MOLECULAR REPRODUCTION AND DEVELOPMENT 33:276-280 (1992)
Increasing Carbon Dioxide From Five Percent to Ten Percent Improves Rabbit Blastocyst Development From Cultured Zygotes KRISTINE HALLDEN, JIANMING LI, EDWARD W. CARNEY, AND ROBERT H. FOOTE Department of Animal Science, Cornell Uniuersity, Ithaca, New York One-cell rabbit zygotes were culABSTRACT tured at 39°C in basal synthetic medium II (BSM-II) with 5%, lo%,or 15%CO, and humidified air t o determine the effect of CO, concentration on development in vitro. After 4 days in culture, 37% of the embryos grown in 10%or 15%CO, had reached the hatching blastocyst stage, but only 10%of the embryos were hatching when cultured under 5% C02 (P = 0.01). Over all blastocysts, cell numbers were 207, 246, and 205 for the 5%, lo%,and 15%CO, treatments, respectively. In a second experiment to determine if there was a beneficial effect, particularly at the blastocyst stage, of a higher concentration of CO,, embryos were cultured 4 days in either 5% or 10% CO, or for 2 days in 5% CO, followed by 2 days in 10% CO,. The numbers of blastomeres per embryo and embryo diameter were greater ( P < 0.05) in embryos cultured continuously in 10%CO, or in 10%CO, only during days 3 and 4 of culture than in embryos cultured continuously in 5% CO,. In a third experiment, one-cell rabbit zygotes were cultured with 5%or 10% COP in a defined, protein-free medium consisting of 1:l RPMl 1640 and Dulbecco's modified Eagle's medium. The proportion of embryos hatching and cell counts were significantly greater (P < 0.01)when cultured in the presence of 10%CO,. These data indicate that a 10%CO, atmosphere exerts a beneficial effect on the development of zygotes into expanding and hatching rabbit blastocysts in vitro. o 1992 Wiley-Liss, tnc.
Key Words: Expanding blastocysts, Culture, CO, INTRODUCTION There is currently a paucity of information on the CO, requirements of preimplantation embryos developing in vitro and on the role that CO, plays in the regulation of early development. Most embryo culture media are modeled after the ionic composition of blood serum (Brinster, 1971). However, studies have demonstrated that the CO, content of reproductive tract fluids may be quite different from that of serum. Vishwakarma (1962) found that total CO, and bicarbonate concentrations in oviductal and uterine fluids of rabbits in estrus are approximately twice those found in blood plasma. Similar findings have been reported by others (Bishop, 1956; David et al., 1969). Kane (1975) studied the culture requirements of rabbit embryos for bicarbonate, the base that together with
0 1992 WILEY-LISS, INC.
CO, forms the bicarbonate-CO, buffer system used in most embryo culture media. Kane found that one-cell zygotes could develop to the morula stage in media devoid of bicarbonate or CO, but that bicarbonate was required for progression to the blastocyst stage. Although culture of rabbit embryos to blastocysts in semidefined medium has been successful, since the report by Kane and Foote (19701, the optimal concentration of CO, for rabbit embryo culture has not been established. The work that has been done in other species and the CO, concentrations found in the rabbit reproductive system suggest a possible benefit to embryos from culturing them at concentrations higher than 5% CO,. In the hamster, a n approximate doubling in the percentage of eight-cell embryos developing to the blastocyst stage was observed when the concentration of CO, in the gas phase was raised from 5% to 10% (Carney and Bavister, 1987). Brinster (19711, however, cultured two-cell mouse embryos with 1%,5%, or 10% CO, and found no significant differences in development to the blastocyst stage. The present report is a study of the effect of different CO, concentrations on the in vitro development of onecell rabbit zygotes to the hatching blastocyst stage, when cultured in media with and without protein. Another objective was to examine the effect of changing the CO, concentration a s embryos progress through the stages of preimplantation development.
MATERIALS AND METHODS Animals and Embryos One-cell zygotes were collected from anesthetized sexually mature superovulated Dutch-belted does (Kennelly and Foote, 1965; Carney and Foote, 1991) 19 h r after artificial insemination and injection of luteinizing hormone by flushing oviducts with phosphatebuffered saline, pH 7.4 (PBS; Gibco, Grand Island, NY) containing 1 mg/ml polyvinyl alcohol (Cat. No. P8136; Sigma, St. Louis, MO), 60 mgiml penicillin-G (Sigma), and 50 mgiliter streptomycin sulfate (Gibco). Embryos
Received March 10,1992; accepted May 6, 1992. Address reprint requests to Dr. Robert H. Foote, Department of Animal Science, Cornell University, Ithaca, NY 14853-4801.
CARBON DIOXIDE AND RABBIT EMBRYO CULTURE recovered were washed through two changes of PBS, and embryos or unfertilized oocytes judged to be abnormal by the presence of a granular cytoplasm, irregular shape, or a thick surrounding mucin layer were discarded. The latter was indicative of premature ovulation.
277
or 15% CO, in air-gas mixtures. In experiment 2, embryos were cultured either for 4 days in 5% CO,, for 4 days in 10% CO,, or for 2 days in 5%CO, followed by 2 days in a chamber gassed with 10% CO,. In experiment 3, the RD medium was tested with 5% and 10% CO, over a 4 day culture period. All gas mixtures were certified by the manufacturer regarding CO, content. All sealed chambers were in a large incubator at 39°C. They were regassed daily.
Culture Conditions Embryos in experiments 1 and 2 were cultured at 39°C in BSM-I1 medium (Kane and Foote, 1970) containing 1.5% bovine serum albumin (Cat. No. A-9647; Evaluation of Embryo Development Sigma) and 2% antibioticiantimycotic solution (Gibco). Each day of culture in experiments 1and 2, with day Bicarbonate-, BSA-, and antibiotic-free stock solutions 0 defined as the day of embryo collection, embryos were were prepared separately from powdered media (Gibco; evaluated according to stage of development and apformula 88-5021EC) by hydration with ultrapure water pearance, and the diameters were measured. The latter (Milli-Q system, 18 MegOhm; Millipore, Bedford, MA) was done because of previous studies indicating that and were stored for up to 2 months a t 4°C. The complete CO, affected blastocyst formation and because rabbit medium with BSM-I1 was prepared the evening before blastocysts expand rapidly. Total time to evaluate emuse, and 200 p.1 drops were dispensed into sterile 35 x bryos and to refill the chambers with gas mixture was 10 mm dishes (Falcon, Oxnard, CAI, each drop overlaid typically 15 min. Embryos in experiments 1and 2 were with 4 ml of Dow 360 Medical Fluid, viscosity 260 cen- given quality scores in 0.5 increments, ranging from 1.0 tistokes (Dow Corning, Midland, MI). Media were for normal embryos to 4.0 for degenerate ones. This equilibrated overnight at 39°C in whatever gas mix- daily attempt to quantify the degree of degeneration tures were used in each experiment. For culture, a max- was found later, upon statistical analysis, to add little imum of five to ten embryos were placed in each drop. information in predicting blastocyst development, so After experiments 1and 2 were completed, a protein- this was discontinued in experiment 3. Also, intrazonal free medium giving improved results for culturing rab- embryo diameters (Carney and Foote, 1990) were meabit zygotes was reported by Carney and Foote (1991). sured in experiments l and 2, using a n inverted microThis medium consisted of a 1 : l mixture of RPMI 1640 scope fitted with a calibrated eyepiece micrometer. with Dulbecco’s MEM (Gibco) and was designated as Hatching blastocysts had a substantial number of cells RD medium. The RD medium was the same as de- extruded through the zona pellucida, giving a n elonscribed by Carney and Foote (1991), except that in the gated rather than a spherical shape. Measurements present study the low-glucose modification of Dulbec- were made across the longest axis. Because cell counts CO’SMEM was used. The pH of the medium was ad- were considered to be a more useful indicator of embryo justed at 25°C to 7.4 with 0.1 N HC1. Groups of five to 15 development than diameter, diameters were not meaone-cell zygotes per donor rabbit were distributed ran- sured in experiment 3. domly into four-well dishes (Nunc, Kamstrup, DenImmediately following any measurements of fresh mark) containing 0.5 ml of medium. No oil or medical embryos on day 4, development was terminated in all fluid was used. The experiment was replicated with six experiments by transferring the embryos into drops donor rabbits, and each day zygotes from a single donor containing 1%formalin. The number of blastomeres were divided equally among treatments. per embryo was then determined by staining embryos Individual screw-cap, 500 ml polymethylpentene for 3 min with the DNA-specific fluorochrome Hoechst plastic jars (Nalgene, Rochester, NY), fitted with gas 33342 (10 p.g/ml), using a 3 min prestain with 0.1% infusion holes and a plug in the lid, were effectively trypan blue to quench fluorescence of sperm heads emutilized as inexpensive incubation chambers for all ex- bedded in the mucin coat (Purse1 et al., 1985). Accurate periments. They were washed with 70% ethanol prior to counting of nuclei was made possible by use of a n ocular each use. Chambers containing media with embryos grid. were gassed for 5 min by inserting a rubber tube attached to a cylinder containing the appropriate gas Statistical Analysis mixture (Matheson, Secaucus, NJ). The tubes were fitAll experiments utilized a randomized complete ted with 0.20 p.m pore diameter filters and washed with 70% ethanol prior to use a s a precaution against micro- block design, with the embryos from each donor reprebial contamination. The chambers contained dishes of senting one complete block. This was done to control for ultrapure water to provide humidity and a bicarbonatel embryo donor (rabbit)effect, which usually was signifiphenol red pH indicator solution as a test for gas reten- cant. Each block was replicated with a t least six donor tion. Threads of the screw cap were lubricated with a rabbits. One-cell zygotes from each superovulated dosmall amount of petroleum jelly to aid in sealing and nor were divided equally and randomly assigned to treatment groups. Any misclassified one-cell zygotes, opening the chambers. In experiment 1,embryos were cultured for exactly 4 as determined later by nuclear staining to be unfertildays in separate chambers containing either 5%, lo%, ized ova, were excluded from the analyses.
278
K. HALLDEN ET AL. TABLE 1. Proportion of Embryos Developing Into Blastocysts, Their Diameters, and Cell Numbers Achieved by Day 4 of Culture Under Various Concentrations of CO, All Became Hatching blastocysts blastocysts No. of blastocysts Diameter Cell (cell embryos (%) Percent (km) number number) Treatment Experiment 1 5% co, 10% co, 15% CO,
89 88 89
87 90 92
10" 37b 37b
139" 157b 160b
208" 281b 229"
207" 246b 205"
5% co, 5-10% CO, 10%
79 76 81
91 86 83
31" 43b 37a.b
250" 280b 265",b
258" 316b 361b
240" 295b 298b
Experiment 2
",bMeanswith different superscripts within columns and experiments differ ( P < 0.05).
Two-way analysis of variance was performed, using the general linear mixed models procedure of the Statistical Analysis System (SAS Institute, Cary, NC). The arcsine transformation of the proportion of blastocysts formed in each droplet or well of media (rabbit x treatment) was used for the percentage data analysis, and diameters and cell numbers for each embryo were transformed logarithmically. These were all considered to be statistically significant at P < 0.05. Rabbits and culture treatments were the main variables in the twoway analysis of variance with the interaction mean square used to test treatment effects. When analysis of variance indicated significant mean differences, these were compared using Duncan's new multiple range test (Snedecor and Cochran, 1980).
RESULTS Experiments 1 and 2: Embryo Development in BSM-I1 Medium Donor rabbits were a substantial source of variation in each experiment (P < 0.01) and this was accounted for in the analysis of variance. The same males were used repeatedly, but donor females could be used only once. The major results of CO, treatment in experiments 1 and 2 are summarized in Table 1. For brevity and because in all treatment groups a high proportion of the embryos developed into blastocysts (83-92%), only embryo development a t the end of culture (4days) is reported. Diameters and cell counts of hatching embryos and cell counts of all embryos were greater in the presence of 10% CO, than 5% CO, (P < 0.05). Also the proportion of embryos with cells migrating through the zona pellucida (hatching blastocysts) generally was greater for those cultured in 10% than in 5% CO, ( P < 0.05). The diameters of early blastocysts before expansion were similar among treatments, as expected, so these data are not presented. Because 10% CO, was superior to 15% CO, in experiment 1, the latter CO, concentration was not studied further. In experiment 2, supplying 10% CO, after the first 2 days of culture gave results equivalent to those obtained with continuous exposure to 10% CO,.
The differences in average cell number over all blastocysts (Table 1)reflect the higher number of cells in the hatching blastocysts with a greater proportion of hatching blastocysts in the 10% CO, groups than in the 5%groups. The differences between experiments 1 and 2 in diameter of embryos and cell numbers simply reflects differences in potential for development among one-cell zygotes obtained from different donors used in the two experiments. Detailed results of the quality scores are not presented, as they showed a pattern similar to the cell number. The lowest scores (best quality) were obtained for the 10% CO, treatment in both experiments 1and 2.
Experiment 3: Embryo Development in RD Medium All embryos developed into either morula or blastocysts (Table 2) and nearly all developed into blastocysts in this protein-free medium. A higher proportion of embryos was hatching in the presence of 10% CO, than in 5% CO, (P < 0.011, with 58% hatching when cultured with 10% CO,. Cell number per embryo cultured with 10% CO, also exceeded the number of cells with 5%CO, (P < 0.01). The number of cells per embryo in the two treatments reflects values similar to those in preceding experiments. DISCUSSION In experiment 1 (Table 11, the 37% hatching rate for embryos cultured in 10% and 15% CO, concentrations was superior (P < 0.051, being almost four times higher than in the 5% CO, environment. No difference was found in the percent that became blastocysts (Table l), suggesting a possible beneficial effect of a higher pC0, for blastocyst growth. This possibility was explored further in experiment 2 in which the CO, concentration was increased from 5% to 10% at the time embryos began to change from morulae to blastocysts. This treatment brought about a significant increase in hatching, although the increase over the results at 5% CO, was not as striking a s in experiment 1,because of the more robust growth of embryos in the presence of
CARBON DIOXIDE AND RABBIT EMBRYO CULTURE
279
TABLE 2. Endpoints After Culturing Rabbit Zygotes for 4 Days in RD Medium With 5%and 10%C0,:Air Mixtures
CO, concentration 5% 10%
No. of embryos 46 48
Embryo development, no. (%I Morulae Expanding Hatching and early blastoblastoblastocysts cysts cysts 15 (33)" 5 (lop
19 (41)" 15 (31)"
12 (26)" 28 (581b
Cell number all blastocy sts" 203" 297b
a*bNumberswithin columns with different superscripts differ ( P < 0.05). "Subsets of 31 and 30 embryos from the 5%and 10%CO, groups, respectively, were stained and counted.
5% C 0 2 in experiment 2. These differences occur despite random sampling from a large colony of rabbits and considerable replication within experiment, but comparable controls were included in each experiment. Also, experiments were conducted in different seasons. Possibly seasonal effects occurred, although light and temperature were controlled. Embryos cultured in the protein-free RD medium in experiment 3 again were more advanced in development in the presence of 10% compared with 5% COP (Table 2). Based on the development into hatching blastocysts in 5% CO, the embryos in experiments 1 and 3 appear to be comparable. The 58% hatching blastocysts in the 10% CO, treatment also is higher than the 40% reported for embryos cultured in RD medium by Carney and Foote (1991) in the presence of 5% CO,. Glucose concentration also was lower in experiment 3 than was used in the earlier study (Carney and Foote, 1991). Cell number data (Tables 1 and 2) from all experiments support the hypothesis that 10% CO, enhances development of rabbit embryos in vitro compared with 5% CO,. It appears that the beneficial effect for elevated pC0, may be related to requirements for blastocyst expansion; embryos cultured either continuously or discontinuously in 10% CO, each had significantly (P < 0.05) higher cell counts than controls for both hatching and all blastocysts combined. The greater cell numbers of embryos cultured in 10%CO, more closely approximate in vivo blastomere numbers, based on extrapolation from a previous study (Carney and Foote, 1990) with 5% CO,. However, even under these improved conditions, cell numbers in vitro are less than those in vivo (Carney and Foote, 1990; Alliston and Pardee, 1973). Blastocyst diameter tended to be greater with 10%C02, but again expansion in vitro was less than the dramatic expansion that occurs in vivo (Adams, 1958; Daniel, 1964). Hatching is partly a n artifact of culture (Kane, 1974), but it is usually associated with a n increase in cell number and accumulation of blastocoele fluid (Kane, 1987), as was generally true here. For this reason, hatching in culture, especially in a protein-free medium, is considered a useful indicator of the general health of embryos in culture. Remodeling of the zona pellucida may occur from within a s the neozona starts
to replace the zona pellucida by day 4 in the rabbit (Denker and Gerdes, 1979). The exact mechanism by which CO, influences rabbit embryo development is not known. In a variety of cell types, CO, has been shown to be a potent regulator of glycolysis and amino acid and pyrimidine synthesis and participates in many other metabolic activities (McLimans, 1972). Graves and Biggers (1970) reported uptake of radioactive bicarbonate into proteins in all stages of preimplantation mouse embryos, with the highest incorporation in early blastocysts. Quinn and Wales (1974) reported that CO, fixation by rabbit embryos was much less than that by mouse embryos. The rate of CO, fixation steadily increased as embryos progressed from the one-cell stage to the blastocyst stage. Much of the CO, was incorporated into aspartic and glutamic acids, both of which were supplied in the media used in the present experiments. In addition to its metabolic functions, CO, is also a n important regulator of intracellular pH. In preimplantation hamster embryos, the stimulatory effects of elevated CO, concentrations appeared to be due to its action as a weak acid in regulating intracellular pH; the effects of CO, could be mimicked by the nonmetabolizable membrane-permeable weak acid 2,4-dimethyloxazolidonedione(Carney and Bavister, 1987). CO, also acts to buffer pH of the culture medium, but the lowest level of CO, (5%)used in present studies is more than adequate to satisfy this function. Embryo metabolism contributes negligibly to changes in pH of media due to the large volume of medium relative to embryo volume. The extremes found in the present experiments of culture media under continuous equilibration with 5% and 10% CO, were 7.2 to 7.4. Kane (1975) has demonstrated normal rabbit blastocyst development over a much wider pH range, including hatching in BSM-I1 over a pH range of 6.9-7.9. Thus, it is doubtful that the CO, effect was due simply to pH. The higher CO, concentrations used may actually not be high, but rather physiological. The pC02 of rabbit oviduct and uterus has been reported to be approximately twice the concentration found in blood plasma (Vishwakarma, 1962). This would make the 10% concentration of CO, (-80 mm Hg) used in this study a close approximation of conditions in vivo. In summary,
280
K. HALLDEN ET AL.
concentrations of C 0 2 higher than the 5% commonly used in practice, particularly 10% C02, resulted in higher frequencies of hatching rabbit blastocysts, and these hatching embryos were larger and contained more cells per blastocyst.
ACKNOWLEDGMENTS The authors thank M. Simkin, C. Tobback, and E. Wilson for surgical assistance. The LH was kindly supplied by Vetrepharm, Inc. This work was supported by a Multicenter Cooperative Program on Non-Human In Vitro Fertilization and Preimplantation Development and was funded by the National Institute of Child Health and Human Development, NIH, through Cooperative Agreement HD21939. Additional funding for honors research for K.H. was provided by the College of Agriculture and Life Sciences, Cornell University. REFERENCES Adams CE (1958): Egg development in the rabbit. The influence of postcoital ligation of the uterine tube and of ovariectomy. J Endocrinol 16:283-293. Alliston CW, Pardee NR (1973):Variability of embryonic development in the rabbit at 19 to 168 hours after mating. Lab Anim Sci 23:665670. Bishop DW (1956):Active secretion in the rabbit oviduct. Am J Physiol 187:347-352. Brinster RL (1971): In vitro culture of the embryo. In Sherman A (ed): “Pathways to Conception.” Springfield, IL: Charles C Thomas, pp 245-277. Carney EW, Bavister BD (1987):Regulation of hamster embryo development in vitro by carbon dioxide. Biol Reprod 36:1155-1163: Carney EW, Foote RH (1990):Effects of superovulation, embryo recovery, culture system and embryo transfer on development of rabbit embryos in vivo and in vitro. J Reprod Fertil89:543-551. Carney EW, Foote RH (1991):Improved development of rabbit one-cell
embryos to the hatching blastocyst stage by culture in a defined, protein-free culture medium. J Reprod Fertil91:113-123. Daniel J C Jr (1964): Early growth of rabbit trophoblast. Am Nat 98:85-96. David A, Brackett BG, Garcia C-R, Mastroianni L Jr (1969): Composition of rabbit oviduct fluid in ligated segments of the fallopian tube. J Reprod Fertil19:285-289. Denker H-W, Gerdes H-J (1979): The dynamic structure of rabbit blastocyst coverings. I: Transformation during regular preimplantation development. Anat Embryo1 157:15-34. Graves CN, Biggers JF (1970): Carbon dioxide fixation by mouse embryos prior to implantation. Science 167:150&1508. Kane MT (1974):The effects of pH on culture of one-cell rabbit ova to blastocysts in bicarbonate-buffered medium. J Reprod Fertil 38:477-480. Kane MT (1975): Bicarbonate requirements of culture of one-cell rabbit ova to blastocysts. Biol Reprod 12:552-555. Kane MT (1987): In vitro growth of preimplantation rabbit embryos. In Bavister BD (ed): “The Mammalian Preimplantation Embryo: Regulation of Growth and Differentiation in Vitro.” New York: Plenum Publishers, pp 193-218. Kane MT, Foote RH (1970): Culture of two- and four-cell rabbit embryos to the expanding blastocyst stage in synthetic media. Proc SOC Exp Biol Med 133:921-925. Kennelly JJ, Foote RH (1965): Superovulatory response to pre- and postpubertal rabbits to commercially available gonadotrophins. J Reprod Fertil9:177-188. McLimans WF (1972): The gaseous environment of the mammalian cell in culture. In Rothblat GH, Cristofalo VJ (ed):“Growth, Nutrition and Metabolism of Cells in Culture.” New York: Academic Press, pp 138-170. Purse1 VG, Wall RJ, Rexroad CE J r , Hammer RE, Brinster RL (1985): A rapid whole-mount staining procedure for nuclei of mammalian embryos. Theriogenology 24:687-700. Quinn P, Wales RG (1974):Fixation ofcarbon dioxide by preimplantation rabbit embryos in vitro. J Reprod Fertil 36:29-39. Snedecor GW, Cochran WG (1980): “Statistical Methods.” Ames, IA: Iowa State University Press, pp 233-236. Vishwakarma P (1962): The pH and bicarbonate-ion content of the oviduct and uterine fluids. Fertil Steril 13:481-485.
Increasing Carbon Dioxide From Five Percent to Ten Percent Improves Rabbit Blastocyst Development From Cultured Zygotes KRISTINE HALLDEN, JIANMING LI, EDWARD W. CARNEY, AND ROBERT H. FOOTE Department of Animal Science, Cornell Uniuersity, Ithaca, New York One-cell rabbit zygotes were culABSTRACT tured at 39°C in basal synthetic medium II (BSM-II) with 5%, lo%,or 15%CO, and humidified air t o determine the effect of CO, concentration on development in vitro. After 4 days in culture, 37% of the embryos grown in 10%or 15%CO, had reached the hatching blastocyst stage, but only 10%of the embryos were hatching when cultured under 5% C02 (P = 0.01). Over all blastocysts, cell numbers were 207, 246, and 205 for the 5%, lo%,and 15%CO, treatments, respectively. In a second experiment to determine if there was a beneficial effect, particularly at the blastocyst stage, of a higher concentration of CO,, embryos were cultured 4 days in either 5% or 10% CO, or for 2 days in 5% CO, followed by 2 days in 10% CO,. The numbers of blastomeres per embryo and embryo diameter were greater ( P < 0.05) in embryos cultured continuously in 10%CO, or in 10%CO, only during days 3 and 4 of culture than in embryos cultured continuously in 5% CO,. In a third experiment, one-cell rabbit zygotes were cultured with 5%or 10% COP in a defined, protein-free medium consisting of 1:l RPMl 1640 and Dulbecco's modified Eagle's medium. The proportion of embryos hatching and cell counts were significantly greater (P < 0.01)when cultured in the presence of 10%CO,. These data indicate that a 10%CO, atmosphere exerts a beneficial effect on the development of zygotes into expanding and hatching rabbit blastocysts in vitro. o 1992 Wiley-Liss, tnc.
Key Words: Expanding blastocysts, Culture, CO, INTRODUCTION There is currently a paucity of information on the CO, requirements of preimplantation embryos developing in vitro and on the role that CO, plays in the regulation of early development. Most embryo culture media are modeled after the ionic composition of blood serum (Brinster, 1971). However, studies have demonstrated that the CO, content of reproductive tract fluids may be quite different from that of serum. Vishwakarma (1962) found that total CO, and bicarbonate concentrations in oviductal and uterine fluids of rabbits in estrus are approximately twice those found in blood plasma. Similar findings have been reported by others (Bishop, 1956; David et al., 1969). Kane (1975) studied the culture requirements of rabbit embryos for bicarbonate, the base that together with
0 1992 WILEY-LISS, INC.
CO, forms the bicarbonate-CO, buffer system used in most embryo culture media. Kane found that one-cell zygotes could develop to the morula stage in media devoid of bicarbonate or CO, but that bicarbonate was required for progression to the blastocyst stage. Although culture of rabbit embryos to blastocysts in semidefined medium has been successful, since the report by Kane and Foote (19701, the optimal concentration of CO, for rabbit embryo culture has not been established. The work that has been done in other species and the CO, concentrations found in the rabbit reproductive system suggest a possible benefit to embryos from culturing them at concentrations higher than 5% CO,. In the hamster, a n approximate doubling in the percentage of eight-cell embryos developing to the blastocyst stage was observed when the concentration of CO, in the gas phase was raised from 5% to 10% (Carney and Bavister, 1987). Brinster (19711, however, cultured two-cell mouse embryos with 1%,5%, or 10% CO, and found no significant differences in development to the blastocyst stage. The present report is a study of the effect of different CO, concentrations on the in vitro development of onecell rabbit zygotes to the hatching blastocyst stage, when cultured in media with and without protein. Another objective was to examine the effect of changing the CO, concentration a s embryos progress through the stages of preimplantation development.
MATERIALS AND METHODS Animals and Embryos One-cell zygotes were collected from anesthetized sexually mature superovulated Dutch-belted does (Kennelly and Foote, 1965; Carney and Foote, 1991) 19 h r after artificial insemination and injection of luteinizing hormone by flushing oviducts with phosphatebuffered saline, pH 7.4 (PBS; Gibco, Grand Island, NY) containing 1 mg/ml polyvinyl alcohol (Cat. No. P8136; Sigma, St. Louis, MO), 60 mgiml penicillin-G (Sigma), and 50 mgiliter streptomycin sulfate (Gibco). Embryos
Received March 10,1992; accepted May 6, 1992. Address reprint requests to Dr. Robert H. Foote, Department of Animal Science, Cornell University, Ithaca, NY 14853-4801.
CARBON DIOXIDE AND RABBIT EMBRYO CULTURE recovered were washed through two changes of PBS, and embryos or unfertilized oocytes judged to be abnormal by the presence of a granular cytoplasm, irregular shape, or a thick surrounding mucin layer were discarded. The latter was indicative of premature ovulation.
277
or 15% CO, in air-gas mixtures. In experiment 2, embryos were cultured either for 4 days in 5% CO,, for 4 days in 10% CO,, or for 2 days in 5%CO, followed by 2 days in a chamber gassed with 10% CO,. In experiment 3, the RD medium was tested with 5% and 10% CO, over a 4 day culture period. All gas mixtures were certified by the manufacturer regarding CO, content. All sealed chambers were in a large incubator at 39°C. They were regassed daily.
Culture Conditions Embryos in experiments 1 and 2 were cultured at 39°C in BSM-I1 medium (Kane and Foote, 1970) containing 1.5% bovine serum albumin (Cat. No. A-9647; Evaluation of Embryo Development Sigma) and 2% antibioticiantimycotic solution (Gibco). Each day of culture in experiments 1and 2, with day Bicarbonate-, BSA-, and antibiotic-free stock solutions 0 defined as the day of embryo collection, embryos were were prepared separately from powdered media (Gibco; evaluated according to stage of development and apformula 88-5021EC) by hydration with ultrapure water pearance, and the diameters were measured. The latter (Milli-Q system, 18 MegOhm; Millipore, Bedford, MA) was done because of previous studies indicating that and were stored for up to 2 months a t 4°C. The complete CO, affected blastocyst formation and because rabbit medium with BSM-I1 was prepared the evening before blastocysts expand rapidly. Total time to evaluate emuse, and 200 p.1 drops were dispensed into sterile 35 x bryos and to refill the chambers with gas mixture was 10 mm dishes (Falcon, Oxnard, CAI, each drop overlaid typically 15 min. Embryos in experiments 1and 2 were with 4 ml of Dow 360 Medical Fluid, viscosity 260 cen- given quality scores in 0.5 increments, ranging from 1.0 tistokes (Dow Corning, Midland, MI). Media were for normal embryos to 4.0 for degenerate ones. This equilibrated overnight at 39°C in whatever gas mix- daily attempt to quantify the degree of degeneration tures were used in each experiment. For culture, a max- was found later, upon statistical analysis, to add little imum of five to ten embryos were placed in each drop. information in predicting blastocyst development, so After experiments 1and 2 were completed, a protein- this was discontinued in experiment 3. Also, intrazonal free medium giving improved results for culturing rab- embryo diameters (Carney and Foote, 1990) were meabit zygotes was reported by Carney and Foote (1991). sured in experiments l and 2, using a n inverted microThis medium consisted of a 1 : l mixture of RPMI 1640 scope fitted with a calibrated eyepiece micrometer. with Dulbecco’s MEM (Gibco) and was designated as Hatching blastocysts had a substantial number of cells RD medium. The RD medium was the same as de- extruded through the zona pellucida, giving a n elonscribed by Carney and Foote (1991), except that in the gated rather than a spherical shape. Measurements present study the low-glucose modification of Dulbec- were made across the longest axis. Because cell counts CO’SMEM was used. The pH of the medium was ad- were considered to be a more useful indicator of embryo justed at 25°C to 7.4 with 0.1 N HC1. Groups of five to 15 development than diameter, diameters were not meaone-cell zygotes per donor rabbit were distributed ran- sured in experiment 3. domly into four-well dishes (Nunc, Kamstrup, DenImmediately following any measurements of fresh mark) containing 0.5 ml of medium. No oil or medical embryos on day 4, development was terminated in all fluid was used. The experiment was replicated with six experiments by transferring the embryos into drops donor rabbits, and each day zygotes from a single donor containing 1%formalin. The number of blastomeres were divided equally among treatments. per embryo was then determined by staining embryos Individual screw-cap, 500 ml polymethylpentene for 3 min with the DNA-specific fluorochrome Hoechst plastic jars (Nalgene, Rochester, NY), fitted with gas 33342 (10 p.g/ml), using a 3 min prestain with 0.1% infusion holes and a plug in the lid, were effectively trypan blue to quench fluorescence of sperm heads emutilized as inexpensive incubation chambers for all ex- bedded in the mucin coat (Purse1 et al., 1985). Accurate periments. They were washed with 70% ethanol prior to counting of nuclei was made possible by use of a n ocular each use. Chambers containing media with embryos grid. were gassed for 5 min by inserting a rubber tube attached to a cylinder containing the appropriate gas Statistical Analysis mixture (Matheson, Secaucus, NJ). The tubes were fitAll experiments utilized a randomized complete ted with 0.20 p.m pore diameter filters and washed with 70% ethanol prior to use a s a precaution against micro- block design, with the embryos from each donor reprebial contamination. The chambers contained dishes of senting one complete block. This was done to control for ultrapure water to provide humidity and a bicarbonatel embryo donor (rabbit)effect, which usually was signifiphenol red pH indicator solution as a test for gas reten- cant. Each block was replicated with a t least six donor tion. Threads of the screw cap were lubricated with a rabbits. One-cell zygotes from each superovulated dosmall amount of petroleum jelly to aid in sealing and nor were divided equally and randomly assigned to treatment groups. Any misclassified one-cell zygotes, opening the chambers. In experiment 1,embryos were cultured for exactly 4 as determined later by nuclear staining to be unfertildays in separate chambers containing either 5%, lo%, ized ova, were excluded from the analyses.
278
K. HALLDEN ET AL. TABLE 1. Proportion of Embryos Developing Into Blastocysts, Their Diameters, and Cell Numbers Achieved by Day 4 of Culture Under Various Concentrations of CO, All Became Hatching blastocysts blastocysts No. of blastocysts Diameter Cell (cell embryos (%) Percent (km) number number) Treatment Experiment 1 5% co, 10% co, 15% CO,
89 88 89
87 90 92
10" 37b 37b
139" 157b 160b
208" 281b 229"
207" 246b 205"
5% co, 5-10% CO, 10%
79 76 81
91 86 83
31" 43b 37a.b
250" 280b 265",b
258" 316b 361b
240" 295b 298b
Experiment 2
",bMeanswith different superscripts within columns and experiments differ ( P < 0.05).
Two-way analysis of variance was performed, using the general linear mixed models procedure of the Statistical Analysis System (SAS Institute, Cary, NC). The arcsine transformation of the proportion of blastocysts formed in each droplet or well of media (rabbit x treatment) was used for the percentage data analysis, and diameters and cell numbers for each embryo were transformed logarithmically. These were all considered to be statistically significant at P < 0.05. Rabbits and culture treatments were the main variables in the twoway analysis of variance with the interaction mean square used to test treatment effects. When analysis of variance indicated significant mean differences, these were compared using Duncan's new multiple range test (Snedecor and Cochran, 1980).
RESULTS Experiments 1 and 2: Embryo Development in BSM-I1 Medium Donor rabbits were a substantial source of variation in each experiment (P < 0.01) and this was accounted for in the analysis of variance. The same males were used repeatedly, but donor females could be used only once. The major results of CO, treatment in experiments 1 and 2 are summarized in Table 1. For brevity and because in all treatment groups a high proportion of the embryos developed into blastocysts (83-92%), only embryo development a t the end of culture (4days) is reported. Diameters and cell counts of hatching embryos and cell counts of all embryos were greater in the presence of 10% CO, than 5% CO, (P < 0.05). Also the proportion of embryos with cells migrating through the zona pellucida (hatching blastocysts) generally was greater for those cultured in 10% than in 5% CO, ( P < 0.05). The diameters of early blastocysts before expansion were similar among treatments, as expected, so these data are not presented. Because 10% CO, was superior to 15% CO, in experiment 1, the latter CO, concentration was not studied further. In experiment 2, supplying 10% CO, after the first 2 days of culture gave results equivalent to those obtained with continuous exposure to 10% CO,.
The differences in average cell number over all blastocysts (Table 1)reflect the higher number of cells in the hatching blastocysts with a greater proportion of hatching blastocysts in the 10% CO, groups than in the 5%groups. The differences between experiments 1 and 2 in diameter of embryos and cell numbers simply reflects differences in potential for development among one-cell zygotes obtained from different donors used in the two experiments. Detailed results of the quality scores are not presented, as they showed a pattern similar to the cell number. The lowest scores (best quality) were obtained for the 10% CO, treatment in both experiments 1and 2.
Experiment 3: Embryo Development in RD Medium All embryos developed into either morula or blastocysts (Table 2) and nearly all developed into blastocysts in this protein-free medium. A higher proportion of embryos was hatching in the presence of 10% CO, than in 5% CO, (P < 0.011, with 58% hatching when cultured with 10% CO,. Cell number per embryo cultured with 10% CO, also exceeded the number of cells with 5%CO, (P < 0.01). The number of cells per embryo in the two treatments reflects values similar to those in preceding experiments. DISCUSSION In experiment 1 (Table 11, the 37% hatching rate for embryos cultured in 10% and 15% CO, concentrations was superior (P < 0.051, being almost four times higher than in the 5% CO, environment. No difference was found in the percent that became blastocysts (Table l), suggesting a possible beneficial effect of a higher pC0, for blastocyst growth. This possibility was explored further in experiment 2 in which the CO, concentration was increased from 5% to 10% at the time embryos began to change from morulae to blastocysts. This treatment brought about a significant increase in hatching, although the increase over the results at 5% CO, was not as striking a s in experiment 1,because of the more robust growth of embryos in the presence of
CARBON DIOXIDE AND RABBIT EMBRYO CULTURE
279
TABLE 2. Endpoints After Culturing Rabbit Zygotes for 4 Days in RD Medium With 5%and 10%C0,:Air Mixtures
CO, concentration 5% 10%
No. of embryos 46 48
Embryo development, no. (%I Morulae Expanding Hatching and early blastoblastoblastocysts cysts cysts 15 (33)" 5 (lop
19 (41)" 15 (31)"
12 (26)" 28 (581b
Cell number all blastocy sts" 203" 297b
a*bNumberswithin columns with different superscripts differ ( P < 0.05). "Subsets of 31 and 30 embryos from the 5%and 10%CO, groups, respectively, were stained and counted.
5% C 0 2 in experiment 2. These differences occur despite random sampling from a large colony of rabbits and considerable replication within experiment, but comparable controls were included in each experiment. Also, experiments were conducted in different seasons. Possibly seasonal effects occurred, although light and temperature were controlled. Embryos cultured in the protein-free RD medium in experiment 3 again were more advanced in development in the presence of 10% compared with 5% COP (Table 2). Based on the development into hatching blastocysts in 5% CO, the embryos in experiments 1 and 3 appear to be comparable. The 58% hatching blastocysts in the 10% CO, treatment also is higher than the 40% reported for embryos cultured in RD medium by Carney and Foote (1991) in the presence of 5% CO,. Glucose concentration also was lower in experiment 3 than was used in the earlier study (Carney and Foote, 1991). Cell number data (Tables 1 and 2) from all experiments support the hypothesis that 10% CO, enhances development of rabbit embryos in vitro compared with 5% CO,. It appears that the beneficial effect for elevated pC0, may be related to requirements for blastocyst expansion; embryos cultured either continuously or discontinuously in 10% CO, each had significantly (P < 0.05) higher cell counts than controls for both hatching and all blastocysts combined. The greater cell numbers of embryos cultured in 10%CO, more closely approximate in vivo blastomere numbers, based on extrapolation from a previous study (Carney and Foote, 1990) with 5% CO,. However, even under these improved conditions, cell numbers in vitro are less than those in vivo (Carney and Foote, 1990; Alliston and Pardee, 1973). Blastocyst diameter tended to be greater with 10%C02, but again expansion in vitro was less than the dramatic expansion that occurs in vivo (Adams, 1958; Daniel, 1964). Hatching is partly a n artifact of culture (Kane, 1974), but it is usually associated with a n increase in cell number and accumulation of blastocoele fluid (Kane, 1987), as was generally true here. For this reason, hatching in culture, especially in a protein-free medium, is considered a useful indicator of the general health of embryos in culture. Remodeling of the zona pellucida may occur from within a s the neozona starts
to replace the zona pellucida by day 4 in the rabbit (Denker and Gerdes, 1979). The exact mechanism by which CO, influences rabbit embryo development is not known. In a variety of cell types, CO, has been shown to be a potent regulator of glycolysis and amino acid and pyrimidine synthesis and participates in many other metabolic activities (McLimans, 1972). Graves and Biggers (1970) reported uptake of radioactive bicarbonate into proteins in all stages of preimplantation mouse embryos, with the highest incorporation in early blastocysts. Quinn and Wales (1974) reported that CO, fixation by rabbit embryos was much less than that by mouse embryos. The rate of CO, fixation steadily increased as embryos progressed from the one-cell stage to the blastocyst stage. Much of the CO, was incorporated into aspartic and glutamic acids, both of which were supplied in the media used in the present experiments. In addition to its metabolic functions, CO, is also a n important regulator of intracellular pH. In preimplantation hamster embryos, the stimulatory effects of elevated CO, concentrations appeared to be due to its action as a weak acid in regulating intracellular pH; the effects of CO, could be mimicked by the nonmetabolizable membrane-permeable weak acid 2,4-dimethyloxazolidonedione(Carney and Bavister, 1987). CO, also acts to buffer pH of the culture medium, but the lowest level of CO, (5%)used in present studies is more than adequate to satisfy this function. Embryo metabolism contributes negligibly to changes in pH of media due to the large volume of medium relative to embryo volume. The extremes found in the present experiments of culture media under continuous equilibration with 5% and 10% CO, were 7.2 to 7.4. Kane (1975) has demonstrated normal rabbit blastocyst development over a much wider pH range, including hatching in BSM-I1 over a pH range of 6.9-7.9. Thus, it is doubtful that the CO, effect was due simply to pH. The higher CO, concentrations used may actually not be high, but rather physiological. The pC02 of rabbit oviduct and uterus has been reported to be approximately twice the concentration found in blood plasma (Vishwakarma, 1962). This would make the 10% concentration of CO, (-80 mm Hg) used in this study a close approximation of conditions in vivo. In summary,
280
K. HALLDEN ET AL.
concentrations of C 0 2 higher than the 5% commonly used in practice, particularly 10% C02, resulted in higher frequencies of hatching rabbit blastocysts, and these hatching embryos were larger and contained more cells per blastocyst.
ACKNOWLEDGMENTS The authors thank M. Simkin, C. Tobback, and E. Wilson for surgical assistance. The LH was kindly supplied by Vetrepharm, Inc. This work was supported by a Multicenter Cooperative Program on Non-Human In Vitro Fertilization and Preimplantation Development and was funded by the National Institute of Child Health and Human Development, NIH, through Cooperative Agreement HD21939. Additional funding for honors research for K.H. was provided by the College of Agriculture and Life Sciences, Cornell University. REFERENCES Adams CE (1958): Egg development in the rabbit. The influence of postcoital ligation of the uterine tube and of ovariectomy. J Endocrinol 16:283-293. Alliston CW, Pardee NR (1973):Variability of embryonic development in the rabbit at 19 to 168 hours after mating. Lab Anim Sci 23:665670. Bishop DW (1956):Active secretion in the rabbit oviduct. Am J Physiol 187:347-352. Brinster RL (1971): In vitro culture of the embryo. In Sherman A (ed): “Pathways to Conception.” Springfield, IL: Charles C Thomas, pp 245-277. Carney EW, Bavister BD (1987):Regulation of hamster embryo development in vitro by carbon dioxide. Biol Reprod 36:1155-1163: Carney EW, Foote RH (1990):Effects of superovulation, embryo recovery, culture system and embryo transfer on development of rabbit embryos in vivo and in vitro. J Reprod Fertil89:543-551. Carney EW, Foote RH (1991):Improved development of rabbit one-cell
embryos to the hatching blastocyst stage by culture in a defined, protein-free culture medium. J Reprod Fertil91:113-123. Daniel J C Jr (1964): Early growth of rabbit trophoblast. Am Nat 98:85-96. David A, Brackett BG, Garcia C-R, Mastroianni L Jr (1969): Composition of rabbit oviduct fluid in ligated segments of the fallopian tube. J Reprod Fertil19:285-289. Denker H-W, Gerdes H-J (1979): The dynamic structure of rabbit blastocyst coverings. I: Transformation during regular preimplantation development. Anat Embryo1 157:15-34. Graves CN, Biggers JF (1970): Carbon dioxide fixation by mouse embryos prior to implantation. Science 167:150&1508. Kane MT (1974):The effects of pH on culture of one-cell rabbit ova to blastocysts in bicarbonate-buffered medium. J Reprod Fertil 38:477-480. Kane MT (1975): Bicarbonate requirements of culture of one-cell rabbit ova to blastocysts. Biol Reprod 12:552-555. Kane MT (1987): In vitro growth of preimplantation rabbit embryos. In Bavister BD (ed): “The Mammalian Preimplantation Embryo: Regulation of Growth and Differentiation in Vitro.” New York: Plenum Publishers, pp 193-218. Kane MT, Foote RH (1970): Culture of two- and four-cell rabbit embryos to the expanding blastocyst stage in synthetic media. Proc SOC Exp Biol Med 133:921-925. Kennelly JJ, Foote RH (1965): Superovulatory response to pre- and postpubertal rabbits to commercially available gonadotrophins. J Reprod Fertil9:177-188. McLimans WF (1972): The gaseous environment of the mammalian cell in culture. In Rothblat GH, Cristofalo VJ (ed):“Growth, Nutrition and Metabolism of Cells in Culture.” New York: Academic Press, pp 138-170. Purse1 VG, Wall RJ, Rexroad CE J r , Hammer RE, Brinster RL (1985): A rapid whole-mount staining procedure for nuclei of mammalian embryos. Theriogenology 24:687-700. Quinn P, Wales RG (1974):Fixation ofcarbon dioxide by preimplantation rabbit embryos in vitro. J Reprod Fertil 36:29-39. Snedecor GW, Cochran WG (1980): “Statistical Methods.” Ames, IA: Iowa State University Press, pp 233-236. Vishwakarma P (1962): The pH and bicarbonate-ion content of the oviduct and uterine fluids. Fertil Steril 13:481-485.
