In the context of this review, we will focus on two major esterase enzymes involved in the cholinergic system, AChE and BuChE, for
in vivo nuclear imaging. These hydrolytic enzymes catalyze the breakdown of cholinergic esters into choline, and the corresponding acetate or butyrate ions. They play a regulatory role in maintaining appropriate levels of acetylcholine (Ach) in the synapse [
80]. Dysregulation of their activities has been linked to different neurogenerative disorders, including AD, which is characterized by severe progressive cognitive and motor impairment. The cholinergic dysfunction hypothesis [
81] has been considered, along with the deposition of extracellular misfolded beta-amyloid (senile plaques), and intraneuronal τ-protein aggregation (neurofibrillary tangles), as one of the crucial causes of this chronic, multifactorial disease, but to date, of unknown etiology [
82,
83]. This hypothesis is supported by previous observations of reduced cholinergic activity, particularly ACh levels, in
post-mortem autopsy of AD’s brain patients.
First class of AChE´s PET probes (labeled AChEIs)
In 1996, Pappata
et al. reported the first
in vivo imaging study of human AChE distribution at different anatomical cerebral levels (striatum ˃ cerebellum ˃ thalamus ˃cerebral cortex) using [
11C]Physostigmine (
79) as a tracer [
89]. The synthesis of this radiolabeled AChE inhibitor was previously described by Bonnot-Lours
et al. in 1993 (
Scheme 21) [
90].
Previously, in 1991, the same group reported
N-[
11C]methyltacrine (
82) as radioligand (
Scheme 22) [
91]. However, this tracer exhibited non-specific binding to brain regions, which could be attributed to the lack of selectivity of AChE over BuChE [
89,
92].
Funaki
et al. described in 2003 the radiosynthesis protocol for [5-
11C-methoxy]-donepezil ([
11C]-donepezil)
84 (Scheme 23), a representative inhibitor radiolabeled with an
N-benzylpiperidine moiety (
83) which was examined
in vitro and
in vivo in rat brain [
93]. The study proposed the use of
84 for
in vivo visualization of AChE in the human brain and for evaluating the efficacy of AChE inhibitor therapies [
93]. In 2007, Okamura
et al. utilized
84 for PET imaging to measure
in vivo AChE density in the brains of patients with AD following 6-months oral administration of donepezil [
94]. The aim of this study was to validate
84 as a tool for pharmacological evaluation of donepezil.
Previously, De Vos
et al. reported the biological evaluation of [
11C]-donepezil as a radiotracer for studying AChE, but with the difference that the methoxy group was the one in position 6 [
95]. This study did not yield conclusive results regarding the distribution of AChE in the brain. Nevertheless, it highlighted the importance of radiolabeling at the appropriate position of the structure [
94].
Only few examples in the literature use
18F for labeling molecules as potential radiotracers to measure acetylcholinesterase in the brain, despite its several advantages over
11C. One such advantage is its ease of cyclotron production and longer half-life (109.8 min vs. 20.4 min) which allows for longer periods of
in vivo scanning of biological processes, providing more time to study and observe the desired phenomenon. Finally, from a drug development perspective, the use of fluorine as a bioisostere of hydrogen offers convenience due to its small van der Waals radius, strong bonding capability with carbon, high electronegativity, and lipophilicity [
85,
96,
97].
One attractive example of using
18F radionucleotide is the work of Lee
et al., who described the synthesis and biological evaluation of halogen-substituted donepezil analogues, including their radiolabeled forms. This research highlights the importance of a drug candidate that not only exhibit good activity but also allows for a non-radioactive synthesis that can be adapted to a feasible radiosynthesis within a short period of time and with good yields [
98]. In this case, although the presence of an unlabeled fluorine into C-3 position of the phenyl ring (
86b) demonstrated the highest AChE inhibitory activity, the [
18F]fluorination at the less reactive
meta-position was limited by labeling methods (Scheme X). Therefore, they ultimately opted to prepare the C-4 (
para-
18F substituted) analogue of donepezil (
87) through a reductive amination with
18F-labeled benzaldehydes, which was previously reported by Wuest (
Scheme 24) [
97].
The same conclusion was also observed in the independent works of Lee [
99] and Ryu [
100], who studied a series of
18F labeled compounds which contains
N-benzyl piperidine benzoisoxazole lactam moiety as a pharmacophore (
89a-b,
Scheme 25).
As a continuation of these results and with the advancement of new radiochemical methodologies, Lee
et al. recently synthesized the
18F labelled
meta-isomer ([
18F]3) of CP-118,954 using diaryliodonium salts precursors for direct nucleophilic
18F-labeling. Their aim was to evaluate the
in vivo affinity of AChE in rat brains, and their findings demonstrated that AChE’s affinity is influenced by the position of aromatic fluorine as represented in
Figure 15 [
101].
Using the same pharmacophore than Lee and Ryu, in 2002, Musachio
et al. reported the radiosynthesis of CP118,954 (
90) labelled with
11C instead of
18F (
Scheme 26) [
102]. In parallel, Bencherif
et al. conducted a study of this inhibitor as PET imaging agent to demonstrate the response to AChEIs such as donepezil and to assess changes in AChE binding sites during the progression of AD [
103].
This type of inhibitors, benzoisoxazole lactam derivatives
95, demonstrated potent
in vitro enzyme inhibition in the sub-nanomolar range compared to other related AChE inhibitors containing
N-benzylpiperidine with indanone (
92) [
95], benzoisoxazoles (
94) [
104] or indoles (
93) [
105] groups (
Figure 16). Most of these compounds also exhibited high selectivity for AChE over BuChE. However, despite the strong binding properties and enzymatic inhibitory activity in the nanomolar range of
92,
93, and
94, these inhibitors failed
in vivo mapping of AChE. The lack of correlation between the radioactivity of these PET probes and the AChE binding affinities values, as well as previous data reported about density of AChE in the brain, shows the importance of selecting a specific pharmacophore to design a suitable radiotracer with proper biodistribution [
103].
Wang
et al. developed a
11C-radiosynthesis method for conformationally restricted quaternary ammonium rivastigmine analogues to image both AChE and BuChE (
Figure 17,
101), based on a newer generator inhibitor with dual activity on both enzymes. These probes, which belong to a different class of enzyme inhibitors, exhibit higher affinity compared to their tertiary amine precursors. However, the presence of a positive charge in their structure hampers their ability to pass through the BBB. Therefore, the authors proposed their potential application as tracers for cardiac imaging of AChE and BuChE. This highlights the importance of designing drugs/tracers that can effectively cross the BBB when targeting the CNS in neurodegenerative disorders [
106].
Optically pure (-)-galanthamine, another traditional AChE inhibitor that has received clinical approval for the treatment of mild to moderate dementia in patients with AD [
107], is also used in the development of potential radiopharmaceuticals probes for imaging brain AChE. Building upon this inhibitor, in 2014, Kimura
et al. described the synthesis and radiolabeling of (-)- and (+)-galanthamines (
104a-b) by
N-methylation using [
11C]methyl triflate of norgalanthamines. These precursors were obtained optically pure by chiral resolution (
Scheme 27) [
108].
In vitro and
in vivo experiments were conducted using these compounds in mice to study the distribution and activity of AChE. Biodistribution studies revealed significant differences in the accumulation of radioactivity in different brain sections (such as the striatum and cerebellum) for both tracers. Furthermore, a different response to pre-treatment with donepezil was also observed in blocking experiments (
Scheme 27). These findings demonstrated that only (-)-[
11C]galanthamine
104a can serve as a PET tracer for imaging regions with abundant AChE, providing insights into the pathogenesis and progression of AD. This can be attributed to its similar AChE inhibitory activity to commercially available (-)-galanthamine hydrobromide, along with its specific binding properties [
108].
Recent studies have focused on the design of PET probes with enhanced selectivity between AChE and BuChE. Despite sharing 65% of the aminoacid sequence, these two enzymes differ in their tissue distribution, kinetic properties, and substrates specificity. AChE is predominantly found in nerve cells, specifically in the synaptic cleft (in its soluble form) and in the synaptic membranes (in its bound form). On the other hand, BuChE is primarily associated with glial cells [
109]. Both enzymes present a catalytic active site situated at the bottom of a hydrophobic gorge, as well as a peripheral anionic site. However, the gorge volume of the catalytic site in BuChE is significantly larger (~200 Å) compared to that of AChE (
Figure 18) [
110]. This conformational disparity enables BuChE to accommodate larger substrates and confers differences in their substrate specificity [
109].
In this context, Sawatzky
et al. were the pioneers in synthesizing a selective and potent inhibitor-type radiotracer for BuChE. They achieved this by incorporating
11C or
18F as radioisotopes into the carbamate moiety (
105) of a tetracyclic precursor (
Scheme 28) [
112].
As shown in
Figure 19, the mechanism of action of these PET probes is based on covalent and pseudo-irreversible radiolabeling through a carbamoylation reaction of a serine hydroxyl residue at the active site of the BuChE enzyme [
68].
In vitro kinetics of enzyme inhibition and a first
ex vivo autoradiography with healthy mouse brain slices were also carried out to demonstrate that this type of tracers could enable the
in vivo mapping of BChE distribution [
112].
The inhibition of this type of ligands is transient due to the chemical instability of carbamates. This fact hinders the design of this carbamate-based PET tracers due to the complexity of their kinetic and binding behaviour. It is also necessary to achieve a precise balance in the carbamoylation rate to accurate reflect the true distribution of BuChE throughout the body and the CNS [
84]. Several publications have reported that the introduction of heterocyclic moieties and polar groups at the end of alkyl chains in these carbamate’s inhibitors might improve physicochemical properties, such as like protein binding, penetration through the BBB, and water solubility [
111].
For this reason, more recently, Gentzsch
et al. have developed a new generation of
18F-PET carbamates tracers with a morpholine moiety at the end of alkyl chains, leading to a significantly prolonged duration of action. The synthesis of these probes has been carried out using a novel protecting group strategy for
18F radiolabeling of carbamate precursors (
Scheme 29) [
84].
Second class of AChE´s PET probes (analogues of ACh)
The design of substrate-type of AChE and/or BuChE radioprobes for
in vivo PET imaging is based on modifying the structure of ACh to obtain a neutral, lipophilic substrate, and permeable to the BBB. As depicted in
Figure 20, once the radioprobe (of lipophilic nature) enters the brain (
k1), it undergoes hydrolysis to form a polar radiometabolite that is unable to cross the membrane (due to its hydrophilic nature). The ratio of the trapped radiometabolite relies on the activities of ChEs (
k3), and the evaluation the generated radioactivity is crucial in elucidating the role of cholinesterase enzymes in AD. According to Kikuchi
et al., when applying the two-tissue compartment kinetic model, it is also important to consider the rate of back diffusion of the radioprobe (
k2) and the rate of metabolite elimination (
kel) to obtain an accurate value for
k3, which is associated with AChE and/or BuChE activities [
88].
In most of examples described in the literature, the basic scaffold consists on a
N-methylpiperidin-4-yl ester with a different acyl group (acetate, propionate, isobutyrate and butyrate), which is hydrolyzed to
N-methylpiperidin-4-ol metabolite. This acyl group determines the specificity for the enzyme. As mentioned earlier, larger substrate sizes provide a better fit to BuChE. Hence, [
11C]MP4A (
118d) and [
11C]MP4P (
118e) exhibit higher specificity for AChE compared to BuChE, whereas [
11C]MP4B (
118f), is considered a potential radiopharmaceutical for BuChE. Currently, the first two
11C labeled analogues are approved for clinical use on neurodegenerative diseases by PET (
Table 1) [
113,
114]
For instance, in a small clinical two-phase study lasting 12 months, Kadir
et al. utilized the [
11C]MP4P (
118 e) to investigate the effect of galantamine on cortical AChE and nicotinic receptor binding in 18 patients with mild AD. The primary focus of the study was to evaluate these effects using PET imaging techniques [
115].
As mentioned in the introduction, authors such as Namba
et al. and Shinotoh
et al. have proven the use of these radioligands as a valid strategy to distinguish between AD (B), Parkinson’s disease (C), and progressive nuclear palsy (D). This distinction is supported by the PET-generated images shown in
Figure 21 [
116,
117].
The design of PET probes with
N-methylpiperidin-3-yl esters have also been studied. However, the presence of an asymmetric carbon in these compounds complicates their handle as radiotracers. Prior separation of the optical isomers is necessary to establish a proper correlation with the described kinetic model. This separation is required due to the different rates of hydrolysis of isomers by AChE, and it is essential for the accurate design of a radioprobe [
118].
In terms of radiosynthesis of these probes, the most commonly used method is the
N-[
11C]methylation using [
11C]methyl iodide or [11C]methyl triflate as radiolabeled precursors. However, when mapping cerebral AChE using the metabolite trapped method, it is not suitable to include the radioisotope into the acyl group. If the acyl group were radiolabeled, the resulting [
11C]acetic acid generated in the blood would enter the brain and be metabolized by glial cells into
11CO
2, which would be rapidly cleared from the brain (
Scheme 30) [
88].
Once again,
18F radiolabeled derivatives were designed to increase the half-life of the radiotracer. The most significant examples are [
18F]FEtP4A (
121a) and [
18F]FEP-4MA (
121b,
Scheme 31). However, the preparation strategy had to be modified for these tracers due to instability of the fluoromethyl group attached to the secondary amine and the adverse effect on AChE activity caused by the undesirable fluorine ion generated through defluorination. Kikuchi
et al. described the successful preparation of these radioligands through the use of radioactive fluoroethylation of
N-piperidin-4-yl esters (
120a-b) with [
18F]fluoroethyl triflate, tosylate or different halogens (-Br, -I) [
88].