J. Embryol. exp. Morph., Vol. 15, 2, pp. 245-260, April 1966
With 1 plate
Printed in Great Britain
245
The effects of erythropoietin on haem synthesis in
mouse yolk sac and cultured foetal liver cells
By R. J. COLE & JOHN PAUL 1
From the Department of Biochemistry, University of Glasgow
Erythropoiesis in the foetal mouse is reported to occur initially in the yolk
sac, followed by the liver, the spleen and finally the bone marrow (Snell, 1956;
Borghese, 1959). At 12 days of gestation 97 % of the circulating erythrocytes
are nucleated (Craig & Russell, 1964) and presumably arise from the yolk sac
blood islands. The cells of this generation persist in decreasing numbers until
the 16th day of gestation. Haemoglobin can first be recognized by histochemical
methods (O'Brien, 1961) in the yolk sac before somite formation at the early
head fold stage and a network of blood islands, almost covering the embryo,
rapidly develops (Plate 1, figs. A-F)'
The rate of erythropoiesis in the adult mammal is controlled by the hormone
erythropoietin, which is released from the kidney as a response to lowered
oxygen tension in the arterial blood supply (Fisher, Schofield & Porteus, 1965).
Erythropoietin, which is glycoprotein, and not species specific, appears to exert
its effect at several levels in the erythropoietic system (Lajtha, 1964). It affects
the erythropoietic stem cell population directly, causing an increased number
of cells to enter the erythrocytic pathway of differentiation (Filaminovicz &
Gurney, 1961; Krantz, Gallien-Lartigue & Goldwasser, 1963), and also increases
haemoglobin synthesis in existing normoblasts (Stahlman, Brecher & Moores,
1962). Further, increases in the absolute number of normoblasts present (Fisher,
Roh, Couch & Nightingale, 1964) and increased release of reticylocytes from
the marrow pool (Fisher, Lajtha, Buttoo & Porteus, 1965) have been observed.
It has recently been found (Gallien-Lartigue, personal communication) that
cells of the erythropoietic series can be maintained in organ cultures of mouse
foetal liver in the presence of anaemic serum but disappear if cultures are grown
with normal serum.
The present communication describes the effects of purified sheep erythropoietin on intact mouse embryos at the yolk sac stages and on cells from foetal
livers disaggregated at various stages of development, both in vitro.
1
Authors' address: The Department of Biochemistry, The University, Glasgow, W. 2,
Scotland.
16-3
246
R. J. COLE & J. PAUL
MATERIALS AND METHODS
All embryos were obtained from random-bred 10- to 12-week-old Swiss
albino mice (Porton strain) after induction of ovulation and mating with
'Gestyl' and 'Pregnyl' (Organon, Ltd.). This permits accurate timing of developmental ages and provides large numbers of embryos of the required stage.
All embryos are timed from ovulation (midnight following the Pregnyl injection) and the morning on which mating plugs are present is day 0.
Embryos to be cultured intact were isolated on the 8th and 9th days of gestation. The uteri were excised aseptically, the decidual masses removed from the
uterine musculature in culture medium based on tris/citrate buffered balanced
salt solution (Paul, 1965) and the embryo, surrounded by trophoblastic giant
cells and Reichert's membrane, isolated (New & Stein, 1964). The embryos
were prepared for culture by opening Reichert's membrane but leaving all the
adhering trophoblast attached to the ectoplacental cone. The embryos were
cultured in 3-5 cm diam. Falcon polystyrene dishes in 5 % CO 2 in air at 36-5 °C.
Each dish contained 1 ml of medium, i.e. just enough to cover the larger embryos. The developmental stage of each embryo was noted and the main axes
were measured, at the beginning and end of the culture period, which was 1621 h. Any markedly abnormal or retarded embryos were rejected from the
final analysis of the data.
Livers from more advanced embryos were removed aseptically in Hanks's
balanced salt solution (Paul, 1965) after checking the developmental stage of
the embryo by its external appearance (Griineberg, 1943). The livers were
disaggregated by exposing them to Difco trypsin 1/250 in isotonic sodium
chloride/sodium citrate, pH 7-8 (Paul, 1965), containing 0-3 % sodium carboxymethyl cellulose (to protect the cell membranes), at 4°C overnight. The supernatant trypsin was then removed and the tubes warmed to 36-5 °C for 5 min.
Culture medium containing 5 % calf serum was added and disaggregation completed by pipetting. The cells were washed once in medium, checked for clumping and for viability (by dye exclusion with 0-1 % naphthalene black) and
counted with a Coulter electronic cell counter, Model D (Threshold 50, Aper-
PLATE 1
A. Surface view of 2-4-somite embryo stained with o-dianisidine to demonstrate the site of
haemoglobin formation. Reichert's membrane removed.
B. Reconstruction of embryo shown in A. ch., Chorion; ect.c, ectoplacental cone (with
maternal erythrocytes); n.f., neural fold; y.s., yolk sac.
C. Surface view of 6-8-somite embryo stained as A.
D. Reconstruction of embryo shown in C.
E. 'Egg cylinder' stage stained with odianisidine. Haemoglobin present only in maternal
erythrocytes attached to surface.
F. Portion of yolk sac of approximately 30-somite embryo, stained with o-dianisidine.
/. Embryol. exp. Morph,, Vol. 15, Part 2
PLATE 1
ect.c.
n.f.
0-25 mm
A
ect.c.
0-25 mm
005
0-2 mm
R. J. COLE & J. P A U L
facing p. 246
Erythropoietin and foetal erythropoiesis
247
ture current 2). This technique regularly produces single cell suspensions of
80-90 % viability at all stages of liver development studied.
The medium used for culturing both the intact embryos and liver cells was
based on Waymouth's MB 752/1 (Waymouth, 1959), which has been found
especially satisfactory for early mammalian embryonic material (Cole & Paul,
1966). This was modified by the addition of 5 % neonatal calf serum, 5 % female
mouse serum, and 0-01 mM-FeCl3, so that the final iron concentration of the
medium was between 1 and 2 /*g/ml. Iron determinations on sera and complete
media, were carried out by the dipyridyl method of Ramsay (1951).
Foetal liver cells were cultured in suspensions of 0-5-1 x 106 cells/ml in 1 ml
aliquots in 10 ml Excello test tubes covered with an Oxoid cap, in 5 % CO2 in
air at 36-5 °C.
Haem synthesized in vitro was labelled by the addition of 59FeCl3 previously
equilibrated with 50 % mouse serum in Hanks's BSS, for at least 12 h at 36-5 °C
at a concentration of 10 /^c/ml. The specific activity of the 59Fe used in these
experiments varied from 3-0 to 12-5 /*c//*g Fe and was added to the cultures at
the level of 1-2 /*c/ml. At the end of the incubation time the embryos or cells
were washed with Hanks's BSS, and 2 parts H 2 0:1 part Drabkin's solution
(Wintrobe, 1961) was added (1 ml final volume per embryo or per cell aliquot).
The samples were lysed by freezing and thawing three times, the debris sedimented by centrifugation and the supernatant decanted and acidified with 0-1 ml
of N-HC1. Total haem was extracted from the supernatant with 1-2 ml of ethyl
methyl ketone (Teale, 1959; Krantz et al. 1963). Aliquots of 0-5 ml were dried
on stainless steel planchettes and counted in a Nuclear Chicago gas-flow counter.
The lysed debris, and the Drabkin's solution supernatant and remaining
ethyl methyl ketone from each intact embryo were recombined, hydrolysed
with perchloric acid and the deoxyribose content of the hydrolysate measured
by the diphenylamine method of Burton (1956). (The presence of ethyl methyl
ketone interferes with the estimation if trichloroacetic acid is used.) This enabled
haem synthesis to be expressed against deoxyribose content per embryo, which
provided a reliable basis for quantitative analysis and overcame the problem of
the range of developmental stages found at the same time of gestation.
The erythropoietin used in these experiments (N.I.H. Haematology Study
Section, step 4, Lot K103,217 A) was prepared from plasma of phenylhydrazinetreated sheep (White, Gurney, Goldwasser & Jacobson, 1960) and had a specific
activity of 22 units/mg protein. It was used at concentrations between 0-15 and
1 unit/ml.
RESULTS
Yolk-sac erythropoiesis
During this early period of development, when morphological changes are
especially rapid, considerable variations in the stages of development reached
occur both within litters and between litters at the same times after ovulation.
248
R. J. COLE & J. PAUL
Each experiment therefore contained embryos of various stages, and the haem
synthesized was labelled throughout the culture period. The yolk sac contains
undifferentiated, or differentiating erythroid precursor cells before the first
fig deoxyribose/embryo when harvested
Text-fig. 1. Haem synthesis during yolk-sac erythropoiesis, labelled with 5BFe for 21 h
in vitro showing relationship between deoxyribose content and haem synthesized in:
(1) control embryos, and (2) embryos exposed to 0 4 unit/ml erythropoietin. (3A)
and (3B) are the 95% confidence limits of the pooled data. O, Control embryo;
x, embryo exposed to erythropoietin. (A-E) Developmental stages of embryos
when explanted, relative to deoxyribose content after 21 h development in vitro.
A, Primitive streak; B, 4-somite; C, 8-somite; D, 14-somite; E, 20-somite.
Erythropoietin and foetal erythropoiesis
249
appearance of haemoglobin in the embryo; hence factors which stimulate
erythropoiesis might bring about haemoglobin synthesis earlier than normal,
or increase the rate of synthesis, or both.
The relationship between the amount of haem synthesized and stage of
development, estimated by deoxyribose content/embryo, is shown in Text-fig. 1.
The thirty embryos used in this experiment were derived from seven litters,
isolated on the late 8th or early 9th day after ovulation, and ranged from presomite to early heart-beat stages. The presence of erythropoietin at 0-2 unit/ml
10 r-
1
0
1
2
3
4
5
6
7
/ig deoxyribose/embryo when harvested
Text-fig. 2. As Text-fig. 1, showing relationship between deoxyribose content and
haem content in: (1) embryos exposed to 1 unit/ml erythropoietin; (2) control
embryos; and (3) pooled embryos. (4 A) and (4 B) are the 95 % confidence units of (3).
did not affect either the total amount of haem synthesized, or cause it to appear
at an earlier stage of development. The stage at which 59Fe incorporation into
haem first occurred, in embryos containing 4-5 fig deoxyribose at the end of the
culture period, is in agreement with the stage at which haemoglobin can first
be detected by histochemical methods. Increasing the concentration of erythropoietin to 1 unit/ml is similarly ineffective (Text-fig. 2). This experiment utilized
twenty embryos from three litters, ranging from pre-somite to 6-8 somite stages
at the time of explantation. The difference in the regression coefficients of the
treated and untreated embryos is not significant and haem synthesis was again
250
R. J. COLE & J. PAUL
initiated at similar embryonic stages. The background values in these experiments were equivalent to less than 1 JUJUM haem formed/embryo.
Foetal liver erythropoiesis
Haem synthesis in duplicate aliquots of the same cell suspension was measured
by determining 59Fe incorporation over periods of 2-4 h and calculated as
fifiu. haem formed per hour per million cells present at time 0, and plotted at the
mid point of the exposure to 59Fe. As far as possible embryos from single litters
Table 1. The numbers of cells per liver obtained after disaggregation with
trypsin, at different developmental stages
Yield of cells/liver
xlO6
Days of gestation
0-4
0-75
5
6
10
10
10*
12*
141
15
18
Neonatal
30 r-
a 20
8
I
10
16
24
32
40
48
56
Hours
Text-fig. 3. Haem synthesis in 10|-day foetal mouse liver cultures: x
x, with
0-15 unit/ml erythropoietin; O
O, control. Solid bars on the time ordinate
indicate period of exposure to 69Fe.
Erythropoietin and foetal erythropoiesis
251
were used for each experiment, although this was not possible with the earlier
developmental stages (Table 1). Any obviously retarded or abnormal embryos
were rejected.
The earliest livers studied were from lOf-day-old embryos. Text-rig. 3 shows
the results obtained from pooled livers from thirty-one embryos from three
litters, all at comparable stages of development, exposed to 0-15 unit erythropoietin/ml. At this stage the livers contain a few faintly pink foci and obviously
50
40
30
I
i
20
10
16
24
32
40
48
56
64
72
Hours
Text-fig. 4. Haem synthesis in 12^-day foetal liver cultures: x
0-4 unit/ml erythropoietin; O
O, control.
x, with
contain little haemoglobin. In the first 8 h of culture both control and erythropoietin-treated cells show a rising rate of haem synthesis. This rise in rate is,
however, only maintained in the presence of erythropoietin, when it reaches a
maximum after approximately 28 h and then declines. The rate of synthesis in
control cultures declines continuously after 8 h. At its maximum, the rate of
haem synthesis in the erythropoietin-treated cells is 19 times that in the control
252
R. J. COLE & J. PAUL
cells at the same time, and more than 6 times the maximum reached by the
control. The failure of late yolk-sac stage embryos to respond to erythropoietin
indicates that the response detected in these liver cultures is by cells which were
an inherent part of the liver structure at the time of explantation and not maturing
cells of the yolk-sac generation of erythropoiesis, reaching it via the circulation.
50 r
40
£ 30
o
©
I"
l H
r
20
I
10
0
8
16
24
32
40
48
56
Hours
Text-fig. 5. Haem synthesis in 14^-day foetal liver cultures. Legend as Text-fig. 2.
A similar response was seen with liver cells from 12^-day embryos exposed
to 0-4 unit of erythropoietin/ml. Text-fig. 4 shows the results obtained from fiftysix embryos from four litters. A maximum rate of incorporation is again
reached after approximately 28 h; this maximum rate is nearly 16 times the
control value at equivalent time, but only 3 times the initial rate.
The rate of haem synthesis in cells from 14^-day embryos (Text-fig. 5)
declines during the first 12 h in culture in the presence of erythropoietin as well
as in control cultures. In the presence of erythropoietin this decline is subsequently reversed, and a maximum rate of synthesis occurs after 24-28 h; in
contrast, it remains very low in the absence of erythropoietin. In these cultures,
stimulated with 0-4 unit of erythropoietin/ml, the maximum rate of synthesis
Erythropoietin and foetal erythropoiesis
253
observed is 45 times that of controls at equivalent times, and approximately
9 times the initial rate.
A marked change in this pattern of response occurs in cultures derived from
15^-day embryonic livers (Text-fig. 6). The mean rate of haem synthesis in the
first 4 h is slightly higher than the maximum erythropoietin-stimulated rate
60
50
40
i
20
10
0
8
16
24
32
40
48
56
64
72
80
Hours
Text-fig. 6. Haem synthesis in 15^-day foetal liver cultures. Legend as Text-fig. 2.
observed in 14^-day embryonic livers. However, the rate of haem synthesis in
both erythropoietin-treated and control cultures falls from the time of explantation and erythropoietin has no significant effect.
A similar decline in the rate of haem synthesis and lack of responsiveness to
erythropoietin has been observed in 16^-day, 18^-day, immediately prenatal
and 1st day post-natal embryos (Text-figs. 7-10). Cells of the 16-day liver
254
R. J. COLE & J. PAUL
showed only very slow rates of haem synthesis and this declined in culture even
in the presence of erythropoietin. The initial rate of synthesis increased with
stage of development until in 19-day embryos, immediately before birth, it was
approximately f that of the highest rate observed, i.e. in cultures from
15^-day livers.
o
o
h
10
o>
a
X
X
1
LJ
8
0
16
24
i
i
32
64
Hours
Fig. 8
Fig. 7
50
r
40
30
30 r
6
8
h
20
I
I
i
10
X
16
24
32
Hours
Fig. 9
40
48
56
20
10
16
24
Hours
Fig. 10
Text-fig. 7. Haem synthesis in 16^-day foetal liver cultures. Legend as Text-fig. 2.
Text-fig. 8. Haem synthesis in 18-day foetal liver cultures. Legend as Text-fig. 2
Text-fig. 9. Haem synthesis in immediately prenatal liver cultures. Legend as Textfig. 2.
Text-fig. 10. Haem synthesis in 1st day postnatal liver cultures. Legend as Text-fig. 2.
32
72
Erythropoietin and foetal erythropoiesis
255
It has been reported (Borghese, 1959) that in the foetal mouse the spleen
becomes an erythropoietic organ at 16 days of gestation. However, no significant incorporation of 59Fe into haem could be detected in cultures derived
from 16^- or 18-day foetal spleen, or neonatal spleen, either initially or after
exposure to erythropoietin. The background values in this series of experiments
were equivalent to 1 fi/m haem/h/106 cells.
DISCUSSION
Several recent studies have suggested that globin synthesis, and hence haemoglobin formation, is controlled at the polysome level by the availability of haem.
The addition of hemin and other tetrapyrrols to isolated avian erythrocyte
nuclei increases the rate of globin synthesis and reverses the inhibition caused
by high O2 tensions (Hammel & Bessman, 1964, 1965). An increased incorporation of 14C valine into haemoglobin in the presence of hemin in rabbit
reticulocytes has also been observed (Bruns & London, 1965). The addition of
the haem precursor #-aminolaevulinic acid to de-embryonated chick blastoderms causes increased haemoglobin synthesis and this effect is actinomycin Dresistant but puromycin-sensitive (Levere & Granick, 1965). Erythropoietin,
acting on potential haemoglobin synthesizing cells in vitro, causes an increase in
the rate of synthesis of RNA (which is probably messenger RNA) within
15min (Krantz & Goldwasser, 1965). The primary effect of erythropoietin
action may therefore be to induce synthesis of messenger RNA for an enzyme
or enzymes responsible for haem synthesis.
In the chick blastoderm, synthesis of messenger RNAs essential for haemoglobin synthesis has occurred before the formation of the head fold stage,
several hours before haemoglobin first becomes detectable, at the 7-8-somite
stage (Wilt, 1965). In even the earliest mouse embryos used in the present study,
transcription of messenger RNA necessary for the initiation of haemoglobin
synthesis may therefore have occurred before the time of explantation and
exposure to erythropoietin in vitro. Since no increase in haem synthesis occurs
in the presence of erythropoietin during the yolk-sac stage, the majority, if not
all, of the potentially erythropoietic cells available at this stage must previously
have been stimulated to differentiate. Whether this initial phase of haemoglobin
synthesis is independent of an erythropoietin-like stimulus, or depends on an endogenous or exogenous, i.e. maternal, hormone supply cannot yet be ascertained.
The results of the experiments with foetal liver cells are summarized in
Text-fig. 11. These show that as early as lOf days of gestation, soon after the
liver can first be recognized as a discrete organ, at least on gross examination,
it contains cells capable of responding to erythropoietin by synthesizing haem.
The maximum rates of synthesis observed, following erythropoietin stimulation
in 12£- and 14^-day embryos and the initial spontaneous rate in 15-day embryos, are all similar. This suggests that, with the doses of erythropoietin used,
256
R. J. COLE & J. PAUL
the sensitive cell population is responding to a similar extent as in vivo, and that
this represents the maximum obtainable rate of synthesis in these conditions.
The temporal pattern of response to erythropoietin seen in the 10-14-day
embryonic livers is very similar to that observed in adult rat bone marrow cells
cultured under similar conditions (Krantz et al. 1963), with a maximum rate of
haem synthesis occurring 28-30 h after the initial stimulation by erythropoietin.
However, the maximum rate of haem synthesis in the liver cells is 3-4 times
higher than that in rat marrow cultures on a per cell basis, with equal doses of
60 r
Rate —Ep when
+ Ep max.—<
10 11 12 13 14 15 16 17 18 1 9 f 20
Birth
Days
Text-fig. 11. The relationship between developmental age, and the rate of haem
synthesis in foetal liver cells in vitro; (1) immediately following disaggregation;
(2) at maximum after erythropoietin stimulation; and (3) in control cultures when
the erythropoietin-induced rate is maximal. (Corrected for differences in erythropoietin levels.)
erythropoietin. The lag period in the response to erythropoietin of approximately 10 h, observed in the cultures of 14^-day liver, was also consistently
found in the rat marrow cultures. Induction of erythropoiesis in vivo in the
polycythaemic mouse takes from 12-18 h and suggests that an initial period of
synthesis of precursors occurs, before haem appears. In this case failure to
demonstrate a similar lag in cultures from 10^-day livers might indicate that
these precursors were present at this stage of embryonic development.
These results suggest that erythropoietin production may be a prerequisite for
hepatic erythropoiesis in the mouse foetus. From the kinetics of the response
to erythropoietin it is likely that erythropoietin production rapidly reaches
a maximum or is actually initiated during the 13th to 14th days of development.
Erythropoietin and foetal erythropoiesis
257
Examination of livers from foetuses between 16 days and birth, and from neonatal animals shows a second phase of haem synthesis with a low rate on the
16th day but increasing until birth. This may indicate cytochrome synthesis in
the liver parenchyma, or as it is generally accepted that bone marrow erythropoiesis begins in the mouse foetus at 16 days it is possible that this secondary
haem synthesizing cell population is derived from circulating maturing cells
produced by the bone marrow.
The stem cell population in a haemopoietic organ may be endogenous in
origin ('extra vascular'), or, alternatively have originated in another site, and
migrated via the circulation to the site of differentiation. Examination of spatial
relationships in fixed material cannot decide between these alternatives, because
the stem cells may possess invasive properties assisting their migration.
Taylor (1965) has shown the presence of pluripotential haemopoietic cells in
12- and 15-day foetal mouse liver, with lymphopoietic capacity in the thymus of
irradiated recipients. Since pluripotential cells cannot be recovered from the
thymus, it is considered that the cells are irreversibly stimulated to differentiate
just before or just after entry. Cells from immediately neonatal mouse liver can
form spleen colonies in irradiated recipients but with only one-third of the
efficiency found with adult marrow (McCulloch & Till, 1963). However, the
proportion of colonies derived from this material representing the different
pathways of haemopoietic differentiation is not stated and the present study
indicates that only few erythropoietin-sensitive stem cells could remain in the
liver at this stage. Therefore, two alternative explanations for the change in the
pattern of haem synthesis, between 14^- and 15^-day foetal mouse livers,
observed in the present study, are possible. The erythropoietin-sensitive stemcell population remaining in the liver may migrate during the 14th day of
development, leaving behind only those cells which have received the stimulus
to differentiate. Alternatively, the stem-cell population originally present, and
its descendants, are entirely switched into the erythropoietic pathway at this
stage of development and the stem cells of the subsequent haemopoietic sites
have a separate origin.
SUMMARY
1. The effects of erythropoietin on haem synthesis during yolk sac and
hepatic erythropoiesis in the foetal mouse have been studied in vitro.
2. Cultures were maintained in Waymouth's medium MB 725/1 with 5 %
calf serum, 5 % mouse serum and 0-01 mM-FeCl3. Yolk-sac erythropoiesis was
studied in embryos explanted at stages between primitive streak and heart beat
and cultured intact, and hepatic erythropoiesis in cell suspensions obtained by
trypsinizing foetal livers.
3. Haem synthesized in vitro was labelled by exposing cultures to 59Fe
equilibrated with mouse transferin, and was separated chemically with ethyl
methyl ketone.
258
R. J. COLE & J. PAUL
4. Haem synthesis does not occur in embryos containing less than 4-5 fig
deoxyribose. Neither the stage at which synthesis is initiated, nor the rate of
synthesis in the yolk sac is affected by erythropoietin.
5. Foetal livers from 10f-, 12£- and 14^-day foetuses contain cells responsive
to erythropoietin by increased rates of haem synthesis. Synthesis is only maintained in vitro in the presence of erythropoietin and maximum rates occur
approximately 28 h after initial exposure.
6. The rate of haem synthesis in 15^-day liver cells is not responsive to erythropoietin and declines in culture, but the initial rate is similar to the maximum
rate obtained from earlier liver cells stimulated by erythropoietin in vitro.
7. The rate of haem synthesis in 16-day liver cells is very low, but rises in
later foetuses until birth, although no response to erythropoietin occurs.
8. The kinetics of response to erythropoietin in 12-14-day mouse liver cells
is similar to that of adult rat bone marrow in vitro.
9. No haem synthesis could be detected in cultures from spleens from 16£- or
18-day foetuses, or from neonatal animals, with or without erythropoietin.
RESUME
Les effets de Verythropoietine sur Phemosynthese dans la vesicule ombilicale
des souris et dans des cultures de cellules hepatiques foetales
1. On a etudie in vitro les effets de Perythropoietine sur Phemosynthese pendant l'erythropoiese dans la vesicule ombilicale et dans le foie des souris
foetales.
2. Les cultures etaient maintenues dans le milieu de Waymouth, MB 725/1
avec 5 % de serum de veau, 5 % de serum de souris et 0-01 HIM FeCl3. On a etudie
l'erythropoiese dans la vesicule ombilicale d'embryons preleves a diverses
etapes entre la ligne primitive et le battement du coeur et maintenues intacts en
la culture. On a etudie l'erythropoiese hepatique au moyen de la suspension de
cellules, obtenue par l'action de la trypsine sur le foie foetal.
3. On a suivi Phemosynthese in vitro en exposant les cultures au 59Fe, equilibre
par de la transferine de souris, et ensuite on separa le heme chimiquement au
moyen d'ethyle methyle cetone.
4. L'hemosynthese n'a pas lieu dans les embryons qui contiennent moins de
4-5 fig de desoxyribose. L'erythropoietine n'a aucun effet ni sur Petape a laquelle
commence la synthese ni sur la vitesse de synthese dans la vesicule ombilicale.
5. Les foies foetaux de lOf jours, de 12^-jours et de 14£ jours, contiennent
des cellules, qui reagissent l'erythropoietine par Pacceleration de leur vitesse
d'hemosynthese. La synthese ne se maintient in vitro qu'en la presence de
l'erythropoietine et les vitesses maximum sont atteintes 28 heures, a peu pres,
apres Pexposition originelle.
6. La vitesse de l'hemosynthese dans les cellules de foie de \5\ jours ne
montre pas de reaction a Perythrbpoietine, et cette vitesse diminue dans les
Erythropoietin and foetal erythropoiesis
259
cultures. Pourtant, la vitesse primaire est semblable a la vitesse maximum
obtenue dans des cellules hepatiques moins developpees, stimulees par l'erythropoietine in vitro.
7. La vitesse de l'hemosynthese dans les cellules de foie de 16 jours reste tres
basse, mais elle s'accelere dans des foetus plus avances et persiste jusqu'a la
naissance bien qu'aucune reaction a l'erythropoietine n'ait lieu.
8. La cinetique de reponse a l'erythropoietine dans les cellules de foie de
souris de 12-14 jours est semblable a celle qui a lieu dans la moelle des rats
adultes in vitro.
9. II est impossible de constater l'hemosynthese ni dans les cultures de rate des
foetus de 16^ ou de 18 jours, ni dans la rate des animaux nouveau-nes, que
l'erythropoietine soit presente ou non.
This work was supported by the National Cancer Institute, United States Public Health
Service (Grant CA-05855). We are grateful to Dr Sanford Krantz of the University of Chicago
for erythropoietin and for his advice, and to Organon Ltd. for gonadotrophins used in this
study.
The histochemical preparations were made by Mr C. F. van der Lugt of the University of
Utrecht, while visiting this laboratory.
We thank Janette Hayes, Margaret Gibson and George Lanyon for their valuable technical
assistance.
REFERENCES
BORGHESE, E. (1959). The present state of research
BRUNS, G. P. & LONDON, I. M. (1965). The effect
on W.W. mice. Acta anat. 36,185-220.
of hemin on the synthesis of globin. Bio-
chem. biophys. Res. Commun. 18, 236-42.
K. (1956). A study of the conditions and mechanism of the diphenylamine reaction
for the colorimetric estimation of DNA. Biochem. J. 62, 315-18.
COLE, R. J. & PAUL, J. (1965). Properties of cultured preimplantation mouse and rabbit embryos
and cell strains derived from them. In Preimplantation Stages of Pregnancy (ed. S. E. W.
Wolstenholme). Ciba Foundation Symposia, p. 82. London: Churchill.
CRAIG, M. E. & RUSSELL, E. S. (1964). A developmental change in hemoglobins correlated
with an embryonic red cell population in the mouse. Dev. Biol. 10, 191-201.
FILAMINOVICZ, E. & GURNEY, C. W. (1961). Studies on erythropoiesis. XVI. Response
to a single dose of erythropoietin in the polycythemic mouse. /. Lab. Clin. Med. 57,
65-72.
FISHER, J. W., LAJTHA, L. G., BUTTOO, A. S. & PORTEUS, D. D. (1965). Direct effects of
erythropoietin on the bone marrow of the isolated perfused hind limb of rabbits. Brit. J.
Haemat. 11, 342-9.
FISHER, J. W., ROH, B. L., COUCH, C. & NIGHTINGALE, W. O. (1964). Influences of cobalt,
sheep erythropoietin and several hormones on erythropoiesis in bone marrows of isolated
perfused limbs of dogs. Blood, 23, 87-98.
FISHER, J. S., SCHOFIELD, R. & PORTEUS, D. D. (1965). Effects of renal hypoxia on erythropoietin production. Brit. J. Haemat. 11, 382-9.
GRUNEBERG, H. (1943). The development of some external features in mouse embryos.
/ . Hered. 34, 88-92.
HAMMEL, C. L. & BESSMAN, S. P. (1964). Hemoglobin synthesis in avian erythrocytes. / . biol.
Chem. 239, 2228-38.
HAMMEL, C. L. & BESSMAN, S. P. (1965). Control of hemoglobin synthesis by oxygen tension
in a cell free system. Arch. Biochem. Biophys. 110, 622-7.
KRANTZ, S. B., GALLIEN-LARTIGUE, O. & GOLDWASSER, E. (1963). The effect of erythropoietin upon heme synthesis by marrow cells in vitro. J. biol. Chem. 238, 4085-90.
BURTON,
260
.
R. J. COLE & J. PAUL
S. B. & GOLDWASSER, E. (1965). On the mechanism of erythropoietin-induced
differentiation. II. The effect on RNA synthesis. Biochim. biophys. Acta, 103, 325-32.
LAJTHA, L. G. (1964). Recent studies in erythroid differentiation and proliferation. Medicine,
43, 625-33.
LEVERE, R. D. & GRANICK, S. (1965). Control of hemoglobin synthesis in the cultured chick
blastoderm by £-aminolaevulinic acid synthetase: increase in the rate of hemoglobin formation with 5-aminolaevulinic acid. Proc. natn. Acad. Sci. U.S.A. 54, 134-7.
MCCULLOCH, E. A. & TILL, J. E. (1963). Repression of colony-forming ability of C57B1
hematopoietic cells transplanted into non-isologous hosts. /. cell comp. Physiol. 61, 301-8.
NEW, D. A. T. & STEIN, K. F. (1964). Cultivation of post implantation mouse and rat embryos on plasma clots. /. Embryol. exp. Morph. 12, 101-11.
O'BRIEN, B. R. A. (1961). Identification of haemoglobin by its catalase reaction with peroxides and o-dianisidine. Stain Technol. 36, 57-61.
PAUL, J. (1965). Cell and Tissue Culture, 3rd edn. Edinburgh: Livingstone.
RAMSAY, W. N. M. (1951). Iron metabolism and haemoglobin formation in the embryonated
hen egg. 2. Some observations on the embryonic blood supply. Biochem. J. 49, 494-9.
SNELL, G. D. (1956). Biology of the Laboratory Mouse. New York: Blakiston Co.
STAHLMAN, F., BRECHER, G. & MOORES, R. R. (1962). Humoral regulation of erythropoiesis.
VIII. The kinetics of red cell production and the effect of erythropoietin. In Erythropoiesis
(eds. O. Jacobson and J. Doyle), p. 162. New York: Grune and Stratton.
TAYLOR, R. B. (1965). Pluripotential cells in mouse embryo liver. Br. J. exp. Path. 46, 376-83.
TEALE, F. W. J. (1959). Cleavage of the haem-protein link by acid methyl ethyl ketone.
Biochim. biophys. Acta, 35, 543.
WAYMOUTH, C. (1959). Rapid proliferation of sublines of NCTC clone 929 (strain L) mouse
cells in a simple chemically defined medium (MB752/1). /. natn. Cancer Inst. 22, 1003-10.
WHITE, W. F., GURNEY, C. W., GOLDWASSER, E. & JACOBSON, L. O. (1960). Studies on erythropoietin. Recent Prog. Horm. Res. 16, 219-59.
WILT, F. (1965). Regulation of the initiation of chick embryo hemoglobin synthesis. /. Mol.
Biol. 12, 331-41.
WINTROBE, M. M. (1961). Clinical Haematology, 5th ed. p. 395. Philadelphia: Lea and
Febiger.
KRANTZ,
{Manuscript received 12 November 1965)