ABSTRACT
A method of labelling 8-celI-stage mouse blastomeres with [3H]thymidine is described, which allows them to be followed to the late blastocyst stage and is compatible with normal postimplantation development. However, the [3H]thymidine does affect the postimplantation vigour of the cells when placed in competition with unlabelled cells in blastomere aggregates. This suggests that caution should be used in interpreting results using [3H]thymidine as a label for early mammalian cells.
INTRODUCTION
In mammalian embryology, tritiated thymidine has been widely used as a marker for studying DNA synthesis (e.g. Mintz, 1964a, 1965; Samoshkina, 1968; Piko, 1970; Barlow, Owen & Graham, 1972) and for following the fate of cells in blastomere aggregation experiments (Mintz, 1964 b; Hillman, Sherman & Graham, 1972; Garner & McLaren, 1974). However, culture of 2-cell mouse embryos for long periods (12-70 h) in low concentrations of [3H]thymidine (0·01-0·1 μCi/ml) has been shown to have a deleterious effect on their development (Snow, 1973a, b). Snow found a sharp reduction in cell number at the 16-to 32-cell stage, and embryos that developed into blastocysts showed little or no inner cell mass (ICM) formation. These trophectoderm vesicles could implant when transferred to pseudopregnant recipients, but the resulting decidua contained only a few trophoblast giant cells and no ICM derivatives. Thus, the implicit assumption of previous workers, that [3H]thymidine did not interfere with development in the experiments they performed, may not be valid. Most authors did not compare the development of labelled and unlabelled blastomeres in their experimental systems, with the exception of Garner & McLaren (1974). They found that cell number in early blastocysts was not reduced by incubation in 0·01 μCi/ml [3H]thymidine from the 2-cell to the 8-cell stage.
In a study specifically designed to test the viability of blastomeres after [3H]-thymidine treatment, Horner & McLaren (1974) reported that embryos cultured continuously from the 2-cell stage in concentrations as low as 5 nCi/ml failed to develop normally after implantation, although cell number was not reduced at the blastocyst stage. Similar culture in 1 nCi/ml allowed normal postimplantation development in 40·5% of cases, but the level of labelling achieved by this procedure was not high enough to be useful for following cell behaviour in developing blastocysts (mean grain count/nucleus = 1·6, after three days in culture). In this respect even the labelling procedure of Garner & McLaren (1974) has limited application since they attained levels of labelling which could only be followed through two cell divisions to the 32-cell stage, but were too weak for labelled cells to be detected at the 64-cell stage. It is therefore important to determine whether a labelling procedure may be devised which allows blastomeres to be labelled sufficiently highly at the 8-cell stage to permit their fate to be followed to the mature blastocyst stage and which is also compatible with normal postimplantation development.
In various recent studies, we have used a procedure by which blastomeres may be labelled during the S-phase of the 8-cell stage at a level which allows their division products to be identified readily three or even four cell divisions later, at the late blastocyst stage. The viability of such labelled blastomeres was tested in two ways. Firstly, blastocysts derived from totally labelled 8-cell-stage embryos were transferred to pseudopregnant recipients to see if they would develop into normal postimplantation embryos. Secondly, labelled blastomeres were combined with unlabelled blastomeres of a different genotype for the isozymal variants of glucose phosphate isomerase (GPI-1) and the resulting chimaeric blastocysts transferred to pseudopregnant recipients. The contribution of the labelled blastomeres to any resulting conceptuses could be assessed by the presence of the appropriate GPI-1 isozyme. This competitive situation provides a very stringent test of their viability and vigour.
MATERIALS AND METHODS
(1) Source of embryos
All embryos were obtained from spontaneously ovulating mice. Embryos either homozygous aa or heterozygous ab at the Gpi-1 locus (De Lorenzo & Ruddle, 1969) were obtained from females of the PO stock (a random-breeding Swiss albino mouse from the Pathology Department, Oxford) mated with CFLP males (Anglia Laboratory Animals Ltd), known to be homozygous aa for Gpi-1. Embryos homozygous for the b allele of Gpi-1 were obtained from mice of the inbred strain C57BL/6J. The day that a copulation plug was found was designated day 1 of pregnancy.
(2) Dissociation and labelling of embryos
Embryos were recovered from the oviduct at the 4-to 8-cell stage on the morning of day 3. All embryos were recovered, stored and cultured in Whitten’s medium (Whitten, 1971), equilibrated with a 5% O2, 5% CO2, 90% N2 gas mixture and kept at 37 °C. Zonae were removed by pronase digestion (in dialysed, prewarmed, 0·25% pronase -Mintz, 1962) and the embryos were allowed to recover in Whitten’s medium for a short period.
PO embryos at the 4-cell stage were dissociated into their component blastomeres by gently sucking in and out of a flame polished micropipette. Each blastomere was then placed in a separate drop of culture medium under paraffin oil and observed every 30 min. When the blastomeres had divided to the 8-cell stage, some were left in culture and some were placed in Whitten’s medium supplemented with [3H]thymidine (specific activity 26 Ci/mw, Radiochemical Centre, Amersham, U.K.) at 0·25 μCi/ml for 2 h. At the end of the 2-hour period the blastomeres were removed and placed in microdrops of Whitten’s medium supplemented with 50% foetal calf serum. They were removed at intervals during the first hour to several successive drops and then cultured for 3-5 h to dilute the unincorporated label. At the end of this period they were rinsed in Whitten’s medium and returned to microdrops of this medium.
(3) Reaggregation of blastomeres
Four types of reaggregates using pairs of 8-cell-stage PO blastomeres were made.
Four pairs of labelled PO blastomeres, forming a totally labelled embryo.
Four pairs of unlabelled PO blastomeres, forming an unlabelled embryo.
One pair of labelled PO blastomeres with six 8-cell stage C57BL/6 blastomeres (unlabelled) (see Fig. 1).
One pair of unlabelled PO blastomeres with six 8-cell stage C57BL/6 blastomeres (unlabelled) (see Fig. 1).
Arrangement of blastomere composites at the 8-cell stage. Two PO blastomeres are surrounded by six C57BL/6 blastomeres.
All aggregates were cultured to the blastocyst stage.
(4) Autoradiography
Fully labelled and composite blastocysts were fixed for cell counts and autoradiography. Following a rinse in phosphate-buffered saline, blastocysts were fixed in fresh Heidenhain’s fixative (Tarkowski & Wroblewska, 1967). They were embedded in agar (Hillman et al. 1972) which was then treated as a large specimen for embedding in wax. Serial sections were cut at 5 μm and, after location of the blastocysts, the slides were dipped in Ilford K2 emulsion and exposed for 2 weeks. After developing, the slides were stained with haemalum and light green. The number of cells in each fully labelled blastocyst and the numbers of labelled and unlabelled cells in each composite blastocyst were counted.
(5) Transfers of the blastocysts
Groups of fully labelled PO blastocysts were transferred to one uterine horn of a pseudopregnant foster mother and, where possible, unlabelled PO blastocysts were transferred, as controls, to the other. Similarly, groups of composites made with labelled PO blastomeres were transferred, using composites made with unlabelled PO blastomeres as controls. Transfers were performed on the evening of the third or the morning of the fourth day of pseudopregnancy.
In addition, one or two experimental blastocysts from each transfer experiment were treated for autoradiography as described above to check the efficacy of the labelling regime.
(6) Assessment of postimplantation development of fully labelled embryos
Fully labelled embryos and unlabelled controls were treated in two ways. One group was recovered on day 6 and the implants fixed in AFA, mounted in wax and sectioned for histological examination. The other group was recovered on day 10 and the implants dissected and scored for the presence of embryonic derivatives.
(7) Gpi-1 analysis of transferred composites
Composite embryos recovered on day 10 were separated into three fractions: embryo, embryonic membranes and trophoblast. Each fraction was analysed for the isozymes of GPI-1 (Chapman, Whitten & Ruddle, 1971). One series of composites using unlabelled PO blastomeres was recovered at term.
RESULTS
(1) Autoradiography and cell counts at the blastocyst stage
Blastomeres fixed at this stage have gone through up to three divisions subsequent to labelling at S-phase of the 8-cell stage. Nevertheless, the levels of labelling are good (often more than 30 grains per nucleus) and labelled cells are readily identified (see Figs. 2 and 3).
Section of fully labelled PO blastocyst. (Two cells of an adjacent blastocyst appear to the right.)
Sections of blastocysts derived from composites made of two labelled PO blastomeres and six unlabelled C57BL/6 blastomeres. Arrows indicate labelled cells. A and B are adjacent sections of the same blastocyst.
Cell counts were made for the fully labelled blastocysts and unlabelled control blastocysts. The mean number for labelled blastocysts is 47·2 (range 32-59), which is similar to the means found in unlabelled blastocysts cultured for similar periods, 47·6 (range 36-59) for PO embryos and 41·9 (range 32-58) for C57BL/6 embryos.
Counts of labelled and unlabelled cells in the composites are shown in Table 1. In 16 out of 17 blastocysts, the labelled cells made up less than one quarter of the total number of cells, but the range was wide. The labelled cells were found in both ICM and trophectoderm of most blastocysts, and were morphologically indistinguishable from the unlabelled cells (see Fig. 3).
(2) Postimplantation development of fully labelled embryos
Normal embryonic development was found at day 6 and day 10 following transfer of fully labelled PO blastocysts and unlabelled controls (Table 2, Fig. 4).
Postimplantation development ofPO embryos labelled with [3H]thymidine at S-phase of the 8-cell compared with the development of unlabelled controls (only pregnant females considered)
![Postimplantation development ofPO embryos labelled with [3H]thymidine at S-phase of the 8-cell compared with the development of unlabelled controls (only pregnant females considered)](https://cob.silverchair-cdn.com/cob/content_public/journal/dev/35/1/10.1242_dev.35.1.95/3/m_develop_35_1_95tb2.png?Expires=1739965100&Signature=E5IZ69i8tmLEXitI4eyxzDaWOKORQEP6WlPJ699SH5sdewGZ2~FnGcO6whtCWGnMjdYr99ZadLQC5SYCuuc-wNH6FKoc04bWSDBv9wyNHmLkMR8JEwJTazIrLAwLXozzlU3NuoIAbBugdAjA8CFI22kyqEN2QoykSStBlO3new41TIYstVXNe3pOKoGndW2INekW-rtoft1~oVj2BhzBXp495TBypbjB9nUCc-GsF9xObmW0ehQ3a4JeiAI7X3NaDDVQ~kuPfUQduxgq9YwsIusPzdWotQwjeGe2TLRw9KrVDKuYe5ZNYSVWFU2orhPEHNIBq6vUJtMiZNcqKMHftw__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
(3) GPI-1 analysis of composites recovered at day 10
The GPI-1 analysis of 13 composites recovered on day 10, which had been made with labelled PO blastomeres, is shown in Table 3. Only one embryo (3) shows a PO-derived contribution to the embryo and membranes fraction. A further four (1, 10, 11, 13) show contributions to the trophoblast fractions.
Embryos recovered at day 10 from composites made at the 8-cell stage with two [3H] thymidine-labelled PO blastomeres and six unlabelled C57BL/6 blastomeres
![Embryos recovered at day 10 from composites made at the 8-cell stage with two [3H] thymidine-labelled PO blastomeres and six unlabelled C57BL/6 blastomeres](https://cob.silverchair-cdn.com/cob/content_public/journal/dev/35/1/10.1242_dev.35.1.95/3/m_develop_35_1_95tb3.png?Expires=1739965100&Signature=wDRI0vkxz0UPLnm1Ru~GGubJ0zY6Km7-ieKMyPi8YImPTT55suoq6ESyz~AYmW2JV~l7GYgrJMhYZnKlvSTSVjpB8GKWFtORP8XC0KvlxlZSKneQhZZYtOPF4HOqJGWmSRLsY5486sK~snRyZ4P~CbQkeQk1HravFC-cmj1Lpws2MUH9Iib2ab07WoAmY63ED7UVIFY~j5rUVzDcye48hNAjOQhuRa~ryDvkxIfXCqnNOSmSOqhb23ZY~iP6FTAxrFpi43UDPu3u4JdJElf7dFSsMZdo6bU37Rbv79irFzfHLgW754fk4umt2jCKebCxXihXVJyrWCASbZ8asBmnAQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
A comparable analysis of composites made with unlabelled PO blastomeres is shown in Table 4. All embryos showed PO-derived contributions; at day 10, when membranes and trophoblast fractions were available, these also showed PO-derived contributions.
DISCUSSION
A method of short-term incubation in [3H]thymidine has been devised which produces a level of label sufficient to follow cells through three or four cell divisions (Figs. 2 and 3). This method does not apparently reduce blastocyst cell number and, more importantly, does not prevent the labelled cells from forming live postimplantation embryos. A total of 7/10 live embryos was obtained from blastocysts which had been fully labelled by this method. In this small series no comment could be made on the relative implantation rate of the labelled and unlabelled control embryos. Nevertheless, these results establish that [3H]thymidine-labelled blastomeres are capable of normal postimplantation development.
When two labelled W blastomeres were combined with six unlabelledC57BL/6 blastomeres there was again no marked effect on cell number at the blastocyst stage. Labelled cells generally contributed just less than one quarter of the total cell number (Table 1). This represents a slight deviation from the expected proportion (one quarter) of the total cell number, which may or may not be a real effect, since labelled and unlabelled cells were derived from different strains. In these composites labelled cells were usually found in both the ICM and the trophectoderm (Fig. 3, Table 1).
When such composites were recovered at day 10 only 1 out of 11 embryo fractions and 4 out of 13 trophoblast fractions analysed contained PO contributions (Table 3). This contrasts strongly with the results from the composites of two unlabelled PO blastomeres and six unlabelled C57BL/6 blastomeres, where all conceptuses showed PO contributions to the embryo fractions (Table 4). Also all four conceptuses analysed at day 10 showed PO contributions to the trophoblast fractions. These results suggest that [3H]thymidine has a deleterious effect on the labelled PO blastomeres which only becomes apparent when the blastomeres are placed in a competitive situation.
The β-irradiation emitted by [3H]thymidine incorporated into DNA is known to produce a variety of deleterious effects on cells (see review by Wimber, 1964). Two of these effects are chromosome breakage and slowing of the mitotic rate. Chromosomes breakage, if not extensive enough to cause cell death, may be repaired. Similar damage caused by X-irradiation is known to be repaired in bacterial (Town, Smith & Kaplan, 1974) and mammalian cells (Wolff, 1972). Repeated repair of damage and rapid dilution of the label during development may explain why normal postimplantation embryos have been obtained from labelled cells in our present experiments. In studies on chick embryos, it has been shown that normal embryonic development is not impaired after [3H]thymidine labelling at various stages (Sauer & Walker, 1961; Weston, 1963).
Slowing of mitotic rate caused by [3H]thymidine provides a possible explanation for the poor development of labelled PO blastomeres in competition with unlabelled cells. While blastocyst cell number is not markedly reduced in fully labelled embryos, a slight retardation in mitotic rate may be indicated by the significant deviation from the expected proportion of labelled to unlabelled cells in composites (Table 1). At implantation a rapid increase in cell number occurs (Buehr & McLaren, 1974), and if the labelled cells are in fact slightly retarded, they will be strongly selected against during this mitotic burst. This will result in the labelled cells being more or less eliminated from the later conceptus. Presumably, when the labelled cells are in a non-competitive situation, such a slight delay prior to implantation is not sufficient to prevent continued development.
These results suggest that caution should be used in interpreting experiments where [3H]thymidine-labelled cells are placed in competition with unlabelled cells. Such a competitive effect might be reduced by lowering the concentration of the [3H]thymidine by half or reducing the time in labelled medium, either of which should be possible in view of the high level of labelling achieved by the present method.
ACKNOWLEDGEMENTS
We should like to thank Drs H. Alexandre, R. L. Gardner and C. F. Graham for valuable discussion during the preparation of this manuscript. Both authors were in receipt of Medical Research Council Research Studentships.