- Email: [email protected]
Induction of commitment of murine erythroleukemia cells (TSA8) to CFU-E with DMSO
Experimental Cell Research 162 (1985) 319-325
Induction
of Commitment of Murine Erythroleukemia Cells (TSA8) to CFU-E with DMSO YUJI MISHINA
Faculty
and MASUO OBINATA*
of Pharmaceutical Sciences, University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113, Japan
The commitment of novel mouse erythroleukemic (MEL) cells (TSAI) to colonyforming units of erythroid (CFU-E) by dimethylsulfoxide (DMSO) was investigated. After exposure to the inducer in liquid culture, the cells were transferred to a semi-solid culture to examine their ability to form erythroid colonies which were dependent on erythropoietin. Exposure to DMSO for 2 days is optimum for CFU-E type colony formation and colonies induced in this manner are equivalent to CFU-E. The induction occurred in a synchronous manner. Partly stained colonies appeared prior to CFU-E formation and are thought to be a result of asymmetric cell division. Appearance of these partly stained colonies suggested that the number of erythropoietin receptors is important in the complete responsiveness to erythropoietin. TSA8 cells constitute a suitable model system in which to anafyse the mechanism of commitment in early erythropoiesis. @ 1986 Academic Ress, Inc.
Mammalian erythropoiesis is an ideal model system in which to study the regulation of determination and differentiation. Regulation of the rate of red blood cell formation may be achieved by control mechanisms that exert their effect at a number of critical steps. These include (I) proliferation of pluripotent hematopoietic stem cells (CFU-S); (ii) commitment of the hematopoietic stem cells to erythropoiesis; (iii) proliferation of committed erythroid precursor cells (namely, BFU-E (burst-forming unit of erythroid) and CFU-E (colony-forming unit of erythroid)); and (iv) commitment of the erythroid precursor cells to express the program of biosynthetic and morphogenetic activities characteristic of terminal differentiation of this specialized lineage [I, 21. Established mouse erythroleukemic (MEL) cells have been used to analyse the commitment to terminal differentiation. MEL cells established from mice infected with Friend virus complex appear to be arrested at the proerythroblastic stage of development [3]. These cells are not responsive to erythropoietin and, in this respect, differ from normal precursor cells. However, MEL cells can be induced in vitro to undergo a coordinated program resembling the final stage of normal erythroid differentiation by the addition of a variety of chemical reagents [3]. Commitment is associated with the entire set of biochemical changes related to the erythroid differentiation. Gusella et al. have shown that the commitment of MEL cells to differentiation * To whom offprint requests should be addressed. Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-%27/86 $03.00
320 Mishina
and Obinata
appears to operate in a stochastic manner [4]. In this work, a specific limitation of proliferative capacity of the MEL cells was used as a commitment assay and the biochemical parameters of differentiation and commitment could be dissociated
151. Recently, Shibuya & Mak established a novel type MEL cell line from the spleen of DBA/2 mice infected with anemia-inducing Friend virus complex (FVA) [6]. This line has the capacity to form colonies in semi-solid culture. In the presence of dimethylsulfoxide (DMSO) or erythropoietin, TSA8 cells form colonies like those derived from CFU-E. Since a continuous 5-day exposure of cells to DMSO in liquid culture induces hemoglobin synthesis [6], this cell line has the ability to differentiate erythroid cells that are similar to the previously established Friend erythroleukemia cells [3, 7, 81. However, this cell line is thought to be arrested at an earlier stage than previously described MEL cells for two reasons: its colony-forming ability in semi-solid culture and its responsibility to erythropoietin. Shibuya & Mak described that the level of hemoglobin-positive colonies of this line is maximum after 4-5 days in semi-solid culture [6]. However, normal CFU-E from mouse bone marrow or fetal liver exhibit maximum levels within 2 days. TSA8 cells may have weak responsiveness to erythropoietin or may be an earlier stage than the CFU-E stage. In any case, these cells may be suitable for the analysis of commitment and proliferation of the earlier precursor cells. In this work, we examined whether DMSO can induce the commitment of TSA8 to erythropoietin-dependent growth and differentiation; in other words, whether DMSO can convert the nature of this line to CFU-E. After DMSO was added to a liquid culture of TSA8 cells, cells were transferred at intervals to semisolid culture to assay the formation of erythropoietin-dependent colonies. The content of CFU-E increased with 2.0 days of incubation with DMSO in a synchronous manner. Partly stained colonies containing hemoglobin-positive and negative cells appeared to precede those of the CFU-E type colony. The former is thought to be the result of asymmetric cell division and could suggest that the number of erythropoietin receptors is important for complete responsiveness to erythropoietin. MATERIALS Cell Culture and Induction
AND METHODS
with DMSO
TSA8 is the MEL cell line established from mouse spleen infected with anemia-inducing Friend virus complex, FV-A [6]. We have obtained the cell line through Drs S. Nomura and M. Oishi, Institute for Applied Microbiology, University of Tokyo, from Dr T. Mak, University of Toronto. The cell line was maintained for one year in the Institute for Applied Microbiology, and then for 5 months in our laboratory. The MEL cells were grown in Iscove’s modified Dulbecco medium supplemented with 15% fetal bovine serum. In this condition, the cells have a saturation density of 3x lo6 cells/ml and the doubling time was approx. 13 h. To optimize the growth rate and the level of the induction, the lot of the fetal bovine serum was checked carefully. For induction, 1.0% (v/v) of DMSO (Merck) was added to the suspension of TSA8 cells one day after their passage at a density of 2x 16 cells/ml. Addition of DMSO at this concentration did not alter Exp CeNRes I62 (1986)
Induction of commitment of MEL cells to CFU-E
321
the growth rate of cells. When the cells were passaged, the cells in a slightly overgrowth gave better induction. On the other hand, the cells in the logarithmic phase gave poor induction. The addition of DMSO at a transition from lag phase to logarithmic phase gave the best induction. During the passage of the cells, the level of the induction sometimes decreased but the level could be maintained by the recloning of cells. The recloning was done in the semi-solid medium used for the assay of CFU-E, but without erythropoietin. The recloning in the semi-solid medium was more efficient than that in a liquid culture to get the clones showing high level of induction.
CFU-E Assay of the MEL Cells The assay was carried out in methylcellulose using the techniques described previously [9]. TSAI cells were transferred to a semi-solid medium at various times after the addition of DMSO. Fourx lo3 nucleated cells were plated in a 24-well multititer dish (Coming) in a mixture containing Iscove’s modified Dulbecco medium, 0.8 % methylcellulose (Tokyo Chemical Ind., Tokyo), 1% bovine serum albumin (Armour, Fraction V, USA), 30% fetal bovine serum (Filtoron, Inc.), 100 mM /I-mercaptoethanol and 0.5 U/ml of step III preparation of sheep plasma erythropoietin (Connaught, Canada). The numbers of hemoglobin-positive colonies and partly stained colonies increased according to the concentration of erythropoietin and reached a plateau level at a concentration of 0.5 U/ml. Therefore 0.5 U/ml erythropoietin was used throughout the experiments. In parallel experiments, 1% (v/v) DMSO was added to the semi-solid culture instead of erythropoietin. Dishes were incubated at 37°C in a humidified atmosphere flushed with 5 % CO2 in air. Colonies were directly stained with benzidine on day 2 and scored with an inverted microscope.
RESULTS The effect of DMSO on the colony-forming ability of the TSA8 cell line was investigated by the following two-phase culture experiment. In the first step, DMSO was added to the liquid culture of TSAS followed by incubation for several days. In the second step, cells were transferred to a semi-solid culture containing methylcellulose in a condition similar to that described by Iscove for mouse CFU-E assay [9], to measure their colony-forming ability. Two days after the transfer, the number of colonies were scored. As shown in fig. 1, colonies derived from TSAS cells could be classified into three types with benzidine dye. In the presence of erythropoietin or DMSO in semi-solid culture, hemoglobin-positive colonies were formed (fig. 1a). Without staining by benzidine dye for the presence of hemoglobins, such colonies were red and morphologically similar to those derived from normal CFU-E obtained from mouse fetal liver or bone marrow. In the absence of erythropoietin or DMSO in semi-solid culture, most of the colonies were not stained by benzidine dye (fig. 1 c). The larger size of the hemoglobin-negative colonies than that of the positive colonies makes it likely that the negative colonies consist of rather premature precursor cells, whose size is larger than the cells in the positive colonies [ 101. Partly stained colonies also appeared shortly after DMSO addition (fig. 1 b), in which hemoglobin-positive cells congregated in a cluster. These clustered positive cells may be derived from one progenitor cell, and the negative cells in other regions from other progenitor cells. TO optimize the condition of induction, we have checked carefully the lot of fetal bovine serum, the growth conditions of the cells, the concentration of Exp Cell Res 162 (1986)
322 Mishina and Obinata
Fig. I. Morphologies of colonies derived from DMSO-induced TSA8 cells. DMSO was added to liquid culture of TSA8 cells and cells were transferred to semi-solid culture containing methylcellulose with erythropoietin (0.5 U/ml) or DMSO (1%). Colonies derived from TSA8 cells could be classified in three types with benzidine dye: (a) hemoglobin-positive colonies like the CFU-E type; (b) partly stained colonies consisting of hemoglobin-positive and -negative clusters; (c) hemoglobin-negative colonies.
erythropoietin in the semi-solid culture and the induction protocol as described in Materials and Methods. In the optimum condition of the induction, the percentage of hemoglobin-positive colonies ranged from 13 to 18% in five trials. During the experiments, the level of the induction decreased, but its level could be recovered by recloning of the cells in the semi-solid medium. Fig, 2 shows the kinetics of the appearance of the hemoglobin-positive colonies following incubation with 1.0% (v/v) DMSO. As is clearly seen in fig. 2A, the number of CFU-E type colonies in the presence of erythropoietin increased upon DMSO addition, to reach a maximum level after 2 days in suspension culture with DMSO: this level was three or four times as high as that achieved in the absence of erythropoietin (fig. 2C). These results indicate that TSAS cells acquired the ability of erythropoietin-dependent colony formation with DMSO addition. This type of colony disappeared after 2 days in culture in semi-solid medium (data not shown). The properties of the CFU-E type colonies of induced TSA8 cells are probably similar to those of CFU-E in the following respect: (i) they consist of more than eight cells; (ii) are hemoglobin-positive; (iii) are erythropoietin-dependent; (iv) appear in 2 days and then disappear. The numbers of the CFU-E type colonies increased, depending on the concentration of erythropoietin, to reach a plateau level at a concentration of 0.5 U/ml. The minimum concentration of erythropoietin (0.5 U/ml) which gave a plateau level in TSAS cells was equivalent to that in the CFU-E from mouse bone marrows. The partly stained colony formation preceded this CFU-E type formation (1.5 days culture with DMSO) and the number of the partly stained colonies decreased with the DMSO addition. It may Exp Cell Res 162 (1986)
Induction of commitment of MEL cells to CFU-E
5I
I/=---
o-=
h
1.
incubation
323
time
I 3 dW
2 with
DMSO
Fig. 2. Kinetics of the appearance of hemoglobin-positive colonies depending on the incubation with
DMSO. DMSO (1.0%) was added to the liquid culture of TSAS cells one day after their passage at a density of 2x 10’ cells/ml, and at various periods after addition of DMSO, cells were transferred to a semi-solid culture containing methylcellulose and cultured at 37°C. After 2 days of culture, colonies were directly stained with benzidine dye and positive colonies were scored with an inverted microscope. Semi-solid culture was prepared with (A) 0.5 U/ml of erythropoietin; (B) 1.0% of DMSO; or (C) no inducer. (0) Content of CFU-E type colonies; (m) content of the partly stained colony against the total plated cell number. Abscissa shows the incubation time of culture with I .O% DMSO before transfer to semi-solid culture.
be a stage midway to the CFU-E-equivalent stage, and perhaps is formed as a result of asymmetric cell division (see Discussion). When the cells in the liquid culture were transferred to the semi-solid culture containing DMSO in place of erythropoietin, CFU-E type colonies were formed similarly (fig. 2 B). Thus, TSA8 cells can form hemoglobin-positive colonies even in the presence of DMSO. Actually, DMSO is more effective than erythropoietin. It is not known whether both have the same effect on the colony formation. In all cases shown in fig. 2A, C, the appearance of all three types of colonies described in fig. 1 was suppressed by more than 3 days’ incubation with DMSO (data not shown). It is likely that the cells lose their colony-forming ability as they differentiate toward a more mature stage by longer exposure to DMSO. Exp Cell Res 162 (1986)
324
Mishina
and Obinata
DISCUSSION We examined the effect of DMSO on the commitment of the novel MEL cell line (TSAS) to erythropoietin-dependent colony formation (CFU-E). The cells were committed to the CFU-E stage of erythroid by DMSO and a maximum level of commitment was observed after 2 days’ exposure to DMSO. This is the first report that responsiveness of the MEL cells to erythropoietin is modulated by a chemical reagent. Under our induced conditions, the entire population failed to respond to the colony formation, even after recloning of the cells: one reason for this is that the induced conditions were not optimal for complete responsiveness. In this respect, it will be interesting to learn whether other inducers used for the induction of differentiation of MEL cells are more effective for colony formation. Shorter exposure to DMSO induces partly stained colonies which contain two clustered cell populations, one cluster hemoglobin-negative and the other hemoglobin-positive. Appearance of the partly stained colony formation preceded formation of the CFU-E type colony. Cells forming the partly stained colonies may be a stage midway between the non-induced stage of TSA8 and the CFU-E equivalent stage. Both colony types appeared in a synchronous manner. When erythropoietin dependency in semi-solid culture was checked, the pattern of appearance of both colony types was similar in the different concentration, but the numbers of colonies in both types increased depending on its concentration. On day 2 when CFU-E type colony formation was maximum, the partly stained colonies decreased to a background level, a further indication that these partly stained colonies are precursor cells to the CFU-E cell stage. The shape of the partly stained colonies suggests that they are formed as a result of the asymmetric division of the TSAS cell. Asymmetric cell division is an attractive model for the generation of various kinds of cells from a multipotential stem cell [ 13, 141.The mechanism of asymmetric cell division for the commitment of TSA8 cells to CFU-E can be explained by the interaction between erythropoietin and its receptor. TSAS cells generated for 1.5 days of culture with DMSO still have a limited number of receptors for erythropoietin-dependent growth and differentiation. When seeded on a semi-solid medium containing erythropoietin, partly stained colonies appeared, when the receptors were localized in one of a pair of daughter cells of the first division; when evenly distributed in the two cells, hemoglobin-negative colonies appeared; the receptors may therefore not be sufficient to achieve the differentiation process in a single cell. Erythropoietin itself can induce CFU-E colonies in semi-solid culture, but requires longer cultivation than normal CFU-E assay and the colony number is low [6]. This may be due to the quite low level of erythropoietin receptors present in the uninduced TSA/8 cells. In a recent experiment we found that the number of receptors to erythropoietin in TSAS cells increased after induction. Before induction, its number was almost at background level, and increased to that of normal CFU-E after induction [15]. Sufficient exposure (2 days) of the cells to DMSO may induce this increase in number of erythropoietin receptors in the cells which is Exp Cell Res 162 (1986)
Induction of commitment of MEL cells to CFU-E
325
required for complete erythropoietin responsiveness; but with shorter exposure the number of receptors is not sufficient and the partly stained colony is then formed. In any case, the asymmetric cell division observed in the commitment of TSAS cells to CFU-E is a stimulating model with which to understand the mechanism of the commitment in hematopoiesis. After induction of TSA8 cells with DMSO in a liquid culture, the cells can form CFU-E colonies with erythropoietin and also with DMSO in a semi-solid medi-, urn. It is not known whether the mechanism of the CFU-E formation caused by erythropoietin differs from that caused by DMSO. Patterns of globin chains synthesized during the erythropoietin-dependent colony formation of bone marrow cells reportedly differ from those in peripheral reticulocytes [l 11or in MEL cells induced with DMSO [ 121.Our preliminary data indicated that the patterns of globin chains synthesized during the differentiation of TSAS are the same with both inducers. Thus, the routes of differentiation of the two inducers may also be the same. We thank Drs T. Mak, M. Oishi and S. Nomura for TSA8 cell lines, and Drs S. Natori, A. Urabe and T. Saito for helpful discussions. This work was partly supported by a grant-in-aid for scientific research from the Ministry of Education, Science, and Culture of Japan.
REFERENCES 1. Marks, P A, Differentiation of normal and neoplastic haematopoietic cells (ed B Clarkson, P A Marks & J E Till) p. 147. Cold Spring Harbor Laboratory, N.Y. (1978). 2. Harrison, P R, Nature 262 (1976) 353. 3. Friend, C, Scher, W, Holland, J G & Sato, T, Proc natl acad sci US 68 (1971) 378. 4. Gusella, J, Geller, R, Clarke, B, Weeks, V & Housman, D, Cell 9 (1976) 221. 5. Levenson, R, Kemen, J & Housman, D, Cell 18 (1979) 1073. 6. Shibuya, T L Mak, T W, Proc natl acad sci US 80 (1983) 3721. 7. Osterstag, W, Melderis, H, Steinheider, G, Kluge, N & Dube, S, Nature new biol239 (1972) 231. 8. Ross, J, Ikawa, Y & Leder, P, Proc natl acad sci US 69 (1972) 3620. 9. Iscove, N N, Sieber, F & Winterhalter, K H, J ceil physiol 83 (1974) 309. 10. Denton, M D & Amstein, H R V, Brit j haemato124 (1973) 7. 11. Alter, B P & Campbell, A S, Exp hematol 12 (1984) 611. 12. Rovera, G, Abramczuk, J & Surrey, S, FEBS lett 81 (1977) 366. 13. Sugimoto, M & Yasuda, T, Thymus 5 (1983) 297. 14. Osgood, E E, J natl cancer inst 18 (1957) 155. 15. Saito, T, Mishina, Y, Fukamachi, H, Obinata, M, Kasuga, M, Urabe, A & Takaku, F. In preparation. Received June 3, 1985 Revised version received August 27, 1985
Printed
in Sweden
Exp Cell Res 162 (1986)
Induction
of Commitment of Murine Erythroleukemia Cells (TSA8) to CFU-E with DMSO YUJI MISHINA
Faculty
and MASUO OBINATA*
of Pharmaceutical Sciences, University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113, Japan
The commitment of novel mouse erythroleukemic (MEL) cells (TSAI) to colonyforming units of erythroid (CFU-E) by dimethylsulfoxide (DMSO) was investigated. After exposure to the inducer in liquid culture, the cells were transferred to a semi-solid culture to examine their ability to form erythroid colonies which were dependent on erythropoietin. Exposure to DMSO for 2 days is optimum for CFU-E type colony formation and colonies induced in this manner are equivalent to CFU-E. The induction occurred in a synchronous manner. Partly stained colonies appeared prior to CFU-E formation and are thought to be a result of asymmetric cell division. Appearance of these partly stained colonies suggested that the number of erythropoietin receptors is important in the complete responsiveness to erythropoietin. TSA8 cells constitute a suitable model system in which to anafyse the mechanism of commitment in early erythropoiesis. @ 1986 Academic Ress, Inc.
Mammalian erythropoiesis is an ideal model system in which to study the regulation of determination and differentiation. Regulation of the rate of red blood cell formation may be achieved by control mechanisms that exert their effect at a number of critical steps. These include (I) proliferation of pluripotent hematopoietic stem cells (CFU-S); (ii) commitment of the hematopoietic stem cells to erythropoiesis; (iii) proliferation of committed erythroid precursor cells (namely, BFU-E (burst-forming unit of erythroid) and CFU-E (colony-forming unit of erythroid)); and (iv) commitment of the erythroid precursor cells to express the program of biosynthetic and morphogenetic activities characteristic of terminal differentiation of this specialized lineage [I, 21. Established mouse erythroleukemic (MEL) cells have been used to analyse the commitment to terminal differentiation. MEL cells established from mice infected with Friend virus complex appear to be arrested at the proerythroblastic stage of development [3]. These cells are not responsive to erythropoietin and, in this respect, differ from normal precursor cells. However, MEL cells can be induced in vitro to undergo a coordinated program resembling the final stage of normal erythroid differentiation by the addition of a variety of chemical reagents [3]. Commitment is associated with the entire set of biochemical changes related to the erythroid differentiation. Gusella et al. have shown that the commitment of MEL cells to differentiation * To whom offprint requests should be addressed. Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-%27/86 $03.00
320 Mishina
and Obinata
appears to operate in a stochastic manner [4]. In this work, a specific limitation of proliferative capacity of the MEL cells was used as a commitment assay and the biochemical parameters of differentiation and commitment could be dissociated
151. Recently, Shibuya & Mak established a novel type MEL cell line from the spleen of DBA/2 mice infected with anemia-inducing Friend virus complex (FVA) [6]. This line has the capacity to form colonies in semi-solid culture. In the presence of dimethylsulfoxide (DMSO) or erythropoietin, TSA8 cells form colonies like those derived from CFU-E. Since a continuous 5-day exposure of cells to DMSO in liquid culture induces hemoglobin synthesis [6], this cell line has the ability to differentiate erythroid cells that are similar to the previously established Friend erythroleukemia cells [3, 7, 81. However, this cell line is thought to be arrested at an earlier stage than previously described MEL cells for two reasons: its colony-forming ability in semi-solid culture and its responsibility to erythropoietin. Shibuya & Mak described that the level of hemoglobin-positive colonies of this line is maximum after 4-5 days in semi-solid culture [6]. However, normal CFU-E from mouse bone marrow or fetal liver exhibit maximum levels within 2 days. TSA8 cells may have weak responsiveness to erythropoietin or may be an earlier stage than the CFU-E stage. In any case, these cells may be suitable for the analysis of commitment and proliferation of the earlier precursor cells. In this work, we examined whether DMSO can induce the commitment of TSA8 to erythropoietin-dependent growth and differentiation; in other words, whether DMSO can convert the nature of this line to CFU-E. After DMSO was added to a liquid culture of TSA8 cells, cells were transferred at intervals to semisolid culture to assay the formation of erythropoietin-dependent colonies. The content of CFU-E increased with 2.0 days of incubation with DMSO in a synchronous manner. Partly stained colonies containing hemoglobin-positive and negative cells appeared to precede those of the CFU-E type colony. The former is thought to be the result of asymmetric cell division and could suggest that the number of erythropoietin receptors is important for complete responsiveness to erythropoietin. MATERIALS Cell Culture and Induction
AND METHODS
with DMSO
TSA8 is the MEL cell line established from mouse spleen infected with anemia-inducing Friend virus complex, FV-A [6]. We have obtained the cell line through Drs S. Nomura and M. Oishi, Institute for Applied Microbiology, University of Tokyo, from Dr T. Mak, University of Toronto. The cell line was maintained for one year in the Institute for Applied Microbiology, and then for 5 months in our laboratory. The MEL cells were grown in Iscove’s modified Dulbecco medium supplemented with 15% fetal bovine serum. In this condition, the cells have a saturation density of 3x lo6 cells/ml and the doubling time was approx. 13 h. To optimize the growth rate and the level of the induction, the lot of the fetal bovine serum was checked carefully. For induction, 1.0% (v/v) of DMSO (Merck) was added to the suspension of TSA8 cells one day after their passage at a density of 2x 16 cells/ml. Addition of DMSO at this concentration did not alter Exp CeNRes I62 (1986)
Induction of commitment of MEL cells to CFU-E
321
the growth rate of cells. When the cells were passaged, the cells in a slightly overgrowth gave better induction. On the other hand, the cells in the logarithmic phase gave poor induction. The addition of DMSO at a transition from lag phase to logarithmic phase gave the best induction. During the passage of the cells, the level of the induction sometimes decreased but the level could be maintained by the recloning of cells. The recloning was done in the semi-solid medium used for the assay of CFU-E, but without erythropoietin. The recloning in the semi-solid medium was more efficient than that in a liquid culture to get the clones showing high level of induction.
CFU-E Assay of the MEL Cells The assay was carried out in methylcellulose using the techniques described previously [9]. TSAI cells were transferred to a semi-solid medium at various times after the addition of DMSO. Fourx lo3 nucleated cells were plated in a 24-well multititer dish (Coming) in a mixture containing Iscove’s modified Dulbecco medium, 0.8 % methylcellulose (Tokyo Chemical Ind., Tokyo), 1% bovine serum albumin (Armour, Fraction V, USA), 30% fetal bovine serum (Filtoron, Inc.), 100 mM /I-mercaptoethanol and 0.5 U/ml of step III preparation of sheep plasma erythropoietin (Connaught, Canada). The numbers of hemoglobin-positive colonies and partly stained colonies increased according to the concentration of erythropoietin and reached a plateau level at a concentration of 0.5 U/ml. Therefore 0.5 U/ml erythropoietin was used throughout the experiments. In parallel experiments, 1% (v/v) DMSO was added to the semi-solid culture instead of erythropoietin. Dishes were incubated at 37°C in a humidified atmosphere flushed with 5 % CO2 in air. Colonies were directly stained with benzidine on day 2 and scored with an inverted microscope.
RESULTS The effect of DMSO on the colony-forming ability of the TSA8 cell line was investigated by the following two-phase culture experiment. In the first step, DMSO was added to the liquid culture of TSAS followed by incubation for several days. In the second step, cells were transferred to a semi-solid culture containing methylcellulose in a condition similar to that described by Iscove for mouse CFU-E assay [9], to measure their colony-forming ability. Two days after the transfer, the number of colonies were scored. As shown in fig. 1, colonies derived from TSAS cells could be classified into three types with benzidine dye. In the presence of erythropoietin or DMSO in semi-solid culture, hemoglobin-positive colonies were formed (fig. 1a). Without staining by benzidine dye for the presence of hemoglobins, such colonies were red and morphologically similar to those derived from normal CFU-E obtained from mouse fetal liver or bone marrow. In the absence of erythropoietin or DMSO in semi-solid culture, most of the colonies were not stained by benzidine dye (fig. 1 c). The larger size of the hemoglobin-negative colonies than that of the positive colonies makes it likely that the negative colonies consist of rather premature precursor cells, whose size is larger than the cells in the positive colonies [ 101. Partly stained colonies also appeared shortly after DMSO addition (fig. 1 b), in which hemoglobin-positive cells congregated in a cluster. These clustered positive cells may be derived from one progenitor cell, and the negative cells in other regions from other progenitor cells. TO optimize the condition of induction, we have checked carefully the lot of fetal bovine serum, the growth conditions of the cells, the concentration of Exp Cell Res 162 (1986)
322 Mishina and Obinata
Fig. I. Morphologies of colonies derived from DMSO-induced TSA8 cells. DMSO was added to liquid culture of TSA8 cells and cells were transferred to semi-solid culture containing methylcellulose with erythropoietin (0.5 U/ml) or DMSO (1%). Colonies derived from TSA8 cells could be classified in three types with benzidine dye: (a) hemoglobin-positive colonies like the CFU-E type; (b) partly stained colonies consisting of hemoglobin-positive and -negative clusters; (c) hemoglobin-negative colonies.
erythropoietin in the semi-solid culture and the induction protocol as described in Materials and Methods. In the optimum condition of the induction, the percentage of hemoglobin-positive colonies ranged from 13 to 18% in five trials. During the experiments, the level of the induction decreased, but its level could be recovered by recloning of the cells in the semi-solid medium. Fig, 2 shows the kinetics of the appearance of the hemoglobin-positive colonies following incubation with 1.0% (v/v) DMSO. As is clearly seen in fig. 2A, the number of CFU-E type colonies in the presence of erythropoietin increased upon DMSO addition, to reach a maximum level after 2 days in suspension culture with DMSO: this level was three or four times as high as that achieved in the absence of erythropoietin (fig. 2C). These results indicate that TSAS cells acquired the ability of erythropoietin-dependent colony formation with DMSO addition. This type of colony disappeared after 2 days in culture in semi-solid medium (data not shown). The properties of the CFU-E type colonies of induced TSA8 cells are probably similar to those of CFU-E in the following respect: (i) they consist of more than eight cells; (ii) are hemoglobin-positive; (iii) are erythropoietin-dependent; (iv) appear in 2 days and then disappear. The numbers of the CFU-E type colonies increased, depending on the concentration of erythropoietin, to reach a plateau level at a concentration of 0.5 U/ml. The minimum concentration of erythropoietin (0.5 U/ml) which gave a plateau level in TSAS cells was equivalent to that in the CFU-E from mouse bone marrows. The partly stained colony formation preceded this CFU-E type formation (1.5 days culture with DMSO) and the number of the partly stained colonies decreased with the DMSO addition. It may Exp Cell Res 162 (1986)
Induction of commitment of MEL cells to CFU-E
5I
I/=---
o-=
h
1.
incubation
323
time
I 3 dW
2 with
DMSO
Fig. 2. Kinetics of the appearance of hemoglobin-positive colonies depending on the incubation with
DMSO. DMSO (1.0%) was added to the liquid culture of TSAS cells one day after their passage at a density of 2x 10’ cells/ml, and at various periods after addition of DMSO, cells were transferred to a semi-solid culture containing methylcellulose and cultured at 37°C. After 2 days of culture, colonies were directly stained with benzidine dye and positive colonies were scored with an inverted microscope. Semi-solid culture was prepared with (A) 0.5 U/ml of erythropoietin; (B) 1.0% of DMSO; or (C) no inducer. (0) Content of CFU-E type colonies; (m) content of the partly stained colony against the total plated cell number. Abscissa shows the incubation time of culture with I .O% DMSO before transfer to semi-solid culture.
be a stage midway to the CFU-E-equivalent stage, and perhaps is formed as a result of asymmetric cell division (see Discussion). When the cells in the liquid culture were transferred to the semi-solid culture containing DMSO in place of erythropoietin, CFU-E type colonies were formed similarly (fig. 2 B). Thus, TSA8 cells can form hemoglobin-positive colonies even in the presence of DMSO. Actually, DMSO is more effective than erythropoietin. It is not known whether both have the same effect on the colony formation. In all cases shown in fig. 2A, C, the appearance of all three types of colonies described in fig. 1 was suppressed by more than 3 days’ incubation with DMSO (data not shown). It is likely that the cells lose their colony-forming ability as they differentiate toward a more mature stage by longer exposure to DMSO. Exp Cell Res 162 (1986)
324
Mishina
and Obinata
DISCUSSION We examined the effect of DMSO on the commitment of the novel MEL cell line (TSAS) to erythropoietin-dependent colony formation (CFU-E). The cells were committed to the CFU-E stage of erythroid by DMSO and a maximum level of commitment was observed after 2 days’ exposure to DMSO. This is the first report that responsiveness of the MEL cells to erythropoietin is modulated by a chemical reagent. Under our induced conditions, the entire population failed to respond to the colony formation, even after recloning of the cells: one reason for this is that the induced conditions were not optimal for complete responsiveness. In this respect, it will be interesting to learn whether other inducers used for the induction of differentiation of MEL cells are more effective for colony formation. Shorter exposure to DMSO induces partly stained colonies which contain two clustered cell populations, one cluster hemoglobin-negative and the other hemoglobin-positive. Appearance of the partly stained colony formation preceded formation of the CFU-E type colony. Cells forming the partly stained colonies may be a stage midway between the non-induced stage of TSA8 and the CFU-E equivalent stage. Both colony types appeared in a synchronous manner. When erythropoietin dependency in semi-solid culture was checked, the pattern of appearance of both colony types was similar in the different concentration, but the numbers of colonies in both types increased depending on its concentration. On day 2 when CFU-E type colony formation was maximum, the partly stained colonies decreased to a background level, a further indication that these partly stained colonies are precursor cells to the CFU-E cell stage. The shape of the partly stained colonies suggests that they are formed as a result of the asymmetric division of the TSAS cell. Asymmetric cell division is an attractive model for the generation of various kinds of cells from a multipotential stem cell [ 13, 141.The mechanism of asymmetric cell division for the commitment of TSA8 cells to CFU-E can be explained by the interaction between erythropoietin and its receptor. TSAS cells generated for 1.5 days of culture with DMSO still have a limited number of receptors for erythropoietin-dependent growth and differentiation. When seeded on a semi-solid medium containing erythropoietin, partly stained colonies appeared, when the receptors were localized in one of a pair of daughter cells of the first division; when evenly distributed in the two cells, hemoglobin-negative colonies appeared; the receptors may therefore not be sufficient to achieve the differentiation process in a single cell. Erythropoietin itself can induce CFU-E colonies in semi-solid culture, but requires longer cultivation than normal CFU-E assay and the colony number is low [6]. This may be due to the quite low level of erythropoietin receptors present in the uninduced TSA/8 cells. In a recent experiment we found that the number of receptors to erythropoietin in TSAS cells increased after induction. Before induction, its number was almost at background level, and increased to that of normal CFU-E after induction [15]. Sufficient exposure (2 days) of the cells to DMSO may induce this increase in number of erythropoietin receptors in the cells which is Exp Cell Res 162 (1986)
Induction of commitment of MEL cells to CFU-E
325
required for complete erythropoietin responsiveness; but with shorter exposure the number of receptors is not sufficient and the partly stained colony is then formed. In any case, the asymmetric cell division observed in the commitment of TSAS cells to CFU-E is a stimulating model with which to understand the mechanism of the commitment in hematopoiesis. After induction of TSA8 cells with DMSO in a liquid culture, the cells can form CFU-E colonies with erythropoietin and also with DMSO in a semi-solid medi-, urn. It is not known whether the mechanism of the CFU-E formation caused by erythropoietin differs from that caused by DMSO. Patterns of globin chains synthesized during the erythropoietin-dependent colony formation of bone marrow cells reportedly differ from those in peripheral reticulocytes [l 11or in MEL cells induced with DMSO [ 121.Our preliminary data indicated that the patterns of globin chains synthesized during the differentiation of TSAS are the same with both inducers. Thus, the routes of differentiation of the two inducers may also be the same. We thank Drs T. Mak, M. Oishi and S. Nomura for TSA8 cell lines, and Drs S. Natori, A. Urabe and T. Saito for helpful discussions. This work was partly supported by a grant-in-aid for scientific research from the Ministry of Education, Science, and Culture of Japan.
REFERENCES 1. Marks, P A, Differentiation of normal and neoplastic haematopoietic cells (ed B Clarkson, P A Marks & J E Till) p. 147. Cold Spring Harbor Laboratory, N.Y. (1978). 2. Harrison, P R, Nature 262 (1976) 353. 3. Friend, C, Scher, W, Holland, J G & Sato, T, Proc natl acad sci US 68 (1971) 378. 4. Gusella, J, Geller, R, Clarke, B, Weeks, V & Housman, D, Cell 9 (1976) 221. 5. Levenson, R, Kemen, J & Housman, D, Cell 18 (1979) 1073. 6. Shibuya, T L Mak, T W, Proc natl acad sci US 80 (1983) 3721. 7. Osterstag, W, Melderis, H, Steinheider, G, Kluge, N & Dube, S, Nature new biol239 (1972) 231. 8. Ross, J, Ikawa, Y & Leder, P, Proc natl acad sci US 69 (1972) 3620. 9. Iscove, N N, Sieber, F & Winterhalter, K H, J ceil physiol 83 (1974) 309. 10. Denton, M D & Amstein, H R V, Brit j haemato124 (1973) 7. 11. Alter, B P & Campbell, A S, Exp hematol 12 (1984) 611. 12. Rovera, G, Abramczuk, J & Surrey, S, FEBS lett 81 (1977) 366. 13. Sugimoto, M & Yasuda, T, Thymus 5 (1983) 297. 14. Osgood, E E, J natl cancer inst 18 (1957) 155. 15. Saito, T, Mishina, Y, Fukamachi, H, Obinata, M, Kasuga, M, Urabe, A & Takaku, F. In preparation. Received June 3, 1985 Revised version received August 27, 1985
Printed
in Sweden
Exp Cell Res 162 (1986)