], [12], and assuming the hydrostatic pressure difference across the membrane to be zero, a pair of coupled differential equations that describe the cell fluid volume and amount of solute in the cell as functions of time are obtained: dv Lp ART[(M i n M e n) σ(m i s m e s)] [13] Figure 2. Schematic drawings of cross-sections of a 2-cell (left) and 4-cell (right) embryo. For one blastomere in the 2-cell embryo, the a and b axis of an oblate spheroid were measured. The effective volume and surface area of the blastomere were obtained by subtracting those of the zone segment from the oblate spheroid. For a 4-cell blastomere, three zones were subtracted instead of the two as depicted in the two-dimensional graph. and dt dm i s (1 V s m i s) 2 dt V V b m i s dv {[ m s(1 σ) ] AP s [(m e s ms)]} i. [14] (1 V s m i s) dt Calculation procedure Oocytes and embryos were perfused with 290 mosm NaCl solution and 1.5 M DMSO solution at 24 C. This temperature was chosen because it is that at which CPA addition is commonly performed. The extracellular DMSO concentration (1.5 M) and NaCl osmolality (290 mosm) were assumed to be invariant with time. At t 0, there Table I. Definitions of major symbols. Symbol Description Units Value e, i Superscripts (e external, i internal to cell) s, n Subscripts (s solute, n non-permeating salts) L p Water permeability µm/min/atm parameter P s Solute permeability cm/min parameter σ Reflection coefficient parameter T Temperature K parameter A Surface area of cell µm 2 V Volume of water plus solute in cell µm 3 variable Ns i Femtomoles of solute in cell at time t femtomoles variable t Time in contact with solute seconds variable M ( ) n Osmolality of initial (isotonic) non-permeating osmoles/kg H 2 O salts inside cell Mn e Osmolality of external nonpermeating salts osmoles/kg H 2 O Mn i Osmolality of internal nonpermeating salts osmoles/kg H 2 O variable m Molality moles/kg H 2 O variable a Activity moles/kg H 2 O variable φ Osmotic coefficient of solute parameter γ Activity coefficient of solute parameter V s Partial molar volume of solute litre/mole R the universal gas constant kcal/mole/k

5 Permeability characteristics of oocytes and early cleavage stage embryos Table II. Osmotically inactive cell volumes (V b ). Values are given as means SD, with the number of observations in parentheses Developmental Stage V b (%) Oocyte (10) Zygote (10) 2-cell (10) 4-cell (15) 8-cell (14) Means do not differ significantly (P 0.1). is no CPA inside the cell, so m i s 0 and V(0) V iso, where V iso, the isosmotic volume, was determined at the beginning of the osmotic response experiments. A commercial software package, MLAB (Civilized Software Inc, Bethesda, MD, USA) was used to solve the differential equations using the Gear method. The Marquard Levenberg method implemented in MLAB was used to perform a three-parameter curve fit calculation to estimate the values of L p, P DMSO, and σ. Statistical analysis The data were subjected to analysis of variance (ANOVA) using the Statistical Analysis System (SAS). Tukey s Studentized Multiple Range Test was used to compare the means of the developmental stages (Steel and Torrie, 1960). Results Microperfusion chamber Immobilization of the embryos with a holding pipette resulted in excellent resolution at high magnification. Although the zona pellucidae of several oocytes and embryos were slightly indented due to the exertion of negative pressure of the holding pipette, the oocytes and embryos themselves maintained a nearly spherical shape throughout the volume excursions. When perfused with 1.5 M DMSO at 24 2 C, oocytes and embryos rapidly decreased in volume and slowly regained their volumes during the perfusion period. The relative cell volumes at the end of the equilibration periods were ~110% in respect to their original volume due to the loaded volume of CPA inside the cell. Viability The viability of oocytes and embryos after the experiment was high. All oocytes fluoresced upon FDA-test. High percentages of embryos developed after perfusion and removal of DMSO during subsequent in-vitro culture to the blastocyst stage; 78.3, 85.7, 87.5 and 90.9% of zygotes, 2-cell, 4-cell, and 8-cell stage embryos developed to blastocysts after treatment. Osmotically inactive volumes The osmotically inactive volumes ranged from 13.7 to 24.2% of isosmotic volumes. The mean values, given in Table II, were not different among developmental stages (P 0.1). Hydraulic conductivity In the absence of DMSO, the L p values for water leaving the cell (water efflux) as well as for water entering the cell (water influx) were significantly higher (P 0.05) for 2-, 4-, and Table III. Water permeability (L p ) in the absence or presence of dimethyl sulphoxide (DMSO). Values are given as means SD, with the number of observations in parentheses Developmental L p -efflux L p -influx L p Stage (µm/min/atm) (µm/min/atm) (µm/min/atm) in absence of in absence of in absence of DMSO DMSO DMSO Oocyte aa (10) aa (10) ab (10) Zygote aa (9) aa (8) ab (10) 2-cell ba (15) ba (13) aba (11) 4-cell ba (10) ba (5) aa (20) 8-cell ba (7) ba (7) ba (20) a,b Means within columns with different superscripts differ (P 0.05). A,B Means within rows with different superscripts differ (P 0.05). Table IV. Water permeability (L p ) in the presence of DMSO based upon measurements conducted on the individual blastomere (method I) or on the entire embryo (method II). Values are given as means SD, with the number of observations in parentheses Developmental L p (Method I) L p (Method II) Stage (µm/min/atm) (µm/min/atm) Oocyte a (10) Zygote a (10) 2-cell aba (11) aa (6) 4-cell aa (20) aa (6) 8-cell ba (20) aa (6) a,b Means within columns with different superscripts differ (P 0.05). A Means within rows with same superscripts do not differ (P 0.05). 8-cell embryos than for oocytes and zygotes (Table III). The L p values for water efflux for all developmental stages (Table III), were not statistically different (P 0.05) from L p values for water influx. The values for hydraulic conductivity in the presence of DMSO were significantly higher (P 0.05) for oocytes and zygotes when compared with the average values of water efflux and water influx. The values for 2-, 4-, and 8-cell embryos were not significantly different (P 0.05) from the values obtained in the absence of DMSO. There were no differences among L p values of cleavage stage embryos (P 0.05) between measurements carried out at the individual cell level (method I) compared with measurements carried out on the entire embryo (method II; Table IV). Permeability to DMSO Membrane permeability to DMSO was highest for 2-cell embryos, but was not significantly higher (P 0.05) than for zygotes and 4-cell embryos. Permeability to DMSO for 8-cell embryos was significantly lower (P 0.05) than for the other (earlier) developmental stages (Table V). There was a significant difference (P 0.05) between measurement methods. Mean values of 2-cell, 4-cell, and 8-cell embryos obtained from measurements carried out on the entire embryo (method II) were higher (P 0.05) than mean values obtained from measurements carried out at the individual cell level (method I; Table V). There was no significant difference among the cleavage stages (2 8-cell stages) when measurements were carried out on the entire embryo (P 0.05), however, their 55

6 R.T.Pfaff et al. Table V. Dimethyl sulphoxide (DMSO) permeability estimates (P DMSO ) based upon measurements conducted on the individual blastomere (method I) or on the entire embryo (method II). Values are given as means SD, with the number of observations in parentheses Developmental P DMSO (Method I) P DMSO (Method II) Stage ( 10 3 /cm/min) ( 10 3 /cm/min) Oocyte ab (10) Zygote ab (10) 2-cell aa (11) cb (6) 4-cell aba (20) cb (6) 8-cell ca (20) cb (6) a,b Means within columns with different superscripts differ (P 0.05). A,B Means within rows with different superscripts differ (P 0.05). Table VI. Reflection coefficient estimates (σ) based upon measurements conducted on the individual blastomere (method I) or on the entire embryo (method II). Values are given as means SD, with the number of observations in parentheses Developmental σ σ Stage (Method I) (unitless) (Method II) (unitless) Oocyte a (10) Zygote a (10) 2-cell aa (11) aa (6) 4-cell aa (20) aa (6) 8-cell aa (20) bb (6) a,b Means within columns with different superscripts differ (P 0.05). A,B Means within rows with different superscripts differ (P 0.05). P DMSO values were higher than those of oocytes and zygotes (P 0.05). Reflection coefficient (σ) The mean values for σ ranged from 0.69 to 0.81 among developmental stages when measurements were carried out on the level of the individual cell, and were not significantly different (P 0.1) (Table VI). The mean values derived from measurements carried out on the entire embryo ranged from 0.64 to 0.90 among developmental stages, and were similar with the exception of the 8-cell stage, which had a significantly higher (P 0.05) σ value Discussion Oocytes and embryos responded osmotically and showed a linear relationship between cell volume and the reciprocal of osmolality when exposed to different osmolalities. The osmotic levels used to construct Boyle van t Hoff plots were well within the viable range, in which cells react osmotically in an ideal (e.g. linear) fashion as described by Mazur and Schneider (1986). The osmotically inactive volumes of ~19 %, are similar to those values that have been reported for zygotes of mice (18%, Leibo, 1980; 22%, Bernard et al. 1988), and for hamster oocytes (21.6%, Shabana and McGrath, 1988). Hydraulic conductivity values were determined at 24 2 C. In the absence of DMSO, there was no significant difference between the values of oocytes and zygotes. Their mean values for water efflux were 0.51 µm/min/atm for zygotes and 0.49 µm/min/atm for oocytes respectively. These values 56 are similar to those values reported by Leibo (1980), i.e µm/min/atm and 0.43 µm/min/atm. The mean values for water efflux of 2-, 4-, and 8-cell stage embryos were in the range µm/min/atm. The means among these developmental stages did not differ significantly, but the means of the cleavage stages were significantly higher than those of oocytes and zygotes with respect to water efflux. The hydraulic conductivity for water influx was, for all developmental stages, slightly higher than the hydraulic conductivity for water efflux. However, the means for water influx and for water efflux were not significantly different. There was no significant difference between measurement methods. In the presence of DMSO, the mean L p value for 8-cell stage embryos was significantly higher than the means of oocytes, zygotes and 4-cell stage embryos, and slightly, but not significantly higher than for 2-cell embryos. The L p mean values for oocytes and zygotes were 0.77 and 0.81 µm/min/ atm respectively. These values are similar to those values reported by McGrath et al. (1992a,b) for oocytes and zygotes of mice, using DMSO. In the presence of DMSO, all L p mean values were elevated, but the increased DMSO permeability was only significantly elevated for oocytes and zygotes. This increase in hydraulic activity is likely due to the presence of DMSO, which has amphiphilic properties. McGrath et al. (1992b) previously reported a similarly elevated L p value when mouse oocytes were perfused with DMSO. Gilmore et al. (1995) found that cryoprotectants affected the hydraulic conductivity of spermatozoa. However, in contrast to the increase in L p in oocytes and embryos seen here, they found that cryoprotectants caused a decrease in water transport through the plasma membrane. The membrane permeability values for DMSO were found to be similar among developmental stages when measurements were made on individual cells. The only exception was at the 8-cell stage, were P DMSO was significantly lower than for the other developmental stages. The P DMSO value for oocytes of cm/min obtained at 24 2 C is similar to that of cm/min at 20 C found by McGrath et al. (1992b). A possible reason for the decreased P DMSO value of 8-cell stage embryos and their increased hydraulic conductivity could be the result of the compaction process (i.e. formation of tight junctions) or the preparation of the embryo to undergo compaction (changes in cytoskeleton and membrane physics). Embryos that were morphologically classified as morulae were deliberately excluded from the experiment to avoid this problem. However, it is possible that ultrastructural changes, that are not discernible at relatively low magnification, had already taken place. Another possible explanation might be the slightly more polymorphic shape of the 8-cell stage embryo compared to the rather regular and consistent shape of embryos of earlier cleavage stages. When measurements were conducted in which the entire embryo was treated as a single osmotic unit, there was no significant difference in P DMSO among 2-, 4-, and 8-cell stage embryos. In contrast, P DMSO values of 2-, 4-, and 8-cell stage embryos were significantly higher than the values for oocytes and zygotes, and ~ times higher than those P DMSO values obtained from measurements which were conducted at

7 Permeability characteristics of oocytes and early cleavage stage embryos Figure 3. Osmotically induced volume changes of an oocyte and cleavage stage embryos after transfer from isotonic phosphatebuffered saline (PBS) solution into 1.5 M dimethyl sulphoxide (DMSO) solution. The curves were reconstructed using parameters obtained from measurements on the individual cell (method I). Figure 4. Simulation of osmotically induced volume changes of oocytes and cleavage stage embryos after transfer from 1.5 M dimethyl sulphoxide (DMSO) solution into isotonic phosphatebuffered saline (PBS) solution using data from measurements on the individual cell (method I). the individual cell level (Table V). Thus, P DMSO coefficients of cleavage stage embryos that are derived from measurements of entire embryos contain inherently inaccurate estimates for both surface area and volume. Additionally, the membrane permeability estimates deviate increasingly with each cleavage division and correspondingly lead to an increasing overestimation of P DMSO. Therefore, surface area/volume measurements at the level of individual cells are important even though they may deviate from the model of a perfect sphere (e.g. for 2-, 4- and 8-cell stage embryos). Measurements of individual cells will yield more accurate estimates of cell volumes and surface areas than estimates derived from measurements treating a cleavage stage embryo as a single unit. Another finding of this study is that the combination of the decreased P DMSO and increased L p values affected the osmotic responses in an additive pattern, resulting in lower minimum volumes during cryoprotectant addition (Figure 3) and higher maximum volumes during cryoprotectant removal (Figure 4). Figure 5 shows a simulation of osmotic responses of 2-, 4-, and 8-cell stage cleavage embryos after transfer into a 1.5 M DMSO solution, using parameter estimates obtained from measurement methods I and II of this study. These data show that both minimum volumes and equilibration times would be considerably underestimated and that this underestimation could result in the development of procedures which cause osmotic damage and cell death (Oda et al., 1992; Gao et al., 1995). The calculated minimum volumes of embryos after transfer into a cryoprotectant solution were similar for oocytes, zygotes, 2-, and 4-cell stage embryos, but were significantly lower than those of 8-cell stage embryos (Figure 3). Embryos at the 8-cell stage also experience greater increases in volume resulting in Figure 5. Comparative simulation of osmotically induced volume changes of 2-, 4-,and 8-cell stage cleavage embryos after transfer from isotonic phosphate-buffered saline (PBS) solution into 1.5 M dimethyl sulphoxide (DMSO) solution. Curves A, B, and C are derived from measurements on the individual cell (method I) of 2-cell ( ), 4-cell ( ) and 8-cell ( ) stage embryos, and curves a, b, c are derived from measurements on the entire embryo (method II) respectively. swelling to more than double their original volume when the parameter estimates obtained in this study were used to simulate the osmotic behaviour after transfer from 1.5 M DMSO into an isotonic PBS solution (Figure 4). Due to the 57

8 R.T.Pfaff et al. presence of the zona pellucida, the expansion of the oocytes and embryos is limited and stopped after completely filling the perivitelline space. The two graphs (Figures 3 and 4) do not suggest that 1.5 M DMSO may be removed in a single step, but provide an understanding of how the cells of oocytes and cleavage stage embryos would respond osmotically when transferred into cryoprotectant containing solutions or when cryoprotectants are removed from the embryos. Knowledge of maximum and minimum volumes of oocytes and embryos during addition and removal of cryoprotectant can be used to determine the time requirements to achieve intercellular CPA equilibration and to develop approaches that allow the cells to stay within their osmotic tolerance limits (Gao et al., 1995). An increased cell surface area/volume ratio of cleavage stage embryos would allow faster cooling rates for 2- and 4-cell stage embryos, assuming that L p and P s values are similar among these developmental stages. However, with each cleavage, less and less surface area is effectively available for the permeation and equilibration process due to alignment and contact of cell membranes between adjacent, neighbouring cells. Thus, the advantage of the theoretical increase in the surface area/volume ratio is balanced by the increasing amount of surface area that is not available in the permeation and equilibration process. After compaction, the parameters affecting the surface area/volume ratio have changed. The blastomeres of the compacted morula are fused with one another and form a single mass. Therefore, we assume that the peripheral surface area primarily contributes to the actual permeation and equilibration process. These data are important both from an applied (i.e. assisted reproductive technology) as well as a basic (i.e. biophysics/ cryobiology) perspective. Many supernumerary human embryos are cryopreserved at the zygote stage, but also often at early cleavage stages (Lassalle et al., 1985; Ashwood-Smith, 1986; Mandelbaum et al., 1987; Testart et al., 1987) to reduce the risk of multiple gestation and to avoid the necessity of discarding these embryos (Fugger 1993). Therefore, for reproductive biologists it is important to understand when different specific cryopreservation protocols should be applied based on the developmental stage of the embryo. For cryobiologists, these data demonstrate that the parameter estimates for membrane permeability are influenced by the technique used to conduct measurements on cells more complex than oocytes and zygotes. The parameter estimate changes are due, in part, to the deviation from a geometrical ideal, i.e. sphere, and to the partial alignment and impingement at the blastomereblastomere contact sites of cleavage stage embryos, which obstruct or prevent permeation and equilibration in certain areas of the embryo. There are many critical steps for cells to survive cryopreservation, as for example addressed for human oocyte cryopreservation by Bernard and Fuller (1996). It is possible to quantify the membrane permeability parameters of oocytes and embryos and to determine viability of these cells as a function of time and osmotic stress. The availability of this information will help in the modelling of more suitable cryopreservation protocols. This study indicates that parameters from cells other than oocytes and zygotes can be obtained. The data show that 58 membrane permeability parameter estimates can be significantly influenced by the measurement methods. The data further show that the water and DMSO membrane permeability of murine oocytes and early cleavage stage embryos (up to the 4-cell stage) are similar. However, the smaller P DMSO value of pre-compaction 8-cell embryos indicates that different stagespecific cryopreservation protocols may be required. Further research needs to be conducted to evaluate the temperature dependent activation energy and the interaction of these parameters as a function of temperature. Acknowledgements The authors thank S.P.Leibo and P.Mazur for critical comments regarding the experimental design and manuscript preparation. References Arakawa, T., Carpenter, J.F., Kita, Y.A. and Crowe, J.H. (1990) The basis for toxicity of certain cryoprotectants: a hypothesis. Cryobiology, 27, Ashwood-Smith, M.J. (1986) The cryopreservation of human embryos. Hum. Reprod., 1, Benson, C.T. and Critser, J.K. (1994) Variation of water permeability (L p ) and its activation energy (E a ) among unfertilized golden hamster and ICR murine oocytes. Cryobiology, 31, Bernard, A. and Fuller, B.J. (1996) Cryopreservation of human oocytes: a review of current problems and perspectives. Hum. Reprod. Update, 2, Bernard, A., McGrath, J.J., Fuller, B.J. et al. (1988) Osmotic response of oocytes using a microscope diffusion chamber: a preliminary study comparing murine and human ova. Cryobiology, 25, Elsden, R.P., Seidel, G.E., Takeda, T. and Farand, G.D. (1982) Field experiments with frozen thawed bovine embryos transferred nonsurgically. Theriogenology, 17, Fahy, G.M. (1984) Cryoprotectant toxicity: biochemical or osmotic? Cryo- Letters, 5, Fugger, E.F. (1993) Current methods for oocyte/embryo cryopreservation. In: Advanced Hands-On Workshop on Cryopreservation of Spermatozoa and Embryos. pp Fuller, B.J., Hunter, J.E., Bernard, A.J. et al. (1992) The permeability of unfertilized oocytes to 1,2 propanediol. A comparison of mouse and human cells. Cryo-Letters, 13, Gao, D.Y., Liu, J., Liu, C. et al. (1995) Prevention of osmotic injury to human spermatozoa during addition and removal of glycerol. Hum. Reprod., 10, Gao, D.Y., McGrath, J.J., Tao, J. et al. (1994) Membrane transport properties of mammalian oocytes: a micropipette perfusion technique. J. Reprod. Fertil., 102, Gilmore, J.A., McGann, L.E., Gao, D.Y. et al. (1995) Effect of cryoprotectant solutes on water permeability of human spermatozoa. Biol. Reprod., 53, Hunter, J.E., Bernard, A., Fuller, B.J. et al. (1992) Measurements of the membrane permeability (L p ) and its temperature dependence (activation energy) in human fresh and failed-to-fertilize oocytes and mouse oocytes. Cryobiology, 29, Jackowski, S.C. and Leibo, S.P. (1976) Response of fertilized mouse ova as function of permeation of glycerol. Cryobiology, 13, 646. Jackowski, S., Leibo, S.P. and Mazur, P. (1980) Glycerol permeabilities of fertilized and unfertilized mouse ova. J. Exp. Zool., 212, Kedem, O. and Katchalsky, A. (1958) Thermodynamic analysis of the permeability of biological membranes to non-electrolytes. Biochim. Biophys. Acta, 27, Lassalle, B., Testart, J. and Renard, J.P. (1985) Human embryo features that influence the success of cryopreservation with the use of 1,2 propanediol. Fertil. Steril., 44, Le Gal, F., Gasqui, P. and Renard, J.P. (1994) Differential osmotic behavior of mammalian oocytes before and after maturation: a quantitative analysis using goat oocytes as a model. Cryobiology, 31, Leibo, S.P. (1980) Water permeability and its activation energy of fertilized and unfertilized mouse ova. J. Membr. Biol., 53,

9 Leibo, S.P. (1986) Cryobiology: preservation of mammalian embryos. Basic Life Sci., 37, Leibo, S.P., Martino, A. and Pollard, J.W. (1995) Chilling injury of mammalian oocytes, zygotes, and embryos. [Abstr.] Cryobiology, 32, 551. Mandelbaum, J., Junca, A.M., Plachot, M. et al. (1987) Human embryo cryopreservation, extrinsic and intrinsic parameters of success. Hum. Reprod., 2, Mazur, P. (1963) Kinetics of water loss from cells at subzero temperatures and the likelihood of intracellular freezing. J. Gen. Physiol., 47, Mazur, P. and Miller, R.H. (1976) Permeability of the human erythrocyte to glycerol in 1 and 2 M solutions at 0 or 20 C. Cryobiology, 13, Mazur, P. and Schneider, U. (1986). Osmotic responses of preimplantation mouse and bovine embryos and their cryobiological implications. Cell Biophysics, 8, McGrath, J.J., Hunter, J., Bernard, J. and Fuller, B. (1992a) On the use of microdiffusion chamber methods to determine the coupled transport of water and cryoprotective agents across biological membranes. In Diller, E.K. and Shitzer, A. (eds), Macroscopic and microscopic heat and mass transfer in biomedical engineering. Elsevier Press, pp McGrath, J.J., Gao, D.Y., Tao, J. et al. (1992b) Coupled transport across the murine oocytes plasma membrane: water and cryoprotective agents. In Toner, M., Flik, M.I., Webb, D.W. et al.(eds), Topics in Heat Transfer. HTD-Vol American Society of Mechanical Engineers Press, New York, pp Miyamoto, H. and Ishibashi T. (1977) Survival of frozen thawed mouse and rat embryos in presence of ethylene glycol. J. Reprod. Fertil., 50, Myers, S.P., Lin, T.T. Pitt, R.E. and Steponkus, P.L. (1987) Cryobehavior of immature bovine oocytes. Cryo-Lett., 8, Oda, K., Gibbons, W.E. and Leibo, S.P. (1992) Osmotic shock of mouse ova. J. Reprod. Fertil., 95, Pollard, J.W. and Leibo, S.P. (1994) Chilling sensitivity of mammalian embryos. Theriogenology, 41, Rotman, B. and Papermaster, B.W. (1965) Membrane properties of living mammalian cells as studied by enzymatic hydrolysis of fluorogenic esters. Proc. Natl. Acad. Sci. USA, 55, Ruffing, N.A., Steponkus, L.P., Pitt, R.E. and Parks, J.E. (1993) Osmotic behavior, hydraulic conductivity, and incidence of intracellular ice formation in bovine oocytes at different developmental stages. Cryobiology, 30, Schneider, U. and Mazur, P. (1984) Osmotic consequences of cryoprotectant permeability and its relation to the survival of frozen thawed embryos. Theriogenology, 21, Shabana, M. and McGrath, J.J. (1988) Cryomicroscope investigation and thermodynamic modeling of freezing of unfertilized hamster ova. Cryobiology, 25, Steel, R.G.D. and Torrie, J.H. (1960) Comparisons involving two sample means and linear correlation. In Principles and Procedures of Statistics. McGraw Hill, NewYork, pp Takeda, T., Elsden, R.P. and Seidel, G.E. (1987) Use of sucrose during removal of cryoprotectants after thawing of eight cell mouse embryos. Theriogenology, 28, Testart, J., Lassalle, B., Belaisch-Allart, J. et al. (1987) Human embryo viability related to freezing and thawing procedures. Am. J. Obstet. Gynecol., 157, Whittingham, D.G., Leibo, S.P. and Mazur, P. (1972) Survival of mouse embryos frozen to 196 and 269 C. Science, 178, Willadsen, S.M., Polge, C. and Rowson, L.E.A. (1978) The viability of deep frozen cow embryos. J. Reprod. Fertil., 52, Received on March 24, 1997; accepted on October 1, 1997 Permeability characteristics of oocytes and early cleavage stage embryos 59