Effect of verapamil on cell cycle transit and c-myc gene expression in normal and malignant murine cells.

Verapamil, the prototype calcium channel blocker, reversibly inhibits cell proliferation in many normal and tumour cell lines (Schmidt et al., Cancer Res., 48, 3617, 1988). We have found that two closely related cell lines - B16 murine melanoma cells and B10.BR normal murine melanocytes growing in culture - behave differently in the presence of verapamil, and we are now utilising these two related cell lines to help elucidate the molecular basis of verapamil's antiproliferative effect. In this study, we studied cell cycle phase distribution and c-myc gene expression in both cell lines in the absence of verapamil, during incubation with verapamil and after the cells were washed free of verapamil. Our studies show that 100 microM verapamil rapidly blocks DNA synthesis in melanocytes but not in B16 cells. Similarly, incubation with verapamil for 6-24 h results in a decreased c-myc signal in melanocytes, but a transient increase in c-myc expression in B16 cells. After verapamil is washed from the cells following a 24-h incubation with drug, c-myc expression increases in melanocytes as they begin again to proliferate, but decreases in B16 cells as they begin to die. Our disparate results with these cell lines suggest that c-myc gene expression, regardless of its known involvement in growth control, is not the immediate target for verapamil's inhibitory action.

The calcium channel blockers have generated much interest in cancer research since the demonstration that at low concentration (5-10Mm) they augment the cytotoxicity of many standard anti-cancer agents in a variety of tumour cell types (Tsuruo et al., 1983a,b;Yalowich & Ross, 1984Robinson et al., 1985;Ince et al., 1986;Merry et al., 1986). Verapamil, the prototype calcium channel blocker, enhances the cytotoxic effects of both vincristine and adriamycin in vitro as well as in vivo in cells previously resistant to these drugs (Tsuruo et al., 1981(Tsuruo et al., , 1983a. Although the precise mechanism of this increased cytotoxic effect is not completely understood, the calcium channel blockers are thought to act by blocking efflux of the chemotherapeutic agents from the cell (Tsuruo et al., 1982).
At higher concentrations (10-100yM), verapamil by itself reversibly inhibits cell growth in several human cell lines (Schmidt et al., 1988). Protein synthesis, DNA synthesis and RNA synthesis are all inhibited within minutes of addition of OO M verapamil to the cells; removal of the drug by simple washing of the cells results in a rapid resumption of cell growth (Schmidt et al., 1988). These reversible antiproliferative effects of verapamil make it an ideal compound to study cell cycle related events. Cell growth is controlled by a cascade of events that ultimately leads to DNA synthesis. Briefly, cell proliferation begins when growth factors interact with the cell membrane, sending a signal via inositol phospholipids to increase cytoplasmic ionised calcium by one pathway and to increase cytoplasmic pH by another pathway (Berridge, 1984). However, stimulation of protein kinase C by phorbol esters (e.g. TPA) directly causes cytoplasmic alkalinisation and in at least some cells this stimulation can by-pass the calcium pathway (Rozengurt & Mendoza, 1985). Furthermore, the involvement of c-onc genes in cell proliferation has been extensively documented (Kahn & Graf, 1986).
One of the first genes linked to cell growth was the protooncogene c-myc. C-myc gene expression is known to be linked tightly to cell proliferation, increasing 10-20 fold in cells treated with some mitogens (Kelly et al., 1983). The cmyc gene, which is expressed in both malignant and normal cells, encodes a protein that is functionally involved in DNA synthesis (Studzinski et al., 1986). This protein is believed to directly regulate the rate at which cells divide (Cole, 1986).
The purpose of this study was to determine the effects of verapamil on cell cycle transit and c-myc gene expression in two closely related cell lines in order to shed further light on the mechanisms of verapamil's antiproliferative effects. B1O.BR normal melanocytes and B16 melanoma cells were chosen for this study because these cells exhibited markedly different responses to incubation with verapamil in preliminary experiments.

Methods
Cell culture Murine melanoma cell lines B16 Fl and B16 FIO were obtained from ATCC. The cells were grown in RPMI medium supplemented with 10% FCS, plus penicillin, streptomycin and fungizone. Melanocytes from B1O.BR mice (Tamura et al., 1987) were kindly provided by Dr Ruth Halaban, Yale University, New Haven, CT. These cells were incubated in Ham's FIO medium supplemented with 15% fetal calf serum and 48 nM TPA (12-O-tetradecanoylphorbol-13-acetate), plus penicillin, streptomycin and fungizone. The growth chamber was maintained at 37°C with 5% CO2. For experiments excluding TPA, cells were incubated in Ham's FIO medium without TPA for 48 h. Cells were used before their 20th generation. The potential for the cells to metastasize was tested according to Fidler & Kripke (1977). After injection of 50,000 to 100,000 cells into the tail veins of C57BL/6 mice, the B16 FIO cells form many more pulmonary metastases than the B16 Fl cells within 2-3 weeks. All experiments were performed at least twice.

Radioisotope incorporation
Actively growing cells (in the exponential phase) were always used when assessing isotope incorporation, Methyl-3Hthymidine was added to cell cultures at a final concentration of 0.5 MCi ml-1. At timed intervals, cell samples were removed after trypsinisation and added to an equal volume Br. J. Cancer (I 989),59,[714][715][716][717][718] of cold 10% trichloroacetic acid. Precipitates were allowed to form for 30min on ice before filtering through Whatman GF/C glass filters mounted in a vacuum manifold. After washing with cold saline solution and ethanol, the filters were dried under a heat lamp, then counted for radioactivity in Ready-Solv-MP (Beckman Instruments) with a scintillation spectrophotometer.
Cell cycle analysis The nuclei isolation medium (NIM) (Thornthwaite et al., 1980) contained per litre: 10mmol phosphate buffer, 146mmol NaCl, 1.Ommol CaCl2, 0.5 mmol MgSO4-7H20, 6.0ml Nonidet NP40 (Sigma) and 700 units RNase (Sigma type lA, boiled for 10min to remove residual DNase activity). The DNA fluorochrome propidium iodide (PI) (Sigma) was dissolved in NIM at a concentration of 50 ug ml-1. Monolayer cells (triplicates) were washed by rinsing with phosphate buffered saline (PBS) before the addition of NIM buffer. The nuclei were kept in NIM buffer for at least 16h and then filtered through a 70pm nylon mesh. Cell cycle analyses on the PI-stained nuclei were performed on a Coulter Electronics Epics V flow cytometer (Coulter Electronics Inc., Hialeah, FL). The instrument was adjusted to achieve coefficients of variation for the nuclei in the range from 3 to 5%. The relative fluorescence intensities of 10,000 PI-stained nuclei were measured and the proportion of nuclei in G1, S and G2-M was calculated using the Para I data analysis program of the Epics flow cytometer.

RNA and DNA isolations and hybridisations
For RNA isolation, cells (107) were washed three times in PBS and transferred to polypropylene tubes. The cell pellet was resuspended in 0.5 ml of extraction buffer (250 mM NaCl, 50mM Tris-hydrochloride (pH 7.4), 5mM EDTA, 1% sodium dodecyl sulphate and 1 mg ml-1 of proteinase K (Sigma)). After incubation for 30 min at 37°C the mixture was sonicated for three 5-s bursts to shear DNA, and extracted with a solution containing 0.5 ml of phenol and 0.25 ml of chloroform. The aqueous phase was extracted again with phenol-chloroform (2:1), washed twice with chloroform, and precipitated with ethanol. DNA was extracted by taking up the pellet in 2 M LiCl solution containing 10 mM EDTA. After centrifugation (10,000 r.p.m., 5min) the RNA pellet was suspended in H20. Total RNA was denatured with 6% formaldehyde and 50% formamide, heated 5min to 65°C, size fractionated on a 1% agarose gel containing 2.2 M formaldehyde and blotted onto nitrocellulose or nylon membranes (Hybond, Amersham) according to Thomas (1980). For DNA preparations, cells were added to extraction buffer, then incubated overnight at 37°C. After phenolchloroform extractions, 10plImlof RNase solution (100 jpgml-1, DNase free) was added and the mixture was incubated for another 30min. After subsequent ammonium acetate (3M) precipitations to remove protein, the DNA was precipitated with ethanol. Aliquots were incubated with 3 units of restriction enzymes (Bam HI, Xba I, Xho I, Bgl II, Sst I) per ,ug of DNA at 37°C overnight, size separated on agarose gels, and blotted on to nitrocellulose or nylon membranes according to Southern (1975).

Results
C-myc expression in normal melanocytes and melanoma cell lines Experiments were first performed to establish base-line c-myc mRNA levels in continuously growing B10.BR melanocytes and B16 melanoma cells. Because B10.BR murine melanocytes require the addition of TPA to their medium for continuous growth (Tamura et al., 1987), total RNA was extracted from cells incubated for 48 h in either the presence or absence of TPA. C-myc expression under these conditions was compared to that obtained in B16 cells with both a high metastatic potential (B16 FIO) and low metastatic potential (B16 F1). The RNA was probed with a 32P-labelled 1,000 b.p. Pst I fragment of a c-myc clone. In one experiment, the filter was double probed for both c-myc and thymidine kinase (tk) gene expression (to confirm increased DNA synthesis). The results of this experiment, as illustrated by Figure  total RNA was isolated and 20pg each resolved on denaturing agarose gels. BlO.BR cells were incubated in medium with or without TPA for 48h before they were harvested, 20pg total RNA each was resolved on denaturing agarose gels. The filters were hybridised with 107 c.p.m. 32P nick translated c-myc cDNA insert and a thymidine kinase (TK) gene specific probe, as described in Materials and methods. Lower panel: the same filters were reprobed with 32P labelled p5B for analysis of 18S rRNA.  Figure 2 Restriction analysis of c-myc from Balb/c spleen and B16 FO0 melanoma cells. DNA (20 pg per lane) was digested to completion with Bam HI (lanes 1 and 6), Xba I (lanes 2 and 7), Xho I (lanes 3 and 8), Bgl II (lanes 4 and 9) and Sst I (lanes S and 10). Digests were fractionated on an agarose gel, transferred to Hybond membranes and hybridised with a 1,400 b.p. Xho I c-myc fragment as described. The size marker is a Hind III digest of phage Lambda DNA.

Southern analysis of c-myc gene sequences
We then assayed for possible rearrangements in the c-myc sequences in the melanoma cells that might account for the constitutive c-myc expression in these cells. DNA was completely digested with enzymes that have recognition sites within the c-myc gene and in the flanking regions as described in Methods. As can be seen in Figure 2, there are no obvious rearrangements in the c-myc gene in the melanoma cells compared to normal control spleen cells. Three separate experiments indicated that no amplification of the c-myc gene in the B16 melanoma cells had occurred.
Effects of verapamil on cell cycle and c-myc expression To monitor the effects of verapamil on DNA distribution and c-myc expression, cells were incubated with verapamil for varying periods of time. Nuclei were analysed on a flow cytometer and RNA was isolated and assayed for c-myc mRNA. Different effects of verapamil on melanoma cells and normal melanocytes were obtained.
Incubation with 100 M verapamil has little effect on B16 cells, the 3H-thymidine incorporation assay shows that DNA synthesis in B16 cells is only 10% inhibited after 3 h (Table II). Longer incubation with 1OM verapamil transiently reduces the number of cells entering from G1 into Sphase, the cells already in S-phase continue their cell cycle transit (Figure 3, Table I). Parallel to the induction of synchronised progression through the cycle there is also a concomitant increase in the expression of the c-myc gene (Figure 4, 12h-lane). A high proportion of the cells in this line continues to traverse the cell cycle even after 24h of incubation with lOOiM verapamil as evidenced by the considerable percentage of cells in S-phase (Table I, 24 h incubation with verapamil). However, drastic changes are seen once the drug is washed from the cells. Approximately 12h after removal of verapamil the B16 cells start to produce melanin and die. Increased cell death is also reflected by the high amount of fluorescent material in front of the GI peak in the DNA histogram (Figure 3, VP release) (indicating that the cells have released nucleic acids while they deteriorated) and by about 50% lower c-myc mRNA levels ( Figure 4). Apparently, addition of verapamil induces differentiation in these cells. By contrast, cell growth in B10.BR normal melanocytes is blocked rapidly and reversibly by 100l M verapamil as occurs also in other cell lines tested previously (Schmidt et al., 1988). DNA synthesis is reduced by 76% within 3h of adding 100Mm verapamil (Table II). The cell cycle phase analyses data (Figure 3 and Table I) show that the cells do not enter into S-phase in the presence of verapamil. Cells already in S-phase, however, proceed through the DNA synthesis phase. Concomitantly, the c-myc signal is decreased (Figure 4). After the B10.BR cells are washed free of verapamil, they rapidly resume growth as also depicted by the DNA histogram ( Figure 3) and increased c-myc mRNA levels ( Figure 4). In summary, verapamil has different effects on the two cell lines studied. The B16 melanoma cells seem to be induced to a differentiation pathway, whereas cell growth of the B10.BR melanocytes is blocked rapidly and reversibly. C-myc mRNA levels parallel the distribution of cells in the cell cycle indicating that changes in c-myc gene expression are secondary effects of the calcium channel blocker.

Discussion
Our work shows that two closely related cell lines are affected differently by verapamil. The proliferation of normal B1O.BR melanocytes is stopped rapidly and rever-B16 F10 Channel number beyond protein kinase C in the signal cascade that ultimately leads to DNA synthesis.
By contrast, melanoma cells continue to proliferate in the presence of 100l M verapamil. This was the first cell line tested so far in this laboratory that continued to proliferate in the presence of verapamil without having been selected for resistance. However, addition of verapamil seems to induce a differentiation pathway because these cells start to produce melanin and die after verapamil is removed. It remains to be established whether this induction of differentiation can be exploited for in vivo treatment.
Since c-myc gene expression is known to be tightly linked to cell proliferation we were interested in the effects of 250 verapamil on the expression of this gene. This linkage is not completely straightforward, however, because in some in vitro differentiation model systems c-myc expression is harvested, total RNA was isolated and 20pg e' denaturing agarose gels. The filters were I 107 c.p.m. 32P nick translated (> 108 c.p.m. per p insert as described in the text. Lower panels: were reprobed with 32P-labelled p5B for analysi The intensities of the signals are: B16FIO: conti VP=145%, 24h VP=92%, VP-release=4 control=100%, 6h VP=37%, 12h VP=44%, VP-release = 58%. sibly by 100 pM verapamil. The results wil parallel our recent findings on a variety of bi lines and normal fibroblasts (Schmidt et contrast to those cells, however, the B10.BR continuously incubated with the mitogen seems that verapamil induces the same re, incubated with and without this mitogen. Ou that verapamil exerts its antiproliferative effe( slightly increased following the induction of differentiation (Curran & Morgan, 1985;Lachman & Skoultchi, 1984), and myc protein levels can stay unchanged while myc mRNA levels are decreased (Wingrove et al., 1988).
Our results show that c-myc mRNA levels parallel the effects of verapamil, decreasing in the melanocytes while the cells are arrested, whereas in the B16 cells which continue to proliferate c-myc expression is transiently increased. Subsequent Southern analyses of c-myc sequences did not indicate any obvious alterations in the c-myc gene in the melanoma cells as compared to normal spleen cells. Obviously, alterations in this gene do not seem to be responsible for the different responses of the respective cell lines to verapamil. It remains to be established whether the different levels of c-myc expression result from changed gene tranmelanocytes and scription or changed mRNA stability, because it is possible cell cycle transit. that verapamil affects post-transcriptional mechanisms that verapamil, cells control the concentration of c-myc mRNA (Blanchard et al., ,000 propidium 1985) differently in these cells. Post-transcriptional modulametry. G1 peaks tion of c-myc mRNA can be mediated by a labile degrada-IR cells, and to tive protein and depends on active protein synthesis (Santos cells. Triplicates et al., 1988). Inhibition of protein synthesis results in superinduction of c-myc mRNA. Because verapamil stops protein synthesis in cells that are arrested (Schmidt et al., 1988), our data indicate that there is rather no post-transcriptional modulation in the B10.BR cells as there is no superinduction of c-myc mRNA. The B16 cells show a transient decline in the number of cells entering into S-phase and a concomitant synchronised progression of the cells that have been in Sphase through the cell cycle. The transient increase in c-mvc expression in this cell line induced by verapamil appears to c-myc be a consequence of the distribution of the cells in the mitotic cycle because c-myc is expressed slightly higher in Gl -18S rRNA phase of the cell cycle, particularly after induction of differentiation (Lachman et al., 1985). After removal of verapamil melanocytes and and extensive cell death there is also a decrease in the c-myc on c-myc gene signal in the B16 cells. cells were incu-Thus, we conclude that the changes in c-myc expression in ated. Cells were B16 and B10.BR cells induced by verapamil are secondary to ach resolved on other effects of the calcium channel blocker paralleling the hybridised with distribution of cells in the cell cycle. g) c-myc cDNA There is recent evidence that verapamil blocks Na+/H+ the same filters exchange in cultured cells thereby interfering with the rolf 100%, 12h alkalinisation of the cytoplasm required for proliferation to 47%; BIO.BR: begin (Hunter et al., 1986;Bhalla & Sharma, 1986). In the 24h VP=42%, model of the signalling cascade that leads to DNA synthesis, alkalinisation of the cytoplasm occurs after increases in cytoplasmic calcium concentration (Berridge, 1984). As we have shown that verapamil seems to exert its effects indepenth this cell line dently of calcium fluxes (Schmidt et al., 1988)