1 1955 Quinidine Impairs Proliferation of Neurofibromatosis Type 2-Deficient Human Malignant Mesothelioma Cells Tamara Utermark, M.Sc. 1 Alexi Alekov, Ph.D. 2 Holger Lerche, M.D. 1,2 Vincent Abramowski 3 Marco Giovannini, Ph.D. 3 Clemens Oliver Hanemann, M.D. 1 1 Department of Neurology, University of Ulm, Ulm, Germany. 2 Department of Applied Physiology, University of Ulm, Ulm, Germany. 3 INSERM U434, Fondation Jean Dausset-CEPH, Paris, France. Supported by grants from Ligue Nationale contre le Cancer and Association pour la Recherche contre le Cancer, the Deutsche Forschungsgemeinschaft (Le1030/5-2 to HL), and the Bundesministerium für Bildung und Forschung/Interdisziplinäres Zentrum für Klinische Forschung Ulm (project B8 to HL). Address for reprints: C.O. Hanemann, M.D., Department of Neurology, Zentrum für klinische Forschung, University of Ulm, Helmholtzstraße 8/1, Ulm, Germany; Fax: ; Received May 28, 2002; revision received December 19, 2002; accepted December 19, BACKGROUND. Human malignant mesotheliomas (HMMs) are aggressive tumors that arise from the mesothelium. They respond poorly to conventional tumor treatment and outcome is often fatal. Inactivating mutations of the neurofibromatosis type 2 (NF2) tumor suppressor gene merlin have been described in nearly 60% of primary malignant mesothelioma and in approximately 20% of the mesothelioma cell lines. Studies regarding human NF2 schwannoma cells revealed a higher proliferation and a larger noninactivating K outward current compared with controls. The enhanced proliferation of merlin-deficient NF2 schwannoma cells could be reduced in the presence of quinidine, a K channel blocker, whereas the proliferation of normal Schwann cells is not affected. The current study was undertaken to evaluate the effect of quinidine on the proliferation of HMM cell lines in relation to their NF2 status. METHODS. Proliferation analyses using bromodeoxyuridine incorporation was performed by immunocytochemical staining and fluorescence assisted cell sorting. The patch-clamp technique was applied for electrophysiologic characterization of the HMM cell lines. The cytochrome P450 2D6 locus, known to be mutated at high frequencies in NF2 patients and to be specifically inhibited by quinidine, was screened for mutations by cycle sequencing. RESULTS. Quinidine selectively reduces the proliferation of merlin-deficient HMM cell lines by causing a G 0 /G 1 arrest, whereas the proliferation rates of merlinexpressing HMM cell lines remain unchanged. The effect of quinidine on the proliferation of HMM cell lines appears to be correlated with the NF2 gene status but not with the K outward current. No relation to cytochrome P450 2D6 mutations was detected. CONCLUSIONS. Quinidine or quinidine analogs are of potential therapeutic interest for the subset of merlin-deficient mesothelioma tumors. Cancer 2003;97: American Cancer Society. DOI /cncr KEYWORDS: human malignant mesothelioma, neurofibromatosis type 2, schwannoma, proliferation, quinidine, K outward current, cytochrome P450 2D6. Human malignant mesotheliomas (HMMs) are aggressive tumors of mesodermal origin that arise from mesothelial cells lining the pleura, peritoneum, or pericardium. Exposure to asbestos has been implicated as a major contributing factor in the development of this malignancy. However, unexposed cases have been reported. Therapy is very difficult as the tumors respond poorly to conventional tumor treatment and outcome is often fatal. Cytogenetic studies have revealed frequent chromosomal abnormalities in HMMs, including rearrangements or losses in the chro American Cancer Society
2 1956 CANCER April 15, 2003 / Volume 97 / Number 8 mosomes 1p, 3p, 4q, 6q, 9p, 14q, and 22q, suggesting that alteration of critical genes in these regions may be instrumental to the tumorigenic mechanism. 1 3 Supporting this hypothesis, inactivating mutations of the tumor suppressor gene, Neurofibromatosis type 2 (NF2), located on chromosome 22q12 have been described in almost 60% of primary malignant mesothelioma and in approximately 20% of the mesothelioma cell lines. 4,5 It has been suggested that loss of the NF2 gene product contributes to the development or progression of a significant subset of HMMs. 6 The NF2 gene encodes a 66-kilo dalton (kda) protein, termed schwannomin 7 or merlin. 8 A germline mutation and a second hit mutation in the NF2 gene lead to NF2, a hereditary disorder characterized by multiple tumors such as bilateral vestibular schwannomas, cerebral meningiomas, and gliomas. Studies on schwannoma cells from NF2 patients revealed a higher proliferation and a larger noninactivating K outward current compared with controls. 9,10 A correlation between high K outward currents and high proliferation rates has been described in literature for several cell types, such as lymphocytes of murine and human origin, lung carcinoma, murine neuroblastoma, human melanoma, and breast carcinoma cells, as well as for rodent Schwann cells Rosenbaum et al. 9 demonstrated that the enhanced proliferation of primary human NF2 schwannoma cells, which do not express schwannomin/merlin, is reduced in the presence of quinidine, a K channel blocker, whereas the proliferation of normal Schwann cells is not affected. However, NF2 schwannoma cells differ from normal Schwann cells not only in K outward current but also in morphology and proliferation rate. 18,19 Therefore, we decided to test whether the reduction in proliferation was related to mutations in the NF2 gene and to evaluate cell lines from HMMs with defined NF2 status. These cells showed the same proliferation rates, K outward currents, and morphology in vitro and differed only in their genotype. This study shows that quinidine reduces the proliferation of merlin-deficient HMM cell lines only by causing G 0 /G 1 arrest, whereas the proliferation rates of merlin-expressing HMM cell lines remain unchanged. In addtion to its function as a channel blocker, quinidine is an inhibitor of a specific isoform of the cytochrome P450 family (CYP2D6), which is involved in the detoxification and metabolism of multiple commonly prescribed drugs as well as numerous endogenous substrates. In NF2 patients, mutations at the CYP2D6 locus leading to the poor metabolizer (PM) phenotype have been found at a significantly increased frequency (13%) compared with controls. 20 We tested the hypothesis that the distinct effect of quinidine on the merlin-deficient HMM cell lines was based on additional mutations in the CYP2D6 locus, causing impaired enzyme function and leading to poor metabolism and therefore decreased proliferation of these cells. However, no correlation of mutations in the cytochrome P450 2D6 gene to merlin deficiency was detected. The response to quinidine appears to be correlated with the NF2 gene status, but not with the K outward current or with cytochrome P450 2D6 mutations. MATERIALS AND METHODS Cell Lines and Cell Culture The characteristics of the HMM cell lines BAR, BLA, HIB, and TRA were reported previously. 21 The cell lines BAR and TRA express the merlin transcript only at a very low level and there is no merlin protein detectable in the cells, whereas HIB and BLA express merlin at normal levels. The culture methods used were based on those described by Zeng et al. 22 The medium for cell culturing consisted of RPMI-1640 (Gibco BRL, Karlsruhe, Germany) supplemented with 10% fetal calf serum (PAA, Cölbe, Germany), 2 mmol/l glutamine (Gibco), 50 U/mL penicillin (Gibco), and 50 g/ml streptomycin (Gibco). Cells were seeded at a concentration of 10,000 cells/cm 2 and maintained in a humidified atmosphere of 5% carbon dioxide at 37 C. The proliferation medium was changed twice a week and cells were subcultured when confluent by detaching them with a trypsin/ethylenediaminetetraacetic acid (EDTA) mixture (Gibco). Western Blot Analysis Protein was prepared from subconfluent cells using a boiling denaturating lysis buffer (20 mm Tris, ph 7.4; 2% sodium dodecyl sulfate [SDS]) and concentrations were determined with a detergent compatible protein assay (Biorad, Munich, Germany) according to the manufacturer s protocol. Proteins were separated by SDS-polyacrylamide gel electrophoresis on an 8% polyacrylamide gel. Blocking was done in Tris-buffered saline (TBS), 0.1% Tween 20 (Sigma, Munich, Germany), 5% milk, and 2% bovine serum albumin (BSA). The membranes were incubated with a polyclonal antibody to NF2 (C18; Santa Cruz Biotechnology, Santa Cruz, CA), diluted 1:250 in blocking solution for 2 hours at room temperature, followed by incubation with the horseradish peroxidase labeled secondary antibody. ECL (AmershamPharmaciaBiotech, Freiburg, Germany) was used for detection. The experiment was performed three times independently.
3 Quinidine Reduces Proliferation in NF2 -/- HMMs/Utermark et al Electrophysiologic Studies on HMM Cell Lines Currents from all four HMM cell lines were recorded using the whole cell configuration of the patch-clamp technique at day 4 7 after splitting. The standard intracellular solution contained (in mm): 150 KCl, 10 Ethylene glycol-bis[beta-aminoethyl ether]-n,n,n N tetraacetic-acid (EGTA), 1 CaCl 2, 5 MgCl 2, 40 HEPES, ph 7.0. The standard extracellular solution contained (in mm): 170 NaCl, 3 KCl, 2.5 CaCl 2, 1.2 MgCl 2, 10 HEPES, ph Patch-clamp pipettes were pulled from borosilicate capillary glass (1.05 mm inner diameter, 1.5 mm outer diameter; Science Products GmbH, Hofheim, Germany) on a two-stage universal puller (Zeitz Instruments, Augsburg, Germany). The pipettes were heat polished and had final resistances between 0.8 and 1.6 MOhm. Currents were recorded at room temperature (21 22 C) using an EPC 7 amplifier (List Medical Electronics, Darmstadt, Germany), pclamp 6 software, and a Digidata 1200 A-D converter (Axon Instruments, Union City, CA). Cells were held at 80 millivolts (mv) and depolarized to potentials between 80 and 120 mv in 10-mV steps. Data were filtered at 0.5 kilohertz (khz) and digitized at a sampling rate of 1 khz. Leakage and capacitative currents were subtracted using a P/4 protocol. The amplitude of quinidine-sensitive K currents ranged from 0.3 to 2 na for all four cell lines. The final voltage error due to series resistance for the quinidine-sensitive current was always smaller than 3 mv (series resistance compensation up to 70%). Quinidine (Sigma, Munich, Germany) was added to the extracellular solution at a concentration of 25 M. Data analysis was performed using pclamp 6, EXCEL 97 (Microsoft, Redmond, WA) and Origin (MicroCal, Northampton, MA) software. For statistical evaluation, the Student t test was applied. Proliferation Assays Proliferation of cells was measured by visualizing the incorporation of bromodeoxyuridine (BrdU; Roche, Mannheim, Germany) using immunostaining and fluorescence-assisted cell sorting analysis (FACS). Immunostaining Analysis Cells were kept in proliferation medium containing different concentrations of quinidine (Sigma) for 4 days. As the quinidine stock solution was dissolved in dimethylsulfoxide (DMSO), a DMSO dilution was used as the control. The medium was changed every second day. 8.5 hours after the addition of 10 M, BrdU cells were fixed for immunocytochemical detection of BrdU as described by Hanemann et al. 23 The first (anti- BrdU, Roche) and secondary antibodies (rabbit-antigoat biotinylated, Vector Laboratories, Burlingame, CA) were diluted 1:200 in TBS with 2% serum. For every cell line, four independent experiments were performed and for every quinidine concentration, about 500 cells were counted at least three times. The Student t test was used for statistical evaluation. Cell Cycle Analysis Supplementary cell cycle analysis was performed for two of the HMM cell lines (HIB and TRA). Cells were cultured on T25 cm 2 flasks in the presence of different concentrations of quinidine for 4 days. Bromodeoxyuridine was added for 8.5 hours, after which cells were trypsinized and spun down for 10 minutes at 1500 revolutions per minute. Cells were washed twice in 1% BSA/phosphate-buffered saline (PBS) and fixed with 70% ethanol on ice. DNA was denatured by 2N HCl/Triton X100 at room temperature for 30 minutes and sodium tetraborate was added to neutralize the ph level. After centrifugation, cells were resuspended in 1% BSA/0.5% Tween 20/PBS and the cell concentration was adjusted to cells per test. For immunofluorescence staining, 20 L of anti-brdu-fluorescein isothiocyanate (Becton Dickinson, Heidelberg, Germany) per 10 6 cells was added and incubated for 30 minutes at room temperature. Cells were washed once with BSA/Tween 20/PBS and collected by centrifugation at 500 g for 5 minutes, followed by resuspension in 1 ml PBS containing 5 g/ml propidium iodide (PI; Serva, Heidelberg, Germany). An unstained probe was used to determine autofluorescence. Data were acquired on a FACScan (FacsCalibur-Becton Dickinson), linked to an Apple MacIntosh computer with CellQuest software (Becton Dickinson), and further analyzed with CellQuest, excluding all dead cells from the results by PI-based gating. Laser excitation was set at 488 nanometer. The experiment was performed independently three times. The Student t test was used for statistical evaluation. Test of Cytotoxicity Cytotoxicity of quinidine was determined by use of an MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] test (Promega, Mannheim, Germany), by which MTS is bioreduced into formazan by metabolically active cells only. The test was performed according to the manufacturer s protocol. Cells were seeded in 96-well plates at a density of 10 4 cells/cm 2 and incubated for 4 days in proliferation medium containing different concentrations of quinidine or DMSO alone as the control medium. The medium was changed every second day. The MTS (Owen s reagent) and phenazine methosulfate (PMS) solutions were mixed immediately before
4 1958 CANCER April 15, 2003 / Volume 97 / Number 8 being added to the cells. Twenty microliters of MTS/ PMS was pipetted into each well containing 100 L medium. Plates were incubated for 4 hours before the absorbency of formazan was recorded at 492 nm using an enzyme-linked immunosorbent assay plate reader (Tecan, Männedorf, Switzerland). For each cell line, four independent MTS tests were performed. Sequence Analysis of CYP2D6 For sequence analysis of the CYP2D6 locus, DNA of all four HMM cell lines was isolated and the exons were amplified using different sets of primer pairs and Platinum Pfx DNA polymerase (Invitrogen, Karlsruhe, Germany). For isolation of DNA, the QIAamp DNA mini kit (Qiagen, Hilden, Germany) was used according to the manufacturer s protocol. The primer sequences used were: Exon 1 sense 5 -GCAAAGGCCATCATCAGC-3, antisense 5 -AAACGGCACTCAGGACTAACTCAT-3 Exon 2 sense 5 -CCGGGGGTCGTCCAAGGTTCAAAT- 3, antisense 5 -GCCCCGCCCACTCGTCACA-3 Exon 3 sense 5 -TCAGGGTGGGCAGAGACGAG-3, antisense 5 -CTGGGGGTGGGAGATCGGGTAA-3 Exon 4 sense 5 -TGCCGCCTTCGCCAACCACT-3, antisense 5 -CTCCTCCAGGCCCTTCTTACA-3 Exon 5 sense 5 -AACGCAGAGCACAGGAGGGAT- TGA-3, antisense 5 -CTGAGCAGGGCCGAGAG- CATACT-3 Exon 6 7 sense 5 -ACCCCGTTCTGTCCCGAGTA-3, antisense 5 -CAGTGTGGTGGCATTGGG-3 Exon 8 9 sense 5 -CCTGCTGTGGGGTCGGAGAG-3, antisense 5 -CAGGCGGTGGGGTAAGC-3 Polymerase chain reaction (PCR) products were purified using the QIAquick PCR purification kit (Qiagen) and concentration was measured on a 1.5% agarose gel with a DNA mass ladder (Gibco) and analyzed with ImageQuant software (Amersham Biosciences, Freiburg, Germany). Cycle sequencing reactions were performed with the ThermoSequenase II dye terminator cycle sequencing kit (AmershamPharmaciaBiotech), using a g template. The primer concentration used was 5 M and the cycling program included 96 C, 30 seconds; 50 C, 15 seconds; and 60 C, 1 minute for 30 cycles. Reactions were cleaned up by sodium acetate/ EDTA buffer and precipitation with ethanol. The pellet was washed once with ethanol, air-dried, and dissolved in formamide loading dye (70% formamide, 1 mm EDTA). Two microliters of each sample was loaded onto a 5% polyacrylamid gel and run on an ABI377 machine for 8.5 hours at 2500 V. For data acquisition the ABI Prism Collection Software (version FIGURE 1. Western blot analysis of merlin in human malignant mesothelioma (HMM) cell lines. BAR, BLA, HIB, and TRA indicate the loading of total protein lysate of the corresponding HMM cell line. Merlin is detected at 70 kilo daltons (arrow). 1.1) and for data analysis the Sequence Analysis Software was used (both Perkin-Elmer, Foster City, CA). RESULTS Quinidine selectively reduces proliferation of merlindeficient human primary NF2 schwannoma cells, but not of normal Schwann cells. 9 We investigated whether this effect of quinidine is related to mutations in the NF2 gene. We chose cells that show identical morphology, same proliferation rates in culture, and identical K outward current (in contrast to NF2 schwannoma and normal Schwann cells), and that differ in their NF2 gene status. Western Blot Analysis of Merlin in HMM Cell Lines To ensure that merlin expression was as stated present in the HMM cell lines BLA and HIB and absent in the HMM cell lines BAR and TRA, we performed Western blot analysis of merlin expression using total cell lysates. Merlin migrates as expected to 70 kda and is only to be detected in lysates originating from the HMM cell lines BLA and HIB (Fig. 1). Electrophysiologic Studies on HMM Cell Lines All four cell lines showed two different types of noninactivating, outward rectifying currents that were well separated by their voltage dependence and current amplitude (Fig. 2). The smaller current (I S ) was activated at less depolarized potentials ( 20 mv and more positive) and was only observed in approximately 40% of all cells investigated. Its amplitude ranged from 0.3 to 2 na at a test potential of 30 mv. The larger current (I L ) was found in all cells and was activated at voltages of 50 mv and more positive values. Its amplitude at 100 mv ranged from 2.5 to 8.5 na. To characterize these two types of currents, two different voltage clamp protocols were used. For
5 Quinidine Reduces Proliferation in NF2 -/- HMMs/Utermark et al FIGURE 2. Electrophysiologic results. (A, B) Families of raw current traces recorded from two mesothelioma cells showing two types of outward rectifying currents (I S and I L ), which were found in all four cell lines to a similar extent. All cells expressed the large current, I L (A), whereas only 30 50% of cells expressed I S (B). According to the voltage dependence of both currents (compare C and D), we applied two pulse protocols: P1 with depolarizations from 80 to 70 millivolts (mv) in 10-mV steps followed by a pulse to 10 mv to record I S (B); P2 with depolarizations from 60 to 120 mv in 10-mV steps followed by a pulse to 90 mv to record I L (A). The holding potential was 80 mv. (A) Currents from a cell expressing only I L and (B) a particularly large I S. (B) At the two most depolarized voltage steps, there is a significant contribution of I L to the outward current. (C) Current-voltage relationship for I S and I L was normalized to the maximum currents at 30 and 120 mv, respectively (n 33 and 11, respectively). (D) Block of I S is shown for seven cells (four TRA, three HIB) in which currents were recorded before and after application of 25 M quinidine. Currents were normalized to the maximum current at 70 mv before the application of quinidine. I L is not affected by quinidine. (E) Current density of I S at 30 mv for merlin-expressing ( : BLA, 23 cells; HIB, 19) and merlin-deficient ( : BAR, 20; TRA, 18) mesothelioma cell lines. the smaller current, I S, cells were held at 80 mv and depolarizing test pulses between 80 and 70 mv were applied, followed by a pulse to 10 mv (P1). For the larger current, I L, cells were equally held at 80 mv and depolarizing test pulses between 60 and 120 mv were applied, followed by a pulse to 90 mv (P2). Figure 2A shows a typical example for a family of whole cell currents recorded with protocol P2 from a mesothelioma cell expressing I L but no I S. Figure 2B shows an example of a cell expressing a relatively large I S recorded with P1. Figure 2C shows the voltage dependence of both currents, demonstrating that they could be well separated. I S was blocked completely upon application of 25 M quinidine in merlin-deficient and merlin-expressing cells, whereas I L was not affected by this drug (Fig. 2D). Recording tail currents at different voltages, the reversal potential of I S was determined to 103 4mV(n 8), which is close to the theoretic potassium equilibrium potential of 101 mv, as calculated by the Nernst equation. I S was thus identified as a delayed rectifier K current with very similar biophysical and pharmacologic properties as those described for Schwann cells. 9,10 Because I L was not sensitive for 25 M quinidine, we did not characterize this current further. A current with very similar biophysical properties was recently characterized in detail in human meningioma cells and identified as a large conductance Ca 2 -activated K current. 24 Because the rate of proliferation was reduced by quinidine exclusively in merlin-deficient schwannoma cells and because a correlation of proliferation with the density of a quinidine-sensitive K current had been shown in Schwann cells, 9,10 we compared the density of I S in cell lines expressing or not expressing merlin. All cell lines were blinded for data evaluation. Of 38 merlin-deficient HMM cells (20 BAR, 18 TRA), 21 expressed a measurable I S (55%) compared with 18 of 47 (38%) merlin-expressing HMM cells (27 BLA, 20 HIB). Figure 2E shows the average densities of I S at 30 mv for all cells investigated (including those without a measurable I S ). There was no significant difference between merlin-expressing and merlin-deficient cells merlin as measured by the Student t test. For I L, there was also no detectable difference between merlin-expressing or merlin-deficient cell lines (results not shown). Both merlin-deficient and merlin-expressing HMM cell lines had identical electrophysiologic properties and were well suited for our study. Studies on Cell Proliferation Cell cultures of all four HMM cell lines were treated with different amounts of quinidine for 4 days before the proliferation rate was determined by immunostaining. They were further evaluated by cell cycle analysis. The proliferation of the merlin-expressing HMM cell lines, BLA and HIB, measured by visualizing the incorporation of BrdU using immunostaining was slightly better and even seemed to be stimulated in the presence of quinidine up to a concentration of 25 M (Fig. 3). In contrast, the proliferation of the merlindeficient HMM cell lines, BAR and TRA, was clearly reduced by quinidine in a dose-dependent manner to a minimum of about 50% of the initial value (Fig. 3). These results were confirmed in four independent experiments. Cell Cycle Analysis Analysis was performed for two of the HMM cell lines, HIB and TRA. The number of cells in S-phase was dramatically reduced to 50% of the initial value when
6 1960 CANCER April 15, 2003 / Volume 97 / Number 8 lines. Thus, no correlation between cytochrome P450 2D6 mutations and NF2 status was found. FIGURE 3. Immunocytochemical analysis of the proliferation of the human malignant mesothelioma (HMM) cell lines in the presence of quinidine. Cells were treated with different amounts of quinidine for 4 days before proliferation was determined. Proliferation rates vary between passages. Therefore, for every experiment, proliferation of cells cultured in the absence of quinidine (dimethylsulfoxide control) was set as 100% for ease of comparison. For each experiment, at least three aliquots ( 500 cells) were counted (n 4). The percentage of bromodeoxyuridine incorporation was determined as means. Columns represent means, error bars indicate the standard error of the mean, and asterisks indicate statistical significance measured by paired t tests of proliferation rates of merlin-expressing versus merlin-deficient HMM cell lines (P 2.5 M 0.003, P 5 M 0.002, P 10 M , P 25 M ). quinidine was added to the merlin-deficient TRA cells but remained almost unchanged for the merlin-expressing HIB cells. The reduction of the cell number in S-phase was correlated with an increase in the number of cells in G 0 /G 1 to almost 200% of the reference, which indicates a G 0 /G 1 arrest as an effect of quinidine (Fig. 4). A cytotoxicity test was performed to exclude the possibility that a reduction in proliferation is due to a toxic effect of quinidine. None of the quinidine concentrations used, ranging from 2.5 to 100 M, revealed any toxicity for the cultures. Sequence Analysis of the CYP450 2D6 Locus To evaluate whether the CYP450 2D6 locus carries any mutations, which could be related to the quinidine effect observed, DNA of all four HMM cell lines was isolated and the nine exons of the CYP450 2D6 locus were amplified. Exonic mutations (three of four) as well as intronic mutations (3 of 10) were found in both of the merlin-deficient and in one of the merlin-expressing HMM cell lines (data not shown). No mutations were found in the second merlin-expressing HMM cell line (BLA). Therefore, we identified mutations in merlin-deficient as well as in one merlin-expressing HMM cell DISCUSSION In this study, we evaluated the effect of quinidine, a broad spectrum channel blocker, on the proliferation of HMM cell lines with a defined NF2 status. We showed that quinidine selectively and significantly reduced proliferation of merlin-deficient HMM cell lines, whereas it had no effect on the proliferation of merlin-expressing HMM cell lines. The detected decrease in proliferation occurs at slightly lower quinidine concentrations in cell cycle analysis by FACS than in immunocytochemical staining of proliferating cells. This is because FACS analysis is more sensitive than immunocytochemical staining. However, as the tendency and the final reduction in proliferation to about 50% of the initial value are almost the same in both experiments, we consider this negligible. Previous studies 9,10 suggested that the reduction in proliferation might be due to the blockage of ion channels by quinidine. Rosenbaum et al. 9 described the inhibition of a noninactivating, outwardly rectifying K current by quinidine in primary cultures of human NF2 schwannoma cells as well as cultures of normal Schwann cells, but the proliferation was reduced only in NF2 schwannoma cells. Besides K channels, quinidine is known to block Na and Ca 2 as well. However, Rosenbaum et al. confined their discussion to K channels as adult human Schwann cells do not express Ca 2 channels and Na channels require a 10-fold higher concentration of quinidine for blocking. 9 The hypothesis proposed by this study suggested that NF2 schwannoma cells respond to quinidine because of their increased K outward current. However, the electrophysiologic studies concerning HMM cell lines showed no significant differences in current density between the merlin-deficient and the merlin-expressing cell lines, although all HMM cell lines expressed a quinidine-sensitive K current with very similar biophysical properties to those described for human Schwann cells. Thus, we conclude that in HMM cell lines, the reduction in cell proliferation by quinidine cannot be attributed to a different expression of potassium channels. The human cytochrome P450 (CYP450) system is encoded by several gene families, and yet more than 20 isoforms have been characterized. A role in the control of growth and differentiation was suggested, 20 as the enzymes catalyze the detoxification of drugs including antiarrhythmics (mono-oxygenase function) 26 and endogenous substrates such as steroids, arachidonates, and nitrosamines. 27,28 Quinidine selectively blocks the CYP450 2D6 isoenzyme (CYP2D6),
7 Quinidine Reduces Proliferation in NF2 -/- HMMs/Utermark et al FIGURE 4. Cell cycle analysis on the proliferation of the human malignant mesothelioma (HMM) cell lines HIB (merlin-deficient) and TRA (merlin-expressing) in the presence of quinidine. Cells were treated with different amounts of quinidine for 4 days before proliferation was determined. As proliferation rates vary between distinct passages, for ease of comparison the proliferation of cells cultured in the absence of quinidine, i.e., the dimethlysulfoxide control was set to 100% in each experiment. For three independent experiments, the number of cells in S-phase (A) and G 0 /G 1 -phase (B) was determined. The percentage of bromodeoxyuridine incorporation was determined as means. Columns represent means, error bars indicate the standard error of the mean, and asterisks indicate statistical significance measured by paired t tests of proliferation rates of merlin-expressing versus merlin-deficient HMM cell lines (P 2.5 M 0.001, P 5 M , P 10 M , P 25 M ). (C) Cell cycle analysis of merlin-deficient TRA cells in the presence of 10 M quinidine. Distinct cell cycle phases are characterized. which has been reported to display reduced activity levels due to specific mutations in the DNA sequence, leading to the PM phenotype. As a consequence of these genetic and metabolic changes, the substrates remain and accumulate in the cytoplasm and may become toxic or be transformed into carcinogens or mutagens. Therefore, the distinct effect of quinidine on the merlin-deficient HMM cell lines may be based on additional mutations in the CYP2D6 locus, causing impaired enzyme function and leading to a PM phenotype for these cells. However, an analysis of the coding sequence of CYP2D6 showed mutations in merlin-deficient as well as merlin-expressing HMM cell lines (data not shown), indicating that there is no link between the quinidine effect and CYP2D6 mutations. Horita et al. 32 described that progesterone is capable of inhibiting proliferation of an HMM cell line (211H) in a dose-dependent manner to about 20% of the initial value at a concentration of 100 M progesterone. The proliferation was measured by the MTS test, but no BrdU incorporation studies or cell cycle analyses were performed. In addition, Horita et al. showed enhanced apoptosis of the progesteronetreated cells, suggesting a regulation of expression of proapoptotic genes by progesterone. Therefore, progesterone has a therapeutic effect for a subset of malignant mesotheliomas, as suggested by the data on the effect of quinidine shown in this study. To summarize, quinidine selectively impairs the proliferation of merlin-deficient HMM cell lines. This effect is not related to the block of K channels as it has been shown for human schwannoma cells. 9,10 No correlation with accompanying cytochrome P450 2D6 mutations was detected. Even though HMM cells might have other genetic differences than expression or deletion of merlin, the common denominator between these cells is the NF2 gene status. Therefore, the effect of quinidine has to be correlated with the loss of merlin expression. Further research is required to elucidate the exact mechanism by which quinidine impairs the proliferation of merlin-deficient cells. In vivo studies, such as xenograft studies, would be helpful to support the significance of the data presented in this study. Quinidine or quinidine analogs are of potential therapeutic interest for merlin-deficient mesotheliomas. REFERENCES 1. Hagemeijer A, Versnel MA, Van Drunen E, et al. Cytogenetic analysis of malignant mesothelioma. Cancer Genet Cytogenet. 1990;47: Taguchi T, Jhanwar SC, Siegfried JM, Keller SM, Testa JR. Recurrent deletions of specific chromosomal sites in 1p, 3p, 6q, and 9p in human malignant mesothelioma. Cancer Res. 1993;53: Bjorkqvist AM, Tammilehto L, Nordling S, et al. Comparison of DNA copy number changes in malignant mesothelioma, adenocarcinoma and large-cell anaplastic carcinoma of the lung. Br J Cancer. 1998;77: Bianchi AB, Mitsunaga SI, Cheng JQ, et al. High frequency of inactivating mutations in the neurofibromatosis type 2 gene (NF2) in primary malignant mesotheliomas (see comments). Proc Natl Acad Sci USA. 1995;92: Sekido Y, Pass HI, Bader S, et al. Neurofibromatosis type 2 (NF2) gene is somatically mutated in mesothelioma but not in lung cancer. Cancer Res. 1995;55: Cheng JQ, Lee WC, Klein MA, Cheng GZ, Jhanwar SC, Testa JR. Frequent mutations of NF2 and allelic loss from chromosome band 22q12 in malignant mesothelioma: evidence for a two-hit mechanism of NF2 inactivation. Genes Chromosomes Cancer. 1999;24: