Embryonal stem (ES) cell lines, established in culture from peri-implantation mouse blastocysts, can colonize both the somatic and germ-cell lineages of chimaeric mice following injection into host blastocysts. Recently, ES cells with multiple integrations of retroviral sequences have been used to introduce these sequences into the germ-line of chimaeric mice, demonstrating an alternative to the microinjection of fertilized eggs for the production of transgenic mice. However, the properties of ES cells raise a unique possibility: that of using the techniques of somatic cell genetics to select cells with genetic modifications such as recessive mutations, and of introducing these mutations into the mouse germ line. Here we report the realization of this possibility by the selection in vitro of variant ES cells deficient in hypoxanthine guanine phosphoribosyl transferase (HPRT; EC 2.4.2.8), their use to produce germline chimaeras resulting in female offspring heterozygous for HPRT-deficiency, and the generation of HPRT-deficient preimplantation embryos from these females. In human males, HPRT deficiency causes Lesch-Nyhan syndrome, which is characterized by mental retardation and self-mutilation.

The numbers of cells in the trophectoderm (TE) and inner cell mass (ICM) of mouse blastocysts were counted by differentially labelling their nuclei with two polynucleotide-specific fluorochromes. Blastocysts recovered from the uterus at intervals between their formation early on Day 4 to the initial stages of implantation on day 5 were analysed. TE cell number increase was initially rapid, indicating some synchronisation of the sixth division, but slowed down progressively and plateaued on Day 5, possibly due to the onset of primary giant cell formation. ICM cell number increase was slower than the corresponding TE cells. As a result, TE cell number more than quadrupled, whereas ICM cell number only doubled over this period. Although the mitotic index of both populations of cells fell steadily, there was no significant difference between them. The decline in the proportion of ICM cells, therefore, is likely to be due to cell death, first detected in early blastocysts and predominantly located in the ICM. In addition, however, a contribution of ICM cells to the overlying polar TE cannot be excluded.

A rapid procedure has been devised to count the numbers of outer trophectoderm (TE) and inner cell mass (ICM) cells of mouse blastocysts by differentially labelling their nuclei in situ with polynucleotide-specific fluorochromes. The TE nuclei were labelled with propidium iodide (PI) by permeabilising the cells using selective antibody-mediated complement lysis (Solter and Knowles, '75). The blastocysts were then fixed in ethanol and the ICM nuclei labelled with bisbenzimide. These two fluorochromes have widely different fluorescent spectra. Thus, by using fluorescence microscopy with appropriate filter combinations, the PI-labelled TE nuclei appeared pink or red; the bisbenzimide-labelled ICM nuclei, blue or unlabelled. The total numbers of blastocyst nuclei and the numbers of ICM nuclei counted by differential labelling were similar to the numbers detected after spreading the nuclei of intact blastocysts or immunosurgically isolated ICMs by air-drying (Tarkowski '66). Differential labelling of TE and ICM nuclei in situ has two important advantages--that the numbers of both these cell types can be determined for individual blastocysts and that spatial relationships are partially preserved so that regional interactions can be studied.