J. Embryol. exp. Morph. 74, 297-309 (1983)
297
Printed in Great Britain © The Company of Biologists Limited 1983
X-chromosome instability in pluripotential stem cell
lines derived from parthenogenetic embryos
By E. J. ROBERTSON 1 , M. J. EVANS 1 AND
M. H. KAUFMAN2
From the Departments of Genetics and of Anatomy, University of Cambridge
SUMMARY
The karyotype of six pluripotential stem cell lines derived from haploid and two additional
lines derived from diploid parthenogenetic embryos is described. All these lines are diploid
and possess a normal autosomal complement. The stage at which diploidization of the haploid
cells occurs is not yet known. The XX-chromosome complement in these lines is unstable,
although in two haploid-derived lines and one diploid-derived line many normal XX-bearing
cells are found in early cultures. All of the lines so far examined either become XO (rarely),
or a single X chromosome shows a deletion in the distal region. The extent of this deletion
varies between lines, but the position of the breakpoint appears to be constant for a given line.
We suggest that these cytogenetic findings raise the possibility that a single deletion event
occurring at an early stage during the isolation of these lines may confer a selective advantage
to those cells carrying the deleted X chromosome.
INTRODUCTION
Although many mouse embryonal carcinoma (EC) cell lines derived from
teratocarcinomas maintain a near normal karyotype, few are euploid. Indeed,
many have been found to be XO, and it has often been assumed that these were
originally derived from XY lines in which the Y chromosome has been lost during
their passage either in vivo or in vitro. Several apparently euploid female
pluripotential lines have also been reported (Papaioannou, Evans, Gardner &
Graham, 1979; McBurney & Strutt, 1980; Mintz & Cronmiller, 1981) and this
has suggested that the XX constitution may be more stable than the XY.
Recently, techniques have been described which allow stem cells to be isolated
directly from mouse embryos (Evans & Kaufman, 1981; Martin, 1981). The
principal advantage of these embryo-derived (EK) cell lines is that, in addition
to their relative ease of isolation, they provide a source of pluripotential stem cell
lines which, at least initially, have a karyotype which is identical to that of the
embryo from which they are derived. Analysis of existing EK fertilized-derived
cell lines has revealed that over 70% are karyotypically stable, and retain a
1
2
Authors' address: Department of Genetics, Downing Street, Cambridge, U.K.
Author's address: Department of Anatomy, Downing Street, Cambridge, U.K.
EMB74
298 E. J. ROBERTSON, M. J. EVANS AND M. H. KAUFMAN
normal chromosome constitution (Robertson, Bradley, Evans & Kaufman,
unpublished).
In the present paper, we describe the analysis of six haploid-derived (HD
lines) and two diploid-derived (DP lines) lines isolated from parthenogenetically
activated eggs of both the 1-pronuclear (uniform haploid) and 2-pronuclear
(heterozygous diploid) classes (Kaufman, 1981). In both the HD and DP lines
the XX condition appeared to be unstable in the undifferentiated cell cultures.
In one line only a single X chromosome was found in all spreads examined, and
two other lines showed some cells with complete loss of one X chromosome.
More frequently, a deletion of the distal segment of one of the X chromosomes
was apparent. The extent of the deletion varied between lines, but remained
constant within a given line. The XX constitution would appear to be unstable
both in fertilized-derived and parthenogenetically-derived lines, but we have
only observed the phenomenon of X deletion in the parthenogeneticallyderived cell lines.
MATERIALS AND METHODS
1. Isolation and culture of parthenogenetically-derived lines
Eggs from 8- to 12-week-old superovulated 129/Sv//Ev and (C57BL x
CBA)Fi female mice were activated parthenogenetically following a brief exposure to 7 % ethanol in phosphate-buffered saline (PBS) (for details of the
procedure, see Kaufman, 1982). Only those activated oocytes that had
developed either a single pronucleus following second polar body extrusion
(uniform haploid class, Kaufman, 1981), or two pronuclei in the absence of
second polar body extrusion (heterozygous diploid class, Kaufman, 1981) were
used subsequently. The Fi activated eggs were cultured in vitro to the expanded
blastocyst stage. Because the 129/Sv//Ev activated eggs generally fail to
progress in culture beyond the 2-cell stage, pronuclear eggs were transferred to
the oviducts of recipients on the afternoon of the first day of pseudopregnancy
(day 1 = day of finding vasectomized plug). The recipients were subsequently
autopsied at about midday on day 4, by which time a high proportion of the
transferred embryos had achieved the blastocyst stage. The haploid and diploid
blastocysts were then treated in one of two ways.
i. in vivo 'delay'
Blastocysts were transferred to the uteri of ovariectomized recipients on day
3 of pseudopregnancy. While anaesthetized, the mice were injected with 1 mg of
Depo-Provera. Autopsies were carried out after 3-6 days, and 'delayed' blastocysts recovered. The latter were explanted into tissue culture, and pluripotential
-cell lines established from individual embryos. Details of the methodology involved have been described elsewhere (from fertilized embryos: Evans & Kaufman,
X-chromosome instability in pluripotential cell lines
299
1981; from haploid parthenogenones: Kaufman, Robertson, Handyside &
Evans, 1983).
ii. direct in vitro culture
A proportion of the expanded non-'delayed' blastocysts were briefly exposed
to acid Tyrode's medium to remove the zona pellucida, then transferred into
individual 1 cm wells (Nunc) containing a preformed feeder layer of inactivated
fibroblasts (Martin & Evans, 1975), and DMEM medium (Gibco) supplemented
with 10 % foetal calf serum, 10 % newborn calf serum, and 10~4M-2-mercaptoethanol. Blastocyst attachment occurred approximately 48 h after explantation.
After an additional 4-day interval, the inner-cell-mass-derived cell clumps were
selectively removed, trypsinized (0-25 % trypsin 1 0 ~ 4 M - E D T A in PBS) and
replated onto feeder layers. The trypsinization and replating procedure was
repeated after 4-6 days and, in successful cultures, nests of stem cells became
visible shortly after this second passage (see Evans & Kaufman, 1981).
All the EK cell lines were maintained exclusively on feeder layers as undifferentiated cultures and passaged at 4-5 day intervals by trypsinization. Cells
were cultured on feeder layers in order to retain their differentiation ability
(Hogan, 1976), and also to minimise alterations occurring in the chromosome
complement (Magrane, 1982). Samples of culture populations were routinely
frozen at intervals of two passage generations.
2. Chromosome analysis
Karyotype analysis was carried out on well-established culture populations
(usually after five to ten passages) in order to ensure that sufficient metaphase
spreads were available for G-banding analysis to be performed.
Exponentially-growing cell cultures were exposed to Colcemid (final concentration 0-02 meg/ml) for 50min. Cells were collected following trypsinization, incubated in hypotonic solution (0-075 M-KCI) for lOmin, pelleted by lowspeed centrifugation (500 r.p.m., 5 min) and fixed in 3:1 methanol: glacial acetic
acid fixative. The latter was changed a further two times, and spreads prepared
immediately by air drying.
The G-banding technique used was a modification of that described for rat
chromosomes by Gallimore & Richardson (1973). Slides were rinsed, air dried
and examined under oil immersion (x650) using a Ziess photomicroscope.
Between 30 and 60 intact G-banded metaphase spreads were analysed for each
of the eight lines studied. In the better preparations an unambiguous identification of all of the chromosomes present could be made. The total number of
chromosomes present was scored, and the morphological appearance of the X
chromosomes determined. Selected metaphase spreads were photographed and
karyograms constructed according to the nomenclature of Nesbitt & Erancke
(1973).
22
XX
16
XX
6
XX
1
XX
6
XX del
9
12
XX del
3
ael
xx
15
XO
7
XO
2
XO
3
X-chromosome constitution
(metaphase scored)
XX
XX del
XO
* Cell lines obtained by direct culture method from the 1-cell stage.
t H D 1 , H D 2 and H D 5 lines isolated from 129 S v / / E v strain embryos, all other lines from (C57BL x CBA)Fi
8
30
30
41
41 (73)
*DP2
is
13
31
40 (58)
*DP1
29
4
35
40 (83)
*HD6
24
7
31
40 (77)
HD5
57
2
60
40 (95)
HD4
37
44
29
40
45
30
40 (93)
40 (98)
40 (97)
HD1
HD2
HD3
Chromosome number
39
40
41
3
1
1
Total metaphases
counted
tCell line
Modal number
(percentage)
Table 1. Chromosome analysis of EK cell lines derived from parthenogenetic embryos
D
en
w
**
v>
O
2
§
w
o
o
w
X-chromosome instability in pluripotential cell lines
301
RESULTS
A total of six haploid-derived and two diploid-derived pluripotential lines have
been isolated from parthenogenetic embryos. The origin, method of isolation
and modal number of these lines are summarized in Table 1.
1. Lines derived from haploid parthenogenetic embryos
Six parthenogenetically-derived lines have been established from the
1-pronuclear (haploid) class of embryos. These have been termed 'haploidderived' or HD lines. A summary of the detailed karyological analyses carried
out on G-banded metaphase spreads collected from early passage cultures are
presented in Table 1.
All six lines were composed solely of diploid cells, and hence provide a source
of completely homozygous diploid cell lines. The modal number of
chromosomes in these lines was 40, with between 77-5 % and 98 % of cells
possessing this count (Table 1).
HDl
HD5, HD6
HD2
HD3
HD4
Fig. 1. A schematic representation of an X chromosome (based on the nomenclature
of Nesbitt & Francke, 1973) to illustrate the approximate position of the breakpoint
for each of the HD cell lines.
302 E. J. ROBERTSON, M. J. EVANS AND M. H. KAUFMAN
Between 10 and 30 spreads from each line were karyotyped, and all found to
have a completely normal diploid autosomal complement. As the original
haploid chromosome complement had been doubled, each cell would be expected to contain two identical X chromosomes. However, changes from the expected XX constitution were apparent in all lines. These changes fell into two
groups: very occasionally, the cells had become XO following the loss of a single
X chromosome, or, much more frequently, a deletion affecting the distal region
of a single X chromosome had occurred.
The presence of a partial deletion involving one of the X chromosomes was
recorded in all six HD lines, whether they had originated from the 129/Sv//Ev
or the Fi hybrid strain of mice, despite the fact that they had been isolated on
separate occasions over a 9 month period. The position of the breakpoint differed considerably between lines, but appeared to be constant for a given line.
A summary of the extent of the deleted segment in the various HD lines is
presented in Fig. 1, and representative examples of XXdel-chromosome pairs
from the various lines are illustrated in Fig. 2.
The most extreme form of the deletion is seen in the HD1 line in which
approximately 70 % of the chromosome is deleted, with only a small fragment of
the X chromosome remaining beyond the centromere. The deletion is least pronounced in the HD4 and HD5 lines in which the deletion involves approximately
35 % and 40 % of the total length of the X chromosome, respectively.
In three of the lines, metaphases were found in which two apparently normal
intact X chromosomes were present. These were present in low numbers in the
HD5 line (3 out of 12 metaphases scored), and HD6 line (1 out of 13 metaphases
scored), but present in approximately half of the cells of the HD4 line (16 out of
31 metaphases scored).
I"
HD1
HD2
-.5 i> II M i
HD3
HD4
HD4
HD5
1
HD6
• Fig. 2. X deletion in haploid-derived cell lines. Examples of pairs of G-banded X
chromosomes from the HD cell lines.
X-chromosome instability in pluripotential cell lines
303
The only haploid-derived line in which XO cells were recorded was the HD5
line (3 out of 12 metaphases scored).
Thus, in all of the lines studied, the majority of cells possessed an XXdel
genotype (Table 1).
All HD lines retained a normal diploid autosomal component. Figure 3 gives
a karyogram from the HD5 cell line, which is of 129/Sv//Ev origin, to illustrate
the possession a normal euploid chromosome complement.
2. Lines derived from diploid parthenogenetic embryos
Two parthenogenetically-derived stem cell lines have been isolated from the
2-pronuclear class of embryos. These lines differ from the haploid-derived lines
in that they contain heterozygous rather than homozygous diploid cells. Details
of their karyotype analysis are presented in Table 1.
Both lines appear to have an unstable XX constitution. The DPI line possesses
a normal diploid autosomal complement but has a bimodal chromosome count
with approximately equal numbers of cells containing 39 and 40 chromosomes.
This is due to the presence of both XO- and XX-genotype cells in the culture
population (see Table 1). A karyogram illustrating the XX genotype is given in
Fig. 4A. No partial deletion involving one X chromosome was observed in this
line. The DP2 line possesses cells containing either 40 or 41 (73 % of spreads)
I) if I! M
II *« II H
K
1 18 M
HD5
Fig. 3. A G-banded karyogram from the HD5 cell line (129/Sv//Ev origin) to show
a normal euploid chromosome complement.
304 E. J. ROBERTSON, M. J. EVANS AND M. H. KAUFMAN
chromosomes. G-banding revealed that all spreads were trisomic for
chromosome 11. Cells of XO, XX and XXdel genotype were recorded, though the
majority of cells was characterized by the possession of an XXdel genotype, in
which approximately 70 % of the material of one of the X chromosomes was lost.
A representative karyogram for this line is presented in Fig. 4B.
DISCUSSION
This paper is the second in a series in which the establishment and properties
III
* 5
If
* *
• •
• •
!r*-:
iif
4
•
t t
mm
m
*
V
ft
M
ft
••
fc
*»
f*
«t
DPI
Fig. 4. Karyograms from the DP lines. (A) DPI. Normal diploid complement. (B)
DP2. The cell line is characterized by possession of trisomy 11 (recorded in all
metaphase spreads examined). This particular metaphase spread illustrates the extent of the deletion of a single X chromosome found in this cell line.
X-chromosome instability in pluripotential cell lines
305
of parthenogenetically-derived pluripotential cell lines are presented. In the first
paper (Kaufman etal. 1983), details of the establishment of lines HD1, 2, 3 and
4 from 'delayed' blastocysts were presented, as well as preliminary observations
on their properties when allowed to differentiate in vivo and in vitro. Minimal
information was provided on the cytogenetic findings, as these had yet to be
investigated in detail. In the present paper, an attempt has been made to clarify
this aspect, with the presentation of details of the karyological analysis of the
original four HD lines, as well as two additional HD lines (HD5 and HD6), and
two diploid-derived lines (DPI and DP2).
However, before discussing the cytogenetic findings in detail, we believe it
worth stressing the fact that, contrary to our previous report (Evans & Kaufman,
1981) and also to the findings of Martin (1981) it is now possible to establish stem
cell lines directly from normal non-'delayed' blastocysts in the absence of
teratocarcinoma-conditioned medium: lines HD6, DPI and DP2, for example,
as well as various fertilized-derived lines (authors, unpublished) were
established in serum-containing tissue-culture medium from embryos that had
been maintained completely in vitro from the 1-cell stage onwards. Thus, while
the enlarged inner-cell-mass-derived component present in 'delayed' blastocysts, and appropriately 'conditioned' medium may both facilitate the isolation
of EK pluripotential lines, they are not necessary.
'A ii H M <
ti It
U
M «*
DP2
Fig. 4B
II
#1*
306 E. J. ROBERTSON, M. J. EVANS AND M. H. KAUFMAN
It is also of interest that the EK cell lines derived from normal non-'delayed'
blastocysts seem to be indistinguishable from those obtained from 'delayed'
blastocysts. Similarly, EK lines derived from parthenogenetic material were
indistinguishable from pluripotent cell lines established from teratocarcinomas
(see Evans & Kaufman, 1981; Kaufman et al. 1983).
Apart from confirming our previous finding (Evans & Kaufman ,1981) that the
autosomal complement of EK cells is apparently quite normal, the present study
has revealed that, contrary to expectation, the X-chromosome complement in
parthenogenetically-derived pluripotent lines does not appear to be stable. Possibly of greatest interest is the observation that in all of the lines so far studied,
a partial deletion involving one of the X chromosomes manifests itself in the lines
established from both haploid- and diploid-derived parthenogenetic material.
Whereas both groups appear to form stable diploid lines, the former are genetically homozygous, while the latter group - which form from embryos which retain
both products of the second meiotic division - are likely to be progressively more
heterozygous at loci increasingly distal from the centromere (Eicher, 1978). In
the most extreme cases, both in the HD and DP lines, an entire X chromosome
is lost from the complement, resulting in the production of an XO line. While the
complete loss of an X chromosome was only rarely encountered in most of the
parthenogenetically-derived lines studied, the presence of a partial deletion of
one of the X chromosomes occurred regularly.
While X-inactivation occurs apparently normally both within the embryo
(Kaufman, Guc-Cubrilo & Lyon, 1978) and in the extraembryonic membranes
(Rastan, Kaufman, Handyside & Lyon, 1980) in heterozygous diploid
parthenogenones, no information appears to be available on X-chromosome
activity in homozygous diploid embryos.
There is considerable evidence that both X chromosomes are active during
early development of female mouse embryos (reviewed by Gartler & Cole, 1981)
and this has also been shown to be the case in some XX EC cell lines (Martin et
al. 1978; McBurney & Strutt, 1980). Presumably both X chromosomes are also
active in XX EK cell lines and indeed in such lines no differential Kanda staining
was observed (unpublished observations A. Stoker). Although no biochemical
information is yet available on X-chromosome activity in the various
parthenogenetically-derived EK lines considered here, there seems no a priori
reason why the X chromosomes should behave differently from the situation
observed in the tumour-derived EC cells.
The female mammal has twice the relative dose of X-linked to autosomal
genes of the male, and X-inactivation provides a mechanism of dosage compensation. That there is a dosage imbalance in XX EC cells is born out by the
observations of a double specific activity of X-linked enzymes in these cells
(Martin et al. 1978; see also Epstein, Travis, Tucker & Smith, 1978; Monk,
1978). This is to be expected as the cell phenotype is of an early embryo epiblast
cell. All X-linked products are, therefore, overproduced and these cells are in
X-chromosome instability in pluripotential cell lines
307
a state of imbalance. Despite this, XX cells not only grow through this stage of
development, but may be isolated and maintained as progressively growing
cultures of EK cells.
Analysis of 21 EK cell lines derived from fertilized embryos shows that 15 of
these are karyotypically normal (authors' unpublished results). To date we have
not observed the loss of a Y chromosome from the 12 XY lines available, however the loss of an entire X chromosome has been observed in two of the three
(otherwise euploid) XX lines studied with the consequent production of both XX
and XO cells in these cultures. Whilst it is clear, therefore, that XX lines can give
rise to XO cells, we have not yet observed a partial deletion of an X chromosome
in any fertilized-derived EK cell line nor has this phenomenon been reported to
occur in any established EC cell line.
It is perhaps relevant that a number of established EC cell lines (which have
been derived from tumours made by ectopic transplantation of early embryos)
are XO. While it has usually been assumed that these arose from XY lines due
to the loss of the Y chromosome, this preliminary EK evidence suggests that this
is perhaps more likely to be due to the loss of an X chromosome. Conceivably,
in contrast to an XX EK line with both X chromosomes active, an XY EK cell
line is carrying little redundant metabolic burden from the small Y chromosome.
The reason why XO lines are found to arise more frequently from XX than from
XY lines could thus be that the removal of one redundant X chromosome confers
a more significant growth advantage on the cell relative to its XX progenitors.
It is noticeable that those long-established EC cell lines which do possess two X
chromosomes also have an increased number of autosomes (McBurney &
Adamson, 1976; Martin et al. 1978) and some tend not to be pluripotential (e.g.
Nulli SCC2A, E. J. Robertson, unpublished results).
Possibly the fact that in most teratocarcinomas one of the X's is maternally
derived and the other paternally derived has some effect on their inactivation
during early cellular differentiation (as is known to be the case in the normal
embryo, West, Frels, Chapman & Papaioannou, 1977; Harper, Fosten & Monk,
1982). Their parental origin may also in some way influence the loss of one or the
other X chromosome in the undifferentiated state. Clearly, this cannot be the
case in the parthenogenones, where both X's are of maternal origin.
Although a variety of breakpoints are observed in the different
parthenogenetically-derived lines studied here, it is unclear at the present time
whether a progressive loss of the distal segment occurs, with the eventual production of an XO line. The uniform state of the deleted segment in each of the lines
so far examined, however, would suggest that this is not the case, but even in the
most well-established lines, cells from only relatively early passages have been
examined. This is in marked contrast to the situation observed in many EC lines
in which cells may have passed through hundreds of generations both as in vivo
tumours and in subsequent growth in vitro. The latter situation would certainly
tend to favour the selective overgrowth and survival of the cells with the most
308 E. J. ROBERTSON, M. J. EVANS AND M. H. KAUFMAN
stable genotype. Indeed, time alone will tell whether the XXdel genotype observed in these lines represents an early stage in the production of the (possibly)
more stable XO genotype. If a variety of distal Xdel conditions are sufficient to
confer the postulated growth advantage on the cell it suggests that either the most
serious imbalance is due to loci which map at the distal region of the X
chromosome, or that deletion of this region removes a controlling locus which
affects the activity of the rest of the chromosome. It would be interesting to
discover if the Xdel chromosome is active in these cells.
Whatever the cause and mechanism, a series of X-chromosome deletions are
now available, and it seems likely that many more of these apparently randomlyoccurring deletions could be produced. These may be extremely useful for genetic mapping of the X chromosome. If the XXdel chromosomes remain active, the
relative specific activities of X-linked enzymes in XX, XXdel and XO cells, which
are identical in all other respects, should provide mapping information. Similarly, DNA from such cells might be used to locate recombinant DNA clones from
mouse X-chromosome libraries.
Further studies will obviously be required to clarify many of the points raised
in this discussion. Does, for example, the XX state in parthenogeneticallyderived EK cells progress to the XXdel state, which then eventually progresses
to the 'stable' XO state? Do fertilized-derived EK XX cells eventually progress
along a similar pathway? A study of homozygous diploid pluripotential lines
established from enucleated fertilized diploidized embryos (Markert & Petters,
1977) might help to clarify some of these issues.
We would like to thank Lesley Cooke and Mary Knox for their excellent technical assistance. This work has been supported by a grant from the Medical Research Council.
REFERENCES
E. M. (1978). Murine ovarian teratomas and parthenotes as cytogenetic tools.
Cytogenetics Cell Genet. 20, 232-239.
EPSTEIN, C. J., TRAVIS, B., TUCKER, G. & SMITH, S. (1978). The direct demonstration of an
X-chromosome dosage effect prior to implantation. In: Genetic Mosaics and Chimeras in
Mammals (ed. L. B. Russell), pp. 261-268. N.Y.: Plenum Press.
EVANS, M. J. & KAUFMAN, M. H. (1981). Establishment in culture of pluripotential cells from
mouse embryos. Nature, Lond. 292, 154-156.
EICHER,
GALLIMORE, P. H. & RICHARDSON, C. R. (1973). An improved banding technique exemplified
in the karyotype analysis of two strains of rat. Chromosoma 41, 259-263.
S. M. & COLE, R. E. (1981). Recent developments in the study of mammalian Xchromosome inactivation. In: Mechanisms of Sex Differentiation in Animals and Man (eds
C. R. Austin & R. G. Edwards), pp. 113-143. N.Y.: Academic Press.
HARPER, M. I., FOSTEN, M. & MONK, M. (1982). Preferential paternal Xinactivation in extraembryonic tissues of early mouse embryos. /. Embryol. exp. Morph. 67, 127-135.
HOGAN, B. L. M. (1976). Changes in the behaviour of teratocarcinoma cells cultivated in vitro.
Nature, Lond. 263, 136-137.
KAUFMAN, M. H. (1981). Parthenogenesis: a system facilitating understanding of factors that
influence early mammalian development. In: Progress in Anatomy Vol. 1 (eds R. J. Harrison & R. L. Holmes), pp. 1-34. Cambridge University Press.
GARTLER,
X-chromosome instability in pluripotential cell lines
309
M. H. (1982). The chromosome complement of single-pronuclear haploid mouse
embryos following activation by ethanol treatment. /. Embryol. exp. Morph. 71,139-154.
KAUFMAN, M. H., GUC-CUBRILO, M. & LYON, M. F. (1978). X-chromosome inactivation in
diploid parthenogenetic mouse embryos. Nature, Lond. 288, 172-174.
KAUFMAN, M. H., ROBERTSON, E. J., HANDYSIDE, A. H. & EVANS, M. J. (1983). Establishment
of pluripotential cell lines from haploid mouse embryos. /. Embryol. exp. Morph. 73,
249-261.
MAGRANE, G. G. (1982). A Comparative Study of Human and Mouse Teratocarcinomas.
Ph.D. Thesis. University of London.
MARKERT, C. L. & PETTERS, R. M. (1977). Homozygous mouse embryos produced by
microsurgery. /. exp. Zool. 201, 295-302.
MARTIN, G. R. (1981). Isolation of a pluripotent cell lines from early mouse embryos cultured
in medium conditioned by teratocarcinoma stem cells. Proc. natn. Acad. Sci., U.S.A. 78,
7634-7638.
KAUFMAN,
MARTIN, G. R., EPSTEIN, C. J., TRAVIS, B., TUCKER, G., YATZIV, S., MARTIN, D. W., CLIFT,
S. & COHEN, S. (1978). X-chromosome inactivation during differentiation of female
teratocarcinoma stem cells in vitro. Nature, Lond. 271, 329-333.
G. R. & EVANS, M. J. (1975). Differentiation of clonal lines of teratocarcinoma
stem cells: formation of embryoid bodies in vitro. Proc. natn. Acad. Sci., U.S.A. 72,14411445.
MCBURNEY, M. W. & ADAMSON, E. D. (1976). Studies on the activity of the X-chromosomes
in female teratocarcinoma cells in culture. Cell 9, 57-70.
MCBURNEY, M. W. & STRUTT, B. J. (1980). Genetic activity of X-chromosomes in pluripotent
female teratocarcinoma cells and their differentiated progeny. Cell 21, 357-364.
MINTZ, B. & CRONMILLER, C. (1981). METT-1: A karyotypically normal in vitro line of
developmental^ totipotent mouse teratocarcinoma cells. Som. Cell Genet. 7, 489-505.
MONK, M. (1978). Biochemical studies on mammalian X-chromosome activity. In: Development in Mammals. Vol. 3 (ed. M. H. Johnson), pp. 189-224. Amsterdam: North Holland.
NESBITT, M. N. & FRANCKE, U. (1973). A system of nomenclature for band patterns of mouse
chrmosomes. Chromosoma 41, 145-158.
PAPAIOANNOU, V. E., EVANS, E. P., GARDNER, R. L. & GRAHAM, C. F. (1979). Growth and
differentiation of an embryonal carcinoma cell line (C145b). J. Embryol. exp. Morph. 54,
277-295.
RASTAN, S., KAUFMAN, M. H., HANDYSIDE, A. H. & LYON, M. F. (1980). X-chromosome
inactivation in extra-embryonic membranes of diploid parthenogenetic mouse embryos
demonstrated by differential staining. Nature, Lond. 288, 172-174.
WEST, J. D., FRELS, W. I., CHAPMAN, V. M. & PAPAIOANNOU, V. E. (1977). Preferential
expression of the maternally derived X-chromosome in the mouse yolk sac. Cell 12,
873-882.
MARTIN,
{Accepted 17 November 1982)