Effect of potassium concentration, type of protein supplement, and embryo density on mouse preimplantation development in vitro *

Transcription

1 FERTILITY AND STERILITY Copyright 1986 The American Ferity Society Printed in U.SA. Effect of potassium concentration, type of protein supplement, and embryo density on mouse preimplantation development in vitro * Lynn M. Wiley, Ph.D.t Stacy Yamami, B.S. David Van Muyden, B.S. Division of Reproductive Biology and Medicine, Department of Obstetrics and Gynecology, University of California, Davis, California Mouse preimplantation embryos were used to test the effects of three culture parameters on development of 2 -cell stage embryos to the late morula stage (4 to 43 hours of culture, 65 to 8 hours after ferization): (1) concentration of potassium (K)-.6 mm, 1.4 mm, and 6. mm; (2) type of protein supplement, bovine serum albumin (BSA), and bovine serum immunoglobulins (BSIG); and (3) number of embryos (2 or 2) cultured per zo- to 12-fLl drop of T6 culture medium under paraffin oil. Cell number per embryo of cultured late morulae was compared with that of same-age embryos developing in vivo. The developmental rate to late morula stage with 2 embryos per drop of medium was media containing.6- and 1.4-mM K,.15% BSA,.15% BSIG > medium containing 1.4-mM K,.3% BSA > all media containing 6.-mM K (p <.1 comparing media containing.6- or 1.4-mM K with media containing 6.-mM K). For a given medium, developmental rate was significantly slower for 2 embryos per drop than for 2 embryos per drop of medium with those media containing either.3% BSA (for all concentrations of K tested) or 6.-mM K (for.3% BSA and for.15% BSA +.15% BSIG). Finally, cell number per embryo was: in vivo-developed embryos> medium containing 1.4-mM K,.15% BSA,.15% BSIG» 6.- or 1.4-mM K,.3% BSA. These data suggest that developmental rate and embryo cell number are enhanced by media containing a reduced concentration of K in combination with a mixture of serum proteins and by culturing several embryos together. Fer SteriI45:111, 1986 A major deficiency in in vitro ferization and embryo transfer (IVF-ET) protocols is the quality of the embryos before transfer. 1 To identify possible ways of improving embryo quality before Received June 3, 1985; revised and accepted September 26, *Supported by grant HD 1633 from the National Institutes of Health, Bethesda, Maryland. treprint requests: Lynn M. Wiley, Ph.D., Division of Reproductive Biology and Medicine, Department of Obstetrics and Gynecology, University of California, Davis, California transfer, we used the mouse preimplantation embryo assay to test the effects of three culture parameters on the rate of development from the 2-cell stage to the late morula stage: (1) concentration of potassium (K); (2) type of protein supplement; and (3) number of embryos cultured per fixed volume of culture medium. The rationale for using mouse preimplantation embryos in this study is twofold: (1) the mouse preimplantation embryo is becoming more frequently used for quality control of embryo culture conditions in IVF-ET programs; and (2) there are no data at present to suggest that the mouse preimplanta- Vo!' 45, No.1, January 1986 Wiley et al Embryo culture for IVF-ET 111

2 tion development profoundly differs physiologically from human preimplantation development, thus making it reasonable, un new data indicate otherwise, to use the needs of the mouse preimplantation embryo as an approximation of the needs of the human preimplantation embryo. The main objective of the culture period is to support embryonic development when the embryo is normally traversing the oviduct. During this time, the embryo must acquire the appropriate number of cells to support implantation and subsequent development upon arrival in the uterus. In the mouse2 and in the human,3, 4 this number is thought to lie between 12 and 18 cells, approximately 6 to 7 hours after ferization, which is when the embryo enters the uterus in both species.2, 4 Previous studies in the rabbit5 and the mouse2 have shown that cleavage-stage embryos with an insufficient number of cells have a significantly reduced survival rate when placed in the uterus. Studies with humans have shown that advanced embryo development 6 and the coordination of embryo development with the appropriate time for arrival in the uterus are important for the subsequent survival of the embryo.4 Collectively, these studies suggest that for the human, the optimal stage for transfer is the late morula stage, when the embryo has at least 16 cells. However, a survey of IVF-ET programs shows that the majority of human embryos are transferred when the embryos are at the 2- to 4-cell stage of development. During a study on mouse blastocoele formation, 7 we found that the rate of mouse development from the 16-cell stage (morula) into expanded blastocysts is sensitive to the concentration of K in the culture medium, such that reducing K to.6 mm doubles the developmental rate of morulae into blastocysts, compared with the developmental rate in 6.-mM K, which is the concentration in the widely used ovum culture medium of Biggers et al. 8 This observation motivated us to investigate whether a reduced concentration of K would also accelerate the development of 2-cell stage embryos to a stage suitable for transfer (a morula containing 16 or more cells). In most IVF -ET programs, a protein supplement is added to the embryo culture media in the form of serum, commonly maternal serum from the ovum donor, fetal cord serum, or some type of bovine serum. However, maternal serum may contain unidentified substances that are embryo- 112 Wiley et al. Embryo culture for IVF-ET toxic,9,1 as do bovine sera and some preparations of serum proteins purified from bovine sera. lo Other studies have shown no advantage of adding maternal serum or fetal cord serum to the culture medium with respect to development of mouse,9 rhesus/lor human12 embryos in vitro. These observations suggest that protein supplements may not be advantageous in embryo culture media. However, mouse preimplantation development in vitro is enhanced by the addition of serum protein (usually bovine serum albumin [BSA]), perhaps because of contaminating trace substances, such as fatty acids,13 or because BSA is performing some supportive action on plasma membrane function. However, the choice ofbsa is arbitrary. It may not be the best choice, based on the protein composition and concentration in oviductal fluid. Currently available data indicate that human oviductal fluid contains a variety of proteins, including serum albumin and immunoglobulins.14 Therefore, as a starting point we studied the effect of supplementing embryo culture medium with a mixture of bovine serum immunoglobulins (BSIG) and BSA. We selected the T6 medium of Quinn et al. 15 as our base culture medium because it is being used in IVF-ET and because of indications that simplified media such as T6 may be more appropriate for culturing preimplantation embryos than more complex media, such as Ham's F-I medium (GIBCO, Grand Island, Ny).16 Finally, it could be important that most mammalian cells do not culture easily as single cells or when their cell density is below a certain threshold that is characteristic for a given cell type. In vivo, the embryo is surrounded by other cells (oviductal epithelium), which might "condition" the extraembryonic milieu so that embryo quality is optimized. To test this possibility, we compared the development of embryos cultured in groups of 2 per 1- to 12-IJ..1 drop of medium with that of embryos cultured 2 per 1- to 12-IJJ drop of medium. MATERIALS AND METHODS SOURCE OF EMBRYOS Two-ceIl-stage embryos were obtained from superovulated random-bred Swiss ICR mice (Charles River Laboratories, Wilmington, MA) 5 to 54 hours after injection with human chorionic Ferity and Sterility

3 Table 1. Composition of Culture Media Stock solution for loo ml of medium NaCI Na2HP4 CaCl 2 2H 2 MgCl 2 6H 2 NaHCOa Glucose Na pyruvate Penicillin G Streptomycin S4 Cholesterol Phenol red Na lactate (6%) gmil ml ml Protein supplements BSA.3 or.15 BSIG.15 Addition of KCl Final concen tration ofk mm gmil ml gonadotropin (hcg). Embryos were flushed from dissected oviducts and rinsed with modified Hanks' balanced salt solution. Morula-stage embryos were obtained from superovulated mice 8 to 82 hours after ovulation (9 to 95 hours after hcg injection) by flushing dissected uterine horns with modified Hanks' balanced salt solution. CULTURE MEDIA We tested five different modifications of the T6 culture medium of Quinn et al. 15 (Tables 1 and 2). We varied the concentration of K and the type of bovine serum protein (BSA, crystallized Pentex, cat. no , and BSIG, fraction II, labile enzyme free, cat. no , Miles Scientific, Inc., Naperville, IL). Unless otherwise specified, all other culture reagents were obtained from Sigma Chemical Company, St. Louis, MO. Water was obtained from a Milli-Q R4 Reagent Water System equipped with ion exchange and charcoal filters (Millipore Corporation, Bedford, MA). All media were used within 3 hours of preparation. EMBRYO CULTURE The different media were compared for their ability to support development of embryos cultured in groups of 2 or 2 embryos/drop of medium (1 to 12 ILl). All cultures were initiated with pooled embryos that were all obtained within a 2-hour period from when female mice were killed. Developmental Rate and Incidence of Expanded Blastocyst Formation in the Different Culture Media At the beginning of the experimental culture period (54 to 56 hours after hcg injection), embryos were placed into microdrops of the desired culture medium with a paraffin oil overlay (Fisher Scientific Company, Fairhaven, NJ, cat. no. -119; not autoclaved) in plastic 35-mm Petri dishes (Falcon Plastics, Oxnard, CA, cat. no. 8) and cultured at 36.5 C under humidified 5% CO2, 95% air. During the culture period (56 to hours after hcg), embryos were periodically scored for developmental progress by observation with an inverted phase contrast microscope. These observations were used for the cavitation rates plotted in Figures 1 to 4. Some of these cultures were allowed to develop to the expanded blastocyst stage so we could determine whether accelerated development to the late morula affected the rate at which late morulae continued development into expanded blastocysts and the incidence of expanded blastocyst formation. Blastocysts were scored as embryos that had formed expanded blastocoeles containing an easily identified, well-developed inner cell mass. Comparison of Cell Number per Embryo of Morulae Developing in Vivo with That of Same Age Cultured Morulae Additional cultures were prepared as described above and incubated un 95 hours after hcg. The cultures were then terminated. The embryos were prepared for cell counts by treatment with a brief exposure to.5% sodium citrate followed by a III mixture of ethanol and glacial acetic acid. Embryos were then deposited singly onto glass slides, blown flat and air-dried. 17 To visualize nuclei, we stained slides of dried embryos with 2% aceto-orcein for 5 minutes. Because these cul- Table 2. Comparison of K Concentration and Type of Protein Supplement in the Five Modifications oft6 Medium Concentration of K Protein supplement (type, gmil ml) mm % 1.4 BSA,.3.6 BSA,.15 + BSIG, BSA,.15 + BSIG, BSA,.3 6. BSA,.15 + BSIG,.15 Wiley et ai. Embryo.culture for IVF-ET 113

4 (j) Ol "Iii Qj ~ "t:l c:: 6 "5 (;. In >-.c E 2 a'! 8 + hours post hcg required as an energy substrate for the first cleavage 18 and enhances the viability of 2- to 8-cell stage embryos, which can use pyruvate and lactate. 19 These comments about pyruvate as well as those about antibiotics suggest that there may be no advantage in "equilibrating" culture medium for 24 hours in an incubator before adding embryos to the medium. We thought it advantageous to add cholesterol to our modification of T6 because of evidence that suggests that it can be used to substitute for the fatty acids that are a common tracp contaminant of commercial preparations of BSA. 12, 2 We wanted to ensure that reducing the concentration of BSA in our media would not produce a fatty acid deficiency effect in our cultures. Figure 1 Developmental rate of 2 embryos/drop of culture media 1, 2, and 3. Medium 1 (-e-) contained 1.4-mM K,.3% BSA; medium 2 (------) contained.6-mm K,.15% BSA,.15% BSIG; and medium 3 (-A-) contained 1.4-mM K,.15% BSA,.15% BSIG. The arrow along the x-axis indicates 95 hours after heg administration (4 to 43 hours of culture, 65 to 8 hours after ovulation). The developmental rates in the three media between the origin and the arrow were compared for significant differences with Student's t-test; there are no significant differences in the developmental rate between the three curves. tured embryos were being processed for cell counts, we killed pregnant mice from the same mating from which the cultured embryos were obtained for in vivo-developed embryos, which were immediately processed as above for cell counts. RESULTS To obtain reproducible results and to minimize developmental asynchrony, we found it essential that dissecting, flushing oviducts, and preparing embryos for the cultures all be completed within 2 hours and that the embryos not be exposed to temperatures cooler than 35 C for more than 2 minutes. These observations should be kept in mind when using the mouse preimplantation embryo assay for quality control in IVF-ET programs. We also found it essential for reproducibility that working stocks of culture media be used within 2 days of preparation. This may be related to the limited half-life of the antibiotics (i.e., for penicillin, tl/2 = 27 hours at incubator temperatures) and the instability of sodium pyruvate. Sodium pyruvate is especially critical because it is DEVELOPMENTAL RATE AND INCIDENCE OF EXPANDED BLASTOCYST FORMATION IN THE DIFFERENT CULTURE MEDIA For developmental rate, we scored the cultures for the percentage of embryos that reached the late morula or a later stage after 7, 76, 95,, 14, and hours after administration of hcg. We determined the incidence of expanded blastocyst formation by counting the number of expanded blastocysts hours after administration of hcg. The data obtained with media 1, 2, (j) Ol 2 In Qj ~ "t:l c:: 6 :; (;. In >-.c E 2 a'! 8 + hours post hcg Figure 2 Developmental rate of 2 embryos/drop of culture media 1, 2, and 3. Information in the legend for Figure 1 applies to this figure, except that regarding Student's t-test comparison of the developmental rate between the origin and the arrow; the developmental rate in medium 1 (--l was significantly slower than the developmental rate in media 2 (-----l and 3 (_A_) (P <.1). 114 Wiley et al. Embryo culture for IVF-ET Ferity and Sterility

5 en Cl <1l 'lii Gi :; " c: <1l <1l :; S 6 (/) >-. E 2 ~ 8 + hours post hcg Figure 3 Developmental rate of2 embryos/drop of culture media 4 and 5. Medium 4 (--) contained 6.-mM K,.3% BSA; and medium 5 (--) contained 6.-mM K,.15% BSA,.15% BSIG. The developmental rate of 2 embryos/drop of medium 1 (_e_) was plotted for purposes of comparison. Arrow along the x-axis indicates 95 hours after hcg administration (4 to 43 hours of culture, 65 to 8 hours after ovulation). Student's t-test was used to compare the developmental rate in media 1, 4, and 5 between the origin and the arrow; the rates in media 4 and 5 were significantly slower than the rate in medium 1 (P <.1). the developmental rate between media 1, 2, and 3 (Fig. 1, Table 3). However, hours after hcg, the incidence of expanded blastocyst formation obtained with medium 1 was significantly lower than that obtained with media 2 and 3 (P <.1; Fig. 2, Table 3). When two embryos were cultured per drop, the developmental rate obtained with medium 1 was significantly slower than that obtained with media 2 and 3 (P <.1; Fig. 3, Table 4). The incidence of expanded blastocyst formation was the same for all three media. Culture Media 4 and 5 When 2 embryos/drop were cultured, the developmental rate obtained with medium 4 was significantly slower than that obtained with medium 5 (P <.1; Fig. 3, Table 4). In addition, the rates obtained with both media 4 and 5 were significantly slower than the rate obtained with medium 1 (P <.1). The incidence of expanded blastocyst formation, however, was similar for media 1, 4, and 5. When 2 embryos/drop were cultured, the developmental rate obtained with medium 4 was again slower than that obtained with medium 5 (P < and 3, which contained 1.4- or.6-mm K (Figs:1 and 2, Table 3), were tabulated and graphed separately from the data obtained with media 4 and 5, both of which contained 6.-mM K (Figs. 3 and 4, Table 4). Within each figure, the segments of the curves between the origin and the points corresponding to 95 hours after hcg (4 to 43 hours of culture; indicated by arrows along the x axes) were compared with Student's t-test for significant differences. The choice to limit comparisons to these segments of the curves was based on the objective of identifying factors that enhanced developmental rate up to the theoretic optimal time for transfer, which corresponded in our experiments to 9 to 95 hours after hcg administration or 4 to 43 hours of culture. The scoring for the ability of embryos to form expanded blastocysts was to ensure that enhanced development to the late morula stage did not compromise continued development into expanded blastocysts. Culture Media 1,2, and 3 When 2 embryos/drop of culture medium were cultured, there was no significant difference in en Cl ~ (/) Gi :; " c: <1l..!!! ::J S 6 (/) >-. E 2 1< 8 + hours post hcg Figure 4 Developmental rate of 2 embryos/drop of culture media 4 and 5. Information in the legend for Figure 3 applies to this figure except that the corresponding data representing 2 embryos/ drop of culture medium 1 was plotted here for comparative purposes. Regarding Student's t-test comparison of the developmental rate between the origin and the arrow, the developmental rate in medium 4 (--) was significantly slower than the rate in medium 1 (-e-) (P <.1), and the rate in medium 5 (--) was similar to that in medium 1. Wiley et ai. Embryo culture for IVF-ET 115

6 Table 3. Developmental Rate in Culture Media 1, 2, and 3: Percent and Statistical Data on the Number of Embryos Attaining the Late Morula or Later Stage After Various Time Periods in Culture a Hours after hcg administration 2 embryos! drop embryos! drop (3. ± 2.5) 53 (1.6 ± 4.53) 88 (17.6 ± 2.61) 97 (19.3 ± 1.5) 95 (19. ± 1.41) 18 (3.5 ± 1.91) 78 (15.5 ± 3.11) 91 (18.3 ±.5) 91 (18.3 ±.58) 94 (18.8 ±.96) Culture medium 2 3 o (.) (.) 17 (3.4 ± 3.58) 17 (3.4 ± 2.3) 8 (16. ± 5.12) 78 (15.6 ± 2.79) 9 (18.4 ± 2.19) 89 (17.8 ± 7.28) 98 (19.5 ± 1.) 98 (19.5 ± 1.) 94 (18.8 ± 1.79) 93 (18.6 ± 1.34) 24 (4.75 ±.5) 85 (17. ± 3.16) 95 (19. ± 1.15) 28 (5.5 ± 3.11) 83 (16.7 ± 4.16) 94 (18.8 ± 1.89) 98 (19.7 ±.58) 98 (19.7 ±.58) anumbers outside the parentheses represent percent of total number of embryos having attained the late morula or later stage by the corresponding hours after hcg administration. These are the values that are plotted in Figures 1 and 2. Numbers inside the parentheses represent the mean number of embryos (± standard deviation) obtained by averaging the data from five separate assays, 2 embryos per assay, for a given culture medium. Consequently, embryos were tested for each percentage value. Each point on the curves plotted in Figures 1 and 2 represents the response of a total of embryos from five separate experiments. Numbers inside the parentheses were used for Student's t-test to determine which culture media yielded developmental rates that were significantly different from the rate obtained using culture medium 1 (1.4-mM K,.3% BSA)'.1; Fig. 4, Table 4). In addition, a comparison of Figures 3 and 4 and corresponding numeric data in Table 4 reveals that for media 4 and 5, the developmental rate was significantly faster for 2 embryos/drop than for 2 embryos/drop. The incidence of expanded blastocyst formation at the end of the culture period, however, was the same for both of these media (Fig. 4, Table 4) and was similar to that obtained for medium 1, regardless of the number of embryos cultured per drop. (Compare data for medium 1 in Figures 1 to 4 and Tables 3 and 4.) The results indicate that when the concentration of K was low (1.4- or.6-mm) and 2 embryos/drop were cultured, there was no difference in the developmental rate to the late morula stage that could be attributed to differences in the type of protein supplement msa or BSA + BSIG). However, when 2 embryos/drop were cultured, there was a significant effect by the type of protein supplement on the developmental rate, with BSA + BSIG producing a faster developmental rate than BSA alone. When the concentration of K was increased to 6. mm, then the type of protein supplement had a significant effect on the developmental rate, regardless of whether 2 or 2 embryos/drop were cultured. Enhanced development to the late morula state did not compromise subsequent development into expanded blastocysts. Finally, the incidence of expanded blastocyst formation at the end of the culture period ( hours after hcg administration) was significantly increased when 2 embryos/drop were cultured of the two culture media that had the low concentrations of K and a protein supplement of BSA + BSIG (media 2 and 3). Comparison of Cell Number per Embryo of Morulae Developing in Vivo with That of Same Age Cultured Morulae At 95 hours after hcg, morulae developing in vivo were prepared for cell counts, along with same-age morulae that had been cultured for 4 Table 4. Developmental Rate in Culture Media 4 and 5: Percent and Statistical Data on the Number of Embryos Attaining the Late Morula or Later Stage After Various Time Periods in Culture a Hours after hcg administration 2 embryos/drop embryos/drop Culture medium (8. ± 2.83) 75 (15. ± 5.68) 95 (19. ±.) 98 (19.5 ±.71) 55 (11. ± 2.83) 78 (15. ± 5.66) o (.) 53 (1.5 ± 2.12) 85(17. ± 1.43) 65 (13. ± 2.83) 88 (17.5 ± 2.12) an umbers outside the parentheses represent percent of total number of embryos having attained the late morula or later stage by the corresponding hours after hcg administration. These are the values that are plotted in Figures 3 and 4. Numbers inside the parentheses represent the mean number of embryos (± standard deviation) obtained by averaging the data from five separate assays, 2 embryos per assay, for a given culture medium. Consequently, embryos were tested for each percentage value. Each point on the curves in Figures 3 and 4 represents the response of a total of embryos from five separate experiments. Numbers inside the parentheses were used for Student's t-test to determine which culture media yielded developmental rates that were significantly different from the rate obtained using culture medium 1 (1.4-mM K,.3% BSA). 116 Wiley et a!. Embryo culture for IVF-ET Ferity and Sterility

7 Table 5. Cell Number per Embryo 95 Hours After hcg Administration a Standard n Mean deviation Range In vivo-developed embryos In vitro-developed embryos Medium Medium Medium Medium athe data are from a matched-parallel experiment that ran concurrently with one of those that was used to obtain data represented in Figures 1 and 3. n, number of embryos. to 43 hours from the 2-cell stage in culture media 1, 2, 3, and 5. Medium 4 was not studied because of the significantly inferior developmental rate that was obtained with this medium (see above). As was expected, the in vivo-developed embryos had the highest cell number per embryo (Tables 5 and 6). Of the media that produced the fastest developmental rates (media 2 and 3), medium 3 produced the highest cell number per embryo of those developed in vitro. This value did not differ substantially from that of same-age embryos developed in vivo (.1 > P <.5). However, the lowest cell number per embryo was obtained with medium 2, which indicates that fast developmental rate does not necessarily correlate with rapid cell proliferation rate. With media 1 and 5, cell number per embryo was significantly lower than that of same-age embryos developed in vivo. DISCUSSION The major objective of this study was to identify ways of accelerating the developmental rate of cultured mouse preimplantation embryos from the 2-cell stage to a stage of development that would correspond with the optimal stage for transfer in the human (morula consisting of 16 or more cells, 65 to 8 hours after ovulation or 9 to 95 hours after hcg administration). We studied three culture parameters for their effect on the developmental rate: concentration of K; type of protein supplement; and number of embryos cultured per 1- to 12-,..d drop of culture medium. The data show that faster developmental rate and higher cell number per embryo at 95 hours after hcg administration was associated with reduced concentrations ofk (1.4 mm rather than 6. mm), with a mixture of serum proteins (.15 grn/loo ml each ofbsa and BSIG, rather than.3 gm/loo ml BSA), and with culturing 2 rather than 2 embryos per drop of culture medium. CONCENTRATION OF POTASSIUM The concentration of K is 6. mm in a culture medium that has been widely used in studies in which mouse preimplantation development serves as an experimental model for mammalian embryogenesis (ovum culture medium of Biggers et al. 8 ). In Ham's F-1 medium, the concentration of K is 4.43 mm. In the T6 medium, which was used for IVF-ET by Quinn et al.,16 the concentration ofk is 1.4 mm; this is the concentration ofk in the modified T6 medium used in our study. Our data indicate that reduced concentration of K can accelerate the developmental rate during cleavage from the 2-cell stage,just as was the case with the developmental rate during cavitation of morulae into blastocysts. 7 The mechanism whereby reduced K accelerates the developmental rate is not obvious, but one possibility is its effect on plasma membrane potential. Decreasing extracellular K hyperpolarizes the cell, causing an influx of sodium that is associated with enhanced mitotic rate in a number of systems Incidentally, the concentration of K in human postovulatory oviductal fluid is 7.7 mm,14 which, in the context of our results, indicates that there are additional factors yet to be elucidated regarding the regulation of the developmental rate in the oviduct. Table 6. Comparisons (Student's t-testj of Cell Number per Embryo Between in Vivo-Developed Embryos and in Vitro Developed Embryos Cultured in Media 1,2,3, and 5 a Comparisons In vivo/medium <.1 In vivo/medium <.1 In vivo/medium >.1 <.5 In vivo/medium <.1 Medium limedium >.1 Medium limedium >.1 <.5 Medium limedium 5.11 >.1 Medium 2/medium <.1 Medium 2/medium >.1 Medium 3/medium >.1 <.5 acalculations based on data shown in Table 5. P Wiley et al. Embryo culture for IVF -ET 117

8 PROTEIN SUPPLEMENT Protein supplements can have detrimental effects on cultured preimplantation embryos, and serum may not be the optimal protein supplement,12 particularly maternal serum,9 which is frequently used in IVF-ET. It seems appropriate, however, to add protein to embryo culture media, because postovulatory oviductal fluid has been estimated to contain 3.5 to 8.6 gm protein/ m Th 1S protem. concentration, however, is ten times the concentration of protein used in T6 medium and other media used for mouse embryo culture, while 2% serum, which is frequently used in IVF-ET, corresponds to about 1.2 to 1.6 gm protein/ ml.24 The protein in oviductal fluid is a mixture that includes albumin and immunoglobulins,14 which was our rationale for investigating the effect of a protein supplement consisting of BSA and BSIG, two serum proteins that can be obtained in relatively pure form from commercial sources. The reason why we obtained faster developmental rates with BSA + BSIG than with BSA alone is not clear. However, we have found that BSIG and the Fc fragment of bovine IgG, but not BSA, can be localized by indirect immunofluorescence on the cell surfaces of mouse preimplantation embryos. This is currently under investigation in our laboratory. Because the replacement of half the BSA with BSIG resulted in significantly improved embryo development, one could argue that we should replace all of the BSA with BSIG in our culture media. However, omitting all BSA from the culture medium may be detrimental because of the uncertainty of the nature of the trace elements that are frequent contaminants of commercially prepared BSA and that may be essential to embryo development in vitro. We have, in fact, performed some preliminary cultures with.3% BSIG and no BSA and found that such a protein supplement significantly impaired embryo viability (data not shown). CULTURING GROUPS OF EMBRYOS IN SMALL VOLUMES OF CULTURE MEDIUM It is common practice to culture ferized human ova either singly or in groups containing all ferized ova from a given patient (usually no more than eight). Frequently, the ova are cultured in relatively large volumes (2 to 3 md of culture medium. However, our results suggest that embryo development is improved by cul- turing larger groups of embryos in smaller volumes of culture medium (1 to 12 /-ll). It is not obvious why embryos appear to benefit from being more crowded, except that metazoan cells do not normally grow in isolation. Relative to human IVF-ET, however, our data suggest that it may be advantageous to consider co culturing human embryos with cellular "feeder layers" (fibroblasts?, oviductal epithelial cells?) or perhaps with 24-hour-old mouse embryos, which could be easily distinguished from the younger human embryos. REDUCING DEVELOPMENTAL LAG WITHOUT SACRIFICING EMBRYO QUALITY Cultured preimplantation embryos lag behind their same-age siblings developing in vivo. This lag was apparent from our cell count data, which showed that embryos removed from oviducts 95 hours after hcg administration had more cells per embryo than same-age embryos removed from oviducts 5 to 54 hours after hcg administration and cultured for 4 to 43 hours, for all five culture media tested. However, of the in vitro-developed embryos, those cultured in medium 3, which contained a reduced level of K (1.4-mM) and was supplemented with.15% BSA +.15% BSIG, produced the highest number of cells per embryo. This value was similar to that of in vivo-developed embryos (.1 > P <.5). This same culture medium also produced one of the highest developmental rates without any decrease in the incidence of blastocyst formation (Fig. 2). Consequently, enhanced developmental rate and cell number per embryo at the late morula stage did not compromise the ability to sustain development into expanded blastocysts. One additional point should be made regarding the incidence of expanded blastocyst formation. Although medium 3 produced a high number of cells per embryo together with a high incidence of expanded blastocyst formation, medium 2, which contained.6-mm K,.15% BSA, and.15% BSIG, also produced a high developmental rate and incidence of blastocyst formation. However, medium 2 also produced the lowest number of cells per embryo. This observation suggests that there is a lower threshold for the concentration of K required by pre implantation embryos to sustain a normal mitotic rate, which has previously been shown to be the general case for cells in culture.22 In addition, this observation em- 118 Wiley et a1. Embryo culture for IVF-ET Ferity and Sterility

9 phasizes that a fast developmental rate is not necessarily accompanied by a rapid cell proliferation rate and that the incidence of blastocyst formation (blastocyst expansion) may not be the best indicator of embryo quality when scoring for the success of development attained by mouse embryos in vitro.25 Acknowledgments. We wish to acknowledge the technical assistance of Ms. Lisa Miller and Ms. Maya Yamagata and the helpful discussions regarding data interpretation with Dr. Michael Femi Obasaju. REFERENCES 1. Speirs A, Trounson A, Warnes GM, Yovich JL, Saunders DM, Chen C: Summary of results. In Clinical in Vitro Ferization, Edited by C Wood, A Trounson. Berlin, Springer-Verlag, 1984, p Mintz B: Mammalian embryo culture. In Methods in Developmental Biology, Edited by FH Wilt, NK Wessells. New York, Thomas Y. Crowell Company, 1967, p Edwards RG: Early human development: from the oocyte to implantation. In Scientific Foundations of Obstetrics and Gynecology, Edited by EE Philipp, J Barnes, M Newton. London, William Heinemann Medical Publishers 1977, p 175 ' 4. Croxatto HB, Ortiz ME, Diaz S, Hess R, Balmaceda J, Croxatto HD: Studies on the duration of egg transport by the human oviduct. II. Ovum location at various intervals following luteinizing hormone peak. Am J Obstet Gynecol 132:629, Chang MC: Development and fate of transferred rabbit ova or blastocyst in relation to the ovulation time of recipients. J Exp Zoo I 114:197, Buslo M, Thorneycroft IH, Simon JA, Cohen SW, Buster JE: Experience with ovum transfer as a therapy for tubal inferity. Fer Steril (Abstr) 41:1s, Wiley LM: Cavitation in the mouse preimplantation embryo: Na/K-ATPase and the origin of nascent blastocoele fluid. Dev BioI 15:33, Biggers JD, Whitten WK, Whittingham DG: The culture of mouse embryo in vitro. In Methods in Mammalian Embryology, Edited by JC Daniel Jr. San Francisco, W. H. Freeman and Co., 1971, p Shirley B, Wortham JWE Jr, Witmyer J, Condon-Mahony M, Fort G: Effects of human serum and plasma on development of mouse embryos in culture media. Fer Steril 43:129, Caro CM, Trounson A: The effect of protein on preimplantation mouse embryo development in vitro. J In Vitro Fer Embryo Trans 1:183, Bavister BD, Boatman DE, Leibfried ML, Vernon MW: Ferization and cleavage of Rhesus monkey oocytes in vitro. BioI Reprod 28:983, Menezo Y, Testart J, Perrone D: Serum is not necessary in human in vitro ferization, early embryo culture, and transfer. Fer Steril 42:75, Kane MT, Headon DR: The role of commercial bovine serum albumin preparations in the culture of one-cell rabbit embryos to blastocysts. J Reprod Fer 6:469, Lippes J, Enders RG, Pragay DA, Bartholomew WR: The collection and analysis of human fallopian tubal fluid. Contraception 5:85, Quinn P, Barros C, Whittingham DG: Preservation of hamster oocytes to assay the ferizing capacity of human spermatozoa. J Reprod Fer 66:161, Quinn P, Warnes GM, Kerin JF, Kirby C: Culture factors in relation to the success of human in vitro ferization and embryo transfer. Fer Steril 41:22, Tarkowski AK: An air-drying method for chromosomal preparations. Cytogenetics 5:394, Whittingham DG, Biggers JD: Fallopian tube and early cleavage in the mouse. Nature 312:942, Brinster RL: Studies on the development of mouse embryos in vitro. II. The effect of energy source. J Exp Zool 158:59, Quinn P, Whittingham DG: Effect of fatty acids on ferization and development of mouse embryos in vitro. J Androl 3:44, Cohen I, Daut J, Noble D: An analysis of the actions of how concentrations of ouabain act on membrane currents in Purkinje fibres. J Physiol 26:75, Kaplan JG: Membrane cation transport and the control of proliferation of mammalian cells. Annu Rev Physiol 4:19, Busa WB, Nuccitelli R: Metabolic regulation via intracellular ph. Am J Physiol 246:R49, Ganong WF: Review of Medical Physiology. Los Altos, California, Lange Medical Publications, Massip A, Van der Zwalmen P, Puissant F, Camus M, Leroy F: Effects of in vitro ferization, culture, freezing and transfer on the ability of mouse embryos to implant and survive. J Reprod Fer 71:199, 1984 Wiley et al. Embryo culture for IVF -ET 119