Intoxicating blood levels of ethanol are maintained for several days in mice housed in an atmosphere of ethanol vapor. On removal from the alcohol, all the mice develop withdrawal signs. The signs can be graded to indicate the time course and intensity of the withdrawal reaction.

Ethanol in vitro increased the fluidity of spin-labeled membranes from normal mice. Membranes from mice that had been subjected to long-term ethanol treatment were relatively resistant to this fluidizing effect. The data suggest that the membranes themselves had adapted to the drug, a novel form of drug tolerance.

Anesthetics almost always disorder or "fluidize" membranes, i.e. the drugs increase the mobility of spin labels and reduce order parameters. This effect is universal at high concentrations above the clinical range, but in some kinds of membranes low concentrations of drugs have an ordering effect. Drugs that carry charges, including many local anesthetics, often stiffen membranes, as do long-chain alcohols or fatty acids that mimic natural membrane components. The potencies of short-chain alcohols correlate well with lipid solubility, but a cutoff is reached at 10-12 carbons, where pharmacological actions become weak or absent despite a progressive increase in lipid solubility. The cutoff is partly explained by the ordering action of the long chains and partly by the difficulty of administering such water-insoluble drugs in vivo. The idea of membrane disorder does not exclude some specificity. Closely related drugs may have different molecular shapes and may be capable of forming hydrogen bonds with different orientations, affecting their ability to make membrane more fluid. Perhaps for this reason, there is a remarkable stereospecificity in the disordering effect of anesthetic steroids, chloralose, and long-chain alkenols. Some specificity is mediated by different membrane environments. The drug action may actually reverse from order to disorder on addition of cholesterol, but in other experimental systems cholesterol blocks a disordering effect, and we cannot yet explain the action of drugs in different biomembranes. Further, drugs may have differential solubilities in membranes of different composition. This cannot always be predicted from octanol:water partition coefficients because branched molecules are differentially excluded from structured bilayers. Charged drugs react quite differently with charged and neutral phospholipids and may have differential actions on the two sides of the bilayer because of the asymmetry of the phospholipid distribution. The deeper reaches of the membrane seem particularly sensitive to disordering, even by drugs that presumably reside near the surface. Thus, proteins whose midregions are sensitive to disordering may be especially disrupted by drugs. This is a new field of pharmacology, currently applied only to a small group of drugs. But an understanding of the physiocochemical actions of drugs in hydrophobic regions of cells will clearly be needed for full understanding of membrane-bound drug receptors, enzymes, and transport systems. This is just a beginning.

No abstract