/ . Embryol. exp. Morph. Vol. 58, pp. 143-155,1980
Printed in Great Britain © Company of Biologists Limited 1980
143
The ontogeny of erythropoiesis in the
mouse detected by the erythroid colony-forming
technique
II. Transition in erythropoietin sensitivity during development
By IVAN N. RICH 1 AND BERNHARD KUBANEK
From the Department of Internal Medicine and Paediatrics, Division of
Haematology, University of Ulm, West Germany
SUMMARY
The erythroid colony-forming technique has been used to study the erythropoietin (Ep)
sensitivity of CFU-E during ontogenesis. Evidence is provided which indicates that an agerelated decreasing Ep sensitivity of CFU-E occurs as the animal develops. In addition, an
intermediate Ep-sensitive CFU-E population represented by the 2-day-old neonate is present
during the transition from foetus to adult. The Ep sensitivity of the foetal, neonatal and adult
CFU-E populations is shown to be an intrinsic property of these populations.
INTRODUCTION
In an earlier report (Rich & Kubanek, 1976) a difference in hepatic and adult
erythroid colony-forming units (CFU-E) was described with respect to erythropoietin (Ep) requirement. It was postulated that the change from foetal to adult
erythropoiesis might be accompanied by a transitional Ep-sensitive phase in
the neonatal animal.
As a first stage in understanding the role of Ep during the ontogeny of the
mouse, the CFU-E technique was used to study hepatic erythropoiesis (Rich &
Kubanek, 1979). It was shown that during this period the Ep sensitivity of the
CFU-E population remains constant throughout the second half of gestation,
indicating that although CFU-E numbers change, Ep sensitivity appears to be
an intrinsic property of the cell.
In this communication the EP sensitivity of the CFU-E populations as the
animal develops is described in detail. It will be shown that the major transitional phase from foetus to adult is accompanied by an intermediate Ep sensitivity of the CFU-E. In addition, perturbation experiments indicate that the
Ep sensitivity, although a relative phenomenon, is an intrinsic property of a
particular CFU-E population.
1
Author's address: Ivan N. Rich, University of Ulm, Zentrum fur Klinische Grundlagenforschung, Parkstrasse 11, D-7900 Ulm/Donau, West Germany (B.R.D.)
144
I. N. RICH AND B. KUBANEK
METHODS AND MATERIALS
Animals
Inbred CBA/Ca mice were used for these studies. Foetuses were obtained as
described by Rich & Kubanek (1979). For experiments involving neonatal
animals, all pregnant mice from one batch were put into a single cage and left
to give birth under normal husbandry conditions. It was found that by this
procedure the incidence of one of the mothers killing her young was considerably reduced. Neonates of 2, 8 and 15 days of age were used. Adult female mice
were employed at an age of 8-12 weeks.
Preparation of suspensions
Preparation of foetal liver and adult bone marrow and spleen cell suspensions
have been described elsewhere (Rich & Kubanek, 1979). However, femora from
2-day neonates as well as all spleens were homogenized using a loose-fitting
glass homogenizer. A single cell suspension of marrow from 8- and 15-day
neonatal femora was produced by flushing out the marrow with medium using
a 22- or 25-gauge syringe needle.
Erythropoietin (Ep)
Two types of Ep were employed, namely that obtained from anaemic sheep
plasma supplied commercially by Connaught Laboratories (Ontario, Canada),
the Step III preparation, and that from human urine. The latter had a specific
activity of 140 units/mg of protein and was prepared as described by Iscove,
Sieber & Winterhalter (1974).
Erythroid colony-forming technique
A modified methyl cellulose method was used (Iscove et al. 1974). This
modified method has been described elsewhere together with the criteria for
CFU-E and BFU-E evaluation (Rich & Kubanek, 1979). For velocity sedimentation studies, however, the technique was scaled down by half and 0-5 ml
of the mixed components were plated in Replica Dishes (Sterilin, England;
Greiner, West Germany).
Irradiation of cell suspension
Two ml of the original foetal liver or adult bone-marrow cell suspensions
were dispensed into plastic tubes and irradiated using a Stabilipan gammairradiator and Duplex diosmeter at 12 mA and 250 kW with a Cu filter. A distance of 45 cm separated the source from the cells, the latter receiving 24-5 rad/
min. After irradiation the suspensions were thoroughly mixed, diluted to the
plating dilution and added to the other, previously mixed culture components
as quickly as possible to avoid changes in physical conditions.
CFU-E erythropoietin sensitivity changes
145
Velocity sedimentation procedure
The Staput method of Miller & Phillips (1969) modified using a foetal calf
serum gradient of 5%, 15% and 30% in Hanks balanced salt solution (BSS)
was used. Cell suspensions were prepared in Hanks BSS containing 3 % foetal
calf serum and 2% L-glutamine. A concentration of 112'5 x 106 cells contained
in 20-25 ml was loaded into an 11-5 cm diameter glass chamber (Glassbauerei,
Weil am Rhein, W. Germany). After 3-5 h the cone volume was discarded and
20 fractions each of 30 ml were collected into 50 ml plastic tubes (Falcon
Plastics, Becton-Dickenson, W. Germany). The tubes were centrifuged at
1200 rev/min for 10 min at 4 °C, after which the supernatant was withdrawn
and the cells resuspended in 1 ml Hanks BSS. The cell count of each fraction
was determined using a Coulter Counter.
Statistical analysis of results
All Ep-dose-response values were normalized as follows. An arbitrary Ep
dose was chosen and the mean CFU-E value for this dose calculated for all dose
responses for a particular tissue. Using this value, the individual values for this
particular Ep dose were recalculated and the difference factor obtained. The
colony values for the complete dose response were multiplied by the factor
specific for that dose response. Linear regressions were then calculated by the
method of least squares. Polynomial regressions were also calculated using this
method. The probability of association was obtained from the correlation coefficient (r) (Bailey, 1959). Testing the significance of differences between
dose-response parameters was performed by regression analysis, whereby
regression line intercepts and regression coefficients were compared.
RESULTS
Transition in erythropoietin sensitivity from foetus to adult
(a) Using Step III erythropoietin
To investigate whether a transitional Ep-sensitive CFU-E population existed
during postnatal development, Ep-dose responses for 2-, 8- and 15-day neonatal
bone marrow and spleen were performed. The results were then compared with
Ep-dose-response curves for CFU-E 14-day foetal liver and adult bone marrow
and spleen by regression analysis. It was found that both 2-day neonatal bone
marrow and spleen exhibited an Ep-dose-response curve which could be interspersed between that for foetal liver and adult bone marrow or spleen. Eightand 15-day neonatal organs exhibited adult CFU-E Ep sensitivity. Figure 1
shows the parallel displacement to a less sensitive CFU-E population as erythropoiesis passes from the hepatic stage at 14 days of gestation to the adult phase
seen in the bone marrow. Similar parallel displaced curves are found for 14-day
foetal liver and 2-day neonatal and adult spleen, but are not shown here. These
146
I. N. RICH AND B. KUBANEK
100 -|
•—•
14-day foetal liver
•A—A 2-day neonatal bone marrow
o
B — • Adult bone marrow
!§• so ^
.s
I
e
70 60 5040 30-
u
D
u
20 H
10 H
0025
0075
'
00125
005 01
Erythropoietin dose (U/ml)
000625
0-2
0-4
0-3
Fig. 1. Dose-response curves for 14-day foetal liver, 2-day neonatal bone marrow
and adult bone marrow CFU-E plotted as the percent response from the maximum
stimulating Ep dose as a function of log Ep Step III concentration.
14-Day foetal liver
2-Day neonatal bone marrow
Adult bone marrow
(y =
r=
O=
r=
(y=
r=
166-5 + 60-9(x);
0-95; P < 0001)
143-7 + 57-6(x);
0-96; P < 0001)
111-9 + 57-OOt);
0-93; P < 0001)
results were obtained using the Step III Ep preparation. Since at higher than
optimal Ep doses a decrease is seen due to impurities in the preparation, linear
regressions have been employed and the CFU-E values evaluated as percent
from the maximum stimulated Ep concentration. The regression coefficients for
all these dose-response curves indicate that the slopes of the linear regressions
are very similar. This is supported by comparing the regresson coefficients;
the dose-response curves are parallel to each other (P < 0-05). In addition,
further regression analysis indicates that each parallel dose response is, within
95 % confidence limits, horizontally displaced from the others. Thus, the CFU-E
giving rise to each of the Ep-dose-response curves represents a different population differing in their 50% Ep values by 0-012 U/ml for 14-day foetal 0-024
U/ml for 2-day neonatal bone marrow and 0-082 U/ml for adult bone marrow.
CFU-E erythropoietin sensitivity changes
147
120 -i
110100-
^
. . ' / » ' ..••••'
A
A
A
9080w
fc
E
o
70 H
Y
/
60 H
/
50-
/ /+
40^
/
ii
•
/
//
/
•
/
/
*
/
/
/
*
•*
K 14-day foetal liver
••
-• 2-day neonatal bone marrow
<*•
-A Adult bone marrow
30 20-
/t
/
100000156 ' 000625 '
0025
' 0075'
0-2 ' 0-4 '0-6 ' 10
000313
00125
005 01
0-3 0-5 0-8
Erythropoietin dose (U/ml.)
' 3-0 ' 6-0 '
2-0
40
80
Fig. 2. Dose-response curves for 14-day foetal liver, 2-day neonatal bone marrow
and adult bone marrow CFU-E plotted as the percent response from the maximum
stimulating Ep dose as a function of log purified human urinary Ep concentrations.
If linear regressions instead of polynomial regressions are calculated the following
dose response parameters between the given Ep concentrations are obtained.
14-Day foetal liver (from 0-00313
(y = 167-8 + 63-3fx);
to 01 U/ml)
r = 0-98; P < 0001)
2-Day neonatal bone marrow (from
(y= 141 -7 + 67-6 (x);
0125 to 0-1 U/ml)
r = 0-91; P= <0-05)
Adult bone marrow (from 0-025 U/ml to
(y = 115-2 + 66-l(x);
0-5 U/ml)
r = 0-97; P < 0001)
(b) Using human urinary erythropoietin
Since the above results could be caused, in part, by the degree of purity of
the Step III Ep preparation, a similar set of experiments were performed using
Ep purified from human urine which produced a plateau in the dose-response
curve at high Ep concentrations. The normalized results, shown in Fig. 2, are
again plotted as the percentage from the maximum stimulating Ep concentration; in this case, the mean plateau value. The results are plotted as. polynomial
regressions. The calculated 50% Ep values from these curves are 0-014, 0-043
and 0-1 U/ml for 14-day foetal liver, 2-day neonatal bone marrow and adult
bone marrow respectively. If, however, only the linear portions of the regressions are considered, the 50% Ep values are 0-014 U/ml for 14-day foetal liver,
0-048 U/ml for 2-day neonatal bone marrow and 0-1 U/ml for adult bone
148
I. N. RICH AND B. KUBANEK
Table 1. Change in concentration and absolute CFU-E of splenic erythropoiesis
during the development of the mouse
17-Day 19-Day 2-Day
prenatal prenatal neonatal
spleen spleen
spleen
No. of experiments
CFU-E/105 cells
CFU-E/organ
2
25
46
2
43
257
4
477
10116
8-Day
neonatal
spleen
4
326
194207
15-Day
neonatal
spleen
Adult
spleen
2
68
34416
4
9
10738
Table 2. Change in concentration, absolute and total CFU-E ofmyeloid erythropoiesis during the development of the mouse
2-Day
neonatal
bone
marrow
8-Day
neonatal
bone
marrow
No. of experiments
4
4
CFU-E/105 cells
258
159
2035
CFU-E/organ
782
CFU-E/animal* 0- 13 x 105 0-34 x 105
15-Day
neonatal
bone
marrow
Adult
bone
marrow
2
4
340
759
17561
80302
2-92 x 105 13-38 xlO 5
* Calculated from one femur being 6% of the total bone marrow (see text).
marrow. Comparison of regression coefficients in this case shows that the doseresponse curves are parallel (P < 0-05) and they are significantly horizontally
displaced from each other (P < 0-05). Thus, as the animal matures, different
and transient CFU-E populations occur exhibiting decreased Ep sensitivity.
Transition in erythropoiesis
Tables 1 and 2 show CFU-E concentrations and absolute CFU-E values for
pre-, neonatal and adult spleen and neonatal and adult bone marrow. Maximum splenic CFU-E values are obtained at about the 8th day of life, thereafter
declining to a constant but low level. Myeloid CFU-E increases continuously
throughout the neonatal period and reaches maximum numbers in adult life.
The number of CFU-E/animal have been calculated by assuming that one femur
is approximately 6% of the total bone marrow (Smith & Clayton, 1970).
Effect of perturbation on the erythropoietin-dose-response curve
A set of experiments were performed in order to test whether mixing different
cell suspensions would affect the Ep sensitivity. The experiments were carried
out as follows: 1 x 105 irradiated (850rad) adult bone marrow cells per ml
were mixed with 1 x 105 normal 14-day foetal liver cells per ml, and the Epdose-response curve so obtained was compared to the Ep-dose-response curve
for 0-5 x 105 normal 14-day foetal liver cells per ml (the usual cell dose plated).
CFU-E erythropoietin sensitivity changes
149
no-.
100-
9080-
n.
70-
W
60Normal 14-day foetal
Liver + 850 rad adult
bone marrow
50^ 4 0 -
Normal adult bone
marrow + 850 rad.
14-day foetal liver
3020100-
1
'
1
'
'
000313 I 0-0125 '
005 ' 0-1
000156
000625
0025
0075
Erythropoietin dose (U/ml)
I 7,1,
' 0-3 ' 0-5
0-2 0-4
Fig. 3. Dose-response curve for normal 14-day foetal liver cells mixed with irradiated adult bone marrow cells 0> = 142-6 + 53-2(x); r = 0-95; P<0001) and normal
adult bone marrow cells mixed with irradiated cells (y = 115-8 + 62-6(x);r = 0-97;
P < 0001) as a function of log Ep concentration. Shaded area represents the doseresponse curve for normal 14-day foetal liver and adult bone marrow cells (see Fig. 1).
110r
100
S
90
3
I
80
x
&
A+
x Adult bone marrow
-a 13-day foetal liver
A 14-day foetal liver
+ 15-day foetal liver
70
60
50
a
W
D
&
40
30
20
10
1
2
3
4
5
6
7
8
9
10 11
Sedimentation velocity (mm/h)
12
13
14
15
Fig. 4. Velocity sedimentation profiles for adult bone marrow and 13-, 14- and 15day foetal liver CFU-E.
16
150
I. N. RICH AND B. KUBANEK
110r
100 90
&
A
+
•5 80
•x Adult bone marrow
B 13-day foetal liver
A 14-day foetal liver
+ 15-day foetal liver
70
60
50
40
30
20
10
0
1
2
3
4
5
6
7
8
9
10
11
Sedimentation velocity (mm/h)
12
13
14
15
16
Fig. 5. Velocity sedimentation profiles for adult bone marrow and 13-, 14- and
15-day foetal liver BFU-E.
The reverse experiment was also performed in which 4 x 105 irradiated (850 rad)
14-day foetal liver cells per ml were mixed with 4 x 105 normal adult bone
marrow cells per ml. The resulting Ep-dose response was compared to that for
normal adult bone marrow cells plated at a cell concentration of 2 x 105/ml.
The results are shown in Fig. 3. There is no significant (P < 0-05) change in the
regression coefficients for the dose response obtained using the mixed cell
suspension compared with the normal cell suspensions. It should, however, be
noted that at all Ep concentrations an increase in CFU-E values was obtained.
Figures 4 and 5 show the velocity sedimentation profiles for CFU-E and
BFU-E for adult bone marrow and 13-, 14- and 15-day foetal liver. There is a
distinct difference in the sedimentation velocities between adult bone marrow
and foetal liver for both CFU-E and BFU-E. Single or pooled fractions from
three different positions of the velocity sedimentation profiles for adult bone
marrow or 14-day foetal liver were tested for their retention of Ep sensitivity
and compared with unfractionated bone marrow or foetal liver CFU-E in Fig. 6.
Points are the mean values of three experiments. The Ep-dose responses for the
unfractionated cell suspensions are essentially parallel. In addition, there is
little or no variation between separated cell suspensions or the unfractionated
cell suspension.
DISCUSSION
The results presented here lead to the conclusion that as the animal develops
from the foetus to the adult, CFU-E populations occur which show a specific
CFU-E erythropoietin sensitivity changes
,.*.--...
•§ 100 H
D.
151
.
w
"•a
80-
601 = 14-day foetal liver (y = 175 + 70-5*)
2 = 12-57 mm/h(> = 182-4 +86-lx)
3 = 9-86 mm/h (y = 194-8 + 87-5x)
4 = 7.26 mm/h O = 1991 + 84-7x)
5 = Adult bone marrow O = 133-8 + 87-9x)
40r;
20u
+
6 = 11-97 mm/h O = 142-3 + 101-8A:)
.;
-j
0
i
00125
i
i
I
I
0025
005 0075 01
Erythropoietin dose (U/ml)
= 7.92 mm/h O = 131-7 + 83-lx)
8 = 5-77 mm/h (y = 145-3 + 104-7*)
7
i
i
0-2
0-4
Fig. 6. Dose-response curves for CFU-E as a function of log Ep concentration
performed on velocity sedimentation fractions from separated 14-day foetal liver and
adult marrow cells. Three fractions from each separated haemopoietic tissue were
taken corresponding to the sedimentation velocity (mm/h) given in the diagram and
compared with CFU-E from unfractionated foetal liver or adult bone marrow. The
50 % Ep values are as follows:
tim/h
U/ml
Unfractionated 14-day foetal liver (1)
0017
Fractionated 14-day foetal liver
12-57
0029
9-86
0022
7-26
0017
—
Unfractionated adult bone marrow (5)
0111
Fractionated adult bone marrow
11-97
015
7-92
0104
5-77
0123
and intrinsic erythropoietin (Ep) responsiveness. This hypothesis has been
postulated previously (Rich & Kubanek, 1976) and results have been summarized
elsewhere (Kubanek, Heit & Rich, 1978; Rich, Heit & Kubanek, 1978, Rich
1978).
The ontogeny of haemopoiesis and erythropoiesis in particular is a process of
orderly and sequential transitions, initiated in the yolk sac. Migration of
haemopoietic stem cells to the liver initiates haemopoiesis in this organ (Johnson & Moore, 1975). Later, erythropoiesis occurs in the spleen and finally in
the bone marrow (Tables 1, 2).
Parallel to the sequential changes of haemopoiesis from the foetal liver to the
bone marrow is an age-related decreasing Ep sensitivity of CFU-E. Concentrations and CFU-E values per organ during this developmental period are
fluctuating continually. Thus for comparison of Ep-dose-response curves results
had to be normalized and presented as the percentage of the maximal stimulating Ep concentration.
152
I. N. RICH AND B. KUBANEK
The Ep-dose-response curves are represented in a similar fashion to that used
by Camiscolli & Gordon (1971) for the polycythaemic mouse bioassay and by
de Klerk, Hart, Kruiswijk & Goudsmit (1978) for the in vitro bioassay of
serum Ep levels. Recently, Haga & Falkanger (1979) have described the in vitro
assay of Ep using newborn mouse livers less than 24 h old as a source of CFU-E.
Figure 2 shows that early foetal liver CFU-E also provide an extremely sensitive,
easily reproducible in vitro bioassay for different Ep preparations using either
fresh or cryopreserved cells (unpublished data). By keeping the Ep preparation
constant and assuming that Ep acts by the same mechanism in all these cell
populations, changing the target cell provides a series of parallel curves from
which the displacement between the linear sections of the dose response provides an estimation of the Ep sensitivity of one cell population compared to
another. Erythropoietin sensitivity in this case is a relative phenomenon since
it may change under different in vitro conditions and with different strains of
mice. Thus 14-day foetal liver CFU-E are more sensitive to Ep than adult bone
marrow CFU-E. In addition, the change from foetal to adult erythropoiesis is
not only accompanied by a change in organ sites but also by a transition in
Ep sensitivity as shown in the example of 2-day neonatal bone marrow. This
phenomenon is not dependent on a particular Ep preparation but is equally
expressed using the commercially prepared Step III or a more purified human
urinary Ep. Therefore, during ontogenesis, different Ep-sensitive CFU-E
populations exist. However, the possibility that this is due to differences in the
age structure of the precursor compartment at different times seems to be
excluded since colonies from foetal and adult erythropoietic tissue are morphologically identical (Rich, 1978). Thus the Ep sensitivity of the CFU-E is an
inherent property which changes during development.
That the Ep sensitivity is a stable and intrinsic property of a particular CFU-E
population has been shown under several different conditions.
During pregnancy, splenic erythropoiesis is greatly stimulated with a 40-fold
increase in CFU-E. This natural perturbation did not change the Ep sensitivity
of adult CFU-E at any time point during this stimulated period (Rich & Kubanek,
1979). It can be concluded that the Ep sensitivity of adult CFU-E remains stable
during a period of considerable stimulation and changing composition of the
erythropoietic tissue. In 1974, Gregory, Tepperman, McCulloch & Till demonstrated that under extreme perturbation the Ep sensitivity of the CFU-E
population was displaced in both the bone marrow and spleen. These results are
in apparent contradiction to those presented here. One reason for this
discrepancy could be the type of perturbation used. However, subjecting mice
to extreme haemorrhage did not shift the Ep-dose-response curve (unpublished
data). Therefore the difference in results still remains unclear.
Retention of parallelism without displacement of the Ep-dose-response
curves was found when living and lethally irradiated haemopoietic cell suspensions of different Ep sensitivities were mixed together. This again shows that
CFU-E erythropoietin sensitivity changes
153
Ep responsiveness is an intrinsic property of the cells. Since a greater number of
CFU-E are stimulated at all Ep concentrations in these mixed suspensions, this
implies that radiation may cause the release of a factor into the medium capable
of stimulating CFU-E. This could be similar to the erythropoietic stimulating
factor released when macrophages are treated with silica. This factor is similar,
if not identical, to Ep as shown by its effect in polycythaemic and bled mice and
its dose-dependent stimulation in the polycythaemic mouse bioassay (Rich,
Heit & Kubanek 1980). These observations lend further support to that proposed by Gruber, Zucali & Mirand (1977), namely that macrophages are
producing and/or storing Ep.
Velocity sedimentation studies indicate that CFU-E from 13- or 14- or 15-day
foetal liver are a larger cell than that from adult bone marrow. The CFU-E from
various fractions of the velocity sedimentation profiles from either 14-day foetal
liver or adult bone marrow show a retention of their specific Ep sensitivity and
are not significantly different to the respective unfractionated cell suspension.
This implies that the whole foetal liver or adult bone marrow CFU-E population is homogeneous with respect to their Ep sensitivity and the difference
is not due to a different composition of the erythroid precursor compartment.
The velocity sedimentation profile of BFU-E derived from foetal liver also
shows a shift to a larger size when compared to adult bone marrow BFU-E.
This is similar to that found by Johnson & Metcalf (1978) for GM-CFC. However, both adult bone marrow and foetal liver BFU-E require the same high Ep
concentrations for growth in order to achieve maximum stimulation and show
similar colony morphology, even though their velocity sedimentation profiles
are different. This observation would imply that BFU-E's may only require
Ep for differentiation.
Our results have shown that foetal liver CFU-E are more sensitive to Ep than
adult bone marrow CFU-E and that during the transition from the foetus to
the adult, CFU-E populations occur (in both femur and spleen) which exhibit
a transitional Ep sensitivity. In addition, the Ep sensitivity shown in vitro by
a specific CFU-E population is an intrinsic property of that population. Earlier
work by Jacobson, Marks & Gaston (1959) and Lucarelli et al. (1968) working
with foetal erythropoietic tissue, as well as Lucarelli, Howard & Stohlman
(1966) and Carmena, Howard & Stohlman (1968) employing the neonate,
concluded that either regulatory mechanisms operating in the adult were different from those in the foetus or early neonatal life, or that foetal erythropoiesis
was regulated autonomously. However, Bleiberg & Feldman (1969) also speculated that foetal liver cells, although being Ep-dependent, are regulated by very
low levels of Ep. Other experiments by Rencricca, Howard, Kubanek & Stohlman (1976) showed that the repopulating capacity of foetal liver cells was
greater than 10-day neonatal bone marrow which in turn was greater than adult
bone marrow. This shows that the same phenomenon can be obtained under
154
I. N. R I C H AND B. KUBANEK
different conditions and supports the proposition of transitional Ep-sensitive
C F U - E populations.
This work was supported by the Deutsche Forschungsgemeinschaft SFB 112/Project A2
and the Volkswagen Foundation.
REFERENCES
BAILEY, N. I. S (1959). Statistical Methods in Biology.
BLEIBERG, I & FELDMAN, M. (1969). On the regulation
English Universities Press.
of hemopoietic spleen colonies produced by embryonic and adult cells. Devi Biol. 19, 566-580.
CAMISCOLLI, J. F. & GORDON, A. S. (1971). Bioassay and standardization of erythropoietin.
From Regulation ofHematopoiesis, vol. 1 (ed. A. S. Gordon), pp. 369-393. N.Y.: Appleton
Century-Crofts.
CARMENA, A. O., HOWARD, D. & STOHLMAN, F. JR. (1968) Regulation of erythropoiesis.
XXII. Erythropoietin production in the new-born animal. Blood 32, 376-382.
DE KLERK, G., HART, A. A. M., KRUISWIJK, C , & GOUDXMIT, R. (1978). Modified method
of erythropoietin (ESF) bioassay in vitro using mouse fetal liver cells. Measurement of ESF
in human serum. Blood 52, 569-577.
GREGORY, C. J., TEPPERMAN, A. D., MCCULLOCH, E. A. & TILL, J. E. (1974). Erythropoietic
progenitors capable of colony formation in culture: Response of normal and genetically
anemic W/Wv mice to manipulations of the erythron. / . cell. Physiol. 84, 1-12.
GRUBER, D. F., ZUCALI, J. R. & MIRAND, E. A. (1977). Identification of erythropoietin pro-
ducing cells in fetal mouse liver culture. Expl Hematol. 5, 392-398.
P. FALKANGER, B. (1979). In vitro assay for erythropoietin: Erythroid colony formation in methyl cellulose used for the measurement of erythropoietin in plasma. Blood 53,
1172-1181.
ISCOVE, N. N., SIEBER, F. & WINTERHALTEP., K. H. (1974). Erythroid colony formation in
cultures of mouse and human bone marrow: Analysis of the requirement for erythropoietin
by gel filtration and affinity chromatography on Agarose-Concanavalin A. J. Cell Physiol.
83, 309-320.
JACOBSON, L. O., MARKS, E. K. & GASTON, E. A. (1959). Studies on erythropoiesis. XII. The
effect of transfusion-induced polycythaemia in the mother on the fetus. Blood 14, 644-652.
JOHNSON, G. R. & METCALF, D. (1978). Characterization of mouse fetal liver granulocytemacrophage colony forming cells using velocity sedimentation. Expl Hematol. 6, 246-256.
JOHNSON, G. R. & MOORE, M. A. S. (1975). Role of stem cell migration in initiation of mouse
fetal liver haemopoiesis. Nature, Lond. 258, 726-728.
KUBANEK, B., HEIT, W. & RICH, I. N. (1978). Change in erythropoietin sensitivity of CFUe
during the ontogeny of the mouse. / . Supramolecular Struct. (Abstract).
LUCARELLI, G., HOWARD, D. & STOHLMAN, F. JR. (1966). Regulation of erythropoiesis. XV.
Neonatal erythropoiesis and the effects of nephrectomy. / . Clin. Invest. 43, 2195-2203.
LUCARELLI, G., PORCELLINI, A., CARNEVALI, C , CARMENA, A. & STOHLMAN, F. JR. (1968).
Fetal and neonatal erythropoiesis. Ann. N.Y. Acad. Sci. 149, 544-559.
MILLER, R. G. & PHILLIPS, R. A. (1969). Separation of cells by velocity sedimentation.
/ . cell Physiol. 73 191-202.
RENCRICCA, N. J., HOWARD, D., KUBANEK, B., & STOHLMAN, F. JR. (1976). Erythroid differentiation of fetal, neonatal and adult hemopoietic stem cells. Scand. J. Hematol. 16,
189-195.
RICH, I. N. (1978). Developmental changes in murine erythropoiesis detected by the erythroid
colony-forming technique. Doctoral Thesis.
RICH, I. N., HEIT, W. & KUBANEK, B. (1978). Regulation of fetal and adult erythropoiesis in
. mouse and man. Expl Hematol. 6, Supp. 3, 59.
RICH, I. N., HEIT, W. & KUBANEK, B. (1980). An erythropoietic factor similar to erythropoietin released by macrophages after treatment with silica. Blut (in the Press).
RICH, I. N. & KUBANEK, B. (1976). Erythroid colony formation (CFUe) in fetal liver and
adult bone marrow and spleen from the mouse. Blut 33,171-180.
HAGA,
CFU-E erythropoietin sensitivity changes
155
I. N. & KUBANEK, B. (1979). The ontogeny of erythropoiesis in the mouse detected by
the erythroid colony-forming technique. I. Hepatic and maternal erythropoiesis. /. Embryo),
exp. Morph. 50, 57-74.
59
SMITH, L. H. & CLAYTON, M L. (1970). Distribution of injected Fe in mice. Expl Hematol.
20, 82-86.
RICH,
{Received 30 October 1979, revised 7 February 1980)
EMB 58