/ . Embryol. exp. Morph. Vol. 61, pp. 117-131, 1981
Printed in Great Britain © Company of Biologists Limited 1981
Expression of Forssman antigen in the
post-implantation mouse embryo
By M. G. STINNAKRE, 1 - 2 M. J. EVANS,1 K. R. WILLISON, 34
AND P. L. STERN 3 5
From the Department of Genetics,
University of Cambridge and MRC Laboratory of
Molecular Biology, Cambridge
SUMMARY
The expression of Forssman antigen on the surface of cells of post-implantation mouse
embryos between 5 and 1\ days old and of cells of the gonads from 10£ days has been
followed using the monoclonal antiserum Ml/22.25. In the early post-implantation embryo
a lineage-related distribution is found.
The inner cell mass of the blastocyst was previously shown to be Forssman antigen
positive and its derivative tissues the epiblast of the 5-day embryo and the primary embryonic
endoderm are also positive. The endoderm cells remain positive both over the embryonic
and extraembryonic portions of the embryo but the epiblast becomes Forssman antigen
negative as it differentiates into embryonic ectoderm. The extraembryonic ectoderm which
is derived from the Forssman negative trophectoderm remains negative throughout. The
primordial germ cells are Forssman antigen positive from their first appearance in the
germinal ridge until day 14 when they become negative but after that time it is other cells
not related by direct lineage which become Forssman antigen positive. These are tentatively
identified as Sertoli cells precursors as it is the Sertoli cells which are the antigen-positive
population in the adult testis.
INTRODUCTION
There is now abundant evidence for cell-surface phenotypic diversity within
a single organism. This diversity can be recognized in a number of different
ways and must reflect the varied functional properties of distinct cell types.
In particular individual plasma membrane components can be recognized
immunologically by specific antisera and one can use such antisera to identify
1
Author's address: University of Cambridge, Department of Genetics, Downing Street,
Cambridge CB2 3EH, U.K.
2
Present address: Institut de Recherche en Biologie Moleculaire, Tour 43, 2 Place
Jussieu, Paris 5e, France.
3
Author's address: Medical Research Council, Laboratory of Molecular Biology, Hills
Road, Cambridge, U.K.
4
Present address: Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 11724,
U.S.A.
5
Present address: University of Oxford, Department of Zoology, South Parks Road,
Oxford, U.K.
118
M. G. STINNAKRE AND OTHERS
and separate cells even though the structure and function of these components
is largely unknown.
Many antisera which recognize antigens shared between embryonal carcinoma
cells (EC) (the stem-cell line of mouse teratocarcinoma) and early mouse
embryo cells have been described (Erickson, 1977). The pattern of reactivity
reported for such antisera falls into two classes: those that recognize antigens
expressed on 'undifferentiated' cells in embryo, EC, and only the germ-cell
line in the adult (Artzt et al. 1973); and those that share essentially the same
embryo-EC distribution of reactivity but also react with cells of other adult
tissues apart from the germ line, for example brain and/or kidney (Stern et al.
1975).
Antisera are complex mixtures of antibodies with multiple specificities and
the two above classes may represent overlapping sets of reactivities. It is
difficult to define the shared and unique specificities of such conventional
antisera precisely. Monoclonal antibody technology allows the production of
large quantities of antibodies to single antigenic determinants (Kohler &
Milstein, 1976). This avoids the problems associated with the interpretation
of the reactions of conventional heterogeneous antisera.
Two monoclonal antibodies have recently been described which belong to
the second class because they react with early embryo cells and EC and some
tissues of the adult mouse (Stern et al. 1978; Willison & Stern, 1978; Solter &
Knowles, 1978). One of these, a rat monoclonal antibody Ml/22.25 was
directed against a Forssman antigenic determinant (FA). This determinant
was shown in previous studies to be first expressed on the trophectoderm of
the early blastocyst (Willison & Stern, 1978). It disappeared from this tissue
at the time of implantation but continued to be present on the cells of the
inner cell mass and primary endoderm. This paper is concerned with the
expression of FA in the post-implantation phase of mouse embryogenesis
with particular emphasis on its expression by the cells of the germ line through
to the adult. The pattern of expression of FA differs markedly from other
EC-embryo antigens, including that of F9 antigen(s) (Buc-Caron, Condamine
& Jacob, 1978) and demonstrates a complex expression in the genital tissues.
Some of these results have been previously described in part. (Evans, LovellBadge, Stern & Stinnakre, 1979).
MATERIALS AND METHODS
Mice
All embryos were taken from an inbred stock of 129 Sv Ev C P mice in
which heterozygosity for the steel gene SI* was maintained. They were caged
in pairs in ambient daylight conditions and examined for mating plugs each
morning. The day when a plug was detected was considered as day 0 of pregnancy and conception was assumed to have occurred midway through the
Expresion of Forssman antigen in mouse embryo
119
dark period. This was not always the case as litters varying in their development
by up to ± % day were found. Allowance was made for this in the interpretation
of the results. Litters examined were either from wild type or from SP/ + x SP/ +
matings. The Wy gene was on the H x 101 stock (Heath, 1978) at F7 and adult
WyIWY males were used for tissue absorption studies.
Isolation and dissection of embryos
Four-day embryos were flushed from the uterus in calcium- and magnesiumfree phosphate-buffered saline (PBS) and 5- to 7|-day embryos were dissected
from their decidua in PBS. Reicherts membrane was opened, most of the
trophoblastic cone discarded and the embryo divided by a transverse cut into
embryonic and extraembryonic regions. These portions were incubated in
cold 0-1 % pancreatin in PBS or 0-25% trypsin for a few minutes, foetal calf
serum added and the ectodermal and endodermal layers dissected apart either
with tungsten needles or by pipetting through a drawn-out, siliconized pasteur
pipette. The isolated tissue layers were further disaggregated when required
by a second treatment with trypsin alone.
Ten and a half-day and later embryos were dissected from their decidual
tissue and extraembryonic membranes. Liver, brain and spinal cord were
isolated and disaggregated by trypsin treatment. The germinal ridges were
dissected and freed from the accessory ducts and blood vessel. Twelve and
a half-day and older embryonic gonads were sexed by their appearance and
the presence of the spermatic blood vessel. Earlier embryos were sexed by
orcein staining of the amnion and examination for the presence or absence of
Barr Bodies (Farias, Kajii & Gardner, 1967). The two methods were crosschecked with each other on representative litters of embryos. In order to
identify SI/SI embryos of between 1O£ and 12^ days one genital ridge or half
of one ridge was stained for the presence of alkaline phosphatase activity by
squashing and incubation at room temperature in staining solution (0-075 MTris/HCl buffer pH 8-6 containing 0-8 % NaCl, 0-05 % disodium napthol ASMX
phosphate, 2 % dimethylformamide and (freshly added) 0-01% fast blue BB
diazo salt). By this method unfixed cells containing alkaline phosphatase are
stained deep blue within 5 min and the germ cells are readily seen (Chiquoine,
1954). The germ ridges of SI'/SI embryos are almost entirely lacking in germ
cells (Bennett, 1956; McCoshen & McCallion, 1975) and are identified by this
technique. The contralateral ridge was used for further examination. Such
germ-cell-deficient ridges were not found in wild-type litters and their correspondence with the SI/SI genotype was confirmed in preliminary experiment
by grafting pieces of dorsal embryo skin to adult hosts and examining the
graft tissue after 3 weeks for hair pigmentation (Bennett, 1956; Stevens, 1967).
Cells from germinal ridges were disaggregated by trypsin treatment and
suspended in Dulbecco-modified Eagles medium containing 18 mM-HEPES,
0-1 % sodium azide and 10% heat-inactivated foetal calf serum. Preparation
120
M. G. STINNAKRE AND OTHERS
of cell suspensions from adult testes and of other tissues for the absorption
assays were as previously described (Stern et al. 1978).
Immunological assays
The monoclonal antibody preparation was the same as previously described
(Stern et al. 1978) and a single batch of antiserum directly conjugated to
fluorescein was used for all the fluorescence studies. Cells were suspended in
i-diluted direct-coupled Ml/22.25 (M1/22.25-F1) for ^ h at 0 °C, washed
twice in suspension medium and examined with a Zeiss photomicroscope
equipped with epifluorescent illumination. Cells were observed by phasecontrast and fluorescence microscopy and only live cells scored. In some
cases cells were also stained for alkaline phosphatase (as above), or for A 5-3 /?
hydroxysteroid-dehydrogenase (Wattenberg, 1958). The deposition of the blue
stain did not obliterate the very bright fluorescence obtainable with the
M1/22.25-F1 reagent. For cytotoxicity assays, the cells were suspended in
^-diluted M 1/22.25 at 0 °C for \ h. The cells were washed in suspension
medium and then incubated at 37 °C in ^-diluted Guinea-pig complement GPC
(L.I.P. Ltd, U.K.) for ^ h . The cells resuspended in PBS containing 0-1%
trypan blue and examined immediately. Killed cells take up the blue stain.
Control incubations without M 1/22.25 treatment but with complement treatment were done in parallel.
RESULTS
1. Expression of FA in 4%- to 6-day embryos
Inner-cell-mass cells isolated from 3Jr- to 4^-day embryos express Forssman
antigen (Willison & Stern, 1978). The inner cell mass differentiates into
embryonic ectoderm and endoderm at about the time of implantation. When
these two tissues are isolated from 4f- to 5^-day embryos they are strongly
Forssman antigen positive. The extraembryonic ectoderm, which is derived
from the trophoblast (Gardner & Papaioannou, 1975), is Forssman antigen
negative during this period. By 5f-days, the intensity of FA labelling of the
embryonic ectoderm is reduced but both the parietal and visceral endoderm
is still brightly fluorescent.
2. Expression of FA in 6- to 7\-day embryos
When 6- to 7f-day egg-cylinder stages are dissected into embryonic and
extraembryonic regions and the germ layers further dissected and examined
either as tissue sheets or cell suspensions by Ml/22.25 immunofluorescence or
cytotoxicity, those dissects containing endodermal cells have significantly
greater proportions of antigen-positive cells. The majority of antigen-expressing
cells have an endodermal morphology; these cells being generally larger and
with more granular cytoplasms than the ectodermal cells. An example
of such an experiment is shown in Table 1. Separated disaggregated
Expression of Forssman antigen in mouse embryo
121
Table 1. Direct immunefluorescence with Ml/22.25 antibodies conjugated to
fluorescein on disaggregated cells of separated germ layers of 7\-day embryos
% cells FA-positive
(Cells nos. in brackets)
Embryonic
portion
Embryonic
Extra-embryonic
Germ layer
Granular
cytoplasm
% cells FA-negative
(Cell nos. in brackets)
Smooth Granular
cytoplasm cytoplasm
Ectoderm*
0
0
Endoderm
76(82)
0
Ectoderm
1 (1)
0
Endoderm
63 (109)
0
* Also includes some mesoderm.
0
4(4)
18 (28)
1 (1)
Smooth
cytoplasm
100(148)
20(22)
81 (126)
36 (62)
embryonic ectoderm from 7-5-day embryos has no antigen-positive cells whereas
embryonic endoderm has 76 % FA-positive cells. The cells which are antigennegative in embryonic endoderm preparations have a morphology consistent
with their being contaminating embryonic ectoderm. Table 1 shows a similar
result for extraembryonic ectoderm and endoderm. Both parietal and visceral
endoderm continue to express the Forssman antigen.
Cytotoxicity experiments confirm the results obtained by immunofluorescence
with M 1/22.25 antibodies. In the experiment shown in Table 1, for example,
70 % of the embryonic ectoderm cells were alive after incubation with M 1/22.25
and guinea-pig complement whereas only 13% of the embryonic endoderm
cells survived. All the latter were small cells with a smooth cytoplasm - a
morphology suggesting that they were contaminating embryonic ectodermal
cells. When portions of a 7-day egg cylinder with both embryonic ectoderm and
endodermal cells were treated with Ml/22.25 antibodies and GPC only the
embryonic ectoderm cells remain viable, as judged by exclusion of trypan blue.
A similar dissect treated only with GPC shows few cells stained. Figure 1
shows the same result by immunofluorescence. In summary, FA appears to
be expressed on the embryonic ectoderm and endoderm before day 6. After
this time the majority of embryonic ectoderm cells no longer have detectable
antigen expression whereas both types of endodermal cells continue to express
it strongly.
3. Expression of FA in liver and brain of 10- to 13-day embryos
From 8 days because of the rapidly increasing complexity of the embryo
it can no longer be easily dissected into component germ layers. In the adult,
FA is detectable in the brain, kidneys and testes, although not in liver. Table 2
shows that cell suspensions from day-10 to -13 brain and spinal-cord preparations
contain increasing proportions of M 1/22.25 reactive cells, whereas those from
liver contain less than 0-1% antigen-positive cells. It appears that cells in
122
M. G. STINNAKRE AND OTHERS
0-01 cm
Fig. 1. A portion of the embryonic region of a 7-day egg cylinder stained with
fluorescein-conjugated Ml/22.25 showing (a) phase-contrast appearance, (b) fluorescence. A sheet of embryonic ectodermal cells are seen protruding from the
edge of the fragment and they are totally unstained in contrast to the brightly
stained embryonic endodermal cells.
Table 2. Tissue distributions
Day
Liver
% Ml/22.25 antigen'
positive cells
Spinal cord/brain
(3)t
10
< 0-1 % (3)
< 01
11
01
0-5
12
40
13
* Indirect immunofluorescence with at least 1000 cells examined using rabbit anti-Rat Ig
conjugated tofluorescein(Stern et al. 1978).
t No. of experiments.
brain/spinal cord begin to express FA at about day 11. Other tissues, with
the exception of genital ridges (see below), were not examined. The antigenpositive cells in brain remain to be identified.
123
1(4,2)
Expression of Forssman antigen in mouse embryo
40
-
—
1 11
i
s
20
CO
i ft
n
d+9
11-day
(3,3)
30
s
w
d'
9
12-day
d
9
13-day
d
9
14-day
6-
9
15-day
11
d
9
16-day
Fig. 2. A histogram showing the percentage of Forssrnan-antigen-positive and
alkaline-phosphatase-positive cells found in disaggregates of genital ridges from
the embryos resulting from both wild-types 129 matings and from SI/+ xS7/+
matings. The numbers in brackets show the number of litters of each class
examined. ( + + x + + and SI/+SI/+ respectively) Where male and female
embryos were separated the results are shown separately.
Key-open bars represent % F.A. +ve cells from + / + x + / + litters.
- hatched bars represent % F.A. +ve cells from 57/+ x 57/+ litters.
- horizontal lines represent % alkaline phosphatase + ve cells.
For each individual sample > 200 cells of each class were counted. The percentage
values have been averaged between litters.
4. Expression of FA in genital tissues
Genital ridges contain increasing proportions of FA-positive cells from
day 11 to about day 14 (Fig. 2). It is known that germ cells migrate from the
allantois to hind-gut endoderm, the dorsal mesentery, the coelomic angles and
finally start to enter the genital ridges at day 10 (Everett, 1943; Chiquoine,
1954; Mintz & Russell, 1955 & 1957; Bennett, 1956; Ozdzenski, 1967). These
cells are easily identified histochemically because they have high levels of
alkaline phosphatase and a distinctive morphology. Figure 2 shows that there
is reasonable correspondence in the proportions of cells which are FA positive
and the proportion with high levels of alkaline phosphatase until day 13.
Figure 3 (a, b) shows FA-positive cells from a 12-day genital ridge which have
typical germ-cell morphology; the cells are large with clear pseudopodia. A
double-labelling experiment with M1/22.25-FL and histochemical stain for
alkaline phosphatase (Table 3) shows that at day 12 the proportion of doublestaining cells can reach 83%. After this time it is known that alkaline phosphatase is a less specific marker for germ cells because other cells in the genital
ridge show slight activity of this enzyme. The proportion of FA-positive cells
in the genital ridges increases until day 14 (Figure 2). Before day 14 there is
a greater proportion of FA-positive cells in the female gonad than in the
124
M. G. STINNAKRE AND OTHERS
\
0 001cm
I
I
Fig. 3. Isolated primordial germ cells from 12-day embryos stained with Ml /22.25-F1
(a) phase contrast (b) fluorescence.
Table 3. Double labelling of genital-ridge cell suspensions for alkaline
phosphatase (AP) and Forssman antigen {FA)
Alkaline-phosphatase-positive
cells % cellsf
Day
11
13
Sex
<? + 9 ( D *
c? + 9(3)
6 (2)
? (4)
6 (2)
FA+
26
83
83
77
64
FA-
74
17
17
23
36
'
9 (1)
40
60
* Numbers in brackets are numbers of litters examined.
t At least 100 cells examined.
male gonad; the relative percentages are reversed after this time (see Figure 2).
Up to day 15, the majority of the FA-positive cells have distinctive germ-cell
morphology (Figure 4). After this time FA is expressed by cells other than
germ cells. These FA-positive non-germ cells are small granular cells and in
the testes could be either Sertoli cells, Leydig or other interstitial cells. Doublelabelling experiments with M1/22.25-F1 and the histochemical stain for A5-3/?
hydroxysteroid dehydrogenase activity (specific for Leydig cells; Niemi &
Ikonen, 1961) on day-16 testicular cells shows that Leydig cells are not the
antigen-positive non-germ cell component. This enzyme activity is already
125
Expression of Forssmart antigen in mouse embryo
100
80
60
40
20
11
12
13
14
15
.16
Day
Fig. 4. Histogram of the percentage of Forssman-antigen-positive cells showing
typical germ-cell morphology when isolated from the gonad of progressively older
embryos.
present in mouse Leydig cells by 13 days (Scheib & Lombard, 1971). After
day 16 some of the antigen-positive cells are found in close contact with germ
cells and there are increasing numbers of antigen-negative germ cells (Figure 4).
Leydig cells, identified histochemically, from new born, post-partum, day-8,
day-23 and adult mice were FA-negative. In two experiments, new-born testes
contained 41 % antigen-positive cells; 3 % of the cells were Leydig cells and
these were FA-negative. All stages of gonocyte differentiation in the postpartum mouse testes are negative; spermatozoa are not labelled with M 1/22.25
antibodies conjugated to fluorescein. Thus it appears that FA is expressed
by the germ cells at least until day 16 of embryonic development but not in
the germ line of the new-born or of the adult male mouse. Leydig cells are
negative and some other cell type(s) is antigen-positive.
5. SI and W loci
The SI and W are complex loci; they are a series of semi-dominant alleles,
some of which, when homozygous, lead to an almost complete absence of
germ cells (reviewed by Heath, 1978). Figure 2 shows that there are consistently
reduced proportions of FA-positive cells in genital ridge preparations from
SI/+ xSl/+ litters compared to + / + X + / + at days 12 to 16. Analysis of
the individual embryos from SI/+ x SI/+ crosses shows three classes of
embryo with respect to the proportions of FA-positive cells in the genital
5
EMB 6l
126
M. G. STINNAKRE AND OTHERS
Table 4. Effect of SlJ locus on % FA cells in genital ridge
Day
12
13
13
13
% FA-positive cells in individual embryos
of litters from 129 SI3/ + x SI V+crosses*
0;0; 6-6; 7-5; 7-9; 9 1 ; 9-6; 17-4
0; 4; 9-4; 10-7;
5-4; 5-7;
2-6; 5-2; 10-6; 119; 30-4
* > 200 cells examined.
1900 Unabsorbcd
51
Cr release
1700 •o 1500 1300 -
y IIOO 900 -
700
500
1:4
1:16
1:64
1:128
1:512
Ratio of volumes of antigen: antibody
Fig. 5. Tissue distribution of Forssman antigen in tissues of an adult Wv/W mouse.
The antigen is present in high levels in brain, kidney and testis as it is in wild-type
animals. Increasing amounts of tissue homogenate were used to absorb an aliquot
of Ml/22.25 whose ability to cause complement mediated lysis of 51Cr-labelled
sheep red blood cells was then tested.
ridge: those with virtually no FA-positive cells (SI/SI), those with normal (+ / +)
and an intermediate class (57/ + ). (Table 4) These SI/SI embryos do not
survive longer than about 14 days in utero.
There are viable alleles, e.g. Wy, which will survive to term but are sterile.
The expression of Forssman antigen in Wx/ Wy homozygote male mice was
examined by absorbtion of M 1/22.25 cytotoxicity for sheep erythrocytes
(which express FA (Stern et al. 1978)). A normal tissue distribution of Forssman
antigen was found (Figure 5). The WvfWv testes which have virtually no germ
cells (Coulombre & Russell, 1954) had only a slightly reduced amount of FA
when compared with other tissues.
Expression of Forssman antigen in mouse embryo
127
0
Extraembryonic ectoderm
Trophectoderm
©y
Morula
©
©
©
Embryonic ectoderm
1 Day
'
Mesoderm
Ectoderm
©
Parietal
©
Endoderm-
©
. Visceral.
•©
Fig. 6. Summary of ontogenetic distribution of cell surface Forssman antigen
during early development of the mouse.
Neither the Sl] nor Wv alleles affect the expression of FA per se since 26/27
isolated ICM's from Slj/+ xSP/+ crosses and 40/40 ICM's from IVV/+ x
Wv/ + crosses were found to be FA-positive.
DISCUSSION
Serological reagents used in the study of embryonic development may
identify antigens which will give further insight into the heterogeneity of the
cells in the developing embryo and in consequence be useful as cell phenotype
markers. Their use to define cell surface markers has, moreover, led to the
supposition that these themselves may provide markers of cell lineage within
the developing embryo. In the case of FA the distribution in the preimplantation
embryo might have suggested that this antigen would be expressed on derivatives
of the ICM. The surprising result was that its expression was not detected
on embryonic ectoderm but on the embryonic endoderm.
The distribution of expression of FA on the cells of the preimplantation
and early post-implantation embryo (to 7 | days) seems, however, to follow
a recognizable pattern. This is depicted schematically in Figure 6. At 2\ days
the morula has no detectable antigen-positive cells. Both cell types derived
from this stage are expressing FA by 3^ days. The trophectoderm loses the
FA just before implantation but it can still be detected on the 4^-day ICM.
Extraembryonic ectoderm derived from the trophectoderm overlying the inner
cell mass continues to have no detectable FA throughout the period studied.
The ICM gives rise to two cell lineages, the embryonic ectoderm and the
endoderm. Up to and including 5 days both tissues are antigen-positive but
after this time only one of the ICM derivatives, the endoderm, retains antigen
5-2
128
M. G. STINNAKRE AND OTHERS
expression. Within the period of embryonic development to day 7f, FA may
be a useful marker of various different lineages. It is clear that FA differs
from the embryonic antigen(s) defined by mouse anti-F9 serum (Buc-Caron
et ah 1978) by being found on the surface only of the endodermal cells in the
6- to 7^-day post-implantation embryo. What is not clear is whether mouse
anti-F9 and other similar antisera might contain some anti-FA-like activity
in addition to other specificities.
The difficulty with a lineage-related view of FA continues in its expression
in later embryos and adults. At day 11 expression of this antigen is first detected
in neural tissues and in the genital ridges. In the latter it is clear that the
antigen-positive cells are germ cells as identified by morphology and histochemically for alkaline phosphatase. The round shape and characteristic
staining of the primordial germ cells - particularly their alkaline phosphatase
histochemistry have previously allowed their lineage to be traced by Ozdzenski
(1967) to a region at the caudal end of the primitive streak in 7^-day embryos,
where the embryonic ectodermal cells form part of the allantoic rudiment.
It is important to remember, however, that the immunological assays used
here do not detect small subpopulations of cells and it cannot be certain that
the embryonic ectoderm after day 6 has absolutely no FA-positive cells.
The view that the antigen-positive cells of the gonadal ridge include a major
contribution from the germ cells is supported by the consistent reduction in
the proportions of FA-positive cells in genital ridges from 129 57/+ x 129 SI/ +
crosses compared with wild type. Also crosses which produce 57/57 homozygotes
have individual embryos which contain few or no germ cells. The trimodal
distribution of numbers of FA-positive cells found in the germinal ridges of
embryos from such crosses is also seen by alkaline phosphatase staining
(McCoshen & McCallion, 1975, and our own observations). The detection of
antigen-positive germ cells in the genital ridge and the increase in the percentage
of positive cells in this tissue until day 14 is entirely consistent with the known
migratory and multiplicative phases of mouse germ cells. The changes in the
relative proportions of positive cells in female versus male genital ridges after
day 14 probably reflect a number of events in the differentiation of testis and
ovary respectively. In the testis at day 15 a separation of the gonocytes and
the supporting cells takes place; this might increase the relative yield of gonocytes from the trypsinized genital ridge. At day 15 in the ovary, most of the
germ cells have already entered meiotic prophase.
It is likely that until day 14 the majority of antigen-positive cells are germ
cells but thereafter the picture becomes more complicated. There are definitely
non-germ-cell FA-positive cells present by day 16 and the primary candidates
for this cell type in the testes would be Leydig or Sertoli cells. The Leydig
cells can be identified histochemically and double-labelling experiments showed
this cell type is not the cell which expresses the FA in the embryonic or adult
testes. Although in the previous study (Stern et ah 1978) spermatozoa from
Expression of Forssman antigen in mouse embryo
129
adults were found to be FA-positive by absorption, no evidence of FA expression
on the germ line after birth could be found. The previous positive results for
sperm can easily be explained by the presence of small numbers of contaminating cells. It seems likely that by birth the major cellular component contributing to the absorptive capacity of the testes is non-germ line. This is
supported by the near normal absorptive capacity of WvfWv homozygote
testes for Ml/22.25. A small change with respect to other tissues could be
expected since the testes of Wy/Wy animals are much smaller than normal
and relative amounts of contributing tissues within the testes are different.
The number of positive cells in the adult mouse testes and the morphology
of the antigen-positive cells in culture (unpublished) makes it likely that this
cell type is a Sertoli cell. In all these experiments with genital tissues it is
difficult to detect small subpopulations of a given cell type which are FA-positive.
Thus although some germ cells are FA-positive in the embryo until at least
day 16, by birth this lineage no longer expresses the antigen; another cellular
component of the testes, probably Sertoli cells, accounts for the positivity of
this tissue.
Considering these results together with those for the early embryo, the
conclusion must be that either it is not possible to interpret FA as a lineagerelated marker, or that our preconceived ideas about the lineages are incorrect.
If FA were to be a lineage-related cell marker it might suggest a derivation
of germ cells either via endoderm or from a subpopulation of ICM cells before
they become committed to ectoderm formation at day 6. It might also suggest
that some accessory gonadal cells could be derived from the primordial germ
cells. This latter possible interpretation is, however, probably ruled out by the
development of FA-positive cells in Wv/Wv testes. Nevertheless, at a given
stage of development FA expression is clearly delineating different subpopulations. Their relationship is unclear but M 1/22.25 antibodies have given
a different view of the cells in the developing embryo.
Although cell antigens shared by different types of cell may not imply a
lineage relationship they might well imply a functional homology of shared,
possibly biologically important, cell-surface molecules. Since there is no apparent
lineage relationship between the different cell types which express this antigen
in the embryo or in the adult it may be useful to look for a functionally common
denominator. The problem is where to start looking. It is known that the
antigen recognized (at least on EC cells and sheep red blood cells) is a glycolipid, and there is some evidence that glycolipids may enhance hormone binding
(Fishman & Brady, 1976), and cell adhesion (Huang, 1978). Both of these
areas might merit further investigation.
The important conclusions from this study are that FA does not give a
readily predictable pattern of expression in the embryo with respect to cell
lineage. Thus using monoclonal antibodies has not really simplified the
situation. However, if such a complex picture emerges from using antibodies
130
M. G. STINNAKRE AND OTHERS
against a single defined antigenic determinant then interpretation of the results
with conventional antisera must be treated with a great deal more caution.
The financial support of the Cancer Research Campaign, the Wellcome Foundation and
the Medical Research Council are very gratefully acknowledged.
REFERENCES
ARTZT, K., DUBOIS, P., BENNETT, D., CONDAMINE, H., BABINET, C. & JACOB, F. (1973).
Surface antigens common to mouse cleavage embryos and primitive teratocarcinoma
cells in culture. Proc. natn. Acad. Sci., U.S.A. 70, 2988-2992.
BENNETT, D. (1956). Development analysis of a mutation with pleiotropic effects in the
mouse. / . Morph. 98, 199-234.
BUC-CARON, M. H., CONDAMINE, H. & JACOB, F. (1978). The presence of F9 antigen on
the surface of mouse embryonic cells until day 8 of embryogenesis. J. Embryol. exp.
Morph. 47, 149-160.
CHIQUOINE, A. D. (1954). The identification, origin and migration of the primordial germ
cells of the mouse embryo. Anat. Record. 118, 135-146.
COULOMBRE J. L. & RUSSELL, E. S. (1954). Analysis of the pleiotropism at the W locus in
the mouse. / . exp. Zool. 126, 277-295.
ERICKSON, R. P. (1977). Differentiation and other alloantigens of spermatozoa. In Immunobiology of Gametes (ed. M. Ediden & M. H. Johnson), pp. 85-114. Cambridge University
Press.
EVERETT, N. B. (1943). Observational and experimental evidences relating to the origin
and differentiation of the definitive germ cells in mice. / . exp. Zool. 92, 49-92.
EVANS, M. J., LOVELL-BADGE, R. H., STERN, P. L. & STINNAKRE, M. G. 1979. Cell lineages
of the mouse embryo and embryonal carcinoma cells: Forssman antigen distribution
and patterns of protein synthesis. In Cell Lineage, Stem Cells and Cell Differentiation.
INSERM Symposium 10 (ed. N. Le Douarin). N. Holland.
FARIAS, E., KAJII, T. & GARDNER, L. I. (1967). Technique for investigation of sex chromatin
in amniotic membrane of rat foetus. Nature 214, 499-500.
FISHMAN, P. H. & BRADY, R. O. (1976). Biosynthesis and function of gangliosides.
Science, 194, 906-915.
GARDNER, R. L. & PAPAIOANNOU, V. E. (1975). Differentiation in the trophectoderm and
inner cell mass. In The Early Development of Mammals (ed. M. Balls & A. Wild), pp. 107—
132. C.U.P.
HEATH, J. K. (1978). Characterisation of a Xenogeneic antiserum raised against the fetal
germ cells of the mouse: cross reactivity with embryonal carcinoma cells. Cell 15, 199—
306.
HUANG, R. T. C. (1978). Cell adhesion mediated by glycolipids. Nature 276, 625-626.
KOHLER, G. & MILSTEIN, C. (1976). Derivation of specific antibody-producing tissue culture
and tumour lines by cell fusion. Eur. J. Immunol. 6, 511-519.
MCCOSHEN, J. A. & MCCALLION, D. J. (1975). A study of primordial germ cells during
their migratory phase in Steel mutant mice. Experientia 31, 589-590.
MINTZ, B. & RUSSELL, E. S. (1955). Developmental modifications of primordial germ cells
induced by the W series genes in the mouse embryo. Anat. Rec. 127, 443.
MINTZ, B. & RUSSELL, E. S. (1957). Gene induced embryological modifications of primordial
germ cells in the mouse. J. exp. Zool. 134, 207-229.
NIEMI, M. &IKONEN, M. (1961). Steroid-3/?-ol-dehydrogenase activity in foetal Leydig's cells.
Nature, Lond. 189, 592-593.
OZDZENSKI, W. (1967). Observations on the origin of primordial germ cells in the mouse.
Zoologica poloniae 17, 367-379.
SCHEIB, D. & LOMBARD, M. N. (1971). Etude histoenzymologique de l'activite de la D5-3/?
hydroxysteroide deshydrogenase du testicule explante en presence d'hydrophyse. Archs
d'Anat. microsc. Morph. exp. 60, 205-218.
Expression of Forssman antigen in mouse embryo
131
D. & KNOWLES, B. (1978). Monoclonal antibody defining a stage-specific mouse
embryonic antigen (SSEA-1). Proc. natn. Acad. Sci., U.S.A. 75, 5565-5569.
STERN, P. L., MARTIN, G. R. & EVANS, M. J. (1975). Cell surface antigens of clonal teratocarcinoma cells at various stages of differentiation. Cell 6, 455-465.
SOLTER,
STERN, P. L., WILLISON, K. R., LENNOX, E., GALFRE, G., MILSTEIN, C , SECHER, D., ZIEGLER,
A. & Springer, T. (1978). Monoclonal antibodies as probes for differentiation and tumour
associated antigens: a Forssman specificity on teratocarcinoma stem cells. Cell 14, 775784.
STEVENS, L. C. (1967). The biology of teratomas. Adv. Morph. 6, 1-31.
WATTENBERG, L. W. (1958). Microscopic histochemical demonstration of steroid 3 /? ol dehydrogenase in tissue sections. /. Histochem. Cytochem. 6, 225-232.
WILLISON, K. R. & STERN, P. L. (1978). Expression of a Forssman antigenic specificity in
the preimplantation mouse embryo. Cell 14, 785-794.
{Received 4 January 1980, revised 24 July 1980)