![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 62, Issue 2, 210-219, August 2002
Laboratory of Retina and Optic Nerve Research, Departments of Ophthalmology (G.E.B., C.S., C.J., S.B., M.E.M.K.), Pharmacology (G.E.B., C.S., L.E., C.J., M.E.M.K.), and Physiology & Biophysics (G.E.B., G.R., S.B.), Dalhousie University, Halifax, Nova Scotia, Canada
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
We show that mitogenic cells expressing T-type Ca2+ channels (T-channels) are more sensitive to the antiproliferative effects of the drugs pimozide and mibefradil than cells without significant T-channel expression. The growth of Y79 and WERI-Rb1 retinoblastoma cells, as well as MCF7 breast cancer epithelial cells, all of which express T-channel current and mRNA for T-channel subunits, is inhibited by pimozide and mibefradil with IC50 values between 0.6 and 1.5 µM. Proliferation of glioma C6 cells, which show little T-channel expression, is less sensitive to these drugs (IC50 = 8 and 5 µM for pimozide and mibefradil, respectively). Neither drug seems to alter cell cycle or the expression of cyclins. Although this strong correlation between T-channel expression and growth inhibition exists, the following results suggest that the drugs inhibit cell growth via different cytotoxic pathways: 1) pimozide and mibefradil have additive effects on T-channel current inhibition, whereas the antiproliferative activity of the drugs together is synergistic; 2) an increase in the number of apoptotic Y79 and MCF7 cells and a decrease in the mRNA for the antiapoptotic gene Bcl-2 is detected only in pimozide-treated cells, whereas in mibefradil-treated cells, the toxicity is primarily necrotic; and 3) growth inhibition by mibefradil is reduced in Y79 cells transfected with T-channel antisense and in differentiated Y79 cells (which have decreased T-channel expression), but growth inhibition by pimozide is affected to a lesser extent. These results suggest that pimozide and mibefradil inhibit cell proliferation via different cytotoxic pathways and that in the case of pimozide, it is unlikely that this effect is mediated solely by T-channel inhibition.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Mibefradil
and pimozide share several common biological properties, which include
Ca2+ channel blockage and, in some cell types,
inhibition of cell growth. Mibefradil, a benzimidazolyl-substituted
tetraline derivative similar in structure to verapamil (Clozel et al.,
1990
), is distinguished from other Ca2+ channel
antagonists because it preferentially blocks low-voltage-activated T-type Ca2+ channels (T-channels) with 10 to 20 times more selectivity than high-voltage-activated (HVA) L-type
Ca2+ channels (Mishra and Hermsmeyer, 1994; see
Lacinová et al., 2000 for a review). Mibefradil seems to bind to
a unique receptor site that overlaps with verapamil and indolizine
sulfone sites and is also able to interfere with the binding of
Ca2+ channel antagonists, such as diltiazem, to
dihydropyridine receptors without affecting dihydropyridine
binding (Rutledge and Triggle, 1995; Bernink et al., 1996; Glasser,
1998). Mibefradil was introduced clinically in 1997 as an antianginal
and antihypertensive agent but was withdrawn from the market less than
a year after its release due to potentially life-threatening
interactions when mibefradil and 
-blockers were taken in combination
with, or acutely replaced by, dihydropyridine
Ca2+ channel blockers (Mullins et al., 1998).
The diphenylbutylpiperidine antipsychotic drug pimozide has also been
shown to be a potent inhibitor of T-type Ca2+
channels but with less selectivity than mibefradil (Galizzi et al.,
1986). In pituitary and heart cells, pimozide inhibits L-type Ca2+ channels (Enyeart et al., 1990), whereas in
adrenal glomerulosa and spermatogenic cells, it blocks T-channels and
Ca2+ influx (Enyeart et al., 1993; Arnoult and
Florman, 1998). The actions of pimozide and mibefradil are not
restricted to Ca2+ channels but may also affect
other ion channels, including K+ (Gomora and
Enyeart, 1999) and Cl
channels (Nilius et al.,
1997).
Besides their capacity to inhibit T-current, both drugs inhibit the
growth of several cell types (Lee and Hait, 1985; Strobl et al., 1990;
Schmitt et al., 1995, 1998). However, the functional role of T-channels
in cell growth and the pharmacological properties of these compounds as
antiproliferative drugs remain unclear. Recently, it was shown that in
cells overexpressing cloned T-channels, there was no alteration in the
cell-doubling population time (Chemin et al., 2000), although it has
been proposed that the inhibition of smooth muscle proliferation and
neointima formation by mibefradil is due to its effect on T-channels
(Schmitt et al., 1995). Pimozide has been shown to inhibit tumor cell
growth of astrocytoma cells, and the apparent mechanism of action may
be via inhibition of calmodulin activity (Lee and Hait, 1985). In
breast cancer epithelial cells, the antiproliferative activity of
pimozide was thought to be associated with its 
-2-binding
capability (Strobl et al., 1998). In addition to actions on
Ca2+ and K+ channels, other
studies documenting the antiproliferative activity of mibefradil on
endothelial cells have also suggested that the effect of this drug
could be associated with actions on Cl channels
(Nilius et al., 1997).
T-type Ca2+ channels are the predominant
Ca2+ channel expressed in mitogenic Y79 cells and
the WERI-Rb1 retinoblastoma cell lines that are derived from human
retinoblastoma tumors. The cells arise from a primitive neuroectodermal
cell and retain the capability to differentiate to neuron or glia-like
cells (Kyritsis et al., 1984). Recently, we demonstrated that T-current
and mRNA expression of T-channel genes is diminished when Y79
retinoblastoma cells are induced to exit the cell cycle and
differentiate (Hirooka et al., 2002). This suggests that T-channels may
be important in membrane potential changes and alterations in
Ca2+ that occur during proliferation and
differentiation of Y79 retinoblastoma cells.
In the present study, we examined the growth-inhibitory actions of the
Ca2+ channel blockers pimozide and mibefradil in
inhibiting cell growth in Y79 and WERI-Rb1 retinoblastoma cells and
compared results with those of MCF7 breast cancer epithelial cells and
C6 glioma cells, two cell lines in which pimozide has been previously
demonstrated to inhibit cell proliferation (Lee and Hait, 1985; Strobl
et al., 1990, 1998). Our studies examined whether mibefradil and
pimozide share common cytotoxic mechanisms and whether the observed
antiproliferative actions of these agents might involve T-channel inhibition.
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Cell Culture. Y79 and WERI-Rb1 (human retinoblastoma), MCF7 (human breast cancer epithelial), and rat glioma C6 cell lines were maintained in Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich, St. Louis, MO) plus 0.02% glutamine supplemented with 10% fetal bovine serum (Invitrogen, Burlington, ON, Canada), without antibiotics. The cells were kept in a humidified 5% CO2-air atmosphere at 37°C, and the medium was changed twice a week.
Chemical differentiation of Y79 cells was performed in 96-well plastic culture dishes. The dishes were pretreated with poly-D-lysine (100 µg/ml) at room temperature for 6 min, washed once with DMEM, and coated with laminin (10 µg/ml) at 37°C for 30 min. The cells were then rinsed with DMEM and incubated with DMEM plus N2 neuronal supplement (insulin, progesterone, transferrin, selenium, and putrescine; Invitrogen) (Albini et al., 1992). This
method yielded approximately 70 to 80% differentiated neuronal-like cells.
Pimozide (Sigma-Aldrich), mibefradil (Hoffman-La Roche, Nutley, NJ),
and nifedipine (ICN Pharmaceuticals Biochemicals Division, Aurora, OH)
were dissolved in dimethyl sulfoxide (DMSO) and diluted to final
concentrations in culture medium. Control cells were exposed to the
same concentrations of DMSO, which never exceeded 0.05%.
For cell transfection, Y79 cells were transfected with
antisense (nucleotide 2084-2103; GenBank accession no. AF126966) or
scrambled fluorescein isothiocyanate-labeled thio-protected oligonucleotides against 
1G (Bijlenga et al., 2000). Cells
(3-4 × 106 cells/ml) were transfected in
serum-free DMEM using Lipofectin (Invitrogen) in the presence of 10 µg of the oligonucleotides according to the manufacturer's protocol.
After incubation with Lipofectin, the cells were cytocentrifuged,
resuspended, and maintained in growth medium for 48 h before
measuring T-current and cell growth. Only cells fluorescing as a result
of the presence of the fluorescein isothiocyanate-labeled
oligonucleotide were chosen for T-current measurements. The
transfection efficiency was approximately 50%.
Measurement of Cell Growth Inhibition and Cytotoxicity. Cell proliferation was quantified using the Cell Titer 96 AQ One Solution Cell Proliferation Assay Kit (Promega, Madison, WI). Cells were plated at 1 × 104 cells/well in 96-well culture plates and treated with different concentrations of pimozide, mibefradil, and nifedipine. The quantity of formazan product formed, which is directly proportional to the number of viable cells, was measured on an enzyme-linked immunosorbent assay reader at 490 nm and converted to cell numbers using a standard curve with different numbers of cells ranging from 1 × 103 to 5 × 104 cells/well.
To measure apoptosis, cytocentrifuged Y79 cells or monolayer cultures for MCF7 cells were stained with a solution composed of acridine orange (100 µg/ml) and ethidium bromide (100 µg/ml) in phosphate-buffered saline (Giuliano et al., 1998). Viable (fluoresce green), necrotic
(fluoresce orange), and apoptotic (fluoresce orange with nuclei
condensed and pyknotic) cells were visualized under a fluorescent
microscope with filters appropriate for fluorescein (495-nm primary
filter and 515-nm secondary filter). The percentage of viable,
apoptotic, and necrotic cells after pimozide and mibefradil treatment
was calculated and analyzed after counting by a person blind to the treatment.
Measurement of Cell Cycle. Cellular DNA content was determined by fluorescence measurement of propidium iodide (PI)-stained cells in a FACScan (BD Biosciences, San Jose, CA). Y79 retinoblastoma cells were treated with 1 M of pimozide or mibefradil for 24 h. Cell were then harvested and fixed in 70% ethanol for 2 h. Subsequently, the fixed cells were incubated in a solution of Ca2+- and Mg2+-free phosphate-buffered saline containing 50 µg/ml RNase A and 200 µg/ml propidium iodide for 1 h. For each cell population, 1 × 104 cells were counted, and the proportions that were in G0/G1, G2/M, and S phases were analyzed using the FCS Express program (De Novo Software, Thornhill, ON, Canada).
RT-PCR.
Total RNA was obtained from cells using TRIzol
(Invitrogen) according to the manufacturer's protocol. Dnase
I-treated RNA samples (2-µg) were used to generate single-stranded
cDNA with a mouse mammary tumor virus reverse transcriptase kit
(Invitrogen). All PCR amplifications were carried out in a total volume
of 25 µl with 200 ng of cDNA, 400 nM primers, 50 mM KCl, 10 mM
Tris-HCl, pH 8.0, 1.5 mM MgCl2, 200 µM dNTP,
and 1.25 units of recombinant Taq polymerase (MBI Fermentas,
Burlington, ON, Canada). Primer sequences have been published
previously for 1G, 1H, and cyclophylin (Hirooka et al., 2002),
Bax and Bcl-2 (Wang and Phang, 1995), cyclin A (Maas et al.,
1995), and cyclins D (D1, D2, and D3; Sola et al., 1999). To compare
the mRNA expression level between treated and untreated cells, the
number of PCR cycles was determined to be within the linear range. The
number of PCR cycles and the annealing temperature for each gene
analyzed was 1) 1G and 1H, 33 cycles at 63°C; 2) cyclophilin,
26 cycles at 50°C; 3) Bax and Bcl-2, 31 and 35 cycles,
respectively, at 65°C; and 4) cyclins D1, D2, and D3 together and
cyclin A, 35 and 24 cycles, respectively, at 51°C.
Electrophysiology.
Whole-cell voltage-clamp recordings were
made from Y79, WERI-Rb1, glioma C6, and MCF7 cells. For recording,
cells were plated on glass coverslips, placed in a 1-ml experimental
chamber, and continuously superfused with external solutions. For
T-current recordings, the cells were superfused with bath solution
containing 150 mM NaCl, 5 mM KCl, 1.5 mM CaCl2,
20 mM BaCl2, 1 mM MgCl2, 5 mM glucose, and 5 mM HEPES, pH 7.4. The patch pipette contained 155 mM
CsCl, 5 mM HEPES, and 1 mM EGTA, pH 7.2. In some recordings of MCF7
cells, 1 mM EGTA was replaced with 5 mM EGTA in the internal pipette
solution. All recordings were made at room temperature (21-24°C).
Pimozide and mibefradil were made as stocks in DMSO, dissolved in
external solutions, and superfused at concentrations cited under
Results. In all cases, the final DMSO concentration did not
exceed 0.1%, a concentration previously shown not to affect Ca2+ currents (Hirooka et al., 2000).
resistance when filled with internal
solutions. The bath solution was connected to the Ag/AgCl bath
electrode by a 1% agar bridge.
Membrane potential and currents were recorded with an Axopatch 1-D
amplifier (Axon Instruments, Inc., Union City, CA). Signals were
filtered at 2 kHz (
3-db, 4-pole Bessel filter) and digitized at 4 kHz
using an Indec Systems interface (Indec Systems, Inc., Capitola, CA)
directly to a PC-compatible computer. Stimulus generation, data
acquisition, and plotting were controlled by BASIC-FASTLAB program
software (Indec Systems). Before seals were made on cells, offset
potentials were nulled using the amplifier circuitry. Capacitance subtraction and series resistance compensation was used in all recordings.
Statistical Analysis. All data are presented as mean ± S.E.M. unless otherwise stated. The statistical methods used were repeated-measures analysis of variance and two-tailed Student's t test for unpaired data, when appropriate. p < 0.05 was considered statistically acceptable.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Pimozide and Mibefradil Inhibit Cell Growth. To examine the effect of the pimozide and mibefradil on growth of retinoblastoma cells and compare this with MCF7 and C6 glioma cells, we first estimated the population-doubling time of each cell line. Retinoblastoma Y79 and WERI-Rb-1 cells showed similar duplication time in the log phase, which was approximately 50 to 60 h, whereas glioma C6 and MCF7 had a shorter time (30-36 h; data not shown). Thus, the effect of the compounds was tested at times corresponding to approximately half of the duplication time for each cell line (24-28 h for retinoblastoma cells and 16-18 h for glioma C6 and MCF7 cells).
Figure 1, A and B, shows that mibefradil and pimozide inhibited retinoblastoma cell proliferation in a dose-dependent manner. The IC50 values for the Y79 cell line were 0.9 ± 0.2 (n = 8) and 1.2 ± 0.4 µM (n = 8) for pimozide and mibefradil, respectively (Fig. 1A). Similar antiproliferative activity was observed with both drugs when tested on WERI-Rb-1 cells. The IC50 for pimozide was 1.2 ± 0.6 (n = 6) and for mibefradil was 1.5 ± 0.5 µM (n = 6) (Fig. 1B). Mibefradil and pimozide also inhibited the proliferation of glioma C6 cells, although at higher doses than that required for retinoblastoma cells. For C6 cells, the IC50 for pimozide was 8.0 ± 2.1 (n = 4), which is consistent with previously reported values (10 µM; Lee and Hait, 1985), whereas the
IC50 we obtained for inhibition of C6 growth by
mibefradil was 5.0 ± 1.0 µM (n = 4) (Fig. 1C).
In MCF7 cells, the IC50 values for inhibition of
proliferation by pimozide and mibefradil were 0.6 ± 0.1 (n = 6) and 1.1 ± 0.2 µM (n = 6), respectively. The IC50 values in MCF7 cells
are similar to values previously reported for pimozide inhibition in
these cells (0.6-3 µM; Strobl et al., 1990, 1998) and similar to the
values we obtained in retinoblastoma cells (pimozide, p > 0.05; mibefradil, p > 0.05) but are significantly lower than the IC50 values we obtained in glioma
C6 cells (pimozide, p < 0.001; mibefradil,
p < 0.001).
|
). Dose-response assays for
growth inhibition in retinoblastoma, C6 glioma, and MCF7 cells
demonstrated that nifedipine did not exhibit a significant growth
inhibitory effect and only inhibited 20 to 40% of the cell proliferation at the maximum concentration tested (15-20 µM) in all
four cell lines (Fig. 1, A-C), possibly reflecting T-channel block or
other nonspecific effects.
We further investigated whether the growth inhibition observed with
pimozide and mibefradil in retinoblastoma and MCF7 cells could be
additive or synergistic. In the case of common mechanisms/pathways of
inhibition, the antiproliferative effect of pimozide and mibefradil should be additive, and the cell growth inhibition in the presence of
both drugs should be close to the sum of the levels measured for
pimozide and mibefradil alone. On the other hand, if the drugs act by
similar and/or independent pathways to produce growth inhibition, then
a synergistic effect would be expected.
Figure 2A shows growth inhibition curves
for Y79 cells in the presence of pimozide and mibefradil alone or when
the two drugs were combined. Inhibition when the drugs were combined
was much greater than the theoretical additive effects, which are shown in the dashed line. At combined concentrations for pimozide and mibefradil of 0.25, 0.5, and 1 µM each, the block was significantly greater than when drugs were applied alone.
|
Cytostatic or Cytotoxic Effects of Pimozide and Mibefradil.
Inhibition of cell growth may be attributed, at least in part, to
arrest of cells in a specific phase of the cell cycle. To determine
whether pimozide and mibefradil could produce alterations in cell cycle
regulation, we investigated whether either of the drugs could inhibit
the proliferation of Y79 cells by arresting and/or killing cells in a
specific phase of the cell cycle. Figure 3A shows the cell cycle analysis of
PI-stained cells using a FACScan. Unsynchronized Y79 cells seeded at a
density >1 × 106 cell/ml were counted
after 24 h of incubation with 1.5 µM pimozide and mibefradil.
The percentage of control cells in
G0/G1, S, and G2/M phase was 46.8 ± 2.9, 15.9 ± 1.0, and 37.2 ± 2.9% (n = 3), respectively.
Similar percentages were obtained in the same cell cycle phases in
pimozide- and mibefradil-treated cells with 45.8 ± 3.9, 15.9 ± 3.6, and 38.2 ± 5.6% of cells in
G0/G1, S, and
G2/M phase in the mibefradil-treated group and
49.0 ± 2.8, 13.7 ± 1.0, and 37.2 ± 2.5% in the
pimozide-treated groups, respectively (n = 3).
|
). We obtained PCR products
for all D cyclins and cyclin A in Y79 retinoblastoma cells using cDNA
from undifferentiated Y79 cells as a template. However, expression of
product for cyclin D1 (428 bp) was lower than cyclins D2 (353 bp) and
D3 (243 bp), and its detection required increased PCR cycles saturating
the densitometric detection for semiquantitative analysis of cyclins D2
and D3 (data not shown). Figure 3, B and C, shows RT-PCR and densitometric analysis of cDNA derived from control and drug-treated Y79 cells. The concentrations of pimozide and mibefradil used (1 µM)
were similar to those that produced 50% growth inhibition in the cell
proliferation assays. A constitutive housekeeping gene, cyclophilin,
was used as an internal control for PCR reactions and produced a 371-bp
PCR product. No significant changes in cyclins D1, D2, or D3 and cyclin
A expression were detected after 8 h of treatment with pimozide
and mibefradil. Similar findings for cyclin expression were obtained in
three other experiments. These results, together with the flow
cytometry results, suggest that pimozide and mibefradil exert their
growth-inhibitory effects on retinoblastoma cells via mechanisms
independent of cell cycle alterations.
Previous studies have suggested that the antiproliferative actions of
pimozide may be attributed to the initiation of apoptotic cell death
pathways (Strobl et al., 1998). Further experiments using a cell
viability assay with acridine orange and ethidium bromide examined the
cytotoxic actions of pimozide and mibefradil in retinoblastoma cells
and compared this with untreated control cells. In this assay, dead
cells stain orange, and viable cells appear green. Apoptotic and
necrotic cells are distinguished by morphological changes (Giuliano et
al., 1998). Figure 4A shows fluorescent
photomicrographs of stained cells in control (C), 1.0 µM pimozide
(P)-treated, and 1 µM mibefradil (M)-treated groups. Control cell
groups consisted of primarily of viable cells, which exhibited green
fluorescence (Fig. 4A, left). In the pimozide group, a number of
apoptotic cells, small cells with condensed fluorescent orange nuclei,
often with yellow shift color, are apparent (Fig. 4A, middle). In the
mibefradil group, cells appeared necrotic with evidence of cell
swelling and both cytoplasmic and nuclear orange fluorescence (Fig. 4A,
right). Mean data from cell viability assays are shown in Fig. 4B. In
control groups, approximately 85 to 90% of the cells were viable,
whereas the rest (10-15%) showed a similar number of necrotic and
apoptotic cells. Three-fold more dead cells (30-40%) were observed in
cells treated with pimozide. This was due almost exclusively to the
higher number of apoptotic cells present in the pimozide-treated group.
In contrast, the increased cell death (30-40%) in mibefradil-treated
cultures seemed to be attributable to an increase (30%) in necrotic
cells.
|
). Figure 4, C and D, shows RT-PCR and densitometry
analysis for Bax and Bcl-2 in Y79 retinoblastoma cells after
pimozide and treatment. A decrease in Bcl-2 mRNA expression
was only detected in pimozide-treated cells. No significant change in
Bcl-2 was observed in mibefradil-treated cells, and no
effect in Bax mRNA expression was detected after 8 h of treatment
with either pimozide or mibefradil. Similar findings were obtained in
two other experiments. These results support the cell viability assays
and suggest that the growth-inhibitory actions of pimozide in
retinoblastoma cells occur via activation of apoptotic pathways of cell
death, whereas the cytotoxic actions of mibefradil may occur via
different mechanisms that result primarily in necrotic loss of cells.
T-Channel Expression and Effects of Pimozide and Mibefradil.
We examined whether the growth-inhibitory effect of pimozide and
mibefradil could be accounted for, at least in part, by block of
T-channels. We first confirmed the presence of T-channels using RT-PCR
with primers specific for 1G and 1H Ca2+
channel subunit RNA, using cDNA derived from undifferentiated cycling
Y79 or WERI-Rb1 cells, and compared this with product obtained with
cDNA from differentiated retinoblastoma cells that had been growing on
a poly-D-lysine/laminin substrate for 6 to 12 days in
defined medium. Under these conditions, retinoblastoma differentiated
primarily into a neuronal phenotype. Because both drugs also inhibited
cell growth in C6 glioma cells and MCF7 cells, we also determined
whether T-channels were expressed in these cell lines. Figure
5A shows that, in accordance with our
previous findings in Hirooka et al. (2002), undifferentiated
retinoblastoma cells (Y79U) and WERI-Rb1 retinoblastoma cells have mRNA
for both 1H (435-bp) and 1G (395-bp) T-type
Ca2+ channels, the expression of which was
decreased on exit of cells from the cell cycle and differentiation into
a neuronal phenotype (Y79D). Under the conditions used, glioma C6 cells
failed to shown any discernible PCR product for either 1H or 1G
T-channels, and patch-clamp recordings from these cells did not detect
measurable T-current (data not shown). PCR product was obtained for
both 1H and 1G T-channels in MCF7 cells (Fig. 5B).
|
80 mV to potentials positive to 60 mV. Figure 5D shows mean
current-voltage relations from 16 cells. Inward current activates at
50 mV, with peak current at 10 mV. The amplitude of T-current in
MCF7 cells varied between 10 and 168 pA (n = 16), and
the mean T-current density determined in a subset of cells was
0.80 ± 0.35 pA/pF (n = 5), which is substantially
less than current density previously reported in retinoblastoma cells
under the same conditions (5.9 ± 1.1 pA/pF; Hirooka et al.,
2002). Mean values for T-current block by mibefradil and pimozide in
MCF7 cells (Fig. 5E), however, were similar to those found in
retinoblastoma (Hirooka et al., 2002), with approximately 50% of
T-current activated at 10 mV inhibited by both mibefradil and
pimozide at 1 µM. Current was almost completely inhibited (>75%) at
5 µM. These values are also consistent with
IC50 values for growth inhibition in
retinoblastoma and MCF7 cells by pimozide and mibefradil. In C6 glioma
cells in which we were unable to detect T-current, both mibefradil and pimozide required 10-fold higher concentrations to produce growth inhibition. Normalized, mean current-voltage plots for MCF7 cells treated with 1 and 5 µM of mibefradil revealed that although current inhibition was observed at all the voltages tested, T-current inhibition by mibefradil was increased at more positive voltages, and a
shift was observed in the current-voltage relationships to more
negative potentials (Fig. 5F).
We next examined whether the blocking effects of pimozide and
mibefradil on T-currents in Y79 retinoblastoma cells were synergistic, as determined for the antiproliferative actions of these agents. Pimozide and mibefradil were administered alone and in combination to
determine the degree to which inhibition of T-currents occurred. Figure
6A shows that pimozide and mibefradil
alone blocked the T-current with IC50 values of
0.83 and 0.60 µM, respectively. In contrast to the synergy in cell
growth inhibition, combined administration of the drugs produced only
additive inhibition of T-current.
|
1G T-channels revealed a >70% reduction in T-current, suggesting that 1G T-type Ca2+ channels may be
the most abundant T-channel in retinoblastoma cells. Figure 6B shows
mean T-current recorded at 10 pA in antisense-transfected cells
(42 ± 8 pA, n = 11) compared with T-current
recorded in sense-transfected (107 ± 20 pA, n = 11) or control cells (105 ± 19 pA, n = 10),
respectively. Figure 6C shows that cells transfected with 1G
antisense oligonucleotide were less sensitive to the growth inhibition
mediated by mibefradil than sense-transfected and control cells. The
inhibitory effect produced by pimozide seemed to be similar in control,
sense, and antisense transfected cells, suggesting that in contrast to
mibefradil, inhibition of T-channels is not essential for the cytotoxic
actions of pimozide.
A test of this result was obtained by comparing the growth inhibition
of both mibefradil and pimozide in undifferentiated cycling
retinoblastoma cells and differentiated noncycling retinoblastoma cells, in which T-channel expression and T-current is decreased (see
Fig. 5A). The cytotoxic effects of pimozide and mibefradil (5 µM)
were examined on both undifferentiated and differentiated cells after
24 h of drug treatment. Figure 6D shows that, as observed in the
1G antisense-transfected retinoblastoma cells, the number of
differentiated retinoblastoma cells diminished by only 40% with
mibefradil treatment compared with approximately 80% in the undifferentiated retinoblastoma cell cultures. The growth
inhibitory-effect of pimozide, however, was similar in both
undifferentiated (89 ± 5% reduction in cell number) and
differentiated (83 ± 8% reduction in cell number) retinoblastoma
cells (Fig. 6D).
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
We observed that the T-channel-blocking drugs pimozide and mibefradil have potent dose-dependent antiproliferative and cytotoxic effects in retinoblastoma and MCF7 cells. The 50% growth inhibition for these drugs occurred at concentrations between 0.6 and 1.5 µM and at lower doses when mibefradil and pimozide were combined to produce synergistic growth inhibition. Examination of the cytotoxic effects of these drugs suggested that mibefradil and pimozide produced cell death via different mechanisms independent of alterations in cell cycle regulation. Pimozide-treated cells underwent apoptosis, whereas mibefradil-treated cells appeared to undergo necrosis.
A common role for T-channels in the antiproliferative effect of these
drugs was suggested by the good correlation between 50% T-current
inhibition and the growth inhibition produced by pimozide and
mibefradil in undifferentiated Y79 cells and MCF7 epithelial cells at 1 µM. However, several findings also supported distinct mechanisms of
cytotoxicity for both mibefradil and pimozide. In the case of pimozide,
these seemed to be independent of T-channel inhibition. First, cell
growth inhibition in Y79 cells with concomitant administration of both
drugs was synergistic, but T-current inhibition was additive. Second,
retinoblastoma cells with 1G T-channel knockdown by antisense
oligonucleotide transfection were less sensitive to growth inhibition
by mibefradil but not pimozide. Third, differentiated retinoblastoma
cells with diminished T-current and with low expression of mRNA for
both the T-type Ca2+ channels 1G and 1H
were less sensitive to the growth inhibition mediated by mibefradil but
not to pimozide. Taken together, these data suggest that pimozide and
mibefradil induce cell death in retinoblastoma and MCF7 cells by
different pathways. In the case of mibefradil, the mechanism of cell
death was primarily necrotic and largely dependent on block of
T-channels. However, cell death caused by pimozide seemed to involve
apoptosis and did not rely on T-channel block.
Pimozide and Mibefradil: Actions on Cell Proliferation.
Pimozide has been reported to inhibit cell proliferation in a number of
cell types and cancer cell lines (Lee and Hait, 1985; Strobl et al.,
1990). In MCF7 cells, the antiproliferative activity of pimozide was
associated with interaction with -2 receptors (Strobl et al., 1998).
-2 receptors are drug-binding proteins that are expressed in high
densities in a variety of tumor types (Vilner et al., 1995) and bind
structurally distinct psychoactive agents, including various
neuropleptic drugs. Investigations of the apoptosis and cytotoxicity
associated with -2 receptor ligands in MCF7 cells revealed that this
was p53-independent and did not seem to be dependent on caspase
activation (Crawford and Bowen, 2002). Studies examining the actions of
-2 agonists in neuroblastoma cells and colon and mammary carcinoma
cells demonstrated that ligands that interacted with -2 receptors
could release Ca2+ stores, which triggered
subsequent extracellular Ca2+ influx after the
depletion of the stores. Prolonged exposure to ligands resulted in
cell death via apoptosis (Brent et al., 1996; Vilner and Bowen, 2000).
In C6 glioma cells, the dose-dependent cytotoxic effect of ligands
was reported to result in receptor-mediated alterations in cellular
morphology, cessation of cell division, and inhibition of calmodulin
activity (Lee and Hait, 1985; Vilner et al., 1995).
), our data would suggest that the pimozide-mediated apoptosis observed was caspase-independent or involved activation of caspases other than
caspase 3. The IC50 for growth inhibition by
pimozide was also in the range of published values for block of
T-channels in a number of other cell types. However, in retinoblastoma
cells treated with 1G T-channel antisense or in differentiated
retinoblastoma cells in which T-channel expression is significantly
decreased, the IC50 for pimozide cytotoxicity was
unchanged. This suggests that although pimozide may block T-channels,
this is not the primary mechanism triggering cell death. Synergistic
growth inhibition by pimozide with mibefradil, another
T-channel-blocking drug, in contrast to additive actions on T-currents
suggests that additional mechanism(s) contribute to the cytotoxicity.
This may also include -2 receptor activation.
Mibefradil has also been reported to inhibit the proliferation of
various cell types, including blood mononuclear cells (Lijnen et al.,
1999), rat smooth muscle cells (Schmitt et al., 1995), mouse liver
cells (Wondergem et al., 2001), and endothelial cells (Nilius et al.,
1997). In hemopoetic cells, the antiproliferative effect of mibefradil
involved 1) a decrease in DNA synthesis as defined by decreased
[3H]thymidine incorporation and 2) was
dependent on the timing of drug administration and involved
Ca2+ channel block (Lijnen et al., 1999).
Mibefradil was reported to prevent neointima formation after vascular
injury in rats (Schmitt et al., 1995). This in vivo effect of
mibefradil was attributed to inhibition of smooth muscle cell
proliferation and to block of T-channels, based on the increased
density of T-channels present in the proliferating muscle cells and the
lack of effect of other dihydropyridine-type HVA
Ca2+ antagonists. In liver, endothelial, and
neuroblastoma cells, mibefradil was also reported to inhibit cell
proliferation, and this was associated with cell swelling and
inhibition of volume-sensitive Cl channels
(Nilius et al., 1997; Rouzaire-Dubois and Dubois, 1998; Wondergem et
al., 2001).
Our data examining the actions of mibefradil on retinoblastoma and MCF7
cells are supportive of an involvement of T-channel inhibition in
contributing to the cytotoxic actions of this drug; mibefradil
inhibited cell growth and T-current at approximately the same
concentration, and the actions of mibefradil were reduced in cells
treated with 1G antisense or in differentiated retinoblastoma cells
with diminished T-channel expression.
T-Channels and the Cell Cycle.
Although the biological role of
T-channels in cell proliferation has yet to be fully defined, the
expression of Ca2+ channels has been reported to
change with cell differentiation and proliferation. Such changes may
contribute to altered membrane excitability during development or
regrowth after disease. T-channels seem to be preferentially expressed
during development in both neurons and muscle, with a decrease in
expression after differentiation (Kuga et al., 1996). In human
myoblasts, T-type Ca2+ currents are expressed
just before fusion, and their inhibition suppresses fusion and the
increase in intracellular Ca2+ that is normally
observed at the onset of fusion (Bijlenga et al., 2000). Re-expression
of T-channels has been reported in remodeled hypertrophied ventricular
myocytes in which DNA replication is reinitiated in the terminally
differentiated phenotype (for review, see Boyden and Jeck, 1995). These
alterations in cardiac T-channels coincided with alterations in the
cell cycle and suggested that cell cycle changes can regulate
expression of T-channels. In aortic smooth muscle cells, expression of
T-channels was reported to increase during the cell cycle, appearing in
G1 phase and increasing in the S phase.
Nonproliferative muscle cells did not express T-channels (Kuga et al.,
1996).
1G and 1H T-channels are the predominant
Ca2+ channels expressed in cycling, nonspiking
undifferentiated cells (Hirooka et al., 2002). Electrophysiological
studies have revealed that T-channels in these cells produce "window
current" at membrane potentials between 50 and 30 mV (Barnes and
Haynes, 1992) and that T-channel activation can lead to
regenerative Ca2+ spikes (Hirooka et al., 2002).
Exit of these cells from the cell cycle and differentiation into a
neuronal phenotype is accompanied by loss of T-channels and increased
expression of voltage-dependent Na+ and HVA
channels. This suggests a functional role for T-channels in the
regulation of membrane excitability during retinoblastoma cell
proliferation. However, it should be noted that a recent study noted no
significant alterations in the cell proliferation rate in human
embryonic kidney cells overexpressing T-channels (Chemin et al., 2000).
Although mibefradil was found to diminish calcium influx in
overexpressing HEK-293 cells, its effect on proliferation of this cell
line was not studied. The authors concluded that although T-channels
can mediate increases in intracellular Ca2+ and
are important in the regulation of basal Ca2+
levels, increased expression does not seem to affect either cell proliferation or cell cycle kinetics. Although this study does not
eliminate a contribution of T-channels to cell proliferation, it does
suggest that these channels are unlikely to be a primary trigger for
stimulation of cell cycle progression.
Our experiments examining the antiproliferative and cytotoxic actions
of the T-channel-blocking drugs pimozide and mibefradil demonstrate
that these Ca2+ channel blockers did not alter
cell cycle regulation. However, T-channels may be one of several
important cellular targets underlying the cytotoxic actions of
mibefradil. The lack of apparent effect of T-channel block in the
cytotoxic actions of pimozide may reflect a number of factors,
including contributions from different cytotoxic pathways, the
dose-range over which these individual pathways are activated, and well
as time- and voltage-dependent differences in the potency of specific
T-channel subtype block with prolonged drug exposure (Arnoult and
Florman, 1998; Rossier et al., 1998; Perchenet et al., 2000). Taken
together, these results suggest that the Ca2+
channel antagonists exert cytotoxic effects via complex interactions that extend beyond their ion channel-blocking activities.
| |
Footnotes |
|---|
Received February 15, 2002; Accepted April 25, 2002
This work was supported by operating grants and salary awards provided by the Canadian Institutes of Health Research, Canadian Breast Cancer Foundation-Atlantic Chapter, and Cancer Research and Education, Nova Scotia. G.E.B. was supported by the Reynolds Fellowship in Pharmacology and a CaRE-Nova Scotia Trainee Award.
Address correspondence to: Melanie E. M. Kelly, Ph.D., Department of Pharmacology, Dalhousie University, Halifax, Nova Scotia, B3S 4H7 Canada. E-mail: mkelly{at}is.dal.ca
| |
Abbreviations |
|---|
T-channel, T-type Ca2+ channel; DMEM, Dulbecco's modified Eagle's medium; DMSO, dimethyl sulfoxide; PI, propidium iodide; HVA, high-voltage-activated; T-current, T-type Ca2+ channel current; RT, reverse transcription; PCR, polymerase chain reaction; bp, base pair(s).
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
origin from a primitive neuroectodermal cell?
Nature (Lond)
307:
471-473[CrossRef][Medline]. channels in macrovascular endothelial cells.
Br J Pharmacol
121:
547-555[CrossRef][Medline].This article has been cited by other articles:
|
|
 |
|
  M.-G. Feng and L. G. Navar Nitric oxide synthase inhibition activates L- and T-type Ca2+ channels in afferent and efferent arterioles Am J Physiol Renal Physiol, April 1, 2006; 290(4): F873 - F879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Rodman, K. Reese, J. Harral, B. Fouty, S. Wu, J. West, M. Hoedt-Miller, Y. Tada, K.-X. Li, C. Cool, et al. Low-Voltage-Activated (T-Type) Calcium Channels Control Proliferation of Human Pulmonary Artery Myocytes Circ. Res., April 29, 2005; 96(8): 864 - 872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. C. Baptiste, A. T. E. Hartwick, C. A. B. Jollimore, W. H. Baldridge, G. M. Seigel, and M. E. M. Kelly An Investigation of the Neuroprotective Effects of Tetracycline Derivatives in Experimental Models of Retinal Cell Death Mol. Pharmacol., November 1, 2004; 66(5): 1113 - 1122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Cataldi, A. Gaudino, V. Lariccia, M. Russo, S. Amoroso, G. di Renzo, and L. Annunziato Imatinib-Mesylate Blocks Recombinant T-Type Calcium Channels Expressed in Human Embryonic Kidney-293 Cells by a Protein Tyrosine Kinase-Independent Mechanism J. Pharmacol. Exp. Ther., April 1, 2004; 309(1): 208 - 215. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. Yoshida, K. Nosaka, J.-i. Yasunaga, I. Nishikata, K. Morishita, and M. Matsuoka Aberrant expression of the MEL1S gene identified in association with hypomethylation in adult T-cell leukemia cells Blood, April 1, 2004; 103(7): 2753 - 2760. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-G. Feng, M. Li, and L. G. Navar T-type calcium channels in the regulation of afferent and efferent arterioles in rats Am J Physiol Renal Physiol, February 1, 2004; 286(2): F331 - F337. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Ferron, V. Capuano, Y. Ruchon, E. Deroubaix, A. Coulombe, and J.-F. Renaud Angiotensin II Signaling Pathways Mediate Expression of Cardiac T-Type Calcium Channels Circ. Res., December 12, 2003; 93(12): 1241 - 1248. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
, significance of
the difference between the drugs combined and alone at
p < 0.05, p < 0.01, and
p < 0.001 levels. A, there was no difference
between pimozide and mibefradil alone at any concentration
(p > 0.05). At 0.25 µM, the combination differed
significantly from pimozide and mibefradil alone (p < 0.01 and p < 0.05, respectively). At 0.5 and 1 µM, the combination differed significantly from pimozide and
mibefradil alone (p < 0.001 and
p < 0.01, respectively, both concentrations). B,
there was no difference between pimozide and mibefradil alone at any
concentration (p > 0.05). At 0.15 µM, the
combination differed significantly from pimozide and mibefradil alone
(p < 0.001 for both). At 0.32 µM, the
combination differed significantly from pimozide and mibefradil alone
(p < 0.01 for both). At 0.62 µM, the combination
differed significantly from pimozide and mibefradil alone
(p < 0.05 for both). At 1 µM, the combination
differed significantly from pimozide and mibefradil alone
(p < 0.05 and p < 0.01, respectively).
