Control of Normal Differentiation of Myeloid Leukemic Cells XI. INDUCTION OF A SPECIFIC REQUIREMENT FOR CELL VIABILITY AND GROWTH DURING THE DIFFERENTIATION OF MYELOID LEUKEMIC CELLS EITAN FIHACH AND LEO SACHS Department of Genetics, Weizmann lnstitute of Science, Rehovot, lsrael
ABSTRACT Normal hematopoietic cells require the presence of a protein (MGI) in the appropriate conditioned medium (CM) for cell viability and growth and for differentiation to mature macrophages and granulocytes. Clones of myeloid leukemic cells have been established in culture ( D clones) which require CM with this protein for differentiation, but not for cell viability and growth. It has been shown that these leukemic cells can be induced by CM to again require, like normal cells, the presence of CM for cell viability and growth. Induction of this requirement, which will be referred to as RVG, occurred before the D' cells differentiated to mature granulocytes. Clones of myeloid leukemic cells (D- clones) that could not be induced to differentiate to mature cells, did not show the induction of RVG. The steroid hormones prednisolone and dexamethasone can induce some, but not all the changes associated with differentiation of D' cells. Incubation with these steroids did not result in the induction of a requirement for these steroids for cell growth and viability. Studies with CM from different sources have shown, that all batches that induced RVG also induced differentiation of D+ cells and that both activities were inhibited after treating the CM with trypsin. It is suggested that the same protein (MGI) may lie involved in both activities. Incubation of D+ cells with CM resulted in an increase in agglutinability by concanavalin A and this increase was maintained even in the absence of CM. This suggests, that the induction of RVG in D' myeloid leukemic cells is associated with a change in the cell surface membrane.
An important question in the control of cell differentiation is the relationship between cell multiplication and differentiation. In order to analyze this relationship, it would be useful to have an experimental system in which the two processes can be experimentally separated. We have now developed such a system with hematopoietic cells. Normal hematopoietic cells are not viable and cannot multiply or differentiate to mature macrophages and granulocytes without the addition of a specific protein which we call (MGI) that can be obtained in the appropriate conditioned medium (CM) (Sachs, '74a). We have established clones of myeloid leukemic cells that are viable and can multiply in the absence of J. CELL. PHYSIOL., 89: 259-266
this protein (Fibach et al., '73). Some of these clones (D+clones) can be induced by MGI to differentiate to mature macrophages and granulocytes while other clones (D- clones) could not be induced to differentiate to mature cells (Fibach et al., '72, '73; Sachs, '74a,b; Fibach and Sachs, '75). This raises the question whether the D cells can be induced to again require, like normal cells, the presence of MGI for cell viability and multiplication. In order to test this possibility, we have cultured D' and D- cells in the presence of the appropriate CM and then transferred the cells to medium without CM. The experiments have shown that D+ cells during the Received Nov. 25, '75. Accepted Feb. 26, '76.
259
260
EITAN FIBACH AND LEO SACHS
process of differentiation again show, like normal cells, a requirement for CM for cell viability and growth. This requirement of CM for viability and growth will be referred to as RVG. MATERIALS AND METHODS
in 1.7 ml 0.33%agar on a 5 ml base of 0.5% agar per 50 mm petri dish in EM with 20% inactivated horse serum (Pluznik and Sachs, '65). The number of colonies with more than 50 cells was counted. Conditioned medium (CM). Serum free CM from lungs and CM with serum from embryo fibroblasts and human spleen were prepared as described (Fibach and Sachs, '75). CM from mature granulocytes was prepared with cells obtained 16 hours after intraperitoneal inoculation into adult mice of 3 ml of 10%sodium caseinate solution (Difco lab., Detroit) in phosphatebuffered saline (PBS).The cells were collected by washing; the peritoneum with PBS and consisted of at least 95% mature granulocytes. After washing with EM, 5 x lo7cells were seeded per 50 mm petri dish in EM without serum and CM was collected three days later. CM from D+ and D- cells was prepared by seeding 5 x lo6 cells per petri dish and the CM was collected after three days incubation in EM with 10% horse serum. Serum from normal mice and from mice after injection of bacterial endotoxin (hletcalf, '71) were prepared as described (Fibach and Sachs, '74). All cultures were incubated in a 37°C incubator with a constant flow of 10%CO,. Concanavalin A (Con A) induced agglutination. Agglutination induced by 200 pgiml Con A (Miles-Yeda, Rehovot) was measured using the Fragiligraph (Model D2). This instrument gives a continuous stirring of the ceill suspension and the degree of cell agglutination is given by the log of light transmission continuously recorded (Vlodavsky and Sachs, '75). Number of experiments. All results are based on at least Ihree separate experiments.
Cells and cell culture. D' and D- clones were isolated and three times clone purified (Fibach et al., '73) without CM from an established cell line derived from a myeloid leukemia in a SL mouse (Ichikawa, '691. Five D' clones (Nos. 3, 9, 11,12 and 24) and five 13- clones (Nos. 2,5, 16, 18 and 19) with stable phenotypes (Hayashi et al., '74) were used in the present experiments. Similar results were obtained with these five independently isolated clones of each cell type. The cells grew in suspension in liquid medium without CM as myeloblasts or some promyelocytes. The blast cells from all ten clones produced myeloid leukemia after intravenous inoculation into adult mice. Cells were generally subcultured every 4-5 days by seeding 5 x 105 cells per 50 mm petri dish in 5 ml Eagle's medium with a 4fold concentration of amino acids and vitamins (EM) (H-21, Grand Island Biological Co., New York) and 10% inactivated (56°C for 30 minutes) horse serum. To determine the CM requirement for viability and growth (RVG) 5 x 105 cells per 50 mm petri dish were incubated in EM with 10%horse serum and 25%conditioned medium (CM) for three days, unless otherwise stated, then washed three times with EM and subcultured at 2 x 106 cells per petri dish in medium with or without the addition of new 25% CM. The cells were then counted, generally three days later, in eosin solution (1:2,000) and the number of non-stained cells was recorded. RESULTS The reproducibility of the results was Znduction of a requirement for CM for generally up to k 20% when the number of viability and growth (RVG) in D+ cells was 4 x lo5 or more per milliliter, myeloid leukemic cells durand up to k 30%when the number of cells ing diffirentiation was less than 4 x 105 per milliliter. Granulocyte differentiation in mass culture was D' and D- myeloid leukemic cells are induced by CM as described (Fibach and viable and can multiply without the addiSachs, '75). Cells were seeded for cloning tion of CM. Three days incubation with
GROWTH AND IXFFERENTIATION OF LEUKEMIC CELLS
25% CM from lungs of mice injected with endotoxin (endotoxin lung CM) gave a 43% decrease in the number of D cells and a 47% increase in the number of D- cells compared to cells growing in medium without CM (Fibach and Sachs, '75).At this time, D cultures contained 80% myeloblasts and promyelocytes and 20% cells in intermediate stages of granulocyte differentiation from myelocytes to stab forms,
but no mature granulocytes. D- cultures showed 100% myeloblasts and promyelocytes and no intermediate stages or mature granulocytes. Transfer of these D' and D- cells to medium with new CM resulted in a continuation of cell multiplication, whereas transfer to medium without CM showed a decrease in the total cell number and a loss of viability in D' but not in D- cells (fig. 1).This requirement of CM
D a y s after subculture
Fig. 1 Number of D and D- cells after transfer to medium with or without CM. The cells were incubated with 25% endotoxin lung CM for three days, before transfer to medium with ( 0 ) or withCM. 0 = D' or D- cells that had never been cultured with CM. out (0)
c u) I
-0 d
E L
2-
26 1
262
EITAN FIHACH AND LEO SACHS
About 8% of the D' cells incubated for three days with CM and then transferred to medium without CM were still present after two days in this medium (fig. 1). Culture of these surkivors for two weeks in medium without CM showed, that these surviving cells were still D' cells as regards their ability to differentiate and their ability to be induced for RVG.
for viability and growth will be referred to as RVG. Transfer of D' cells to medium without CM after one, two or three days incubation with CM has shown, that the optimal induction of RVG was found after three days incubation with CM (fig. 2). D' cells after three days incubation with CM had completely lost their ability to form colonies in agar without CM and required los instead of 1 0 2 cells to induce 100%leukemia in isoloeous mice. Incubation of D- cells with C 6 did not result in any decrease in colony formation in agar or leukemogenicity in vivo (table 1).
Znduction of RVG and induction of cell d iffc'rentiation Incubation of I)' cells for three days with 25% endotoxin lung CM in addition to inducing RVG, induces rosette formation
TAHLE 1
Cloning efficiency in agar and leukemogenicity in vivo of D and I, cells hefore and after treatment with endotoxin lung CM Percent animals with leukemia Cell type Incubation with CM
Cloning efficiency
102
Number af cells inoculated 103 104
10'
I%)
DD' DD-
19+ 4 0
. .
+ __ +
100 0 100 100
22+ 5 20+ 7
100 0 100 100
100 80 100 100
100 100 100 100
D' and D- cells were incubated for three days with 25% endotoxin lung CM. The (cells were washed with PBS, counted and seeded in agar at 1,000 cells per 50 mm petri dish or injected intravenously into 2-month old female mice, five mice per point. Colonies in agar were counted after 12 days and animals were observed for three months after cell inoculation. All experiments were repeated three times. TAHLE 2
Znduction of RVG and other cellular changes associated with differentiation Granulocyte differentiation to Cell type Incubation with
D-
D D'
CM CM Trypsinized CM
Intermediate stages
Mature granulocytes
(%)
I%)
8 42 5
0
33 0
Rosette formation (percent) Fc
C3
Cells with phagocytosis
I %)
15 25 2
40 61
27
3
47 7
Cell at1:achnient tcl the Fetri dish
Cloning efficiency in agar
I%) 2 20 1
f%) 20 0 18
RVG
(Sbl
80 12
67
The cells were incubated with 25% endotoxin lung CM. Morphological differentiation to intermediate stages (myelocyte to stab form) or mature granulocytes was determined after ten days incubation. All the other properties were determined after three days incubation. Rosette formation was tested with sheep erythrocytes coated with rabbit anti sheep antibody (Fc) or with this antibody plus mouse complement (C3) (Lotem and Sachs, '74). About 1,000 cells were counted in each experiment for rosette formation. The reproducibility of rosette formation was with lower frequengenerally + 10%when the percentage of rosette forming cells was higher than 10%.and r ~ 25% cies of rosette forming cells. Phagocytosis of latex particles (Hacto-latex 0.81 p) and cell attachment to the surface of the petri dish, after staining with May-Grunwald Giemsa, were tested as described (Fihach and Sachs, '75). RVG is expressed as a percentage of the number of cells after transfer to medium without CM compared to the number of cells in medium with CM.
263
GROWTH AND DIFFERENTIATION O F LEUKEMIC CELLS
3
c u)
I
0
2
2
'5 L
aI
a
0
I
I
I
2
I 3
Doys ofter subcullure
Fig. 3 Number of D' cells after incubation in medium containing prednisolone. The cells were transferred to medium with (01 or without prednisolone after three days incubation with 0.4 pg/ml prednisolone ( 1. Similar results were obtained with 0.4pg/ml dexamethasone. TABLE 3
Induction of RVG and differentiation in D cells and colony formation with normal bone marrow cells by CM and serum from different sources Granulocyte differentiation to Source of CM or serum
Normal mouse serum D' or D- myeloid leukemic cells Mouse embryo fibroblasts Mouse normal granulocytes Lungs from endotoxin injected mice Serum from endotoxin injected mice Rat embryo fibroblasts Human spleen cells
No. of colonies per 4 x 1O'hone marrow cells
RVG
Intermediate stages
Mature manulocvtes
(%I 94 89
(%I 0 0
(76)
0 0
0 0
19 27 11
12 25 38
2 32 47
92 0 145
17
27
5
112
14 13
33 35
41 40
21 ' 32
RVC was determined after incubation of D cells for three days with 25% CM or 2.5%mouse serum and then transferring the cells to medium with or without CM or serum of the same type used for the primary culture. The results are expressed as in the legends to table 2. Granulocyte differentiation was determined after 10 days incubation with 25% CM or 2.5%mouse serum. Colony formation with normal hematopoietic cells (Pluznik and Sachs, '65, Ichikawa et al., '66; Bradley and Metcalf, '66; Worton et al., '69) was assayed by seeding 4 x lWbone marrow cells from 2 month old C57B1/6 mice in 0.33% agar on an 0.5% agar base containing 25%CM or 2.5%mouse serum. Colonies containing more than 50 cells were counted after seven days incubation. I Results from the two out of five batches of CM from rat fibroblasts and the three out of six batches of CM from human spleen, that induced colonies with normal bone marrow cells.
264
EITAN FIBACH A N D LEO SACHS
for Fc and C3 receptors (Lotem and Sachs, '741, phagocytosis of latex particles and cell attachment to the surface of a petri dish (Fibach and Sachs, '75) and 20% cells in intermediate stages of granulocyte differentiation. Ten days incubation with this CM resulted in 42% cells in intermediate stages and 33% mature granulocytes (table 2). Dcells showed a lower induction of Fc and C3 rosettes and almost none of the other changes found in D' cells (table 2). The serum free endotoxin lung CM that induced RVG was incubated with 400 pg/ml crystalline trypsin (Sigma Chemical Co., St. Louis) at 37°C for two hours followed by addition of the same concentration of soybean trypsin inhibitor (Sigma Chemical Co., St. Louis). This resulted in an inhibition of the ability of the CM to induce RVG and inhibition of the other cellular changes associated with cell differentiation (table 2). Incubation of this CM with trypsin and trypsin inhibitor without pre-incubation with trypsin, did not change the activity of this CM. These results indicate that the factor in CM that induces RVG, like the protein that induces differentiation, is trypsin sensitive. The steroid hormones prednisolone and dexamethasone can induce some, but not
all the changes associated with differentiation of D' leukemic cells (Lotem and Sachs, '75). Incubation of D+ cells with 0.4-4 pg/ml of these hormones induced these changes, but did not induce a requirement for these hormones for growth and viability (fig. 3).
CM from dijferent sources CM from different sources, including different species, were tested for their ability to induce RVG and differentiation in D' cells and collony formation with normal bone marrow cells. The results indicate (table 3) that CM from rat and mouse embryo fibroblasts, human spleen and normal mouse granulocytes, like endotoxin lung CM, induced RVG and various degrees of differentiation to mature granulocytes. There was no induction of RVG or differentiation by CM from D' or Dmyeloid leukemic cells. Serum from mice injected with bacterial endotoxin, but not normal mouse serum, also induced RVG and differentiation. The majority {of CM from different sources and serum from endotoxin injected mice induced the formation of colonies with mature macrophages and granulocytes from normal bone marrow cells.
T I me (minutes)
Fig. 4 Con A induced agglutinability of D' cells after transfer to medium with or without CM. Cells were incubated for three days with 25% endotoxin lung CM and assayed for agglutinability one day after (A) transfer to medium with 25% CM (B) transfer to medium without CM, (C) cells which were not treated with CM. Con A solution was added at a final concentration of 200 &ml and the light transmission automatically recorded.
GROWTH AND DIFFERENTIATION OF LEUKEMIC CELLS
However three out of five batches of CM from rat embryo fibroblasts, three out of six batches of CM from human spleen and six out of six batches of CM from normal mature granulocytes, induced RVG and differentiation in D' cells, but did not induce the formation of colonies from normal bone marrow cells (table 3 ) .
Agglutinability by concanavalin A (Con A ) Incubation of D' cells for three days with 25% endotoxin lung CM results in an increase in the ability of the cells to be agglutinated by Con A (Vlodavsky et al., '76). Cells tested one day after transfer to medium without CM, before there was a considerable loss of cell viability, still showed the high Con A agglutinability of the cells incubated in medium with CM (fig. 4). DISCUSSION
Normal hematopoietic cells require a protein, which we call MGI, for cell viability, growth and differentiation to mature macrophages and granulocytes (Sachs, '74a). Clones of myeloid leukemic cells have been established (D+1 which require CM with this protein for differentiation to macrophages and granulocytes but do not require the CM for cell viability and growth. The present experiments have shown, that D' myeloid leukemic cells during the process of differentiation again require the presence of CM for cell viability and growth (RVG).The induction of RVG occurred before the appearance of mature cells and resulted in a loss of cloning efficiency in agar in the absence of CM. The development of leukemia in 100% of the animals required the inoculation of lo5 cells induced for RVG compared to lo2 control cells. D- myeloid leukemic cells which could not be induced to differentiate to mature cells by the same CM, were also not induced for RVG. Incubation of serum-free CM with trypsin inhibited both the ability to induce RVG and the ability to induce other cellular changes associated with differentiation. It may be that the
265
same protein (MGI) induces RVG and the differentiation of D' cells. It can also be suggested that the change in the cells resulting in a decrease or lack of requirement for this protein for viability and growth, may be one of the causes of myeloid leukemia. Experiments with CM from different sources and species have shown, that RVG in mouse D+ cells can also be induced b y CM from rat and human cells. All the CM from different sources and serum from mice injected with bacterial endotoxin that induced RVG also induced differentiation of D+ cells. This supports the suggestion that the same protein may be involved in both processes. Although the majority of CM from different sources also induced the formation of colonies with macrophages and granulocytes from normal bone marrow cells, some batches of CM from rat embryo fibroblasts and human spleen cells and all batches of CM from normal mature granulocytes, induced RVG and differentiation in D' cells, but did not induce the fomation of colonies from normal bone marrow cells, This suggests that there may be co-factor(s) (Landau and Sachs, '71a,b) required for colony formation with normal bone marrow cells that are absent in these batches of CM. D' but not D- cells show an increased cell agglutinability by Con A after incubation with the appropriate CM (Vlodavsky et al., '76). The present experiments have shown that this increase in cell agglutinability was not reversed after the cells were transferred to medium without CM. This suggests that the specific requirement for viability and growth induced in D+ cells during differentiation, is associated with a change in the cell surface membrane. ACKNOWLEDGMENT
We thank Mrs. Esther Gerassi for skillful technical assistance. This work was supported by grants from the Talisman Foundation and the Jerome and Estelle R. Newman Assistance Fund, New York.
266
EITAN FIBACH AND LEO SACHS
LITERATURE CITED Bradley, T. R., and D. Metcalf 1966 The growth of mouse bone marrow cells in uitro. Aust. J. Exp. Biol. Med. Sci., 44: 287-299. Fibach, E., M. Hayashi and I,. Sachs 1973 Control of normal differentiation of myeloid leukemic cells to macrophages and granulocytes. Proc. Natl. Acad. Sci. ( U S A . ) , 70: 343-346. Fibach, E., T. Landau and L. Sachs 1972 Normal differentiation of myeloid leukemic cells induced by a differentiation inducing protein. Nature, New Biol., 237: 276-278. Fihach, E., and L. Sachs 1974 Control of normal differentiation of myeloid leukemic cells. IV. Induction of differentiation by serum from endotoxin treated mice. J. Cell. Physiol., 83: 177-186. 1975 Control of normal differentiation of myeloid leukemic cells. VIII. Induction of differentiation to mature granulocytes in mass culture. J. Cell. Physiol., 86: 221-230. Hayashi, M., E. Fibach and L. Sachs 1974 Control of normal differentiation of myeloid leukemic cells. V. Normal differentiation in aneuploid leukemic cells and the chromosome banding pattern of D’ and Dclones. Int. J. Cancer, 14: 40-48. Ichikawa, Y. 1969 Differentiation of a cell line of myeloid leukemia. J. Cell. Physiol., 74: 223-234. Ichikawa, Y., D. H. Pluznik and L. Sachs 1966 In uitro control of the development of macrophage and granulocyte colonies. Proc. Natl. Acad. Sci. ( U S A . ) , 56: 488-495. Landau, T., and L. Sachs 1971a Characterization of the inducer required for the development of macrophage and granulocyte colonies. P r k . Natl. Acad. Sci. ( U S A . ) , 68: 2540-2544.
1971b Activation of a differentiation inducing protein by adenine and adenine containing nucleotides. FEBS Letters, 17: 339-341. Lotem, J,, and L. Sachs 1974 Different blocks in the differentiation of myeloid leukemic cells. Proc. Natl. Acad. Sci. (USA,), 71: 3507-3511. 1975 Induction of specific changes in the surface membrane of myeloid leukemic cells by steroid hormones. Int. J. Cancer, 15: 731-740. Metcalf, D. 1971 Acute antigen induced elevation of serum colony stimlulating factor (CSF)levels. Immunology, 21: 427-436. Pluznik, D. H., and L. Sachs 1965 The cloning of normal “mast” cells in tissue culture. J. Cell. Comp. Physiol., 66: 319-324. Sachs. L. 1974a Regulation of membrane changes, differentiation and inalignancy in carcinogenesis. Harvey Lectures, Academic Press, New York, 68: 1-35. 19741, Control of growth and differentiation in normal hematopaietic and leukemic cells. In: Control of Proliferation in Animal Cells. Cold Spring Harbor Lab. New York, pp. 915-925. Vlodavsky, I., E. Fibach and L. Sachs 1976 Control of normal differentiation of myeloid leukemic cells. X. Glucose utilization, cellular ATP and associated membrane changes in D and D- cells. J. Cell. Physiol., 87: 167-177. Vlodavsky, I., and L. Siichs 1975 Lectin receptors on the cell surface membrane and the kinetics of lectin induced agglutination. Exp. Cell Res., 93: 11 1-1 19. Worton, R. G., E. A. McCulloch and J. E. Till 1969 Physical separation of hematopoietic stem cells from cells forming colonies in culture. 1. Cell. Physiol., 74: 171-182
~
ABSTRACT Normal hematopoietic cells require the presence of a protein (MGI) in the appropriate conditioned medium (CM) for cell viability and growth and for differentiation to mature macrophages and granulocytes. Clones of myeloid leukemic cells have been established in culture ( D clones) which require CM with this protein for differentiation, but not for cell viability and growth. It has been shown that these leukemic cells can be induced by CM to again require, like normal cells, the presence of CM for cell viability and growth. Induction of this requirement, which will be referred to as RVG, occurred before the D' cells differentiated to mature granulocytes. Clones of myeloid leukemic cells (D- clones) that could not be induced to differentiate to mature cells, did not show the induction of RVG. The steroid hormones prednisolone and dexamethasone can induce some, but not all the changes associated with differentiation of D' cells. Incubation with these steroids did not result in the induction of a requirement for these steroids for cell growth and viability. Studies with CM from different sources have shown, that all batches that induced RVG also induced differentiation of D+ cells and that both activities were inhibited after treating the CM with trypsin. It is suggested that the same protein (MGI) may lie involved in both activities. Incubation of D+ cells with CM resulted in an increase in agglutinability by concanavalin A and this increase was maintained even in the absence of CM. This suggests, that the induction of RVG in D' myeloid leukemic cells is associated with a change in the cell surface membrane.
An important question in the control of cell differentiation is the relationship between cell multiplication and differentiation. In order to analyze this relationship, it would be useful to have an experimental system in which the two processes can be experimentally separated. We have now developed such a system with hematopoietic cells. Normal hematopoietic cells are not viable and cannot multiply or differentiate to mature macrophages and granulocytes without the addition of a specific protein which we call (MGI) that can be obtained in the appropriate conditioned medium (CM) (Sachs, '74a). We have established clones of myeloid leukemic cells that are viable and can multiply in the absence of J. CELL. PHYSIOL., 89: 259-266
this protein (Fibach et al., '73). Some of these clones (D+clones) can be induced by MGI to differentiate to mature macrophages and granulocytes while other clones (D- clones) could not be induced to differentiate to mature cells (Fibach et al., '72, '73; Sachs, '74a,b; Fibach and Sachs, '75). This raises the question whether the D cells can be induced to again require, like normal cells, the presence of MGI for cell viability and multiplication. In order to test this possibility, we have cultured D' and D- cells in the presence of the appropriate CM and then transferred the cells to medium without CM. The experiments have shown that D+ cells during the Received Nov. 25, '75. Accepted Feb. 26, '76.
259
260
EITAN FIBACH AND LEO SACHS
process of differentiation again show, like normal cells, a requirement for CM for cell viability and growth. This requirement of CM for viability and growth will be referred to as RVG. MATERIALS AND METHODS
in 1.7 ml 0.33%agar on a 5 ml base of 0.5% agar per 50 mm petri dish in EM with 20% inactivated horse serum (Pluznik and Sachs, '65). The number of colonies with more than 50 cells was counted. Conditioned medium (CM). Serum free CM from lungs and CM with serum from embryo fibroblasts and human spleen were prepared as described (Fibach and Sachs, '75). CM from mature granulocytes was prepared with cells obtained 16 hours after intraperitoneal inoculation into adult mice of 3 ml of 10%sodium caseinate solution (Difco lab., Detroit) in phosphatebuffered saline (PBS).The cells were collected by washing; the peritoneum with PBS and consisted of at least 95% mature granulocytes. After washing with EM, 5 x lo7cells were seeded per 50 mm petri dish in EM without serum and CM was collected three days later. CM from D+ and D- cells was prepared by seeding 5 x lo6 cells per petri dish and the CM was collected after three days incubation in EM with 10% horse serum. Serum from normal mice and from mice after injection of bacterial endotoxin (hletcalf, '71) were prepared as described (Fibach and Sachs, '74). All cultures were incubated in a 37°C incubator with a constant flow of 10%CO,. Concanavalin A (Con A) induced agglutination. Agglutination induced by 200 pgiml Con A (Miles-Yeda, Rehovot) was measured using the Fragiligraph (Model D2). This instrument gives a continuous stirring of the ceill suspension and the degree of cell agglutination is given by the log of light transmission continuously recorded (Vlodavsky and Sachs, '75). Number of experiments. All results are based on at least Ihree separate experiments.
Cells and cell culture. D' and D- clones were isolated and three times clone purified (Fibach et al., '73) without CM from an established cell line derived from a myeloid leukemia in a SL mouse (Ichikawa, '691. Five D' clones (Nos. 3, 9, 11,12 and 24) and five 13- clones (Nos. 2,5, 16, 18 and 19) with stable phenotypes (Hayashi et al., '74) were used in the present experiments. Similar results were obtained with these five independently isolated clones of each cell type. The cells grew in suspension in liquid medium without CM as myeloblasts or some promyelocytes. The blast cells from all ten clones produced myeloid leukemia after intravenous inoculation into adult mice. Cells were generally subcultured every 4-5 days by seeding 5 x 105 cells per 50 mm petri dish in 5 ml Eagle's medium with a 4fold concentration of amino acids and vitamins (EM) (H-21, Grand Island Biological Co., New York) and 10% inactivated (56°C for 30 minutes) horse serum. To determine the CM requirement for viability and growth (RVG) 5 x 105 cells per 50 mm petri dish were incubated in EM with 10%horse serum and 25%conditioned medium (CM) for three days, unless otherwise stated, then washed three times with EM and subcultured at 2 x 106 cells per petri dish in medium with or without the addition of new 25% CM. The cells were then counted, generally three days later, in eosin solution (1:2,000) and the number of non-stained cells was recorded. RESULTS The reproducibility of the results was Znduction of a requirement for CM for generally up to k 20% when the number of viability and growth (RVG) in D+ cells was 4 x lo5 or more per milliliter, myeloid leukemic cells durand up to k 30%when the number of cells ing diffirentiation was less than 4 x 105 per milliliter. Granulocyte differentiation in mass culture was D' and D- myeloid leukemic cells are induced by CM as described (Fibach and viable and can multiply without the addiSachs, '75). Cells were seeded for cloning tion of CM. Three days incubation with
GROWTH AND IXFFERENTIATION OF LEUKEMIC CELLS
25% CM from lungs of mice injected with endotoxin (endotoxin lung CM) gave a 43% decrease in the number of D cells and a 47% increase in the number of D- cells compared to cells growing in medium without CM (Fibach and Sachs, '75).At this time, D cultures contained 80% myeloblasts and promyelocytes and 20% cells in intermediate stages of granulocyte differentiation from myelocytes to stab forms,
but no mature granulocytes. D- cultures showed 100% myeloblasts and promyelocytes and no intermediate stages or mature granulocytes. Transfer of these D' and D- cells to medium with new CM resulted in a continuation of cell multiplication, whereas transfer to medium without CM showed a decrease in the total cell number and a loss of viability in D' but not in D- cells (fig. 1).This requirement of CM
D a y s after subculture
Fig. 1 Number of D and D- cells after transfer to medium with or without CM. The cells were incubated with 25% endotoxin lung CM for three days, before transfer to medium with ( 0 ) or withCM. 0 = D' or D- cells that had never been cultured with CM. out (0)
c u) I
-0 d
E L
2-
26 1
262
EITAN FIHACH AND LEO SACHS
About 8% of the D' cells incubated for three days with CM and then transferred to medium without CM were still present after two days in this medium (fig. 1). Culture of these surkivors for two weeks in medium without CM showed, that these surviving cells were still D' cells as regards their ability to differentiate and their ability to be induced for RVG.
for viability and growth will be referred to as RVG. Transfer of D' cells to medium without CM after one, two or three days incubation with CM has shown, that the optimal induction of RVG was found after three days incubation with CM (fig. 2). D' cells after three days incubation with CM had completely lost their ability to form colonies in agar without CM and required los instead of 1 0 2 cells to induce 100%leukemia in isoloeous mice. Incubation of D- cells with C 6 did not result in any decrease in colony formation in agar or leukemogenicity in vivo (table 1).
Znduction of RVG and induction of cell d iffc'rentiation Incubation of I)' cells for three days with 25% endotoxin lung CM in addition to inducing RVG, induces rosette formation
TAHLE 1
Cloning efficiency in agar and leukemogenicity in vivo of D and I, cells hefore and after treatment with endotoxin lung CM Percent animals with leukemia Cell type Incubation with CM
Cloning efficiency
102
Number af cells inoculated 103 104
10'
I%)
DD' DD-
19+ 4 0
. .
+ __ +
100 0 100 100
22+ 5 20+ 7
100 0 100 100
100 80 100 100
100 100 100 100
D' and D- cells were incubated for three days with 25% endotoxin lung CM. The (cells were washed with PBS, counted and seeded in agar at 1,000 cells per 50 mm petri dish or injected intravenously into 2-month old female mice, five mice per point. Colonies in agar were counted after 12 days and animals were observed for three months after cell inoculation. All experiments were repeated three times. TAHLE 2
Znduction of RVG and other cellular changes associated with differentiation Granulocyte differentiation to Cell type Incubation with
D-
D D'
CM CM Trypsinized CM
Intermediate stages
Mature granulocytes
(%)
I%)
8 42 5
0
33 0
Rosette formation (percent) Fc
C3
Cells with phagocytosis
I %)
15 25 2
40 61
27
3
47 7
Cell at1:achnient tcl the Fetri dish
Cloning efficiency in agar
I%) 2 20 1
f%) 20 0 18
RVG
(Sbl
80 12
67
The cells were incubated with 25% endotoxin lung CM. Morphological differentiation to intermediate stages (myelocyte to stab form) or mature granulocytes was determined after ten days incubation. All the other properties were determined after three days incubation. Rosette formation was tested with sheep erythrocytes coated with rabbit anti sheep antibody (Fc) or with this antibody plus mouse complement (C3) (Lotem and Sachs, '74). About 1,000 cells were counted in each experiment for rosette formation. The reproducibility of rosette formation was with lower frequengenerally + 10%when the percentage of rosette forming cells was higher than 10%.and r ~ 25% cies of rosette forming cells. Phagocytosis of latex particles (Hacto-latex 0.81 p) and cell attachment to the surface of the petri dish, after staining with May-Grunwald Giemsa, were tested as described (Fihach and Sachs, '75). RVG is expressed as a percentage of the number of cells after transfer to medium without CM compared to the number of cells in medium with CM.
263
GROWTH AND DIFFERENTIATION O F LEUKEMIC CELLS
3
c u)
I
0
2
2
'5 L
aI
a
0
I
I
I
2
I 3
Doys ofter subcullure
Fig. 3 Number of D' cells after incubation in medium containing prednisolone. The cells were transferred to medium with (01 or without prednisolone after three days incubation with 0.4 pg/ml prednisolone ( 1. Similar results were obtained with 0.4pg/ml dexamethasone. TABLE 3
Induction of RVG and differentiation in D cells and colony formation with normal bone marrow cells by CM and serum from different sources Granulocyte differentiation to Source of CM or serum
Normal mouse serum D' or D- myeloid leukemic cells Mouse embryo fibroblasts Mouse normal granulocytes Lungs from endotoxin injected mice Serum from endotoxin injected mice Rat embryo fibroblasts Human spleen cells
No. of colonies per 4 x 1O'hone marrow cells
RVG
Intermediate stages
Mature manulocvtes
(%I 94 89
(%I 0 0
(76)
0 0
0 0
19 27 11
12 25 38
2 32 47
92 0 145
17
27
5
112
14 13
33 35
41 40
21 ' 32
RVC was determined after incubation of D cells for three days with 25% CM or 2.5%mouse serum and then transferring the cells to medium with or without CM or serum of the same type used for the primary culture. The results are expressed as in the legends to table 2. Granulocyte differentiation was determined after 10 days incubation with 25% CM or 2.5%mouse serum. Colony formation with normal hematopoietic cells (Pluznik and Sachs, '65, Ichikawa et al., '66; Bradley and Metcalf, '66; Worton et al., '69) was assayed by seeding 4 x lWbone marrow cells from 2 month old C57B1/6 mice in 0.33% agar on an 0.5% agar base containing 25%CM or 2.5%mouse serum. Colonies containing more than 50 cells were counted after seven days incubation. I Results from the two out of five batches of CM from rat fibroblasts and the three out of six batches of CM from human spleen, that induced colonies with normal bone marrow cells.
264
EITAN FIBACH A N D LEO SACHS
for Fc and C3 receptors (Lotem and Sachs, '741, phagocytosis of latex particles and cell attachment to the surface of a petri dish (Fibach and Sachs, '75) and 20% cells in intermediate stages of granulocyte differentiation. Ten days incubation with this CM resulted in 42% cells in intermediate stages and 33% mature granulocytes (table 2). Dcells showed a lower induction of Fc and C3 rosettes and almost none of the other changes found in D' cells (table 2). The serum free endotoxin lung CM that induced RVG was incubated with 400 pg/ml crystalline trypsin (Sigma Chemical Co., St. Louis) at 37°C for two hours followed by addition of the same concentration of soybean trypsin inhibitor (Sigma Chemical Co., St. Louis). This resulted in an inhibition of the ability of the CM to induce RVG and inhibition of the other cellular changes associated with cell differentiation (table 2). Incubation of this CM with trypsin and trypsin inhibitor without pre-incubation with trypsin, did not change the activity of this CM. These results indicate that the factor in CM that induces RVG, like the protein that induces differentiation, is trypsin sensitive. The steroid hormones prednisolone and dexamethasone can induce some, but not
all the changes associated with differentiation of D' leukemic cells (Lotem and Sachs, '75). Incubation of D+ cells with 0.4-4 pg/ml of these hormones induced these changes, but did not induce a requirement for these hormones for growth and viability (fig. 3).
CM from dijferent sources CM from different sources, including different species, were tested for their ability to induce RVG and differentiation in D' cells and collony formation with normal bone marrow cells. The results indicate (table 3) that CM from rat and mouse embryo fibroblasts, human spleen and normal mouse granulocytes, like endotoxin lung CM, induced RVG and various degrees of differentiation to mature granulocytes. There was no induction of RVG or differentiation by CM from D' or Dmyeloid leukemic cells. Serum from mice injected with bacterial endotoxin, but not normal mouse serum, also induced RVG and differentiation. The majority {of CM from different sources and serum from endotoxin injected mice induced the formation of colonies with mature macrophages and granulocytes from normal bone marrow cells.
T I me (minutes)
Fig. 4 Con A induced agglutinability of D' cells after transfer to medium with or without CM. Cells were incubated for three days with 25% endotoxin lung CM and assayed for agglutinability one day after (A) transfer to medium with 25% CM (B) transfer to medium without CM, (C) cells which were not treated with CM. Con A solution was added at a final concentration of 200 &ml and the light transmission automatically recorded.
GROWTH AND DIFFERENTIATION OF LEUKEMIC CELLS
However three out of five batches of CM from rat embryo fibroblasts, three out of six batches of CM from human spleen and six out of six batches of CM from normal mature granulocytes, induced RVG and differentiation in D' cells, but did not induce the formation of colonies from normal bone marrow cells (table 3 ) .
Agglutinability by concanavalin A (Con A ) Incubation of D' cells for three days with 25% endotoxin lung CM results in an increase in the ability of the cells to be agglutinated by Con A (Vlodavsky et al., '76). Cells tested one day after transfer to medium without CM, before there was a considerable loss of cell viability, still showed the high Con A agglutinability of the cells incubated in medium with CM (fig. 4). DISCUSSION
Normal hematopoietic cells require a protein, which we call MGI, for cell viability, growth and differentiation to mature macrophages and granulocytes (Sachs, '74a). Clones of myeloid leukemic cells have been established (D+1 which require CM with this protein for differentiation to macrophages and granulocytes but do not require the CM for cell viability and growth. The present experiments have shown, that D' myeloid leukemic cells during the process of differentiation again require the presence of CM for cell viability and growth (RVG).The induction of RVG occurred before the appearance of mature cells and resulted in a loss of cloning efficiency in agar in the absence of CM. The development of leukemia in 100% of the animals required the inoculation of lo5 cells induced for RVG compared to lo2 control cells. D- myeloid leukemic cells which could not be induced to differentiate to mature cells by the same CM, were also not induced for RVG. Incubation of serum-free CM with trypsin inhibited both the ability to induce RVG and the ability to induce other cellular changes associated with differentiation. It may be that the
265
same protein (MGI) induces RVG and the differentiation of D' cells. It can also be suggested that the change in the cells resulting in a decrease or lack of requirement for this protein for viability and growth, may be one of the causes of myeloid leukemia. Experiments with CM from different sources and species have shown, that RVG in mouse D+ cells can also be induced b y CM from rat and human cells. All the CM from different sources and serum from mice injected with bacterial endotoxin that induced RVG also induced differentiation of D+ cells. This supports the suggestion that the same protein may be involved in both processes. Although the majority of CM from different sources also induced the formation of colonies with macrophages and granulocytes from normal bone marrow cells, some batches of CM from rat embryo fibroblasts and human spleen cells and all batches of CM from normal mature granulocytes, induced RVG and differentiation in D' cells, but did not induce the fomation of colonies from normal bone marrow cells, This suggests that there may be co-factor(s) (Landau and Sachs, '71a,b) required for colony formation with normal bone marrow cells that are absent in these batches of CM. D' but not D- cells show an increased cell agglutinability by Con A after incubation with the appropriate CM (Vlodavsky et al., '76). The present experiments have shown that this increase in cell agglutinability was not reversed after the cells were transferred to medium without CM. This suggests that the specific requirement for viability and growth induced in D+ cells during differentiation, is associated with a change in the cell surface membrane. ACKNOWLEDGMENT
We thank Mrs. Esther Gerassi for skillful technical assistance. This work was supported by grants from the Talisman Foundation and the Jerome and Estelle R. Newman Assistance Fund, New York.
266
EITAN FIBACH AND LEO SACHS
LITERATURE CITED Bradley, T. R., and D. Metcalf 1966 The growth of mouse bone marrow cells in uitro. Aust. J. Exp. Biol. Med. Sci., 44: 287-299. Fibach, E., M. Hayashi and I,. Sachs 1973 Control of normal differentiation of myeloid leukemic cells to macrophages and granulocytes. Proc. Natl. Acad. Sci. ( U S A . ) , 70: 343-346. Fibach, E., T. Landau and L. Sachs 1972 Normal differentiation of myeloid leukemic cells induced by a differentiation inducing protein. Nature, New Biol., 237: 276-278. Fihach, E., and L. Sachs 1974 Control of normal differentiation of myeloid leukemic cells. IV. Induction of differentiation by serum from endotoxin treated mice. J. Cell. Physiol., 83: 177-186. 1975 Control of normal differentiation of myeloid leukemic cells. VIII. Induction of differentiation to mature granulocytes in mass culture. J. Cell. Physiol., 86: 221-230. Hayashi, M., E. Fibach and L. Sachs 1974 Control of normal differentiation of myeloid leukemic cells. V. Normal differentiation in aneuploid leukemic cells and the chromosome banding pattern of D’ and Dclones. Int. J. Cancer, 14: 40-48. Ichikawa, Y. 1969 Differentiation of a cell line of myeloid leukemia. J. Cell. Physiol., 74: 223-234. Ichikawa, Y., D. H. Pluznik and L. Sachs 1966 In uitro control of the development of macrophage and granulocyte colonies. Proc. Natl. Acad. Sci. ( U S A . ) , 56: 488-495. Landau, T., and L. Sachs 1971a Characterization of the inducer required for the development of macrophage and granulocyte colonies. P r k . Natl. Acad. Sci. ( U S A . ) , 68: 2540-2544.
1971b Activation of a differentiation inducing protein by adenine and adenine containing nucleotides. FEBS Letters, 17: 339-341. Lotem, J,, and L. Sachs 1974 Different blocks in the differentiation of myeloid leukemic cells. Proc. Natl. Acad. Sci. (USA,), 71: 3507-3511. 1975 Induction of specific changes in the surface membrane of myeloid leukemic cells by steroid hormones. Int. J. Cancer, 15: 731-740. Metcalf, D. 1971 Acute antigen induced elevation of serum colony stimlulating factor (CSF)levels. Immunology, 21: 427-436. Pluznik, D. H., and L. Sachs 1965 The cloning of normal “mast” cells in tissue culture. J. Cell. Comp. Physiol., 66: 319-324. Sachs. L. 1974a Regulation of membrane changes, differentiation and inalignancy in carcinogenesis. Harvey Lectures, Academic Press, New York, 68: 1-35. 19741, Control of growth and differentiation in normal hematopaietic and leukemic cells. In: Control of Proliferation in Animal Cells. Cold Spring Harbor Lab. New York, pp. 915-925. Vlodavsky, I., E. Fibach and L. Sachs 1976 Control of normal differentiation of myeloid leukemic cells. X. Glucose utilization, cellular ATP and associated membrane changes in D and D- cells. J. Cell. Physiol., 87: 167-177. Vlodavsky, I., and L. Siichs 1975 Lectin receptors on the cell surface membrane and the kinetics of lectin induced agglutination. Exp. Cell Res., 93: 11 1-1 19. Worton, R. G., E. A. McCulloch and J. E. Till 1969 Physical separation of hematopoietic stem cells from cells forming colonies in culture. 1. Cell. Physiol., 74: 171-182
~
