FUNDAMENTAL AND APPLIED TOXICOLOGY 36, 6 2 - 7 0 (1997) ARTICLE NO. FA962252 Correlation between Hydrophobicity of Short-Chain Aliphatic Alcohols and Their Ability to Alter Plasma Membrane Integrity SUSAN C. MCKARNS,* CORWIN HANSCH,! WILLIAM S. CALDWELL,* WALTER T. MORGAN,* SARAH K. MOORE,* AND DAVID J. DOOLITTLE* *Research and Development, R. J. Reynolds Tobacco Company, Winston-Salem, North Carolina; and ^Department of Chemistry, Pomona College, Claremont, California Received December 12, 1995; accepted September 10, 1996 ical effects of other, as yet untested, aliphatic alcohols and aliphatic alcohol-like compounds (e.g., anesthetics) on the plasma Correlation between Hydrophobicity of Short-Chain Aliphatic Alcohols and Their Ability to Alter Plasma Membrane Integrity. membrane, o 1997 society of Toxkoiogr. MCKARNS, S. C , HANSCH, C, CALDWELL, W. S., MORGAN, W. T., MOORE, S. K., AND DOOLITTLE, D. J. (1997). Fundam. AppL Tox- icol. 36, 62-70. The quantitative relationship between chemical structure and biological activity has received considerable attention in the fields of pharmacology and drug development More recently, quantitative structure-activity relationships (QSARs) have been used for predicting chemical toxicity. It has been proposed that alcohols may elicit their toxic effects through hydrophobic interactions with the cellular membrane. The objective of this study was to evaluate the role of hydrophobicity in the loss of membrane integrity following acute exposure to short-chain aliphatic alcohols in rat liver epithelial cells in vitro. The series of alcohols studied included methanol, ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol, 1-heptanol, 1-octanol, 2-butanol, 2-methyl-l-propanol, and 2methyl-2-propanol. The lactate dehydrogenase (LDH) assay was used to quantify membrane integrity. The logarithm of the octanol/ water partition coefficient (log P) was used to quantify hydrophobicity. LDH50 values, representing alcohol concentrations yielding a 50% increase in LDH release relative to untreated controls (i.e., mild disruption of membrane integrity), and EC50 values, representing alcohol concentrations yielding 50% of the maximal release of LDH (i.e., moderate disruption of LDH release), were experimentally determined for each alcohol. The LDH50 and EC50 values were then used to derive the QSAR relationship. The aqueous alcohol concentrations yielding LDH50 or EC50 values ranged from 8.9 X 10~4 M (LDH50 for octanol) to 3.5 M (EC50 for methanol), and the log P of the alcohols ranged from -0.77 (methanol) to 3.00 (octanol). From these data, we have derived two QSAR equations describing the role of hydrophobicity in the release of LDH from rat liver epithelial cells following a 1-hr alcohol exposure. The QSAR equation for LDH50 values, log (1/LDH50) = 0.896 log P + 0.117 (n = 11, SD = 0.131), was nearly identical to the QSAR equation for EC50 values, log (1/EC50) = 0.893 log P + 0.101 (n = 11, SD = 0.133], suggesting that similar structureactivity relationships exist at both mild and moderate levels of membrane disruption. Our data indicate that an increase in LDH release was positively and linearly correlated with the hydrophobicity (r = 0.993). These data may help predict the potential biolog- 0272-0590/97 $25.00 Copyright © 1997 by the Society of Toxicology. All rights of reproduction in any form reserved. 62 The pathologies associated with ethanol toxicity, including neurological impairments, liver disease, and certain cancers, are relatively well defined (Lieber, 1994, review; Mufti, 1991; Mufti and Sipes, 1991). However, the mechanisms which underlie these cytotoxic events are not clearly understood. Altered membrane integrity, increased cell membrane fluidity, lipid peroxidation, oxidative stress, impaired protein secretion, mitochondrial injury, acetaldehyde interaction with cellular protein and membrane lipids, and cellular hypoxia are all proposed potential mechanisms of ethanol toxicity (for review, see Zakin and Boyer, 1990). The importance of each of these potential mechanisms of action in the cascade of ethanol induced toxicity is currently under investigation. A large body of evidence suggests that the primary site of action of ethanol in the liver is the cell membrane (for reviews, see Lieber, 1994; Hunt, 1985; Hunt, 1975). Interestingly, an apparent adaptive response to this increased membrane fluidity effect exists and reportedly involves an increase in cell membrane rigidity (Lyon and Goldstein, 1982). Recent studies suggest that biochemical changes (e.g., protein alteration, lipid peroxidation, and oxidative stress) may also be associated with ethanol-induced hepatic cellular injury (Ties and Nagy, 1995; Shaw et al., 1995; Sergent et al, 1995; Chiarpotto et al, 1995; Devi et al., 1993). Likewise, specific biochemical actions involving GAB A, receptor-coupled chloride channels and glutamate receptors (Hoffman and Tabakoff, 1993; Dietrich et al., 1989, review) have been postulated to mediate the effects of ethanol within the central nervous system. Thus, it appears that the potential biological effects of ethanol may be mediated by biochemical (e.g., receptor-mediated) and/or biophysical (e.g., physical disruption of the membrane) processes, and these processes may be cell-type-, concentration-, and time-dependent. HYDROPHOBICITY OF SHORT-CHAIN ALIPHATIC ALCOHOLS 63 TABLE 1 Short-Chain Aliphatic Alcohols Studied Alcohol Structure Methanol Common applications* Antifreeze, octane booster for gasoline, fuel for model airplanes and picnic stoves, paint removers, adhesives, inks, alternative motor fuel, solvent for manufacture of Pharmaceuticals OH Ethanol Alcoholic beverages, lotion, tonics, colognes, perfumes, octane booster for gasoline, topical antiseptic, solvent for pharmaceutical preparations OH 1-Propanol Industrial solvent for resins and cellulose esters, laboratory solvent OH 1-Butanol Industrial solvent for fats, waxes, resins, shellac and varnish, manufacture of lacquers, rayon and detergents, laboratory solvent OH 2-Butanol Industrial cleaners, paint removers, solvent for the manufacture of flavors, perfumes, and dyestuffs OH 2-Methy 1-1 -propanol Paint removers, varnish removers, manufacture of esters for fruit flavor essences CH, 2-Methyl-2-propanol 1-Pentanol H3C Denaturant for ethanol, octane booster for gasoline, manufacture of flavors and perfumes OH Laboratory solvent, solvent for organic synthesis OH 1-Hexanol Manufacture of Pharmaceuticals such as antiseptics amd hypnotics 1 -Heptanol Laboratory solvent 1-Octanol OH Manufacture if perfumes and plasticizers ' Budavari « a/. (1989). Although ethanol has been the most widely studied, many short-chain aliphatic alcohols are toxic as well (Nelson et al., 1990). As a class, short-chain aliphatic alcohols are relatively small molecules consisting of a polar, hydrophilic hydroxyl group and a nonpolar, hydrophobic hydrocarbon chain that has a high affinity for membranes (Hunt, 1975). Therefore, these compounds have the potential to act by physical insertion into the target membrane, disrupting the structural organization of the lipid bilayer, altering plasma membrane integrity, and subsequently disrupting normal membrane function (Lyon et al., 1981). It has been hypothesized that aliphatic alcohols dissolve in hydrophobic regions of the membrane and increase membrane fluidity (Goldstein et al., 1981; McCreery and Hunt, 1978; Chin and Goldstein, 1977; Roth and Seeman, 1972). According to this proposed model, the potency of an alcohol to alter membrane integrity should correlate with the hydrophobicity of the alcohol. The more hydrophobic the alcohol, the more readily it should disrupt membrane integrity. The hydrophobicity of an alcohol can be determined by measuring its ability to partition between hydrophobic and hydrophilic solvents. The most widely used measure of hydrophobicity is the logarithm of the octanol/water partition coefficient, log P (Lindenburg, 1951). Log P can be determined by first allowing the alcohol to equilibrate between water and octanol phases, and then measuring its concentration in each phase. The magnitude of log P is dependent upon a number of variables including the size and molecular structure of the compound (Lindenberg, 1951). For instance, the magnitude of log P of aliphatic alcohols increases with the number of carbons (Paterson et al., 1972). The magnitude of log P for structural isomers of aliphatic alcohols is straight-chain primary > branchedchain primary > secondary > tertiary (Lyon et al., 1981). The depressant, hypnotic, and membrane-disruptive potencies of short-chain aliphatic alcohols have been highly correlated with hydrophobicity in neuronal membranes (Lyon et al., 1981). However, the role that hydrophobicity plays on the toxicity of these compounds in liver cells has not yet been extensively characterized. The objective of this study was to develop a quantitative structure-activity relationship (QSAR) to characterize the 64 McKARNS ET AL. B. Ediancl 150- 125/ 100- i 75- 50- M- 1 4 5 * * 0.0 0.5 * 1.0 1.5 1.0 0.0 0.1 CUJ 0,» 0.4 0.5 0 . 1-S<rtano< i I T —-" i 0.00 0.01 0.02 0.04 0.04 0.05 O.M 0.000 0.005 0.010 0.015 CO X Q 0000 0.002 0.004 O.OOt 0.00* 0.010 0.00000 0.0002J O.OOOSO 0.00075 0.00100 J. 2-H«0iy<- 1-prop«>ol 2S0- 200- I 150- I 100- j 50- 0- * - — 0.0 0.1 O.J O.J 0.4 0.5 0.0 0 1 0.2 O.I Alcohol (M) 0.4 0 5 0.0 0.1 0.2 0.1 0.4 0.6 65 HYDROPHOBICITY OF SHORT-CHAIN ALIPHATIC ALCOHOLS TABLE 2 Experimentally Determined LDH50 and EC50 Values Alcohol" R group" Methanol Ethanol 1 -Propanol 2-Methyl-2-propanol 2-Butanol 2-Methyl-1 -propanol 1-Butanol 1-Pentanol 1-Hexanol 1-Heptanol 1-Octanol CH 3 C2H5 CjH, C(CH 3 ) 3 CH(CH3)C2H5 CH 2 CH(CH 3 ), CjH, C 3 H,, C«H13 C 7 H, 5 C,H 17 LDH50* EC5(T (M) (M) 3.4 1.7 0.47 0.36 0.19 0.16 0.16 0.047 0.012 0.0045 0.00089 3.5 1.6 0.46 0.32 0.16 0.15 0.16 0.047 0.011 0.0042 0.00090 " The short-chain aliphatic alcohols tested in this study. b The hydrophobic R groups of the alcohols tested in this study. c Values derived from the concentration—response curves. Data expressed in molar units. correlation between an alcohol's hydrophobicity and its potential to disrupt plasma membrane integrity in rat liver epithelial cells following acute exposure in vitro. The release of endogenous lactate dehydrogenase (LDH) was used to quantify disruption of membrane integrity. From these data, we have developed a QSAR describing a positive and linear (r = 0.993 and p < 0.001) correlation between the disruption of plasma membrane integrity and hydrophobicity for short-chain, aliphatic alcohols in rat liver cells following acute exposure in vitro. MATERIALS AND METHODS Cells and cell culture. The WB rat hepatic epithelial cell line (Tsao et al., 1984) was used in this study. Cells were grown and maintained in Ham's F12-K nutrient mixture (Kaighn's modified) supplemented with 5% (v/v) heat-inactivated fetal bovine serum (FBS), 1% L-glutamine, and 50 /ig/ml gentamycin. The cell cultures were grown and maintained at 37°C in a humidified incubator containing 5% CO 2 . Stock cultures were maintained in Corning 75-cm 2 plastic tissue culture flasks. F12-K growth medium, FBS, and Hank's balanced salt solution (HBSS) were purchased from JRH Biosciences (Lenexa, KS). Gentamycin was purchased from Sigma Chemical Co. (St. Louis, MO). Chemicals. 1-Butanol (99.5%), 99.5% 2-butanol, 99.5+% ethanol, 98% 1-heptanol, 98% 1-hexanol, 99.5+% 2-methyl-l-propanol, 99.5+% 2methyl-2-propanol, 99+% 1-pentanol, 9 9 + % 1-propanol, and 99+% 1octanol, were purchased from Aldrich Chemical Company (Milwaukee, WI). Methanol (100%) was purchased from Sigma. Alcohol treatment Two x 103 cells suspended in 2 ml culture medium, supplemented with 5% FBS, were pipetted into Corning 35-mm-diameter plastic tissue culture plates. Medium was aspirated from each plate and 2 ml of fresh growth medium (without FBS) containing the test compound was added to confluent monolayers for 1 hr at 37°C. Alcohol dosing solutions were sonicated in autoclaved Pyrex glass flasks at 37°C during preparation, maintained in covered vials in order to minimize evaporation, and used within 4 hr of preparation. Lactate dehydrogenase release. Quantification of lactate dehydrogenase (LDH) in cell culture medium following a 1-hr chemical exposure was used to quantify plasma membrane integrity. An increase of LDH in the medium after exposure to a xenobiotic is correlated with the breakdown of the plasma membrane integrity (Danpure, 1984). It should be noted that an increase of LDH release is not always associated with irreversible cell death, and may, in some instances, represent reversible graded increases in plasma membrane permeability (Danpure, 1984). In our assays, LDH was quantified using a commercially available method based on previously published procedures (Azuma et al, 1986). LDH was measured in cell culture supematants after a 1-hr incubation at 37°C in serum-free growth medium with or without the test compound. Immediately following the 1-hr treatment, a 100-/il aliquot of culture medium from each plate was analyzed for LDH activity using an optimized LDH test kit (DG1340-K; Sigma). This assay is based on the reduction of NAD to NADH by LDH. The reaction was carried out at room temperature and absorbance was monitored at 340 nm using a Beckman DU-70 spectrophotometer. An increase in LDH activity in the extracellular medium is interpreted as a loss of plasma membrane integrity. A minimum of nine readings (three plates, three readings per plate) were collected for each data point. OctanoUwater partition coefficient (P). Published octanol/water partition coefficients for the alcohols were obtained from Hansch et al. (1995). Derivation of the LDH50 and EC50 Values. LDH50 and EC50 endpoints were used to develop the QSAR of the alcohols. The LDH50 values are defined as the concentrations which elicited a 50% increase of LDH release relative to the untreated control. The EC50 values are defined as the concentrations which elicited 50% of the maximal cellular LDH release. The two measures were obtained to compare very subtle changes of membrane integrity (LDH50) with marked cytotoxic effects (EC50). To determine LDH50, the experimental data were fit using a Boltzmann sigmoidal model (Eq. 1). y = (A, - A2y(\ + (exp((AT - x^dx)) + (1) In this model, A, and A2 are values corresponding to the lower and upper limits on the dose-response curve, x is the concentration, Xo is the dose corresponding to the midpoint (50%) between A, and A2, and dx denotes the slope of the dose-response curve. A,, A2, xc, and dx were obtained using the computer-generated Boltzmann model. Equation (1) was rearranged and solved for x to yield Equation (2). LDH50 = *50 = dx*\a((A, - - 1) + (2) In this model, _y50 is the amount of LDH release corresponding to a 50% increase from the untreated control (v50 = ].5*A,). A,,A2,xo,dx, and >>50 were determined from Eq. (1). The EC50 values were obtained by determining rhe alcohol concentrations which yielded 50% of the maximum LDH release. The concentrationresponse data for each alcohol were fit to a Boltzmann sigmoidal model FIG. 1. Concentration-response curves generated for each alcohol following a 1-hr in vitro exposure in WB-344 rat liver epithelial cells. Each data point represents a minimum of nine readings (three plates, three readings per plate). The data are expressed as the mean ± standard deviation. 66 McKARNS ET AL. TABLE 3 Parameters Used To Derive the Quantitative Structure-Activity Relationships (QSARs) Alcohol* f Log/* Observed log (1/LDH50)1' Predicted log (1/LDH50)' A Log (1/LDH50/ Observed log (l/EC50)» Predicted log (1/EC50)* A Log (1/EC50)' Methanol Ethanol 1-Propanol 2-Methyl-2-propanol 2-Butanol 2-Methyl-l-propanol 1-Butanol 1-Pentanol 1-Hexanol 1-Heptanol 1-Octanol 0.17 0.49 1.8 2.2 4.1 5.8 7.9 36 110 520 1,000 -0.77 -0.31 0.25 0.35 0.61 0.76 0.88 1.6 2.0 2.7 3.0 -0.53 -0.22 0.33 0.45 0.73 0.81 0.81 1.3 1.9 2.3 3.1 -0.59 -0.18 0.33 0.41 0.65 0.78 0.89 1.5 1.9 2.5 2.8 0.06 -0.04 0.01 0.04 0.08 0.03 -0.08 -0.2 0 -0.2 0.3 -0.54 -0.20 0.34 0.50 0.79 0.81 0.81 1.3 2.0 2.4 3.1 -0.57 -0.16 0.34 0.43 0.66 0.80 0.91 1.5 1.9 2.6 2.8 -0.03 -0.04 0 0.07 0.13 0.01 -0.1 -0.2 0.1 -0.2 0.3 * The short-chain aliphatic alcohols tested in this study. * P values represent the octanol/water partition coefficient for each alcohol. These published values were obtained from Hansch et al. (1995). c Log P values represent the logarithm of the octanol/water partition coefficient for each alcohol. ' T h e observed log (1/LDH50) values represent the experimentally determined values. These values were derived from the concentration-response curves (Fig. 1). ' The predicted log (1/LDH50) values represent the predicted values. These values were computed using Eq. (3) (text). /r These values represent [observed log (1/LDH50) - predicted log (1/LDH50)]. ' The observed log (1/EC50) values represent the experimentally determined values. These values were derived from the concentration-response curve (Fig- 1). * The calculated log (1/EC50) values represent the predicted values. These values were computed using Eq. (4) (text). ' These values represent [observed log (1/EC50) - predicted log (1/EC50)]. using Origin version 3.5 software (Microcal Inc.). The EC50 values were computer-generated using the Origin software. RESULTS among the alcohols studied (Table 3). These P values were obtained from published work by Hansch et al. (1995). LDH50 and EC50 values were transformed to the logarithms of (1/LDH50) and (1/EC50) (Table 3). Log (1/LDH50) and log (1/EC50) were plotted against log P to develop the QSARs describing the relationship of hydrophobicity to loss of membrane integrity (Fig. 2). These QSARs are described by Eqs. (3) and (4). A series of short-chain aliphatic alcohols (i.e., methanol, ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol, 1heptanol, 1-octanol, 2-butanol, 2-methyl-1-propanol, and 2methyl-2-propanol) were studied (Table 1). Each alcohol tested increased LDH release in a concentration-dependent log 1/LDH50 = 0.896 log P + 0.117 manner (Fig. 1). „ = n , r = 0.993, SD = 0.131 (3) LDH50 and EC50 values were determined to quantify and compare the potency of each alcohol (Table 2). The LDH50 log 1/EC50 = 0.895 log P + 0.101 and EC50 values were calculated from the concentration n= 11, r = 0.993, SD = 0.133, (4) response curves (Fig. 1). These LDH50 and EC50 values were then used to formulate QSARs to describe mild versus moderate disruption of membrane integrity (Eqs. 3 and 4). where n is the number of data points, r is the correlation The LDH50 and EC50 values were nearly identical for each coefficient, and SD is the standard deviation. alcohol, indicative of the steep concentration-response Both log (1/LDH50) and log (1/EC50) (r = 0.993) were curves. The aqueous alcohol concentrations resulting in the positively and linearly correlated with hydrophobicity (Figs. LDH50 and EC50 endpoints varied approximately 4000-fold 2a and 2b, respectively). The slopes (0.896 and 0.895, reamong the alcohols and ranged from 8.9 X 10"4 M (octanol) spectively) and y intercepts (0.117 and 0.101, respectively) to 3.5 M (methanol), indicative of the wide range of toxicity differed little between the LDH50 and EC50 data, suggesting of the series of alcohols tested. a possible similarity in the mechanism(s) of biological action A 6000-fold range, from 0.17 (methanol) to 1000 (octa- between these two portions of the concentration-response nol), in octanol/water partition coefficients (/>) was present curve. The potency to disrupt membrane integrity (expressed 67 HYDROPHOBICITY OF SHORT-CHAIN ALIPHATIC ALCOHOLS a l-octanol % 3- s'^ 2- t3 l-heptanol l-hcxanol I*' 2-methyl-1-propanol A t-l s' 1 ~ y ^ * 1-pentanol 2-butanol A # # i -butanol 2-methy1-2-propanol % y/^\ -propanol 0- /•m«nan < | t h a n 0 1 -1 0993,n-ll,SD-0.131 p<00001 -2- 1 ' 1 ' 1 ' 1 1 -1 1 ' LogP b l-octanol^ 3- s^ y / ^ 1-heptanol 1-hexanol ^ / 22-methyl-1-propanol Log 1- 0- -1 - -2- 2 butanol Lf , //^ 1-pentanol tanol 2-methyl-2-propanol J} y/^ 1 -propanol /""ethano! ^— ^^methanol l ' l y-0.895 log P +0 101 r-O993,n-ll,SD-O133 P< 0.0001 ' 1 1 | 1 | 1 -1 LogP FIG. 2. The correlation of (a) mild, i.e., LDH50, and (b) moderate, i.e., EC50, disruption of membrane integrity with octanol:water partition coefficients. Note die log scales and the correlation (r = 0.993) over a wide range of lipid solubilities and alcohol potencies. 68 McKARNS ET AL. 1-octanol 3- ^ / • / • 1-heptanol 2- ^n-hexanol ^* 1 - 2-butanol l-propanol - / 1-pentanol y ' w 1 -butanol, 2-methy 1-1 -propanol ™2-methyl-2-propanol 0J*ethanol jnethanol y-0.518logP-l 257 r = 0990.n-ll,SD 0.159 P-<0.0001 -1 • 0 1 l i I • I 2 i 1 3 ' 4 1 ' 5 1 ' 1 6 ' 7 1 8 ' 1 9 Total Number of Carbons in Alcohol FIG. 3. The correlation of disruption of membrane integrity with the number of carbons in the alcohol. A linear regression of alcohol potency expressed as the logarithm of (1/LDH50) and the number of carbons (r = 0.990). as 1/LDH50) increased logarithmically with the number carbon atoms in the molecule (slope = 0.52, r = 0.99, and p < 0.0001) (Fig. 3). A plot (not shown) of the EC50 versus the number of carbon atoms resulted in a similar pattern with the same slope (i.e., 0.52). The relative contribution of the molecular shape of the alcohols to alter LDH release is unclear from these data. The order of potency for these compounds in neuronal membranes has been reported to be straight-chain primary > branched-chain primary > secondary > tertiary (Lyon et ai, 1981). Figure 3 suggests a trend toward straight-chain > tertiary (1-butanol > 2-methyl-2-propanol); however, we are unable to differentiate any difference between isoprimary and secondary alcohols (2-methyl-l-propanol = 2-butanol). The log P for each alcohol was inserted in Eqs. (3) and (4) to generate the predicted LDH50 and EC50 values for disruption of membrane integrity. The difference between the predicted and experimentally determined LDH50 values (ALDH50) and the difference between the calculated and experimentally determined EC50 values (AEC50) values are listed in Table 3. The negligible difference between the experimentally determined and predicted values is suggestive of the accuracy of these QSAR equations to predict the potential of alcohols, ranging in log P from 0.17 to 3.0, to disrupt membrane integrity under the conditions studied. DISCUSSION It is well established that the hydrophobicity of a xenobiotic is often associated with biological action (Hansch and Dunn, 1972). A number of models for assessing hydrophobic interactions have been reported. A particularly simple yet effective model is the disaggregation of silanized glass beads (Cecil, 1967). Silanized glass beads have a tendency to cluster together when placed in water and these clusters can be broken up by the addition of hydrophobic compounds like alcohols. The equation describing this purely physical effect of alcohols on silanized glass bead aggregation is similar to equations 3 and 4 and other equations describing alcoholmediated membrane damage (Hansch et ai, 1989). A direct correlation between membrane fluidity and aliphatic alcohol hydrophobicity has been documented for neuronal membranes (Lyon et al, 1981). The objective of this study was to develop a quantitative structure-activity relationship to describe the relationship between hydrophobicity and disruption of nonneuronal membrane integrity for a series of shortchain aliphatic alcohols following acute exposure in vitro. Our results correlating alcohol potency with hydrophobicity (slope = 0.895) and the number of carbons in the alcohol chain (slope = 0.520) nearly parallel similar studies in neuronal membranes (Lyons et ai, 1981), and are in close HYDROPHOBICITY OF SHORT-CHAIN ALIPHATIC ALCOHOLS agreement with other studies suggesting that alcohol potency on membrane damage is mediated via biophysical processes (Hunt, 1985, review; McCreery and Hunt, 1978; and Hansch and Glave, 1971). Overall molecular shape does not appear to be a major contributor to the potency to disrupt membrane integrity in our proposed model. The behavior of the three branchedchain alcohols (2-butanol, 2-methyl-l-propanol and 2methyl-2-propanol) appears to be described by QSAR Eqs. (3) and (4), albeit the behavior of 2-methyl-2-propanol (Fig. 3) suggests that the steric effects may require some additional investigation. Our results are not in agreement with studies suggesting that steric hindrance is important in the ability of aliphatic alcohols to alter membrane fluidity in neuronal membranes (Lyon et al., 1981). Although cell type may be an important factor in this discrepancy, changes in enzyme activity and cell differentiation during cell culture should not be overlooked as possible contributing factors as well ales and Nagy, 1995). In contrast to the toxic effect examined in the present study, several studies report a beneficial or protective effect of alcohols (Seeman, 1972; Roth and Seeman, 1972, 1971). The protective effects of alcohols have been reported to occur at lower concentrations than needed to cause cell lysis, to be concentration-dependent, and to increase with an increase in the numbers of carbons in the alcohol. These reports are consistent with our data which suggest a possible membrane stabilizing effect (i.e., a concentration-dependent inhibition of LDH release) of 1-octanol (Fig. 1). In conclusion, under the conditions tested, an alcohol's potential to increase the release of LDH is positively and linearly correlated with its hydrophobicity in rat liver cells (r = 0.993). Our results are consistent with numerous other investigations suggesting that biophysical processes are a major contributor to the biological effects of alcohols on membranes. One could speculate that while biophysical processes may be largely responsible for the toxicity of hydrophobic alcohols (e.g., heptanol and octanol), biochemical mechanisms may play a more essential role in mediating the toxicity of alcohols with lower hydrophobicity (e.g., ethanol). Clarification of the mechanisms of actions of individual alcohols is necessary to characterize fully the relevance of hydrophobicity in the toxicity of these compounds. The QSAR developed from this study may help predict the potential biological effects of other, as yet untested, aliphatic alcohols and aliphatic alcohol-like compounds (e.g., anesthetics). Furthermore, these data may help direct future efforts in understanding the mechanisms underlying changes in plasma membrane integrity following acute alcohol exposure. REFERENCES Azuma, Y., Ozasa, N., Ueda, Y., and Takagi, N. (1986). Pharmacological studies on the anti-inflammatory action of phenolic compounds. J. Dent. Res. 65(1), 53-56. 69 Budavari, S., O'Neil, M. J., Smith, A., and Heckelman, P. E. (Eds.) (1989). Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals, 1 lth ed. Merck, Rahway, NJ. Cecil, R. (1967). Model system for hydrophobic interactions. Nature 214, 369. Chiarpotto, E., Biasi, F., Aragano, M., Scavazza, A., Danni, O., Dianzani, M. U., and Poli, G. (1995). Ethanol-induced potentiation of rat hepatocyte damage due to 1,2-dibromoethane. Alcohol Alcoholism 30(1), 37-45. Chin, J. H., and Goldstein, D. B. (1977). Drug tolerance in biomembranes. A spin label study of the effects of ethanol. Science 196, 684-685. Danpure, C. J. (1984). Lactate dehydrogenase and cell injury. Cell Biochem. Funct. 2, 144. Devi, B. G., Henderson, G. I., Frosto, T. A., and Schenker, S. (1993). Effect of ethanol on rat fetal hetatocytes: Studies on cell replication, lipid peroxidation, and glutathione. Hepatology 18, 648-659. Dietrich, R. A., Dunwiddle, T. V., Harris, R. A., and Erwin, V. G. (1989). Mechanism of action of ethanol: Initial central nervous system actions. Pharmacol. Rev. 41(4), 489-537. Hansch, C , and Dunn, W. J., Ill (1972). Linear relationships between lipophilic character and biological activity of drugs. J. Phamu Sci. 61, 1 — 19. Hansch, C , and Glave, W. R. (1971). Structure-activity relationships and membrane-perturbing agents. Mol. Pharmacol. 7, 337-354. Hansch, C , Kim, D., Leo, A. J., Novellino, E., Silipo, C , and Vittoria, A. (1989). Toward a quantitative comparative toxicology of organic compounds. Crit. Rev. Toxicol. 19, 185-226. Hansch, C , Leo, A., and Hoekman, D. (1995). In Explonng QSAR. Hydrophobic, Electronic and Steric Constants. American Chemical Society, Washington, DC. Hoffman, P. L., and Tabakoff, B. (1993). Ethanol, sedative hypnotics and glutamate receptor function in brain and cultured cells. Alcohol Alcoholism Suppl. 2, 345-351. Hunt, W. A. (1975). The effects of aliphatic alcohols on the biophysical and biochemical correlates of membrane function. Adv. Exp. Med. Biol. 56, 195-210. Hunt, W. A. (1985). Membrane interactions. In: Alcohol and Biological Membranes (H. T. Blane and D. W. Goodwin, Eds.). Guilford, New York. lies, K. E., and Nagy, L. E. (1995). Chronic ethanol feeding increases the quantity of gas-protein in rat liver plasma membranes. Hepatology 21, 1154-1160. Jain, M. K., Wu, N. Y.-M., and Wray, L. V. (1975). Drug induced phase change in bilayer as possible mode of action of membrane expanding drugs. Nature (London) 255, 494-496. Leiber, C. S. (1994). Alcohol and the liver 1994 update. Gastroenterology 106(4), 1085-1105. Lindenberg, B. A. (1951). Sur la solubility des substances organiques amphipatiques dans les glycerides neutres et hydroxyles. J. Chim. Phys. 48, 350-355. Lyon, R. C , and Goldstein, D. B. (1982). Increased membrane order after chronic ethanol treatment Alcohol Clin. Exp. Res. 6, 148. Lyon, R. C , McComb, J. A., Schreurs, J., and Goldstein, D. B. (1981). A relationship between alcohol intoxication and the disordering of brain membranes by a series of short-chain alcohols. J. Pharmacol. Exp. Ther. 218(3), 669-675. McCreery, M. J., and Hunt, W. A. (1978). Physico-chemical correlates of alcohol intoxication. Neuropharmacology 17, 451-461. Mufti, S. I. (1991). Free radicals generated in ethanol metabolism may be 70 McKARNS ET AL. responsible for tumor promoting effects of ethanol. Adv. Exp. Med Biol 283, 777-783. Mufti, S. I., and Sipes, I. G. (1991). A reduction in mixed function oxidases and in tumor promoting effects of ethanol in a NDEA-initiated hepatocarcinogenesis model. Adv. Exp. Med- Biol. 283, 347-352. Nelson, B. K., Brightwell, W. S., and Krieg, E. F. (1990). Developmental toxicology of industrial alcohols: A summary of 13 alcohols administered by inhalation to rats. Toxicol. Ind. Health 6, 373-387. Paterson, S. J., Butler, K. W., Huang, P., Labelle, J., Smith, I. C. P., and Schneider, H. (1972). The effects of alcohols on lipid bilayers: A spin label study. Biochim. Biophys. Ada 266, 597-602. Roth, S., and Seeman, P. (1971). All lipid-soluble anesthetics protect red cells. Nature 231, 284-285. Roth, S., and Seeman, P. (1972). Anesthetics expand erythrocyte mem- branes without causing the loss of K+. Biochim. Biophys. Acta 255, 190198. Seeman, P. (1972). The membrane actions of anesthetics and tranquilizers. Pharmacol. Rev. 24, 583-655. Sergent, O., Morel, I., Chevanne, M., Cilliard, P., and Cillard, J. (1995). Oxidative stress induced by ethanol in rat hepatocyte cultures. Biochem. Mol. Biol. Int. 35(3), 575-583. Shaw, S., Eng, J., and Jayatilleke, E. (1995). Ethanol-induced free radical injury to the hepatocyte glucagon receptor. Alcohol 12(3), 273-277. Tsao, M. S., Smith, J. D., Nelson, K. C , and Grisham, J. W. (1984). A diploid epithelial cell line from normal rat liver with phenotypic properties of "oval" cells. Exp. Cell Res. 154, 38-52. Zakin, D., and Boyer, T. D. (1990). Alcoholic liver disease. In Hepatology: A Textbook of Liver Disease, 2nd ed. (D. Zakin and D. Boyer, Eds.), pp. 821-869. W. B. Saunders, Philadelphia.