Letters to the editor |
Reply: One-step versus two-step culture of mouse preimplantation embryos
1 Department of Cell Biology, Harvard Medical School, Boston, MA 2 Department of Molecular and Integrated Physiology, University of Kansas Medical School, Kansas City, KS 3 Reproductive Science Center, Lexington, MA and 4 Department of Obstetrics and Gynecology, University of Massachusetts Medical School, Worcester, MA, USA
5 To whom correspondence should be addressed at: Department of Cell Biology, Harvard Medical School, Boston, MA 02115, USA. E-mail: john_biggers{at}hms.harvard.edu
Sir,
David Gardner and Michele Lane, who have been strong advocates of two-step protocols for the culture of human preimplantation embryos, have written the accompanying letter in which they offer a ‘different perspective’ on our recent paper (Biggers et al., 2005
). We feel that we must respond to this letter since their so-called ‘different perspective’ is an attempt to challenge the validity of our work by claims that are misleading and confusing. Also, there are several instances where we are seriously misquoted or misinterpreted. For example, Gardner and Lane, in the fourth paragraph of their letter, make the following statement:
So irrespective of the paper by Biggers et al. (2005)
Nowhere does our work support changing media at 48 h. It also is especially disturbing to read a statement that a scientific view is absolute. Scientific theories and practical applications change as new evidence accumulates, and new interpretations are made. Nothing is absolute in science. We hope that our reply will be regarded as a non-acrimonious reply in a debate which will help embryologists decide whether a one-step protocol or a two-step protocol for the culture of human embryos is sufficient for their needs.
Two-step culture protocols for the culture of human preimplantation embryos from the zygote to the blastocyst, in which the medium is changed midway through the culture period, are used almost universally. There is little direct experimental evidence that two-step culture protocols are better than one-step culture protocols (Biggers, 1998, 2002, 2003; Biggers and Racowsky, 2002; Summers and Biggers, 2003). Our 2005 paper reported experimental studies on the development of the mouse zygote to the blastocyst stage comparing one-step and two-step protocols. Very similar results were found between one-step and two-step protocols (Biggers et al., 2005).
Before responding to Gardner and Lane’s comments in detail, we would like to clarify the objectives of the experiments reported in our 2005 paper and the null hypotheses we set out to test. These experiments were designed specifically to compare one-step and two-step protocols, defined jointly by the different media used and the volume of the culture droplet. We emphasize the use of the word ‘protocol’ to identify the treatments because they were not concerned solely with the effects of different media. The volumes of the droplets used depended on whether the protocol was a one-step (50 µl) or two-step protocol (20 µl). The volumes adopted were based on the recommendations of Gardner and Lane (1996). In the first experiment, the overall null hypothesis was that there are no significant differences between the effects of four protocols using KSOMgAA and G1.2 (one-step, droplet volume 50 µl), and KSOMgAA/KSOMgAA and G2.1/G2.2 (two-step, droplet volume 20 µl). In the second experiment, the overall null hypothesis was that there are no significant differences between the effects of four protocols using KSOMgAA and DM2 (one-step, droplet volume 50 µl), and KSOMgAA/EDTA-free KSOMgAA and DM2/DM1 (two-step, droplet volume 20 µl). The four protocols in each experiment were compared simultaneously over the same period and were replicated several times.
Challenges to our methodology
The questions raised by Gardner and Lane about our 2005 paper concern the method we used to denote the developmental age of the embryos produced in culture and the quality of the oil under which the embryos were cultured.
In the fifth paragraph of their letter, Gardner and Lane make the statement:
In the paper by Biggers et al. (2005)
This statement is incorrect and has led to the erroneous conclusion that we observed our blastocysts on day 6 of in vivo development. We state in our paper that the zero time used was the time when HCG was injected, and that the zygotes were placed in culture 24 h later. Thus, 144 h post-HCG is equivalent to day 5 of in vivo development which Gardner and Lane deem acceptable. In their own work, Gardner and Lane define time development as days from placing the embryos in culture. It is difficult to control exactly the developmental stage reached when zygotes are placed in culture so that there is inevitably some variation in the development of embryos in culture between laboratories and in experiments within laboratories. This criticism of our work is therefore not valid.
The opening paragraph of the letter by Gardner and Lane challenges the validity of our paper on the grounds that the oil overlay used in our cultures was not sold for human IVF and therefore had sub-optimum properties. They raise the quality of oil issue by citing evidence that when a sub-optimum oil is used, the degree of development of preimplantation mouse embryos increases with the volume of the droplet of medium under the oil. The oil we used was M8410 (Sigma Chemical, St Louis, MO, USA), which is embryo tested. The data sheet supplied by the company, which describes their quality-control methods, states that 80% of 1-cell F1 hybrid mouse embryos must develop into blastocysts during 96 h of culture in Human Tubal Fluid medium supplemented with 4% BSA before the oil is approved for sale. Almost identical tests using mouse embryos are used for the quality control of oils used for the culture of human preimplantation embryos (see data sheets from Vitrolife Inc., Englewood, CO, USA; Sage IVF Inc., Trumbull, CT, USA). Thus, we do not know of any a priori reason why the oil we used in experimental work with mice should have been the oil approved for culturing human preimplantation embryos. We have used this oil (M8410, Sigma) after sterile filtration in our laboratory for many years, as have many other research laboratories throughout the world for the culture of mammalian preimplantation embryos. In fact, Gardner and Lane have used Sigma oil many times in their own studies (e.g. Lane and Gardner, 1998; Lane et al., 2001; Zander et al., 2006). Do they wish to challenge the validity of their own experiments?
Challenges to our interpretations
The issues raised by Gardner and Lane concern the three main areas they use to support their advocacy of two-step protocols for the culture of preimplantation embryos: (1) the elimination of EDTA toxicity, (2) the elimination of glutamine breakdown which generates toxic ammonium and (3) the ‘back to nature’ philosophy used to guide the concentrations of media components.
EDTA toxicity
EDTA is included in media for the culture of mouse embryos in order to overcome the 2-cell block. Lawitts and Biggers (1991), in the initial work that led to the KSOM family of media for the culture of mouse embryos, showed that a concentration of only 10 µmol/l was sufficient. Lane and Gardner (2001) demonstrated that this concentration of EDTA has no effect on the activity of the citric acid cycle, and therefore the utilization of pyruvate as the main energy source in the initial cleavage stages of the mouse. Moreover, they found that 10 µmol/l EDTA did not affect the implantation rate, fetal development and fetal weight. The work in our recent paper demonstrates further that a concentration of 10 µmol/l EDTA does not interfere with development of the mouse zygote to the blastocyst stage and the development of young after these blastocysts are transferred to surrogate mothers. Despite this direct evidence that 10 µmol/l EDTA does not impair the post-implantation development of mouse embryos, Gardner and Lane argue in their letter that it is necessary to remove EDTA in a second phase of culture even if the concentration of EDTA in the medium is only 10 µmol/l because this concentration of EDTA inhibits 3-phosphoglycerate kinase, a key enzyme in glycolysis on which the blastocyst is very dependent (Lane and Gardner, 2001). They demonstrated this inhibitory effect of EDTA on the enzyme using cell-free extracts of mouse embryos, where the compound has direct access to the enzyme. Under these conditions, it is likely that the EDTA will be inhibitory in much lower concentrations than in the intact cells of the preimplantation embryo and thus the fact that 10 µmol/l EDTA inhibits 3-phosphoglycerate kinase is a poor indicator of the concentration of EDTA that can be safely added to a medium for the culture of intact embryos. In view of the direct evidence that 10 µmol/l EDTA does not impair development of mouse embryos under culture conditions, there is no clear reason to remove it by the use of a two-step protocol.
Glutamine breakdown
Gardner and Lane write, in the fourth paragraph of their letter:
Therefore, three independent laboratories, including that of Biggers, have demonstrated that ammonium can induce exencephaly in mouse fetuses following the exposure of the preimplantation embryo.
This statement is a serious misquotation of the work we have described elsewhere (Biggers et al., 2004a, b; Summers et al., 2005). We have never studied the addition of ammonium to media for the culture of mouse preimplantation embryos. The only single case of exencephaly that we have observed was in a group of embryos cultured in KSOMgAA in which there was no glutamine since it had been replaced with glycylglutamine that does not readily breakdown producing ammonium (Biggers et al., 2004a). In their letter, Lane and Gardner selectively report those findings of Sinawat et al. (2003) which favour their argument. Sinawat et al. (2003) have shown that exencephaly can occur when preimplantation mouse embryos are cultured in media, including KSOM, supplemented with ammonia, but do not report any case of exencephaly in the control cultures containing 1 mmol/l glutamine but where no ammonia was added. Thus, we reiterate (review: Biggers et al., 2004b) that there is no compelling published evidence that excencephaly occurs in mice, following culture in media containing 1 mmol/l glutamine not supplemented with ammonia.
Gardner and Lane have made extensive studies on the action of ammonium on preimplantation development, and they have used their results to support the need for a two-step protocol to remove the ammonium which arises in culture media by the spontaneous breakdown of glutamine into ammonium and 5-pyrrolidone-2-carboxylic acid (Biggers et al., 2004b; Summers et al., 2005). We have discussed in other papers (Biggers et al., 2004b; Summers et al., 2005) why experiments studying the effects of adding ammonium instead of glutamine to culture media do not help in deciding whether two-step protocols are necessary to culture mouse embryos in media containing glutamine. The interpretation of experiments in which ammonium is added to culture media is not simple, but depends on taking account of the kinetics of the reaction involved in the breakdown of glutamine. Very recently, it has been reported that the preimplantation mouse embryo is more sensitive to ammonium at the zygote and 2-cell stage than at later stages (Zander et al., 2006). The amount of ammonium accumulated from the breakdown of glutamine will be low when the embryos are the most sensitive, while at the end of the culture period, when more ammonium has accumulated, the embryo is relatively insensitive. This debate, however, is now irrelevant. Glutamine can be replaced with stable dipeptides of glutamine, such as glycylglutamine or alanylglutamine (Biggers et al., 2004b), which do not give rise to ammonium. Thus a primary reason for adopting a two-step protocol has been eliminated.
Back-to-nature principle
In the last paragraph of their letter, Gardner and Lane write
There can be no discussion on the changing physiology and metabolism of the embryo during the preimplantation period, and that the zygote and blastocyst are as different as skin and heart muscle.
The analogy made in this statement makes little sense. Presumably, it is made to provide another argument in support of two-step protocols, in which adjustments in the compositions of the first and second media are attempts to imitate prior knowledge on the physiology of the mother and embryo. It is true that there are well-known changes in the concentrations of glucose and pyruvate in the female genital tract during the initial stages of pregnancy which correlate with the changes in the utilization of these compounds by the embryo. It is also likely that there are similar associated changes with the amino acids, although these are still incompletely understood. This ‘physiological’ argument is weak for at least two reasons. First, in a two-step protocol, the adjustments made in the second medium involve only a small subset of all the compounds in the first medium, and so there is relatively little change in the overall artificial environments provided by chemically defined media. Second, it ignores the importance of the stresses to which the embryos have to adapt in artificial environments provided by chemically defined media. The adaptations must occur twice in two-step protocols versus once when using one-step protocols. Full discussions of this topic are found in Biggers (2003) and Summers and Biggers (2003).
Implications for the culture of preimplantation human embryos
In the last paragraph of their letter, Gardner and Lane state
Whatever the outcome of the debate, there can be no question that sequential media formulations used by Biggers et al. (2005)
Balaban and Urman’s paper is irrelevant to the present debate since they did not compare one-step and two-step protocols. They compared two two-step protocols: G2.1/G2.2 with a new two-step formulation (series GIII) whose composition is a trade secret. Gardner and Lane have missed the point that the present debate requires data in which one-step and two-step protocols are simultaneously compared.
It is instructive to summarize the history of two-step protocols. Gardner and Lane’s proposal to use a two-step protocol was based on their work on mice on the inhibitory effects of EDTA to preimplantation development. This was soon followed by similar work on the toxic effects of ammonium. Later attempts were made to imitate natural physiological changes associated with energy and amino acid metabolism. The appearance of the G1.1/G1.2 two-step protocol stimulated several companies to market other media for use in a two-step protocol (Biggers, 2000). Unfortunately for science, all these companies have kept the concentrations of the compounds in their media secret. Advertising during this period has given the impression that the superior value of a two-step protocol was settled for the culture of human preimplantation embryos. Our results using mice show that one-step protocols can be developed. If they can be developed for human use, unnecessary handling of the embryos can be eliminated. We know of only two papers that provide the results of a direct comparison of one-step and two-step protocols for the culture of human preimplantation embryos (Biggers and Racowsky, 2002; Macklon et al., 2002). The results of both studies suggest that a one-step protocol is sufficient for the culture of human zygotes to the blastocyst stage at rates comparable with other two-step protocols. In a discussion of protocols for culturing human preimplantation embryos, Summers and Biggers (2003) wrote, ‘Further investigations comparing development using either two- or one-step protocols may show that although two-step procedures are sufficient to support development, they may not be necessary’. It is time that such studies be undertaken.
We wish to acknowledge several helpful suggestions made by Dr Betsey Williams.
References
Balaban B and Urman B. (2005) Comparison of two sequential media for culturing cleavage stage embryos and blastocysts: embryo characteristics and clinical outcome. Reprod Biomed Online 10:485–491.[Medline]
Biggers JD. (1998) Reflections on the culture of the preimplantation embryo. Int J Dev Biol 42:879–884.[ISI][Medline]
Biggers JD. (2000) Ethical issues and the commercialization of embryo culture media. Reprod Biomed Online 1:74–76.[Medline]
Biggers JD. (2002) Thoughts on embryo culture conditions. Reprod Biomed Online 4:Suppl. 1, 30–38.
Biggers JD. (2003) Fundamentals of the design of culture media that support human preimplantation development. In Van Blerkom J (Ed.). Essential IVF (Kluwer Academic Press, Norwell, MA, USA) pp. 291–332.
Biggers JD and Racowsky C. (2002) The development of fertilized human ova to the blastocyst stage in medium KSOMAA: is a two-step protocol necessary? Reprod Biomed Online 5:133–140.[Medline]
Biggers JD, McGinnis LK, Lawitts JA. (2004a) Enhanced effect of glycyl-L-glutamine on mouse preimplantation embryos in vitro. Reprod Biomed Online 9:59–69.[ISI][Medline]
Biggers JD, McGinnis LK, Summers MC. (2004b) Discrepancies between the effects of glutamine in cultures of preimplantation mouse embryos. Reprod Biomed Online 9:70–73.[ISI][Medline]
Biggers JD, McGinnis LK, Lawitts JA. (2005) One-step versus two-step culture of preimplantation embryos: is there a difference? Hum Reprod 20:3376–3384.
Gardner DK and Lane M. (1996) 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 11:2703–2712.
Lane M and Gardner DK. (1998) Amino acids and vitamins prevent culture-induced metabolic perturbations and associated loss of viability of mouse embryos. Hum Reprod 13:991–997.
Lane M and Gardner DK. (2001) Inhibiting 3-phosphoglycerate kinase by EDTA stimulates the development of the cleavage stage mouse embryo. Mol Reprod Dev 60:233–240.[CrossRef][ISI][Medline]
Lane M, Hooper K, Gardner DK. (2001) Effect of essential amino acids on mouse embryo viability and ammonium production. J Assist Reprod Genet 18:519–525.[CrossRef][ISI][Medline]
Lawitts JA and Biggers JD. (1991) Optimization of mouse embryo culture media using simplex methods. J Reprod Fertil 91:543–556.[Abstract]
Macklon NS, Pieters MHEC, Hassan MA, Jeucken PHM, Eijkemans MJC, Fauser BCJM. (2002) A prospective randomized comparison of sequential versus monoculture systems for in-vitro human blastocyst development. Hum Reprod 17:2700–2705.
Sinawat S, Wei-Chih Hsaio, Flockhart JH, Kaufman MH, Keith J, West JD. (2003) Fetal abnormalities produced after preimplantation exposure of mouse embryos to ammonium chloride. Hum Reprod 18:2157–2165.
Summers MC and Biggers JD. (2003) Chemically defined media and the culture of mammalian preimplantation embryos: historical perspective and current issues. Human Reprod Update 9:557–582.
Summers MC, McGinnis LK, Lawitts JA, Biggers JD. (2005) Mouse embryo development following IVF in media containing either L-glutamine or glycyl-L-glutamine. Hum Reprod 20:1364–1371.
Zander DL, Thompson JG, Lane M. (2006) Perturbations in mouse embryo development and viability caused by ammonium are more severe after exposure at the cleavage stages. Biol Reprod 74:288–294.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||