EXPERIMENTAL
CELL RESEARCH
189,64-68
(1990)
The Influence of Ethanol on Cell Membrane Fluidity, Migration, and Invasion of Murine Melanoma Cells SIMONE SILBERMAN,**~ Departments
of *Pathology,
TERENCE W. MCGARVEY,~ $Biochemktry,
and tdnatomy,
should be addressed at Department of of Chicago, 2160 South Fist Ave., May-
$3.00
PERSKY? Illinois
60153
AND
METHODS
Materials. B16FlO and K1735 murine melanoma cells were a gift from I. Fidler (University of Texas and M. D. Anderson Hospital, Houston). The metastatic capability of the B16FlO and K1735 cell lines have been described by Fidler [7] and Kripke [S], respectively. Cells were maintained at 37’C, 5% CO*, as preconfluent monolayers in Eagle’s Minimal essential medium (MEM; GIBCO Laboratories, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS, GIBCO) and penicillin-streptomycin sulfate (GIBCO). Ethyl alcohol (ZOO proof, Quantum Chemical Corp., Tucsola, IL) was diluted (v/v) in MEM to yield concentrations of 35.4 mA4 (0.5%), 170.8 mM (l.O%), and 256.2 mM (1.5%). CeU viability and adhesh. Tumor cells (5 X 10’ per milliliter) were incubated in medium or medium supplemented with the various concentrations of ethanol. Cells were allowed to attach to polycarbonate membranes for 4 h or to petri dishes prior to being harvested with 1 n&f EDTA in nhosnhate-buffered saline (PBS). Cells were counted by hemacytometer and viability was determined.by trypan blue exclusion. Cellproliferation. Cell proliferation was determined by labeling the cells with 1 &ml of 5-iodo-2’-deoxy[G-sH]uridine ([8H]IdUR) (specific activity of 5.0 CifmMol, 1 &i/& Amersham Corp., Arlington Heights, IL). Each Transwell insert (Costar Corp., Cambridge, MA), 6.5 mm in diameter, received 5 X 10’ BlGFlO-labeled cells. Cell proliferation was determined at 4, 12, 24, 36, 48, 60, and 72 h. Cells were maintained at 37°C in a humidified 5% COx incubator. At the conclusion of each time period, attached cells were harvested by the addition of 1 mM EDTA and filtered. The supematant (free 3H) and cellular (bound 3H) radioactivities were then counted separately in a Beckman beta counter.
Alterations of membrane fluidity have been correlated with cellular changes following chemical and viral transformation. Poste and Nicolson [l] described shedding of cholesterol-rich membrane vesicles from tumor cells, leaving behind a more fluid membrane. A direct correlation has been shown between increase in membrane fluidity and the metastatic potential of tumor cells [2]. Cellular migration encompasses cell-substrate interaction, motility and deformability among other factors. Alterations of membrane fluidity can be induced in vitro by aliphatic alcohols [3]. Ethanol and other short-chain alcohols penetrate the lipid bilayer of membranes to increase fluidity [4]. Ethanol also induced changes in the
Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
of Chicago, Maywood,
MATERIAL
INTRODUCTION
00144S27/90
University
AND BRUCE
structure and organization of fatty acid moieties of membrane phospholipids [5]. Specific interactions of ethanol with cytoskeletal proteins, such as F-actin, have been described in hepatocytes [6]. The occurrence of ethanol-induced changes in cell surface properties such as cell membrane fluidity and composition suggests that ethanol can conceivably affect cellular migration and invasion of neoplastic cells treated with ethanol. In this in vitro study we examined the effect of three concentrations of ethanol on the migration and invasion of B16FlO and K1735 murine melanoma cells. Uncoated polycarbonate membranes and a reconstituted basement membrane-like matrix were used to study migration and invasion, respectively. The actin cytoskeleton of untreated and ethanol-treated cells was examined by staining with Bodipy phallacidin.
The short-term effects of ethanol (85.4, 170.8, and 256.2 m&f) on cellular viability, proliferation, migration, and invasion were investigated on murine melanoma cells. Experiments with the fluorescent probe l,B-diphenyl1,3,5-hexatriene indicated that the two highest concentrations of ethanol induced low microviscosity (high lipid fluidity). Cellular viability and proliferation, as determined by the incorporation of [‘H]IdUR, were unaffected by all three concentrations of ethanol. A membrane migration assay and a collagen type IV invasion assay evaluated cellular migration and invasion, respectively. For Bl6FlO and K1735 cells, the migration rate was significantly increased by 170.8 and 256.2 mM concentrations of ethanol. Although the invasion of Bl6FlO cells was not affected, invasion of K1735 cells was inhibited by 170.8 and 256.2 mMethanol. The effect of ethanol on the cytoskeleton was monitored by fluorescent staining of F-a&in. In contrast to untreated cells, F-actin staining of 256.2 m&f ethanoltreated cells showed spike-like projections from the cell surface. Our findings suggest that ethanol can influence cell migration and invasion in vitro, as well as F-actin 0 1990 Academic Preso, Inc. organization.
i To whom correspondence Pathology, Loyola University wood, IL 60153.
Loyoh
ERIC COMRIE,$
64
ETHANOL-INDUCED (DPH, AldFluidity measurement. 1,6-Diphenyl-1,3,5hexatriene rich Chemical Co., Milwalkee, WI) was used as a fluorescent probe for measurement of steady-state fluorescent polarization. The stock solution contained 1 mZtf DPH in acetonitrile. B16FlO cells, 3 X lo’, were labeled by incubating 20 pl DPH/ml of medium for 1 h at 37%. Cells were detached with EDTA solution, washed, and resuspended in PBS to a concentration of 1 X lOa cells/ml. Ethanol was added at the appropriate concentrations during fluidity measurements, Fluorescent polarization was measured with an SPF-5OOCspectrofluorometer (SLM Aminco Instruments, Inc., Urbana, IL). Excitation and emission wavelengths were 360 and 430 nm, respectively. A slit of 5 nm was used for both the emission and excitation wavelengths. Fluorescent polarization (P) was obtained from fluorescent measurements via the equation
p = I”,” - Al,”U”,h/&J &,v
+
1h.v (zv,h,zh,b)
where Z is fluorescent intensity and the subscripts indicate the orientation of the excitation and emmission polarizers, respectively (v, vertical, h, horizontal). The factor Zv,,/Zh,his the transmission efficiency correcting factor. Migration assay. A O.l-ml suspension of cells (5 X 106 cells/ml) was plated into each Transwell insert (top well) while 600 ~1of MEM was added to the bottom wells. Inserts contained a 6.5mm-diameter polyvinyl pyrrolidone-free polycarbonate filter having S-pm-diameter pores. Media were changed daily with the appropriate concentrations of ethanol. Filters were processed for scanning electron microscopy (SEM) as described by McGarvey and Persky [9]. Cells were counted at 500~ for B16FlO cells and at 1200X for the more highly migratory K1735 cells. Znuasim assay. Polycarbonate filters (Transwell inserts) were coated with Matrigel (Collaborative Research, Lexington, MA) diluted 1:20 in MEM. Matrigel contains collagen type IV, heparan sulfate proteoglycan, entactin, and laminin. Coated filters were dried overnight in a laminar flow hood. The following day, filters were rehydrated with medium to form a uniform layer of Matrigel. The rehydrating medium was removed and 5 X 10’ tumor cells were seeded with the appropriate ethanol concentration. Triplicate samples were seeded simultaneously. Cells were refed with media, including ethanol, every 24 h. The filters were processed for SEM. Ethzol concentrations. The amount of ethanol remaining in MEM after 24 h with or without 5 X 10’ B16FlO cells was quantitated by the Paramax analytical system (Baxter Healthcare Corp., Irvine, CA). Fluorescent staining. B16FlO cells were cultured on coverslips for 24 and 48 h in the presence or absence of 256.2 mM ethanol. Bodipy phallacidin (Molecular Probes Inc., Eugene, OR) staining for F-actin was performed as described by Barak et al. [lo]. Statistics. A paired Student’s t-test was performed on the polarixation data. A P value of 0.05 was defined as significant. Migration and invasion assays, which were performed at least in triplicate, were quantified by counting with SEM the number of migratory or invasive cells within 50 consecutive fields on the bottom of each Transwell insert membrane, each field at 500 or 1200~. “Percentage migration” or “percentage invasion” was defined as the mean number of migratory or invasive cells counted per field times 50 (the number of fields counted) times 15.36 (50 fields at 500X is equal to l/15.36 of the filter surface area) or 36.86 (50 fields at 1200X is equal to l/36.86 of the filter surface area) divided by 50,000 (the number of cells originally plated). This definition assumes a uniform cell proliferation on the polycarbonate membrane. Mean and standard error (SE) were calculated for each filter. A mean number of cells for the 150 fields was obtained for each time period (24,48, and 72 h) and for each experimental parameter (0,85.4,170.8, and 256.2 m&f ethanol). Means were compared un-
65
CELL MOTILITY
loo’ cl- -00.6%BtoH
A, . A 1.0%LtOii 80.. o-o 1.5%MOH O-O Control ,p+$-* A ; 20.. 0’ 2 / g 40.. /IO
28
cl
.’ \
0
4
0
12
42
24
20
72
Times~hrs) FIG. 1. Kinetics of B16FlO proliferation in the absence or preeence of ethanol. There is an increase in cell-bound [3H]IdUR incorporation with time. The growth curves for control and ethanol-treated cells are similar. It can be concluded that ethanol had no major adverse effect on proliferation. der parametric conditions with samples not paired. A P value of 0.05 was defined as significant. RESULTS
Ethanol concentrations decreased approximately 40% after 24 h of incubation with or without cells (data not shown). This suggests that ethanol loss was the result of evaporation rather than cellular metabolism. Cell viability and adhesion. Viability was determined at the time of plating cells (0 h) on the polycarbonate membrane and again 4 h later following cell attachment. Cell viability in MEM alone and in MEM supplemented with 85.4,170.8, and 256.2 mikf ethanol was greater than 90%. Viability after incubation for 4 h was 94, 89, 84, and 88%) respectively. None of the examined concentrations of ethanol affected cell viability following a 4-h incubation period. Experiments involving cell detachment from tissue culture-treated petri dishes indicated that 256.2 mM ethanol increased the length of time needed to detach the cells by EDTA (data not shown). Cellproliferation. For each time period, radioactivity was determined for both cell-bound [3H]IdUR (Fig. 1) and unbound [3H]IdUR (not shown). The data indicate that radioactivity for attached cells was greatest between 24 and 48 h of incubation, after which it declined. A concomitant decrease in the radioactivity of the supernatant (unbound [3H]IdUR) was observed. Fluidity measurement. The mean polarization values and standard error of the mean (SE) of three independent samples of DPH-labeled cells were: 0.261 + 0.003 for 0 mkf (control), 0.256 + 0.009 for 85.4 n&f, 0.234 + 0.009 for 170.8 mkf, and 0.229 + 0.009 for 256.2 mM ethanol. The 170.8 and 256.2 mM ethanol-treated cells were significantly more fluid than control cells (P = 0.045 and 0.024, respectively). The 85.4 mM ethanol concentration did not significantly affect fluidity (P = 0.630).
SILBERMAN MIGRATION RATES
FOR BlEiFiO
ET AL.
CELLS
0
0.0
mM EtoH
m
85.4
mM EtOH
m
170.8
mM EtOH
m
256.2
mM Eton
MIGRATION a.5 8.0
48
FOR K1735
CELLS *
Tb
LO.0
7.5
b 4.“+
24 hrs
RATES
mM EtOH
m
85.4
m
170.8
mM EtOH mt4 EtOH
256.2
mH EtOH
T
hrs
Time
Time
(24 hrs)
FIG. 2. Time course of ethanol-mediated B16FlO (a) and K1735 (b) cell migration in the Transwell polycarbonate membrane system. Values are the mean + SE number of cells in three separate experiments. Asterisks represent a significant change from the appropriate 24- or 48-h control.
Migration assay, As shown in Fig. 2, after 24 h culture, 85.4 mM ethanol did not significantly affect B16FlO and K1735 cell migration whereas the higher concentrations of 170.8 and 256.2 mM ethanol showed progressive increase in migration. For B16FlO cells, the increase was significantly greater (155 and 220% for 170.8 and 256.2 mM ethanol, respectively) than that of untreated cells. For K1735 cells, the increase was significantly greater (202 and 223% for 170.8 and 256.2 mM ethanol, respectively). At 48 h, untreated B16FlO cells showed a significant increase in migration compared to that of untreated cells at the 24-h incubation period. All three concentrations of ethanol-treated B16FlO cells, however, showed a significant increase in migration at 48 h (126,133, and 244%) over that of the 48-h control. The highest migration rate was observed in the presence of 256.2 mM ethanol. The percentage migration of untreated cells after 24 and 48 h of incubation was 0.8 and 2.0%, respectively. Invasion assay. Untreated B16FlO cells, which were cultured for 48 h, showed no significant change in invasion from ethanol-treated cells (Fig. 3). The percentage invasion of untreated B16FlO cells after 48 h of incubation was 0.002%. The invasion assay was repeated with the K1735 cell line, but for a 72-h incubation period (Fig. 4). The percentage invasion of the untreated K1735 cells after 72 h of incubation was 7.1%. Ethanol at 85.4 mM concentration did not produce a significant increase in invasion. The 170.8 and 256.2 mM ethanol concentra-
tions induced a progressive and significant decrease in cell invasion, i.e., 44% (P = 0.001) and 74% (P = 1 X lo-s), respectively.
0.00
0
05.4 00s~
170.8 EtOH
256.2
(mM)
FIG. 3. The effect of ethanol on B16FlO cell invasion on Matrigel-coated polycarbonate membranes after 48 h. Values are means f SE of three separate experiments All three concentrations of ethanol have no significant effect on invasion.
ETHANOL-INDUCED eT
CELL
67
MOTILITY
treated cells suggest that assembly of microfilaments may be altered by ethanol. DISCUSSION
6
f x I n ', s i 0 n
f 4
2
o-
0
05.4 Oosa
EtOH
ImMl
FIG. 4. The effect of ethanol on K1735 cell invasion on Matrigelcoated polycarbonate membranes after 72 h. The assay was performed in triplicate. Values are means + SE. Asterisks represent a significant change (i.e., decrease) from the control.
Fhorescent staining. Fluorescent staining of control B16FlO cells for F-actin showed characteristic aggregates and some stress fibers (Fig. 5a). In 256.2 miV ethanol-treated cells, actin appeared to protrude beyond the cell surface, forming fine spike-like projections (Fig. 5b). Differences detected in the staining pattern for ethanol-
The Nuclepore polycarbonate membrane is chemically compatible with ethanol [ll]. Therefore, differences among control and ethanol-treated cells for the migration and invasion assays cannot be attributed to swelling or shrinking of the polycarbonate membrane. The maximum concentration of ethanol, 256.2 mA4 (1.5%), was nontoxic to B16FlO cells over a 4- to 72hr incubation period. In addition, cell detachment from tissue culture-treated petri dishes, which were examined over a 30-min period, was inhibited by 256.2 mM ethanol. The mode of cell adhesion to substrate and the assembly of actin bundles have been shown to be closely related to cellular locomotion [ 121. Migration was significantly increased for both cell lines by 170.8 and 256.2 mA4 ethanol but not significantly altered by 85.4 n&f ethanol. Ethanol has been reported to affect cytoskeletal function. For example, French et al. [6] have shown inhibition of microfilament contraction as well as induction of abnormal intermediate filaments in hepatocytes exposed to ethanol. Ethanol has also been reported to increase the receptor-mediated activation of phospholipase C that leads to increased intracellular calcium in intact hepatocytes [13]. This increase in intracellular calcium could affect actin polymerization and could lead to the altered distribution of F-actin. In our study, the microspikes, which were only seen in the ethanol-treated B16FlO cells, may be due to
FIG. 5. Fluorescent microscopy of control (a) and 256.2 mM ethanol-treated (b) cells stained for F-actin with Bodipy phallacidin. F-actin aggregates are seen in both control and ethanol-treated cells (arrowheads). Microspikes are noted only on the ethanol-treated cells (white arrow). Micrographs are at the same magnification. Bar, 20 pm.
68
SILBERMAN
ET AL
CELL RESEARCH
189,64-68
(1990)
The Influence of Ethanol on Cell Membrane Fluidity, Migration, and Invasion of Murine Melanoma Cells SIMONE SILBERMAN,**~ Departments
of *Pathology,
TERENCE W. MCGARVEY,~ $Biochemktry,
and tdnatomy,
should be addressed at Department of of Chicago, 2160 South Fist Ave., May-
$3.00
PERSKY? Illinois
60153
AND
METHODS
Materials. B16FlO and K1735 murine melanoma cells were a gift from I. Fidler (University of Texas and M. D. Anderson Hospital, Houston). The metastatic capability of the B16FlO and K1735 cell lines have been described by Fidler [7] and Kripke [S], respectively. Cells were maintained at 37’C, 5% CO*, as preconfluent monolayers in Eagle’s Minimal essential medium (MEM; GIBCO Laboratories, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS, GIBCO) and penicillin-streptomycin sulfate (GIBCO). Ethyl alcohol (ZOO proof, Quantum Chemical Corp., Tucsola, IL) was diluted (v/v) in MEM to yield concentrations of 35.4 mA4 (0.5%), 170.8 mM (l.O%), and 256.2 mM (1.5%). CeU viability and adhesh. Tumor cells (5 X 10’ per milliliter) were incubated in medium or medium supplemented with the various concentrations of ethanol. Cells were allowed to attach to polycarbonate membranes for 4 h or to petri dishes prior to being harvested with 1 n&f EDTA in nhosnhate-buffered saline (PBS). Cells were counted by hemacytometer and viability was determined.by trypan blue exclusion. Cellproliferation. Cell proliferation was determined by labeling the cells with 1 &ml of 5-iodo-2’-deoxy[G-sH]uridine ([8H]IdUR) (specific activity of 5.0 CifmMol, 1 &i/& Amersham Corp., Arlington Heights, IL). Each Transwell insert (Costar Corp., Cambridge, MA), 6.5 mm in diameter, received 5 X 10’ BlGFlO-labeled cells. Cell proliferation was determined at 4, 12, 24, 36, 48, 60, and 72 h. Cells were maintained at 37°C in a humidified 5% COx incubator. At the conclusion of each time period, attached cells were harvested by the addition of 1 mM EDTA and filtered. The supematant (free 3H) and cellular (bound 3H) radioactivities were then counted separately in a Beckman beta counter.
Alterations of membrane fluidity have been correlated with cellular changes following chemical and viral transformation. Poste and Nicolson [l] described shedding of cholesterol-rich membrane vesicles from tumor cells, leaving behind a more fluid membrane. A direct correlation has been shown between increase in membrane fluidity and the metastatic potential of tumor cells [2]. Cellular migration encompasses cell-substrate interaction, motility and deformability among other factors. Alterations of membrane fluidity can be induced in vitro by aliphatic alcohols [3]. Ethanol and other short-chain alcohols penetrate the lipid bilayer of membranes to increase fluidity [4]. Ethanol also induced changes in the
Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
of Chicago, Maywood,
MATERIAL
INTRODUCTION
00144S27/90
University
AND BRUCE
structure and organization of fatty acid moieties of membrane phospholipids [5]. Specific interactions of ethanol with cytoskeletal proteins, such as F-actin, have been described in hepatocytes [6]. The occurrence of ethanol-induced changes in cell surface properties such as cell membrane fluidity and composition suggests that ethanol can conceivably affect cellular migration and invasion of neoplastic cells treated with ethanol. In this in vitro study we examined the effect of three concentrations of ethanol on the migration and invasion of B16FlO and K1735 murine melanoma cells. Uncoated polycarbonate membranes and a reconstituted basement membrane-like matrix were used to study migration and invasion, respectively. The actin cytoskeleton of untreated and ethanol-treated cells was examined by staining with Bodipy phallacidin.
The short-term effects of ethanol (85.4, 170.8, and 256.2 m&f) on cellular viability, proliferation, migration, and invasion were investigated on murine melanoma cells. Experiments with the fluorescent probe l,B-diphenyl1,3,5-hexatriene indicated that the two highest concentrations of ethanol induced low microviscosity (high lipid fluidity). Cellular viability and proliferation, as determined by the incorporation of [‘H]IdUR, were unaffected by all three concentrations of ethanol. A membrane migration assay and a collagen type IV invasion assay evaluated cellular migration and invasion, respectively. For Bl6FlO and K1735 cells, the migration rate was significantly increased by 170.8 and 256.2 mM concentrations of ethanol. Although the invasion of Bl6FlO cells was not affected, invasion of K1735 cells was inhibited by 170.8 and 256.2 mMethanol. The effect of ethanol on the cytoskeleton was monitored by fluorescent staining of F-a&in. In contrast to untreated cells, F-actin staining of 256.2 m&f ethanoltreated cells showed spike-like projections from the cell surface. Our findings suggest that ethanol can influence cell migration and invasion in vitro, as well as F-actin 0 1990 Academic Preso, Inc. organization.
i To whom correspondence Pathology, Loyola University wood, IL 60153.
Loyoh
ERIC COMRIE,$
64
ETHANOL-INDUCED (DPH, AldFluidity measurement. 1,6-Diphenyl-1,3,5hexatriene rich Chemical Co., Milwalkee, WI) was used as a fluorescent probe for measurement of steady-state fluorescent polarization. The stock solution contained 1 mZtf DPH in acetonitrile. B16FlO cells, 3 X lo’, were labeled by incubating 20 pl DPH/ml of medium for 1 h at 37%. Cells were detached with EDTA solution, washed, and resuspended in PBS to a concentration of 1 X lOa cells/ml. Ethanol was added at the appropriate concentrations during fluidity measurements, Fluorescent polarization was measured with an SPF-5OOCspectrofluorometer (SLM Aminco Instruments, Inc., Urbana, IL). Excitation and emission wavelengths were 360 and 430 nm, respectively. A slit of 5 nm was used for both the emission and excitation wavelengths. Fluorescent polarization (P) was obtained from fluorescent measurements via the equation
p = I”,” - Al,”U”,h/&J &,v
+
1h.v (zv,h,zh,b)
where Z is fluorescent intensity and the subscripts indicate the orientation of the excitation and emmission polarizers, respectively (v, vertical, h, horizontal). The factor Zv,,/Zh,his the transmission efficiency correcting factor. Migration assay. A O.l-ml suspension of cells (5 X 106 cells/ml) was plated into each Transwell insert (top well) while 600 ~1of MEM was added to the bottom wells. Inserts contained a 6.5mm-diameter polyvinyl pyrrolidone-free polycarbonate filter having S-pm-diameter pores. Media were changed daily with the appropriate concentrations of ethanol. Filters were processed for scanning electron microscopy (SEM) as described by McGarvey and Persky [9]. Cells were counted at 500~ for B16FlO cells and at 1200X for the more highly migratory K1735 cells. Znuasim assay. Polycarbonate filters (Transwell inserts) were coated with Matrigel (Collaborative Research, Lexington, MA) diluted 1:20 in MEM. Matrigel contains collagen type IV, heparan sulfate proteoglycan, entactin, and laminin. Coated filters were dried overnight in a laminar flow hood. The following day, filters were rehydrated with medium to form a uniform layer of Matrigel. The rehydrating medium was removed and 5 X 10’ tumor cells were seeded with the appropriate ethanol concentration. Triplicate samples were seeded simultaneously. Cells were refed with media, including ethanol, every 24 h. The filters were processed for SEM. Ethzol concentrations. The amount of ethanol remaining in MEM after 24 h with or without 5 X 10’ B16FlO cells was quantitated by the Paramax analytical system (Baxter Healthcare Corp., Irvine, CA). Fluorescent staining. B16FlO cells were cultured on coverslips for 24 and 48 h in the presence or absence of 256.2 mM ethanol. Bodipy phallacidin (Molecular Probes Inc., Eugene, OR) staining for F-actin was performed as described by Barak et al. [lo]. Statistics. A paired Student’s t-test was performed on the polarixation data. A P value of 0.05 was defined as significant. Migration and invasion assays, which were performed at least in triplicate, were quantified by counting with SEM the number of migratory or invasive cells within 50 consecutive fields on the bottom of each Transwell insert membrane, each field at 500 or 1200~. “Percentage migration” or “percentage invasion” was defined as the mean number of migratory or invasive cells counted per field times 50 (the number of fields counted) times 15.36 (50 fields at 500X is equal to l/15.36 of the filter surface area) or 36.86 (50 fields at 1200X is equal to l/36.86 of the filter surface area) divided by 50,000 (the number of cells originally plated). This definition assumes a uniform cell proliferation on the polycarbonate membrane. Mean and standard error (SE) were calculated for each filter. A mean number of cells for the 150 fields was obtained for each time period (24,48, and 72 h) and for each experimental parameter (0,85.4,170.8, and 256.2 m&f ethanol). Means were compared un-
65
CELL MOTILITY
loo’ cl- -00.6%BtoH
A, . A 1.0%LtOii 80.. o-o 1.5%MOH O-O Control ,p+$-* A ; 20.. 0’ 2 / g 40.. /IO
28
cl
.’ \
0
4
0
12
42
24
20
72
Times~hrs) FIG. 1. Kinetics of B16FlO proliferation in the absence or preeence of ethanol. There is an increase in cell-bound [3H]IdUR incorporation with time. The growth curves for control and ethanol-treated cells are similar. It can be concluded that ethanol had no major adverse effect on proliferation. der parametric conditions with samples not paired. A P value of 0.05 was defined as significant. RESULTS
Ethanol concentrations decreased approximately 40% after 24 h of incubation with or without cells (data not shown). This suggests that ethanol loss was the result of evaporation rather than cellular metabolism. Cell viability and adhesion. Viability was determined at the time of plating cells (0 h) on the polycarbonate membrane and again 4 h later following cell attachment. Cell viability in MEM alone and in MEM supplemented with 85.4,170.8, and 256.2 mikf ethanol was greater than 90%. Viability after incubation for 4 h was 94, 89, 84, and 88%) respectively. None of the examined concentrations of ethanol affected cell viability following a 4-h incubation period. Experiments involving cell detachment from tissue culture-treated petri dishes indicated that 256.2 mM ethanol increased the length of time needed to detach the cells by EDTA (data not shown). Cellproliferation. For each time period, radioactivity was determined for both cell-bound [3H]IdUR (Fig. 1) and unbound [3H]IdUR (not shown). The data indicate that radioactivity for attached cells was greatest between 24 and 48 h of incubation, after which it declined. A concomitant decrease in the radioactivity of the supernatant (unbound [3H]IdUR) was observed. Fluidity measurement. The mean polarization values and standard error of the mean (SE) of three independent samples of DPH-labeled cells were: 0.261 + 0.003 for 0 mkf (control), 0.256 + 0.009 for 85.4 n&f, 0.234 + 0.009 for 170.8 mkf, and 0.229 + 0.009 for 256.2 mM ethanol. The 170.8 and 256.2 mM ethanol-treated cells were significantly more fluid than control cells (P = 0.045 and 0.024, respectively). The 85.4 mM ethanol concentration did not significantly affect fluidity (P = 0.630).
SILBERMAN MIGRATION RATES
FOR BlEiFiO
ET AL.
CELLS
0
0.0
mM EtoH
m
85.4
mM EtOH
m
170.8
mM EtOH
m
256.2
mM Eton
MIGRATION a.5 8.0
48
FOR K1735
CELLS *
Tb
LO.0
7.5
b 4.“+
24 hrs
RATES
mM EtOH
m
85.4
m
170.8
mM EtOH mt4 EtOH
256.2
mH EtOH
T
hrs
Time
Time
(24 hrs)
FIG. 2. Time course of ethanol-mediated B16FlO (a) and K1735 (b) cell migration in the Transwell polycarbonate membrane system. Values are the mean + SE number of cells in three separate experiments. Asterisks represent a significant change from the appropriate 24- or 48-h control.
Migration assay, As shown in Fig. 2, after 24 h culture, 85.4 mM ethanol did not significantly affect B16FlO and K1735 cell migration whereas the higher concentrations of 170.8 and 256.2 mM ethanol showed progressive increase in migration. For B16FlO cells, the increase was significantly greater (155 and 220% for 170.8 and 256.2 mM ethanol, respectively) than that of untreated cells. For K1735 cells, the increase was significantly greater (202 and 223% for 170.8 and 256.2 mM ethanol, respectively). At 48 h, untreated B16FlO cells showed a significant increase in migration compared to that of untreated cells at the 24-h incubation period. All three concentrations of ethanol-treated B16FlO cells, however, showed a significant increase in migration at 48 h (126,133, and 244%) over that of the 48-h control. The highest migration rate was observed in the presence of 256.2 mM ethanol. The percentage migration of untreated cells after 24 and 48 h of incubation was 0.8 and 2.0%, respectively. Invasion assay. Untreated B16FlO cells, which were cultured for 48 h, showed no significant change in invasion from ethanol-treated cells (Fig. 3). The percentage invasion of untreated B16FlO cells after 48 h of incubation was 0.002%. The invasion assay was repeated with the K1735 cell line, but for a 72-h incubation period (Fig. 4). The percentage invasion of the untreated K1735 cells after 72 h of incubation was 7.1%. Ethanol at 85.4 mM concentration did not produce a significant increase in invasion. The 170.8 and 256.2 mM ethanol concentra-
tions induced a progressive and significant decrease in cell invasion, i.e., 44% (P = 0.001) and 74% (P = 1 X lo-s), respectively.
0.00
0
05.4 00s~
170.8 EtOH
256.2
(mM)
FIG. 3. The effect of ethanol on B16FlO cell invasion on Matrigel-coated polycarbonate membranes after 48 h. Values are means f SE of three separate experiments All three concentrations of ethanol have no significant effect on invasion.
ETHANOL-INDUCED eT
CELL
67
MOTILITY
treated cells suggest that assembly of microfilaments may be altered by ethanol. DISCUSSION
6
f x I n ', s i 0 n
f 4
2
o-
0
05.4 Oosa
EtOH
ImMl
FIG. 4. The effect of ethanol on K1735 cell invasion on Matrigelcoated polycarbonate membranes after 72 h. The assay was performed in triplicate. Values are means + SE. Asterisks represent a significant change (i.e., decrease) from the control.
Fhorescent staining. Fluorescent staining of control B16FlO cells for F-actin showed characteristic aggregates and some stress fibers (Fig. 5a). In 256.2 miV ethanol-treated cells, actin appeared to protrude beyond the cell surface, forming fine spike-like projections (Fig. 5b). Differences detected in the staining pattern for ethanol-
The Nuclepore polycarbonate membrane is chemically compatible with ethanol [ll]. Therefore, differences among control and ethanol-treated cells for the migration and invasion assays cannot be attributed to swelling or shrinking of the polycarbonate membrane. The maximum concentration of ethanol, 256.2 mA4 (1.5%), was nontoxic to B16FlO cells over a 4- to 72hr incubation period. In addition, cell detachment from tissue culture-treated petri dishes, which were examined over a 30-min period, was inhibited by 256.2 mM ethanol. The mode of cell adhesion to substrate and the assembly of actin bundles have been shown to be closely related to cellular locomotion [ 121. Migration was significantly increased for both cell lines by 170.8 and 256.2 mA4 ethanol but not significantly altered by 85.4 n&f ethanol. Ethanol has been reported to affect cytoskeletal function. For example, French et al. [6] have shown inhibition of microfilament contraction as well as induction of abnormal intermediate filaments in hepatocytes exposed to ethanol. Ethanol has also been reported to increase the receptor-mediated activation of phospholipase C that leads to increased intracellular calcium in intact hepatocytes [13]. This increase in intracellular calcium could affect actin polymerization and could lead to the altered distribution of F-actin. In our study, the microspikes, which were only seen in the ethanol-treated B16FlO cells, may be due to
FIG. 5. Fluorescent microscopy of control (a) and 256.2 mM ethanol-treated (b) cells stained for F-actin with Bodipy phallacidin. F-actin aggregates are seen in both control and ethanol-treated cells (arrowheads). Microspikes are noted only on the ethanol-treated cells (white arrow). Micrographs are at the same magnification. Bar, 20 pm.
68
SILBERMAN
ET AL
