We have examined the effects of a number of derivatives of ryanodine on K+ conduction in the Ca2+ release channel purified from sheep cardiac sarcoplasmic reticulum (SR). In a fashion comparable to that of ryanodine, the addition of nanomolar to micromolar quantities to the cytoplasmic face (the exact amount depending on the derivative) causes the channel to enter a state of reduced conductance that has a high open probability. However, the amplitude of that reduced conductance state varies between the different derivatives. In symmetrical 210 mM K+, ryanodine leads to a conductance state with an amplitude of 56.8 +/- 0.5% of control, ryanodol leads to a level of 69.4 +/- 0.6%, ester A ryanodine modifies to one of 61.5 +/- 1.4%, 9,21-dehydroryanodine to one of 58.3 +/- 0.3%, 9 beta,21beta-epoxyryanodine to one of 56.8 +/- 0.8%, 9-hydroxy-21-azidoryanodine to one of 56.3 +/- 0.4%, 10-pyrroleryanodol to one of 52.2 +/- 1.0%, 3-epiryanodine to one of 42.9 +/- 0.7%, CBZ glycyl ryanodine to one of 29.4 +/- 1.0%, 21-p-nitrobenzoyl-amino-9-hydroxyryanodine to one of 26.1 +/- 0.5%, beta-alanyl ryanodine to one of 14.3 +/- 0.5%, and guanidino-propionyl ryanodine to one of 5.8 +/- 0.1% (chord conductance at +60 mV, +/- SEM). For the majority of the derivatives the effect is irreversible within the lifetime of a single-channel experiment (up to 1 h). However, for four of the derivatives, typified by ryanodol, the effect is reversible, with dwell times in the substate lasting tens of seconds to minutes. The effect caused by ryanodol is dependent on transmembrane voltage, with modification more likely to occur and lasting longer at +60 than at -60 mV holding potential. The addition of concentrations of ryanodol insufficient to cause modification does not lead to an increase in single-channel open probability, such as has been reported for ryanodine. At concentrations of > or = 500 mu M, ryanodine after initial rapid modification of the channel leads to irreversible closure, generally within a minute. In contrast, comparable concentrations of beta-alanyl ryanodine do not cause such a phenomenon after modification, even after prolonged periods of recording (>5 min). The implications of these results for the site(s) of interaction with the channel protein and mechanism of the action of ryanodine are discussed. Changes in the structure of ryanodine can lead to specific changes in the electrophysiological consequences of the interaction of the alkaloid with the sheep cardiac SR Ca2+ release channel.