Drs. Lieberman, Javitch, and Moore Reply

To the Editor: Dr. Pomara raises a very important issue regarding the evaluation of the potential efficacy of xanomeline, or any candidate molecule, as an antipsychotic therapy in the treatment of schizophrenia, particularly pertaining to its effects on cognition. The standard preclinical screening strategies often do not include evaluation of the effects of chronic administration of a candidate drug. Although the reduced throughput and increased immediate costs would seem prohibitive, the actual cost of failing to assess a drug’s chronic effects on physiological and cognitive processes relevant to the disease is a reduction in predictive validity, particularly with regard to cognitive, behavioral, and/or physiological effects that may interfere with the drug’s therapeutic effects (1) .

We were able to find very few studies examining the behavioral, neurochemical, or neurophysiological effects of the chronic administration of xanomeline or other M ₁ agonists. Reported effects of the chronic administration of xanomeline and other compounds with M ₄ agonism include a reduction in the proportion of spontaneously active dopamine neurons in the ventral tegmental area (2 , 3) —an effect that would be predicted based on studies of other antipsychotic drugs—leading to decreased dopamine efflux in the striatum (4) . In the case of M ₄ agonists, since medial dopamine neurons are more affected, the action of the drug in “clamping” dopamine release would be predicted to affect the medial striatum more than the lateral nigrostriatal pathway. This effect on mesostriatal dopamine release, which may be mediated via xanomeline’s actions on both midbrain dopamine neurons and in the striatum, must be considered as a potential mechanism underlying the ability of the drug to control psychosis.

Another issue to consider in assessing the mechanisms underlying the effects of xanomeline is the extent to which these effects result from M ₁ /M ₄ agonism as opposed to effects of the compound on other molecular targets. For example, similar to clozapine, xanomeline is a relatively potent antagonist at serotonin 5-HT _2A , 5-HT _2C , and 5-HT ₇ receptors, with affinities for these receptors comparable with those for muscarinic receptors. Xanomeline also has moderate affinity for dopamine D ₃ receptors. The recent discovery of highly selective allosteric potentiators of M ₁ and M ₄ receptors (5 – 7) presently allows for a determination of the contribution of M ₁ /M ₄ agonism to the beneficial effects of xanomeline.

An important “site of action” of xanomeline, as highlighted by Dr. Pomara, is the chronic regulation of acetylcholine release in the neocortex and hippocampus. In a number of studies, schizophrenia patients have exhibited decreased binding at M ₁ and M ₄ receptors, an effect consistent with chronically increased extracellular levels of acetylcholine (8) . Many of these studies were conducted among medicated patients, and we have no way of knowing whether chronic “overstimulation” at M ₁ and M ₄ receptors is the mechanism by which receptor binding is reduced in schizophrenia. Nevertheless, this finding, as well as other findings, has led some investigators to postulate that chronic and/or intermittent hyperactivity of cortical cholinergic transmission plays a role in the attentional deficits and psychotic symptoms in schizophrenia (9) . However, more recent studies have indicated that a higher resting “set point” for extracellular acetylcholine in the cortex may be beneficial as long as the system retains the capacity for phasic increases in cholinergic transmission (10) . Increases in acetylcholine in response to (and contingent upon) cognitive demands are essential for normal attentional processing. Disruptions of attentional mechanisms produced by a loss of responsivity of cholinergic transmission in the cortex may contribute to inappropriate encoding of environmental contingencies.

Until the appropriate studies are conducted, we have no way of knowing the effects of chronic M ₁ receptor stimulation on cognitive modulation of acetylcholine release in the cortex. However, it is encouraging to note that chronic administration of the M ₁ agonist CI1017 resulted in improvement in the acquisition of a hippocampally dependent Pavlovian association, possibly through downregulation of an M ₁ -mediated after-hyperpolarization of hippocampal neurons. Thus, selective downregulation of M ₁ receptor-mediated effects through chronic stimulation (and possible internalization [11] ) of M ₁ receptors may serve to increase responsivity of neocortical and hippocampal neurons to acetylcholine through nicotinic and other muscarinic mechanisms and leave intact modulation by other monoaminergic systems, including dopamine. Consequently, we might be tempted to postulate that xanomeline may represent an improvement over dopamine D ₂ antagonists in that it controls striatal dopamine release without blocking D ₂ receptors but also maintains—perhaps even therapeutically adjusts—a set point for cortical cholinergic transmission that also permits dynamic cholinergic transmission. The latter mechanism may serve to help improve or preserve attentional abilities and, in turn, reduce the probability of inappropriate association formation that may contribute to psychosis and reduced functional outcome (12) . We hope that those vested in this therapy are moved to test these hypotheses with chronic drug studies in animal models.