Endogenous folic acid is essential for normal development of preimplantation embryos

Transcription

1 Human Reproduction vol.13 no.5 pp , 1998 Endogenous folic acid is essential for normal development of preimplantation embryos C.O Neill Human Reproduction Unit, Department of Physiology, University of Sydney, Royal North Shore Hospital of Sydney, St Leonards NSW 2065, Australia Preimplantation mammalian embryos develop with a high degree of autonomy. To date, there have been no unequivocal demonstrations of a requirement for vitamins in preimplantation embryo development. Reduced folic acid acts as an important methyl donor in many reactions including the synthesis of thymidine. Thymidine does not accumulate in cells so it might be expected that significant amounts of reduced folate would be required to support the exponential increase in DNA synthesis that occurs during early embryo development. The reduction of folate is catalysed by dihydrofolate reductase (EC ) which is selectively inhibited by the anti-cancer drug methotrexate. Methotrexate caused a dose-dependent inhibition of cell division in 1-cell, 2-cell and 8-cell mouse embryos with 50% inhibition of division occurring at concentrations of 1 10 µm. At a concentration of 0.1 µm only minimal inhibition of the initial cell division occurred, but continuous culture in this concentration of methotrexate completely inhibited further cell divisions. This suggests that most of the exogenous store of reduced folates was used in the first round of cell division. The effects of methotrexate were apparently primarily due to thymidine starvation, since a 10-fold excess of thymidine over methotrexate in culture media reversed the inhibition of development. Supplementing media with folic acid had no beneficial effect on the rate at which zygotes produced by in-vitro fertilization developed to the blastocyst stage. It is concluded that the development of the early embryo has an absolute requirement for reduced folate for thymidine synthesis which is met entirely by endogenous sources. Key words: dihydrofolate reductase/dna/embryo/folic acid/ methotrexate/methylation/thymidine During this phase of development (from the time of fertilization until blastocyst formation) the embryo undergoes an exponential increase in cell numbers and hence DNA synthesis (Barlow et al., 1972). Recent studies also show that embryos produce their own growth factors and receptors thus creating an autocrine loop which stimulates development (Rappolee et al., 1988; Paria and Dey, 1990; O Neill, 1997). There is no overall increase in the diameter of the embryo and there is marked degradation of oocyte-derived proteins (Latham et al., 1992) and RNA from the 2-cell stage (Schultz, 1993). Many of the substrates for the cleaving embryo s biosynthetic needs can therefore be met by degradation and salvage pathways. It is not clear, however, that such mechanisms account for DNA synthesis. It was shown (Alexiou and Leese, 1992) that the salvage pathways were insufficient to fulfil the demand for purines in the early mouse embryo. Production of the pyrimidines may be limited by the absence of a pool from which thymidine can be salvaged in the gametes. The necessary increase in the rate of DNA synthesis required as development proceeds would therefore demand an exponential increase in de-novo thymidine synthesis. Thymidine monophosphate is produced by the methylation of uridine monophosphate by thymidylate synthase (EC ; 5,10-methylene tetrahydrofolate:dump C-methyltransferase). The co-enzyme 5,10-methylene tetrahydrofolic acid (THF) is produced by the reduction of the vitamin, folic acid (pteroylglutamic acid), and by dihydrofolate reductase (EC ). Methotrexate is a potent and selective inhibitor of dihydrofolate reductase and results in inhibition of DNA synthesis (Metzler, 1977) due to thymidine starvation. If the embryo synthesizes thymidine by this pathway it would have a significant requirement for reduced folic acid, yet there are no reports of an essential requirement for folic acid in the development of the preimplantation embryo. In this study the necessity for folic acid in the development of the early embryo was examined by observing the effects of methotrexate. Introduction The preimplantation mouse embryo develops in relatively simple defined media. An optimal ionic composition of such media has recently been defined (Lawitts and Biggers, 1993), while the minimum energy substrate requirements have long been known (for review see Kaye, 1986). While an exogenous fixed nitrogen source seems not to be an absolute requirement for preimplantation stage development (Cholewa and Whitten, 1970), the addition of exogenous mixed amino acids does improve embryo viability (Lane and Gardner, 1994). Materials and methods Embryo culture and collection media In this study oocyte and embryo collection were performed in HEPESbuffered synthetic human tubal fluid medium (HEPES HTF) (Quinn et al., 1985) and embryo culture was performed in modified synthetic human tubal fluid medium (modified HTF) (O Neill, 1997) which is the same as HTF but also has 0.11 mmol/l EDTA and 1 mmol/l glutamine (based on CZB medium; Chatot et al., 1989). Modified HTF has the benefit of supporting development of zygotes through the 2-cell block. All components of media were tissue culture grade from Sigma Chemical Co (St Louis, MO, USA). Medium was 1312 European Society for Human Reproduction and Embryology

2 Folic acid and embryo development supplemented with bovine serum albumin 3 mg/ml (BSA, Fraction V; CSL Ltd, Melbourne, Australia). All cultures were performed in 10 µl drops in 60-well HLA plates (LUX 5260; Nunc Inc, Naperville IL, USA) overlayed with 10 ml heavy paraffin oil (BDH Scientific, Sydney, Australia). Embryos Random-bred Swiss albino mice (Sydney University, NSW, Australia), 6 8 weeks old, were superovulated by i.p. injection of 10 IU pregnant mares serum gonadotrophin (Folligon; Intervet International, Boxmeer, The Netherlands) followed 48 h later by an i.p. injection of 10 IU human chorionic gonadotrophin (HCG, Chorulon; Intervet). They were either left unmated or paired overnight with males of proven fertility and the presence of a copulation plug indicated day 1 of pregnancy. Cumulus masses or embryos were flushed from the reproductive tract with HEPES HTF. Embryos at the 2-cell stage were recovered by flushing the oviducts of mated animals 42 h after HCG injection. Zygotes fertilized in situ (ISF) were collected from the oviducts h after HCG. They were freed of any remaining cumulus cells by brief exposure to 300 IU hyaluronidase (Sigma) in HEPES HTF. They were then thoroughly washed in five changes of HEPES HTF. In-vitro fertilization (IVF) was performed as previously described (O Neill, 1997) with modifications. Two males, aged weeks and of proven fertility were used for each IVF procedure. Following cervical dislocation the epididymides were removed and placed into 1 ml of pre-equilibrated modified HTF in a Petri dish. The epididymides were punctured with a sterile needle and the spermatozoa gently squeezed out into medium. The dish was placed into an incubator at 37 C with 5% CO 2 in air for 40 min to allow the spermatozoa to disperse. Cumulus masses containing oocytes were then collected h after HCG and extensively washed in HEPES HTF. Groups of ~30 oocytes with their associated cumulus masses were placed into 5 ml plastic tubes (Falcon, Becton Dickinson Labware, Lincoln Park, NJ, USA) in 1 ml of HTF. Following dispersal of the spermatozoa, they were thoroughly mixed and the concentration assessed using a haemocytometer. Motile spermatozoa ( ) were added to each tube containing oocytes. The fertilization rate was assessed at 5 6 h after insemination by visualization of pronuclei. All fertilized oocytes were extensively washed in HEPES HTF to remove spermatozoa and cumulus cells and then pooled. Embryos of all types were recovered in a minimal volume and assigned to various treatments as required. Methotrexate hydrate (lot 102H10995), folic acid (lot 13H04085) and thymidine (lot 69F-08795) were purchased from Sigma. Each was prepared as 10 mm solutions fresh daily and diluted serially in modified HTF to achieve the desired concentrations. Monitoring embryo development Cell division of zygotes and 2-cell embryos was monitored visually using an Olympus dissecting microscope at 80 magnification. After the 8-cell stage the number of cells present in the embryo cannot be readily identified by this means, therefore cell counts were performed by visualization of cell nuclei following staining with 4 µg/ml Hoechst dye (bisbenzimide; Sigma). Nuclei were visualized using mercury lamp UV illumination and epifluorescence on a Nikon Optiphot microscope with Nikon filter block UV-1A. Statistical analysis The effect of treatments on cell division and progression to the blastocyst stage was assessed using χ 2 analysis. Where the assumptions of the χ 2 test were not met the conditional binomial exact test was used. The effect of methotrexate on cell division Table I. The number of cells present in mouse embryos collected fresh from the reproductive tract at the zygote, 2-cell and 8-cell stage of development following 20 h culture in increasing concentrations of methotrexate. There were 30 embryos in each treatment, and each experiment was repeated twice Methotrexate (µm) Number of cells per embryo after 20 h (mean SD) Zygotes 2-cell 8-cell ND a a a b Degenerate Degenerate ND not determined. Degenerate cells that had undergone extensive fragmentation with complete loss of normal morphology. a P 0.001; b P The superscript a shows the first concentration of methotrexate which caused inhibition of cell division compared with control. b Shows that inhibition at 100 µm was significantly greater in 8-cell than 1- or 2-cells. in cultured zygotes was assessed by logistic regression analysis, and 2-cell and 8-cell embryos by analysis of variance. The effect of folate supplementation on the number of cells in blastocyst after culture was assessed by analysis of variance. Individual comparisons of means after analysis of variance was performed by the least significant difference test. Results Methotrexate caused a significant (P 0.001) dosedependent inhibition of the cleavage of 1-cell (zygote), 2- cell and 8-cell mouse embryos (Table I) collected fresh from the reproductive tract demonstrating an essential role for dihydrofolate reductase at these stages of development. At each developmental stage, significant inhibition of cell division occurred at 1 µm. The concentration which caused a 50% reduction in the rate of cell division was between 1 10 µm for each developmental stage examined. While a concentration of 10 mm methotrexate was required to totally inhibit the division of fresh zygotes and 2-cell embryos, 8- cell embryos were significantly (P 0.01) more sensitive to the drug with complete inhibition of cleavage occurring at a concentration of 100 µm. For zygotes and 2-cell embryos cultured continuously in methotrexate, the cell divisions after the first division were more sensitive to inhibition. Approximately 50% of zygotes cultured in 1 µm methotrexate divided from the 1 2-cell stage overnight. However, when culture continued in methotrexate, none of these 2-cell embryos progressed to the 4-cell stage. A similar pattern of sensitivity to methotrexate occurred for embryos at the 2-cell stage. At a concentration of 1 µm methotrexate, slightly more than half the embryos divided; however, these embryos displayed no further cell division in the presence of this concentration of methotrexate. To determine whether the primary effect of methotrexate was due to inhibition of thymidine synthesis, embryos were cultured in the presence or absence of methotrexate in 1313

3 C.O Neill Table II. The effect of increasing concentrations of thymidine in embryo culture media on the first cell division and development of embryos to the blastocyst stage in the presence and absence of methotrexate Thymidine (µm) No methotrexate Methotrexate (10 µm) Proportion 2-cell Proportion blastocysts Proportion 2-cell Proportion blastocysts (at 20 h) (at 96 h) (at 20 h) (at 96 h) 0 15/20 15/20 5/30 0/ /20 12/20 2/30 0/ /20 12/20 14/30 13/30 a /20 13/20 19/30 a 15/30 a a No significant difference (P 0.05) in development compared with corresponding treatment in the absence of methotrexate. Table III. Effect of exogenous folic acid on in-vitro fertilization (IVF) zygote development after 96 h in culture Folic acid (µg/ml) Blastocysts (96 h) 35/50 40/50 38/50 37/50 31/50 No. of cells per blastocyst (mean SD) increasing concentrations of exogenous thymidine. Exogenous thymidine had no effect on the rate of zygote development in the absence of methotrexate (Table II). Methotrexate (10 µm) significantly (P 0.001) inhibited the proportion of zygotes dividing to the 2-cell stage and completely inhibited development to blastocysts, confirming the results in Table I. In the presence of methotrexate (10 µm), thymidine present in media supported a dose-dependent increase (P 0.01) in the rate of cleavage of zygotes to both 2-cell and to blastocyst stages. At a concentration of 100 µm, thymidine reversed the effects of methotrexate (Table II). Zygotes produced by IVF and cultured in media supplemented with folic acid at a concentration of µg/ml folic acid developed to blastocysts at the same rate as control embryos (Table III). Furthermore, the number of cells present in each blastocyst was unaffected by the presence of exogenous folic acid. Thus, exogenous folic acid had no effect on cell cycle progression or the rate of development up to the blastocyst stage. Discussion Folic acid in its reduced form serves as an important donor of methyl groups in methylation reactions. A rich source of this vitamin is leafy vegetables. It is known that dietary folic acid can be important for normal fetal development (Miller et al., 1989), and folate deficiency may limit fertility (Lindemann, 1993). The reduced form of the vitamin is tetrahydrofolic acid (THF). Folic acid is reduced to THF in two steps. Initially it is transformed spontaneously in the presence of NADPH 2 to 7,8-dihydrofolate. This intermediate is converted to 5,6,7, tetrahydrofolic acid by dihydrofolate reductase. The polyglutamyl derivatives of tetrahydrofolic acid are also important methyl donors. The anti-cancer drug methotrexate is a potent competitive inhibitor of dihydrofolate reductase. Since THF is oxidized in biosynthetic reactions the presence of methotrexate prevents its reduction back to THF. The resulting starvation of reduced folate limits methylation reactions including the synthesis of thymidine, methionine and purines (Metzler, 1977). It might also influence events such as polyamine synthesis and epigenetic marking of DNA by methylation through limiting the production of the methyl donor S-adenyl-methionine. Since thymidine does not accumulate in cells an important immediate consequence of methotrexate treatment is inhibition of cell division due to inhibition of DNA synthesis, secondary to thymidine starvation. In blastocysts, times more thymidine is required for DNA replication than is the case in the zygote, because of the increased number of cells. It is reasonable to ask how embryos meet this increased demand for reduced folate for thymidine synthesis. Adequate concentrations of folate co-enzymes might be achieved during early development by: (i) an increased uptake of folate by the embryo as it develops; (ii) increased activity of dihydrofolate reductase, giving faster turnover of oxidized folate; (iii) a reduced rate of DNA synthesis in embryo cells as development progresses (which would lead to progressively slower cell cycle times); or (iv) provision of sufficient pools of folate in the gametes to support subsequent development throughout this period. The extensive observations of adequate development of preimplantation embryos in defined media without folic acid

4 Folic acid and embryo development suggests that increased uptake of folate with embryo development is not required. Furthermore, addition of folic acid to media had no significant effect on the development of 8-cell hamster embryos (Kane and Bavister, 1988) or morulae-stage rabbit embryos (Kane, 1988) to the blastocyst stage. This study showed that exogenous folic acid had no beneficial effect on development of IVF zygotes to the blastocyst stage, implying that uptake of folates by the early embryo is not necessary for development. Despite this lack of effect of exogenous folate, a requirement for reduced folate in the cell division of zygotes, 2-cell and 8-cell mouse embryos is inferred by the dosedependent inhibition of cell division by methotrexate. However, it was interesting to note that compared with 8- cell embryos, both zygotes and freshly collected 2-cell embryos were less sensitive to methotrexate. Both zygotes and 2-cells were collected before the expected time of completion of DNA synthesis so the relative insensitivity was probably not due to the requirements for thymidine synthesis being completed prior to treatment. Rather the comparative resistance of zygotes and 2-cells to methotrexate may be due to the presence of relatively high concentrations of reduced folate within the embryo prior to treatment. If this was the case synthesis of thymidine (and hence DNA) would require little dihydrofolate reductase activity for the first round of DNA replication. The observation that zygotes that progressed to 2-cell embryos in the presence of methotrexate (0.1 1 µm) failed to undergo further cell division is consistent with this hypothesis. In this case, continual exposure of embryos to methotrexate meant that as THF and derivatives were oxidized to dihydrofolate, the pool of reduced folate was depleted to the extent that insufficient thymidine synthesis could occur to support complete DNA duplication. The observation that freshly collected 8-cell embryos were at least as sensitive to methotrexate as zygotes and fresh 2-cell embryos argues that the activity or concentration of dihydrofolate reductase was not increased at these later development stages. Indeed, by the time the 8-cell stage is reached, the components of the cytoplasm of the originating zygote have been partitioned into 8 cells, without an overall increase in the size of the embryo. This infers that there would be at least a 1/8 reduction in the available folates relative to the amount of DNA synthesis required in each blastomere of the 8 cells compared with zygotes. The steeper slope of the dose-response of inhibition by methotrexate in 8-cells compared with 1-cell and 2-cells may reflect this relative decline in available folate. However, the length of the cell cycle does not decline in 8-cell embryos, rather the first and second cell cycles are generally slower than later divisions, suggesting that there is significant reserve capacity of either folate, dihydrofolate reductase or both during early development. The inhibition of cell division by methotrexate was apparently largely due to the inhibition of thymidine synthesis, since supplementation of embryo culture media with thymidine could significantly reverse the inhibition of cell division in zygotes caused by methotrexate. Thymidine supplementation also caused a significant improvement in the proportion of embryos developing to the blastocyst stage in the presence of methotrexate. Together these results suggest that the inhibition of thymidine synthesis is the most important product involving the folate co-enzymes. It also suggests that there is an intact thymidylate synthase (EC ) pathway present in the early embryo. The lack of effect of exogenous thymidine on the rate of development of zygotes to the blastocyst stage may suggest that thymidine synthesis is not normally limiting in preimplantation development. The addition of folic acid to culture media was also without beneficial effect on either the rate of cell-cycle progression or the proportion of embryos produced by IVF that developed to blastocysts. It can be concluded from this study that reduced folic acid co-enzymes are essential for all stages of normal mouse preimplantation embryo development. All of the folic acid required for this development is present within the gametes at the time of fertilization. This apparent accumulation of folate within the zygote is sufficient to satisfy the requirements of the embryo during the preimplantation phase of development. The major role for the folate is in the synthesis of thymidine required for DNA synthesis and repair. The preimplantation mouse embryo appears to meet its biosynthetic needs by uptake of simple exogenous carbohydrate energy sources, the degradation and salvaging of macromolecules of gametic origin (which may be supplemented by uptake of exogenous substrates such as amino acids), the synthesis of autocrine growth factors and their receptors, and the accumulation within gametes of sufficient pools of recycled co-enzymes, such as folic acid. The ability of human embryos to develop in similar defined media in vitro, suggests that they also possess sufficient pools of folic acid to support early development obviating any need for its use as a media supplement. Acknowledgements I thank Adam Koch for expert technical assistance, Northern Sydney Area Health Service for financial assistance and Kathryn O Neill for editorial assistance. References Alexiou, M. and Leese, H.J. (1992) Purine utilisation, de novo synthesis and degradation in mouse preimplantation embryos. Development, 114, Barlow, P., Owen, D.A.J. and Graham, C. (1972) DNA synthesis in the preimplantation mouse embryo. J. Embryol. Exp. Morphol., 27, Chatot, C.L., Ziomek, C.A., Bavister, B.D. et al. (1989) An improved culture medium supports development of random-bred 1-cell mouse embryos in vitro. J. Reprod. Fertil., 86, Cholewa, J.A. and Whitten, W.K. (1970) Development of two-cell mouse embryos in the absence of a fixed-nitrogen source. J. Reprod. Fertil., 22, Kane, M.T. (1988) The effects of water-soluble vitamins on the expansion of rabbit blastocysts in vitro. J. Exp. Zool., 245, Kane, M.T. and Bavister, B.D. (1988) Vitamin requirements for development of eight-cell hamster embryos to hatching blastocysts in vitro. Biol. Reprod., 39, Kaye, P.L. (1986) Metabolic aspects of the physiology of the preimplantation embryo. In Rossant, J. and Pederson, R.A. (eds), Experimental Approaches to Mammalian Embryonic Development. Cambridge University Press, Cambridge, UK, pp Lane, M. and Gardner, D.K. (1994) Increase in postimplantation development 1315

5 C.O Neill of cultured mouse embryos by amino acids and induction of fetal retardation and exencephaly by ammonium ions. J. Reprod. Fertil., 102, Latham, K.E., Garrels, J.I., Chang, C. et al. (1992) Analysis of embryonic mouse development: construction of a high-resolution, two-dimensional gel protein database. Appl. Theoret. Elect., 2, Lawitts, J.A. and Biggers, J.D. (1993) Culture of preimplantation embryos. Methods Enzymol., 225, Lindemann, M.D. (1993) Supplemental folic acid: a requirement for optimizing swine reproduction. [Review.] J. Anim. Sci., 71, Metzler, D.E. (1977) Biochemistry, the Chemical Reactions of Living Cells. Academic Press Inc, New York, USA. Miller, P.N., Pratten, M.K. and Beck, F. (1989) Growth of 9.5-day rat embryos in folic-acid-deficient serum. Teratology, 39, O Neill, C. (1997) Evidence for the requirement of autocrine growth factors for development of mouse preimplantation embryos in vitro. Biol. Reprod., 59, Paria, B.C. and Dey, S.K. (1990) Preimplantation embryo development in vitro: Cooperative interactions among embryos and the role of growth factors. Proc. Natl. Acad. Sci. USA, 87, Quinn, P., Warnes, G.M., Kerin, J.F. et al. (1985) Culture factors affecting the success rate of IVF and embryo transfer. Ann. N.Y. Acad. Sci., 442, Rappolee, R.A. Brenner, C.A., Schultz R. et al. (1988) Developmental expression of PDGF, TGF-α and TGF-β genes in preimplantation mouse embryos. Science, 241, Schultz, R.M. (1993) Regulation of zygotic gene activation in the mouse. BioEssays, 15, Received on August 27, 1997; accepted on February 13,