JOIJRNAL OF CET,T,UI,AR PHYSIOLOGY 142:410-419 (1990)
Menthol Blocks Dihydropyridine-Insensitive Ca2+ Channels and Induces Neurite Outgrowth in Human Neuroblastoma Cells N E I L SIDELL,*, M. A N T H O N Y VERITY, AND E D W A R D P. N O R D D,v,von of Neuropathology and Krain Research Institute iN 5 , M A V ) and Ilepartment of Nephrology i t P N ), the loncson Comprehensive Cancer Center, UCLA School ot Medfcfne, 10s Angeles, C'a/,forn,a 90024
Voltage-gated CaL+ channels were identified in LA-N-5 human neuroblastorna cells using the Ca2+ sensitive fluorescent probe, tura-2. Using a variety of "classical" Caz+ channel blockers, we have demonstrated the presence oC both dihydropyridine (DHP)-sensitive and -insensitive channel types that can be activated by depolarization of the cells with either high K ' or gramicidin in the bathing solution. Brief exposure of LA-N-5 cells to menthol blunted the depolarization-induced CaL+ influx though both DHP-sensitive and DHP-insensitive channels. This effect i s concentration dependent (50% maximal blocking effect with 0.25 m M menthol), rapid in onset, and readily reversible. The specificity of the Ca' +-channel blocking effect of menthol was demonstrated in parallel studies using compounds with similar structures: menthone blocked Ca' ' channels with about half the potency of menthol, while cyclohexanol was without effect. Addition of either menthol or menthone to LA-N-5 cultures induced neurite outgrowth, cellular clustering, and reduction of cell growth in a dose-dependent fashion that correlated with the ability of these compounds to inhibit the DHPinsensitive Ca2+ influx. Cyclohexanol had no biologic activity. Taken together, the parallel potency for blockade of DHP-insensitive Ca2 influx with the biologic activity of menthol suggests a role for certain types of Ca" channels in triggering growth and morphologic changes in LA-N-5 cells. +
Modification of the activity of voltage-gated Ca2+ channels has been shown t o pla a n important role in controlling a number of Ca2+ -dependent cellular events. Of the three Ca'+ channel types identified by electrophysiological techniques (L-, N-, and T-type), the biological function of the L-type channel has been best characterized (McClesky et al., 1986; Miller, 1987 for reviews). In this regard, the identification and characterization of the physiologic role of L-type calcium channels has been facilitated by their sensitivity to pharmacological manipulations through specific binding of phenylalkylamines such as verapamil, and a variety of dihydropyridine (DHP) derivatives. These latter compounds can act both as agonist or antagonist to channel function (Nowycky et al., 1985; Catterall et al., 1988). In contrast, T-type Ca2 channels are relatively insensitive to these agents and their physiologic role remains speculative. Recently, it has been reported that menthol, a cyclic alcohol widely appreciated for its ability to produce a cooling sensation, blocks Ca2+ currents through voltage-activated channels in cultured dorsal root ganglion cells from chick and r a t embryos (Swandulla et al., 1987). This compound was found to be unique in its pharmacologic action in that it both accelerated inactivation of L-type channels and selectively blocked Ttype channels. In the present study, voltage-gated Ca2+ channels that are both sensitive and insensitive +
"-: 1990 WILEY-LISS.
INC.
to verapamil and dihydropyridines and that are reversibly blocked by menthol are identified in human neuroblastoma cells. Furthermore, we show that treatment of neuroblastoma cells in culture with menthol results in morphologic differentiation and inhibition of proliferation which correlate with the ability of this compound to block the dihydropyridine-insensitive Ca2+ current. These findings confirm and extend the Ca-channel blocking activity of menthol to neuronal cells of human origin and demonstrate a new biologic effect of this agent.
MATERIALS AND METHODS Materials and cell cultures Fura-2 AM was obtained from Molecular Probes (Eugene, OR). Ionomycin was purchased from Calbiochem (La Jolla, CAI. Gramicidin, verapamil, nicardipine, ni-
Received February 28, 1989; accepted October 25, 1989.
"To whom reprint requestdcorrespondence should be addressed at Dept. of Pathology, 18-170 Center far Health Sciences, UCLA School of Medicine, Los Angeles, CA 90024. Edward P. Nord's present address is Division of Nephrology, Dept. of Medicine, School of Medicine, SUNY a t Stony Brook, Stony Brook, NY.
MENTHOL BLOCKS Ca CHANNELS
41 1
modipine, and (- )menthol (2-isopropyl-5-methyl-cyclo- lets (10-150 pM, Tsien et al., 1985). Preparatory exhexanol) were obtained from Sigma Chemical Go. periments including fluorescence microscopy estab(St. Louis, MO) and (Flmenthone, thymol, eugenol, lished that this loading procedure yields excellent and cyclohexanol were obtained from Aldrich Chemi- homogeneous distribution of probe in the cytoplasm cal Co. (Milwaukee, WI). Omega-conotoxin GVIA with no apparent compartmentalization. (wCTx), was obtained from Bachem Inc. (Torrance, Changes in ICa2+1, were monitored on a PerkinCA). Ethyleneglycol-O,O'-bis(2-aminoethyl)-N,N,N'N~Elmer LS-5 fluorometer (Perkin-Elmer, Norwalk, CT) tetracetic acid (EGTA) was purchased from Fluka equipped with a thermostatically controlled chamber Chemical Gorp (Ronkonkoma, NY). All other chemi- (37"C), using excitation wavelengths 340/360 nm, cals used were obtained from Sigma and were of the emission wavelength 510 nm, with slit widths set a t 5 highest commercial grade available. nm. Details of the experimental methods have been Studies were performed on LA-N-5 human neuro- described elsewhere (Aboolian and Nord, 1988). blastoma cells originally obtained from the laboratory Briefly, to initiate an experiment a n aliquot of furaof Dr. Robert Seeger, Dept. of Pediatrics, UCLA School 2-loaded cells (approximately lo6 cells) was pelleted in of Medicine. The cells were cultured in 75 cm2 tissue a Beckman microfuge B (Beckman Instruments, Palo culture flasks (Falcon Plastics, Oxnard, CA) with Alto, CA) for 10 sec, rinsed twice to remove all extraRPMI 1640 containing 10% heat-inactivated fetal calf cellular dye, and introduced into a cuvette to which serum, 50 IU penicilliniml, 50 pg streptomyciniml, and assay buffer had previously been added. The composi1 pg fungizone/ml (Flow Labs, McLean, VA). Cells tion of the Ca2 ' replete assay solution was identical to were grown and subcultured a s previously described that of the loading solution detailed earlier. Where in(Sidell, 1982). In experiments where the effects of dicated a Ca2 ' -free medium was employed, which was (-)menthol, (-)menthone, and cyclohexanol on cell devoid of added Ca2', and buffered with 0.5 mM growth and morphology were tested, the compounds EGTA. The final osmolarity of all solutions was 300 were added to the growth medium from stock solutions mOsm. The cell suspension was continually stirred us(in DMSO) so that the final solvent concentration was ing a spectrophotometer cuvette stirring system (Spec0.1% (volivol). Control experiments contained 0.1% tracell, Oreland, PA). Despite this stirring system, (vol/vol) DMSO alone. This concentration of DMSO did large clumps of cells had a tendency to settle within the not affect the plating efficiency, growth rate, or cell first 3 min after being added t o the cuvette. This was evident by a gradual decline in emission intensity with morphology of LA-N-5 cells. both 340 and 360 nm excitation wavelengths which Measurement of cytosolic leveled off after this initial time period. Thus, in most free Ca2+ concentration cases, readings were begun approximately 3 min after Changes in cytosolic free Ca2+ concentration the cells had been added to the cuvette. Except where [Ca2'],, were f luorometrically monitored using the indicated, test compounds were added to the cuvette in Ca2+-sensitivedye fura-2 (Grynkiewicz et al., 1985) a s a 1:1,000 dilution. Preparatory experiments estabpreviously described by one of us (Aboolian and Nord, lished that compounds and vehicles, by themselves, did 1988). To load cells with the fluorescent probe, mono- not alter the nature of the fluorescent reading. layers were dislodged by briskly tapping the culture Fluorescence intensity was continually monitored at flask with the palm of the hand. For most experiments a Ca" ' -sensitive wavelength (340 nm) and intermita single 75% confluent 75 cm2 flask was used; how- tent readings obtained at a Ca2+-insensitive waveever, in some experiments the contents of two such length (360 nm), thereby allowing the determination of flasks were pooled. The cell suspension was centri- ratio 340/360 (Aboolian and Nord, 1988). Ratio 3401360 fuged at 200g for 5 min and the pellet washed and was therefore not dependent upon cell number nor inre-suspended in a final volume of 1 ml of a solution ter-experimental discrepancies in intracellular dye composed of (in mM) 140 NaC1, 3 KC1, 1 CaCl,, 1 concentration. [Ca2 ' 1, was derived using the formula MgCI2, 10 glucose, and 10 Hepes, brought to pH 7.4 described by Grynkiewicz and coworkers (1985). R,,, with NaOH a t 37°C (standard solution). Fura-2-AM (in and R,,,, were determined a s previously described DMSO) was added to the cell suspension in a 1 : l O O (Aboolian and Nord, 1988). Figures presented are data dilution to give a final dye concentration of 10 pM. from typical representative experiments. Each experiLoading proceeded at 37°C for 30 min. To determine ment was performed on at least four separate cell susintracellular fura-2 concentration, a n aliquot of loaded pensions, each obtained from a different passage of cells (5 x lo6 cells) was washed (3 X ) and added to a cells. Where indicated, experimental error was reknown volume of de-ionized water. Membrane frag- ported as the standard error of the mean (+-SEM). ments were pelleted and the fluorescence value of the r3H]thymidineincorporation supernatant obtained a t 360 nm excitation, 510 nm emission wavelengths. The 360 nm excitation waveL'HIThymidine incorporation of treated and control length is Ca2+-independentand is dependent on fura-2 LA-N-5 cells, a s a n index of DNA synthesis, was perconcentration (Grynkiewicz et al., 1985). The value ob- formed a s previously described (Sidell e t al., 1981, tained was compared to known standards derived from 1983). Briefly, cells (10' cells/well) were plated in quache K+-salt of fura-2 diluted in a n identical volume of druplicate wells of flat-bottomed microtest plates de-ionized water. The intracellular dye concentration (M.A. Bioproducts). One day after plating, medium was calculated a t -50 pM based on an estimated cell containing the compound to be tested or vehicle was volume of 1 pl. This fura-2 value is within the range added to each well. Four days later, the plates were determined for smooth muscle cells isolated from toad pulsed with 1.0 pCi of [methyl-3H~thymidine/well (6.7 stomach (30 pM, Becker e t al., 1989) or human plate- CiimM) (New England Nuclear, Claremont, CA). After
412
SIDELL ET AL.
B
A
.-
+
K
N
1
H 1rnin
Verapamil
Nicardipine
Nimodipine
H 1min Fig. 1. Effects of KCI and gramicidin on [Caz ' II in LA-N-5 cells. Cells, loaded with fura-2 as described in Materials and Methods, were pelleted, washed ( 2 x I , and suspended in either Ca2+-replete (standard solution) (A,€%) or a Ca"-free solution (C) of identical composition but without Ca2+ and to which EGTA (0.5 mM) had been added. Where indicated KCI (K), gramicidin (G),or ionomycin (I) were added to the cuvettc from a stock solution in 1 : l O O or 1:1,OOO dilution (ionophores t , respectively. Final concentration of additives was KC1 50 mM, gramicidin 1 pM, and ionomycin 1 pM.
Fig. 2. Inhibition by Ca2 ' channel blockers of the high K ' -induced [Caz ' 1, rise. Fura-2-loaded cells were incubated in the presence of 10 p.M verapamil, 10 IJ.Mnicardipine, or 5 p M nimodipine (as indicated) for 3 min prior to challenge with 50 mM KC1. Verapamil, nicardipine, and nimodipine were added t o the cuvette from stock solutions (in DMSOt at 1:1,000 dilutions. A control experiment is included for comparison.
413
MENTHOL BLOCKS Ca CHANNELS
a n additional 18 hours of incubation at 37"C, trichloroacetic acid (TCAI-insoluble material was harvested for scintillation counting by conventional methods (Sidell et al., 1981).
RESULTS
Determination of voltage-sensitive Ca"' channels Steady-state [Ca"], of LA-N-5 neuroblastoma cells Determination of acetylcholinesterase in the Ca2 ' -replete solution was determined to be 90 k (AChE) activity Specific AChE activity was measured as a biochem- 2 nM in = 163). To test for the presence of voltage-senical index of the relative state of differentiation of sitive Ca2 ' channels in their plasma membrane, the treated and control LA-N-5 cells (Sidell et al., 1984). cells were depolarized by two independent modaliFor measurement of AChE activity, cells were grown ties-namely, challenge of fura-2-loaded cells with eiKC1 or the N a + , K + ionophore gramicidin. Reprein 75 cm2 tissue culture flasks for 4 days in the absence ther or presence of the test reagents, washed twice with sentative results from such experiments are illustrated in Figure 1. In the Ca-replete medium, rapid exposure isotonic saline, and harvested by vigorous shaking of of fura-2-loaded cells to either 50 mM KC1 (Fig. 1A) or the culture flask. After removal of saline, cells were 1 pM gramicidin (Fig. 1B) resulted in a brisk and imfrozen a t -2O"C, thawed by the addition of ice-cold 10 mediate increment in [Ca2+lito approximately 2.5mM sodium phosphate buffer (pH 7.4) containing 0.5% Triton X-100 (1.5 m1/106cells), and sonicated for 10 sec. 3.5-fold the resting value. ICa2+li subsequently deSamples of the homogenate were assayed for AChE creased and attained a new steady-state level within 2-3 min which was greater than baseline [Ca"tli and activity as previously described (Sidell et al., 1984). maintained through 6 min. Tracings were not Protein concentrations were determined with a Bio- was monitored beyond this time period. Note that quantiRad Coomassie protein assay kit using bovine serum tative differences were observed between the two methalbumin a s the standard. Student's t-test was used to ods of depolarization. First, the rise in [Ca"Ii induced assess the statistical significance of differences be- by high K + was more rapid in onset than that induced tween specific AChE activity in control and treated by gramicidin since with g-ramieidin a finite time pecultures. riod is required for insertion of the pore-forming ionophore into the membrane (Rink et al., 1980). Second, the magnitude of the Ca2+ increment was greater with 1 p,M gramicidin (269 nM) than with 50 mM KC1 (195 VER nM) as would be predicted withgreater depolarization. The mean fold increment in LCa +Ii on exposure of cells t o KC1 was 2.5 i 0.2 ( n = 2 0 ) versus 3.0 ? 0.3 ( n = 8 ) when challenged with gramicidin. When the identical maneuvers were executed on cells suspended in a Cafree medium (Fig. lC), no increment in LCa"tli was observed. To verify the functional integrity of the plasma membrane in Ca2 -free conditions, the Ca2 ionophore ionomycin (1 pM) was introduced into the cuvette (Fig. 1C). A prompt and large increment in lCa"1, was observed which declined to below the iniFig. 3. Blockade of open voltage-gated Ca2' channels. Fura- tial resting ICa2+li.Note also that resting [Ca2 ' Ii in 2-loaded cells were depolarized using 50 mM KCI. As indicated nimo+
+
dipine (5 p M ) INIM) and verapamil(10 pM) (VER) were added to the cuvette during the sustained elevation of Lea'+ J, following depolarization.
B
A
C
w CTX
n H 1min Fig. 4. Inhibition by o-conotoxin of' the [Ca"], peak induced by high K'. Fura-%loaded cells were exposed to 2 I*.MwCtx after challenge with 50 mM KC1 (A) or 1 min before challenge (B).In C and L), cells were continuously exposed to 5 pM nimodipine in the cuvette for > 5 min before KCI depolarization in the absence (C) or presence (D) of 2 pLM wCTx.
D
414
SIDELL ET AL.
A
M
.L
60
A
longed increase in Ca2+permeability induced by membrane depolarization. In subsequent experiments, 50 mM KC1 was used for investigating the nature of the induced [Ca2 1, rise. +
H Imin
Fig. 6. Menthol inhibition of the dihydropyridine-insensitivecomponent of the K+-induced[Ca"'], rise. LA-N-5 cells were exposed to 5 KM nimodipine (saturating concentration) during the fura-2 loading procedure and throughout all subsequent steps. Cells were depolarized using 50 mM KCl in the absence (A) or presence (B) of menthol (0.5 mM, 2 min) in the cuvette.
the Ca"-free medium was markedly lower than in the Ca-replete medium (49 nM versus 97 nM, respectively). The nature and kinetics of the Ca2+ efflux mechanism in LA-N-5 cells was not further evaluated. Taken together, these data demonstrate that LA-N-5 cells possess voltage-sensitive Ca2+ channels in their plasma membrane. The increment in LCa2+],on depolarization is totally dependent upon entry of Ca2 ' into the cell from the extracellular compartment. The protracted time period that [CaZ' ], remained elevated above pre-depolarization baseline level suggests a pro-
Inhibition of the voltage-sensitive Ca2+influx by "classical"Ca2+channel blockers The nature of the KCl-induced Ca2+ influx into LAN-5 cells was further characterized by testing the effect of various known Ca2+ channel blockers on [Ca2+1,. Brief (3 min) exposure of cells to verapamil (10 pM), nicardipine (10 pM), or nimodipine (5 pM) blunted the maximal increment in 1Ca2+l,induced by 50 mM KC1 by 50-70% and resulted in a return to baseline values within 2 min (Fig. 2). This effect was qualitatively and quantitatively similar for the three drugs tested. The mean value obtained was 59 i. 4%blockade (n = 9 with three replicates for each of the three drugs). Exposure of cells to either of the three compounds for longer time periods (>30 min) yielded identical results (data not shown). Thus, although these Ca" channel blockers inhibited the initial rise in [Ca2 ' I,, a DHP- and phenylalkylamine-insensitive component to the [Ca2+I, increment exists. As illustrated in Figure 3, attainment of a new elevated steady-state [Ca"], level after depolarization by KC1 was rapidly reversed upon addition of nimodipine or verapamil to the cuvette. These obser+
415
MENTHOL BLOCKS Ca CHANNELS
D
C
B
A 250
f
60 K
H 1rnin
Fig. 7. Complete block of‘ the dihydropyridine-insensitive lCa2 1, increment by menthol plus wCTx. LA-N-5 cells were exposed to 5 p M nimodipine during the fura-2 loading procedure and throughout all subsequent steps. Cells were depolarized using 50 pM KCI (K) in the absence of additional channel blockers (A),or in the presence of saturating concentrations of wCTx 12 pM)(B), menthol 10.5 mM1 (C), or wCTx (2 pM)plus menthol (0.5 mMi (D).
100
A-A
7
B
A
C
9)
VI
0
a m
e
H lmin
X
E ’c
ap
0.05
1.oo
0.10 Blocker (mM)
Fig. 8. Concentration-dependent inhibition by menthol !*l, ment,hone (oi, and cyclohexanol ( & of the DHP-insensitive component of the [Ca’+Ji increment induced by high K’ . LA-N-5 cells were exposed to 5 pM nimodipine during the fura-2-loadingprocedure and throughout all subsequent steps. Results are expressed as % of the maximum K+-induced ICa”], rise of nimodipine-treated cells in the absence of the test compound. In all cases, the test compound was added to the cuvette in a 1:1,000 dilution 2 min prior to depolarization with the KCI (50 mM final concentration).
vations further support the existence of voltage-sensitive Ca2 ’ channels in the plasma membrane of LA-N-5 cells. Indeed, the prolonged elevation of lCa2+l, after depolarization, and the rapid reversal of this “plateau” phase of the response by verapamil and the dihydropyridine derivative, suggests the presence of classical “long-lasting” or L-type Ca’+ channels (McClesky et al., 1986; Miller, 1987). It has been shown that wCTx, a 27 amino acid peptide from the venom of the marine snail Conus geopraphus, blocks specific types of calcium channels depending upon the target tissue. Figure 4 shows that in LA-N-5 cells, brief exposure (1min) to wCTx preferentially blocked the initial [Ca2-1, peak induced by K + depolarization but had little effect on the plateau phase of the [Ca2 ‘I, rise. This blocking effect of wCTx on the Ca2 ’ peak was found to be saturating at concentrations greater than 0.1 pM. When
Fig. 9. Reversibility of the menthol blockade of’the DHY-insensitive [Ca2’ 1, rise induced by high K + . Ki-induced ICa’- Ii rise in LA-N-5 cells treated with 6 FM nimodipine in the absence (A) and presence (B)of menthol !0.5 mM?2 min).In C cells were treated with both 5 p M nimodipine and 0.5 mM menthol during the fura-2-loadingprocedure, hut only the nimodipine was included in the assay buffer in the cuvette during the [Ca” 1, measurements.
wCTx and nimodi ine were added together before the high K’, the [Ca ‘1, rise was substantially reduced from that seen with nimodipine alone (Fig. 4C,D). Thus the channels blocked by wCTx include a subset that are DHP insensitive. Similar findings of wCTx’s spectrum of action on a different neuroblastoma cell line have recently been reported (Sher et al., 1988). Blockade of Ca2+ channels by menthol It has recently been demonstrated that menthol blocks Ca2+ channels in dorsal root ganglion cells (Swandulla et al., 1987). To test whether menthol influences Ca2+ channel activity in LA-N-5 cells, the K + -induced [Ca”], increment was monitored in cells preincubated for 2 rnin in 0.5 mM menthol. As shown in Figure 5, menthol preincubation did not affect resting [Ca2+1,; however, the depolarization-induced [Ca2+],rise was markedly attenuated. The mean value in four ex eriments was 78 2 6%,blockade of the maximal [Ca’ ‘1, rise induced by 50 mM KC1. Moreover, menthol added to depolarized cells induced a rapid decline of the “plateau” phase of [Ca2+], (Fig. 5C), as observed with verapamil and the DHP Ca“’ channel blockers (Fig. 3). To test whether menthol attenuated the DHP-insen-
T
P “
416
SIDELL ET AL.
sitive component of the K ' -induced [Ca"Ii increment, cells were exposed to 5 pM nimodipine during the fura2-loading procedure and throughout all subsequent steps. This concentration of nimodipine was found to be saturating for its ability to block the K -induced [Ca2+Iirise. Figure 6 demonstrates that exposure of nimodipine-pretreated cells to 0.5 mM menthol blunted the nimodipine-insensitive K -induced [Ca2 ' Ii increment by 80% (83 t 4% in 9 experiments) in a manner similar t o that observed with wCTx. With either of these compounds, a minor residual nimodipine-insensitive component usually remained which was completely eliminated when saturating concentrations of menthol (0.5mM) and wCTx (2 pM) were added together (Fig. 7). Thus, menthol's target spectrum of action ap ears to include a subset of nimodipine-insensitive Ca+ ' channels that are also resistant t o blockade by wCTx. The inhibition of the DHP-insensitive component of the [Ca"'], rise by menthol was studied in further detail. The dependence of the K+-induced TCa2'1, increment on menthol concentration is depicted in Figure 8. A clear inhibitory effect was observed a t menthol concentrations >O. 1mM with maximum inhibition occurring a t 0.5 mM. This menthol-induced inhibitory effect was extremely rapid in onset and was observed after only 15 sec of preincubation, the shortest time period required to technically add menthol to the cuvette and be assured of adequate mixing (data not shown). Blockade by menthol was completely reversible; when 0.5 mM menthol was present during the 30 min fura-2loading period, but not added to the cuvette, the K ' induced [Ca2+lirise was essentially the same as that observed in cells never exposed to menthol (Fig. 9). To evaluate the specificity of the menthol effect on the DHP-insensitive [Ca2+Ii rise, identical experiments were conducted using cyclohexanol, a highly lipid soluble alcohol from which menthol is derived, and menthone, where the hydroxy group of menthol is replaced by a keto group. As illustrated in Figure 8, cyclohexanol failed to inhibit the K ' -induced [Ca' ' Ii rise even a t millimolar concentrations. On the other hand, menthone inhibited this rise in a concentration-dependent manner, but with dose-response characteristics approximately one-half as effective as menthol. Thymol and eugenol, two other compounds with structures similar to menthol, inhibited the K+-induced [Ca2'Ii rise but were also less effective than menthol (data not shown).
ent as early as 24 hours. In parallel experiments, menthone increased cellular clustering and neurite outgrowth in a manner qualitatively similar to that seen with menthol but required a concentration of approximately 1mM to produce maximal effects. On the other hand, addition of cyclohexanol t o the culture medium a t concentrations as high a s 4 mM had no effect on cell morphology. Similarly, no effect was seen with wCTx and nimodipine, either alone or together, a t concentrations that maximally inhibited the K'-induced LCa2' 1, rise (2 pM, wCTx; 5 pM, nimodipine). The evaluation of the growth modulating effects of menthol was difficult to quantify from numbers of viable cells due to the extensive cellular clustering that occurred in the presence of this agent (see Fig. 101. Therefore, the effects of menthol on growth rate was assessed by monitoring incorporation of r3Hlthymidine. Figure 11 shows the concentration-dependent effects of menthol on incorporation of' ['Hjthymidine into LA-N-5 cells after 4 days of treatment. It is evident that 80% inhibition was achieved a t the highest menthol concentration (1mM) tested, while no effect was seen below 0.25 mM menthol. Figure 11 also shows that menthone caused inhibition of L3H1thymidine incorporation into LA-N-5 cells, but less effectively than menthol. Cyclohexanol was without effect.
Effects of menthol on cell g r o w t h and morphology The effect of menthol on cell growth and morphology was tested by addition of the compound to the cultural medium as described in Materials and Methods. As shown in Figure 10, culturing LA-N-5 cells with menthol caused a n increase in cellular clustering and neurite outgrowth. The menthol concentration that caused maximal increase in the formation of neurites was 0.5 mM, while concentrations less than 0.1 mM produced no noticeable morphologic effects (data not shown). No decrease in the percentage of viable cells was detected in the treated cultures as compared to that of control cultures, although higher menthol concentrations (>1 mM) were found to be toxic over the extended culture period. The time required for maximal increase in the formation of neurites was 3-4 days with effects appar-
This report identifies voltage-gated Ca2 ' channels in LA-N-5 human neuroblastoma cells and demonstrates blockade of these channels by menthol. Both DHP-sensitive and DHP-insensitive Ca2 ' channels are blocked, and the effect is concentration dependent, rapid in onset, and readily reversible. These latter findings may suggest a menthol binding site on the outside of the cell membrane. Interestingly, patch-clump studies have similarly inferred a n external binding site for menthol's action on L- and T-type Ca2' channels in chick dorsal root ganglion cells (Swandulla et al., 1987). We cannot, at this time, make definitive conclusions a s to the specific channel types which may be affected in LA-N-5 cells since insensitivity to dihydropyridines are not limited to T-type channels in some systems (Yaari et al., 1987). Menthol, a n alcohol derivative, could possibly act by inducing general mem-
+
Acetylcholinesterase (AChE) activity It has previously been demonstrated that, concomitant with neurite outgrowth, AChE activity increased in LA-N-5 cells induced to differentiate by treatment with retinoic acid (RA) (Side11 et al., 1984). To assess whether this biochemical index of neuroblastoma differentiation was also associated with menthol-induced neurite outgrowth, AChE activity was measured in cells treated with various concentrations of menthol for 4 days (the time when a maximal increase in neurite formation was observed). The results in Table 1 show that the AChE activities in menthol-treated cells ranged from 1.3 t o 1.6 nmiminimg protein. These values are not significantly different from a n average AChE activity of 1.2 nmiminimg protein measured in untreated cells. As positive controls, LA-N-5 cells were treated with 1 pM RA over the same time period and, as reported, showed almost a twofold increase in AChE activity as compared to untreated cells.
DISCUSSION
Fig. 10. Effects of menthol on the morphology uf LA-N-5 cells. Solvent-treated controls (A); cells cultured in the presence of 0.5mM menthol for 4 days (B) showing cellular clustering and increased neurite outgrowth. x 200.
418
SlUELL ET AL T
TABLE 1. Effect of menthol on specific AChE activity in LA-N-5 cells
T
Treatment’ Control Menthol (mM) .5 .25 .1P5 RA tl uM1
AChE2 (nmole/min:mg protein) 1.2 -+ 0.2 1.5 -t 0.3 I .6 ? 0.4 1.3 I 0.3 2.1 ? 0.2”
‘Four day treatment period as described in Materials and Methods. *Values are means t SEM of‘four independent experiments each done
111
dupli-
CSte.
0
‘Significantly different (P*:0.051 from vehicle control.
4/
1 .o
0.5
2.0
Blocker (mM)
due to some, as yet, unknown action that is not causally connected to Caz ’ influx. Resolution of these questions will require detailed voltage clamp studies to delineate the Ca2+channel types present in LA-N-5 cells and their functional response to menthol, wCTx, and dihydropyridines, a s well as determination of the stability of these compounds in long term culture. The pattern of morphologic differentiation (i.e., neubrane labilization or by modifying membrane fluidity due to its lipophilic properties (Grisham and Barnett, rite outgrowth and cellular clustering) exhibited by 1973; Haydon et al., 1977). However, cyclohexanol, the menthol-treated LA-N-5 cells was not a s elaborate a s highly lipid soluble cyclic alcohol from which menthol that typically observed when these cells are induced to is derived, failed to show any effects on Ca2+ influx differentiate with RA (Sidell et al., 1983; Robson and even at millimolar concentrations. On the other hand, Sidell, 1985). Furthermore, while RA induces a n inmenthone, in which the hydroxy group of menthol is crease in AChE (Sidell et al., 1984 and Table 11, this replaced by a keto group, was about half as effective as activity was not significantly affected by menthol. menthol in blocking the K’-induced Ca2+ influx. Two These results might be best interpreted if differentiaother menthol derivatives, thymol and eugenol, were tion of LA-N-5 cells is viewed a s occurring through multiple steps that are each characterized by specific similarly less effective. LA-N-5 cells grown in the presence of menthol in the events. Thus, a limited degree of neurite extension and culture medium (at the concentrations shown to a moderately decreased growth rate might reflect early acutely inhibit voltage-gated Ca2 influx) exhibited events in differentiation that are induced by menthol, a n increase in cellular clustering, neurite outgrowth, while the cessation of cellular growth, increased AChE and a reduction in growth rate as reflected by reduced activity, and the extensive degree of neurite outgrowth thymidine incorporation. A similar correlation held seen with RA-treated cells are characteristic of more with methone, while cyclohexanol had no effect on LA- advanced stages of differentiation. Interestingly, RA N-5 cell-growth or morphology a t the concentrations has been shown to block T-type (DHP-insensitive) tested. The parallel potency for blockade of K ’ -induced Ca2+ channels in a mouse B cell hybridoma a t concenCa2 ’ influx by these agents with their ability to affect trations which correlate with its ability to retard the proliferation of the cells (Bosma and Sidell, 1988). It is LA-N-5 cells biologically suggests that functional Ca2 channels play a role in determining the phenotypic ex- currently not known whether RA blocks similar chanpression of these cells. However, a t the present time, nels in LA-N-5 cells; however, in the context of the this possibility must be interpreted with caution in the present work, i t is tempting to speculate that funcabsence of any demonstrated effects of nimodipine and tional Ca2- channels might also play a role in the moroCTx (which also blocked nimodipine-insensitive phologic response of these cells to RA or other inducing [Ca2+],rises) on LA-N-5 growth and morphology. Pre- agents. In support of this concept, earlier studies have vious electrophysiologic studies have documented that shown that Ca2 ’ channel modulators can also affect oCTx can block Ca2’ channels of the L- and N-type, neurite outgrowth from normal (nontransformed) neubut spares the T-type (McCleskey et al., 19871, while rons (Mattson and Kater, 1987) and have demonstrated menthol can affect L- and T-type but has no action on strong correlations between the [Ca2 1, levels in these N-type (Swandulla et al., 1987). Our results showing cells and their growth status (Connor, 1986; Cohan et that menthol blocks nimodipine-insensitive Ca2+-in- al., 1987). Together with such reports, the present work flux that is resistant to wCTx (Fig. 7) suggests that the further supports the hypothesis that voltage-gated spectrum of action of these agents are also different in Ca2 channels can influence the morphological develLA-N-5 cells. Taken together, it appears that the bio- opment of certain types of nerve cells. logic action of menthol cannot be simply attributed to ACKNOWLEDGMENTS reduction of total Ca2 influx through DHP-sensitive and/or -insensitive channels but may depend upon inThis work was supported by National Institutes of volvement of a specific channel type that is not affected Health grants CA 43503, CA 30515, DK 36351, and by the other “classical” Ca2+ channel blockers tested. DK 41585. The authors wish to thank Bieshia Chang Alternatively, the biologic effects of menthol may be and Guangyang Han for excellent technical assistance Fig. 11. Dose-dependent effects of menthol (01, menthone (21, and cyclohexanol (A) on the incorporation o f L3Hlthymidine into LA-N-5 cells after 4 days o f treatment. Symhok represent the averages + SEkI of quadruplicate cultures from a representative experiment. The control value in the absence of blacker is shown by .
+
+
+
+
MENTHOL BLOCKS Ca CHANNELS
and to Ms. Han for her help in preparation of the figures.
LITERATURE CITED Ahoolian, A., and Nord, E.P. (1988) Bradykinin increases cytosolic free [Ca'+ I, in proximal tubule cells. Am. cJ, Physiol. (Renal Fluid Electrolyte Physiol. 1, 24tF486-F493. Becker, P.L., Singer, J.J., Walsh, J.V. Jr., and Fay, F.S.(1989) Regulation of calcium concentration in voltage-clamped smooth muscle cells. Science, 244t211-214. Bosrna, M.. and Sidell. N. (1988) Retinoic acid inhibits CaZLcurrents and cell proliferation in a B-lymphocyte cell line. J. Cell. Physiol., 135:317-323. Catterall, W.A., Seagar, M.J.; and Takahashi, M. (19x8) Molecular properties of dihydropyridine-sensitivecalcium channels in skeletal musclc. J. J3iol. Chem., 263.3535-3536. Cohan, C.S., Connor, J.A., and Kater, S.B.(1987) Electrically and chemically mediated increases in intracellular calcium in neurunal growth cones. J. Neurosci., 7t3588-3599. Connor, J . A . (1986) Digital imaging of free calcium changes and of spatial gradients in growing processes in single, mammalian central nervous system cells. R o c . Natl. Acad. Sci. USA, 83:61796183. Grisham, C.M., and Barnett, R.E. (19731 The effect of long chain alcohols on membrane lioids and the (Na- + K'bA'Wase. Biochin~.-Biophys.Acta, 31 1 .'417-422. Grynkiewicz, G., Poene, M., and Tsien, R.Y. (1985)A new generation of Ca2+indicators with greatly improved fluorescence . properties. J. . Biol. Chem., 260t3440-3450.Haydon, D.A., Hendry, B.M., Levison, S.R., and Requena, J. (1977) Anesthesia by the n-alkanes. A comparative study of nerve impulse blockage and the properties of black lipid hilayer membrane. Biochim. Biophys. Acta, 470t17-34. Mattson, M.P., and Kater, S.B. (1987) Calcium regulation of neurite elongation and growth cone motility. J. Neurosci., 7:4034-4043. McClesky, E.W., Pox, A.P., Feldman, D., and Tsien, R.W. (1986) Different types of calcium channels. J . Exp. Biol., 124t177-190. McClesky, E.W., Fox, A.P., Feldman, D., Cruz, L.J., Olivera? B.M., ~~~
~
~
419
I'sien, R.W., and Yoshikami, D. (1987) Omega-Conotoxin: Direct and persistent blockade of specific types of calcium channels in neurons but not musclc. Proc. Natl. Acad. Sci. USA. 84:4327-4331. Miller, R.J. (1987) Multiple calcium channels and neuronal function. Science, 235t46-52. Nowvcky, M.C.. Fox. A.P.. and Tsien. R.W. (1985) Three tvDes o f neuronal calcium channels' with different calcium agonisi"&nsitivity. Nature, 316:440-443. Rink, T.J., Montecucco, C., Hesketh, T.R., and Tsien, R.Y. (19801 Lymphocyte membrane potential assessed with fluorescent probes. Biochim. Biophys. Acta, ,5Y5:15-30. Rohson, J.A., and Sidell, N. 1198.5) Ultrastructural features of a human neuroblastoma cell line treated with retinoic acid. Neuroscience, 24: 1149-1 162. Sher, E., Pandiella, A,, and Clementi, F. (1988). Omega-Conotoxin binding and effects on calcium channel function in human neuroblastoma and rat pheochrornocytoma cell lines. FEBS Lett., 235: 178-182. Sidell, N. (19821Retinoic acid-induced growth inhibition and morphologic differentiation o f human neuroblastoma cells in vitro. JNCI 68:589-596. Sidell, N., Altman, A,, Haussler, M.R., and Seeger, K.C. (1983)Effects of retinoic acid (KA) on the growth and phenotypic expression of several human neurohlastoma cell lines. Exp. Cell Res., 148.21-30. Sidell, N., Pamatiga. E., and Golub, S.H. (1981) Augmentation of human thymocyte proliferative responses by retinoic acid. Exp. Cell Biol., 49t239-245. Sidell, N., Lucas, C.A., and Kreutzherg, G.W. (1984) Regulation of acetylcholinesterase activity by retinoic acid in a human neurohlastoma cell line. Exp. Cell Res., 155:305-309. Swandulla; D., Carhone, E., Schafer, K., and Lux, H.D. (1987) Effect of menthol on two types of Ca currents in cultured sensory neurons of vertebrates. Pflugers Arch., 409:52-59. Tsien, R.Y., Rink, T.J., and Poenie, M. (1985). Measurement of cytosolic free Ca2 ' in individual small cells using fluorescence microscopy with dual excitation wavelengths. Cell Calcium, 6t145-157. Yaari, Y., IIamon, B., and Lux, H.D. (1987) Development of two types of calcium channels in cultured mammalian hippocampal neurons. Science, 235580-682.
Menthol Blocks Dihydropyridine-Insensitive Ca2+ Channels and Induces Neurite Outgrowth in Human Neuroblastoma Cells N E I L SIDELL,*, M. A N T H O N Y VERITY, AND E D W A R D P. N O R D D,v,von of Neuropathology and Krain Research Institute iN 5 , M A V ) and Ilepartment of Nephrology i t P N ), the loncson Comprehensive Cancer Center, UCLA School ot Medfcfne, 10s Angeles, C'a/,forn,a 90024
Voltage-gated CaL+ channels were identified in LA-N-5 human neuroblastorna cells using the Ca2+ sensitive fluorescent probe, tura-2. Using a variety of "classical" Caz+ channel blockers, we have demonstrated the presence oC both dihydropyridine (DHP)-sensitive and -insensitive channel types that can be activated by depolarization of the cells with either high K ' or gramicidin in the bathing solution. Brief exposure of LA-N-5 cells to menthol blunted the depolarization-induced CaL+ influx though both DHP-sensitive and DHP-insensitive channels. This effect i s concentration dependent (50% maximal blocking effect with 0.25 m M menthol), rapid in onset, and readily reversible. The specificity of the Ca' +-channel blocking effect of menthol was demonstrated in parallel studies using compounds with similar structures: menthone blocked Ca' ' channels with about half the potency of menthol, while cyclohexanol was without effect. Addition of either menthol or menthone to LA-N-5 cultures induced neurite outgrowth, cellular clustering, and reduction of cell growth in a dose-dependent fashion that correlated with the ability of these compounds to inhibit the DHPinsensitive Ca2+ influx. Cyclohexanol had no biologic activity. Taken together, the parallel potency for blockade of DHP-insensitive Ca2 influx with the biologic activity of menthol suggests a role for certain types of Ca" channels in triggering growth and morphologic changes in LA-N-5 cells. +
Modification of the activity of voltage-gated Ca2+ channels has been shown t o pla a n important role in controlling a number of Ca2+ -dependent cellular events. Of the three Ca'+ channel types identified by electrophysiological techniques (L-, N-, and T-type), the biological function of the L-type channel has been best characterized (McClesky et al., 1986; Miller, 1987 for reviews). In this regard, the identification and characterization of the physiologic role of L-type calcium channels has been facilitated by their sensitivity to pharmacological manipulations through specific binding of phenylalkylamines such as verapamil, and a variety of dihydropyridine (DHP) derivatives. These latter compounds can act both as agonist or antagonist to channel function (Nowycky et al., 1985; Catterall et al., 1988). In contrast, T-type Ca2 channels are relatively insensitive to these agents and their physiologic role remains speculative. Recently, it has been reported that menthol, a cyclic alcohol widely appreciated for its ability to produce a cooling sensation, blocks Ca2+ currents through voltage-activated channels in cultured dorsal root ganglion cells from chick and r a t embryos (Swandulla et al., 1987). This compound was found to be unique in its pharmacologic action in that it both accelerated inactivation of L-type channels and selectively blocked Ttype channels. In the present study, voltage-gated Ca2+ channels that are both sensitive and insensitive +
"-: 1990 WILEY-LISS.
INC.
to verapamil and dihydropyridines and that are reversibly blocked by menthol are identified in human neuroblastoma cells. Furthermore, we show that treatment of neuroblastoma cells in culture with menthol results in morphologic differentiation and inhibition of proliferation which correlate with the ability of this compound to block the dihydropyridine-insensitive Ca2+ current. These findings confirm and extend the Ca-channel blocking activity of menthol to neuronal cells of human origin and demonstrate a new biologic effect of this agent.
MATERIALS AND METHODS Materials and cell cultures Fura-2 AM was obtained from Molecular Probes (Eugene, OR). Ionomycin was purchased from Calbiochem (La Jolla, CAI. Gramicidin, verapamil, nicardipine, ni-
Received February 28, 1989; accepted October 25, 1989.
"To whom reprint requestdcorrespondence should be addressed at Dept. of Pathology, 18-170 Center far Health Sciences, UCLA School of Medicine, Los Angeles, CA 90024. Edward P. Nord's present address is Division of Nephrology, Dept. of Medicine, School of Medicine, SUNY a t Stony Brook, Stony Brook, NY.
MENTHOL BLOCKS Ca CHANNELS
41 1
modipine, and (- )menthol (2-isopropyl-5-methyl-cyclo- lets (10-150 pM, Tsien et al., 1985). Preparatory exhexanol) were obtained from Sigma Chemical Go. periments including fluorescence microscopy estab(St. Louis, MO) and (Flmenthone, thymol, eugenol, lished that this loading procedure yields excellent and cyclohexanol were obtained from Aldrich Chemi- homogeneous distribution of probe in the cytoplasm cal Co. (Milwaukee, WI). Omega-conotoxin GVIA with no apparent compartmentalization. (wCTx), was obtained from Bachem Inc. (Torrance, Changes in ICa2+1, were monitored on a PerkinCA). Ethyleneglycol-O,O'-bis(2-aminoethyl)-N,N,N'N~Elmer LS-5 fluorometer (Perkin-Elmer, Norwalk, CT) tetracetic acid (EGTA) was purchased from Fluka equipped with a thermostatically controlled chamber Chemical Gorp (Ronkonkoma, NY). All other chemi- (37"C), using excitation wavelengths 340/360 nm, cals used were obtained from Sigma and were of the emission wavelength 510 nm, with slit widths set a t 5 highest commercial grade available. nm. Details of the experimental methods have been Studies were performed on LA-N-5 human neuro- described elsewhere (Aboolian and Nord, 1988). blastoma cells originally obtained from the laboratory Briefly, to initiate an experiment a n aliquot of furaof Dr. Robert Seeger, Dept. of Pediatrics, UCLA School 2-loaded cells (approximately lo6 cells) was pelleted in of Medicine. The cells were cultured in 75 cm2 tissue a Beckman microfuge B (Beckman Instruments, Palo culture flasks (Falcon Plastics, Oxnard, CA) with Alto, CA) for 10 sec, rinsed twice to remove all extraRPMI 1640 containing 10% heat-inactivated fetal calf cellular dye, and introduced into a cuvette to which serum, 50 IU penicilliniml, 50 pg streptomyciniml, and assay buffer had previously been added. The composi1 pg fungizone/ml (Flow Labs, McLean, VA). Cells tion of the Ca2 ' replete assay solution was identical to were grown and subcultured a s previously described that of the loading solution detailed earlier. Where in(Sidell, 1982). In experiments where the effects of dicated a Ca2 ' -free medium was employed, which was (-)menthol, (-)menthone, and cyclohexanol on cell devoid of added Ca2', and buffered with 0.5 mM growth and morphology were tested, the compounds EGTA. The final osmolarity of all solutions was 300 were added to the growth medium from stock solutions mOsm. The cell suspension was continually stirred us(in DMSO) so that the final solvent concentration was ing a spectrophotometer cuvette stirring system (Spec0.1% (volivol). Control experiments contained 0.1% tracell, Oreland, PA). Despite this stirring system, (vol/vol) DMSO alone. This concentration of DMSO did large clumps of cells had a tendency to settle within the not affect the plating efficiency, growth rate, or cell first 3 min after being added t o the cuvette. This was evident by a gradual decline in emission intensity with morphology of LA-N-5 cells. both 340 and 360 nm excitation wavelengths which Measurement of cytosolic leveled off after this initial time period. Thus, in most free Ca2+ concentration cases, readings were begun approximately 3 min after Changes in cytosolic free Ca2+ concentration the cells had been added to the cuvette. Except where [Ca2'],, were f luorometrically monitored using the indicated, test compounds were added to the cuvette in Ca2+-sensitivedye fura-2 (Grynkiewicz et al., 1985) a s a 1:1,000 dilution. Preparatory experiments estabpreviously described by one of us (Aboolian and Nord, lished that compounds and vehicles, by themselves, did 1988). To load cells with the fluorescent probe, mono- not alter the nature of the fluorescent reading. layers were dislodged by briskly tapping the culture Fluorescence intensity was continually monitored at flask with the palm of the hand. For most experiments a Ca" ' -sensitive wavelength (340 nm) and intermita single 75% confluent 75 cm2 flask was used; how- tent readings obtained at a Ca2+-insensitive waveever, in some experiments the contents of two such length (360 nm), thereby allowing the determination of flasks were pooled. The cell suspension was centri- ratio 340/360 (Aboolian and Nord, 1988). Ratio 3401360 fuged at 200g for 5 min and the pellet washed and was therefore not dependent upon cell number nor inre-suspended in a final volume of 1 ml of a solution ter-experimental discrepancies in intracellular dye composed of (in mM) 140 NaC1, 3 KC1, 1 CaCl,, 1 concentration. [Ca2 ' 1, was derived using the formula MgCI2, 10 glucose, and 10 Hepes, brought to pH 7.4 described by Grynkiewicz and coworkers (1985). R,,, with NaOH a t 37°C (standard solution). Fura-2-AM (in and R,,,, were determined a s previously described DMSO) was added to the cell suspension in a 1 : l O O (Aboolian and Nord, 1988). Figures presented are data dilution to give a final dye concentration of 10 pM. from typical representative experiments. Each experiLoading proceeded at 37°C for 30 min. To determine ment was performed on at least four separate cell susintracellular fura-2 concentration, a n aliquot of loaded pensions, each obtained from a different passage of cells (5 x lo6 cells) was washed (3 X ) and added to a cells. Where indicated, experimental error was reknown volume of de-ionized water. Membrane frag- ported as the standard error of the mean (+-SEM). ments were pelleted and the fluorescence value of the r3H]thymidineincorporation supernatant obtained a t 360 nm excitation, 510 nm emission wavelengths. The 360 nm excitation waveL'HIThymidine incorporation of treated and control length is Ca2+-independentand is dependent on fura-2 LA-N-5 cells, a s a n index of DNA synthesis, was perconcentration (Grynkiewicz et al., 1985). The value ob- formed a s previously described (Sidell e t al., 1981, tained was compared to known standards derived from 1983). Briefly, cells (10' cells/well) were plated in quache K+-salt of fura-2 diluted in a n identical volume of druplicate wells of flat-bottomed microtest plates de-ionized water. The intracellular dye concentration (M.A. Bioproducts). One day after plating, medium was calculated a t -50 pM based on an estimated cell containing the compound to be tested or vehicle was volume of 1 pl. This fura-2 value is within the range added to each well. Four days later, the plates were determined for smooth muscle cells isolated from toad pulsed with 1.0 pCi of [methyl-3H~thymidine/well (6.7 stomach (30 pM, Becker e t al., 1989) or human plate- CiimM) (New England Nuclear, Claremont, CA). After
412
SIDELL ET AL.
B
A
.-
+
K
N
1
H 1rnin
Verapamil
Nicardipine
Nimodipine
H 1min Fig. 1. Effects of KCI and gramicidin on [Caz ' II in LA-N-5 cells. Cells, loaded with fura-2 as described in Materials and Methods, were pelleted, washed ( 2 x I , and suspended in either Ca2+-replete (standard solution) (A,€%) or a Ca"-free solution (C) of identical composition but without Ca2+ and to which EGTA (0.5 mM) had been added. Where indicated KCI (K), gramicidin (G),or ionomycin (I) were added to the cuvettc from a stock solution in 1 : l O O or 1:1,OOO dilution (ionophores t , respectively. Final concentration of additives was KC1 50 mM, gramicidin 1 pM, and ionomycin 1 pM.
Fig. 2. Inhibition by Ca2 ' channel blockers of the high K ' -induced [Caz ' 1, rise. Fura-2-loaded cells were incubated in the presence of 10 p.M verapamil, 10 IJ.Mnicardipine, or 5 p M nimodipine (as indicated) for 3 min prior to challenge with 50 mM KC1. Verapamil, nicardipine, and nimodipine were added t o the cuvette from stock solutions (in DMSOt at 1:1,000 dilutions. A control experiment is included for comparison.
413
MENTHOL BLOCKS Ca CHANNELS
a n additional 18 hours of incubation at 37"C, trichloroacetic acid (TCAI-insoluble material was harvested for scintillation counting by conventional methods (Sidell et al., 1981).
RESULTS
Determination of voltage-sensitive Ca"' channels Steady-state [Ca"], of LA-N-5 neuroblastoma cells Determination of acetylcholinesterase in the Ca2 ' -replete solution was determined to be 90 k (AChE) activity Specific AChE activity was measured as a biochem- 2 nM in = 163). To test for the presence of voltage-senical index of the relative state of differentiation of sitive Ca2 ' channels in their plasma membrane, the treated and control LA-N-5 cells (Sidell et al., 1984). cells were depolarized by two independent modaliFor measurement of AChE activity, cells were grown ties-namely, challenge of fura-2-loaded cells with eiKC1 or the N a + , K + ionophore gramicidin. Reprein 75 cm2 tissue culture flasks for 4 days in the absence ther or presence of the test reagents, washed twice with sentative results from such experiments are illustrated in Figure 1. In the Ca-replete medium, rapid exposure isotonic saline, and harvested by vigorous shaking of of fura-2-loaded cells to either 50 mM KC1 (Fig. 1A) or the culture flask. After removal of saline, cells were 1 pM gramicidin (Fig. 1B) resulted in a brisk and imfrozen a t -2O"C, thawed by the addition of ice-cold 10 mediate increment in [Ca2+lito approximately 2.5mM sodium phosphate buffer (pH 7.4) containing 0.5% Triton X-100 (1.5 m1/106cells), and sonicated for 10 sec. 3.5-fold the resting value. ICa2+li subsequently deSamples of the homogenate were assayed for AChE creased and attained a new steady-state level within 2-3 min which was greater than baseline [Ca"tli and activity as previously described (Sidell et al., 1984). maintained through 6 min. Tracings were not Protein concentrations were determined with a Bio- was monitored beyond this time period. Note that quantiRad Coomassie protein assay kit using bovine serum tative differences were observed between the two methalbumin a s the standard. Student's t-test was used to ods of depolarization. First, the rise in [Ca"Ii induced assess the statistical significance of differences be- by high K + was more rapid in onset than that induced tween specific AChE activity in control and treated by gramicidin since with g-ramieidin a finite time pecultures. riod is required for insertion of the pore-forming ionophore into the membrane (Rink et al., 1980). Second, the magnitude of the Ca2+ increment was greater with 1 p,M gramicidin (269 nM) than with 50 mM KC1 (195 VER nM) as would be predicted withgreater depolarization. The mean fold increment in LCa +Ii on exposure of cells t o KC1 was 2.5 i 0.2 ( n = 2 0 ) versus 3.0 ? 0.3 ( n = 8 ) when challenged with gramicidin. When the identical maneuvers were executed on cells suspended in a Cafree medium (Fig. lC), no increment in LCa"tli was observed. To verify the functional integrity of the plasma membrane in Ca2 -free conditions, the Ca2 ionophore ionomycin (1 pM) was introduced into the cuvette (Fig. 1C). A prompt and large increment in lCa"1, was observed which declined to below the iniFig. 3. Blockade of open voltage-gated Ca2' channels. Fura- tial resting ICa2+li.Note also that resting [Ca2 ' Ii in 2-loaded cells were depolarized using 50 mM KCI. As indicated nimo+
+
dipine (5 p M ) INIM) and verapamil(10 pM) (VER) were added to the cuvette during the sustained elevation of Lea'+ J, following depolarization.
B
A
C
w CTX
n H 1min Fig. 4. Inhibition by o-conotoxin of' the [Ca"], peak induced by high K'. Fura-%loaded cells were exposed to 2 I*.MwCtx after challenge with 50 mM KC1 (A) or 1 min before challenge (B).In C and L), cells were continuously exposed to 5 pM nimodipine in the cuvette for > 5 min before KCI depolarization in the absence (C) or presence (D) of 2 pLM wCTx.
D
414
SIDELL ET AL.
A
M
.L
60
A
longed increase in Ca2+permeability induced by membrane depolarization. In subsequent experiments, 50 mM KC1 was used for investigating the nature of the induced [Ca2 1, rise. +
H Imin
Fig. 6. Menthol inhibition of the dihydropyridine-insensitivecomponent of the K+-induced[Ca"'], rise. LA-N-5 cells were exposed to 5 KM nimodipine (saturating concentration) during the fura-2 loading procedure and throughout all subsequent steps. Cells were depolarized using 50 mM KCl in the absence (A) or presence (B) of menthol (0.5 mM, 2 min) in the cuvette.
the Ca"-free medium was markedly lower than in the Ca-replete medium (49 nM versus 97 nM, respectively). The nature and kinetics of the Ca2+ efflux mechanism in LA-N-5 cells was not further evaluated. Taken together, these data demonstrate that LA-N-5 cells possess voltage-sensitive Ca2+ channels in their plasma membrane. The increment in LCa2+],on depolarization is totally dependent upon entry of Ca2 ' into the cell from the extracellular compartment. The protracted time period that [CaZ' ], remained elevated above pre-depolarization baseline level suggests a pro-
Inhibition of the voltage-sensitive Ca2+influx by "classical"Ca2+channel blockers The nature of the KCl-induced Ca2+ influx into LAN-5 cells was further characterized by testing the effect of various known Ca2+ channel blockers on [Ca2+1,. Brief (3 min) exposure of cells to verapamil (10 pM), nicardipine (10 pM), or nimodipine (5 pM) blunted the maximal increment in 1Ca2+l,induced by 50 mM KC1 by 50-70% and resulted in a return to baseline values within 2 min (Fig. 2). This effect was qualitatively and quantitatively similar for the three drugs tested. The mean value obtained was 59 i. 4%blockade (n = 9 with three replicates for each of the three drugs). Exposure of cells to either of the three compounds for longer time periods (>30 min) yielded identical results (data not shown). Thus, although these Ca" channel blockers inhibited the initial rise in [Ca2 ' I,, a DHP- and phenylalkylamine-insensitive component to the [Ca2+I, increment exists. As illustrated in Figure 3, attainment of a new elevated steady-state [Ca"], level after depolarization by KC1 was rapidly reversed upon addition of nimodipine or verapamil to the cuvette. These obser+
415
MENTHOL BLOCKS Ca CHANNELS
D
C
B
A 250
f
60 K
H 1rnin
Fig. 7. Complete block of‘ the dihydropyridine-insensitive lCa2 1, increment by menthol plus wCTx. LA-N-5 cells were exposed to 5 p M nimodipine during the fura-2 loading procedure and throughout all subsequent steps. Cells were depolarized using 50 pM KCI (K) in the absence of additional channel blockers (A),or in the presence of saturating concentrations of wCTx 12 pM)(B), menthol 10.5 mM1 (C), or wCTx (2 pM)plus menthol (0.5 mMi (D).
100
A-A
7
B
A
C
9)
VI
0
a m
e
H lmin
X
E ’c
ap
0.05
1.oo
0.10 Blocker (mM)
Fig. 8. Concentration-dependent inhibition by menthol !*l, ment,hone (oi, and cyclohexanol ( & of the DHP-insensitive component of the [Ca’+Ji increment induced by high K’ . LA-N-5 cells were exposed to 5 pM nimodipine during the fura-2-loadingprocedure and throughout all subsequent steps. Results are expressed as % of the maximum K+-induced ICa”], rise of nimodipine-treated cells in the absence of the test compound. In all cases, the test compound was added to the cuvette in a 1:1,000 dilution 2 min prior to depolarization with the KCI (50 mM final concentration).
vations further support the existence of voltage-sensitive Ca2 ’ channels in the plasma membrane of LA-N-5 cells. Indeed, the prolonged elevation of lCa2+l, after depolarization, and the rapid reversal of this “plateau” phase of the response by verapamil and the dihydropyridine derivative, suggests the presence of classical “long-lasting” or L-type Ca’+ channels (McClesky et al., 1986; Miller, 1987). It has been shown that wCTx, a 27 amino acid peptide from the venom of the marine snail Conus geopraphus, blocks specific types of calcium channels depending upon the target tissue. Figure 4 shows that in LA-N-5 cells, brief exposure (1min) to wCTx preferentially blocked the initial [Ca2-1, peak induced by K + depolarization but had little effect on the plateau phase of the [Ca2 ‘I, rise. This blocking effect of wCTx on the Ca2 ’ peak was found to be saturating at concentrations greater than 0.1 pM. When
Fig. 9. Reversibility of the menthol blockade of’the DHY-insensitive [Ca2’ 1, rise induced by high K + . Ki-induced ICa’- Ii rise in LA-N-5 cells treated with 6 FM nimodipine in the absence (A) and presence (B)of menthol !0.5 mM?2 min).In C cells were treated with both 5 p M nimodipine and 0.5 mM menthol during the fura-2-loadingprocedure, hut only the nimodipine was included in the assay buffer in the cuvette during the [Ca” 1, measurements.
wCTx and nimodi ine were added together before the high K’, the [Ca ‘1, rise was substantially reduced from that seen with nimodipine alone (Fig. 4C,D). Thus the channels blocked by wCTx include a subset that are DHP insensitive. Similar findings of wCTx’s spectrum of action on a different neuroblastoma cell line have recently been reported (Sher et al., 1988). Blockade of Ca2+ channels by menthol It has recently been demonstrated that menthol blocks Ca2+ channels in dorsal root ganglion cells (Swandulla et al., 1987). To test whether menthol influences Ca2+ channel activity in LA-N-5 cells, the K + -induced [Ca”], increment was monitored in cells preincubated for 2 rnin in 0.5 mM menthol. As shown in Figure 5, menthol preincubation did not affect resting [Ca2+1,; however, the depolarization-induced [Ca2+],rise was markedly attenuated. The mean value in four ex eriments was 78 2 6%,blockade of the maximal [Ca’ ‘1, rise induced by 50 mM KC1. Moreover, menthol added to depolarized cells induced a rapid decline of the “plateau” phase of [Ca2+], (Fig. 5C), as observed with verapamil and the DHP Ca“’ channel blockers (Fig. 3). To test whether menthol attenuated the DHP-insen-
T
P “
416
SIDELL ET AL.
sitive component of the K ' -induced [Ca"Ii increment, cells were exposed to 5 pM nimodipine during the fura2-loading procedure and throughout all subsequent steps. This concentration of nimodipine was found to be saturating for its ability to block the K -induced [Ca2+Iirise. Figure 6 demonstrates that exposure of nimodipine-pretreated cells to 0.5 mM menthol blunted the nimodipine-insensitive K -induced [Ca2 ' Ii increment by 80% (83 t 4% in 9 experiments) in a manner similar t o that observed with wCTx. With either of these compounds, a minor residual nimodipine-insensitive component usually remained which was completely eliminated when saturating concentrations of menthol (0.5mM) and wCTx (2 pM) were added together (Fig. 7). Thus, menthol's target spectrum of action ap ears to include a subset of nimodipine-insensitive Ca+ ' channels that are also resistant t o blockade by wCTx. The inhibition of the DHP-insensitive component of the [Ca"'], rise by menthol was studied in further detail. The dependence of the K+-induced TCa2'1, increment on menthol concentration is depicted in Figure 8. A clear inhibitory effect was observed a t menthol concentrations >O. 1mM with maximum inhibition occurring a t 0.5 mM. This menthol-induced inhibitory effect was extremely rapid in onset and was observed after only 15 sec of preincubation, the shortest time period required to technically add menthol to the cuvette and be assured of adequate mixing (data not shown). Blockade by menthol was completely reversible; when 0.5 mM menthol was present during the 30 min fura-2loading period, but not added to the cuvette, the K ' induced [Ca2+lirise was essentially the same as that observed in cells never exposed to menthol (Fig. 9). To evaluate the specificity of the menthol effect on the DHP-insensitive [Ca2+Ii rise, identical experiments were conducted using cyclohexanol, a highly lipid soluble alcohol from which menthol is derived, and menthone, where the hydroxy group of menthol is replaced by a keto group. As illustrated in Figure 8, cyclohexanol failed to inhibit the K ' -induced [Ca' ' Ii rise even a t millimolar concentrations. On the other hand, menthone inhibited this rise in a concentration-dependent manner, but with dose-response characteristics approximately one-half as effective as menthol. Thymol and eugenol, two other compounds with structures similar to menthol, inhibited the K+-induced [Ca2'Ii rise but were also less effective than menthol (data not shown).
ent as early as 24 hours. In parallel experiments, menthone increased cellular clustering and neurite outgrowth in a manner qualitatively similar to that seen with menthol but required a concentration of approximately 1mM to produce maximal effects. On the other hand, addition of cyclohexanol t o the culture medium a t concentrations as high a s 4 mM had no effect on cell morphology. Similarly, no effect was seen with wCTx and nimodipine, either alone or together, a t concentrations that maximally inhibited the K'-induced LCa2' 1, rise (2 pM, wCTx; 5 pM, nimodipine). The evaluation of the growth modulating effects of menthol was difficult to quantify from numbers of viable cells due to the extensive cellular clustering that occurred in the presence of this agent (see Fig. 101. Therefore, the effects of menthol on growth rate was assessed by monitoring incorporation of r3Hlthymidine. Figure 11 shows the concentration-dependent effects of menthol on incorporation of' ['Hjthymidine into LA-N-5 cells after 4 days of treatment. It is evident that 80% inhibition was achieved a t the highest menthol concentration (1mM) tested, while no effect was seen below 0.25 mM menthol. Figure 11 also shows that menthone caused inhibition of L3H1thymidine incorporation into LA-N-5 cells, but less effectively than menthol. Cyclohexanol was without effect.
Effects of menthol on cell g r o w t h and morphology The effect of menthol on cell growth and morphology was tested by addition of the compound to the cultural medium as described in Materials and Methods. As shown in Figure 10, culturing LA-N-5 cells with menthol caused a n increase in cellular clustering and neurite outgrowth. The menthol concentration that caused maximal increase in the formation of neurites was 0.5 mM, while concentrations less than 0.1 mM produced no noticeable morphologic effects (data not shown). No decrease in the percentage of viable cells was detected in the treated cultures as compared to that of control cultures, although higher menthol concentrations (>1 mM) were found to be toxic over the extended culture period. The time required for maximal increase in the formation of neurites was 3-4 days with effects appar-
This report identifies voltage-gated Ca2 ' channels in LA-N-5 human neuroblastoma cells and demonstrates blockade of these channels by menthol. Both DHP-sensitive and DHP-insensitive Ca2 ' channels are blocked, and the effect is concentration dependent, rapid in onset, and readily reversible. These latter findings may suggest a menthol binding site on the outside of the cell membrane. Interestingly, patch-clump studies have similarly inferred a n external binding site for menthol's action on L- and T-type Ca2' channels in chick dorsal root ganglion cells (Swandulla et al., 1987). We cannot, at this time, make definitive conclusions a s to the specific channel types which may be affected in LA-N-5 cells since insensitivity to dihydropyridines are not limited to T-type channels in some systems (Yaari et al., 1987). Menthol, a n alcohol derivative, could possibly act by inducing general mem-
+
Acetylcholinesterase (AChE) activity It has previously been demonstrated that, concomitant with neurite outgrowth, AChE activity increased in LA-N-5 cells induced to differentiate by treatment with retinoic acid (RA) (Side11 et al., 1984). To assess whether this biochemical index of neuroblastoma differentiation was also associated with menthol-induced neurite outgrowth, AChE activity was measured in cells treated with various concentrations of menthol for 4 days (the time when a maximal increase in neurite formation was observed). The results in Table 1 show that the AChE activities in menthol-treated cells ranged from 1.3 t o 1.6 nmiminimg protein. These values are not significantly different from a n average AChE activity of 1.2 nmiminimg protein measured in untreated cells. As positive controls, LA-N-5 cells were treated with 1 pM RA over the same time period and, as reported, showed almost a twofold increase in AChE activity as compared to untreated cells.
DISCUSSION
Fig. 10. Effects of menthol on the morphology uf LA-N-5 cells. Solvent-treated controls (A); cells cultured in the presence of 0.5mM menthol for 4 days (B) showing cellular clustering and increased neurite outgrowth. x 200.
418
SlUELL ET AL T
TABLE 1. Effect of menthol on specific AChE activity in LA-N-5 cells
T
Treatment’ Control Menthol (mM) .5 .25 .1P5 RA tl uM1
AChE2 (nmole/min:mg protein) 1.2 -+ 0.2 1.5 -t 0.3 I .6 ? 0.4 1.3 I 0.3 2.1 ? 0.2”
‘Four day treatment period as described in Materials and Methods. *Values are means t SEM of‘four independent experiments each done
111
dupli-
CSte.
0
‘Significantly different (P*:0.051 from vehicle control.
4/
1 .o
0.5
2.0
Blocker (mM)
due to some, as yet, unknown action that is not causally connected to Caz ’ influx. Resolution of these questions will require detailed voltage clamp studies to delineate the Ca2+channel types present in LA-N-5 cells and their functional response to menthol, wCTx, and dihydropyridines, a s well as determination of the stability of these compounds in long term culture. The pattern of morphologic differentiation (i.e., neubrane labilization or by modifying membrane fluidity due to its lipophilic properties (Grisham and Barnett, rite outgrowth and cellular clustering) exhibited by 1973; Haydon et al., 1977). However, cyclohexanol, the menthol-treated LA-N-5 cells was not a s elaborate a s highly lipid soluble cyclic alcohol from which menthol that typically observed when these cells are induced to is derived, failed to show any effects on Ca2+ influx differentiate with RA (Sidell et al., 1983; Robson and even at millimolar concentrations. On the other hand, Sidell, 1985). Furthermore, while RA induces a n inmenthone, in which the hydroxy group of menthol is crease in AChE (Sidell et al., 1984 and Table 11, this replaced by a keto group, was about half as effective as activity was not significantly affected by menthol. menthol in blocking the K’-induced Ca2+ influx. Two These results might be best interpreted if differentiaother menthol derivatives, thymol and eugenol, were tion of LA-N-5 cells is viewed a s occurring through multiple steps that are each characterized by specific similarly less effective. LA-N-5 cells grown in the presence of menthol in the events. Thus, a limited degree of neurite extension and culture medium (at the concentrations shown to a moderately decreased growth rate might reflect early acutely inhibit voltage-gated Ca2 influx) exhibited events in differentiation that are induced by menthol, a n increase in cellular clustering, neurite outgrowth, while the cessation of cellular growth, increased AChE and a reduction in growth rate as reflected by reduced activity, and the extensive degree of neurite outgrowth thymidine incorporation. A similar correlation held seen with RA-treated cells are characteristic of more with methone, while cyclohexanol had no effect on LA- advanced stages of differentiation. Interestingly, RA N-5 cell-growth or morphology a t the concentrations has been shown to block T-type (DHP-insensitive) tested. The parallel potency for blockade of K ’ -induced Ca2+ channels in a mouse B cell hybridoma a t concenCa2 ’ influx by these agents with their ability to affect trations which correlate with its ability to retard the proliferation of the cells (Bosma and Sidell, 1988). It is LA-N-5 cells biologically suggests that functional Ca2 channels play a role in determining the phenotypic ex- currently not known whether RA blocks similar chanpression of these cells. However, a t the present time, nels in LA-N-5 cells; however, in the context of the this possibility must be interpreted with caution in the present work, i t is tempting to speculate that funcabsence of any demonstrated effects of nimodipine and tional Ca2- channels might also play a role in the moroCTx (which also blocked nimodipine-insensitive phologic response of these cells to RA or other inducing [Ca2+],rises) on LA-N-5 growth and morphology. Pre- agents. In support of this concept, earlier studies have vious electrophysiologic studies have documented that shown that Ca2 ’ channel modulators can also affect oCTx can block Ca2’ channels of the L- and N-type, neurite outgrowth from normal (nontransformed) neubut spares the T-type (McCleskey et al., 19871, while rons (Mattson and Kater, 1987) and have demonstrated menthol can affect L- and T-type but has no action on strong correlations between the [Ca2 1, levels in these N-type (Swandulla et al., 1987). Our results showing cells and their growth status (Connor, 1986; Cohan et that menthol blocks nimodipine-insensitive Ca2+-in- al., 1987). Together with such reports, the present work flux that is resistant to wCTx (Fig. 7) suggests that the further supports the hypothesis that voltage-gated spectrum of action of these agents are also different in Ca2 channels can influence the morphological develLA-N-5 cells. Taken together, it appears that the bio- opment of certain types of nerve cells. logic action of menthol cannot be simply attributed to ACKNOWLEDGMENTS reduction of total Ca2 influx through DHP-sensitive and/or -insensitive channels but may depend upon inThis work was supported by National Institutes of volvement of a specific channel type that is not affected Health grants CA 43503, CA 30515, DK 36351, and by the other “classical” Ca2+ channel blockers tested. DK 41585. The authors wish to thank Bieshia Chang Alternatively, the biologic effects of menthol may be and Guangyang Han for excellent technical assistance Fig. 11. Dose-dependent effects of menthol (01, menthone (21, and cyclohexanol (A) on the incorporation o f L3Hlthymidine into LA-N-5 cells after 4 days o f treatment. Symhok represent the averages + SEkI of quadruplicate cultures from a representative experiment. The control value in the absence of blacker is shown by .
+
+
+
+
MENTHOL BLOCKS Ca CHANNELS
and to Ms. Han for her help in preparation of the figures.
LITERATURE CITED Ahoolian, A., and Nord, E.P. (1988) Bradykinin increases cytosolic free [Ca'+ I, in proximal tubule cells. Am. cJ, Physiol. (Renal Fluid Electrolyte Physiol. 1, 24tF486-F493. Becker, P.L., Singer, J.J., Walsh, J.V. Jr., and Fay, F.S.(1989) Regulation of calcium concentration in voltage-clamped smooth muscle cells. Science, 244t211-214. Bosrna, M.. and Sidell. N. (1988) Retinoic acid inhibits CaZLcurrents and cell proliferation in a B-lymphocyte cell line. J. Cell. Physiol., 135:317-323. Catterall, W.A., Seagar, M.J.; and Takahashi, M. (19x8) Molecular properties of dihydropyridine-sensitivecalcium channels in skeletal musclc. J. J3iol. Chem., 263.3535-3536. Cohan, C.S., Connor, J.A., and Kater, S.B.(1987) Electrically and chemically mediated increases in intracellular calcium in neurunal growth cones. J. Neurosci., 7t3588-3599. Connor, J . A . (1986) Digital imaging of free calcium changes and of spatial gradients in growing processes in single, mammalian central nervous system cells. R o c . Natl. Acad. Sci. USA, 83:61796183. Grisham, C.M., and Barnett, R.E. (19731 The effect of long chain alcohols on membrane lioids and the (Na- + K'bA'Wase. Biochin~.-Biophys.Acta, 31 1 .'417-422. Grynkiewicz, G., Poene, M., and Tsien, R.Y. (1985)A new generation of Ca2+indicators with greatly improved fluorescence . properties. J. . Biol. Chem., 260t3440-3450.Haydon, D.A., Hendry, B.M., Levison, S.R., and Requena, J. (1977) Anesthesia by the n-alkanes. A comparative study of nerve impulse blockage and the properties of black lipid hilayer membrane. Biochim. Biophys. Acta, 470t17-34. Mattson, M.P., and Kater, S.B. (1987) Calcium regulation of neurite elongation and growth cone motility. J. Neurosci., 7:4034-4043. McClesky, E.W., Pox, A.P., Feldman, D., and Tsien, R.W. (1986) Different types of calcium channels. J . Exp. Biol., 124t177-190. McClesky, E.W., Fox, A.P., Feldman, D., Cruz, L.J., Olivera? B.M., ~~~
~
~
419
I'sien, R.W., and Yoshikami, D. (1987) Omega-Conotoxin: Direct and persistent blockade of specific types of calcium channels in neurons but not musclc. Proc. Natl. Acad. Sci. USA. 84:4327-4331. Miller, R.J. (1987) Multiple calcium channels and neuronal function. Science, 235t46-52. Nowvcky, M.C.. Fox. A.P.. and Tsien. R.W. (1985) Three tvDes o f neuronal calcium channels' with different calcium agonisi"&nsitivity. Nature, 316:440-443. Rink, T.J., Montecucco, C., Hesketh, T.R., and Tsien, R.Y. (19801 Lymphocyte membrane potential assessed with fluorescent probes. Biochim. Biophys. Acta, ,5Y5:15-30. Rohson, J.A., and Sidell, N. 1198.5) Ultrastructural features of a human neuroblastoma cell line treated with retinoic acid. Neuroscience, 24: 1149-1 162. Sher, E., Pandiella, A,, and Clementi, F. (1988). Omega-Conotoxin binding and effects on calcium channel function in human neuroblastoma and rat pheochrornocytoma cell lines. FEBS Lett., 235: 178-182. Sidell, N. (19821Retinoic acid-induced growth inhibition and morphologic differentiation o f human neuroblastoma cells in vitro. JNCI 68:589-596. Sidell, N., Altman, A,, Haussler, M.R., and Seeger, K.C. (1983)Effects of retinoic acid (KA) on the growth and phenotypic expression of several human neurohlastoma cell lines. Exp. Cell Res., 148.21-30. Sidell, N., Pamatiga. E., and Golub, S.H. (1981) Augmentation of human thymocyte proliferative responses by retinoic acid. Exp. Cell Biol., 49t239-245. Sidell, N., Lucas, C.A., and Kreutzherg, G.W. (1984) Regulation of acetylcholinesterase activity by retinoic acid in a human neurohlastoma cell line. Exp. Cell Res., 155:305-309. Swandulla; D., Carhone, E., Schafer, K., and Lux, H.D. (1987) Effect of menthol on two types of Ca currents in cultured sensory neurons of vertebrates. Pflugers Arch., 409:52-59. Tsien, R.Y., Rink, T.J., and Poenie, M. (1985). Measurement of cytosolic free Ca2 ' in individual small cells using fluorescence microscopy with dual excitation wavelengths. Cell Calcium, 6t145-157. Yaari, Y., IIamon, B., and Lux, H.D. (1987) Development of two types of calcium channels in cultured mammalian hippocampal neurons. Science, 235580-682.
