Previous investigations have shown that the teratocarcinoma line PCC3/A/1 is able to differentiate in vitro to produce red blood cells of the primitive hemopoietic cell population of the mouse embryo. The present paper demonstrates the presence in the same cultures of multipotential cells capable of giving rise to colonies containing erythrocytes, macrophages, and megakaryocytes when stimulated by pokeweed-mitogen spleen-cell conditioned medium.

Teratocarcinoma cells have been shown to be able to differentiate along hemopoietic lineages either in vitro (Cudennec & Nicolas, 1977) (see also Graham, 1977), or in vivo (Mintz & Illmensee, 1975; Papaioannou et al. 1975; Illmensee & Mintz, 1976; Papaioannou et al. 1978; Cudennec & Salaün, 1979). As previously described (Cudennec & Nicolas, 1977) blood-island formation regularly takes place in organ cultures of the clonal teratocarcinoma line PCC3/A/1 (Jakob et al. 1973; Nicolas et al. 1976). Morphological observation of blood-cell differentiation in the cultures showed that large primitive erythrocytes developed but non-nucleated erythroid cells never appeared under these conditions. The erythropoietic foci were, in all cases, associated with vesicles lined by an endodermal-like epithelium. In addition, the blood islands that arose in the cultures were transitory structures in which erythropoiesis ceased completely after 5 to 10 days.

Biochemical analysis showed that the red cells synthesize a set of embryonic hemoglobins characteristic of the primitive, yolk-sac type, generation of erythrocytes (Cudennec, Delouvee & Thiery, 1979). Consequently, it appears that the blood foci formed in organ culture are the expression of a yolk-sac hematopoietic potential by teratocarcinoma cells.

Such a developmental capability is never expressed in the histiotypic tissue cultures of this teratocarcinoma line although various differentiated tissues, especially endodermal epithelium, have been shown to arise under these conditions (Nicolas et al. 1975). All the attempts we have made to detect hemoglobin synthesis in such cultures have been unsuccessful.

To further characterize the developmental capabilities of the hemopoietic cells which differentiate in organ cultures of teratocarcinoma cells, we have determined the responsiveness of such cells to stimulating factors known to be operative on hemopoietic cells of both adult and fetal origin.

Media conditioned by pokeweed-mitogen-stimulated spleen cells (SCM), when used in soft agar cultures, stimulates the formation of colonies of neutrophils, macrophages, eosinophils and megakaryocytes fiom mouse bonemarrow cells (Metcalf, Parker, Chester & Kincade, 1974; Metcalf, MacDonald, Odartchenko & Sordat, 1975). In addition SCM contains (a) factor(s) capable of stimulating multipotential hemopoietic cells to form clones containing erythrocytes, neutrophils, macrophages, eosinophils and megakaryocytes (Johnson & Metcalf, 1977a). These mixed hemopoietic-colony-forming cells occur most frequently in yolk sac, early fetal liver and peripheral blood cultures but are also present in cultures of adult spleen or bone-marrow cells (Johnson & Metcalf, 1977a, b).

The effect of the various colony stimulating factors (Burgess, Metcalf & Russel, 1978) present in SCM was tested in agar cultures of cells obtained either from the hemopoietic foci present in organ cultures of teratocarcinoma cells or from histiotypic cultures.

Cultures of PCC3/A/1 cell line (see Fig. 1)

Teratocarcinoma cells were maintained undifferentiated by serial transplantations in 10 cm plastic Petri dishes (Nunclon). The medium (Dulbecco’s modified Eagle’s medium: DMEM) was supplemented by 15 % FCS (GIBCO). To set up organ cultures, the cells of a confluent culture were detached from the plastic by gentle pipetting. The resultant cell suspension was spun for 10 min at 800 g and the pellet divided into 4 parts (approx. 107 cells each) which were individually transferred onto a Millipore filter (0·22 μm pore size). These cultures were kept at the gas-medium (DMEM + FCS) interphase (Primary Organotypic cultures :PO cultures). Fifty percent of the PO cultures showed blood foci between day 10 and 30 of incubation (Cudennec & Nicolas, 1977). After 10 days some PO cultures were subcultured in secondary organotypic (SO) cultures by fragmentation. Each PO culture was divided into about fifty 1 mm3 pieces which were individually transferred onto a new Millipore filter. The percentage of SO cultures showing red-cell foci was 85 % and hemopoietic foci occurred sooner in SO cultures than in PO cultures. In spite of these differences between PO and SO cultures, there were no differences in the nature and relative percentages of the various differentiated hemopoietic cells identified by analysis of blood cell smears (Cudennec et al. 1979). All cultures were performed at 37 °C in an atmosphere composed of 10 % CO2 in air.

Fig. 1.

Different types of culture used to allow erythroid capabilities to be expressed by cells of the PCC3/A/I teratocarcinoma line.

Fig. 1.

Different types of culture used to allow erythroid capabilities to be expressed by cells of the PCC3/A/I teratocarcinoma line.

Preparation of blood-cell suspension

Tissues obtained from blood-forming areas of approximately fifty 9-day-old SO cultures were collected and transferred into Eisen’s balanced salt solution (EBSS). The aggregates were disrupted by gentle pipetting, the cells were washed twice in EBSS and resuspended in 2 ml cold EBSS. The cells were layered onto 10 ml EBSS containing 50 % (w/v) FCS and allowed to sediment for 75 min. In order to obtain a pure single cell suspension, we discarded the lower fraction of 2 ml containing cell aggregates and the upper fraction of 2 ml containing debris. Viability of cells was determined by Trypan blue exclusion and the cell numbers evaluated in a hemocytometer. This suspension was the source of the cells subsequently transferred into agar culture containing pokeweed-mitogen-stimulated spleen-conditioned medium (SCM).

Preparation of spleen-conditioned medium

C57B1 spleen cells were incubated for 5 days at a concentration of 2 × 106 cells per ml in RPMI-1640 containing 5 % heat-inactiváted human plasma and 0·05 ml of a 1:15 dilution of pokeweed mitogen per ml of culture medium (GIBCO). After incubation, the media were centrifuged for 10 min at 3000 g. The supernatant fluid was then harvested and Millipore filtered.

Agar cultures

Cells were cultured in agar medium as described previously (Metcalf et al. 1974; Metcalf et al. 1975; Johnson & Metcalf, 1977a). Briefly, cells were suspended in 0-3 % agar in DMEM supplemented with -20 % human plasma. One ml suspensions were plated into plastic Petri dishes containing 0·2 ml SCM (or 0·2 ml DMEM in control experiments). These cultures were incubated for 7 days.

Scoring of agar cultures

Cultures were scored for the presence of erythroid (red or pink) colonies or non-erythroid (white) colonies using a dissection microscope at × 40 magnification.

The cellular composition of the colonies was determined by picking off individual colonies, smearing them on glass slides and staining them either with benzidine and Giemsa or hematoxylin (McLeod, Shreeve & Axelrod, 1974) or for acetylcholinesterase detection (Karnovsky & Roots, 1964).

Some erythroid colonies were fixed for 20 min in 3 % glutaraldehyde in cacodylate buffer and post fixed for 1 h in 1 % OsO4. After alcohol dehydration, the colony was embedded in Epon 812, and ultrathin sections cut with a Reichert 0MU2 ultramicrotome. The sections were double stained with uranyl acetate and lead citrate, and observed in a Hitachi HS9 electron microscope.

Undifferentiated tumor cells (Embryonal Carcinoma cells:EC cells, the stem cells of the tumor) were harvested from exponentially growing cell cultures in which confluency had not yet been reached. The cell suspension was plated in soft agar culture with or without spleen-conditioned medium (SCM). After seven days of incubation, similar results were obtained irrespective of the presence of SCM in the cultures. Compact aggregates developed in which differentiated tissues, such as rhythmically beating muscle cells could sometimes be identified. But in no instance were hemopoietic colonies observed.

A second series of experiments was performed using cells harvested from organ cultures of PCC3/A/1 cell line (Fig. 1). As previously mentioned this tumor maintained under such conditions gave rise to various differentiated tissues including blood islands. The spatial organization of the differentiated tissues was chaotic. Consequently the mechanical picking-up of cells from blood-forming areas led to heterogenous cell suspensions. These suspensions were composed not only of hemopoietic cells but also of cells, or clumps of cells, from surrounding tissues -namely endodermal-like epithelium – as well as EC cells, the stem cells of the tumor.

In order to separate the single cells and the clumps, the suspensions were sedimented in 50 % FCS-containing culture medium. As described in the ‘Materials and methods’ section, only the single cells were used in the clonal agar culture system.

In addition to aggregates similar to those obtained in soft agar cultures of EC cells from histiotypic cultures, colonies of hemopoietic cells developed in 7-day soft-agar cultures of cells obtained from SO cultures. Both erythroid and non-erythroid colonies could be distinguished according to their size, colour and cellular composition. No hemopoietic colonies developed in control cultures without SCM. No colonies of erythroid cells developed from suspensions of cells prepared from SO cultures older than 20 days in which hemopoietic activity had ceased.

Erythroid colonies were characterized by the presence of erythroid cells easily visualized in living cultures due to extensive hemoglobinization. These cells were evenly distributed in colonies composed of up to several hundreds of cells and represented 40 to 70 % of the total colony cellularity. When examined by electron microscopy, the erythroid colonies contained erythroid cells at all developmental stages of hemoglobinization. In some colonies, the erythroid cells could be seen mixed with megakaryocytes (Figs. 2 and 3), macrophages and occasional neutrophils. These colonies appeared to be similar to those which develop from embryonic hemopoietic cells described previously (Johnson & Metcalf, 1977a, b).

Fig. 2.

Electron micrograph of a mixed-erythroid colony derived from teratocarcinoma cells, showing two different cell types: Erythroblast (E) and Megakaryocyte (M). (The colony was picked out of a day-7 agar culture containing SCM.) (Scale bar = 2 μm.)

Fig. 2.

Electron micrograph of a mixed-erythroid colony derived from teratocarcinoma cells, showing two different cell types: Erythroblast (E) and Megakaryocyte (M). (The colony was picked out of a day-7 agar culture containing SCM.) (Scale bar = 2 μm.)

Fig. 3.

Detail of the cytoplasm of a Megakaryocyte developed in a mixed-erythroid colony derived from teratocarcinoma cells. Note the presence of typical α granules. (Scale bar = 0·5 μm.)

Fig. 3.

Detail of the cytoplasm of a Megakaryocyte developed in a mixed-erythroid colony derived from teratocarcinoma cells. Note the presence of typical α granules. (Scale bar = 0·5 μm.)

Non-erythroid hemopoietic colonies were numerous in agar cultures (Table 1). They appeared most commonly unicentric but polymorphic. Some were composed of a tight center and a loose outer mantle of cells, the remainder being composed of dispersed cells.

Table 1.

Frequency of erythroid and non-erythroid colony-forming cells in agar cultures of organ-cultured PCC3/A/1 cells*

Frequency of erythroid and non-erythroid colony-forming cells in agar cultures of organ-cultured PCC3/A/1 cells*
Frequency of erythroid and non-erythroid colony-forming cells in agar cultures of organ-cultured PCC3/A/1 cells*

To determine the cellular composition of the erythroid and non-erythroid hemopoietic colonies, individual colonies were stained with either benzidine, Giemsa or acetylcholinesterase stains. Thus, after seven days of culture, individual red-coloured colonies were sequentially removed and the cells smeared onto microsome slides, fixed with methanol and stained with benzidine and Giemsa stains. When examined microscopically, stained smears of erythroid colonies contained benzidine-positive erythroid cells at all stages of maturation, intermingled with macrophages, megakaryocytes and occasionally, neutrophils. To determine the proportion of erythroid colonies that contained megakaryocytes, 30 colonies were sequentially removed from day-7 cultures and placed onto chick-egg-albumin-coated slides, allowed to air dry, followed by acetone fixation and acetylcholinesterase staining. Twenty of the 30 colonies contained acetylcholinesterase-positive megakaryocyte. A similar procedure was applied to both erythroid and non-erythroid colonies in order to determine the overall incidence of megakaryocyte clones. A total of 96 sequential colonies were removed and stained with acetylcholinesterase and Giemsa stains. Eleven of the colonies contained only megakaryocytes, 21 contained megakaryocytes and erythroid cells and the remaining 64 colonies contained only macrophages.

The frequency on day 7 of culture, of colonies developing from suspensions of blood-forming SO cultured cells is reported in Table 1. Five series of cultures were performed from five different sets of teratocarcinoma cultures, using a total of four batches of human plasma as medium supplement. The cell density at seeding varied from 2 × 104 to 2 × 105 cells per 1 ml culture but did not affect the frequency of colonies in the cultures of a replicate series.

The differences in colony frequency from experiment to experiment were probably due to the different batches of human plasma used and the composition of the teratocarcinoma cell suspension cultured.

Human plasma number 1 was pretested in Melbourne on normal fetal liver cells before being used in teratocarcinoma cell cultures. Batch numbers 2, 3 and 4 were purchased from a hospital in Paris. Mixed erythroid colony formation obtained from teratocarcinoma cell cultures appeared to be highly dependent on the human-plasma batch (some of them were absolutely insufficient for the agar colony growth). The frequency of erythroid colony precursors obtained in cultures containing human plasma number 1 corresponds approximately to those found in mouse embryonic yolk sac on day 12 of gestation (Johnson & Metcalf, 1977b).

Previous results have shown that the PCC3/A/1 teratocarcinoma line is able to produce spontaneously in vitro red cells pertaining to the primitive, yolk-sactype, generation (Cudennec & Nicolas, 1977; Cudennec et al. 1979). This transient differentiation represents the unique erythroid capability expressed by PCC3/A/1 in mass culture. The lack of definitive erythrocyte formation in such a situation might result either from the inadequacy of the culture system to provide the specific stimulation to the erythroid precursors or from the depletion of hemopoietic stem-cell compartment due to the differentiation of each stem cell toward the primitive hemopoietic generation.

Present results show that, in addition to well-differentiated hemopoietic cells, blood islands differentiating in organ culture of teratocarcinoma cells contain precursors of several hemopoietic lineages. It also implies that the short life span of erythropoiesis in these cultures is not due to the depletion of the hemopoietic stem-cell compartment, but possibly due to the lack of appropriate stimulation for further differentiation. During normal murine development the fetal liver succeeds the yolk sac as the primary fetal hemopoietic organ. No tissues structurally similar to fetal liver have been observed in organ cultures of PCC3/A/1 cells. Consequently, failure of hepatic differentiation may result in a lack of the necessary regulatory signals for further hemopoietic differentiation of precursors present in the yolk sac. This hypothesis is consistent with data showing that fetal liver hemopoiesis is sustained by the migration of cells into the liver primordium at the 28-somite stage (Johnson & Moore, 1975).

Burgess
,
A. W.
,
Metcalf
,
D.
&
Russel
,
S.
(
1978
).
In Differentiation of Normal and Neoplastic Cells
(ed.
B.
Clarkson
,
P. A.
Marks
and
J. E.
Till
).
Cold Spring Harbor Conference on Cell proliferation
, vol.
5
, p.
339
.
Cudennec
,
C. A.
,
Delouvee
,
A.
&
Thiery
,
J. P.
(
1979
).
Embryonic hemoglobin production by teratocarcinoma-derived cells in in vitro cultures
.
In Cell Lineage, Stem Cells and Cell Determination
(ed.
N.
Le Douarin
), INSERM Symposium, no.
10
, p.
163
.
Amsterdam
:
Elsevier/North-Holland Biomedical Press
.
Cudennec
,
C. A.
&
Nicolas
,
J. F.
(
1977
).
Blood formation in a clonal cell line of mouse teratocarcinoma
.
J. Embryol. exp. Morph
.
38
,
203
210
.
Cudennec
,
C. A.
&
Salaün
,
J.
(
1979
).
Definitive red blood cell differentiation in a clonal line of mouse teratocarcinoma cultured in vivo in the chick embryo
.
Cell Diff
.
8
,
75
82
.
Graham
,
C. F.
(
1977
).
Teratocarcinoma cells and normal mouse embryogenesis
.
In Concepts in Mammalian Embryogenesis
(ed.
M. I.
Sherman
), p.
315
.
Cambridge, M.A
. :
M.I.T. Press
.
Illmensee
,
K.
&
Mintz
,
B.
(
1976
).
Totipotency and normal differentiation of single teratocarcinomacells cloned by injection into blastocysts
.
Proc. natn. Acad. Sci., U.S.A
.
73
,
549
553
.
Jakob
,
J.
,
Bonn
,
T.
,
Gaillard
,
J.
,
Nicolas
,
J. F.
&
Jacob
,
F.
(
1973
).
Tératocarcinome de la souris: Isolement, culture et propriétés de cellules à potentialités multiples
.
Ann. Microbiol. Inst. Pasteur
124B
,
269
282
.
Johnson
,
G. R.
&
Metcalf
,
D.
(
1977a
).
Pure and mixed erythroid colony formation in vitro stimulated by spleen conditioned medium with no detectable erythropoietin
.
Proc. natn. Acad. Sci., U.S.A
.
74
,
3879
3882
.
Johnson
,
G. R.
&
Metcalf
,
D.
(
1977b
).
Nature of cells forming erythroid colonies in agar after stimulation by spleen conditioned medium
.
J. Cell Physiol
.
94
,
243
252
.
Johnson
,
G. R.
&
Moore
,
M.A. S.
(
1975
).
Role of stem cell migration in initiation of mouse foetal lines haemopoiesis
.
Nature
258
,
726
728
.
Karnovsky
,
M. J.
&
Roots
,
L.
(
1964
).
A direct-coloring thiocholine method for cholinesterase
.
J. Histochem. Cytochem
.
12
,
219
221
.
McLeod
,
D. L.
,
Shreeve
,
M. M.
&
Axelrad
,
A. A.
(
1974
).
Improved plasma culture system for production of erythrocytic colonies in vitro: quantitative assay method for CFU-E
.
Blood
44
,
517
534
.
Metcalf
,
D.
,
MacDonald
,
H. R.
,
Odartchenko
,
N.
&
Sordat
,
B.
(
1975
).
Growth of mouse megakaryocyte colonies in vitro
.
Proc. natn. Acad. Sci., U.S.A
.
72
,
1744
1748
.
Metcalf
,
D.
,
Parker
,
J.
,
Chester
,
H. M.
&
Kincade
,
P. W.
(
1974
).
Formation of eosinophilic like granulocytic colonies by mouse bone marrow cells in vitro
.
J. Cell Physiol
.
84
,
275
290
.
Mintz
,
B.
&
Illmensee
,
K.
(
1975
).
Normal genetically mosaic mice produced from malignant teratocarcinoma cells
.
Proc. natn. Acad. Sci. U.S.A
.
76
,
3585
3589
.
Nicolas
,
J. F.
,
Anver
,
P.
,
Gaillard
,
J.
,
Guenet
,
J. L.
,
Jakob
,
H.
&
Jacob
,
F.
(
1976
).
Cell lines derived from teratocarcinomas
.
Cancer Res
.
36
,
4224
4231
.
Nicolas
,
J. F.
,
Dubois
,
P.
,
Jakob
,
H.
,
Gaillard
,
J.
&
Jacob
,
F.
(
1975
).
Tératocarcinome de la souris: Différenciation en culture d’une lignée de cellules primitives à potentialités multiples
.
Ann. Microbiol. Inst. Pasteur
126A
,
3
22
, 3-22.
Papaioannou
,
V. E.
,
Gardner
,
R. L.
,
McBurney
,
M. W.
,
Babinet
,
C.
&
Evans
,
M. J.
(
1978
).
Participation of cultured teratocarcinoma cells in mouse embryogenesis
.
J. Embryol. exp. Morph
.
44
,
93
104
.
Papaioannou
,
V. E.
,
McBurney
,
M. W.
,
Gardner
,
R. L.
&
Evans
,
M. J.
(
1975
).
Fate of teratocarcinoma cells injected into early mouse embryo
.
Nature, Lond
.
258
,
70
73
.