I-iododexetimide Binds Preferentially to the Muscarinic M 1 Receptor In-vivo, but Is It Also a Potential Tool to Assess the Occupancy of Muscarinic M 1 Receptors by Agonists? Cognitive Deterioration in Neuropsychiatric Disorders Is Associated with High Attrition Rates Giving an Urgent Need to Develo

To cite this article: Geor Bakker, et al. The antagonist SPECT tracer 123 I-iododexetimide binds preferentially to the muscarinic M 1 receptor in-vivo, but is it also a potential tool to assess the occupancy of muscarinic M 1 receptors by agonists? Licensed under a Creative Commons Attribution 4.0 International License which allows users including authors of articles to copy and redistribute the material in any medium or format, in addition to remix, transform, and build upon the material for any purpose, even commercially, as long as the author and original source are properly cited or credited. Introduction There are longstanding implications that the muscarinic system, as part of the cholinergic system, plays a critical role in cognition [1]. Administration of cholinergic receptor antagonists give profound cognitive deficits, and cholinesterase inhibitors are broadly prescribed to maintain and improve cognition in neuropsychiatric disorders like Alzheimer's disease [2]. It is thought these deficits are mediated, at least partly, by the muscarinic receptor subtype 1 (M 1 receptor) due to its high expression in prefrontal cortex, hippocampus, and striatum, which are brain regions critical for cognition [3-5]. Moreover, muscarinic M 1 receptor knock out animals showed deficits in learning and memory [6-8]. There is also preliminary evidence that changes in M 1 expression may be the underlying pathophysiology of cognitive deterioration in schizophrenia and related disorders. Importantly, a hallmark post mortem study found 75% reduction in M 1 receptor density in a frontal brain area in a subgroup of patients with schizophrenia compared to healthy controls, which they termed muscarinic receptor deficiency schizophrenia or MRDS [9, 10]. Also, it has been suggested that particularly this subgroup of schizophrenic patients may suffer from cognitive deficits. In addition, clinical pilot studies examining effects of M 1 agonists and positive allosteric modulators (PAM) show improved scores on cognitive test batteries [11-13]. However, measuring the M 1 receptor selectively over the other 4 subtypes of muscarinic receptors in-vivo has proven challenging due to a lack of a well characterised selective radio ligand and the stereotactic homogeneity of the 5 muscarinic receptor subtypes (for a review see [14]). Dexetimide is a candidate compound to image the muscarinergic system as it is a muscarinic antagonist prescribed in clinical practice for neuroleptic-induced Parkinsonism. Already back in the 1990's, two studies were published investigating radiolabelled dexetimide as a radio ligand to image muscarinic receptors in-vivo [15, 16]. These studies showed promising results …


Introduction
There are longstanding implications that the muscarinic system, as part of the cholinergic system, plays a critical role in cognition [1] .Administration of cholinergic receptor antagonists give profound cognitive deficits, and cholinesterase inhibitors are broadly prescribed to maintain and improve cognition in neuropsychiatric disorders like Alzheimer's disease [2] .It is thought these deficits are mediated, at least partly, by the muscarinic receptor subtype 1 (M 1 receptor) due to its high expression in prefrontal cortex, hippocampus, and striatum, which are brain regions critical for cognition [3][4][5] .Moreover, muscarinic M 1 receptor knock out animals showed deficits in learning and memory [6-8]   .There is also preliminary evidence that changes in M 1 expression may be the underlying pathophysiology of cognitive deterioration in schizophrenia and related disorders.Importantly, a hallmark post mortem study found 75% reduction in M 1 receptor density in a frontal brain area in a subgroup of patients with schizophrenia compared to healthy controls, which they termed muscarinic receptor deficiency schizophrenia or MRDS [9,10] .Also, it has been suggested that particularly this subgroup of schizophrenic patients may suffer from cognitive deficits.In addition, clinical pilot studies examining effects of M 1 agonists and positive allosteric modulators (PAM) show improved scores on cognitive test batteries [11][12][13] .However, measuring the M 1 receptor selectively over the other 4 subtypes of muscarinic receptors in-vivo has proven challenging due to a lack of a well characterised selective radio ligand and the stereotactic homogeneity of the 5 muscarinic receptor subtypes (for a review see [14] ).
Dexetimide is a candidate compound to image the muscarinergic system as it is a muscarinic antagonist prescribed in clinical practice for neuroleptic-induced Parkinsonism.Already back in the 1990's, two studies were published investigating radiolabelled dexetimide as a radio ligand to image muscarinic receptors in-vivo [15,16] .These studies showed promising results with high binding of 123 I-iododexetimide in M 1 -rich brain areas and a high brain uptake (7-8% of injected dose) [16] , however it is unclear whether this was specific binding to the M 1 receptor.Consequently, we conducted a series of experiments to determine selectivity and binding profile of 123 I-iododexetimide to all muscarinic receptor subtypes.The current research highlight will address the key findings of these experiments previously published [17] together with ongoing experiments assessing 123 I-iododexetimide as single photon emission computed tomography (SPECT) tracer to measure occupancy of the M 1 receptor by M 1 agonists.These additional experiments resulted from promising results of clinical pilot studies which demonstrated that M 1 agonists and PAMs may improve cognition [11][12][13].
Binding profile of 127/123 I-iododexetimide to muscarinic receptor subtypes Series of in-vitro competitive binding studies were conducted to assess binding affinity and functional antagonism of 127 I-iododexetimide for all five human muscarinic receptor subtypes overexpressed on Chinese hamster ovarian (CHO) cell membranes [17] .The affinity was determined by the displacement of 3 H-n-methylscopolamine, a highly selective M 1 antagonist [18] , by 127 I-iododexetimide.Results revealed that the affinity of 127 I-iododexetimide of binding to the M 1 receptor subtype was in the Pico molar range.
Regarding selectivity, the affinity of 127 I-iododexetimide towards the M 1 receptor was much higher compared to the other subtypes.In addition, 127 I-iododexetimide binding to the M 1 receptor showed the highest affinity to antagonize acetylcholine activated receptor subtypes.Bio distribution studies in rats corroborated these findings by showing that 127 I-iododexetimide could be displaced by the M 1/4 selective agonist xanomeline in a dose dependent manner.To validate binding selectivity of 127 I-iododexetimide to muscarinic receptors, further studies were conducted in control and KO mice for each muscarinic receptor subtype.Results showed that only in KO mice of the M 1 receptor the 127 I-iododexetimide binding was significantly decreased in the M 1 receptor-rich frontal cortex (Figure 1).Finally, bio distribution studies in rats were performed to evaluate whether the antipsychotic olanzapine, which has a high affinity for M 1 receptors (K i = 1.9 nM) [19] and acts as an antagonist, was able to block 123 I-iododexetimide binding in M 1 -rich brain areas ex-vivo.Phosphor storage imaging was conducted to measure brain distribution of 123 I-iododexetimide concurrent with administration of olanzapine [20] .As expected, acute administration of the M 1 antagonist olanzapine resulted in a significant decrease of 123 I-iododexetimide binding in M 1 -rich brain areas.
All things considered, it was concluded that 123 I-iododexetimide preferentially binds to M 1 receptors in-vivo as an antagonist SPECT tracer.

Background
Since M 1 agonists and PAMs may improve cognition [11-13]   , we were interested to test whether the antagonist M 1 SPECT tracer 123 I-iododexetimide may be useful in future imaging studies to evaluate the occupancy of M 1 receptors by M 1 agonists, like xanomeline.As described earlier, we already showed that the acute administration of xanomeline was able to block 127 I-iododexetimide binding dose dependently ex-vivo in rats [17] .In these experiments, liquid chromatography-mass spectroscopy (LC-MS/MS) was used to assess 127 I-iododexetimide binding.An advantage of this technique is that the measurement is not influenced by metabolites that are formed after injection in rats.Contrary, using storage phosphor imaging, the formation of 123 I-labelled metabolites in rats could influence the outcome measurement.
However, in clinical practice, In brief, 16 male Wistar rats (average weight approximately 320 gram) received a single dose of xanomeline (n=8; dose 3 mg/kg body weight intraperitoneally) or placebo (0.3 ml saline; n=8) acutely, whereas 16 other male rats were pre-treated with xanomeline (twice a day 3 mg/kg) or placebo for 14 days.One hour after drug treatment in the acute group, and 24 h after the final injection in the chronic group the rats were anesthetized, injected intravenously in a tail vein with approximately 50 MBq 123 I-iododexetimide (synthesis, specific activity and radiochemical purity as previously described; [17,21] ) and sacrificed as previously described [17] .Then, binding of 123 I-iododexetimide was determined with storage phosphor imaging as earlier described [17] .For analysis, regions of interest (ROIs) were drawn manually for the prefrontal cortex, hippocampus and striatum, areas rich in M 1 receptors, as earlier described [3,17] .Binding in the cerebellum was chosen as the non-specific region because of the low muscarinic acetylcholine receptor expression in this area.According to our previous study, the ratio of specific to non-specific binding was used as the outcome measure [17] .Differences in hippocampal, prefrontal and striatal 123 I-iododexetimide binding ratios for both the xanomeline and placebo treatment was analysed using a one-way multivariate analysis of variance (MANOVA).

Results
The acute group showed no significant decrease in 123 I-iododexetimide binding ratios for all M 1 receptor-rich brain areas examined, whereas the chronic group did show significantly lower binding ratios in all these brain areas (Figure 2).

Discussion
The current bio distribution studies in rats showed that 123 I-iododexetimide binding ratios were not significant lower after acute administration of xanomeline as compared to the placebo condition.This finding was unexpected, since we previously showed that the acute administration of xanomeline was able to block 127 I-iododexetimide binding dose dependently as assessed ex-vivo in rats.In more detail, a single dose of 127 I-iododexetimide decreased e.g., the specific to non-specific binding ratio (which is the outcome measure of our current storage phosphor imaging study) in the frontal binding by approximately 20% (Figure 3).In addition, in our previous study, 127 I-iododexetimide binding was determined 40 min after injection, while in the current study, the rats were killed 2 h after injection of 123 I-iododexetimide.Also, in our previous study LC-MS/MS was used (which measurement is not influenced by metabolites formed after injection in rats), while in the current study we used storage phosphor imaging.These factors might explain why we did not observe a decreased 123 I-iododexetimide binding ratio in our present study after an acute dose of 3 mg/kg xanomeline.We cannot exclude, however, that we will find reduced 123 I-iododexetimide binding ratios after administration of a higher dose than 3 mg/kg.However, the question then remains whether such results would be translatable to humans, because the dose used in this study was already high compared to doses used in human trials [13,22] .So, since the storage phosphor measurements using the SPECT tracer 123 I-iododexetimide may reflect the clinical practice better than results obtained with 127 I-iododexetimide, we conclude that it is not likely that 123 I-iododexetimide SPECT is a useful tool to assess the occupancy of M 1 receptors after acute administration of an agonist like xanomeline.
Since 123 I-iododexetimide itself is a M 1 receptor antagonist, this might explain why 123 I-iododexetimide may not be the ideal radiotracer to assess occupancy of the M 1 receptor by M 1 agonists.Commonly, the occupancy of receptors is much higher when therapeutic doses of antagonists are used as compared to agonists.It is therefore possible that a M 1 receptor agonist radiotracer could better serve as a potential tracer to assess the occupancy of M 1 receptors by M 1 agonists, and further research is needed to explore this possibility.Importantly, the development of M 1 agonist radiotracers for positron emission tomography (PET) imaging has started [23,24] .Interestingly, (sub) chronic administration of xanomeline did induce significantly lower 123 I-iododexetimide binding ratios, possibly reflecting down-regulation of M 1 receptors.It is well known that agonists can induce down-regulation of receptors on the cell membrane [25,26] .Consequently, 123 I-iododexetimide SPECT might be a promising tool to assess the long-term effects of M 1 agonists on M 1 receptor expression.

Future directions
PET or SPECT imaging of M 1 receptors is highly important to fully understand the role of M 1 receptors in cognitive symptoms such as seen in schizophrenia.Cognitive deficits are the best established predictors of functional disability in this disorder [27] .In this regard, our data suggest that 123 I-iododexetimide SPECT may be a useful imaging tool to further evaluate M 1 receptor changes in neuropsychiatric disorders, as a potential stratifying biomarker, or to assess the occupancy of M 1 receptors of M 1 antagonists or after chronic treatment with M 1 agonists like xanomeline, although it may be less suited to evaluate efficacy of agonist drugs.Indeed, we recently started a clinical study in which we will examine the existence of MDRS using 123 I-iododexetimide SPECT.However cognitive deficits in schizophrenia as well as in other neuropsychiatric disorders have proven difficult to treat and therefore more research on M 1 agonists is needed.Future studies should concentrate efforts towards developing an adequate M 1 agonist radiotracer to get more insight into M 1 agonist functioning.
Finally, regarding studies on 123 I-iododexetimde, it may be of interest in future studies to evaluate whether also the acute administration of other M 1 agonists than xanomeline will influence 123 I-iododexetimide.Also, it may be of interest to test whether the 123 I-iododexetimide binding is sensitive to changes in acetylcholine concentrations e.g., induced by cholinesterase inhibitors.

Conclusions
In conclusion, extensive characterisation of 123 I-iododexetimide validates that its antagonistic in-vivo binding predominantly reflects binding to the M 1 receptor.Consequently 123 I-iododexetimide SPECT may a useful means to assess M 1 receptors in-vivo related to cognitive deterioration in neuro-psychiatric disorders, such as Parkinson's disease, Alzheimer's disease and psychotic disorders, and to assess occupancy of M 1 receptors by antagonist M 1 drugs, although it may be less suited to assess efficacy and occupancy of the M 1 receptor of M 1 agonist drugs.

Figure 1 .
Figure 1.Validation of specific binding to M1 receptors of 127 I-iododexetimide in M1-M5 receptor knock out mice, 40 min after injection.Binding potential was calculated as specific binding in frontal cortex (total binding minus nonspecific binding) divided by nonspecific binding.This research was originally published in J Nucl Med.Bakker et al. 123 I-Iododexetimide Preferentially Binds to the Muscarinic Receptor Subtype M1 in Vivo.J Nucl Med 2015; 56:317-22.© By the Society of Nuclear Medicine and Molecular Imaging, Inc.

Figure 2 .
Figure 2. Binding potential of 123 I-iododexetimide in ROIs of prefrontal cortex, striatum and hippocampus.Binding potential was calculated as specific binding (total binding minus nonspecific binding) in ROI divided by nonspecific binding (measured in cerebellum). 123I-iododexetimide binding was measured 2 h after intravenous injection of 123 I-iododexetimide.Upper panel (acute experiment): Rats (n= 8/group) were pre-treated with 1 dose of saline or xanomeline (3 mg/kg) 1 h before injection with radiotracer.Lower panel (chronic experiment): Rats (n= 8/group) were pre-treated for 14 days with 2 doses of saline or xanomeline (3 mg/kg) per day until 24 h before injection with radiotracer.*Statistically significantly lower as compared with control group.

Figure 3 .
Figure 3. Distribution of 127 I-iododexetimide in cerebellum, frontal cortex, striatum, and plasma (n= 3-4) in response to dose increase of agonist xanomeline (1-60 mg/kg) 40 min after injection of 127 I-iododexetimide.Mice were pre-treated with vehicle (0) or xanomeline 30 min before injection of radiotracer.Data points represent mean specific binding ± SEM.This research was originally published in J Nucl Med.Bakker et al. 123 I-Iododexetimide Preferentially Binds to the Muscarinic Receptor Subtype M1 In Vivo.J Nucl Med 2015; 56:317-22.© By the Society of Nuclear Medicine and Molecular Imaging, Inc.