Stimulation of fetal hemoglobin synthesis in bone marrow cultures

Transcription

1 Proc. Natl. Acad. Sci. USA Vol. 73, No. 6, pp , June 1976 Cell Biology Stiulation of fetal heoglobin synthesis in bone arrow cultures fro adult individuals (huan erythroblasts/erythroid colonies/fetal heoglobin regulation/heoglobinopathies) TH. PAPAYANNOPOULOU, M. BRCE, AND G. STAMATOYANNOPOULOS Divisions of Heatology and Medical Genetics, Departent of Medicine, University of Washington School of Medicine, Seattle, Washington Counicated by Cleent A. Finch, March 22, 1976 ABSTRACT The regulation of fetal heoglobin in adult erythroid cells was investigated with bone arrow cultures. Fetal heoglobin (Hb F) was identified in individual erythroid colonies with fluorescent antibodies against Hb F and synthesis of y chains was deterined with analyses of radioactive globins. The appearance of fetal heoglobin in erythroid colonies was clonal. All the cells of the Hb F synthesizing colonies contained fetal heoglobin. The frequency of erythroid colonies showing Hb was higher than expected copared to the frequency of Hb F containing cells in the blood. Production of Hb F in culture, as shown by analysis of the radioactive globins, was 5 to 14 ties higher than baseline Hb F synthesis. These results suggest that the ability for y chain synthesis in erythroid cells is deterined at or above the level of the precursor cell fro which the erythroid colonies, in vitro, derive (probably an erythropoietin responsive ste cell), and that stiulation of fetal heoglobin synthesis in adult erythroid cells is possible. n an, and a few other aals, fetal heoglobin (Hb F, chain coposition ay272, is the predoinant heoglobin of the fetus and it is replaced by adult heoglobin (Hb A: a2.2) soon after birth. Although hypotheses to explain this switch fro y to fi globin foration in an have been proposed (1-5), the factors involved in this process are unknown. Delineation of the echaniss of regulation of fetal heoglobin in an, however, ight be of considerable biological interest and of possible practical iportance for treatent of sickle cell aneia and hoozygous,f-thalasseia. The replaceent of fetal by adult heoglobin after birth is not coplete, since sall aounts of fetal heoglobin are consistently present in the blood of noral adults (6-9). t is, therefore, of interest to learn how continued synthesis of fetal heoglobin in the adult is regulated and whether it can be further stiulated. We exained this in vitro, with a bone arrow culture syste which perits selective proliferation of erythroid cells (1). We report here data which suggest appearance of increased and clonal synthesis of fetal heoglobin in the adult erythroid cells, in vitro. METHODS Bone Marrow Cultures. Bone arrow aspirates fro healthy volunteers without a heoglobinopathy and fro a patient hoozygous for heoglobin S were used; on a few occasions ribs surgically reoved fro patients were the source of arrow cells. Cells were washed in ediu NCTC-19 (Microbiological Associates, Bethesda, Md.) containing 2% (vol/vol) fetal calf seru, penicillin (5 units/l) and streptoycin (5,gg/l); the buffy coat layers were reoved, and resuspended in fresh edia. Nucleated cells (5 X 14 or 1 X 15) Abbreviations: CM-cellulose, carboxyethylcellulose; CM-Sephadex, carboxyethyl-sephadex; Hb F, fetal heoglobin; Hb A, adult heoglobin; Hb S, heoglobin S; FTC, fluorescein isothiocyanate; anti-hb F-FTC, anti-hb F antibody conjugated with FTC; anti-hb A-FTC, anti-hb A antibody conjugated with FTC. 233 were cultured as plasa clots in icrotiter wells, in the presence of beef ebryo extract (Grand sland Biological Co., Grand sland, N.Y.), fetal calf seru (Flow Laboratories, Rockville, Md.), bovine seru albuin (fraction V, Siga Cheical Co., St. Louis, Mo.), erythropoietin (aneic sheep plasa, Step preparation, Connaught Research Laboratories, Toronto, Canada) or huan urine erythropoietin (specific activity 7 international units/g of protein), and citrated bovine plasa (Grand sland Biological). After clots fored, the cultures were incubated in a huidified CO2 incubator at 37. Seven to 9 days later, the plasa clots were reoved, flattened on glass slides, fixed in 5% glutaraldehyde, and stained with benzidine. Benzidine-positive colonies were scored using accepted criteria (1-12). Antibodies and unofluorescent Labeling. Antibodies against heoglobin F (anti-hb F) or A (anti-hb A) were raised in rabbits (anti-hb F) or horses (anti-hb A) and purified by affinity chroatography. The purification procedure, the conjugation with fluorescein isothiocyanate (FTC), and evidence of specificity of the anti-hb F antibodies have been described (9). Peripheral blood sears were fixed and labeled with anti-hb F antibodies conjugated with FTC (anti-hb F-FTC), as previously described (9); sears prepared fro bone arrow cell suspensions were treated siilarly. Flattened plasa clots were fixed for 12 in in an acetone-ethanol (9:1 vol/vol) solution, rinsed for 2 in in phosphate-buffered saline at ph 7. (9), rinsed in distilled water for another 2 in, and dried. They were labeled with anti-hb F-FTC in the sae fashion as the sears, except that the whole area of the clot was covered with the fluorescent antibody. After incubation for 1 hr in a huidified chaber at 37, the slides with clots were rinsed in phosphate-buffered saline, then in water, and then dried. They were ounted with clear seru, and viewed in the fluorescent icroscope (Zeiss, Universal fluorescence icroscope with reflected light excitation) under 15-ties agnification. Freshly thawed antibody solutions were used for each experient and conditions of fixation, staining, and viewing were kept constant. Evaluation of fluorescent labeling was best when plasa clots contained fewer erythroid colonies: cultures inoculated with 5 X 14 nucleated cells per clot were optial. The reaction of the fixed plasa clots with the fluorescent antiheoglobin antibodies preserved the orphological characteristics of the cells and peritted identification and quantitative estiates of total erythroid colonies under fluorescent light. Erythroid cells and colonies were classified into categories according to intensity of fluorescent labeling. Fluorescence of oderate to bright intensity was considered as evidence for intracellular presence of the heoglobin detected by the fluorescent antibody. Globin Synthesis. Fresh bone arrow cell suspensions or blood were incubated in the presence of 5,uCi of [3H]leucine (New England Nuclear, Boston, Mass.) as described (13) with

2 234 Cell Biology: Papayannopoulou et al. the exception that incubations were carried out for 2 hr. After addition of nonradioactive Hb F, part of the saple was converted to globin and chains were separated on carboxyethyl(cm)-cellulose coluns (14, 15). Another part was subjected to ion exchange chroatography on carboxyethyl(cm)-sephadex (16); heoglobins F and A were isolated, converted to globin, and globin chains separated. The globin coluns were onitored for radioactivity in the eluates by adding 1 l of each fraction to 1 l of Biofluor (New England Nuclear, Boston, Mass.) and radioactivity was deterined in a Packard scintillation spectroeter. Fractions fro heoglobin coluns were processed for liquid scintillation counting, as previously described (13). For studies of globin synthesis in clot cultures, 1 X 17 to 1.5 X 17 nucleated cells were cultured in plasa clots. Cultures were labeled by adding 1,gCi of [3H]leucine (specific activity 8 Ci/ol) per clot at day 6 or 7. Twenty-four hours later, the cultures were harvested; clots were extensively rinsed in large volues of phosphate-buffered saline at ph 7., suspended in a sall volue of distilled water, disrupted by sonication followed by freeze-thawing, and centrifuged at 23, X g. Nonradioactive Hb Fo and Ao (or So) were added to the supernates. The lysates were first subjected to gel filtration, and then to isolation of heoglobins by CM-Sephadex chroatography; globin chains were then separated by CM-cellulose chroatography (14, 15). Alternatively, after CM-Sephadex chroatography of lysates, gel filtration of heoglobins or globins and separation of chains were perfored. Gel filtration of heoglobins was carried out at 4 in 2.5 X 1 c coluns of Sephadex G-1 equilibrated and developed with.5 M Tris-HCl, at ph 7.4 (13). Gel filtration of globins was perfored in 2.5 X 1 c coluns of Sephadex G-1 equilibrated and developed with 2% (vol/vol) foric acid (15). Radioactivity was onitored in the chroatographic fractions as entioned above. RESULTS unofluorescent labeling of the plasa clots peritted the identification of heoglobins synthesized in individual colonies. After reaction with anti-hb A antibody conjugated with FTC (anti-hb A-FTC), all the erythroid colonies fluoresced, indicating that, as expected, Hb A is synthesized in all the erythroid cells in culture. n contrast, reaction with anti-hb F-FTC revealed a different picture: a heterogeneous distribution of Hb F aong the erythroid colonies, with only a few colonies binding the antibody and eitting oderate to bright fluorescence (Fig 1). n addition to its restriction to a sall proportion of erythroid colonies, fetal heoglobin synthesis appeared to be clonal in the sense that all the cells of each Hb F synthesizing colony bound the anti-hb F-FTC, displaying the sae intensity of fluorescence. Occasionally when advanced aturation produced saller noroblasts within a positive colony, the cytoplas of these saller cells was soewhat brighter in fluorescence than the rest of the reaining cells. There was no relationship between presence or absence of Hb F and size of the erythroid colonies. The frequency of erythroid cells synthesizing Hb F in tivo was copared with the frequency of Hb F synthesizing colonies in vitro in 14 individuals without heoglobinopathy and (in two experients) in a person with sickle cell aneia. The proportions of Hb F synthesizing cells in Vivo was deterined after labeling the peripheral blood (9) or the uncultured bone arrow speciens of these individuals with anti-hb F-FTC. Fro.5% to 8% of the erythrocytes in the peripheral blood Proc. Natl. Acad. Sci. USA 73 (1976) FG. 1. Photoicrographs showing erythroid colonies fro 8-day cultures in plasa clots fixed and reacted with anti-hb F-FTC as described in the text. White arrows point to erythroid colonies which fail to bind the fluorescent antibody. (a) Shows three colonies, two of which contain fetal heoglobin; (b) shows the contrast between a bright Hb F containing colony and an erythroid colony failing to bind the anti-hb F-FTC (agnification X 13). contained Hb F (Table 1). n the individuals without a heoglobinopathy, fro 4.3 to 29.6% of the erythroid colonies displayed bright fluorescence after they were labeled with anti-hb F-FTC. The proportion of Hb F synthesizing colonies was several ties higher than that expected fro the frequency of Hb F containing red cells or erythroblasts, in vo (Table 1). To exaine the possibility that induced synthesis of Hb F ight have occurred in culture, the radioactive globins synthesized in culture were analyzed. n three experients, the arrow cells were of noral heoglobin phenotype while in the fourth they were fro a patient with hoozygous Hb S disease. The synthesis of y chain was barely detectable in the uncultured arrow cells fro the noral individuals (Table 2). n the uncultured arrow fro the individual hoozygous for Hb S, y chain synthesis was 1.3 to 1.9% of the total non-a chain synthesis. After 7-8 days in culture, significant aounts of fetal heoglobin (Fig. 2) and of y globin (Fig. 3) were produced. The data fro the four experients are suarized in Table 2. n cultured erythroid cells fro the noral individuals, y chain synthesis increased fro 5-fold to 14-fold above the in vivo baseline levels. n the person with sickle cell aneia,

3 Table 1. Cell Biology: Papayannopoulou et al. Proc. Natl. Acad. Sci. USA 73 (1976) 235 Coparison between the proportion of Hb F synthesizing colonies in plasa clot cultures and the frequency of Hb F containing erythrocytes in the peripheral blood of the bone arrow donors No. of Erythroid colonies % Hb F- % Hb F- Exp. plasa clots Total erythroid with oderate and synthesizing containing no.* counted colonies scored bright fluorescence colonies cells in blood AA AA AA SS t AA AA AA AA AA t AA AA SS t AA AA AA AA * The letters denote the heoglobin genotype of the bone arrow donor. AA: noral; SS: hoozygote for Hb S. t The donor of SS 32 and SS 51 has 21% Hb F containing cells in his blood, but about 1% Hb F containing noroblasts in the bone arrow; this reflects the better survival of red cells containing Hb F in hoozygous Hb S disease. $ Measureents were not done. 'y chain synthesis increased fro the approxiate baseline value of 1.6% to 18.3% after 8 days in culture. Globin synthesis was balanced. The ratio of a to 3+-y chains produced in culture was identical to the a:# chain ratio estiated with incubations of fresh bone arrow or blood saples (Table 2). DSCUSSON Most of the experiental work on the regulation of globin synthesis during developent has been perfored in the ouse, chicken, and aphibia. Work has focused on the transition fro ebryonic to the definitive pattern of erythropoiesis (17). The ajor conclusion was that the switch fro ebryonic to the adult globin foration in these species is ediated through an irreversible replaceent of cell lines which are prograed for different patterns of globin gene transcription. The regulation of fetal heoglobin synthesis and the switch fro y to,b globin chain foration in an appear ore coplex. n contrast to the ebryonic globin foration, there is no relationship between fetal heoglobin foration and the site of erythropoiesis, and there is evidence that during fetal developent, fetal and adult heoglobins are synthesized by the sae erythroid cells (18, 19). The transition fro fetal to adult heoglobin foration is, thus, one fro a developental stage in which ostly -y but also f, chains are synthesized in the cell, to a later stage in which there is exclusive synthesis of, chains. This switch is not coplete, since fetal heoglobin, though in inial quantities, continues to be synthesized in the adult, and ay be elevated in several acquired and hereditary disorders. Several questions arise. Do these developental patterns in Hb F synthesis reflect differences in regulation of transcription of the closely linked ly and, loci, which occurs in erythroid cells of coon ste cell origin? Or do they represent replaceent, during developent, of heopoietic ste cell lines that ostly direct Hb F synthesis, by later lines coitted to exclusive synthesis of Hb A? Which are the factors responsible for the aintenance of Hb F in the fetus? What are the cellular and olecular echaniss of continued fetal Hb synthesis in the adult? s Hb F synthesis in the adult regulated at the level of translation in the fully differentiated erythroid cell? Or is it cellularly restricted, with the ability for y RNA synthesis having been deterined at a ste cell level? s the pattern of globin chain foration in the adult heopoietic tissue reversible? This study provides soe answers, for it presents evidence for both stiulation and cellular restriction Table 2. Globin chain synthesis before culture Globin chain synthesis in huan arrow plasa clot cultures Globin chain synthesis in culture Ratios of [3H]leucine incorporation Ratios of [3H] leucine ncreent [3H ] leucine incorporation total cp (x 1-3) incorporation in 7 chain Exp. synthesis no.* cx/f3 'yb+' ct a/j3 a/f3+y 7/t3~7 in culture AA AA AA SS * Letters denote heoglobin genotype. AA: noral; SS: hoozygote for Hb S.

4 236 Cell Biology: Papayannopoulou et al. Proc. Natl. Acad. Sci. USA 73 (1976) T Fraction nuber w 2 r C) ~ FG. 2. ncorporation of [3H]leucine into heoglobin F and S (separated by carboxyethyl-sephadex chroatography) of an individual with hoozygous heoglobin S disease (SS 51 of Tables 1 and 2). (a) Control 2-hr bone arrow incubation; (b) 7-day plasa clot cultures. The heoglobins used in (b) have been purified by gel filtration in Sephadex G-1. The radioactivity incorporated into Hb F accounts in (a) for less than 2% while in (b) for approxiately 2% of the total heoglobin radioactivity. of fetal heoglobin synthesis in the heopoietic tissue of the adult. There is a great deal of evidence for increased y chain synthesis in this study; in the erythroid cell cultures there is increased synthesis of a protein which (a) chroatographs with heoglobin on gel filtration; (b) elutes with Hb F on ion exchange chroatography of heoglobins; (c) chroatographs with the y globin chain on globin chain chroatography; and (d) has the iunological characteristics of fetal heoglobin. Further direct evidence (to be presented in detail elsewhere) was obtained by radiosequence of the y chain fractions obtained fro carboxyethylcellulose chroatography of purified globin received fro culture lysates. ndirect evidence for increased chain synthesis is obtained fro the estiates of the To - x co CN 1..5 %, Fraction nuber r c 1 c) z -V 5 r x FG. 3. Carboxyethylcellulose chroatography of globin chains showing incorporation of [3H]leucine into y and a chains of heoglobin F purified by CM-Sephadex and Sephadex G-1 chroatography. Eight-day plasa clot cultures of bone arrow of an individual with noral heoglobin genotype (AA 5, Table 2). ratios of a and non-a chains produced in the arrow cells before and after culture. n cultures with the higher Hb F production, the ratio of synthesized a and,3 chains indicates considerable,b chain deficit; the ratio of a/,3+y chain production is, however, identical to the a:f3 chain ratio before culture, and suggests that the,3 chain deficit is copensated by the newly produced y chains. The increased y chain synthesis does not result fro a disordered production of y chains, but rather increased y RNA and decreased,3 RNA production are so regulated as to keep the total globin chain synthesis balanced. The evidence for cellular restriction of Hb F is provided by the results of iunofluorescent labeling with anti-hb F- FTC, which show a clonal appearance of fetal heoglobin in erythroid colonies. n a given Hb F containing colony, all the cells reacted with the fluorescent antibody to a siilar extent which indicates that they had siilar Hb F contents. Because, aong the Hb F containing colonies there were sall ones consisting of 8-16 cells and large colonies containing over 1 cells, production of fetal heoglobin appears independent of the nuber of cell divisions or the rate of erythroid cell proliferation. t sees, thus, that all cells of each colony were prograed for siilar aounts of y chain foration. This suggests that the inforation for the turning on of y chain synthesis was present in the original cells fro which each colony was derived. The erythroid colonies in plasa clots are, ost probably, initiated by erythropoietin-responsive coitted ste cells (112); the findings of the iunocheical study suggest that the ability for y gene transcription is deterined at least at the level of this erythroid cell precursor. ncreased synthesis of Hb F (2) or persistence of y chain synthesis (21) in intro has been reported but was not confired (22). The discrepancy between the present and previous (22) data ay reflect differences in the experiental syste. Previous work has utilized suspension bone arrow culture systes in which erythroid cell proliferation is liited and globin chain synthesis progressively declines (2-22). n contrast, the culture syste used in this study perits selective proliferation of erythroid cells and, thus, offers the opportunity for detection of changed patterns of globin synthesis even if these have occurred in a inor population of erythroid precursors. The developental origin and the echanis of this cellular

5 Cell Biology: Papayannopoulou et al. restriction of increased Hb F production in vitro is, at present, -speculative. f we assue that the precursor cells fro which the Hb F containing colonies arise represent the progeny of a ste cell line prograed for fetal heoglobin synthesis, the increased frequency of Hb F containing colonies in vitro will suggest that the culture conditions provide a survival or a proliferative advantage to these early cells. Alternatively, huoral or icroenvironental stiuli in the culture ight have induced gene transcription in erythroid precursor cells prograed before culture to produce only adult heoglobin; the clonal appearance of Hb F in the erythroid colonies, in the latter case, ight reflect quantitative or qualitative variations of these stiuli in the icroenvironents of the culture. Although there are no answers at present, the in vitro syste we have used provides the possibility for experients along these lines. The evidence for increased Hb F synthesis in seisolid cultures adds to the arguent, based on in vivo observations of subjects with acquired elevations of Hb F, that the switch fro Hb F to Hb A synthesis is reversible, and that induction of fetal heoglobin in the adult is possible. The cooperation of Drs. W. Mills, R. Pinkha, and P. Fluvog, who provided us with freshly reoved rib preparations is thankfully acknowledged. We thank Ms. S. Kurachi, Mr. G. Li, and Ms. E. Garzel for expert technical assistance. The huan urinary erythropoietin (Pool H-11-TaLSL, partially purified by Dr. P. Dukes, Heatology Research Laboratory, Childrens Hospital of Los Angeles) was generously supplied by the Blood Resources Branch of the National Heart and Lung nstitute. This study was supported by Grants GM15253 and HL6242 fro the U.S. Public Health Service. A portion of this work was conducted through the Clinical Research Center facility of the University of Washington supported by the National nstitutes of Health (Grant RR-37). Proc. Natl. Acad. Sci. USA 73 (1976) Neel, J. V. (1961) Blood 18, Motulsky, A. G. (1962) Nature 194, Baglioni, C. (1963) in Molecular Genetics, Part, ed. Taylor, J. H. (Acadeic Press, New York), pp Kabat, D. (1972) Science 175, ngra, V. M. (1972) Nature 235, Bertles, J. F. (1974) Ann. N.Y. Acad. Sci 241, Boyer, S. H., Belding, T. K., Margolet, L. & Noyes, A. N. (1974) Science 188, Boyer, S. H., Belding, T. K., Margolet, L., Noyes, A. N., Burke, P. J. & Bell, W. R. (1975) Johns Hopkins Med. J. 137, Wood, W. G., Staatoyannopoulos, G., Li, G. & Nute, P. E. (1975) Blood 46, Stephenson, J. R., Axelrad, A. A., McLeod, D. L. & Shreeve, M. M. (1971) Proc. Natl. Acad. Sci. USA 68, Tepperan, A. D., Curtis, J. E. & McCulloch, E. A. (1974) Blood 44, McLeod, D. L., Shreeve, M. M. & Axelrad, A. A. (1974) Blood 44, Wood, W. G. & Staatoyannopoulos, G. (1975) J. Clin. nvest. 55, Clegg, J. B., Naughton, M. A. & Weatherall, D. J. (1966) J. Mol. Biol. 19, Chalevelakis, G., Clegg, J. B. & Weatherall, D. J. (1975) Proc. Natl. Acad. Sci. USA 72, Dozy, A. M. & Huisan, T. H. J. (1969) J. Chroatogr. 4, Marks, P. A. & Rifkind, R. A. (1972) Science 175, Shepard, M. K., Weatherall, D. J. & Conley, C. L. (1962) Bull. Johns Hopkins Hosp. 11, Rosenberg, M. (197) Proc. Natl. Acad. Sci. USA 67, Hall, J. G. & Motulsky, A. G. (1968) Nature 217, Gabuzda, T. G., Silver, R. K., Chui, L. C. & Lewis, H. B. (197) Br. J. Haeatol. 19, Wood, W. G. (1974) Br. J. Haeatol. 26,