American Journal of Hematology 7:45-51 (1979)
Chemical Modification of Nuclear Proteins by Erythropoietin Jerry L. Spivak and Lorna Peck The Clayton Laboratories of the Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland
The spleen of the exhypoxic polycythemic mouse was employed as a model system to study the effect of erythropoietin on enzymes that chemically modify nuclear proteins. At selected time intervals after in vivo administration of erythropoietin, acetyltransferase and methyltransferase activity were measured in nuclei isolated from the spleens of treated mice. In addition, the incorporation of labeled methyl and acetate groups into individual histone proteins was also examined. A 36% increase in nuclear acetyltransferase activity was observed eight hours after administration of erythropoietin, whereas nuclear methyltransferase activity increased by 42% 24 hours after administration of the hormone. Selective acetylation or methylation of individual histone proteins was not observed, and it is concluded that activation of transcription by erythropoietin is not the result of acetylation or methylation of nuclear proteins. Key words: erythropoietin, acetylation, methylation, chromatin
INT RODUCTION
Erythropoiesis is regulated by the hormone erythropoietin, but the mechanism by which the hormone interacts with its target cells is unknown. An early effect of erythropoietin in vivo or in vitro is an increase in RNA synthesis [ I , 21. This is not due to an alteration in the intracellular pool size of nucleic acid precursor molecules and is not dependent on DNA [3] or protein synthesis [4]. Studies in our laboratory have indicated that, following administration of the hormone in vivo, there is a sequential activation of different forms of nuclear RNA polymerase in erythropoietin-responsive cells [51 . Similar observations have been made by Goldwasser and Inana employing isolated nuclei and a marrow cytoplasmic factor obtained from erythropoietin-treated cells 161 . Since the marrow cytoplasmic factor had no direct effect on the behavior of partially purified RNA polymerase 161, an explanation for the activation of transcription by erythropoietin must be sought elsewhere. Takaku and coworkers have reported that histone acetylation precedes RNA synthesis in spleen cells exposed to erythropoietin [7] . Because chemical
Received for publication November 30, 1978; accepted April 1 7 , 1979. Address reprint requests to Dr. Spivak, Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205.
0361-8609/79/0701-0045$01.70 0 1979 Alan
R. Liss, Inc.
46
Spivak and Peck
modification of nuclear proteins might be involved in the activation of transcription by the hormone, we examined the effect of erythropoietin on nuclear acetyltransferase and methyltransferase activity in the spleen of the exhypoxic mouse. MATERIALS AND METHODS Preparation of Tissue
Exhypoxic polycythemic Swiss-Webster female mice with hematocrits of 0.65 or greater were injected in groups of four with either 5 units of erythropoietin (Step 111, 3.2 units/mg protein, Connaught Medical Research Laboratory, Willowdale, Ontario, Canada) in 0.5 ml of 0.5% bovine serum albumin in saline (BSA-NSS) or 0.5 ml of 0.8% BSA-NSS. At selected time intervals, mice were killed by cervical dislocation and the spleens were removed. Spleen cell nuclei were isolated by a modification [8] of the method of Blobel and Potter [9]. DNA content was determined by the method of Burton [lo]. Assay of Nuclear Acetyltransferase
The assay is a modification of the procedure described by Bondy et a1 [ 1 11 . Spleen cell nuclei were suspended in 0.05M Tris HCl (pH 7.0), 25% glycerol, 0.005 M MgClz, 0.001 M EDTA (TGME) to give a DNA concentration of 0.5-1 .O mg/ml. The nuclei (50-100 pg DNA) were incubated in a reaction mixture that contained, in a final volume of 250 0.2 pCi 3H-acetyl coenzyme A (acetyl CoA, specific activity 0.6-3.07 Ci/mmole), 0.01 pmole unlabeled acetyl CoA, 10 pg pyruvate kinase, and 0.002 mmole phosphoenolpyruvate in TGME. When incorporation of acetate into individual histone proteins was examined, the reaction mixture was scaled up fourfold. Incorporation of labeled acetate into nuclear protein had a pH optimum of 7.0, was temperature sensitive, was linear until the concentration of nuclei expressed as p g of DNA exceeded 100 pg, and was complete within 12 minutes. The reaction was stopped by chilling the reaction tubes and adding trichloroacetic acid (TCA) to a final concentration of 20%. The precipitated nuclei were collected on nitrocellulose filters, which were washed with chloroform: Ethanol (1 : 1) and then dried for liquid scintillation counting in a toluene-based scintillation fluor. Counting efficiency for tritium was 30%. Incorporation of acetate into nuclear protein was not due to protein synthesis since the reaction was not inhibited by puromycin and no incorporation of acetate occurred when labeled acetate was substituted for labeled acetyl CoA. Since nuclei heated to 65°C did not incorporate acetate into nuclear protein, it is likely that acetate incorporation was due to an enzymatic reaction. In addition, when histones and nonhistone proteins were isolated from nuclei after incubation with labeled acetyl CoA, 90% of the label was found in the histone proteins. Assay of Nuclear Methyltransferase
The assay is a modification of the procedure of Sekeris et a1 [ 121 . Spleen cell nuclei were suspended in 0.05 M Tris HCl (pH 8.5), 0.25 M sucrose, 0.005 M MgClz, 0.001 M EDTA, and 0.001 M dithiothreitol (TSMED) to give a DNA concentration of 1.O-2.0 mg/ml. The nuclei (100-200 pg DNA) were incubated at 37°C in a reaction mixture that contained, in a final volume of 250 pl, 1 pCi (methyL3H) S-adenosyl-
Erythropoietin and Nuclear Proteins
47
methionine (specific activity, 8.8-1 1.9 Ci/mmole) and 0.002 mmole unlabeled S-adenosylmethionine in TSMED buffer. Incorporation of methionine into the isolated nuclei had a pH optimum of 8.5, was linear until the concentration of nuclei expressed as pg of DNA exceeded 400 pg, and was complete within eight minutes. After the reaction was complete, the nuclei were processed as described above for the acetylation experiments. Methylation of nuclear protein was not due to protein synthesis, since the reaction was not inhibited by puromycin and no methylation occurred when methionine was substituted for S-adenosylmethionine. Since nuclei heated to 65°C did not incorporate methyl groups into nuclear protein, it is likely that methylation was due t o an enzymatic reaction. The methylated product from the reaction mixture was not precipitated by 10%perchloric acid, indicating that the product was not methylated DNA. Approximately 20% of the labeled methyl groups were soluble in 20% TCA after incubation with ribonuclease. Consequently, ribonuclease (69 pg) was routinely included in the reaction mixture. In contrast to acetylation, histone and nonhistone proteins were methylated t o the same extent. Preparation of Asialoerythropoietin [ 13)
Sheep erythropoietin was dissolved 0.03 M HCl and heated at 80°C for 30 minutes. After cooling on ice, the pH was adjusted to 7.4 and the solution was dialyzed against phosphate-buffered saline before use. Erythropoietin treated in this fashion fails to stimulate incorporation of 59Fe into red cells in vivo in exhypoxic polycythemic mice but is still capable of stimulating the proliferation of erythroid colonies (CFU-E) in vitro in plasma clot cultures.
RESULTS Nuclear Acetyltransferase Activity
The effect of erythropoietin on spleen nuclear acetyltransferase activity is shown in Table I. Only at eight hours after administration of the hormone was there a significant increase in acetyltransferase activity. The effect appeared to be both specific for erythropoietin and organ specific, since asialoerythropoietin, prepared by acid hydrolysis, failed to stimulate nuclear acetyltransferase activity (Table IIA), nor was liver nuclear acetyltransferase activity affected at eight hours (Table IIB) or at other time intervals (1, 12, 24, 36, and 70 hours, data not shown). The extent to which individual histone proteins were acetylated was investigated by employing a double label technique [ 141 . Nuclei from control animals were incubated with either l4 C-acetyl CoA or H-acetyl CoA, and nuclei from erythropoietin-treated animals were incubated with l4 C-acetyl CoA. After incubation, H-acetate-labeled control nuclei were mixed with equal amounts of either l4 C-acetate-labeled control nuclei or l 4 C-acetate labeled nuclei from erythropoietin-treated animals. The nuclear proteins were extracted from the mixed nuclei as previously described [ 141 , and the histone proteins were isolated by polyacrylamide gel electrophoresis for determination of the 14C/3Hratio. As shown in Table 111, at the time of maximal nuclear acetyltransferase activity there was no selective acetylation of the individual histone proteins. The quantity
48
Spivak and Peck TABLE 1. Effect of Erythropoietin on Splenic Nuclear Acetyltransferase Activity Time after injection (hours)
Erythropoietin
Percent change
cpm/gg D N A ~ 148.2 f 3.0 142.G-t 6.6 80.3 f 4.9 65.8 i- 2.4 147.0 i 6.8 111.0 t 7.3 160.3 f 3.8 165.0 c 41.2 53.8 t 5.6 73.6 -e 5.8 120.0 f 2.8 163.5 t 2.2 63.7 i 5.5 76.7 c 1.7 143.5 * 2.7 153.0 t 2.1 208.2 c 3.2 172.0 2 1.4 168.5 c 4.4 184.5 -i 3.1 222.7 t 1.8 212.2 t 2.9 297:6 2 14.1 282.0 t 13.4 77.0 * 4.5 75.9 t 2.4 182.2 f 6.7 194.0 t 6.5 8’1.3 -t 2.5 96.7 i 7.1 87.9k 5.1 78.0 c 3.8 96.2 t 6.8 84.8 * 2.8 161.0 t 6.3 174.0 f 5.0
NSb NS NS NS 36 * 36* NS NS NS NS NS NS NS NS NS NS NS NS
BSA-NSS
1 4 8 12 18 24 36 48 70
aMean c SEM. bNo significhnt difference. *P < 0.05 TABLE 11. Effect of Erythropoietin on Splenic and Liver Nuclear Acetyltransferase Activity Time after injection (hours)
A. Spleen B. Liver
8 8
BSA-NSS
Erythropoietin cpm/gg DNAa 15.0 -t 1.7b 147.5 t 6.8
23.8 c 1.0 181.8 t 3.9
aMean i- SEM bDesialated.
of nonhistone proteins was too small for analysis of acetate incorporation into the individual proteins. Nuclear Methyltransferase Activity
Since erythropoietin-stimulated nuclear acetyltransferase activity did not precede hormone-stimulated activation of transcription, we examined the effect of erythropoietin on nuclear methyltransferase activity. As shown in Table IV, a substantial increase in nuclear methyltransferase activity was observed 24 hours after administration of the hormone. The effect appeared to be specific for erythropoietin as well as organ specific, since no effect was seen with asialoerythropoietin (Table VA) and the hormone did
E r y t h r o p o i e t i n and Nuclear Proteins TABLE 111. Effect of Erythropoietin on Acetylation of Splenic Histone Proteins* Histone Proteins 14C/3H ratio
Fl
Control (8hours) 0.09 Control Er ythropoietin (8 hours) 0.08 Control
F3
F2b
F2a2
F2al
0.09
0.10
0.08
0.12
0.09
0.09
0.07
0.09
*Nuclei from erythropoietin-treated animals labeled with ''C-acetyl CoA were mixed with control nuclei labeled with 3H-acetyl CoA. The histone roteins were isolated from the mixed nuclei for determination of the p4C/3H ratio. A mixture of labeled control nuclei was analyzed in the same manner. Although not shown, similar results were obtained at 1 2 and 48 hours after administration of erythropoietin.
TABLE IV. Effect of Erythropoietin on Splenic Nuclear Methyltransferase Activity Time after injection (hours)
BSA-NSS
10.4 t 1.3 5.0 f 0.3 6.3 i: 0.2 6.6 i: 0.1 7.4 f 0.4 10.4 f 0.4 13.4 t 0.1 28.2 i: 1.0 17.0 t 1.0 10.8 i 0.7 16.0 i 0.5 6.9 f 0.4
1
4 8 24 48 70
Erythropoietin cpm/pg DNAa 10.4 t 0.3 6.0 i: 0.1 4.6 i: 0.3 7.3 t 0.3 8.4 f 0.7 11.3 i: 0.02 17.4 i 1.1 43.4 t 1.7 18.9 i: 1.9 11.8 t 0.7 14.6 t 0.9 6.5 i 0.4
aMean t SEM. bNo significant difference. *P < 0.05.
TABLE V. Effect of Erythropoietin on Splenic and Liver Nuclear Methyltransferase Activity Time after injection (hours) A . Spleen
B. Liver
24 24
ahlean t SEM bDesialated.
BSA-NSS
Erythropoietin
cpm/pg D N A ~ 11.0 f 0.8 8.4 -r 0.6b 45.5 t 1.5 46.8+ 1.8
Percent change NSb NS NS NS NS NS 29* 53* NS NS NS NS
49
50
Spivak and Peck
not stimulate liver methyltransferase activity (Table VB). Isolation and separation of the histone proteins by polyacrylamide gel electrophoresis failed to reveal selective methylation of any of these proteins (data not shown). DISCUSSION Following administration of erythropoietin, a well-defined sequence of biochemical events occurs in the spleen of the exhypoxic polycythemic mouse, culminating in the production of mature erythrocytes [ S , 14, 151. In the spleen, the earliest observed effect of erythropoietin is an increase in the activity of nuclear RNA polymerase [5] which occurs within one hour after administration of the hormone. In the present study, we have employed this model in order to determine whether the activation of RNA polymerase by erythropoietin is associated with chemical modification of nuclear proteins, as suggested by Takaku et a1 [7]. In accordance with the observations of those investigators, we found that nuclear acetyltransferase activity was increased eight hours after administration of erythropoietin. However, in contrast to their findings, we did not observe an increase in nuclear protein acetylation at an earlier time period, nor did we observe increased acetylation of histone proteins. These conflicting observations could be due to differences in the erythropoietin preparation employed or the method of preparing polycythemic mice, but they are more likely due to the method of measuring nuclear protein acetylation. Takaku et a1 employed a technique that does not distinguish between enzymatic acetylation of existing nuclear proteins and incorporation of acetate during the de novo synthesis of nuclear proteins. The time course of acetate incorporation into histones that they describe corresponds to the demonstrated increase in synthesis of histone proteins in the spleen, which occurs within three hours after administration of erythropoietin [ 141 . Using an assay that specifically quantitates nuclear acetyltransferase, we failed to demonstrate that acetyltransferase activity increased prior to the activation of transcription. This is not surprising since acetylation of histones has not been strictly correlated with RNA synthesis in avian erythrocytes [16], and in phytohemagglutinintreated lymphocytes, RNA synthesis was associated with a decrease in histone acetylation [17]. In contrast to acetyltransferase activity, maximal nuclear methyltransferase activity was a relatively late event following administration of erythropoietin. Methylation of histones is a late event in the cell cycle of the regenerating hepatocyte and may be involved in the condensation of chromatin and cessation of nucleic acid synthesis in that cell [ 181 . The changes that we observed in methyltransferase activity may reflect a similar phenomenon in differentiating erythroid cells. Our data indicate that an explanation for the activation of transcription by erythropoietin cannot be ascribed to acetylation or methylation of nuclear proteins. Recently, Adamson and his coworkers have presented evidence that erythropoietinstimulated erythroid cell differentiation can be potentiated by agents that activate the enzyme adenylate cyclase [19] . However, whether CAMPfunctions as a second messenger for erythropoietin is unsettled [20]. Recent studies from our laboratory suggest that divalent cations are also important in the interaction of erythropoietin and its target cells [21]. The interrelationship of divalent cations and cyclic nucleotides in this process and how they influence the intranuclear events observed after administration of erythropoietin are subjects for further study.
Erythropoietin and Nuclear Proteins
51
ACKNOWLEDGMENTS
This work was supported by U.S.P.H.S. grant AM 16702. REFERENCES 1. Rudolph W, Perretta M: Effect of erythropoietin on ‘‘C-formate uptake by spleen and bone marrow nucleic acids of erythrocyte-transfused mice. Proc SOCExp Biol Med 124:1041, 1967. 2. Gross M, Goldwasser E : On the mechanism of erythropoietin-induced differentiation. V. Characterization of the ribonucleic acid formed as a result of erythropoietin action. Biochemistry 8:1795, 1969. 3. Gross M, Goldwasser E: On the mechanism of erythropoietin-induced differentiation. VII. The relationship between stimulated deoxyribonucleic acid synthesis and ribonucleic acid synthesis. J Biol Chem 245:1632, 1970. 4. Gross M, Goldwasser E: On the mechanism of erythropoietin-induced differentiation. XI. Stimulated RNA synthesis independent of protein synthesis. Biochim Biophys Acta 287:5 14, 1972. 5. Piantadosi CA, Dickerman HW, Spivdk JL: Sequential activation of splenic nuclear RNA polymerases by erythropoietin. J Clin Invest 57:20, 1976. 6. Goldwasser E, Inana G: Molecular aspects of the initiation of erythropoiesis. In Golde DW, Cline MJ, Metcalf D, Fox CF (eds): “Hematopoietic Cell Differentiation.” New York: Academic Press, 1978, p 15. 7. Takaku F, Nakao K, Ono T, et al: Changes in histone acetylation and RNA synthesis in the spleen of polycythemic mouse after erythropoietin injection. Biochim Biophys Acta 195: 396,1969. 8. Spivak JL, Toretti D, Dickerman HW: Effects of phenylhydrazine-induced hemolytic anemia on nuclear RNA polymerase activity of the mouse spleen. Blood 42:257, 1973. 9. Blobel G, Potter VR: Nuclei from rat liver: Isolation method that combines purity with high yield. Science 154:1662, 1966. 10. Burton K: A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem J 62:315, 1956. 11. Bondy SC, Roberts S, Morelos BS: Histone-acetylating enzyme of brain. Biochem J 119:665, 1970. 12. Sekeris CE, Sekeri KE, Gallwitz D: The methylation of the histones of rat liver nuclei in vitro. Hoppe-Seylers Z Physiol Chem 384:1660, 1967. 13. Goldwasser E, Kung CK-H, Elidson J : On the mechanism of erythropoietin-induced differentiation. XIII. The role of sialic acid in erythropoietin action. J Biol Chem 245:4204, 1974. 14. Spivak JL: Effect of erythropoietin on chromosomal protein synthesis. Blood 4 7 5 8 1 , 1976. 15. Roodman GD, Hutton JJ, Bollum FT: DNA polymerase activities during erythropoiesis: Effects of erythropoietin, vinblastine, colcemid and daunomycin. Exp Cell Res 91:269, 1975. 16. Sanders LA, Schechter NM, McCarty KS: Comparative study of histone acetylation, histone deacetylation and ribonucleic acid synthesis in avian reticulocytes and erythrocytes. Biochemistry 12:783, 1973. 17. Monjardino JPPV, MacGillivray AJ: RNA and histone metabolism in small lymphocytes stimulated by phytohemagglutinin. Exp Cell Res 60:1, 1970. 18. Tidwell T, Allfrey VG, Mirsky AE: The methylation of histones during regeneration of the liver. J Biol Chem 243:707, 1968. 19. Brown JE, Adamson JW: Modulation of in vitro erythropoiesis: The influence of p-adrenergic agonists on erythroid colony formation. J Clin Invest 60:70, 1977. 20. Graber SE, Carriilo M, Krantz SB: Lack of effect of erythropoietin on cyclic adenosine - 3‘,5’ monophosphate levels in rat fetal liver cells. J Lab Clin Med 83:288, 1974. 21. Misiti J, Spivak JL: Erythropoiesis in vitro: Requirement for calcium. Clin Res 26:620A, 1978.
Chemical Modification of Nuclear Proteins by Erythropoietin Jerry L. Spivak and Lorna Peck The Clayton Laboratories of the Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland
The spleen of the exhypoxic polycythemic mouse was employed as a model system to study the effect of erythropoietin on enzymes that chemically modify nuclear proteins. At selected time intervals after in vivo administration of erythropoietin, acetyltransferase and methyltransferase activity were measured in nuclei isolated from the spleens of treated mice. In addition, the incorporation of labeled methyl and acetate groups into individual histone proteins was also examined. A 36% increase in nuclear acetyltransferase activity was observed eight hours after administration of erythropoietin, whereas nuclear methyltransferase activity increased by 42% 24 hours after administration of the hormone. Selective acetylation or methylation of individual histone proteins was not observed, and it is concluded that activation of transcription by erythropoietin is not the result of acetylation or methylation of nuclear proteins. Key words: erythropoietin, acetylation, methylation, chromatin
INT RODUCTION
Erythropoiesis is regulated by the hormone erythropoietin, but the mechanism by which the hormone interacts with its target cells is unknown. An early effect of erythropoietin in vivo or in vitro is an increase in RNA synthesis [ I , 21. This is not due to an alteration in the intracellular pool size of nucleic acid precursor molecules and is not dependent on DNA [3] or protein synthesis [4]. Studies in our laboratory have indicated that, following administration of the hormone in vivo, there is a sequential activation of different forms of nuclear RNA polymerase in erythropoietin-responsive cells [51 . Similar observations have been made by Goldwasser and Inana employing isolated nuclei and a marrow cytoplasmic factor obtained from erythropoietin-treated cells 161 . Since the marrow cytoplasmic factor had no direct effect on the behavior of partially purified RNA polymerase 161, an explanation for the activation of transcription by erythropoietin must be sought elsewhere. Takaku and coworkers have reported that histone acetylation precedes RNA synthesis in spleen cells exposed to erythropoietin [7] . Because chemical
Received for publication November 30, 1978; accepted April 1 7 , 1979. Address reprint requests to Dr. Spivak, Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205.
0361-8609/79/0701-0045$01.70 0 1979 Alan
R. Liss, Inc.
46
Spivak and Peck
modification of nuclear proteins might be involved in the activation of transcription by the hormone, we examined the effect of erythropoietin on nuclear acetyltransferase and methyltransferase activity in the spleen of the exhypoxic mouse. MATERIALS AND METHODS Preparation of Tissue
Exhypoxic polycythemic Swiss-Webster female mice with hematocrits of 0.65 or greater were injected in groups of four with either 5 units of erythropoietin (Step 111, 3.2 units/mg protein, Connaught Medical Research Laboratory, Willowdale, Ontario, Canada) in 0.5 ml of 0.5% bovine serum albumin in saline (BSA-NSS) or 0.5 ml of 0.8% BSA-NSS. At selected time intervals, mice were killed by cervical dislocation and the spleens were removed. Spleen cell nuclei were isolated by a modification [8] of the method of Blobel and Potter [9]. DNA content was determined by the method of Burton [lo]. Assay of Nuclear Acetyltransferase
The assay is a modification of the procedure described by Bondy et a1 [ 1 11 . Spleen cell nuclei were suspended in 0.05M Tris HCl (pH 7.0), 25% glycerol, 0.005 M MgClz, 0.001 M EDTA (TGME) to give a DNA concentration of 0.5-1 .O mg/ml. The nuclei (50-100 pg DNA) were incubated in a reaction mixture that contained, in a final volume of 250 0.2 pCi 3H-acetyl coenzyme A (acetyl CoA, specific activity 0.6-3.07 Ci/mmole), 0.01 pmole unlabeled acetyl CoA, 10 pg pyruvate kinase, and 0.002 mmole phosphoenolpyruvate in TGME. When incorporation of acetate into individual histone proteins was examined, the reaction mixture was scaled up fourfold. Incorporation of labeled acetate into nuclear protein had a pH optimum of 7.0, was temperature sensitive, was linear until the concentration of nuclei expressed as p g of DNA exceeded 100 pg, and was complete within 12 minutes. The reaction was stopped by chilling the reaction tubes and adding trichloroacetic acid (TCA) to a final concentration of 20%. The precipitated nuclei were collected on nitrocellulose filters, which were washed with chloroform: Ethanol (1 : 1) and then dried for liquid scintillation counting in a toluene-based scintillation fluor. Counting efficiency for tritium was 30%. Incorporation of acetate into nuclear protein was not due to protein synthesis since the reaction was not inhibited by puromycin and no incorporation of acetate occurred when labeled acetate was substituted for labeled acetyl CoA. Since nuclei heated to 65°C did not incorporate acetate into nuclear protein, it is likely that acetate incorporation was due to an enzymatic reaction. In addition, when histones and nonhistone proteins were isolated from nuclei after incubation with labeled acetyl CoA, 90% of the label was found in the histone proteins. Assay of Nuclear Methyltransferase
The assay is a modification of the procedure of Sekeris et a1 [ 121 . Spleen cell nuclei were suspended in 0.05 M Tris HCl (pH 8.5), 0.25 M sucrose, 0.005 M MgClz, 0.001 M EDTA, and 0.001 M dithiothreitol (TSMED) to give a DNA concentration of 1.O-2.0 mg/ml. The nuclei (100-200 pg DNA) were incubated at 37°C in a reaction mixture that contained, in a final volume of 250 pl, 1 pCi (methyL3H) S-adenosyl-
Erythropoietin and Nuclear Proteins
47
methionine (specific activity, 8.8-1 1.9 Ci/mmole) and 0.002 mmole unlabeled S-adenosylmethionine in TSMED buffer. Incorporation of methionine into the isolated nuclei had a pH optimum of 8.5, was linear until the concentration of nuclei expressed as pg of DNA exceeded 400 pg, and was complete within eight minutes. After the reaction was complete, the nuclei were processed as described above for the acetylation experiments. Methylation of nuclear protein was not due to protein synthesis, since the reaction was not inhibited by puromycin and no methylation occurred when methionine was substituted for S-adenosylmethionine. Since nuclei heated to 65°C did not incorporate methyl groups into nuclear protein, it is likely that methylation was due t o an enzymatic reaction. The methylated product from the reaction mixture was not precipitated by 10%perchloric acid, indicating that the product was not methylated DNA. Approximately 20% of the labeled methyl groups were soluble in 20% TCA after incubation with ribonuclease. Consequently, ribonuclease (69 pg) was routinely included in the reaction mixture. In contrast to acetylation, histone and nonhistone proteins were methylated t o the same extent. Preparation of Asialoerythropoietin [ 13)
Sheep erythropoietin was dissolved 0.03 M HCl and heated at 80°C for 30 minutes. After cooling on ice, the pH was adjusted to 7.4 and the solution was dialyzed against phosphate-buffered saline before use. Erythropoietin treated in this fashion fails to stimulate incorporation of 59Fe into red cells in vivo in exhypoxic polycythemic mice but is still capable of stimulating the proliferation of erythroid colonies (CFU-E) in vitro in plasma clot cultures.
RESULTS Nuclear Acetyltransferase Activity
The effect of erythropoietin on spleen nuclear acetyltransferase activity is shown in Table I. Only at eight hours after administration of the hormone was there a significant increase in acetyltransferase activity. The effect appeared to be both specific for erythropoietin and organ specific, since asialoerythropoietin, prepared by acid hydrolysis, failed to stimulate nuclear acetyltransferase activity (Table IIA), nor was liver nuclear acetyltransferase activity affected at eight hours (Table IIB) or at other time intervals (1, 12, 24, 36, and 70 hours, data not shown). The extent to which individual histone proteins were acetylated was investigated by employing a double label technique [ 141 . Nuclei from control animals were incubated with either l4 C-acetyl CoA or H-acetyl CoA, and nuclei from erythropoietin-treated animals were incubated with l4 C-acetyl CoA. After incubation, H-acetate-labeled control nuclei were mixed with equal amounts of either l4 C-acetate-labeled control nuclei or l 4 C-acetate labeled nuclei from erythropoietin-treated animals. The nuclear proteins were extracted from the mixed nuclei as previously described [ 141 , and the histone proteins were isolated by polyacrylamide gel electrophoresis for determination of the 14C/3Hratio. As shown in Table 111, at the time of maximal nuclear acetyltransferase activity there was no selective acetylation of the individual histone proteins. The quantity
48
Spivak and Peck TABLE 1. Effect of Erythropoietin on Splenic Nuclear Acetyltransferase Activity Time after injection (hours)
Erythropoietin
Percent change
cpm/gg D N A ~ 148.2 f 3.0 142.G-t 6.6 80.3 f 4.9 65.8 i- 2.4 147.0 i 6.8 111.0 t 7.3 160.3 f 3.8 165.0 c 41.2 53.8 t 5.6 73.6 -e 5.8 120.0 f 2.8 163.5 t 2.2 63.7 i 5.5 76.7 c 1.7 143.5 * 2.7 153.0 t 2.1 208.2 c 3.2 172.0 2 1.4 168.5 c 4.4 184.5 -i 3.1 222.7 t 1.8 212.2 t 2.9 297:6 2 14.1 282.0 t 13.4 77.0 * 4.5 75.9 t 2.4 182.2 f 6.7 194.0 t 6.5 8’1.3 -t 2.5 96.7 i 7.1 87.9k 5.1 78.0 c 3.8 96.2 t 6.8 84.8 * 2.8 161.0 t 6.3 174.0 f 5.0
NSb NS NS NS 36 * 36* NS NS NS NS NS NS NS NS NS NS NS NS
BSA-NSS
1 4 8 12 18 24 36 48 70
aMean c SEM. bNo significhnt difference. *P < 0.05 TABLE 11. Effect of Erythropoietin on Splenic and Liver Nuclear Acetyltransferase Activity Time after injection (hours)
A. Spleen B. Liver
8 8
BSA-NSS
Erythropoietin cpm/gg DNAa 15.0 -t 1.7b 147.5 t 6.8
23.8 c 1.0 181.8 t 3.9
aMean i- SEM bDesialated.
of nonhistone proteins was too small for analysis of acetate incorporation into the individual proteins. Nuclear Methyltransferase Activity
Since erythropoietin-stimulated nuclear acetyltransferase activity did not precede hormone-stimulated activation of transcription, we examined the effect of erythropoietin on nuclear methyltransferase activity. As shown in Table IV, a substantial increase in nuclear methyltransferase activity was observed 24 hours after administration of the hormone. The effect appeared to be specific for erythropoietin as well as organ specific, since no effect was seen with asialoerythropoietin (Table VA) and the hormone did
E r y t h r o p o i e t i n and Nuclear Proteins TABLE 111. Effect of Erythropoietin on Acetylation of Splenic Histone Proteins* Histone Proteins 14C/3H ratio
Fl
Control (8hours) 0.09 Control Er ythropoietin (8 hours) 0.08 Control
F3
F2b
F2a2
F2al
0.09
0.10
0.08
0.12
0.09
0.09
0.07
0.09
*Nuclei from erythropoietin-treated animals labeled with ''C-acetyl CoA were mixed with control nuclei labeled with 3H-acetyl CoA. The histone roteins were isolated from the mixed nuclei for determination of the p4C/3H ratio. A mixture of labeled control nuclei was analyzed in the same manner. Although not shown, similar results were obtained at 1 2 and 48 hours after administration of erythropoietin.
TABLE IV. Effect of Erythropoietin on Splenic Nuclear Methyltransferase Activity Time after injection (hours)
BSA-NSS
10.4 t 1.3 5.0 f 0.3 6.3 i: 0.2 6.6 i: 0.1 7.4 f 0.4 10.4 f 0.4 13.4 t 0.1 28.2 i: 1.0 17.0 t 1.0 10.8 i 0.7 16.0 i 0.5 6.9 f 0.4
1
4 8 24 48 70
Erythropoietin cpm/pg DNAa 10.4 t 0.3 6.0 i: 0.1 4.6 i: 0.3 7.3 t 0.3 8.4 f 0.7 11.3 i: 0.02 17.4 i 1.1 43.4 t 1.7 18.9 i: 1.9 11.8 t 0.7 14.6 t 0.9 6.5 i 0.4
aMean t SEM. bNo significant difference. *P < 0.05.
TABLE V. Effect of Erythropoietin on Splenic and Liver Nuclear Methyltransferase Activity Time after injection (hours) A . Spleen
B. Liver
24 24
ahlean t SEM bDesialated.
BSA-NSS
Erythropoietin
cpm/pg D N A ~ 11.0 f 0.8 8.4 -r 0.6b 45.5 t 1.5 46.8+ 1.8
Percent change NSb NS NS NS NS NS 29* 53* NS NS NS NS
49
50
Spivak and Peck
not stimulate liver methyltransferase activity (Table VB). Isolation and separation of the histone proteins by polyacrylamide gel electrophoresis failed to reveal selective methylation of any of these proteins (data not shown). DISCUSSION Following administration of erythropoietin, a well-defined sequence of biochemical events occurs in the spleen of the exhypoxic polycythemic mouse, culminating in the production of mature erythrocytes [ S , 14, 151. In the spleen, the earliest observed effect of erythropoietin is an increase in the activity of nuclear RNA polymerase [5] which occurs within one hour after administration of the hormone. In the present study, we have employed this model in order to determine whether the activation of RNA polymerase by erythropoietin is associated with chemical modification of nuclear proteins, as suggested by Takaku et a1 [7]. In accordance with the observations of those investigators, we found that nuclear acetyltransferase activity was increased eight hours after administration of erythropoietin. However, in contrast to their findings, we did not observe an increase in nuclear protein acetylation at an earlier time period, nor did we observe increased acetylation of histone proteins. These conflicting observations could be due to differences in the erythropoietin preparation employed or the method of preparing polycythemic mice, but they are more likely due to the method of measuring nuclear protein acetylation. Takaku et a1 employed a technique that does not distinguish between enzymatic acetylation of existing nuclear proteins and incorporation of acetate during the de novo synthesis of nuclear proteins. The time course of acetate incorporation into histones that they describe corresponds to the demonstrated increase in synthesis of histone proteins in the spleen, which occurs within three hours after administration of erythropoietin [ 141 . Using an assay that specifically quantitates nuclear acetyltransferase, we failed to demonstrate that acetyltransferase activity increased prior to the activation of transcription. This is not surprising since acetylation of histones has not been strictly correlated with RNA synthesis in avian erythrocytes [16], and in phytohemagglutinintreated lymphocytes, RNA synthesis was associated with a decrease in histone acetylation [17]. In contrast to acetyltransferase activity, maximal nuclear methyltransferase activity was a relatively late event following administration of erythropoietin. Methylation of histones is a late event in the cell cycle of the regenerating hepatocyte and may be involved in the condensation of chromatin and cessation of nucleic acid synthesis in that cell [ 181 . The changes that we observed in methyltransferase activity may reflect a similar phenomenon in differentiating erythroid cells. Our data indicate that an explanation for the activation of transcription by erythropoietin cannot be ascribed to acetylation or methylation of nuclear proteins. Recently, Adamson and his coworkers have presented evidence that erythropoietinstimulated erythroid cell differentiation can be potentiated by agents that activate the enzyme adenylate cyclase [19] . However, whether CAMPfunctions as a second messenger for erythropoietin is unsettled [20]. Recent studies from our laboratory suggest that divalent cations are also important in the interaction of erythropoietin and its target cells [21]. The interrelationship of divalent cations and cyclic nucleotides in this process and how they influence the intranuclear events observed after administration of erythropoietin are subjects for further study.
Erythropoietin and Nuclear Proteins
51
ACKNOWLEDGMENTS
This work was supported by U.S.P.H.S. grant AM 16702. REFERENCES 1. Rudolph W, Perretta M: Effect of erythropoietin on ‘‘C-formate uptake by spleen and bone marrow nucleic acids of erythrocyte-transfused mice. Proc SOCExp Biol Med 124:1041, 1967. 2. Gross M, Goldwasser E : On the mechanism of erythropoietin-induced differentiation. V. Characterization of the ribonucleic acid formed as a result of erythropoietin action. Biochemistry 8:1795, 1969. 3. Gross M, Goldwasser E: On the mechanism of erythropoietin-induced differentiation. VII. The relationship between stimulated deoxyribonucleic acid synthesis and ribonucleic acid synthesis. J Biol Chem 245:1632, 1970. 4. Gross M, Goldwasser E: On the mechanism of erythropoietin-induced differentiation. XI. Stimulated RNA synthesis independent of protein synthesis. Biochim Biophys Acta 287:5 14, 1972. 5. Piantadosi CA, Dickerman HW, Spivdk JL: Sequential activation of splenic nuclear RNA polymerases by erythropoietin. J Clin Invest 57:20, 1976. 6. Goldwasser E, Inana G: Molecular aspects of the initiation of erythropoiesis. In Golde DW, Cline MJ, Metcalf D, Fox CF (eds): “Hematopoietic Cell Differentiation.” New York: Academic Press, 1978, p 15. 7. Takaku F, Nakao K, Ono T, et al: Changes in histone acetylation and RNA synthesis in the spleen of polycythemic mouse after erythropoietin injection. Biochim Biophys Acta 195: 396,1969. 8. Spivak JL, Toretti D, Dickerman HW: Effects of phenylhydrazine-induced hemolytic anemia on nuclear RNA polymerase activity of the mouse spleen. Blood 42:257, 1973. 9. Blobel G, Potter VR: Nuclei from rat liver: Isolation method that combines purity with high yield. Science 154:1662, 1966. 10. Burton K: A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem J 62:315, 1956. 11. Bondy SC, Roberts S, Morelos BS: Histone-acetylating enzyme of brain. Biochem J 119:665, 1970. 12. Sekeris CE, Sekeri KE, Gallwitz D: The methylation of the histones of rat liver nuclei in vitro. Hoppe-Seylers Z Physiol Chem 384:1660, 1967. 13. Goldwasser E, Kung CK-H, Elidson J : On the mechanism of erythropoietin-induced differentiation. XIII. The role of sialic acid in erythropoietin action. J Biol Chem 245:4204, 1974. 14. Spivak JL: Effect of erythropoietin on chromosomal protein synthesis. Blood 4 7 5 8 1 , 1976. 15. Roodman GD, Hutton JJ, Bollum FT: DNA polymerase activities during erythropoiesis: Effects of erythropoietin, vinblastine, colcemid and daunomycin. Exp Cell Res 91:269, 1975. 16. Sanders LA, Schechter NM, McCarty KS: Comparative study of histone acetylation, histone deacetylation and ribonucleic acid synthesis in avian reticulocytes and erythrocytes. Biochemistry 12:783, 1973. 17. Monjardino JPPV, MacGillivray AJ: RNA and histone metabolism in small lymphocytes stimulated by phytohemagglutinin. Exp Cell Res 60:1, 1970. 18. Tidwell T, Allfrey VG, Mirsky AE: The methylation of histones during regeneration of the liver. J Biol Chem 243:707, 1968. 19. Brown JE, Adamson JW: Modulation of in vitro erythropoiesis: The influence of p-adrenergic agonists on erythroid colony formation. J Clin Invest 60:70, 1977. 20. Graber SE, Carriilo M, Krantz SB: Lack of effect of erythropoietin on cyclic adenosine - 3‘,5’ monophosphate levels in rat fetal liver cells. J Lab Clin Med 83:288, 1974. 21. Misiti J, Spivak JL: Erythropoiesis in vitro: Requirement for calcium. Clin Res 26:620A, 1978.
