Positron Emission Tomography Imaging of Synaptic Dysfunction in Parkinson’s Disease

Abstract

Parkinson’s disease (PD) is one of the most common neurodegenerative diseases with a complex pathogenesis. Aggregations formed by abnormal deposition of alpha-synuclein (αSyn) lead to synapse dysfunction of the dopamine and non-dopamine systems. The loss of dopaminergic neurons and concomitant alterations in non-dopaminergic function in PD constitute its primary pathological manifestation. Positron emission tomography (PET), as a representative molecular imaging technique, enables the non-invasive visualization, characterization, and quantification of biological processes at cellular and molecular levels. Imaging synaptic function with PET would provide insights into the mechanisms underlying PD and facilitate the optimization of clinical management. In this review, we focus on the synaptic dysfunction associated with the αSyn pathology of PD, summarize various related targets and radiopharmaceuticals, and discuss applications and perspectives of PET imaging of synaptic dysfunction in PD.

Introduction

Parkinson's disease (PD) is the second most common neurodegenerative disease. Its pathogenesis is complex and multifaceted. But what is clear is that the synapse serves as the fundamental functional unit of neurons and plays the crucial node within neural circuits, and its degree of dysfunction determines disease symptoms and severity. The misfolding and deposition of alpha-synuclein (αSyn) lead to irreversible synapse dysfunction of dopamine and non-dopamine systems in PD. The progressive degeneration of dopamine terminals in the substantia nigra (SN) is the main pathological feature. When a certain number of dopaminergic neurons are lost, the typical motor symptoms in PD occur. Furthermore, in response to the dramatic alterations in dopamine function, various non-dopaminergic systems change accordingly. It is increasingly acknowledged that the presence of many PD symptoms is also attributed to disturbances in non-dopaminergic regulatory systems. For example, abnormal activation of non-dopamine systems such as the serotonin or cholinergic system is involved in the pathogenesis of Parkinsonian motor behavior [1, 2]. Non-dopamine system dysfunction is also a main cause of non-motor symptoms in PD, such as depression, anxiety, and insomnia. These motor and non-motor symptoms resulting from synaptic dysfunction greatly impact the quality of life of PD patients and may even elevate the risk of disability in severe cases [3]. At present, the traditional diagnosis of PD mainly relies on medical history, clinical manifestations, and neurological examination, which has a lag and low accuracy. A large amount of data indicates that early diagnosis and intervention in PD can evidently improve patients’ quality of life and prolong their lifespan. Therefore, it is important to diagnose PD as early as possible for effective prevention and treatment.

With the development of precision medicine, molecular imaging has gained increasing attention, by which detecting early subtle changes in neurological diseases can be supporting evidence in guiding clinical management [4–6]. As a representative molecular imaging technology, positron emission tomography (PET) can measure the distribution of radiotracers in vivo, with outstanding capabilities to visualize, characterize, and quantify biological processes at molecular and cellular levels. Increasingly, PET molecular imaging has been successfully applied in clinical diagnosis, treatment, and research on PD. To date, a series of imaging probes have been developed for the visualization of synaptic function in PD.

In this review, we focus on the synaptic dysfunction related to the αSyn pathology of PD. Next, we summarize various targets and radiopharmaceuticals related to dopaminergic and non-dopaminergic synaptic dysfunction in PD (Table 1, Fig. 1), and discuss their applications. Finally, we present the current limitations of PET imaging of synaptic dysfunction in PD and analyze future directions of development, providing new perspectives and suggestions for further research and applications.

Table 1.

PET radiopharmaceuticals for synaptic dysfunction of dopamine and non-dopamine systems in Parkinson’s disease.

	Targets	Radiopharmaceuticals	References
Dopamine system
Dopaminergic system	Aromatic L-amino acid decarboxylase	[18F]DOPA	Ref.[18]
	Vesicular monoamine transporter 2	[11C]TBZ, [11C]MTBZ, [11C]DTBZ	Ref.[23, 92, 93]
		[18F]FP-DTBZ	Ref.[25]
	Dopamine transporter	[11C]beta-CFT, [11C]PE2I	Ref.[28, 30]
		[18F]FP-CIT, [18F]FE-PE2I	Ref.[76, 82]
	Dopamine receptor	[11C]raclopride, [11C]NMSP, [11C]NPA, [11C]MNPA, [11C]PHNO	Ref.[33, 34, 37, 92, 93]
		[18F]Fallypride	Ref.[35]
Non-dopamine system
Cholinergic system	Vesicular acetylcholine transporter	[18F]FMV, [18F]FAA, [18F]NEFA, [18F]FBT, [18F]FEOBV, [18F]FBVM	Ref.[38, 94–96]
	Nicotinic acetylcholine receptor	[11C]nicotine, [11C]MPA, [11C]ABT-418, [11C]MeQAA	Ref.[39, 40]
		2-[18F]FA-85380, 6-[18F]FA-85380, [18F]Flubatine, [18F]AZAN, [18F]XTRA, [18F]ASEM	Ref.[42, 43, 46, 97, 98]
	Acetylcholinesterase	[11C]AMP, [11C]PMP, [11C]donepezil	Ref.[44, 46]
Serotonergic system	Serotonin transporter	[11C]MCN-5652, [11C]DASB	Ref.[54, 57]
	5-Hydroxytryptamine receptor	[11C]WAY-100635	Ref.[58]
		[18F]FCWAY, [18F]MPPF, [18F]MeFWAY	Ref.[59, 60, 99]
Norepinephrine system	Norepinephrine transporter	[11C]Methylreboxetine, [11C]NS8880	Ref.[63, 65]
		(S, S)-[18F]FMeNER-D2	Ref.[64]
Cannabinoid system	Cannabinoid receptor	[11C]JHU75528, [11C]MePPEP	Ref.[100, 101]
		[18F]MK-9470, [18F]FMPEP-d2, [18F]FPATPP	Ref.[67, 68, 101]

Fig. 1.

Clinical symptoms in Parkinson’s disease and schematic of PET targets related to synaptic dysfunction. DAT, dopamine transporter; VMAT2, vesicular monoamine transporter 2; AADC, aromatic L-amino acid decarboxylase; VAChT, vesicular acetylcholine transporter; AChE, acetylcholinesterase; SERT, serotonin transporter; NET, norepinephrine transporter; CB1/CB2R, cannabinoid type 1 and type 2 receptors; D1/D2R, dopamine type 1 and type 2 receptors; nAChR, nicotinic acetylcholine receptor; 5-HT(1A)R, 5-hydroxytryptamine 1A receptor. (Parkinson’s patient section created with BioRender.com).

The Key Mechanism of Synaptic Dysfunction in PD: Alpha-synuclein (αSyn)

PD is essentially an alpha-synucleinopathy. αSyn is a natural presynaptic neuronal protein encoded by the SNCA gene and is widely expressed in the healthy central nervous system (CNS) [7]. In PD, αSyn undergoes misfolding and pathological deposition to form Lewy body aggregates. These aggregates subsequently spread throughout the brain in a prion-like manner [8]. The substance has different biological effects on the PD mechanism by localizing in different parts. In the mitochondrion, αSyn aggregates can cause mitochondrial dysfunction and induce the production of reactive oxygen species, which greatly increases the level of oxidative stress [9]. This is a common feature of neurodegenerative diseases. The existence of αSyn inclusions has also been confirmed in astrocytes and microglia, which leads to immune system disorders and neuroinflammation [10]. This pathological process causes glial cells to lose their neuroprotective properties and trigger synaptic dysfunction. Importantly, the potential synergistic effect between mitochondrial dysfunction and neuroinflammation greatly prompts the deterioration of synaptic function and eventual apoptosis of neurons as a whole [11] (Fig. 2). In addition, mutations in αSyn-related genes, such as SNCA and A30P, result in increased concentrations of αSyn and faster aggregation, which are also strongly associated with early-onset or hereditary PD [12, 13].

Fig. 2.

Pathological mechanism of synaptic dysfunction in PD. The core pathology of PD is characterized by the misfolding and aggregation of alpha-synuclein, resulting in the formation of Lewy bodies. This aggregation induces neuroinflammation through the activation of microglia and astrocytes, as well as causing mitochondrial dysfunction and oxidative stress. These changes lead to synaptic dysfunction and neuronal degeneration or loss ultimately. ROS, reactive oxygen species.

αSyn enables early diagnosis by targeting the fundamental mechanisms of PD, making it the most valuable and direct biomarker for monitoring disease. However, the development of αSyn-targeted probes has been in its infancy. Although little progress has been made in this field, such as [¹⁸F]BQ2 and [¹¹C]MODAG-001, these compounds have the limitation of poor specificity and cannot provide satisfactory image quality [14, 15]. Recently, Ye's team developed the specific probe [¹⁸F]F0502B for αSyn for the first time and successfully verified it in animals, filling in the gap in the field of αSyn molecular probes at last [16]. This radiopharmaceutical can achieve priority and tight binding with αSyn, and it has no affinity for Abeta and tau fibers, showing high specificity to be a promising radiopharmaceutical for αSyn. At present, the team is conducting clinical transformation and application research in Huashan Hospital. Anyway, the development of an αSyn-specific PET probe remains a challenging task.

Dopaminergic Synapse Dysfunction in PD

Dopaminergic impairment is the most evident alteration resulting from synaptic dysfunction in PD. The role of the dopamine system in synapses generally consists of four components, that is synthesis, transport, binding, and reuptake. Initially, L-tyrosine is hydroxylated to form L-levodopa, which is subsequently decarboxylated by aromatic L-amino acid decarboxylase (AADC) to produce dopamine. Next, dopamine is transported into vesicles by a membrane protein located on the vesicles of monoaminergic neurons, vesicular monoamine transporter protein 2 (VMAT2). After depolarizing neuronal excitation, dopamine in vesicles is released into the synaptic gap and acts by binding to dopamine receptors in the postsynaptic membrane. Parts of the dopamine in the synaptic gap can be reuptaken into the presynaptic neuron cytoplasm through active transport of the dopamine transporter (DAT) on the presynaptic membrane for degradation or reuse. Dopaminergic impairment is the core mechanism of PD, which has a major impact on the nigrostriatal circuit responsible for motor coordination and gives rise to main extrapyramidal symptoms. In the potential mechanism of 1-methyl-4-phenyl-1,2,3,6-tetra-hydropyridine, which is the common method for modeling PD, the lethal toxin, 1-methyl-4-phenylpyridinium, is mistakenly taken up by the DAT, inhibiting the mitochondrial respiratory chain. This mainly leads to the damage or loss of dopaminergic neurons and the nigrostriatal axons, inducing irreversible PD [17]. Many anatomical pathological and imaging studies have confirmed the loss of dopaminergic neurons and the reduction in dopaminergic activity in early PD patients. Therefore, it can be inferred that targets on dopaminergic synapses play a pivotal role in the pathogenesis of PD.

Dopamine Synthesis: Aromatic L-amino Acid Decarboxylase (AADC)

Dihydroxyphenylalanine (DOPA) is the precursor of dopamine, and its accumulation can reflect the activity of AADC and the integrity of presynaptic dopaminergic function. [¹⁸F]DOPA is a fluorinated analog of natural levodopa. After being injected and transported into cells, it is decarboxylated by AADC to form [¹⁸F]dopamine. Since the 1980s, [¹⁸F]DOPA has been successfully synthesized and used in clinical examinations [18]. It was initially used to evaluate the integrity of the dopaminergic system in PD. [¹⁸F]DOPA has a maximal uptake range and the highest standardized uptake value in the basal ganglia, but very low uptake in normal tissues, providing an outstanding lesion/background ratio and better capacity for disease monitoring [19]. However, early decarboxylase frequently exhibits a compensatory phenomenon that makes its fluctuations inconsequential or even normal, limiting the sensitivity of [¹⁸F]DOPA [20]. Moreover, its radiosynthesis process is complex, and finding a simpler and more effective manufacturing strategy has been a challenge [21]. There are no other radiopharmaceuticals targeting AADC, and this area is continuously evolving.

Dopamine Transport: Vesicular Monoamine Transporter Protein 2 (VMAT2)

The initial development of VMAT2 radiopharmaceuticals relied upon the metabolites of tetrabenazine (TBZ) or its analog (2-O-methyl)dihydrotetrabenazine (MTBZ). [¹¹C]TBZ, the earliest radioactive tracer for PET imaging of VMAT2, has been developed and approved for use in many countries since the 1960s. In 1993, Kilbourn and colleagues effectively conducted VMAT2 imaging in the human brain using [¹¹C]TBZ. Subsequently, the case of the highly specific binding of TBZ and MTBZ has aroused other exploration. Dihydrotetrabenzoxazine (DTBZ) is one of the metabolites of TBZ that blocks the storage of monoamine neurotransmitters in presynaptic vesicles by binding to VMAT2 [22]. After modification and common ¹¹C labeling, it has been used in imaging studies of neurodegenerative diseases. [¹¹C]DTBZ is highly sensitive to dopaminergic terminal loss and remains unaffected by compensation or drug regulation. This makes it the most widely used VMAT2 PET molecular probe to evaluate the integrity of presynaptic dopamine terminals [23]. However, this substance is essentially generated through the complete reduction and rapid formation of TBZ under the action of carbonyl reductase. Consequently, its imaging effect does not significantly differ from TBZ-related radiopharmaceuticals [24]. Afterward, a novel radioligand [¹⁸F]FP-DTBZ ([¹⁸F]AV-133) was developed and showed excellent kinetics and affinity in both animal and human brains, offering an improved approach for detecting VMAT2 [25, 26]. The diversity of PET probes for VMAT2 has helped to comprehensively understand dopaminergic presynaptic function.

Dopamine Reuptake: Dopamine Transporter (DAT)

Because of the high affinity and sensitivity of cocaine and its derivatives for the DAT, many DAT imaging ligands have been developed based on phenyltropane cocaine analogs. 2beta-carbomethoxy-3beta-(4-fluorophenyl)tropane (beta-CFT or FP-CIT) is an effective ligand developed earlier for quantifying DAT density and is purified from nor-beta-CFT. We usually label ¹¹C at 0-methyl as a simpler and easier method with a shorter synthesis time. However, there is a serious limitation in that the kinetics of [¹¹C]beta-CFT are so slow that it is only absorbed in the striatum 24 h after injection, which leads to the DAT only being adequately imaged the next day [27]. Later, [¹⁸F]FP-CIT became an excellent substitute taking only 2 h after injection to reach a plateau [28]. Thus, [¹⁸F]FP-CIT has become an excellent substitute because of its more convenient performance in clinical examination. In 1999, a new ligand, N-(3-iodoprop-2E-enyl)-2beta-carbomethoxy-3beta-(4'-methylphenyl)nortropane (PE2I), was developed in an autoradiography study of normal postmortem human brain, showing the main distribution of the DAT in the basal ganglia. It has high structural homology with CIT compounds. At the same time, it has a lower binding force to norepinephrine and the serotonin transporter (SERT), due to the combination of 4-methyl substituted by a benzene ring with 3-iodopropyl-2-enyl [29]. In a competitive in vitro study, it was confirmed that [¹¹C]PE2I has high selectivity and affinity for the DAT, and serves as one of the most selective ligands. However, [¹¹C]PE2I has significant variability in the quantitative measurement of the DAT, and the radioactive metabolites produced by 4-hydroxymethyl bind to the DAT in the striatum, interfering with quantitative monitoring [30]. In comparison, [¹⁸F]FE-PE2I, a novel probe, was evaluated to have a faster elution rate and stronger kinetic activity [31]. More importantly, it produces fewer radiometabolites with intermediate lipophilicity that interferes with quantification [31, 32]. So [¹⁸F]FE-PE2I may be a more suitable choice for the quantitative display of the DAT in vivo.

Binding to Dopamine: Dopamine Receptor

Dopamine receptors are mainly categorized into D1 receptors (D1, D5) and D2 receptors (D2, D3, and D4), and most research is still focused on D2 receptors. [¹¹C] Raclopride is a common D2 antagonist imaging agent with high selectivity and sensitivity, and it shows strong reliability in many clinical trials. [¹¹C] Raclopride PET was used for the first time to report the release of DA from the striatum in humans in 1998. Subsequently, Volkow et al. used modafinil to occupy the DAT and raise dopamine levels and successfully presented this fluctuation with [¹¹C]raclopride, which proved the feasibility of [¹¹C]raclopride as an indicator of changes in extracellular dopamine [33]. However low affinity is a disadvantage that cannot be ignored. In comparison, the radioligand [¹¹C]NMSP has a higher affinity (0.1–0.3 nmol/L) for the D2 receptor. However, it also has the capability to bind to both 5-HT receptors simultaneously, which poses a major concern [34]. With further exploration, [¹⁸F]Fallypride was reported as a D2 radioligand with high affinity and specificity at the same time and is capable of detecting complex defects within the cortico-striato-thalamic circuit [35]. PET studies have reported that D2 agonists are more susceptible to endogenous competition than traditional D2 antagonists, resulting in more sensitive detection of extracellular DA changes [36, 37]. And its affinity is higher and more saturated in vivo. A series of agonists for the quantitative assessment of D2 receptors, such as [¹¹C]NPA, [¹¹C]MNPA, and [¹¹C]PHNO, have been developed as excellent PET radiotracers. In rodent studies, [¹¹C]PHNO showed strong binding signals in the striatum, with a specific/non-specific ratio >90% [37]. Since 2008, researchers have evaluated the performance of these agonist probes in humans, all of which display fast dynamics and an ideal imaging effect with low noise and variability.

Non-dopaminergic Synaptic Dysfunction in PD

It is well known that PD is highly heterogeneous with multiple phenotypes; this is the combined effect of diverse systems and neurotransmitter factors caused by a complex neurochemical imbalance in this disease. Although dopaminergic neuron depletion is the main feature of PD, it is undeniable that the non-dopaminergic system also has significant changes in its pathogenesis (Fig. 3). Non-dopaminergic synaptic dysfunction in PD involves many aspects, such as the cholinergic, serotonergic, norepinephrine, and cannabinoid systems.

Fig. 3.

Applications of PET imaging of non-dopamine systems in PD. A Cholinergic PET images of [¹⁸F]FEOBV VAChT (upper row) and [¹⁸F]Flubatine α4β2 nAChR (lower row) in a single PD patient. In columns A and B, both radioligands show increased uptake in the anterior frontotemporal cortex. In columns C and D, [¹⁸F]FEOBV shows increased uptake in the striatum and thalamus, and [¹⁸F]Flubatine shows increased uptake in the thalamus but not significantly in the striatum. In columns E–G, [¹⁸F]FEOBV shows significantly enhanced radioactivity in the whole cerebellum, while [¹⁸F]Flubatine shows diffusely reduced uptake in the cerebellum. Adapted with permission from Ref.[94]. B Axial, sagittal, and coronal [¹¹C]DASB PET imaging among healthy controls, early unilateral idiopathic Parkinson’s disease (IPD), and advanced bilateral IPD patients, evaluating the availability of 5-HT_1ARs. Adapted with permission from Ref.[102]. C [¹¹C]MeNER PET images of NET sites in healthy controls and patients with PD. Adapted with permission from Ref.[103].

Cholinergic System

For imaging synaptic function in the brain, the cholinergic system consists of the vesicular acetylcholine transporter (VAChT) in the presynaptic terminals, nicotinic acetylcholine receptors (nAChRs) in the postsynaptic terminals, and acetylcholinesterase (AChE) in the synaptic cleft. Neuroimaging studies have shown reduced cholinergic neurons in cortical and subcortical regions of PD patients. The principal non-motor symptom in PD is cognitive impairment, which is widely recognized as a highly incapacitating process involving the cholinergic neurotransmission system. In addition, the cholinergic system partly influences the development of motor symptoms in PD. It has been shown that mice with VAChT deletion are more vulnerable to neurotoxic damage that mimics the dopamine imbalance pathogenesis of PD. This evidence indicates a close relationship between cholinergic dysfunction and PD.

Vesicular Acetylcholine Transporter (VAChT)

In the early stage, a series of vesical analogs labeled with radionuclides, such as [¹⁸F]FMV, [¹⁸F]FAA, and [¹⁸F]NEFA, were developed as PET agents in the quantitative study of VAChT. However, they have little or no value for human application. [¹⁸F] Fluorobenzyltrozamicol ([¹⁸F]FBT) is an effective ligand for cholinergic terminal density. Its PET imaging can detect both acute and chronic changes in cholinergic activity. Related studies using the [¹⁸F]FBT analog (+)-meta-125I-iodobenzyltrozamicol also suggested the consistency of its fluctuations with cholinergic changes, which provides evidence for the clinical imaging application of neuropsychiatric disease in the future. (-)-5-[¹⁸F]fluoroethoxybenzovesamicol ([¹⁸F]FEOBV), a visamicol derivative, is a novel VAChT-specific radioligand that was recently developed. Currently, it is widely used in the quantitative study of cholinergic function in clinical practice to demonstrate substantial efficacy in the assessment of neurodegenerative dementias such as PD and AD.

Recently, a number of promising radiopharmaceuticals have been further developed and promoted for VAChT, such as [¹⁸F]FBVM. It can sensitively detect cholinergic deprivation even in brain regions with low VAChT density, such as the hippocampus and cerebral cortex. Compared with [¹⁸F]FEOBV, it exhibits up to 50 times higher affinity, an enhanced signal-to-noise ratio for brain uptake, and superior cholinergic PET images [38]. However, [¹⁸F]FBVM is currently limited to preclinical stages and is waiting to be applied in clinical practice.

Nicotinic Acetylcholine Receptor (nAChR)

The nAChR, a barrel-shaped hetero-oligomer, is most abundant in its heteromeric α4β2 and α7 subtypes in the CNS. It acts on many physiological and behavioral processes by modifying the effects of nicotine. [¹¹C] Nicotine was developed as the first nAChR radioligand to be used in neurodegenerative disorders, but it is affected by various limitations in the brain like blood flow and the blood-brain barrier (BBB) [39]. Most fatally, it is dominated by non-specific binding, resulting in low-quality images. Two other nicotine analogs, [¹¹C]MPA and [¹¹C]ABT-418, have a similar disadvantage [40]. Later, 3-pyridyl ether derivatives such as 2-[¹⁸F]FA-85380 and 6-[¹⁸F]FA-85380 were developed. They are nicotine agonists with higher affinity and specificity for α4β2 nAChRs, and the toxicity to the organism is greatly curtailed [41]. But they also have a slow kinetics limitation.

In order to solve these problems, novel α4β2 subtype radioligands with fast kinetics have emerged, representatively [¹⁸F]Flubatine, [¹⁸F]AZAN, and [¹⁸F]XTRA. Due to the effective anesthetic and analgesic properties of nAChRs, this kind of radiopharmaceutical was developed on the basis of epibatidine and its derivatives isolated from Ecuadorian poison frogs. In the first human quantitative study, [¹⁸F]Flubatine showed high metabolic stability and a rapid kinetic response [42]. However, [¹⁸F]AZAN and [¹⁸F]XTRA are mostly metabolized after injection, resulting in a serious problem of instability [43]. For the α7 subtype, only a few agents have been developed, such as [¹¹C]MeQAA and [¹⁸F]ASEM, leaving a lot of space to be explored.

Acetylcholinesterase (AChE)

AChE regulates and maintains the concentration of ACh, which keeps the cholinergic system in a state of dynamic equilibrium. Once the hydrolysis of AChE is blocked, ACh does not dissociate from receptors of the postsynaptic membrane, which leads to many neurocognitive disorders in PD. Here are the two most common AChE radioactive probes, [¹¹C]AMP ([¹¹C]MP4A) and [¹¹C]PMP ([¹¹C]MP4P). They both belong to substrate-type ligands for AChE PET imaging. In addition, [¹¹C]AMP and [¹¹C]PMP are small molecular lipophilic analogs of ACh that can easily cross the BBB [44]. It was found that the radioactivity uptake after injection was significantly correlated with the AChE density in different brain regions. Besides, the pharmacokinetics of these compounds are so fast that they can enter the brain and be completely hydrolyzed into radioactive metabolites by AChE only 1 min after injection. The permeability of the BBB to this radioactive metabolite and its elimination is so slow that it can be retained in the brain for a long time, providing a stable time window for observation [45]. At present, these two radiopharmaceuticals are extensively used in clinical and research settings. However, their inability to provide peripheral imaging remains a major limitation to be considered [45].

According to the quantitative results of [¹¹C]AMP on cholinergic inhibition, we found that the inhibitory effect induced by the AChE inhibitor, donepezil, is closely related to the plasma IC₅₀ in a dose-dependent manner. Subsequently, based on the strong affinity and specificity of this reversible inhibitor, a new radiopharmaceutical, [¹¹C]donepezil, has been developed. It is classified as a ligand-type probe and enables visualization of AChE density in peripheral organs, giving access to more extracerebral information than substrate-based probes [46].

Serotonergic System

The serotonergic system, including SERT and 5-hydroxytryptamine (5-HT) receptors, engages in the regulation of cognition, emotion, and motor behavior. Alterations in serotonin signaling are a non-neglectable component in PD. Pathology reports have confirmed that the serotonergic system is damaged in PD at the early stage, which is characterized by decreased expression in several basal ganglia nuclei, such as the putamen, caudate, and ventral caudate-putamen [47, 48]. This may be related to the negative modulation of serotonergic activity by αSyn in PD [49]. Multiple pieces of evidence suggest that serotonergic terminals can contribute to the development of levodopa-induced dyskinesia by modulating the dopamine system and promoting the non-physiologic release of dopamine in the striatum [50, 51]. Besides, some non-motor symptoms of PD, such as cognition, depression, and anxiety, also have a tight association with serotonergic dysfunction [52, 53].

Serotonin Transporter (SERT)

[¹¹C]MCN-5652 is the first successfully applied PET radiotracer for the SERT. It was found to have a good correlation with serotonin uptake sites (r = 0.86) in mice 1 h after injection. At present, [¹¹C]MCN-5652 has been applied to several diseases involving the serotonin system by evaluating the level of the SERT. However, it was found that [¹¹C]MCN-5652 is only suitable for quantitatively visualizing regions with high SERT concentrations. Nonetheless, the specific binding may be omitted in areas with relatively low density, such as the prefrontal cortex [54]. This limitation makes it difficult to obtain reliable quantitative SERT data.

Later, [¹¹C]3-amino-4-(2-dimethylaminomethylphenylsulfanyl)-benzonitrile ([¹¹C]DASB) was discovered as a better probe for visualizing serotonergic function in the brain. [¹¹C]DASB presents clear advantages over [¹¹C]MCN-5652, including a higher plasma-free fraction and faster kinetics. Its specificity and non-specific binding rate to SERT are higher, resulting in a stronger contrast imaging effect [55, 56]. The feasibility of monitoring SERT by [¹¹C]DASB has been evaluated in many rat studies, and the results show extremely high sensitivity and saturation binding (ED₅₀ = 56 nmol/kg) to SERT in the brain and have good reliability and reproducibility [57].

5-Hydroxytyrptamine (5-HT) Receptor

[¹¹C]WAY-100635 is a 5-HT_1A antagonist with high affinity and selectivity that displays a significant specific signal in the serotonergic regions of the human cerebral cortex and raphe nuclei. In a study related to PD tremor, it was used to detect a generalized decrease in postsynaptic 5-HT_1A receptor binding potentials (BP_ND) [58]. Due to the limitations of the ¹¹C marker, researchers developed [¹⁸F]FCWAY and [¹⁸F]MPPF based on the analogs of WAY-100635, which have been used in human research. However, the huge defluorination of [¹⁸F]FCWAY hinders the accurate quantification of 5-HT_1A receptor binding in the brain [59]. In order to solve this problem, a new radiopharmaceutical, [¹⁸F]MeFWAY, was developed with excellent metabolic stability. In the first human study, it showed excellent measurement of 5-HT_1A receptors [60]. Although in a comparative study, the kinetic indicators, the distribution volume ratio, and the area under the regional time-activity curve ratio in [¹⁸F]MeFWAY were not as good as [¹⁸F]FCWAY, it is still a promising radiopharmaceutical [61].

Norepinephrine System

The relationship between the norepinephrine system and PD is multifaceted. In PD, the density and expression of norepinephrine neurons are generally reduced at the locus coeruleus (LC). In the meantime, the presynaptic transporter-mediated function of rapid norepinephrine clearance is diminished as well. Depletion of norepinephrine terminals may also be one of the reasons for the progression of symptoms. A PET imaging study revealed that norepinephrine-related LC degeneration is consistent with postmortem pathology in early PD, pointing to the possibility that norepinephrine injury may result in nonmotor symptoms in PD [62]. Therefore, it is widely agreed that norepinephrine synaptic dysfunction may contribute to the development and progression of PD.

The norepinephrine transporter (NET) regulates neural transmission by reabsorbing up to 80–90% of norepinephrine in the extracellular space. Based on reboxetine analogs known as norepinephrine reuptake inhibitors, many NET ligands have been developed and entered the field of molecular imaging. [¹¹C]Methylreboxetine ([¹¹C]MRB or [¹¹C]MeNER) is a mature probe on the NET occupancy that has a very high affinity for the NET, making its affinity with other targets negligible [63]. PET imaging with [¹¹C]MRB can effectively draw the map of norepinephrine defects in vivo. However, it has certain limitations in application due to its low absorption rate, sensitivity, and reliability. (S, S)-[¹⁸F]FMeNER-D2, as a new type of radioactive ligand, has a more stable determination time and peak uptake than [¹¹C]MRB. But its BP_ND is so low that it is only 0.4–0.6 in the anterior cingulate cortex [64]. There are also many other NET radioligands, like [¹¹C]NS8880, all of which have greatly improved properties over previous developers [65]. However, they all need to be validated in subsequent clinical trials.

Cannabinoid System

The endogenous cannabinoid system (ECS) was discovered in humans and other mammals as early as 1992. At present, the two main known receptors are cannabinoid type 1 (CB1R) and cannabinoid type 2 (CB2R), which play important roles in regulating various psychological and physiological functions through their interaction with endocannabinoids. CB1R is mainly found on the presynaptic terminals of dopaminergic neurons, and its activation increases the release of dopamine. It has been found that the overexpression of CB1R in PD may contribute to neuroinflammation, oxidative stress, and cell death processes. The above sections are all vital contributors to synaptic dysfunction. CB2R exists in peripheral immune cells and can reduce inflammation in the brain, implying an intrinsic molecular link to PD. Through the above mechanisms, the activity of cannabinoid receptors may influence the symptoms of PD to a large extent.

Due to the lipid signal transduction features of the ECS, the successful preparation of cannabinoid receptor radioligand requires moderate lipophilicity [66]. The first generation of imaging agents based on rimonabant are unsuitable for clinical application due to their high lipophilicity or poor BBB penetration. By continuously improving the lipophilicity of imaging agents, a series of second-generation CB1R PET agents have been successfully developed. [¹⁸F]MK-9470 is one of the most representative imaging agents and is widely used. As a potent reverse agonist, it has high selectivity (Ki = 0.7 nmol/L) and moderate lipophilicity (logD_7.4 = 4.7) and can accomplish a forceful but reversible binding to CB1Rs. It is known that CB1Rs are mainly distributed in the SN, globus pallidus, putamen, hippocampus, and cerebellum of the human brain. A study of [¹⁸F]MK-9470 imaging of the rhesus monkey brain shows that the binding site of this radioactive ligand is highly compatible with CB1Rs. However, in the process of non-invasive imaging in humans, the kinetics of [¹⁸F]MK-9470 are so slow that it takes 2 h to reach the plateau after injection [67]. There are also some other second-generation imaging agents, including [¹¹C]JHU75528 ([¹¹C]OMAR), [¹¹C]MePPEP, and [¹⁸F]FMPEP-d₂, that balance selectivity, affinity, and lipophilicity but have the limitations of elution speed and defluorination capacity. Recently, it has been reported that a novel CB1R radioligand [¹⁸F]FPATPP has high specificity, high stability, and low defluorination, but only one case has been studied in mice, and it still needs our continuous exploration [68].

Applications of Synaptic Dysfunction PET Imaging in PD

Diagnostic Performance

According to the latest diagnosis criteria for PD, clinical symptoms still play a leading role. The major pathological changes in PD are the degeneration and loss of dopaminergic neurons in the SN pars compacta. However, due to the substantial compensatory effect of PD in its initial stage, typical symptoms frequently manifest after the loss of ~50% of dopaminergic neurons, which greatly contributes to the diagnostic lag. Radionuclide imaging has proven to be a potent tool for PD diagnosis; it can recognize early changes from an internal molecular perspective and greatly improve the diagnosis rate.

The axons of dopaminergic neurons are the most vulnerable because they are the earliest target of αSyn accumulation in PD, thus, presynaptic dopaminergic molecular imaging is recommended here for diagnosis. The Movement Disorder Society also regards normal presynaptic dopaminergic PET images as the absolute exclusion criterion of PD [69]. Initially, the diagnosis of PD was made by [¹⁸F]DOPA, reflecting AADC activity, to assess dopaminergic synthesis. In a [¹⁸F]DOPA PET study, the caudate nucleus and anterior thalamus were found to be as accurate as 0.93 in the diagnosis of PD by analyzing the striatal subregions [70]. However, due to the compensatory increase of the synthesis and dopa decarboxylase activity of residual dopamine neurons, the level of dopamine does not decrease significantly in the early stage. Therefore, [¹⁸F]DOPA PET imaging only has good diagnostic value in patients with middle and advanced PD. In contrast, the diagnostic power of DAT imaging is particularly outstanding in early PD (Fig. 4). Recognized as the most sensitive molecular imaging marker of PD, it displays a significant reduction in uptake early and tends to further decrease due to compensatory mechanisms [71]. A study used [¹⁸F]FE-PE2I PET imaging to measure the availability of DAT. The loss of DAT was found to be significantly detectable in de novo PD patients, even before the appearance of clinical symptoms [72]. Another quantitative brain imaging study in PD patients suggested that the correlation between BP_ND and the standardized uptake value ratio in [¹⁸F]FE-PE2I PET images could already be satisfactory earlier on [73]. Recently, other dopaminergic presynaptic targets have appeared to have greater performance in PD. For example, VMAT2 is less susceptible to dopamine drugs and compensatory regulation than AADC and DAT, and it is almost always concentrated in dopaminergic axons [74]. These make VMAT2 a more objective target for presynaptic dopaminergic imaging. A PET study with [¹⁸F]FP-DTBZ to diagnose PD, and its blinded validation, showed a high clinical diagnostic compliance rate of 0.933, pointing out that targeted VMAT2 imaging is also of high application potential in effectively distinguishing PD from healthy individuals [75].

Fig. 4.

PET images of the dopamine transporter in PD. A [¹⁸F]FE-PE2I cross-sectional images of SRTM BP_ND and SUVR parameters, involving 2 reference areas of the cerebellum and occipital cortex, and 3 acquisition time windows. (a) a 52-year-old female with a healthy control. (b) a 67-year-old male PD patient (H-Y = 2), [¹⁸F]FE-PE2I uptake in the striatum is significantly reduced. Adapted with permission from Ref.[73] B Comparison of [¹⁸F]FP-CIT PET images in the striatum between a healthy control, an early-stage PD patient, and a late-stage PD patient. (a) healthy control, [¹⁸F]FP-CIT PET/MR image; (b) healthy control; (c) early-stage PD, [¹⁸F]FP-CIT uptake in the striatum is significantly reduced; (d) late-stage PD, as the disease progresses, [¹⁸F]FP-CIT uptake is further reduced. SRTM, simplified reference tissue model; BP_ND, binding potential; SUVR, standardized uptake value ratio; H–Y, Hoehn–Yahr stage. Adapted with permission from Ref.[104].

Differential Diagnosis of Parkinsonian Syndromes

Due to the significant overlap of clinical manifestations and mechanisms in Parkinsonian disorders, the differentiation of PD is also a link that cannot be ignored in the diagnostic process. Many PET scanning studies of dopamine synapses have demonstrated their high clinical value for the differential diagnosis of PD (Fig. 5). In a dual-phase PET/CT imaging study, [¹⁸F]FP-CIT was noted to achieve 91.7% specificity and 80% accuracy in differentiating idiopathic PD from atypical Parkinson's syndrome [76]. The semi-quantitative calculation of depth information from DAT PET images showed a significant difference between PD, multiple system atrophy (MSA), and progressive supranuclear palsy (PSP) (P <0.001) [77]. A comparative study found that in a series of Parkinsonian syndromes, a single [¹¹C]PE2I PET examination can effectively replace the dual examination of [¹²³I]FP-CIT SPECT and [¹⁸F]FDG PET in identifying them [78].

Fig. 5.

[¹⁸F]FP-CIT PET images for differential diagnosis of Non-PD, IPD, MSA-P, MSA-C, and PSP. A, B Axial and coronal slices of delayed phase [¹⁸F]FP-CIT PET. IPD patients show decreased metabolism of the bilateral dorsal posterior putamen and ventral putamen. For MSA-C and MSA-P, decreased metabolism is also found in the posterior putamen and ventral putamen. PSP patients were characterized by extensive reduction in DAT binding of the putamen and caudate nuclei. C In the early phase [¹⁸F]FP-CIT PET, IPD has normal or hypermetabolic striatal metabolism in the bilateral dorsolateral putamen. MSA shows hypometabolism in the cerebellar cortex or basal ganglia region, which can differentiate between MSA-C and MSA-P. PSP shows hypometabolism in the medial frontal cortices and midbrain. Non-PD, people without Parkinson’s disease; IPD, idiopathic Parkinson’s disease; MSA-C, multiple system atrophy-cerebellar type; MSA-P, multiple system atrophy-Parkinson type; PSP, progressive supranuclear palsy; DAT, dopamine transporter. Adapted with permission from Ref.[105].

In addition, screening for non-dopaminergic alterations can also provide additional information for the diagnosis of PD. A [¹¹C]DASB PET study targeting SERT successfully explored the imaging differences between PD and MSA [79]. Compared to PD, MSA patients have more severe serotonergic deficits in the brainstem and some subcortical regions. Other studies have found that PD and PSP can be distinguished by quantifying AChE density by [¹¹C]PMP PET scanning. This method revealed a decrease in cholinergic AChE activity in the cerebral cortex in PD patients, while the decline is more remarkable in the subcortical region, such as the thalamus, in PSP and MSA patients [80, 81].

Efficacy Detection and Prognosis Evaluation

In terms of monitoring treatment changes and predicting prognosis, synaptic PET imaging also has tremendous advantages. In a group of early and stable PD patients, Jessia et al. synthetically analyzed dopaminergic and serotonergic states in combination with [¹¹C]DTBZ, [¹¹C]methylphenidate, [¹¹C]DASB, and [¹¹C]raclopride. The results showed a significant correlation with the degree of risk of developing dyskinesia after 3 years of levodopa treatment [51]. In another longitudinal trial, [¹⁸F]FE-PE2I PET was performed in non-advanced PD patients following 2 years of treatment, and little change was found in DAT availability in the SN, suggesting slow progression or compensatory changes [82]. Based on the monitored treatment in real-time, we can also adjust and optimize therapy regimens promptly. A PET study targeting VMAT2 by [¹¹C]DTBZ suggested that the synaptic density had a determination on the decay time of the levodopa response, so that PD patients with a lower synaptic density may necessitate alternative or more aggressive therapies [83]. These precise and prompt assessments by PET molecular imaging of synaptic function will greatly benefit individuals with PD.

Exploration of Mechanism and Heterogeneity

PD is a complex and long-progressive disease with high heterogeneity that is affected by clinical, pathological, genetic, and other factors. Different individuals or stages lead to differences in phenotype and prognosis. Pathologically, these manifestations depend on the damage to specific brain regions caused by the abnormal deposition of αSyn. However, although a specific probe for αSyn has been developed recently, it is still in the preclinical trial stage [16]. In order to better understand intrinsic pathology and stratify PD patients, imaging of different kinds of synaptic function provides excellent options, which allows us to visualize the pathological condition more comprehensively.

In the Dutch Parkinson cohort study protocol, cholinergic and dopaminergic PET imaging were involved as one of the criteria for classifying PD subtypes [84]. AChE-related cholinergic PET imaging is commonly used to assess symptoms associated with visceral parasympathetic nervous system disorders. Studies have shown that [¹¹C]donepezil binding is significantly reduced in the small intestine and pancreas in PD patients with symptoms such as constipation and gastroparesis [85]. This method provides feasibility for studying the temporal and spatial heterogeneity in PD. Knudsen et al. combined [¹¹C]donepezil, [¹¹C]MeNER, and [¹⁸F]DOPA PET to evaluate the Braak I, II, and III stages of PD, respectively [86]. Another study also revealed two types of body priority and brain priority in PD through cholinergic PET imaging [87]. Molecular imaging based on different synapses brings great convenience to exploring the whole picture and heterogeneity of PD pathology.

As the premise for in-depth analysis of all the above applications, quantitative PET molecular imaging should not be neglected; it helps to establish the direct relationship between imaging parameters and biofunctional indicators. Manual piece-by-piece analysis is a typical and traditional strategy. However, this method is time-consuming and laborious and depends on the operator’s judgment. By contrast, semi-automated or fully-automated analysis methods serve as commendable alternatives. Many comparative studies have demonstrated that the automatic definition of regions of interest has lower test-retest variability and better reliability than manual methods [88]. Nevertheless, in these automated approaches, special attention should be paid to target area conditions and the preprocessing steps. An inappropriate situation or treatment will lead to a deviation in the results. Besides, in constructing classification or prediction models, fully automated feature extraction by deep learning can achieve performance comparable to or superior to traditional radionics [89, 90]. However, this advantage is always based on the large scale of datasets, which severely limits the practical application of fully automatic modeling. Nowadays, multi-center collaborative research has become a trend. However, while sharing imaging data, we also need to consider several problems of data privacy and security. All in all, weighing convenience and accuracy, low data costs, optimal performance, and expanding sample centers under compliance remain worth thinking about, especially in the era of big data.

In addition, there are some limitations of PET imaging in PD applications. Firstly, PET technology has an obvious inherent defect in its low spatial resolution, which is not conducive to recognizing some subtle anatomical structures. Secondly, due to the short half-life (T_1/2), most radiopharmaceuticals, such as ¹¹C (T_1/2 = 20 min), ¹³N (T_{1/2 =} 10 min), and ¹⁵O (T_1/2 = 2 min), must rely on an internal cyclotron for immediate production, which greatly limits their availability. In contrast, ¹⁸F has a long imaging window (T_1/2 = 110 min) and convenient synthesis conditions that can be readily supplied by commercial radiopharmacies. Nevertheless, some targets still lack radiotracers labeled with ¹⁸F or other isotopes possessing a long T_1/2. Finally, synaptic dysfunction in PD is fundamentally attributed to the misfolding and deposition of αSyn, but the development of molecular probes related to αSyn is currently not adequate for clinical trials.

Conclusions and Future Perspectives

PD is the second-most common neurodegenerative disease after Alzheimer's disease, with complex pathological mechanisms that are closely associated with different targets on synapses. It has typical motor symptoms, accompanied by some non-motor symptoms. However, there is often a latent period in the clinical diagnosis based on symptoms, which makes it impossible for doctors to make a timely diagnosis. PD not only affects the quality of life of a large number of patients but also imposes a heavy burden on society in many aspects. Therefore, it is important to make an early diagnosis and risk assessment of PD so as to further develop treatment strategies and greatly improve the prognosis.

PET is an excellent molecular imaging technique that enables non-invasive, dynamic visualization of metabolic and functional changes in vivo [91]. Given the succession of pathological events associated with synaptic dysfunction in PD, we should focus on various crucial molecular imaging targets, comprising not only αSyn but also the dopamine and non-dopamine systems. PET imaging of these biomarkers will be helpful for the early diagnosis of PD and is also of great importance in its treatment and prognostic evaluation.

In summary, nuclear molecular imaging technology has great potential in PD. Our research direction should focus on developing αSyn-specific probes to delve into PD fundamentally. We must continue to enrich novel PD-related synaptic function imaging tracers, especially those with a long T_1/2. Moreover, the use of multimodal imaging in PD applications should be increased to address the intrinsic weaknesses of various imaging techniques and to acquire diverse dimensional data from images. In the context of the big data era, we should actively summarize the experience of PET quantification, look for a balance between data cost and performance, and safeguard data privacy. In the future, further studies, including in vitro screening and in vivo verification, are needed in this field in order to accelerate clinical application and find more possibilities for PET applications in PD and even other kinds of neurodegenerative disorders, so that patients can receive more benefits quickly.

Conflict of interest

The authors declare no competing interests.