Calcium Ionophore-induced Transient Down-regulation of c-myb mRNA Levels in Friend Erythroleukemia Cells*

The effects of calcium ionophores A23187 and iono- mycin on the c-myb and c-myc mRNA levels have been investigated in the Friend erythroleukemia cell line F4-6 using Northern blot analysis. Treatment of the cells with 0.5-4 pM A23187 or 1-4 p~ ionomycin induced a concentration-dependent decrease in c-myb mRNA; this decrease was abolished by EGTA. c-myc mRNA levels were only moderately affected. After 12-24 h of calcium ionophore exposure, c-myb mRNA returned to pretreatment levels. No similar decrease in c-myb mRNA was seen with the sodium ionophore monensin (up to 16 WM). The dimethyl sulfoxide-in- duced suppression of c-myb and also of c-myc mRNA levels was not prevented in Ca2+-free medium and thus appeared Ca2+-independent. A23187 and ionomycin were capable of inducing @-globin mRNA synthesis in F4-6 cells. Prolonged calcium ionophore exposure, however, strongly reduced cell viability and resulted only in a slight hemoglobin increase at lower concen- trations. These results suggest that a rise in [Ca2+Ji may be a signal leading to a transient decrease in c-myb mRNA and the initiation of erythroid differentiation in Friend cells. The transient suppression of c-myb mRNA levels represents a common feature of the action of dimethyl sulfoxide

firmed in subsequent studies (13)(14)(15), these experiments suggest that Ca2+ influx caused by the calcium ionophore A23187 may have effects on the early cellular events involved in the induction of differentiation (6,7,9,10). In fact, it has been demonstrated recently that A23187 as well as another chemically distinct calcium ionophore, ionomycin, can trigger commitment to differentiation in Friend cells (16).
Modulation of the expression of nuclear protooncogenes such as c-myc, c-myb, and c-fos, may be involved in the early cellular changes leading to differentiation in Friend cells (5). A characteristic transient decrease in c-myc and c-myb mRNA levels has been observed in Friend cells during the first few hours after treatment with the known chemical inducers of differentiation, Me2S0, hexamethylene bisacetamide (HMBA), and hypoxanthine (17-22) as well as with the potent differentiation-inducing anthracycline antitumor antibiotic, aclacinomycin A (23). However, the biochemical mechanisms underlying these rapid changes in the oncogene expression as well as their involvement in the commitment process have not been defined.
Since calcium ionophores induce early biochemical changes in Friend cells that may result in commitment to differentiation (16), the present experiments were undertaken to investigate whether calcium ionophores can produce early effects on the levels of c-myc and c-myb mRNAs in Friend erythroleukemia cells.

MATERIALS AND METHODS
Chemicals"A23187, ionomycin, and monensin were purchased from Sigma. The ionophores were dissolved in ethanol. Experiments demonstrated that the resulting solvent concentration in the cell culture medium had no effect on the mRNA levels or cell differentiation.
For the experiments, exponentially growing cells were plated at 6-7 X 10' cells/ml. Approximately 16 h later, the cells were treated with the test substances, and at defined time points thereafter cells were harvested and stored at -80 "C before RNA isolation. To measure hemoglobin production, cells were incubated for 4 days after treatment, and hemoglobin content of the cells was determined by the benzidine technique of Luftig et al. (24) using bovine hemoglobin as standard. Cell viability was examined by trypan blue exclusion.
Northern Blot Analysis-Total cellular RNA from F4-6 cells was isolated by the acid guanidinium thiocyanate-phenol-chloroform e-myb mRNA Decrease in Friend Cells by Calcium Ionophores method of Chomczynski and Sacchi (25). 30 pg samples of RNA were denaturated by glyoxylation, size-separated by electrophoresis through 1% agarose gel, and transferred to Biodyne A membrane (Pall, Portsmouth, United Kingdom) by the capillary blotting technique (26) using 20 X standard saline citrate solution (SSC, 1 X SSC contains 0.15 M sodium chloride and 0.015 M sodium citrate). RNA was immobilized by baking the membrane a t 80 "C for 1.5 h. The cDNA probes for hybridization were labeled with [w3*P]dCTP by the random primer method (27) using the multiprime DNA-labeling system from Amersham Buchler (Braunschweig, FRG). Membrane prehybridization (6 h) and hybridization (16-18 h) were carried out at 42 "C in 50% (c-myc, @-actin, and @-globin) or 40% (v-myb) formamide, 50 mM sodium phosphate buffer, pH 6.5, 5 X SSC, 5 X Denhardt's solution (0.1% bovine serum albumin, 0.1% Ficoll, 0.1% polyvinylpyrrolidone), 0.1% SDS, and 250 pg/ml denatured herring sperm DNA (Boehringer, Mannheim, FRG). Hybridized blots were washed initially in 2 X SSC containing 0.1% SDS at room temperature for 30 min and a t 50 "C twice for 30 min. More stringent washing was performed after hybridization with c-myc and @-actin cDNA probes in 0.2 X SSC and 0.1% SDS at 55 "C twice for 30 min, after hybridization with @-globin cDNA in 0.1 X SSC and 0.1% SDS at 55 "C twice for 30 min and after hybridization with v-myb in 0.5 X SSC and 0.1% SDS at 50 "C twice for 30 min. The membranes were exposed at -80 "C to an x-ray film using intensifying screens. Following stripping in 50% formamide and 10 mM sodium phosphate buffer, pH 6.5, at 65 "C for 30 min, the membranes were subsequently hybridized with the cDNA probes.

RESULTS
Effect of A23187 on the c-myb and c-myc mRNA Levels-F4-6 cells were exposed to the calcium ionophore A23187 at the concentrations of 0.5, 1, 2, and 4 p~. Three h after treatment the cells were harvested, and c-myb and c-myc mRNA levels were measured using Northern blot analysis (Fig. 1). The levels of &actin mRNA were also determined as an internal control for RNA loading. The results show a concentration-dependent decrease in c-myb mRNA levels after A23187 treatment. The c-myc mRNA levels were only moderately affected at the two highest concentrations tested. The effect of A23187 was studied also in the presence of EGTA, 3 mM (Ca2+ concentration in the medium used was -2.1 mM). The results demonstrate that the effect of A23187 Total cellular RNA was isolated and tested using Northern blot analysis. Amounts of 30 pg of denatured RNA were size fractioned on 1% agarose gel and transferred by the capillary blotting technique to a nylon filter. Blots were hybridized sequentially with v-myb, third exon c-myc, and @-actin cDNA probes. @-Actin mRNA was determined as a control to verify the amount of RNA in each lane. on the c-myb expression was abolished in the presence of EGTA. This indicates that the observed effect of A23187, indeed, is depending on the presence of extracellular Ca2+.
T o further characterize the effect of A23187, 1 pM, on the c-myb mRNA levels, a time course analysis was performed over a treatment period of 12 h. As shown in Fig. 2, c-myb mRNA levels began to decrease a t 2 h after treatment. c-myb mRNA levels showed a minimum at 3-6 h. After 9 h, c-myb mRNA levels recovered and approached pretreatment levels a t 12 h.
Effect of Ionomycin on the c-myb and c-myc mRNA Levek-In further experiments the effect of ionomycin, another calcium ionophore, was investigated on c-myb and c-myc mRNA levels in F4-6 cells. Fig. 3 shows the effect of ionomycin at concentrations of 1, 2, and 4 p~ 1 and 3 h after treatment and at concentrations of 1 and 2 p~ 24 h after treatment. Similarly to the results obtained with A23187, ionomycin induced a concentration-dependent decrease in c-myb mRNA levels 3 h after drug exposure. No changes in the levels of cmyb mRNA could be observed 24 h after treatment as compared with the untreated control. c-myc mRNA levels remained unchanged or only slightly affected during incubation with ionomycin. As shown in Fig. 4, the effect of ionomycin, 2 p~ on the c-myb mRNA levels was antagonized in the presence of EGTA, 3 mM.
Effect of Monensin on c-myb and c-myc mRNA Levels-To characterize the specificity of the action of calcium ionophores on the c-myb expression, we studied the effect of the sodium ionophore monensin on the c-myb and c-myc mRNA levels in F4-6 cells. As A23187, monensin acts as a H+ exchanger, leading to an increase in cytoplasmic pH. This has also been discussed as being responsible for the effects of A23187 in Friend erythroleukemia cells (15). The results showed that monensin, studied at the same concentrations as the calcium ionophores (1-4 p M ) or higher (up to 16 p M ) did not induce a decrease in c-myb mRNA levels (Fig. 5). Moreover a slight increase in c-myb and a decrease in c-myc transcript levels could be observed in F4-6 cells after 3 h of treatment with monensin.
Effect of EGTA on the MepSO-induced Changes in c-myb and c-myc mRNA Levels-The results with the two calcium ionophores suggest that Ca2+ influx in Friend erythroleukemia cells can induce a transient decrease in c-myb mRNA levels. Because chemical inducers of the differentiation, such as Me2S0, are known to induce similar changes (19-23), we studied the effect of EGTA, 3 mM, on the MepSO-induced decrease in c-myb and c-myc transcript levels (Fig. 6). In accordance with previous data (17-23), Me2S0, 1.5%, induced a strong decrease in c-myc and c-myb mRNA levels within 1 and 3 h, respectively. These changes were not antagonized by the presence of EGTA in the incubation medium. A slight increase in c-myb (at 3 h) and decrease in c-myc mRNA levels was observed with EGTA alone. Similar negative results were obtained when the cells had been incubated in Ca2+ free medium (3 mM EGTA) for 18 h before Me2S0 treatment (Fig.  7A) or when the EGTA concentration was raised from 3-5 or 10 mM (Fig. 7B). These results strongly suggest that the Me2SO-induced suppression of c-myb expression occurs in a Ca2+-independent manner. Thereafter they were exposed to Me2S0, 1.5%, for additional 3 h. B, cells were exposed for 3 h to Me2S0, 1.5%, in the presence and absence of EGTA, 5 and 10 mM. For experimental conditions see Fig. 1.
Effect of Calcium Ionophore Treatment on @-Globin mRNA Expression-The calcium ionophores A23187 and ionomycin have been shown in previous studies to induce commitment to differentiation in Friend cells at concentrations of 1-2 pM (16) a t which these agents decreased c-myb mRNA levels in the present experiments (Figs. 1-3). T o study a specific cellular event in the Friend cell differentiation process, we analyzed the induction of @-globin mRNA synthesis, which begins 2 days after inducer treatment (5, 17). Cells were treated with A23187 and ionomycin for 36 h a t concentrations of 0.5 and 1 p~ and 1 and 1.5 PM, respectively. After this incubation period, the substances were removed and an additional 24 h later the induction of P-globin mRNA synthesis was determined using Northern blot analysis. The effect of Me2S0, 1.5%, was comparatively studied. The results show that both A23187 and ionomycin induced @-globin mRNA synthesis in F4-6 cells in this early phase of treatment (Fig.  8). Me2S0, however, induced a stronger increase in 6-globin mRNA than the calcium ionopores.
Effect of Calcium Ionophore Treatment on the Hemoglobin Synthesis-To study the effect of calcium ionophores on the terminal differentiation status of F4-6 cells, hemoglobin production was estimated after 4 days of drug treatment. The calcium ionophores were studied in a concentration range in which the cell viability was reduced. Again, the effects of Me2S0, 1.5%, were comparatively studied. After this longer  period of treatment with the calcium ionophores, cell viability was reduced at concentrations below 0.5 ~L M (Table I). A23187 proved to be more effective in reducing cell viability and showed toxic effects a t 0.5 p~. A slight increase in hemoglobin production could be observed with the ionophores which was accentuated with ionomycin a t 250-500 nM. These results show that although calcium ionophores lead to the initiation of the terminal differentiation program of F4-6 cells during the early phase of treatment, prolonged incubation with calcium ionophores above 0.5 p~ is associated with toxic effects.

DISCUSSION
The present experiments demonstrate that two chemically distinct calcium ionophores, A23187 and ionomycin, induce a dramatic fall in c-myb mRNA levels in Friend erythroleukemia cells. In the studied concentration range, these substances have been shown to elevate the intracellular Ca2+ concentration in Friend cells (16). The sodium ionophore monensin, tested up to 16 p~, did not induce early decrease in c-myb mRNA levels. The effect of the calcium ionophores on the cmyb expression could be antagonized by EGTA in the incu-bation medium. Therefore, the results suggest that an increase in intracellular Ca2+ concentration is responsible for the observed effects on c-myb mRNA levels.
Transcriptional arrest has been shown to be one of the major mechanisms that switches off c-myb expression (30). The time course of the decrease in c-myb mRNA levels induced by the calcium ionophore A23187 (Fig. 2) can well be explained with such a mechanism, since inhibition of the total RNA synthesis with actinomycin D in Friend cells reduces cmyb mRNA levels with similar kinetics (21, 23). However, increased mRNA turnover may also contribute to the regulation of c-myb mRNA levels in Friend erythroleukemia cells (21). On the other hand, calcium ionophores induced a transient decrease in c-myb mRNA levels. By 12-24 h after treatment, c-myb mRNA levels returned to about pretreatment levels (Figs. 2 and 3). This is interesting in view of the sustained increase in cytosolic Ca2+ concentration in Friend cells in the presence of calcium ionophores (16). Further experiments are necessary to elucidate the biochemical mechanisms by which a rise in intracellular Ca2+ concentration may suppress c-myb mRNA levels in Friend erythroleukemia cells.
A23187 and ionomycin have been shown to induce commitment to differentiation of Friend cells a t concentrations of 1-2 p~ (16) a t which these agents were active in decreasing cmyb mRNA levels in the present experiments. Our observations that calcium ionophores induce &globin mRNA synthesis in the early phase of treatment, which was measured 24 h after drug removal (Fig. 8), are consistent with these results. It should be noted, however, that Friend cells can be induced to synthesize globin mRNA also in the absence of commitment ( 5 ) . Although the increased @-globin mRNA expression indicates that the cells were executing the terminal differentiation program, longer incubation of F4-6 cells with A23187 and ionomycin strongly reduced cell viability and thus only slight increases in hemoglobin production could be demonstrated a t lower concentrations (Table I). It is noteworthy that the changes in c-myb mRNA levels after calcium ionophore treatment are similar to those after exposure of Friend cells to chemical inducers of differentiation (19-23). On the other hand, the rapid transient decrease in c-myc expression, which is observed after treatment with Me2S0 and other inducers (17-23), was not seen with the calcium ionophores. These results may point to the importance of the decrease in c-myb expression for the early cellular events of Friend cell differentiation.
In contrast, expression of transfected c-myb cDNAs during the early phase of inducer treatment does not block differentiation or early commitment effects in Friend cells (31,32). It has been suggested that the postcommitment down-regulation of c-myb mRNA a t a later time, not the early transient decrease, is related to terminal differentiation. However, it is important to consider that the expression of transfected cmyb in Friend cells does not prevent the early effects of Me2S0 or HMBA on the endogenous c-myb and c-myc mRNA regulation (31, 33). Therefore, these experiments do not exclude the possibility that the early down-regulation of endogenous c-myb expression is involved in the commitment program. Furthermore, in c-myc-transfected Friend cells, the levels of exogenous c-myc protein synthesis are decreased by up to 50% or more upon exposure t o M e 8 0 during the commitment period (22). The question arises whether in cmyb-transfected cells a similar decrease in c-myb protein synthesis occurs during the commitment phase, in spite of the high levels of exogenous c-myb mRNA (31-33).
The effect of Me2S0 on c-myb and c-myc mRNA levels was not prevented by the presence of EGTA (Figs. 6 and 7). Accordingly, the MezSO-induced decrease in c-myb mRNA levels appears independent of Ca2+. This observation is consistent with previous results showing that MezSO causes a slight decrease in cytosolic Ca2+ concentration (15,16). Therefore, M e 8 0 and calcium ionophores decrease c-myb mRNA levels via different mechanisms. In addition, the effect of MezSO on c-myb mRNA levels, but not that of the calcium ionophores is accompanied by a similar effect on the c-myc expression (19,(21)(22)(23). However, c-myb mRNA levels may represent a common target of the action of MezSO and calcium ionophores in Friend erythroleukemia cells.
Erythropoietin-induced differentiation in SKT6, J2E, and Rauscher mouse erythroleukemia cells has been shown to lead to a decline in c-myb mRNA levels and an increase in c-myc transcripts in the early stage of differentiation (34-36). On the other hand, MezSO or sodium butyrate treatment results in a decrease of both c-myb and c-myc mRNA levels in these cells. It was concluded that down-regulation of c-myb gene expression is an important prerequisite for either erythropoietin initiated or chemical-induced erythroid differentiation (34,35). The early transient down-regulation of c-myb mRNA levels in J2E cells by erythropoietin (35) showed a similar kinetic to that observed with calcium ionophores in the present experiments (Fig. 2). Earlier studies demonstrated that calcium ionophores enhance colony formation by erythroid colony-forming units and that EGTA inhibits such colony formation (37). A series of subsequent studies suggest that calcium may be involved in the erythropoietin-mediated signal transduction, although there are also conflicting results (reviewed in Ref. 38). Considering our present results, therefore, the question arises whether the early down-regulation of c-myb expression by erythropoietin is mediated by a calcium signal. In this case, the effect of calcium ionophores on c-myb expression shown in the present experiments may represent a part of the erythropoietin signal pathway in erythroid precursor cells. The stimulation of c-myc expression by erythropoietin (34-36) has been shown to occur through other mechanisms (36) and may also be related to the cell proliferation-stimulating activity of erythropoietin (35,39). Further experiments with erythropoietin responsible cell lines are necessary to reveal this possibility. However, if the erythropoietin-induced decline of c-myb expression occurs in a Ca2+-independent manner, our results with calcium ionophores would represent another alternative mechanism of c-myb mRNA down-regulation with inducers of the commitment to terminal differentiation (16) and would still underline the importance of the regulation of c-myb expression in the early phase of erythroid differentiation.