British Journal of Haematology, 1976, 33, 387.

Haemopoietic Progenitor Cells in Prenatal Congenitally Anaemic ‘Flexed-Tailed’

(fif> Mice

R. J. COLE AND T. REGAN Developmental Genetics Group, School ofBiologica1 Sciences, University of Sussex, Falmer, Brighton (Received 19 Septeniber 1975 ; acceptedfor publication 4 October 1975)

SUMMARY.The incidences of erythroid colony forming cells (CFU,) and granulocyte-macrophage colony forming cells (CFU,) have been measured in I 1-1 8 d prenatal livers of mice of genotypej’and nearly congenic +/ controls. In normal fetal livers numbers of CFU, (cells able to form colonies of 16 or more cells after 72 h in vitro) rise to a maximum on day 14 of gestation and represent c 1% of total fetal liver cells. InJ’fetal livers, peak values for numbers and proportions of CFU, are 50% of normal. Thej’lesion does not reduce the numbers of CFU, in fetal liver. Since this deficiency in CFU, parallels deficiencies of similar magnitude in spleen-colony forming units (CFU,) and erythroblasts in the liver, and erythrocytes in the blood, offlffetuses it is concluded that thefl’lesion is expressed at an early stage of haemopoietic development in prenatal life. The possibility that restricted haem synthesis is the primary effect of thefljgenotype and responsible for disturbances of both haemopoietic cellular proliferation and haenioglobin synthesis is examined.

+

Perturbations of blood cell production caused by mutant genes have proved useful in the analysis of normal haemopoietic regulation, particularly during prenatal life, where direct physiological intervention is difficult (Russell, 1970; Cole & Tarbutt, 1973 ; Cole, 1975). The severe anaemia of prenatal and neonatalfljmice is caused by a dual effect of this genotype (Gruneberg, 1942; Russell et al, 1968) expressed at the cellular level. Erythropoietic development of the early fetal liver is retarded, limiting the production and release of erythrocytes (Bateman et al, 1972; Bateman & Cole, 1972; Tarbutt & Cole, 1972). In addition, haemoglobin synthesis in fetal-liver derived reticulocytes is reduced (Cole et al, 1974) as a result of lowered activity of 8-aminolaevulinate synthetase and d-aminolaevulinate dehydratase (Cole et al, 1g75a). This restriction of haem synthesis with unimpaired cellular iron uptake (Cole et al, 1972) leads to the accumulation of siderotic granules which are characteristic of neonatal erythrocytes in this mutant. Recent observations that haem may be critically involved in the initiation of synthesis of a variety of non-globin proteins (Beuzard et al, 1970 ; Matthews et al, 1973 ; Raffel et al, 1974) and therefore also in the control of cell multiplication, may provide a unifying explanation for the apparent disparate cellular effects of the .fljgenotype (Cole et a!, 1g7sa). Correspondence: Dr R. J. Cole, Developmental Genetics Group, School of Biological Sciences, University of Sussex, Falmer, Brighton.

387

388

R. J. Cole and

T.Regan

Functional analyses of prenatal haemopoietic tissue have been extended by the development of in vitro clonal assays for granulocyte-macrophage progenitor cells (CFU,) (Bradley& Metcalf, 1966) and erythroid progenitor cells (CFU,) (Stephenson et al, 1971), allowing quantitation of cells belonging to populations intermediate between the multipotent stem cell and histologically recognizable ‘blast’ cells. In the present study the numbers of such progenitor cells in the prenatal liver o f j j m i c e have been compared with normal, in order to establish the stage of haemopoietic cell production at which the restrictive effects of the lesion first become apparent and to provide further data on the characteristics of CFU, in prenatal haemopoietic tissue.

f7f

MATERIALS AND METHODS

+

Animals. FL/I-RejjLvb/Luband FL/4-Re +/ Lub/Lub mice (Coleman et nl, 1969) were bred from stocks obtained from the Jackson Laboratory, Bar Harbor, U.S.A. Homozygous mutant and wild-type fetuses were obtained from timed natural matings, the morning on which mating plugs were found beginning day 0. Erythroid colonyforming cells. Fetal livers were disaggregated by repeated pipetting and the cells cultured in plasma clots according to the method of Stephenson et al(1971) in medium NCTC 109, supplemented with 10% fetal calf serum, 1% mouse serum, 0.02 mg/mlL-asparagine, 0.85% bovine embryo extract, 2% bovine serum albumin, and 10% citrated bovine plasma. I ml cultures were initiated with I-4x 105 cells in 35 mm ‘Nunc’ plastic Petri dishes and maintained for 3 d in 5% C 0 2 in air. Human urinary erythropoietin N.I.H. lot EP M 10Ta LSL (75.4 u/mg) was used at 0.6 units/ml. Colonies were characterized, after fixation and drying, by staining with Lepehne stain or by staining in sit# with 0.2% benzidine in 0.5% acetic acid +0.4% 30% H,O, (Cooper et a2, 1974), but counting was performed routinely without fixation or staining, at x roo magnification. Colonies containing more than 16 cells and of characteristic erythroid appearance were scored. Graniilocyte-macrophage colony forming cells. Cells were cultured in 3 5 mm diameter Petri dishes in 1.5 ml of modified supplemented Eagles M.E.M. containing 0.3% agar as described by Metcalf & Moore (1971). Colony Stimulating Factor was derived from medium conditioned by mouse strain LgzgS cells, grown in suspension to I x Io6/ml, added to a final concentration of 10%. This semi-solid agar medium was supplemented by the addition of 4% rat erythrocyte lysate (Bradley et a / , 1974) which greatly enhances colony growth and CFU, expression. Cultures were maintained at 37°C in 5% COP in air for 8 d in a freshly sterilized incubator, and scored at x 50 for the incidence of colonies of more than 40 cells. RESULTS

CFU, in Normal andff Fetal Liver The erythropoietin dose responsiveness of prenatal normal and J’jerythroid progenitor cells in terms of their ability to forni colonies in uitro is shown in Fig I. Up to 45% of the total colonies exceeding eight cells appear without addition of erythropoietin. Such ‘spontaneous’ colonies may represent cells responding to erythropoietic stimulatory factors in the serum and plasma, which are essential components of the culture medium, but most probably

Haemopoiesis in Prenataljf Mice

3 89

develop from relatively late cells already stimulated to differentiate while in the liver. A much smaller proportion of colonies of 16 or more cells forms ‘spontaneously’, usually less than I 5 % , and the erythropoietin-dependent formation of a colony of at least 16 cells in 72 h has been used to define the fetal liver CFU, described in this report. The erythropoietin sensitivity of prenatal J’CFU, is not distinguishable &.om normal. In normal FL/4-Re fetal livers the number of CFU, capable of forming colonies of 16 or more cells, rises to a maximum on day 14 of gestation and then declines sharply (Fig 2). On day 14 CFU, represent c 1% of total nucleated cells in the fetal liver, falling rapidly during day I S , and being essentially absent from the fetal liver during the last 3-4 days of gestation (Fig 3). Absolute numbers and proportions of CFU, rise more slowly inflffetal livers and peak values are c 50% of those achieved in normal mice (Figs 2 and 3). CFU, persist for longer in 100

-

00

-

u

2

60-

z2 -

.-

40-

$

20

-

0.001

0005

001

0.05

0.1

0.5

1

2

3

jfthan in normal livers. This is in accordance with earlier observations of extended sensitivity to erythropoietin ofj’fetal liver cells in vitro, and the presence of higher than normal proportions of erythroblasts in thejfliver late in gestation (Bateman et al, 1972; Tarbutt & Cole, 1972). Such differences are consistent with prenatal physiological compensation within fetuses for the deficit in tissue oxygenation caused by reductions in both erythrocytes and their haemoglobin content.

8’

CFU,in Normal a n d f f Fetal Livers and Adult Bone Marrow Changes in numbers of granulocyte-macrophage progenitors follow a similar pattern in both normal andjffetal livers, so production of such cells in the mutant fetuses is unimpaired (Fig 4). There is a rapid rise in number during the first 4-5 days of liver development,

R. J. Cole and T.Regan

390

I

10

Fetal age (days)

Fetal age (days)

FIG 2

FIG 3

FIG 2 . Numbers of CFU, (able to form colonies of 16 or more cells) in livers of prenatalJ”(.) and normal (0) mice and numbers of colonies formed from normal ( 0 )andJ”(w) livers without additional erythropoietin. FIG 3. Proportions of CFU. in livers of prenatalj’jand normal mice (symbols as Fig 2).

followed by a decline during days 15 and 16, with a second sharp increase in the period immediately preceding birth. The maximum proportion of granulocyte-macrophage progenitors occurs on days 12-13 when such cells represent approximately 0.2% of total liver cell numbers in both genotypes. In contrast to the extreme demand for erythrocytes during prenatal development, granulocyte production in the fetal liver is relatively low, and may be physiologically repressed in normal fetuses, relative to the potential expression of fetal CFU, in vitm (Moore & Williams, 1973). Relatively trivial compensatory multiplication of granulocytic precursors could therefore overcome the effects of restricted input into this cellular compartment. The incidence of CFU, is the same in the femoral bone marrows of adult malejjmice as in the wild type, and the responses of these cells to varying concentrations of colony stimulating factor are almost identical in terms of the minimum concentration of CSF eliciting colony formation, and the maximum response obtained (Fig 5 ) . DISCUSSION The cell type (CFU,) giving rise to colonies of erythroblasts after 2-3 d in culture appears

Huemopoiesis in Prenutalflf Mice

391

4c

1

P

3c 9

G X

v

L 0)

t 2c \ U

3

L

V

1c

F e t a l age (days)

FIG 4. Numbers of CFU, in normal (D)prenatal livers.

( 0 ) and-(!!(.)

and proportions of CFU, in normal ( 0 ) andflf

%conditioned medium

FIG 5 . The response of CFU. from adultJ'(.) factor derived from mouse L9zgS cells.

and normal ( 0 ) bone marrow to colony stimulating

to be a relatively differentiated precursor to histologically recognizable erythroblasts since the populations of CFU, in postnatal spleen and bone marrow are reduced when the level of circulating erythropoietin is lowered (Gregory et al, 1973). In prenatal mouse liver the maximum incidence of CFU, occurs 2-3 d after the maximum incidence of proerythroblasts, but their frequency broadly follows that of early erythroblasts (Cole et a!, 197jb, c). Stimulation of granulocyte formation by fetal liver cell suspensions in vitro does not reduce the number of erythroid colonies formed in the presence of erythropoietin (Stephenson et a2, 1971) confirniing the independence of these cell lineages in vitio.

R. I. Cole and T. Regan

392

The blood of unstressed adultj’mice contains normal concentrations of erythrocytes, and their bone marrow normal numbers of CFU, although the spleen content of CFU, is reduced. However, adult’j haemopoietic tissue is particularly deficient in its ability to give rise to transient endogenous erythropoietic spleen colonies (TE-CFU) in irradiated animals (Gregory et al, 1975). Compensatory erythropoiesis in phenylhydrazine treated adult jj” mice is also delayed (Coleman et al, 1969) and, although unstressedfj-adult haemopoietic tissue contains normal numbers of multipotent stem cells (CFU,), the growth rate off’spleen colonies in irradiated host-mice is lower than normal (Thompson et a f , 1966; Fowler et a f , 1967). Therefore, in adult mice the effect of the jflesion appears to be restricted to limiting enhanced proliferation of 200

-

? I

I I

180.

I

I

160

I

I I I I

~

- 140-

I

I

E“b 120-

I

I I

12

13

14 15 16 17 Fetal age (days)

18

19

FIG 6. Comparison of the course of erythropoiesisin prenatalfl’and normal mice. Haematological parameters observed inJj fetuses are expressed as a percentage of the values observed in normal fetuses on the same day of development, CFU, ( A ) ; CFU. (A); and erythroblasts in fetal liver (0); concentration of erythrocytes in blood (a); concentration of haemoglobin in blood (w). (Including data from Bateman et al, 1972; Bateman & Cole, 1972; Tarhutt & Cole, 1972.)

erythropoietic precursors in response to stress. Haemoglobin synthesis in circulating reticulocytes produced in such stress conditions is quantitatively normal (Cole et al, 1972). In contrast to the reported situation in adult j f mice, the present study shows a clear deficiency in both the absolute number, and relatiie frequency of CFU, in prenataljffetal liver, up to the fifteenth or sixteenth day of gestation. Previous studies (Bateman & Cole, 1972) have shown that the absolute number of CFU, injffetal livers is reduced during the same period, although the ratio of CFU, to total liver cells and the rate of increase of CFU, are normal. Therefore during prenatal life the lesion in production of erythroid cells caused by thej’genotype, is expressed in a cell type ancestral to CFU,, and perhaps at the level of the multipotent stem cell, or its transition to erythroid differentiation.

Haemopoiesis in PrenataljfMice

393

When the progress of erythropoiesis in prenatalj j m i c e is compared with normal (Fig 6) it is apparent that changes in the ratios of mutant to normal values relating to CFU,, CFU,, erythroblast numbers in the fetal liver, and erythrocyte concentrations in the peripheral blood follow broadly similar trends, providing further evidence that an early haemopoietic cell population is effected by thejjlesion. The greatest deficiency is seen early in development of the mutant fetal liver, and normal values are approached by the seventeenth or eighteenth day of development. The anaemia of prenataljjmice is complicated by a further deficiency in haemoglobin content caused by the production of hypochromic erythrocytes. In neonatal j j mice such erythrocytes contain only c 66% of the normal complement of haemoglobin. The primary cause of this lesion appears to be reduced activity of 8-aminolaevulinate synthetase and dehydratase, enzymes early in the haem synthetic pathway, which becomes critical after the cessation of transcription (Cole et al, 1975a). Circulating reticulocytes of prenataljjmice also show unbalanced globin synthesis, with a deficiency in D-chain production (Cole et al, 1974).However, although this lesion can be repaired in vitro by addition of haem it is possible that this disturbed globin synthesis is a consequence of disturbed proliferation, since human congenital dyserythropoietic anaemia is also associated with deficient &chain synthesis (Weatherall et al, 1973), while the disturbed globin synthesis in human sideroblastic anaemias and in lead poisoning (both associated with deficient haem synthesis) is expressed by reduced a-chain production (Piddington & White, 1974). It has been suggested (Gregory et al, 1975) that the effects of thejjlesion may be restricted to particular lineages of erythroid cells, which expand and differentiate only during erythropoietic stress. Our analysis of erythropoiesisin prenatalj'mice, and recent investigation of the general role of haem in the control of protein biosynthesis, indicate that a single lesion in porphyrin synthesis could be responsible for the range of disturbances seen in haemopoiesis and growth o f j j m i c e (Cole et al, 197sa). However, unless it proves possible to determine levels of haem synthesis in early haemopoietic progenitor cells the possibility that the primary effect of thejjlesion is on cell proliferation with secondary effects on haemoglobin synthesis cannot be excluded. ACKNOWLEDGMENTS

This work was supported by the Medical Research Council. Human urinary erythropoietin was supplied by the Committee on Erythropoietin of the U.S. National Heart and Lung Institute; it was procured by the Department of Physiology, University of the North-East, Corrientes, Argentina, and processed by the Hematology Research Laboratory, Childrens Hospital of Los Angeles. REFERENCES BATEMAN, A.E. & COLE,R.J. (1972)Colony forming cells in the livers of prenatal flexed ( 8 ' ) anaemic mice. Cell and Tissue Kinetics, 5 , 165. BATEMAN, A.E., COLE,R.J., REGAN, T. & TARBUTT, R.G. (1972)The role of erythropoietin in prenatal erythropoiesis of congenitally anaemic flexed-tailed (f()mice. Brifish]otrrnal ofHaematology, 22, 415.

BRADLEY, T.R. & WTCALF, D. (1966)The growth of mouse bone marrow cells in vitro. Australian Journal of Experimental Biology and Medical Science, 4, 287. BRADLEY, T.R., SUMNER, M.A. & McI"EY, E. (1974) Observations on enhancement of mouse marrow cell proliferation by erythrocyte extracts.

394

R.]. Cole and T.Regan

In:Haemopoiesis in Culture (ed. by W. A. Robinson). D.H.E.W. U.S.A. (NIH) 74-20s. BEUZARD, Y., RODVIEN, R. & LONDON,I.M. (1970) Effect of hemin on the synthesis of haemoglobin and other proteins in niainnialian cells. Proceedings ofthe National Academy of Sciences of the United States of America, 70, 1022. COLE,R.J. (1975) Regulatory functions of inicroenvironmental and hormonal factors in prenatal haemopoietic tissues. In: Early Mammalian Development (ed. by M. Balls and A. Wild). Cambridge University Press, London. COLE,R.J., GARLICK, J. & CHEEK,E.M. (1g75a) Activities of haem synthetic enzymes in blood cells of prenatal flexed-tailed (J’) anaemic mice. Journal .f Embryology and Experimental Morphology, 34, 373. COLE,R.J., GARLICK, J. & TARBUTT, R.G. (1974) Disturbed haem and globin synthesis in reticulocytes of prenatal flexed-tailed (flf) anaemic mice. Genetical Research, 23, 125. COLE, R.J., REGAN, T. & TARBUTT, R.G. (1972) Haemoglobin synthesis in reticulocytes of prenatal J” anaemic mice. British Journal of Haematology, 23, 443. COLE,R.J., REGAN,T., WHITE,S.L. & CHEBK,E.M. (1975~)The relationship between erythropoietin dependent cellular differentiation and colony forming ability in prenatal haemopoietic tissues. Journal of Embryology and Experimental Morphology, 34, 575. COLE,R.J., REGAN,T., WHITE,S.L. & CHEEK,E.M. (1975b) Expression of congenital defects in the haemopoietic microenvironment: erythroid and granulocyte-macrophage progenitor cells in prenatal ‘Steel’ (Slj/SP) anaemic mice. Cell and Tissue Kinetics, 8, 479. R.G. (1973) Kinetics of cell COLE,R.J. & TARBUTT, multiplication and differentiation during adult and prenatal haemopoiesis. In: The Cell Cycle in Development and D$erentiation fed. by M. Balls and F. s. Billett). Cambridge University Press, London. COLEMAN, D.L., RUSSELL, E.S. & LEVIN,E.Y. (1969) Enzymatic studies of the hemopoietic defect in flexed mice. Genetics, 61, 63 I. COOPER, M.C., LEVY,J., CANTOR, L.N., MARKS,P.A. & RIFKINU, R.A. (1974) The effect of erythropoietin on colonial growth of erythroid precursor cells in vitro. Proceedings of the National Academy of Sciences ofthe United States ofAmerica, 71,1677. FOWLER,J.H., TILL,J.E., MCCULLOCH,E.A. & SIMINOVITCH, L. (1967) The cellular basis for the defect in haemopoiesis in flexed-tailed mice. 11. The specificity of the defect for erythropoiesis. British Journal ofHaematology, 13,256. C.J., MCCULLOCH, E.A. &TILL,J.E. (1973) GREGORY,

Erythropoietic progenitors capable of coIony formation in culture: state of differentiation. Journal of Cellular Physiology, 81, 411. GREGORY, C.J., MCCULLOCH, E.A. & TILL,J.E. (1975) The cellular basis for the defect in haemopoiesis in flexed-tailed mice. 111. Restriction of the defect to erythropoietic progenitors capable of transient colony formation in vivo. BritishJournal of Haematology, 30. 401. GRUNEBERG, H. (1942) The anaemia of flexed-tailed mice. 11. Siderocytes.Journal of Genetics, 4, 246. MATTHEWS, M.B., HUNT, T. & BRAYLEY, A. (1973) Specificity of the control of protein synthesis by haemin. Nature: N e w Biology, 243, 230. METCALP, D. & MOORE,M.A.S. (1972) Haemopoietic Cells. North Holland Publishing Co., Amsterdam. MOORE,M.A.S. & WILLIAMS, N. (1973) Analysis of proliferation and differentiation of foetal granulocyte-macrophage progenitor cells in haemopoietic tissue. Cell and Tissue Kinetics, 6, 461. PIDDINGTON, S.K. & WHITE, J.M. (1974) The effect of lead on total globin and ct- and &chain synthesis; in vitro and in vivo. British Journal ofHaematology, 27, 415. P. (1974) Role for RAFFEL,C., STEIN,S. & KAEMPFER, heme in mammalian protein synthesis: activation of an initiation factor. Proceedings of the National Academy ofsciences of the United States ofAmerica, 71,4020. RUSSELL, E.S. (1970) Abnormalities of erythropoiesis associated with mutant genes in mice. Regulation of Hematopoiesis (ed. by A. S. Gordon), Vol. I, pp 649-975. Appleton-Century-Crofts, New York. RUSSRLL, E.S., THOMPSON, M.W. & MCFARLAND, E.C. (1968) Analysis of the effects of Wandf genic substitutions on fetal mouse hematology. Genetics, 58, 2.59-

STEPHENSON, J.R., AXBLRAD, A.A., MCLBOD,D.L. & SHRJWE, M.M. (1971) Induction of colonies of hemoglobin-synthesizing cells by erythropoietin in vitro. Proreedings of the National Academy of Sciences of the United States ofAmerica, 68, 1542. TARBUTT, R.G. & COLE,R.J. (1972) Foetal erythropoiesis in genetically anaemic, flexed-tailed (J’) mice. Cell and Tissue Kinetics, 5,491. THOMPSON, M.W., MCCULLOCH, E.A., SIMINOVITCH, L. & TILL,J.E. (1966) The cellular basis for the defect in haemopoiesis in flexed-tailed mice. I. The nature and persistence of the defect. BritishJournal of Haematology, 12, 152. WEATHERALL, D.J., CLEGG,J.B., KNOX-MCCAULAY, H.H.M., BUNCH,C., HOPKINS, C.8. & TEMPERLEY, I.J. (1973) A genetically determined disorder with features both of thalassaemia and congenital diserythropoietic anaemia. British Journal offlaematology, 24, 681.