J. Embryol. exp. Morph. 83, 15-31 (1984)
Printed in Great Britain © The Company of Biologists Limited 1984
Initial phases of the rat testis differentiation in vitro
By ROXANE AGELOPOULOU, SOLANGE MAGRE,
EVANGELIE PATSAVOUDI AND ALFRED JOST
Laboratoire de Physiologie du Developpement, College de France, 11, place
Marcelin Berthelot, 75231 Paris Cedex 05, France
SUMMARY
Rat gonadal primordia with their supporting mesonephroi were explanted in vitro at the
undifferentiated stage (12 days 16h after fertilization), at the outset of testicular differentiation (13 days 9 h) or when already containing seminiferous cords. The younger foetuses were
sexed with the sex chromatin test in the amniotic membrane. The basal medium was CMRL
1066 and the culture period, 1 to 4 days. Testicular differentiation resulted from the appearance of large clear cells, the primordial Sertoli cells, and from their aggregation into
seminiferous cords. Addition of 15 % foetal calf serum to the medium prevented the differentiation of seminiferous cords, but large clear cells appeared. In testes from 14- or 15-day-old
foetuses, the seminiferous cords disintegrated under the influence of serum. The serum did not
prevent the differentiation of Sertoli cells, but impaired organogenesis or maintenance of the
early seminiferous cords. The results support previous histological observations on the initial
stages of testicular differentiation.
INTRODUCTION
Several years ago we undertook a study of the very first histological or cytological signs of testicular differentiation among the cells of the undifferentiated
primordium of the rat foetus (Jost, 1972; Jost, Vigier, Prepin & Perchellet, 1973;
Jost, Magre & Cressent, 1974). This study resulted in two findings: 1) testicular
differentiation begins at earlier stages than was previously (Thomson, 1942;
Torrey, 1945) or subsequently (Merchant-Larios, 1976) reported by others; 2)
the first morphologically recognizable event is the differentiation of a few
primordial Sertoli cells, 13 days 9 h after fertilization. The number of these cells
increases very fast during the few following hours. These cells progressively
aggregate into forming seminiferous cords and include germ cells. In electron
microscopy they are seen to establish complex membrane contacts with deep
infoldings and interdigitations (Magre & Jost, 1980, 1983).
In order to investigate these processes experimentally, we performed in vitro
cultures. Differentiation of testes from undifferentiated mouse or rat primordia
during in vitro culture has been reported in the past (Wolff, 1952; Asayama &
Furusawa, 1960, 1961; Taketo & Koide, 1981; Byskov & Saxen, 1976; Byskov
& Grinsted, 1981; Stein & Anderson, 1981). However, the authors did not sex
the foetuses before cultivating their primordia and they did not focus their
16
R. AGELOPOULOU AND OTHERS
interest on the initial stages and the histological mechanisms of gonadal sex
differentiation. In our experiments the foetuses were sexed with the sex
chromatin test in the amniotic membrane, enabling us to study the initial phases
of testicular differentiation.
The initial formation of testicular cords could be followed in primordia from
male rat foetuses cultured a few days in a synthetic medium. Testicular organogenesis, however, did not take place when foetal calf serum was present in the
medium. Instead, large clear cells resembling primordial Sertoli cells differentiated, but remained scattered through the developing gonad. We hypothesized
that these cells actually were primordial Sertoli cells which did not aggregate as
they normally do (Magre, Agelopoulou & Jost, 1980, 1981; Jost, Magre &
Agelopoulou, 1981). It was decided to study in further details the in vitro model
of testicular differentiation. This is the aim of the present paper.
Most of the experiments were done on primordia explanted at 12 days 16 h or
at 13 days 9 h, i.e. either clearly before or just at the outset of the differentiation
of the first Sertoli cells preceding the formation of the seminiferous cords (Jost,
1972). Complementary experiments made at later stages up to 16 days 9h, will
also be described and compared with the preceding ones.
MATERIALS AND METHODS
Animals
The study was made with Wistar CF rats (stock maintained by the CNRS,
Colony of R. Janvier, 53680 Le Genest, France).
Foetal age
Five or six females were caged for one night with a male. The pregnant animals
were detected by palpation 12 or 14 days later. It was assumed that ovulation and
fertilization took place at 1 a.m. This hour was taken as time zero of pregnancy
(Jost & Picon, 1970). The age of the foetuses was calculated in days and hours
after time zero. For the sake of sparing space in the text but not in the headings
and legends to figures, we shall write 12-day-old foetuses for those preserved at the
stage of 12 days 16 h (at 17 h in the afternoon) and 13- to 16-day-old foetuses for
those preserved at 10 o'clock in the morning (for instance stage of 13 days 9h).
Sexing the young
The foetal masses with intact foetal membranes and placenta were aseptically
removed from the uterus of anaesthetized females (3mg/100g body weight
sodium pentobarbital). The sex of the young was recognizable at dissection in
foetuses over 14 days old; for younger stages, the sex chromatin test was used:
a piece of amniotic membrane was carefully spread over a slide and treated with
acetic orcein (1 g synthetic orcein Gurr in 45 ml glacial acetic acid, diluted by one
Early testicular organogenesis
in vitro
17
half with distilled water before use) (Farias, Kajii & Gardner, 1967; Jost, 1972).
The foetuses were sexed extemporaneously by one person, before or while
another person secured the explants to be cultivated in vitro. Another piece of
amniotic membrane was fixed for 30min with Carnoy's fluid, stained with
toluidine blue, and kept for control.
Explants
The foetuses were aseptically dissected under a dissecting microscope. On 12day and 13-day-old foetuses, the intestine and mesentery were discarded and the
two mesonephroi with the gonadal primordia on them were removed, together
with part of the dorsal wall, generally comprising the aorta and the cardinal veins
(Fig. 1A). In such a preparation, the gonadal primordia remain untouched and
the block of tissue permits further differentiation.
In older foetuses, each mesonephros (which is rudimentary in rats) with its
attached gonad was cultured separately. Sometimes the gonads were dissected
free of other structures. Before being introduced into the culture dishes, the
explants were rinsed in RPMI - Hepes Buffer 0-02M (Moore, Gerner & Franklin, 1967) (purchased at Eurobio, Paris).
In vitro cultures
The explants were placed in Falcon dishes (Falcon 3037) containing
approximately 0-7 ml medium, on grids (Falcon 3014) previously soaked in a 1 %
agar solution. The medium was level with the grid. The temperature was
3 6 ° C ± 1 and the gas mixture was 19 %O 2 + 76 %N 2 + 5 %CO 2 . The culture
period was from 1 to 4 days.
In a few instances the cultures were immersed in Falcon dishes without the grid
and the oxygen pressure was increased. The results were essentially the same as
those reported here and will not be presented.
The medium used was CMRL 1066 (Parker, 1961) (purchased at Eurobio,
Paris) completed extemporaneously with glutamine (3 jul of a 200 HIM commercial solution per ml), 250i.u. penicillin (Specilline, Specia) and 100/ig streptomycin (Diamant) per ml. In certain experiments foetal calf serum (Eurobio)
was added to the basic medium. The concentration was expressed as percentage
of the final volume (0-1 to 15 % ) .
The medium was renewed after 2 days, for cultures grown for 3 or 4 days.
The protocol of the experiments including the number and characteristics of
the cultures is given in Tables 1 and 2.
Histology
At the end of the culture period, explants were fixed for 24 h in Bouin's fluid,
paraffin-embedded and sectioned at 5/mi thickness. Every fifth section in the
series was mounted on a slide and stained with haemalum or haematoxylin plus
eosin.
18
R. AGELOPOULOU AND OTHERS
9
Fig. 1. Sections through explants from male foetuses before culture in vitro. (A)
Low-power view of an explant from a 12 days 16 h old foetus, showing the two
gonads (g) over the mesonephroi (m) and the tissues around the aorta (a). (B)
Higher magnification of the gonad on the right side of the same explant. (C) Similar
section of a gonad from 13 days 9h old foetus; a few Sertoli cells can be seen
(arrows), d: dorsal mesentery; c: cardinal vein; w: Wolffian duct. A x220, B and
C X500.
Early testicular organogenesis in vitro
19
Table 1. Number of cultures studied, classified according to age at explantation,
medium, sex and duration of the culture period
Age at
explantation
Duration of culture
Medium
1 day
2 days
3 days
4 days
cf + 9
12dayl6h
CMRL 1066
serum added
13day9h
CMRL 1066
serum added
CMRL 1066
serum added
CMRL 1066
serum added
14 day 9 h
15 day 9 h
16 day 9 h
6+6
7+7
3+3
3+3
14 + 9
8+6
9+7
10 + 7
16 + 11
7+5
15 + 10
15 + 13
8+6
8+6
13 + 10
13 + 15
4+2
10 + 3
11 + 3
14 + 2
6+6
6+ 6
2+2
4+3
CMRL 1066
serum added
17 + 12
14+12
21 +16
14 + 6
15 + 15
1.7 +12
10 + 3
14 + 4
18 + 9
23 + 9
The medium was CMRL 1066; the concentration of foetal calf serum was 15 %.
Each culture contained both gonads in the explants from 12- and 13-day-old foetuses and
only one gonad in the explants from older foetuses.
Necrotic explants are not included in the table: 17 explants from 12-day-old foetuses and 9from 13-day-old foetuses.
RESULTS
I. Testicular differentiation in vitro
It is evident that great attention should be paid to chronology and to developmental stages in studies concerning initial events in differentiation processes.
This explains some of the following precisions.
a) 12 days 16 h old undifferentiated explants
The development of these young gonads (Fig. IB) was slower in vitro than in
vivo. After 24 h in vitro, the explants still provisionally maintained their original
shape; many mitoses were obvious, but no sign of testicular differentiation could
be recognized with certainty. After two days in vitro, the explants tended to
flatten. A few large clear cells, the first Sertoli cells, were present, usually deep
in the gonad near the mesonephros (Fig. 2A), in a location similar to that
described at 13 days 9 h in vivo (Jost, 1972). In some places these cells aggregated
and became polarized, their smoothened basal surface indicating the incipient
differentiation of the outer surface of the forming cord (Fig. 2A). This process
had progressed after 3 days and more so after 4 days in vitro.
After 4 days of culture, the size of the whole explant remained small, its free
surface was flattened and usually the gonad did not project over the mesonephros,
but it was included in the surrounding tissues; however, the volume of the gonad
20
R. AGELOPOULOU AND OTHERS
Early testicular organogenesis in vitro
21
Fig. 3. Explants from 12 days 16h old female foetuses, cultured for 4 days either
without (A) or with serum (B). The structure of the presumptive ovaries remains
undifferentiated. (x500).
itself had somewhat increased as seen in serial sections. The best developed
explants contained structures similar to seminiferous cords, consisting of clear
large cells, the Sertoli cells, which surrounded the germ cells; most of the Sertoli
cells were polarized, their nucleus being located at the periphery of the cord,
while the cytoplasm extended inward (Fig. 2C). The germ cells were not very
numerous. The cords were surrounded by an eosinophilic line that corresponded
to a basal membrane. The extracordal cells had a scanty cytoplasm and an
irregular rather dense nucleus; some of them were flattened around the cord and
looked like connective tissue cells. Almost half the explants were less well
Fig. 2. Explants from 12 days 16 h old male foetuses cultured either without serum
in the medium (A and C, left side) or with 15 % foetal calf serum (B and D, right
side). In the absence of serum, primordia from male foetuses organize structures
resembling seminiferous cords ((A) after 2 days; (C) after 4 days in vitro); the
external palisade of Sertoli cells is easily recognizable in (C) (arrows). In the
presence of serum, no cords are recognizable but many clear cells differentiate ((B)
after 2 days; (D) after 4 days). Some germ cells were labelled g (x500). Note that
the size and general condition of the explants improved in cultures with serum.
22
R. AGELOPOULOU AND OTHERS
developed, and their seminiferous cord-like structures were either scanty or
incompletely differentiated; in the latter case, part of the contour of the developing seminiferous cord could have differentiated, while on the opposite side of its
section it could merge into an undifferentiated area. Similar aspects were
previously observed in vivo at early stages of testis differentiation (Jost, 1972).
The germ cells were not counted in these gonads; the impression from the
histological study of the sections is that their number increased during the two
first days of culture and decreased thereafter.
In explants taken from 12-day-old female foetuses the gonadal primordium
remained composed of one apparent cell type only (in addition to the germ cells),
namely small cells with an irregular nucleus, even after 4 days in vitro (Fig. 3A).
b) 13 days 9 h old explants, at incipient testicular differentiation
At the time of explantation, gonads of 13-day-old foetuses are larger than
those of 12-day foetuses (Fig. 1C). Very careful study is necessary to recognize
the first few Sertoli cells which appear in the deep anterior part of the gonad, near
the mesonephric structures (Jost, 1972).
After 3 or 4 days in vitro, the gonads became larger than those explanted at
12 days, but were of course much smaller than gonads developed in vivo for the
same period. Distinct seminiferous cords had formed, with polarized Sertoli cells
at their periphery and centrally located germ cells (Fig. 4C). These cords were
limited by a well-defined basal membrane and surrounded by small dark cells
resembling connective tissue cells. In some explants, testicular development was
not as good as that just described but quite recognizable.
It is important to notice that one day after explantation no seminiferous cords
had yet differentiated. However, many large cells with a clear cytoplasm were
scattered throughout the gonad, either isolated or in clusters (Fig. 4A). Cords
appeared two days after explantation at first in the vicinity of the mesonephric
structures. They apparently resulted from the aggregation of the clear Sertoli
cells, in a way similar to that described in vivo.
Female gonads from 13-day-old foetuses behaved like those explanted one day
previously.
c) 14 days 9 h to 16 days 9h testicular explants
On day 14 the testis already contains well-differentiated seminiferous cords in
its anterior part, the posterior part remaining partly undifferentiated. After the
first day of in vitro culture, some cords were well organized while others were still
Fig. 4. Explants from male 13 days 9h old foetuses, cultured either without serum
in the medium ((A) and (C) left side) or with 15 % foetal calf serum ((B) and (D)
right side). After one day in vitro (A and B) large cells with abundant cytoplasm
appear in both media. After 4 days, well-organized seminiferous cords differentiate
in the absence of serum (C) but not in its presence (D). (x500).
Early testicular organogenesis in vitro
23
24
R. AGELOPOULOU AND OTHERS
Early testicular organogenesis in vitro
25
Fig. 6. Testes explanted at 16 days 9 h and cultured for 4 days. The two testes from
the same foetus were cultured in different media: without serum (A) and with serum
in the medium (B). Seminiferous cords persist in the testis cultured with serum.
(X500).
ill defined. After two days in vitro, the seminiferous cords were more numerous
and better organized, and even more so after 4 days. The structure of the cords
was typical, with peripheral Sertoli cells whose nuclei were near the surface of
the cord, while the cytoplasm extended inwards; the germ cells were located in
the centre (Fig. 5A). The cords were circled with a basal membrane, clearly
revealed by Mallory trichrome staining, and were surrounded by small flattened
fibroblasts. Testes explanted at 15 or 16 days of age (Figs 5C and 6A) maintained,
after two or four days in vitro, a testicular structure that did not deviate very
much from normal. But of course they did not notably grow.
In conclusion, the testicular organization of 14- to 16-day-old testes was
preserved after 4 days of culture in the synthetic medium.
Fig. 5. Testes explanted at 14 days 9 h (A,B), at 15 days 9h (C,D) and cultured for
4 days. The two testes from the same foetus were cultured in different media. Left
side offigureshows testes cultured without serum (A,C); right side shows the contralateral testes from the same foetuses (B,D) cultured in medium containing serum.
Note the disaggregation in the presence of serum of the seminiferous cords. (x500).
26
R. AGELOPOULOU AND OTHERS
II. Effect of serum on testicular differentiation
The effect of adding 15 % foetal calf serum to the 1066 medium on either the
initial differentiation of seminiferous cords or on already formed cords was
studied systematically (Table 1).
a) Effect on initial stages of testicular differentiation
The presence of serum in the medium seemed to improve the general appearance of the explants from 12- or 13-day-old foetuses as regards growth and the
number of mitoses. However, even after four days of culture, no seminiferous
cords or similar structures had differentiated (Figs 2D and 4D). A new cell type
had appeared, characterized by its large size, abundant and clear cytoplasm, and
large feebly stained nuclei. These cells were either isolated or grouped into small
clusters, among dense undifferentiated cells. This cell type was completely
absent in female gonads cultured in the same way, in the absence or presence of
serum (Fig. 3).
Special attention was paid to short-term cultures, in view of their importance
for the comprehension of the initial stages of gonadal differentiation. Gonads
explanted at 12 days 16 h and studied 24 h later, were very similar to those
obtained in the absence of serum. After 48 h (Fig. 2B) and 72 h, many of the new
cells were seen, but they were not polarized and they did not aggregate into
seminiferous cords; cultures with or without serum had become clearly recognizable.
In male gonads explanted at 13 days 9 h and cultured for 24 h, many of the large
clear cells were present and the gonads were similar to those obtained in the
absence of serum (Fig. 4A and B); however, this no longer applied to gonads
cultured for 48 or 72 h in which seminiferous cords failed to differentiate (Fig.
4D).
Comparison of the gonads from male and female foetuses (Figs 2, 3 and 4)
demonstrated that a sex difference had become obvious, even if the serum
prevented true testicular organogenesis.
b). Effect of serum on already differentiated testes
Testicular explants from 14-, 15- and 16-day-old foetuses were cultured for up
to 4 days in the presence of serum and compared with similar explants cultured
in serum-free medium. In some cases the two gonads from the same foetus were
cultured in different media (Figs 5 and 6). The serum exerted a disintegrating
activity on the seminiferous cords, that was all the more pronounced as the testis
was younger.
The 14-day testes showed progressive changes which became very evident
after 48 h in vitro when the seminiferous cords had, to a large extent, disappeared.
Clusters of Sertoli cells showing no polarization were separated by small dark
Early testicular organogenesis in vitro
27
Table 2. Effect of various concentrations of normal foetal calf serum in 1066
medium, on 13 days 9h male explants cultured for 4 days
Effect of serum
Concentration
of serum
5%
1%
0-5%
0-1%
Number
of assays
10
17
10
7
10
13
4
2
0
4
3
2
0
0
3
3
The number of cultures showing a strong effect (+), an intermediary effect (±) or no effect
(—) is given for each concentration.
cells. During the 3rd and 4th days, the dispersion of the former Sertoli cells and the
germ cells was accentuated. Only a few sections of seminiferous cords remained.
The changes were not as extensive in the testes explanted at 15 days. After two
days in vitro, several seminiferous cords were still recognizable, but after 3 or 4
days in vitro, the occurrence of structures similar to seminiferous cords was low.
Testes from 16-day-old foetuses explanted with or without serum did not look
very different. Even if after 4 days the seminiferous cords may have been
somewhat underdeveloped, it was evident that the 16-day-old testes were resistant to the disorganizing effect of serum.
c) Concentration of serum necessary
The preceding experiments were made with an arbitrarily fixed concentration
of 15 % foetal calf serum. We studied the effect of lower concentrations in order
to determine a threshold concentration (Table 2).
A concentration of 5 % serum produced a full serum effect (total absence of
seminiferous cords in all cultures). With 1 %, most of the cultures exhibited a
clear serum effect, some were dubious, but none was free of any effect. With the
lowest concentrations used (0-5 % and 0-1 %) some explants showed the serum
effect, others did not. These low concentrations can be considered threshold
concentrations. Similar results were obtained with human serum, except that the
threshold concentrations seemed somewhat higher (Chartrain, Magre, Maingourd &Jost, 1984).
DISCUSSION
Although the effect of serum on testicular differentiation, that we discovered
by chance, was used mainly as a tool to scrutinize the normal differentiation of
the testis, it is necessary to discuss some aspects of this effect, in addition to the
bearing of the present experiments for testicular differentiation.
28
R. AGELOPOULOU AND OTHERS
1 The effect of serum on in vitro organogenesis
The 'serum effect', described in the present paper, is so obvious that one may
wonder whether previous authors studying testicular differentiation in vitro encountered it. Therefore the scanty literature in the field was carefully scrutinized.
To the best of our knowledge, only three papers were devoted to the in vitro
study of testicular differentiation from undifferentiated mammalian gonadal
primordia. All three papers deal with mouse gonads. Wolff (1952) using his own
culture technique on a medium enriched with embryo extract but containing no
serum, obtained sterile ovaries or testes from primordia taken from 11-5-day-old
mouse foetuses. Asayama & Furusawa (1961) cultured 12- or 13-day-old primordia in either glucose- or horse-serum-enriched Tyrode. In the former medium,
testicular differentiation proceeded for 48 h. In the latter medium on the contrary, horse serum 'acted as a retarding and modifying factor'. It is difficult to
judge whether this effect was the same as the 'serum effect' reported in the
present paper, but this seems not unlikely. According to Taketo & Koide (1981),
who examined explants taken from 11-day-old mouse embryos and cultured in
a medium containing heat-inactivated horse serum, Sertoli cells and true
seminiferous cords only formed after 5 days in vitro i.e. very belatedly. After
three days, these authors observed incompletely differentiated structures
reminiscent of the transient aspects described in the developing rat testis (Jost,
1972) or of those reported here in older testes which disintegrated in vitro under
the influence of serum.
The work of Stein & Anderson (1981) on rats can hardly be used in the present
discussion, because they pooled the results for cultures of non-sexed undifferentiated gonads and differentiating testes, thus constituting a group of 34 gonads
comprising 14 cases that could not be identified as ovary or testis after 6 or 7 days
of culture. The testes they obtained were not illustrated. A few other papers
devoted to the reciprocal influence of testes and ovaries in vitro mention testicular differentiation from undifferentiated primordia, but unfortunately give
no precise descriptions. In Byskov & Saxen's (1976) cultures of 11-day mouse
gonads in a medium containing 10 % foetal calf serum, the influence of serum on
testicular differentiation, if any, could hardly be recognized, since the male
gonads were only identified by the presence of testicular cords in the explants at
the end of the experiment. Grinsted, Byskov & Andreasen (1979) observed the
absence of seminiferous cords in cultures assumed to be testes, grown in used
media that had contained adult testes or epididymis; the structure of the control
presumed testes was not reported in detail and the foetuses were not sexed.
Subsequently, Byskov & Grinsted (1981) reported that early mouse gonadal
primordia separated from their mesonephros could differentiate testes in vitro
(in medium containing 10 % serum), while those cultured with their mesonephros
were feminized. Absence of histologically recognizable seminiferous cords and
the occurrence of meiotic prophase aspects in the germ cells were considered
Early testicular organogenesis in vitro
29
signs of feminization. O & Baker (1976) cultivated hamster gonadal primordia
from 12-day-old foetuses that were sexed (according to sex chromatin or
karyotype), in a medium containing 20% foetal calf serum. The testes were
disorganized when co-cultured with ovaries; four male gonads cultured in the
absence of ovaries contained seminiferous cords (not illustrated).
In conclusion, except for the work of Asayama & Furusawa (1961) who may
have observed what we called the 'serum effect' but lacked knowledge of the
normal differentiation of the foetal testis and did not sex the young, the results
of the other authors cannot easily be compared with our own. Moreover, it is not
certain that serum affects testicular differentiation in other animal species in the
same way as in the rat. More research is still needed.
Another question deserves consideration: why does foetal blood not prevent
testicular differentiation in the foetus itself, when the testicular cords form? One
may imagine that the serum component involved is not yet present at an appreciable concentration in foetal plasma when the testes differentiate, or that it is
confined to the vascular system and does not reach the cells and the structures
responsible for the 'serum effect' observed in vitro. This question awaits further
investigation.
It is too early to discuss the mode of action of the serum in our system. Many
effects of foetal calf serum or other sera on cells cultured in vitro have been
reported. Curtis & Greaves (1965) isolated a protein from horse serum that
prevented in vitro aggregation of chick embryonic cells. Several reports deal with
in vitro cultures of Sertoli cells from young rats and concern impairment by serum
of either hormonal secretion (Rommerts, Kriiger-Sewnarian, Grootegoed, de
Jong & Van der Molen, 1979), or cellular processes and contacts (Tung,
Dorrington & Fritz, 1975; Hutson, Garner & Stocco, 1980). In our system,
organogenesis of the seminiferous cords is prevented, perhaps not only because
cell contacts are disturbed, but possibly also because some characteristics of the
extracellular matrix are affected. Efforts are underway in our laboratory to
isolate the active serum fraction.
2) The initial phases of testicular organogenesis
The significance of the present experiments for the problem of testicular
differentiation results from the agreement between the facts observed earlier in
vivo and now in vitro. The first recognizable event in testicular differentiation in
vivo is the differentiation of a new cell type, the primordial Sertoli cells: only a
few of these cells were seen at the stage of 13 days 9h; their number increased
and they aggregated during the next hours to form seminiferous cords (Jost,
1972; Magre & Jost, 1980, 1983). In vitro, in serum-free medium, the initial
testicular differentiation also began with the appearance of large clear cells,
readily visible after 48 h in the 12-day-old explants, and after 24 h in the 13-dayold explants. These cells aggregated to form seminiferous cords the next day. A
longer period elapsed in vitro between the appearance of the new type of cells
EMB83
30
R. AGELOPOULOU AND OTHERS
and their aggregation than in vivo. Because, the new cell type also appeared in
the explants cultured in the presence of serum, this initial stage did not seem to
be disturbed by serum. In the medium containing serum, however, the clear cells
remained scattered or distributed into groups of various sizes, and did not
aggregate and become polarized in forming seminiferous cords. The assumption
that the isolated clear cells which thus appeared in explants from 12- or 13-dayold male foetuses are Sertoli cells is supported by the present observations on 14or 15-day-old testicular explants cultured in vitro with serum; their seminiferous
cords disintegrated and released cells similar to those appearing in the younger
explants. Additional evidence was obtained for the Sertoli nature of the clear
cells that appeared in male gonads cultured in the presence of serum; it was
shown that these gonads exerted a Mullerian inhibiting activity when placed in
contact with foetal Mullerian ducts. It is known that the Mullerian inhibiting
factor or Anti-Miillerian Hormone is produced by the Sertoli cells (Blanchard &
Josso, 1974; Price, 1979). These results were alluded to earlier (Jost, Magre &
Agelopoulou, 1981; Magre, Jost & Valentino, 1982) and will be reported in
detail elsewhere. The in vitro model of differentiation of the testicular
seminiferous cords thus confirms the description which concluded that the
progressive differentiation and aggregation of Sertoli cells constitutes the initial
phase of normal testicular differentiation in the rat foetus.
The technical assistance of Odette Locquet and Marthe Solvar for histology and Robert
Urbe for animal care is gratefully acknowledged.
REFERENCES
S. & FURUSAWA, M. (1960). Culture in vitro of prospective gonads and gonad
primordia of mouse embryos. Zool. Mag. 69, 283.
ASAYAMA, S. & FURUSAWA, M. (1961). Sex differentiation of primordial gonads of the mouse
embryo after cultivation in vitro. Jap. J. exp. Morph. 15, 34-47.
BLANCHARD, M. G. & Josso, M. (1974). Source of the Anti-Mullerian Hormone synthetized
by the fetal testis: Mullerian inhibiting activity of fetal bovine Sertoli cells in tissue culture.
Pediat. Res. 8, 968-971.
BYSKOV, A. G. & GRINSTED, J. (1981). Feminizing effect of mesonephros on cultured differentiating mouse gonads and ducts. Science 212, 817-818.
BYSKOV, A. G. & SAXEN, L. (1976). Induction of meiosis in fetal mouse testis in vitro. Devi
Biol. 52, 193-200.
CHARTRAIN, I., MAGRE, S., MAINGOURD, M. & JOST, A. (1984). Effect of serum on organogenesis of the rat testis in vitro (in preparation).
CURTIS, A. S. G. & GREAVES, M. F. (1965). The inhibition of cell aggregation by a pure serum
protein. /. Embryol. exp. Morph. 13, 309-326.
FARIAS, E., KAJII, T. & GARDNER, L. I. (1967). Technique for investigation of sex chromatin
in amniotic membrane of rat fetuses. Nature 214, 499-500.
GRINSTED, J., BYSKOV, A. G. & ANDREASEN, M. P. (1979). Induction of meiosis in fetal mouse
testis in vitro by rete testis tissue from pubertal mice and bulls. /. Reprod. Fert. 56,653-656.
HUTSON, J. C , GARNER, C. W. & STOCCO, D. M. (1980). Effects of serum components on the
morphology of Sertoli cells in culture. Anat. Record 197, 205-211.
JOST, A. (1972). Donnees preliminaries sur les stades initiaux de la differenciation du testicule
chez le rat. Archs Anat. microsc. Morph. exp. 61, 415-438.
ASAYAMA,
Early testicular organogenesis in vitro
31
A., MAGRE, S. & AGELOPOULOU, R. (1981). Early stages of testicular differentiation in
the rat. Hum. Genet. 58, 59-63.
JOST, A., MAGRE, S. & CRESSENT, M. (1974). Sertoli cells and early testicular differentiation.
Male Fertility and Sterility, (eds R. E. Mancini & L. Martini), pp. 1-11. New York: Acad.
Press.
JOST, A. & PICON, L. (1970). Hormonal control of fetal development and metabolism.
Advances in Metabolic Disorders Volume 4, pp. 123-184. New York: Acad. Press Inc.
JOST,
JOST, A., VIGIER, B., PRE>IN, J. & PERCHELLET, J. P. (1973). Studies on sex differentiation in
Mammals. Rec. Prog. Horm. Res. 29, 1-41.
S., AGELOPOULOU, R. & JOST, A. (1980). Cellules de Sertoli et organogenese du
testicule foetal. Ann. Endocrinol. 41, 531-537.
MAGRE, S., AGELOPOULOU, R. & JOST, A. (1981). Action du s6rum de foetus de veau sur la
differentiation in vitro ou le maintien des cordons s6miniferes du testicule du foetus de Rat.
C.r. hebd. Acad. ScL, Paris 292, 85-89.
MAGRE, S. & JOST, A. (1980). The initial phases of testicular organogenesis in the rat. An
electron microscopy study. Archs Anat. microsc. Morph. exp. 69, 297-318.
MAGRE, S. & JOST, A. (1983). Early stages of the differentiation of the rat testis: relations
between Sertoli and germ cells. Current Problems in Germ Cell Differentiation, (eds A.
McLaren & C. C. Wylie), pp. 201-214. Cambridge University Press.
MAGRE, S., JOST, A. & VALENTINO, O. (1982). Morphogenese testiculaire in vitro et differentiation fonctionnelle des cellules de Sertoli. Influence du serum de foetus de veau. Colloque
de PINSERM: Ontogenese du systeme endocrinien. INSERM 109, p. 556 (abstract).
MERCHANT-LARIOS, H. (1976). The onset of testicular differentiation in the rat: an ultrastructural study. Am. J. Anat. 145, 319-330.
MOORE, G. E., GERNER, R. E. & FRANKLIN, H. A. (1967). Culture of normal human
leukocytes. J. Amer. Med. Ass. 199, 519-524.
O, W. & BAKER, T. C. (1976). Initiation and control of meiosis in hamster gonads in vitro. J.
Reprod. Fert. 48, 399-401.
PARKER, R. C. (1961). Methods of Tissue Culture, 3rd Edition. New York: Harper (Hoeber).
PRICE, J. M. (1979). The secretion of Miillerian inhibiting substance by cultured isolated
Sertoli cells of the neonatal calf. Am. J. Anat. 156, 147-157.
MAGRE,
ROMMERTS, F. F. G., KRUGER-SEWNARIAN, B. C , GROOTEGOED, J. A., DE JONG, F. H. & VAN
DER MOLEN, H. J. (1979). Secretion of proteins, androgen binding protein and oestradiol
by Sertoli cells in culture: effects of serum. Acta Endocrinol. 90, 552-556.
L. E. & ANDERSON, E. (1981). In vitro analysis of ovarian differentiation and the
initiation of meiosis in the rat. Acta Anat. 110, 189-205.
TAKETO, T. & KOIDE, S. S. (1981). In vitro development of testis and ovary from indifferent
fetal mouse gonads. Devi Biol. 84, 61-66.
THOMSON, J. D. (1942). Comparative studies on gonad development in the rat, the pig and in
cattle. Proc. Iowa Acad. Sci. 49, 475-501.
TORREY, T. W. (1945). The development of the urinogental system of the albino rat. II. The
gonads. Am. J. Anat. 76, 375-397.
TUNG, P. S., DORRINGTON, J. H. & FRITZ, I. B. (1975). Structural changes induced by folliclestimulating hormone or dibutyryl cyclic AMP on presumptive Sertoli cells in culture. Proc.
natn. Acad. Sci., U.S.A. 72, 1838-1842.
WOLFF, E. (1952). Sur la differentiation sexuelle des gonades de souris explant6es in vitro.
C.r. hebd. Acad. Sci., Paris 234, 1712-1714.
STEIN,
(Accepted 21 May 1984)