REPRODUCTIVE BIOLOGY. New media for culture to blastocyst. Martha Hentemann, M.D., and Kjell Bertheussen, Ph.D.

Transcription

1 REPRODUCTIVE BIOLOGY New media for culture to blastocyst Martha Hentemann, M.D., and Kjell Bertheussen, Ph.D. Department of Obstetrics and Gynecology, In Vitro Fertilization Unit, University Hospital of Northern Norway, and Institute of Clinical Medicine, University of Tromsø, Tromsø, Norway Objective: To develop sequential media for extended culture of embryos in human IVF, with use of a mouse embryo assay. The BlastAssist system was developed. Blastocyst culture has been our routine method since 1994, using earlier media IVF and M3 in sequence. Design: Experimental laboratory study. Setting: University Hospital and university research unit. Animal(s): F1 hybrids between CD1 female and BDF male mice. Intervention(s): Two-cell mouse embryos were cultured 4 days in test media; 1 day in phase 1 medium followed by 3 days in phase 2 medium. Main Outcome Measure(s): Percentage of expanded and hatched blastocysts. Result(s): Compared with a two-cell block in the old culture system, more than 80% went to the blastocyst stage in the new media. The new medium system was based on a heavily modified low-glucose Earle s balanced salt solution. Additions were pyruvate, lactate, amino acids, human serum albumin, and Synthetic Serum Replacement. In contrast to regular tissue culture cells, embryos required an increased concentration of EDTA. The final media produced a blastocyst frequency of 94% and a day 4 hatching frequency of 71%. Conclusion(s): With the high blastocyst frequency demonstrated here, the combination of blastocyst culture and single ET should be an effective means of treatment in patients and could help to eliminate multiple gestations. The BlastAssist system is commercially available. (Fertil Steril Ò 2009;91: Ó2009 by American Society for Reproductive Medicine.) Key Words: Mouse embryo development, blastocyst culture, SSR, sequential culture media The quality of the culture system is a crucial factor influencing the success of IVF. During the early years of IVF, 2-day culture was established as a routine with use of simple serum-free media based on albumin as a serum substitute. We developed the Universal IVF medium in 1988, which became a popular commercially available medium, and later the M3 medium for human blastocyst culture (in regular clinical use since 1994). Both media contained the Synthetic Serum Replacement Type 2 (SSR 2) (1, 2). This sequential system, Universal IVF for 2 days, then M3 for the next 3 days, allowed the successful transfer of human blastocysts in an IVF program (2). David Gardner used the mouse model to develop a sequential culture system for human blastocyst culture (3). It became clear that earlier media developed from experience in mammalian tissue culture did not satisfy the specific metabolic needs of embryos. Received June 1, 2007; revised November 28, 2007; accepted December 5, M.H. has nothing to disclose. K.B. is shareholder in Medi-Cult, Denmark. Supported in part by MediCult, Jyllinge, Denmark. Dr. Bertheussen is a shareholder in MediCult. Reprint requests: Kjell Bertheussen, Ph.D., In Vitro Fertilization Department, University Hospital of Northern Norway, N-9038 Tromsø, Norway (FAX: ; kjell.bertheussen@unn.no). The mouse embryo has been of vital importance in both quality control and research to improve the outcome of IVF. Although there are certain differences between mouse and human embryos that affect physiologic parameters, the size of mouse embryos makes them useful as practice models for various research studies. The aim of this study was to optimize experimentally the composition of our sequential media and monitor the subsequent effect on mouse embryo morphology, cleavage, and development of blastocysts. We continued our use of albumin and SSR 2 as a combined serum replacement. Searching the published literature for sensitive blocking mouse strains (John D. Biggers and Piyush R. Bhatnagar, Harvard Medical School, personal communication), we selected F1 hybrids between CD1 female and BDF male mice as our mouse model. The result, BlastAssist sequential system, has been in regular clinical use since MATERIALS AND METHODS Animals Six- to 12-week-old female CD1 mice were maintained under controlled lighting conditions (light: 14 hours, dark: 10 hours). They were superovulated by injection of 7.5 IU 878 Fertility and Sterility â Vol. 91, No. 3, March /09/$36.00 Copyright ª2009 American Society for Reproductive Medicine, Published by Elsevier Inc. doi: /j.fertnstert

2 pregnant mare serum gonadotropin (Sigma Chemical Co., Germany) followed by 7.5 IU hcg (Sigma) at 48 hours after pregnant mare serum gonadotropin injection. At the same time mating with BDF males was initiated. All experimental protocols were approved by the National Animal Research Authority in Oslo, Norway. Culture Conditions The embryos were cultured at 37 Cin7%CO 2. Embryos developed to the two-cell stage were collected from the oviducts of female mice at 42 to 44 hours after hcg injection. After washing once in preparing medium (HEPES buffered BBEM) the embryos were divided and placed into different culture media after washing them twice in the respective medium. All embryos were cultured in groups of 5 to 15 in 60- ml drops of medium under paraffin oil (mineral oil, embryo tested, M 8410; Sigma) in mm Petri dishes or in 0.5 ml medium in Nunc four-well plates (VWR, Oslo, Norway). After 24 hours of culture (phase 1) media were changed, and embryos were washed twice and transferred to phase 2 test media. Evaluation of Development The number of embryos developed to four-cell, eight-cell, morula, blastocyst, expanded blastocyst, and hatched blastocyst stages were monitored under an inverted microscope once daily and scored morphologically. Light micrographs of living embryo cultures were obtained with use of an inverted photomicroscope equipped with phase-contrast objectives. Experiments published here were performed in triplicates and with a total of 60 embryos or more in each test medium. Standard deviation <10% of quantities is given in the results. Chemicals Except as noted, biochemicals and medium powders were purchased form Sigma-Aldrich (Oslo, Norway). Synthetic Serum Replacement 2 (1,000 stock solution) was prepared as shown earlier (1). Insulin (500 ng/ml final concentration) used in SSR 2 was human recombinant (catalog No ; Boehringer Mannheim, Mannheim, Germany). Human serum albumin (HSA) was purchased from MediCult, Jyllinge, Denmark. Media Cell culture media were prepared each month from medium powder, sodium bicarbonate, penicillin G (50 IU/mL), streptomycin sulfate (50 mg/ml), and high-quality sterile water (Sigma). After addition of all components, media were sterile-filtered through a membrane filter and stored at 4 C in the dark. Final osmolarity was 280 mosm with standard bicarbonate concentration, 2.2 g/l NaHCO 3. There was no attempt to adjust the ph, because media were equilibrated in a CO 2 incubator before the addition of cells. The basal medium (BBEM) for embryo culture developed through this study is based on a modified low-glucose Earle s balanced salt solution (EBSS), to which has been added pyruvate, lactate, amino acids including glutamine, and some other components as listed below. For serum-free culture of embryos, SSR 2 (1 ml/l) and HSA (2 mg/ml) were always added to the basal medium (Table 1). TABLE 1 Media used in this study: Phase 1 Phase 2 IVF: EBSS solution with SSR 2 M3: Hams MCDB 302 with SSR mmoi/l pyruvate 2.2g/L NaHCO3 l0 mg/ml HSA 2 mg/ml HSA BBEM: EBSS 2.2g/LNaHCO3 BL-2: based on BBEM all amino acids low glucose, pyruvate low lactate phosphate not reduced taurine low glutamine non-essential amino acids low HEPES without HEPES/taurine some components at higher concentrations (i.e. glucose) B1A M1: BBEM B1A M2: BL-2 SSR 2 SSR 2 EDTA EDTA HSA HSA G-l: Vitrolife G-2: Vitrolife Fertility and Sterility â 879

3 The new sequential medium system was designated the BlastAssist (BlA) system (available from MediCult), consisting of two media: M1 and M2. The corresponding basal media were designated BBEM (phase 1) and blastocyst medium, phase 2 (BL-2), the latter based on the BBEM formulation. RESULTS Glucose, Pyruvate, and Two-cell Block The first study showed an arrest at the two-cell stage in mouse embryos cultured in the old IVF medium. By lowering glucose and pyruvate in the IVF medium (i.e., early stage BBEM medium), the embryos developed normally with a blastocyst frequency of 50%. Both factors were inhibitory at original concentrations. Glucose was reduced from 5 mmol/l to 0.5 mmol/l, and pyruvate was reduced from 0.8 mmol/l to 0.2 mmol/l. The chelator EDTA improved development in BBEM medium (Fig. 1) but did not change the two-cell block induced in the other media. Optimizing SSR for Embryo Culture After a series of titration experiments a combination of SSR and EDTA was found that gave a higher percentage of blastocysts and hatched blastocysts compared with basal medium without these components. The final concentration of EDTA for mouse embryo culture was set at 100 mmol/l in phase 1 and 10 mmol/l in phase 2 (Fig. 1). The New Sequential IVF Media The sequential system, BlA, consists of two media, denoted M1 and M2. Phase 1: human IVF day 0 2, mouse two-cell day 0 1 BlA M1 ¼ BBEM þ SSR 2 þ 100 mmol/l EDTA þ HSA Phase 2: human IVF day 2 5, mouse two-cell day 1 4 BlA M2 ¼ BL-2 þ SSR 2 þ 10 mmol/l EDTA þ HSA Ethylenediaminetetraacetic acid has been reduced in the human IVF media. The final media produced a blastocyst frequency of 94% and a day 4 hatching frequency of 71% for the mouse CD1 strain used here (average results from all experiments). Phase-contrast morphology of CD1 embryos cultured in the new BlA system is shown in Figure 2. Most embryos can be seen in the hatched stage. BlastAssist M1 and M2 versus Gardner s G1/G2 Next a comparison between two media in two culture systems was carried out: closed drops under oil and open wells. Commercially available sequential media G-1 and G-2 (Vitrolife, Kungsbacka, Sweden) were compared with BlA M1 and M2. We found an increased rate of hatched blastocysts in BlA M1 and M2 as compared with G-1 and G-2 in both culture systems (Fig. 3). Performance of Three Medium Systems in Mouse Embryo Development After final development of the new media a qualitative picture became evident (Fig. 4). Although development was arrested at the two-cell stage in the original IVF/M3 system, exchange of one medium to yield the BlA M1/M3 system improved developmental capacity but ending at the morula stage. However, when embryos were cultured in both optimized media (BlA M1 and BlA M2), a full development to the hatched blastocyst was noted. FIGURE 1 Mouse embryo culture: optimizing SSR for embryo culture, percent embryos forming blastocysts/ hatched blastocysts. Basal media: BBEM/BL % embryos forming blastocyst FIGURE 2 Mouse embryos cultured 4 days in final BlA system. Most embryos have hatched. Phase-contrast micrograph (original magnification 100) day 3 blastocysts day 4 hatched 20 0 basal medium w/hsa + SSR + EDTA + SSR + EDTA 880 Hentemann and Bertheussen New media for culture to blastocyst Vol. 91, No. 3, March 2009

4 FIGURE 3 Performance of BlA M1/M2 and Gardner s G-1/G-2 in two mouse culture systems: drops (under oil) and open wells. The percentage of hatched blastocysts on day 4 was determined % embryos hatched blastocyst day 4 FIGURE 5 Human IVF: comparison of three medium systems. Percentage of patients with blastocyst transfer on day 5 morning (110 hours of culture after ovum pickup, at least one blastocyst transferred) (Hentemann M, Bertheussen K, unpublished observations) ET day 5: patients with 1 blastocyst 88, drops wells ,7 0 BlastAssist Gardner 20 Performance of Three Medium Systems in Human IVF The new media were introduced in our IVF clinic for a period of 2 years ( ), where the performance of three medium systems was compared. In an accompanying communication detailed clinical results will be shown (Hentemann M, Bertheussen K, unpublished observations). The group with culture in new BlA M1 and M2 showed significantly more patients with blastocyst transfer on the morning of day 5 compared with results with BlA M1 and old M3 or IVF and old M3 (Fig. 5). The low blastocyst figure in the IVF old M3 group in part is caused by the delayed FIGURE 4 Mouse embryo culture: qualitative comparison of three medium systems. Final stage of development 2-cell morula hatching blastocyst IVF/ old M3 BlA M1/ old M3 BlA M1/BlA M2 0 5,2 IVF/ old M3 BlA M1/ old M3 BlA M1/BlA M2 embryo development in this old medium system; thus many patients delivered compact morula stages the morning of day 5, which subsequently developed to blastocysts later on day 5 (as seen in surplus embryos). DISCUSSION The main result reported in this article is the formulation of an optimized serum-free medium system for the culture of human and animal embryos to the blastocyst stage. It contains one protein, HSA; it is otherwise protein free, except a low concentration of human insulin (500 ng/ml). Amino acids have been added on the basis of an understanding of the specific requirements of embryos at early and later stages, leading to a sequential system of two media: BlA M1 and M2. Since its development in 1998, the BlA system has been commercially available (MediCult) and is being used by an increasing number of IVF clinics. The media M1 and M2 are based on two basal media developed through this study and contain EDTA, SSR 2, and HSA as a combined serum replacement. The similarity in metabolic parameters between mouse and human embryos has facilitated the use of the mouse as a model to study the formulation of culture media to be used at different stages over the preimplantation period from fertilization to the fully expanded blastocyst stage (4). The so-called two-cell block, which is observed in most inbred and outbred strains of mice, makes a clear quality indicator during development of improved media (5, 6). Fertility and Sterility â 881

5 In vitro blocks to development often occur at the stage where the transcription of the embryonic genome begins: in the mouse at the two-cell stage and in humans between the four-cell and eight-cell stage, where a weaker block may be observed (7, 8). These blocks occur in embryos cultured under suboptimal conditions. With the new BlA system, two-cell block is abolished, whereas our older media (Universal IVF and old M3, used in sequence) lead to a full arrest at the two-cell stage in our sensitive mouse strain. A parallel effect was noted in human IVF, where BlA gave an improved development to the blastocyst stage (Hentemann M, Bertheussen K, unpublished observations). The arrest at the mouse two-cell stage was found to be caused by increased concentrations of glucose and pyruvate. The combination BlA M1 þ old M3 also demonstrated a problem inherent in the old M3 medium. Based on Ham s MCDB 302, M3 is a very complex and rich medium optimized for mammalian tissue culture, and many components may exert an inhibitory effect on cleavage embryos at high concentrations. Different needs of the embryo throughout its development are the underlying logic for sequential media. Most favorable conditions for the early embryo do not support optimal blastocyst development and differentiation. On the other hand the conditions supporting blastocyst development and differentiation are detrimental to the early embryo. In vivo, dramatic changes in embryo physiology and nutrient requirements go along with environmental changes in the female reproductive tract. To minimize intracellular stress when embryos are moved from one to the next phase of culture, sequential media are developed to exist as a series of media, each one sharing formulation similarities (9). With regard to embryo physiology it is appropriate to consider the preimplantation period in at least two phases: precompaction and postcompaction, characterizing the cleavage stage embryo by relatively low levels of biosynthesis, low respiratory rates, and limited capacity to use glucose as an energy source. In the postcompaction embryo, biosynthetic rates are increasing, along with an increased respiratory capacity and an ability to use glucose (3). It has been demonstrated that the embryo exhibits a switch in amino acid requirement. Up to the eight-cell stage nonessential amino acids are important, whereas all 20 amino acids become necessary after compaction. These principles were implemented in the BlA system, where M1 contains nonessential amino acids and a very low glucose concentration, whereas M2 contains all amino acids and a higher concentration of glucose. Each medium, M1 and M2, shares formulation similarities. Synthetic Serum Replacement 2 was developed for long-term culture of cells in protein-free media (1). The most important component to replace serum was found to be an optimized metal ion buffer, which maintains trace metal (free) concentrations at steady levels, largely independent of environmental metal impurities. This metal buffer is based on a ligand mixture of 40 mmol/l citrate and 4.3 mmol/l EDTA intended to control a mixture of iron and necessary trace metals. Without a stable metal buffer, iron and trace metals will be precipitated by phosphate and bicarbonate, and this may explain the occurrence of several quite modern embryo medium formulas low in bicarbonate and with phosphate low or absent. Such designs are clearly unphysiologic and were avoided in BlA, as well as in the earlier media described in this report. Ethylenediaminetetraacetic acid has been reported to improve embryo cleavage at considerably higher concentrations than that in SSR 2 (3, 5, 10, 11). A similar finding is reported here for the BlA system for mouse embryo culture, the optimum concentrations being 100 mmol/l in M1 and 10 mmol/l in M2. Ethylenediaminetetraacetic acid was reduced in the media for human IVF. In the absence of added trace elements, EDTA may show negative side effects in phase 2 as reported by Gardner and Lane (even though this finding is being disputed by others) (3, 5, 11, 12). Ethylenediaminetetraacetic acid should be combined with a balanced iron trace metal system. In the BlA media a clear improvement in blastocyst morphology, as well as development, was noted by the presence of EDTA in both culture phases. One possible role of EDTA in the culture medium is to act as a chelator of transition metal ions to prevent extracellular generation of oxygen radicals (13, 14). In comparison with Gardner s G-1/G-2 system (Vitrolife), BlA produced superior results in the mouse, especially during culture in open wells. The SSR 2 serum replacement may be responsible in part for the difference, but there are also considerable differences in the design of the basal media. There are several routes for further development of good IVF media. New basal media will come with a more extensive composition (including vitamins) in concordance with improved knowledge of embryo metabolism. To avoid the ammonia problem, glutamine should be replaced by more stable derivatives without compromising medium performance. A future step would also be the elimination of the last animal-derived protein, albumin, by the use of either a recombinant product or a synthetic replacement. Acknowledgments: We are indebted to Professor Jan M. Maltau (Institute of Clinical Medicine, Tromsø, Norway) for his encouragement and help during this study. REFERENCES 1. Bertheussen K. Growth of cells in a new defined protein-free medium. Cytotechnology 1993;11: Bertheussen K, Forsdahl F, Maltau JM. In vitro fertilization. New media for embryo culture to the blastocyst stage. Proceedings of the 10th World Congress of In Vitro Fertilization and Assisted Reproduction; 1997 May 24 28; Vancouver, British Columbia, Canada: Gardner DK, Lane M. Culture and selection of viable blastocysts: a feasible proposition for human IVF? Hum Reprod Update 1997;3: Hentemann and Bertheussen New media for culture to blastocyst Vol. 91, No. 3, March 2009

6 4. Quinn P, Horstman FC. Is the mouse a good model for the human with respect to the development of the preimplantation embryo in vitro? Hum Reprod 1998;13(Suppl 4): Gardner DK, Lane M. Alleviation of the 2-cell block and development to the blastocyst of CF1 mouse embryos: role of amino acids, EDTA and physical parameters. Hum Reprod 1996;11: Lawitts JA, Biggers JD. Overcoming the 2-cell block by modifying standard components in a mouse embryo culture medium. Biol Reprod 1991;45: Dobson AT, Raja R, Abeyta MJ, Taylor T, Shen S, Haqq C, et al. The unique transcriptome through day 3 of human preimplantation development. Hum Mol Genet 2004;13: Hadi T, Hammer MA, Algire C, Richards T, Baltz JM. Similar effects of osmolarity, glucose, and phosphate on cleavage past the 2-cell stage in mouse embryos from outbred and F1 hybrid females. Biol Reprod 2005;72: Gardner DK. Textbook of assisted reproductive techniques. Laboratory and clinical perspectives. London: Taylor and Francis, Abramczuk J, Solter D, Koprowski H. The beneficial effect EDTA on development of mouse one-cell embryos in chemically defined medium. Dev Biol 1977;61: Biggers JD, McGinnis LK, Lawitts JA. One-step versus two-step culture of mouse preimplantation embryos: is there a difference? Hum Reprod 2005;20: Biggers JD, McGinnis L, Summers MC. One-step versus two-step culture of mouse preimplantation embryos [reply]. Hum Reprod 2006;21: Matsukawa T, Ikeda S, Imai H, Yamada M. Alleviation of the two-cell block of ICR mouse embryos by polyaminocarboxylate metal chelators. Reproduction 2002;124: Menezo Y, Chouteau J, Guyader-Joly C, Veiga A. [Sequential media: why and how?]. Contracept Fertil Sex 1999;27: Fertility and Sterility â 883