According to the modulated receptor hypothesis, sodium channels have a specific receptor site for local anesthetic and antiarrhythmic drugs. Thus, in the presence of a high concentration of two drugs, competitive displacement of one drug by another may occur. Furthermore, if a drug that has relatively rapid post-stimulation recovery kinetics (e.g., lidocaine) displaces another drug with relatively slow recovery kinetics (i.e., quinidine or bupivacaine), then a net reduction in sodium channel blockade is expected at certain stimulation rates. We tested this prediction, using the maximum upstroke velocity of the ventricular action potential as an indicator of drug-free sodium channels. A single sucrose gap technique was used to stimulate guinea pig papillary muscles, and to control membrane voltage at all times except during the action potential upstroke. Drug-induced inhibition of maximum upstroke velocity increased as the stimulation rate was increased, and was significant (P less than 0.05) at stimulation rates between 2.5 and 4 Hz in the presence of 43 microM lidocaine (n = 5), and between 0.15 and 4 Hz in the presence of 3.5 microM bupivacaine (n = 4). The addition of 43 microM lidocaine to a perfusate containing 3.5 microM bupivacaine resulted in a net increase in maximum upstroke velocity that was significant at rates between 1 and 3.3 Hz, with a maximum increase of 25 +/- 6% at 1.6 Hz. In contrast, addition of 43 microM lidocaine to a perfusate containing 15 microM quinidine did not result in a significant change in maximum upstroke velocity at driving rates between 0.05 and 3.3 Hz (P greater than 0.2; n = 4). However, evidence for displacement of quinidine by lidocaine could be demonstrated by measuring post-stimulation recovery after a conditioning train of 19 10-msec pulses applied at 28 Hz. With this stimulation protocol, 41 +/- 4% of maximum upstroke velocity recovered slowly from block with a time constant of 3.7 +/- 1.2 seconds at - 100 mV in the presence of 15 microM quinidine (n = 5). In the presence of a mixture of 43 microM lidocaine and 15 microM quinidine, this slow component was significantly reduced to 16 +/- 7% (n = 5; P less than 0.01), while 71 +/- 13% of maximum upstroke velocity recovered with a time constant of 115 +/- 21 msec, typical of lidocaine-blocked channels. A two-drug version of the modulated receptor theory was formulated. The effects of drug mixtures could be accounted for by this model.(ABSTRACT TRUNCATED AT 400 WORDS)