To the Editor:
We have several concerns about the data and the conclusions of the article by Fagerlund et al. 1that reported on block of adult human muscle acetylcholine receptors (nAChR) by nondepolarizing neuromuscular blockers (NDMBs). Overall, the study by Fagerlund et al. 1confirms that nondepolarizing neuromuscular blocking drugs have both competitive and noncompetitive blocking actions at neuromuscular nicotinic receptors. However, the study does not have the resolution to define the time or receptor state dependence of the block and, hence, provides no insights into the relative roles of the mechanisms in the clinically relevant actions of NDMBs. Accordingly, it is not possible to conclude that NDMBs have a qualitatively different mode of action for blocking the human neuromuscular junction. Indeed, the current results have a great similarity to the results obtained with similar techniques on receptors derived from other species. Given the wealth of information on receptors from other species and the strong support for the idea that the competitive mechanism underlies functional block for these species, it is still most likely that these relaxants are functionally competitive blockers in humans.
Our major concern is that their experimental protocol does not produce a known set of receptor states; moreover, these states change with time during an application. Several investigators2,3have noted that whole frog oocytes cannot be superfused quickly enough with agonist to make reliable determinations of receptor kinetics. As a result, the acetylcholine (ACh)-activated current at any time point represents a combination of open, closed (≥3), and desensitized (≥2) receptor states. Experiments performed in the presence of antagonists produce at least two additional states. Unless measurements are made after the system has equilibrated among the states or at times when the states are clearly defined, the results cannot be interpreted. The traces shown by Fagerlund et al. 1are complex (e.g. , in fig. 1, there are multiple phases to the responses, whereas in fig. 5, the time course is slow) but suggest that the time course for solution exchange is relatively slow (seconds) and that the exchange may not be uniform over the entire oocyte surface.
A consequence of slow solution exchange is that receptor desensitization complicates the interpretation of the measured currents. Desensitization of muscle nAChR proceeds mainly from the open state,4and therefore, both the rate and the extent of desensitization are enhanced at higher ACh concentrations. In human adult ACh receptor, 10 μm ACh activates approximately 60% of the receptors, and fast desensitization proceeds with a time constant of approximately 100 ms, and results in 90% desensitization (Mandy Liu, Ph.D., unpublished data, November 2005).
A final kinetic factor that can affect measurements with relatively undefined concentration change time courses is the establishment of the competitive equilibrium between agonist and antagonist at the agonist-binding site. Depending on concentration and protocol (e.g. , preapplication vs. coapplication), it can take some time for equilibrium competition to be established. As exemplified by controlled perfusion protocols, the resultant time course can be biphasic.5,6
The idea that NDMBs exhibit noncompetitive actions is not new. The classic experiments by Colquhoun et al. 7definitively demonstrated that d-tubocurarine has both competitive and noncompetitive inhibitory actions on frog muscle receptors. Furthermore, they reported that the noncompetitive mechanism was more apparent at higher agonist concentrations (i.e. , at higher levels of channel activation). From their careful analysis of the concentration and voltage dependence of the inhibition, they concluded that the noncompetitive mechanism reflected open-channel block. The open-channel block actually had a higher affinity than the competitive block; however (as they point out), the open-channel block is not of major functional or clinical importance. The reason for this is that the channel must be open to be blocked, whereas during normal physiologic function, the channels are open very briefly and significant block does not develop. Accordingly, the competitive block, which is established for resting receptors, provides the clinically relevant muscle relaxation. Further studies8,9reported similar (but less comprehensive) observations at mammalian nAChR. At normal membrane potentials, channel block by curare develops at a rate of about 107m−1· s−1, for a channel when it is open, and therefore, at the higher concentrations of agonist and NDMB, the block will be significant during the initial seconds of the slow applications used by Fagerlund et al. 1Hence, an agonist concentration activating more than half the receptors (e.g. , 10 μm ACh) would be expected to show a significant contribution from open-channel block, as is observed.1
Fagerlund et al. 1suggest that noncompetitive block is related to receptor desensitization. Possible interactions between desensitization and block do not seem to be required to explain previously published observations. In addition, other studies of the interaction between d-tubocurarine and the mouse fetal muscle nicotinic receptor have not found any indication that d-tubocurarine desensitizes nAChR.10
One way to determine the half maximal inhibitory concentration (IC50) values of competitive antagonists is to perfuse outside-out patches rapidly with saturating concentrations of ACh and to assess the number of activatable channels before significant desensitization, channel block, or dissociation of antagonist.5,6,11–14This method avoids the complications of multiple receptor states: at the time of the peak current, nearly all the receptors are either in the open state or in one of the antagonist-bound (nonconducting) states. An alternative approach is to activate channels with low concentrations of agonist; this reduces the effects of both desensitization and channel block.15Neither method, however, directly addresses the question of whether inhibition is competitive with agonist. That question is addressed with α-bungarotoxin binding experiments. Importantly, the antagonist affinity derived from the binding experiments are fully able to predict functional block of responses15and are in agreement with more quantitative studies of functional block.5,6,11–14In other words, the functional consequences of NDMB agents can be quantitatively explained by competitive inhibition of ACh binding.
In summary, the study by Fagerlund et al. 1confirms that nondepolarizing neuromuscular blocking drugs have both competitive and noncompetitive blocking actions at neuromuscular nicotinic receptors. However, the experimental protocols do not have sufficient definition to allow quantitative analysis.
*Stony Brook University, Stony Brook, New York. james.dilger@stonybrook.edu