Resynchronization of microglial activity in the brain is associated with restoration of motor function in Parkinson’s disease

Abstract

Neuroinflammation is a key factor in Parkinson’s disease (PD) pathogenesis. However, the regional heterogeneity of biomarkers related to inflammation in PD is less well defined. We developed [¹⁸F]GSK PET imaging to quantify neuroinflammation via the P2X7 receptor (P2X7R) in A53T PD male mice and wild-type (WT) male mice. Montelukast (MK) was administered to mice, and weekly behavior tests confirmed MK’s efficacy. [¹⁸F]L-DOPA/[¹⁸F]GSK PET, motor testing, autoradiography, and immunofluorescence were performed after MK treatments. MK improved motor function and reduced the brain uptake of [¹⁸F]GSK, indicating resynchronization of regional microglial activity. The whole brain uptake of [¹⁸F]GSK was correlated with motor functional restoration, while [¹⁸F]L-DOPA PET was not. Overall, our study indicated that brain mapping of [¹⁸F]GSK PET is beneficial for exploring P2X7R-related neuroinflammation, which is correspondent to motor function in PD.

Brain PET-MRI imaging shows higher [¹⁸F]GSK uptake across multiple regions in A53T compared to WT mice. Correlation analyses reveal that resynchronization of microglia in Parkinson’s disease is associated with motor function restoration.

Introduction

Parkinson’s disease (PD) is one of the most common neurodegenerative diseases, affecting more than 6 million patients worldwide¹. Patients with PD suffer from motor symptoms, which seriously affect their quality of life. Therefore, optimal treatments for PD are urgently needed. The pathology of PD is characterized by widespread aggregation of α-synuclein (α-syn) in Lewy bodies (LBs), damage to dopaminergic neurons (DA neurons), and neuroinflammation^2,3. However, the development of diagnostic and therapeutic agents for aggregated α-syn and damaged DA neurons is currently limited. The heterogeneity of α-syn hinders the development of ligands⁴. Treatments targeting DA neurons are not effective once the honeymoon period has elapsed⁵.

As a major pathological feature of PD, neuroinflammation plays a prominent role in its pathogenesis⁶. Inflammation may lead to metabolic and functional deficits in various brain regions, resulting in different motor symptoms and cognitive deficits^7,8. Microglial activation is the key element of neuroinflammation. Activated microglia might have negative influences on the progression of PD⁹. Microglia, the resident immune cells of the central nervous system (CNS), are pivotal mediators of neuroinflammation in PD. These cells dynamically shift between pro-inflammatory (M1) and anti-inflammatory (M2) phenotypes, with chronic M1 activation driving neuronal damage through excessive release of cytokines (e.g., TNF-α, IL-1β) and reactive oxygen species (ROS)¹⁰. In PD, activated microglia cluster around LBs and degenerating DA neurons, perpetuating α-syn pathology and accelerating disease progression¹¹. BV2 cells, a mouse microglial cell line, are widely used as an in vitro model for studying microglial activation and neuroinflammation¹². They can be stimulated with lipopolysaccharide (LPS) to adopt a pro-inflammatory state, mimicking the inflammatory response seen in neurodegenerative diseases¹². Previous studies have shown regional heterogeneity in the expression of biomarkers related to inflammation⁷. For instance, Pintado et al. revealed that LPS-induced neuroinflammatory cascades display regional divergences, with microglial susceptibility to LPS stimulation differing substantially across the substantia nigra and striatum¹³. Additionally, Ji et al. highlighted spatial disparities in neutrophil infiltration, showing preferential accumulation in the substantia nigra pars compacta (SNpc) compared to cortical regions in neurodegenerative models¹⁴. These collective findings underscore the brain region-specific distribution and activation states of inflammatory cells. Regional positron emission tomography (PET) imaging enables dynamic monitoring of these spatiotemporal variations, providing non-invasive assessment of localized pathological changes through spatially resolved imaging. This approach enhances precision in mapping neuroinflammatory trajectories, thereby improving the prediction of disease progression trajectories and therapeutic outcomes.

The P2X7 receptor (P2X7R) is expressed mainly in activated microglia and, to a lesser extent, in neurons and astrocytes¹⁵. Under normal physiological conditions, P2X7R is considered a “silent transporter” but can be activated by high levels of extracellular ATP (>0.1 mM) under pathological conditions in the CNS^16–18. The expression of P2X7R is upregulated in the brains of PD patients and in 6-hydroxydopamine-lesioned rats^19,20. PET is a noninvasive imaging tool that reflects receptor expression levels in vivo via radiolabeled ligands. [³H]A804598 was a P2X7R tracer reported to be able to cross the blood‒brain barrier (BBB), but it has been limited by high levels of nonspecific binding¹⁶. Janssen et al. developed P2X7R-targeting tracers, including [¹¹C]SMW139²¹. [¹¹C]SMW139 has been validated for its specific binding to P2X7R in mouse model PET studies²². However, the [¹¹C]SMW139 is metabolized fast and to a greater extent in female than male wild-type (WT) mice²². GSK1482160 (GSK in short) has demonstrated robust binding affinity for the P2X7R. The carbon-11-labeled analog, [¹¹C]GSK, has been shown to exhibit high affinity for brain P2X7R, with a reported dissociation constant (Kd) of 1.15 ± 0.12 nM^23,24. This radioligand has been successfully employed in PET studies, including in LPS-induced mouse models²⁴ and experimental autoimmune encephalomyelitis (EAE) rat models²⁵, further supporting its potential as a P2X7R tracer. To address the short physical half-life of carbon-11, we synthesized a fluorine-18-labeled analog, [¹⁸F]GSK^16,26. Preclinical evaluations in healthy nonhuman primates (NHPs) using PET imaging have confirmed that [¹⁸F]GSK is a promising radioligand for clinical applications in P2X7R imaging^16,18,26. Additionally, [¹⁸F]GSK exhibits several advantageous properties, including the ability to cross the BBB, a high radiochemical yield (RCY of 14.05–16.42%), and excellent binding affinity (IC₅₀ = 2.12 ± 0.06 nM)²⁷.These characteristics position [¹⁸F]GSK as a strong candidate for translational research and clinical use in P2X7R imaging. Although [¹⁸F]GSK PET has demonstrated elevated neuroinflammation in Alzheimer’s disease (AD) transgenic mice²⁸, there have been no studies on [¹⁸F]GSK PET in the brains of A53T mice.

In recent years, drugs that suppress neuroinflammation have received increasing amounts of attention²⁹. Minotetracycline³⁰ and acetaminophen³¹ have been tested in clinical trials; however, they did not result in significant improvements in PD symptoms. Montelukast (MK), a potent antagonist of cysteinyl leukotriene receptor 1 (CysLTR1), is routinely used in the management of asthma and allergic rhinitis. MK treatment of AD provided novel modes of action such as the modulation of microglia phenotypes in the diseased brain³². MK exhibited a significant neuroprotective effect by reducing the proliferation of microglia and astrocytes, preserving regional brain metabolism and metabolic connectivity from damage caused by quinolinic acid (QA) in Huntington’s disease (HD)³³. Additionally, MK may directly modulate microglia since CysLTR1 is expressed on microglia³⁴. MK promotes M2 subtype polarization in BV2 microglia and ameliorates symptoms³⁴. Recently, studies have reported that MK may improve motor deficits and reduce the incidence of PD³⁵. MK can cross the BBB³⁶. However, neuroinflammatory activity, particularly microglial activation within inflammatory foci, induces BBB integrity disruption and subsequent permeability elevation³⁷. This phenomenon enhances anti-inflammatory drug accessibility to lesioned areas while simultaneously altering their spatiotemporal distribution patterns across brain regions³⁸. Consequently, such regional pharmacokinetic variations may underlie divergent therapeutic outcomes in distinct neural circuits. Nevertheless, systematic investigations remain scarce regarding two key aspects: (1) in vivo imaging-based evaluation of anti-inflammatory drug biodistribution, and (2) spatial mapping of MK’s modulatory effects on region-specific neuroinflammation in PD pathogenesis. Moreover, conventional evaluation methodologies remain inadequate, and there currently exists a paucity of quantitative approaches for assessing anti-inflammatory therapies in clinical practice.

Although existing studies have provided important clues linking neuroinflammation to motor symptoms in PD, their relationship remains a subject of considerable debate. Ferrari et al. demonstrated that overexpression of IL-1β in the substantia nigra correlates with motor symptoms in rat models³⁹. However, Lavisse et al. conducted detailed neuroinflammatory assessments in PD patients using PET targeting the 18 kDa Translocator Protein (TSPO) in microglia⁴⁰. Their findings revealed no significant correlation between microglial activation levels and either motor symptom severity or disease duration⁴⁰. Moreover, PD-associated neuroinflammation extends beyond classical pathology-rich regions like substantia nigra and striatum⁴¹. Nevertheless, how neuroinflammatory processes in distinct brain regions specifically contribute to motor manifestations remains poorly understood, underscoring the critical need to unravel the complex spatiotemporal interplay between neuroinflammation and PD motor symptomatology.

In this study, we performed PET imaging and quantification to assess the neurodegenerative brain regions involved in PD and the anti-inflammatory response of MK. We performed [¹⁸F]L-DOPA PET scans (for DA neurons), motor testing, [¹⁸F]GSK PET scans, autoradiography and immunofluorescence staining after 5 weeks of MK treatment in transgenic A53T PD mice and WT mice. The relationships between brain regions and different behavior tests were also investigated. For the first time, a correlation was found between increased P2X7R expression in various brain regions and different motor symptoms. [¹⁸F]GSK PET imaging delineates regional disparities in intracranial inflammation, facilitating the evaluation of disease progression and responses to anti-inflammatory therapies.

Results

Radiochemical quality for brain PET imaging

In this study, [¹⁸F]GSK was prepared via a manual or automatic synthesis module following our previous protocol (Fig. 1a)^16,27. Gamma spectrometry was performed 12 hours after radiochemical labeling, which confirmed the energy spectrum (511 kev) of the fluorine-18 nuclide. The total RCY was 14.05–16.42% (decay corrected to the end of synthesis), the molar activity was 6.8–23.62 GBq/μmol (n = 10, corrected for decay at the end of synthesis), and the radiochemical purity was > 98%. The raw product was further separated and purified by high-performance liquid chromatography (HPLC) (Fig. 1b), and [¹⁸F]GSK was collected from 8.1 min to 13.5 min in 35% acetonitrile at 4 mL/min (Fig. 1b). The radio HPLC peak of [¹⁸F]GSK ranged from 13 min to 14 min in the presence of 35% acetonitrile at 1 mL/min (Fig. 1c, d). Quality control of [¹⁸F]GSK injection was carried out, as shown in Fig. 1c, which revealed that the product of [¹⁸F]GSK satisfied the quality requirements for PET imaging.

Fig. 1. Radiosynthesis workflows of [¹⁸F]GSK and experimental design.

a Radiosynthesis of [¹⁸F]GSK. The precursor reacted at 120 °C to generate the crude product, which was purified after semi preparation and confirmed to be the target product after quality control and co-injection by HPLC. b Semi preparation results of [¹⁸F]GSK. The [¹⁸F]GSK fraction was eluted and collected between 8.1 and 13.5 minutes using 35% acetonitrile as the mobile phase, delivered at a flow rate of 4 mL/min. c Quality control results of [¹⁸F]GSK. The quality control analysis of [¹⁸F]GSK exhibited a distinct peak eluting between 13 and 14 minutes under chromatographic conditions of 35% acetonitrile and a flow rate of 1 mL/min. d Co-injection results of [¹⁸F]GSK. The co-injection analysis of [¹⁸F]GSK exhibited a distinct peak eluting between 13 and 14 minutes under chromatographic conditions of 35% acetonitrile and a flow rate of 1 mL/min.

Motor behaviors were improved in A53T/MK mice

The grip tests were performed to measure the maximum force of specific limbs (forelimbs), while the pole tests assessed basal ganglia-related motor coordination. The inverted screen tests were performed to evaluate whole-body muscular endurance, particularly measuring persistence in inverted postures. The illustrations of the motor tests were shown in Fig. S1. The number of the mice was shown in Table S1.

Throughout the treatment process, weekly motor tests were conducted to assess the efficacy of MK. In the 1^st week of treatment, the forelimb muscle strength of A53T mice was weaker than that of WT mice (A53T: 0.565 ± 0.066 N vs. WT: 0.718 ± 0.098 N, adjusted p = 0.0261) (Fig. 2a, Table S2). However, the forelimb strength in A53T mice was improved by MK after two weeks. In the 3^rd week, the grip strength of the A53T/MK mice increased (0.666 ± 0.047 N). Moreover, in the 5^th week, the forelimb muscle strength of the A53T/MK mice further increased, which was significantly different from that of the A53T mice (A53T: 0.630 ± 0.049 N vs. A53T/MK: 0.736 ± 0.027 N, adjusted p = 0.0003) (Table 1).

Fig. 2. Motor behaviors were improved by MK treatment in A53T mice.

a Quantitative results of the grip tests. Two-way ANOVA and Tukey’s multiple comparisons test were performed in each week. Statistical significance between A53T and WT groups (adjusted p < 0.05) was denoted by black asterisks, while significant differences between A53T and A53T/MK mice (adjusted p < 0.05) were marked with red asterisks. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 were considered significantly different. b, c T-Turn and T-Descend of the pole tests over five weeks (5^th week: T-Turn: A53T: 60.208 ± 45.035; A53T/MK: 2.917 ± 0.376; WT: 3.208 ± 0.954; WT/MK: 3.000 ± 0.894) (5^th week: T-T-Descend: A53T: 63.833 ± 17.449; A53T/MK: 7.597 ± 0.570; WT: 10.458 ± 0.845; WT/MK: 9.208 ± 0.778). Two-way ANOVA and Tukey’s multiple comparisons test were performed in each week. Statistical significance between A53T and WT groups (adjusted p < 0.05) was denoted by black asterisks, while significant differences between A53T and A53T/MK mice (adjusted p < 0.05) were marked with red asterisks. *p < 0.05, and **p < 0.01 were considered significantly different. d–h Quantitative results of the inverted screen tests over five weeks. The decreasing trend of endurance in the A53T mice was greater than that in the A53T/MK, WT, and WT/MK mice. i Decline ratio of the latency time during five weeks of MK treatment. In the first week of treatment, a significant difference in the decline ratio was observed between the A53T mice and the WT (A53T: 14.346 ± 6.858 vs. WT: 2.345 ± 0.811, adjusted p = 0.0317, gray asterisks) and WT/MK mice (A53T: 14.346 ± 6.858 vs. WT/MK: 2.724 ± 1.192, adjusted p = 0.0387, black asterisks). However, no significant difference was found between the A53T mice and the A53T/MK mice (A53T: 14.346 ± 6.858 vs. A53T/MK: 16.254 ± 11.861, adjusted p = 0.9627). After five weeks of MK treatment, the decline ratio of the A53T mice showed a significant difference compared to that of the A53T/MK mice (A53T: 11.538 ± 4.473 vs. A53T/MK: 5.445 ± 3.016, adjusted p = 0.0136, red asterisks). Two-way ANOVA and Tukey’s multiple comparisons test were performed. *p < 0.05, **p < 0.01, and ***p < 0.001 were considered significantly different.

Table 1.

Quantification data and p values of the motor test results after 5 weeks of MK treatment

Motor tests	Measurement	Mean ± SD				Adjusted p value
		A53T (n = 6)	A53T/MK (n = 6)	WT (n = 6)	WT/MK (n = 6)	A53T vs. A53T/MK	A53T vs. WT
Grip tests	Grip strength (N)	0.630 ± 0.049	0.736 ± 0.027	0.693 ± 0.030	0.729 ± 0.033	0.0294*	0.0003***
Pole tests	T-Turn (s)	60.208 ± 45.035	2.917 ± 0.376	3.208 ± 0.954	3.000 ± 0.894	0.0015**	0.0014**
	T-Descend (s)	63.833 ± 42.740	7.597 ± 1.396	10.458 ± 2.070	9.208 ± 1.907	0.0018**	0.0010**
Inverted screen tests	Decline ratio	11.538 ± 4.473	5.445 ± 3.016	4.061 ± 2.110	3.523 ± 0.974	0.0015**	0.0136*

The data are shown as the means ± SD. Two-way ANOVA and Tukey’s multiple comparisons test was performed. *p < 0.05, **p < 0.01 and ***p < 0.001 were considered significantly different.

In the 1^st week of treatment, the A53T mice showed poorer motor coordination in comparison to the WT mice (Fig. 2b, c, Tables S3-S4). WT mice flexibly turned (5.361 ± 1.497 s) and descended (12.958 ± 2.580 s) on a climbing pole. A53T mice were unable to turn smoothly and were afraid to descend, requiring a significantly longer latency to turn over (38.042 ± 41.383 s) and descend (40.333 ± 40.037 s). MK treatment restored the motor coordination of A53T/MK mice. In the 2^nd week, the A53T/MK mice exhibited improved coordination and a faster speed of turning over (A53T: 54.722 ± 42.044 s vs. A53T/MK: 11.000 ± 11.382 s, adjusted p = 0.0119) and descending (A53T: 57.319 ± 39.954 s vs. A53T/MK: 16.264 ± 12.221 s, adjusted p = 0.0144). Continued improvement in motor behaviors in the A53T/MK mice were also observed during subsequent MK treatment (Fig. 2b, c, Table 1).

Compared with WT mice, A53T mice had reduced muscular endurance in inverted screen tests (Fig. 2d–i, Tables S5-S12). The latency time of the A53T mice tended to decrease faster, and the descending ratio of the A53T mice (14.346 ± 6.858) and A53T/MK mice (16.254 ± 11.861) was greater than that of the WT mice (2.345 ± 0.811) in the 1^st week, indicating poor endurance of the A53T mice. The descending ratio of A53T mice also exhibited a similar trend (2^nd week: 13.728 ± 16.929; 3^rd week: 12.500 ± 8.337; 4^th week: 12.507 ± 9.230; 5^th week: 11.538 ± 4.473), whereas the descending ratio of A53T/MK mice decreased in subsequent weeks of MK treatment (2^nd week: 6.887 ± 4.565; 3^rd week: 5.140 ± 2.306; 4^th week: 4.664 ± 1.582; 5^th week: 5.445 ± 3.016), implying that muscular endurance decay was slowed by MK treatment (Fig. 2i, Table 1, Table S10).

[¹⁸F]GSK PET and autoradiography revealed reduced inflammation in A53T/MK mice

Compared with those in the brains of WT mice, increased [¹⁸F]GSK uptake was observed in most brain regions of A53T mice, including the cortex, basal forebrain septum, striatum, hypothalamus, amygdala, hippocampus, midbrain, and whole brain (Fig. 3, Table 2). In addition, most of these brain regions were significantly lower in the A53T/MK group than in the A53T group. [¹⁸F]GSK autoradiography is shown in Fig. 4, Table S13. Compared with WT mice, A53T mice presented increased [¹⁸F]GSK uptake in the cortex, basal forebrain septum, striatum, thalamus, hypothalamus, amygdala, hippocampus, midbrain, and brainstem. Moreover, the autoradiography results were consistent with the PET results, which revealed that brain [¹⁸F]GSK uptake was reduced in most brain regions of the A53T/MK group mice.

Fig. 3. Representative [¹⁸F]GSK PET-MRI images and quantification.

a Representative images of [¹⁸F]GSK PET-MRI in four groups of mice (each group n = 6). b–m Quantitative results of [¹⁸F]GSK PET-MRI. [¹⁸F]GSK PET revealed that neuroinflammation was significantly increased in most brain regions (cortex, basal forebrain septum, striatum, thalamus, hypothalamus, amygdala, hippocampus, midbrain, brainstem and whole brain) in A53T mice compared to A53T/MK mice (red asterisks), WT mice (gray asterisks) or WT/MK mice (black asterisks). Two-way ANOVA and Tukey’s multiple comparisons test were performed. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 were considered significantly different.

Table 2.

[¹⁸F]GSK SUV_R^LV (mean ± SD) and adjusted p value in the brain regions after MK treatment

	SUVRLV (Mean ± SD)				Adjusted p value
Brain regions	A53T (n = 6)	A53T/MK (n = 6)	WT (n = 6)	WT/MK (n = 6)	A53T vs. A53T/MK	A53T vs. WT
Whole brain	1.639 ± 0.115	1.437 ± 0.070	1.401 ± 0.068	1.387 ± 0.153	0.0191*	0.0053**
Amygdala	1.905 ± 0.234	1.549 ± 0.062	1.538 ± 0.166	1.525 ± 0.183	0.0096**	0.0074**
Basal forebrain septum	1.727 ± 0.123	1.443 ± 0.131	1.433 ± 0.088	1.480 ± 0.204	0.0128*	0.0097**
Brainstem	1.815 ± 0.142	1.676 ± 0.175	1.598 ± 0.070	1.528 ± 0.225	0.4699	0.1308
Cortex	1.614 ± 0.138	1.338 ± 0.083	1.310 ± 0.078	1.329 ± 0.145	0.0026**	0.0010***
Hippocampus	1.525 ± 0.139	1.241 ± 0.149	1.244 ± 0.081	1.295 ± 0.153	0.0076**	0.0083**
Hypothalamus	1.796 ± 0.185	1.531 ± 0.062	1.501 ± 0.075	1.408 ± 0.093	0.0041**	0.0015**
Inferior colliculi	1.324 ± 0.242	1.256 ± 0.131	1.243 ± 0.226	1.242 ± 0.182	0.9356	0.8945
Midbrain	1.548 ± 0.165	1.337 ± 0.175	1.284 ± 0.067	1.314 ± 0.185	0.1615	0.0472*
Superior colliculi	1.362 ± 0.184	1.260 ± 0.230	1.119 ± 0.144	1.109 ± 0.185	0.9297	0.2009
Striatum	1.616 ± 0.139	1.328 ± 0.154	1.306 ± 0.063	1.291 ± 0.084	0.0018**	0.0009***
Thalamus	1.410 ± 0.133	1.264 ± 0.103	1.267 ± 0.064	1.202 ± 0.131	0.1390	0.1496

The data are shown as the means ± SD. Two-way ANOVA and Tukey’s multiple comparisons test was performed. *p < 0.05, **p < 0.01, and ***p < 0.001 were considered significantly different.

Fig. 4. Ex vivo autoradiography of [¹⁸F]GSK.

a Autoradiography images after MK treatment. b–k Increased [¹⁸F]GSK uptake was observed in the cortex, basal forebrain septum, striatum, thalamus, hypothalamus, amygdala, hippocampus, midbrain, and brainstem of A53T mice compared with A53T/MK mice (red asterisks), WT mice (gray asterisks), and WT/MK mice (black asterisks). Two-way ANOVA and Tukey’s multiple comparisons test was performed. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 were considered significantly different.

[¹⁸F]L-DOPA PET revealed that it partially ameliorated DA dysfunction in A53T/MK mice

Representative [¹⁸F]L-DOPA PET- Magnetic resonance imaging (MRI) images and analysis results are displayed in Fig. S2 and Tables S14, S15. The results of [¹⁸F]L-DOPA PET-MRI revealed impaired dopamine metabolism in A53T mice. Compared with WT mice, there was decreased DA synthesis observed in the cortex, basal forebrain septum, striatum, superior colliculi, and whole brain in the brains of A53T mice. Moreover, DA neuron function was improved in A53T/MK mice. Compared with A53T mice, the increased [¹⁸F]L-DOPA uptake was observed in the basal forebrain septum, striatum, and superior colliculi in the brains of A53T/MK mice.

Immunofluorescence confirmed deactivated microglia markers by MK

The double immunofluorescence results for IBA1 and P2X7R are shown in Fig. 5, Tables S16, 17. Colocalization staining revealed that IBA1 (green) and P2X7R (red) were expressed on the microglia of the cortex. The morphology of the activated microglia was round or rod-shaped, with an enlarged soma size and shorter protrusions. A decrease in activated microglia was observed in the striatum of A53T/MK, WT, and WT/MK mice compared to A53T mice. The IBA1 and P2X7R immunofluorescence intensities in the brains of A53T mice were greater than those in the brains of A53T/MK, WT, and WT/MK mice.

Fig. 5. Immunofluorescence staining of the cortex revealed the colocalization of IBA1 and P2X7R.

a Immunofluorescence staining was performed to analyze the colocalization of IBA1 (green), P2X7R (red), and DAPI (blue) in the cortex of the four groups of mice. Increased numbers of activated microglia (white arrows) were observed in A53T mice compared with A53T/MK, WT, and WT/MK mice. b Quantification of the fluorescence intensity of DAPI. c Quantification of the fluorescence intensity of IBA1. The IBA1 fluorescence intensity was greater in A53T mice than in A53T/MK (red star), WT (gray star), and WT/MK (black star) mice. d Quantification of the fluorescence intensity of P2X7R. Increased P2X7R fluorescence intensity in A53T mice was observed. Two-way ANOVA and Tukey’s multiple comparisons test was performed. *p < 0.05, and **p < 0.01 were considered significantly different.

Correlations between brain PET imaging and motor symptoms

Since the brain regions related to different motor tests may differ, we focused on the relationships between motor test results and different brain regions. Pearson correlation analysis was performed between [¹⁸F]GSK and the motor test results (Figs. 6–8), as well as between [¹⁸F]L-DOPA and the motor test results (Figs. 6–8). The brain regions with high [¹⁸F]L-DOPA correlations differed among the three behavioral tests. For the grip tests, [¹⁸F]L-DOPA uptake in the striatum, basal forebrain septum and the whole brain was negatively correlated with muscle strength disorders (1-Grip strength) (Fig. 6). For the pole tests, [¹⁸F]L-DOPA uptake in four brain regions (striatum, basal forebrain septum, midbrain, and whole brain) was negatively correlated with motor coordination disorders (T-Turn) (Fig. 7). For the inverted screen tests, [¹⁸F]L-DOPA uptake in five brain regions (whole brain, hypothalamus, basal forebrain septum, cortex, and striatum) was negatively correlated with endurance disorders (Decling ratio) (Fig. 8). The striatum and basal forebrain septum were overlapping brain region with a high [¹⁸F]L-DOPA correlation.

Fig. 6. Correlation analysis between muscle strength disorder and the PET results.

a Distribution of brain regions with strong correlations between muscle strength disorder (1-Grip strength) and neuroinflammation in the mouse brain. Brain regions with r values greater than 0.4 and p values less than 0.05 are colored purple. b, c The r values and p values of the correlation analysis between grip strength disorder and [¹⁸F]GSK SUV_R^LV in brain regions. d Correlation analysis between grip strength disorder and [¹⁸F]GSK SUV_R^LV in various brain regions. [¹⁸F]GSK uptake is correlated with muscle strength disorder in most brain regions. e. Distribution of brain regions with strong correlations between muscle strength disorder (1-Grip strength) and [¹⁸F]L-DOPA uptake in the mouse brain. Brain regions with r values less than -0.4 and p values less than 0.05 are colored orange. f, g The r values and p values of the correlation analysis between grip strength disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. h Correlation analysis between grip strength disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. [¹⁸F]L-DOPA uptake in the striatum, basal forebrain septum and the whole brain is correlated with muscle strength. Keywords: BS brainstem, CB cerebellum, CTX cortex, HIP hippocampus, HYT hypothalamus, IC inferior colliculi, MB midbrain, STR striatum, THA thalamus.

Fig. 8. Correlation analysis between muscle endurance disorder and the PET results.

a Distribution of brain regions with strong correlations between muscle endurance disorder (Decling ratio) and neuroinflammation in the mouse brain. Brain regions with r values greater than 0.4 and p values less than 0.05 are colored purple. b, c The r values and p values of the correlation analysis between muscle endurance disorder and [¹⁸F]GSK SUV_R^LV in brain regions. d Correlation analysis between muscle endurance disorder and [¹⁸F]GSK SUV_R^LV in brain regions. [¹⁸F]GSK uptake is correlated with muscle endurance disorder in six brain regions. e Distribution of brain regions with strong correlations between muscle endurance disorder (Decling ratio) and [¹⁸F]L-DOPA uptake in the mouse brain. Brain regions with r values less than -0.4 and p values less than 0.05 are colored orange. f, g The r values and p values of the correlation analysis between muscle endurance disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. h Correlation analysis between muscle endurance disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. [¹⁸F]L-DOPA uptake is correlated with muscle endurance disorder in the whole brain, hypothalamus, basal forebrain septum, cortex, and striatum. Keywords: BS brainstem, CB cerebellum, CTX cortex, HIP hippocampus, HYT hypothalamus, IC inferior colliculi, MB midbrain, STR striatum, THA thalamus.

Fig. 7. Correlation analysis between motor coordination disorder and the PET results.

a Distribution of brain regions with strong correlations between motor coordination disorder (T-Turn) and neuroinflammation in the mouse brain. Brain regions with r values greater than 0.4 and p values less than 0.05 are colored purple. b, c The r values and p values of the correlation analysis between motor coordination disorder and [¹⁸F]GSK SUV_R^LV in brain regions. d Correlation analysis between motor coordination disorder and [¹⁸F]GSK SUV_R^LV in brain regions. [¹⁸F]GSK uptake is correlated with motor coordination disorder in most brain regions. e Distribution of brain regions with strong correlations between motor coordination disorder (T-Turn) and [¹⁸F]L-DOPA uptake in the mouse brain. Brain regions with r values less than -0.4 and p values less than 0.05 are colored orange. f, g The r values and p values of the correlation analysis between motor coordination disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. h Correlation analysis between motor coordination disorder and [¹⁸F]L-DOPA SUV_R^CB in brain regions. [¹⁸F]L-DOPA uptake is correlated with motor coordination disorder in striatum, basal forebrain septum, midbrain, and the whole brian. Keywords: BS: brainstem; CB: cerebellum; CTX: cortex; HIP: hippocampus; HYT: hypothalamus; IC: inferior colliculi; MB: midbrain; STR: striatum; THA: thalamus.

Moreover, although the high [¹⁸F]GSK correlation brain regions for the three behavioral tests mostly overlap, the highest correlation brain regions for inflammation were different (Figs. 6–8). For muscle strength testing, the cortex, hippocampus, and striatum were the brain regions with the highest [¹⁸F]GSK correlations (Fig. 6). The averaged whole brain uptake of [¹⁸F]GSK was correlated to grip tests (p = 0.0005, r = 0.6594). For the pole tests, the cortex, midbrain, and striatum were the brain regions with the highest [¹⁸F]GSK correlations (Fig. 7). The averaged whole brain uptake of [¹⁸F]GSK was correlated to pole tests (p = 0.0031, r = 0.5777). For the inverted screen tests, the hypothalamus, basal forebrain septum, and striatum were the brain regions most strongly correlated with [¹⁸F]GSK (Fig. 8). The averaged whole brain uptake of [¹⁸F]GSK was correlated to inverted screen tests (p = 0.0453, r = 0.4212). In addition, [¹⁸F]GSK high-correlation brain regions outnumbered and contained most of the [¹⁸F]L-DOPA high-correlation brain regions.

Discussion

In this study, we assessed the disease progression and efficacy of MK in various brain regions of PD mice via [¹⁸F]L-DOPA PET and [¹⁸F]GSK PET. [¹⁸F]L-DOPA PET demonstrated DA neuronal damage in A53T mice brains (cortex, basal forebrain septum, striatum, superior colliculi and whole brain). The decrease in [¹⁸F]L-DOPA levels in the striatum is directly linked to the emergence of motor symptoms of PD. The cortex and basal forebrain septum play crucial roles in the development of PD. The primary motor cortex is associated with motor planning and execution in PD patients, and its dysfunction depends on the severity and/or duration of dopamine depletion⁴². The basal forebrain septum is associated with cognitive impairment in advanced PD⁴³. In the brains of A53T mice, dopamine depletion occurs not only in the striatum and midbrain but also in other regions. Thus, attention to neurons with reduced [¹⁸F]L-DOPA uptake in PD patients should not be limited to the striatum and substantia nigra but should also focus on these brain regions.

There are more brain regions with P2X7R changes in [¹⁸F]GSK PET than brain regions with changes in [¹⁸F]L-DOPA uptake. We observed extensive neuroinflammation in the brains of A53T mice. [¹⁸F]GSK PET revealed increased uptake in eleven brain regions. This result was consistent with previous findings that neuroinflammation occurs in the neurodegenerative brains of PD animal models and patients^44,45. Neuroinflammation has been detected in some areas where damage to DA neurons has not been identified. After MK treatment, inflammation in these brain regions decreased. This may suggest damage to non-DAergic neurons. Non-DAergic neurons, such as those in the noradrenergic system, serotoninergic system, and acetyl cholinergic system, also play important roles in the progression of PD. However, PET imaging of DA neurons ([¹⁸F]L-DOPA, [¹⁸F]DTBZ) does not indicate damage to other neurons. Therefore, in vivo imaging targeting neuroinflammation is beneficial for evaluating disease progression. Microglial P2X7R is activated upon cell injury and death and can be recognized by [¹⁸F]GSK. Previously, PET imaging targeting TSPO was carried out for both PD patients and healthy controls⁴¹. The amygdala, cortex, hippocampus, midbrain, striatum, and thalamus show high uptake in neurodegenerative brain regions⁴¹. Our study highlights the brain regions where inflammation is elevated in those PD patients. We observed that the basal forebrain and hypothalamus were additional brain regions associated with increased inflammation in the brains of A53T mice.

The motor symptoms of PD are diverse. To our knowledge, there has been no studies on whether changes in [¹⁸F]L-DOPA uptake in different brain regions are associated with these various motor symptoms. Correlation analysis between motor tests and [¹⁸F]L-DOPA revealed that [¹⁸F]L-DOPA uptake in the A53T mice brain is negatively correlated with motor dysfunction. The striatum is the only brain region that showed a negative correlation with three behavioral tests. The superior colliculi play a critical role in the neural control of saccadic eye movements⁴⁶.

In the early stages of PD, degenerative changes occur in the basal forebrain⁴⁷. Current studies have demonstrated that alterations in the basal forebrain during PD primarily involve the degeneration of cholinergic neurons, leading to cognitive impairment⁴⁸. Although no definitive research has linked dopaminergic changes in the basal forebrain to muscle strength, motor coordination and endurance, our findings revealed a significant correlation between alterations in dopamine uptake within this region and motor performance. This study may enhance the comprehensive understanding of [¹⁸F]L-DOPA PET imaging in PD, further guiding new directions for diagnosis and treatment on the basis of DA neurons in specific brain regions.

The observed neuroinflammation-motor behaviors association exhibited regional specificity within distinct cerebral regions. Although inflammation spreads throughout almost the entire brain, our results demonstrated that the effects of inflammation on behavioral tests are not homogeneous among brain regions (Figs. 6–8). The brain regions with the highest correlation between muscle strength and inflammation include the cortex, hippocampus, and striatum, among others (Fig. 6). A previous study revealed a significant association between grip strength and gray matter volume in these brain regions, including the temporal pontine cortex, hippocampus, striatum, and brain stem⁴⁹. Gray matter atrophy in PD is driven by neuronal loss and structural damage⁵⁰. Although compensatory mechanisms or interventions may slow disease progression⁵¹, there was no evidence identify that PD brain atrophy is a reversible pathological outcome. In the present study, we observed that inflammation was present in the brains of A53T mice and the behavior was improved by MK. Thus, inflammation in brain regions such as cortex, hippocampus, and striatum may serve as an early warning sign of muscle strength disorders in PD. The brain regions with the highest correlation between restored motor coordination and inflammation, including the cortex, midbrain, striatum, and hippocampus (among others), were identified as special centers for motor coordination function (Fig. 7). The hippocampus may be involved in the encoding or retrieval of spatial information, whereas higher-order motor integration and planning rely on the inferior parietal cortex⁵². The striatum, thalamus, and midbrain are also involved in the projection of motor coordination functions⁵³. Our study revealed the brain regions with the greatest correlation between muscular endurance and inflammation (Fig. 8), which further contributes to our understanding of endurance disorders. When fatigue develops during endurance exercises, there is an increased voluntary drive from the cortex to the spinal alpha motor neurons (which control the working muscles) to maintain the target force⁵⁴. Although the projections between brain regions have not yet been explained, we suggest for the first time that other brain regions, such as the hypothalamus, basal forebrain septum, striatum, cortex, and amygdala, may also be involved in muscle fatigue. The [¹⁸F]GSK brain regions associated with endurance exhibited overlap with the [¹⁸F]L-DOPA brain regions. In summary, different motor behaviors may be affected by specific brain regions. [¹⁸F]GSK neuroimaging findings may reveal patterns of attributable brain changes due to inflammatory responses.

[¹⁸F]GSK may reveal more neural trauma, including but not limited to dopamine neurons. This may suggest a link between the occurrence of motor disturbances and the impairment of non-DAergic neurons caused by α-syn in PD patients. Additionally, analysis of inflammation in these brain regions may help predict Deep brain stimulation (DBS) treatment responses. DBS, which targets and disrupts dysfunctional brain regions or networks, offers important therapies for PD. However, functional imaging studies using [¹⁸F]L-DOPA PET failed to show any neuroprotective effect of DBS in the subthalamic nucleus (STN) in PD patients⁵⁵. An increasing number of studies have suggested that DBS can alleviate chronic inflammation and that the anti-inflammatory effect may be a mechanism for its efficacy⁵⁶. Advances in our understanding of neuroanatomical networks and neuroinflammation may expand the use of DBS or other neuromodulation therapies in the future.

Nevertheless, our study has several limitations. First, despite the absence of significant differences in uptake between groups in certain brain regions, there is a correlation with motor behaviors. The lack of statistical differences between groups may be attributed to an insufficient sample size. Second, although MK is effective in restore motor functions and suppressing neuroinflammation in PD mice, the detailed molecular mechanism remains to be further elucidated in future.

In the present study, we combined motor testing, [¹⁸F]L-DOPA PET, which targets DA neurons, and [¹⁸F]GSK PET, which targets microglial P2X7R, to map the damage of DA neurons and the microglial activity of brain regions in PD. Motor behaviors and partial DA deficits improved, and microglial activity was resynchoroizeded by MK. Furthermore, [¹⁸F]GSK PET revealed altered microglial activations in MK-treated A53T mice and revealed correlations between motor behaviors and inflammation in brain regions. This study contributes to revealing the connection between neurodegenerative PD symptoms and P2X7R-related neuroinflammation, providing strong evidence for the use of [¹⁸F]GSK PET in the comprehensive evaluation and treatment of PD.

Methods

Experimental design

A53T transgenic mice overexpressing human α-syn with PD-associated mutation⁵⁷ were used in this study. A53T mice were purchased from the Jackson Laboratory (strain 039167). C57BL/6 J mice were used as WT mice as controls for A53T mice. All studies were conducted in accordance with the Provision and General Recommendation of the Chinese Experimental Animals Administration Legislation. The mice were housed in groups under standard laboratory conditions (25 °C, 12-hour light/12-hour dark cycle) in plastic cages with sawdust bedding. Only mice demonstrating statistically significant differences in motor behavior (grip tests, pole tests and inverted screen tests) were subsequently enrolled in the intervention study. The sample size was determined as the minimal necessary number approved by the Institutional Animal Care and Use Committee (IACUC), in compliance with the 3R principles (replacement, reduction, refinement) for rodent studies, combined with constraints imposed by the production costs of PET/CT tracers. No animals, experimental units, or data points were excluded from the analysis in any experimental group. To mitigate potential confounding factors, comprehensive measures were implemented including standardized environmental controls (temperature, humidity, lighting cycle, and diet), randomized cage rotation protocols, blinded operational procedures, and matching of baseline biometric parameters (body weight ±10%).

Six-month-old male mice were used in this study. The animals were randomly divided into four groups. Given the limited cohort size (n = 6/group), manual randomization was performed using Fisher-Yates permutation from Cambridge Statistical Tables (Series B), where unique animal IDs were sequentially matched to random digits through modular arithmetic, ultimately allocating the first six ID-matched subjects to treatment and the remainder to control groups. Group I: A53T mice were treated daily with normal saline (NS) via intragastric gavage (i.g.) (A53T, n = 6). Group II: A53T mice were treated (i.g.) daily with MK (10 mg/kg, Macklin Inc.) dissolved in NS (A53T/MK, n = 6). Group III: WT mice were treated (i.g.) daily with NS (WT, n = 6). Group IV: WT mice were treated (i.g.) daily with MK dissolved in NS (WT/MK, n = 6). All the groups were treated with 100 μL of solution. The dosage of MK was 10 mg/kg, as previously described⁵⁸. During the MK treatment period, weekly behavior tests were conducted to confirm the stability of the drug’s efficacy before proceeding with the PET scans. All the groups were treated for 5 weeks. In the 5^th week, the mice were subjected to behavior tests (n = 6 per group), [¹⁸F]GSK PET-MRI scans (n = 6 per group), [¹⁸F]L-DOPA PET-MRI scans (n = 3-4 per group), autoradiography (n = 2 per group), and immunofluorescence (n = 2 per group)(Table S1). Mice were randomized by independent personnel not involved in subsequent experimental procedures, while all operators performing motor tests and PET imaging remained rigorously blinded to group assignments throughout data acquisition and analysis phases. Animals were observed daily for normal activity, coat condition, and food/water intake. No animals reached a state necessitating early euthanasia. A summary of the reagents and antibodies used is shown in Table S18. A summary of the equipment and software is shown in Table S19.

Radiochemistry

[¹⁸F]L-DOPA was obtained from the Guangdong Huixuan Pharmaceutical Company. [¹⁸F]GSK was prepared as described in previous studies⁵⁹. Fluoride-18 was purchased from Guangzhou Atomic High Tech Pharmaceutical Company and trapped on a preconditioned Sep-Pak Accell Plus QMA Light cartridge (WAT023525, Waters, USA). The QMA cartridge loaded with fluoride-18 was rinsed with a preprepared tetraethylammonium bicarbonate (TEAB) solution (1 mL, 14 mg/mL, acetonitrile:purified water = 1:1). The system was dried at 120 °C under nitrogen, followed by azeotropic drying with the addition of 1 mL of acetonitrile to make the system as anhydrous as possible. The predissolved precursor (2 mg) was added to the dried reaction system and reacted in a bottle with a pressure cap at 120 °C for 15 minutes. At the end of the reaction, the mixture was passed through a high-performance column (ZORBAX SB-C18, Agilent, USA) for radio semipreparative HPLC (P230A/P, Elite, China) to purify the target product. The final product was loaded on a Sep-Pak Plus C18 cartridge (WAT020515, Waters, USA) and concentrated with ethanol. The concentrated product was injected into a Shim-pack HPLC-packed column (227-30017-08, Shimadzu, Japan) for quality control and coinjection. The reaction and HPLC procedure are shown in Fig. 1a–d.

Estimation of motor activity

Since the motor symptoms of PD vary, we performed motor tests (including grip tests, pole tests, and inverted screen tests) to investigate multiple motor behsviors in A53T mice. The illustrations of the motor tests were shown in Fig. S1. Grip tests Muscle weakness is a common symptom in PD patients⁶⁰. Grip tests were performed to assess the grip strength of the forelimbs. The mice were placed on top of the grid, and the tails were held gently to ensure that only the forelimbs were able to grip the grid. The peak force of the forelimbs was recorded in 5 trials per mouse on an electronic grip strength meter (Huayonbio Co Ltd., China). The mean peak force was calculated. The measurements were invalidated if the mice struggled or if their hind limbs touched the grid. The results of the grip tests after treatment are shown in Fig. 2. Pole tests PD patients with stiffness in the limbs exhibit worse motor coordination⁶¹. Pole tests were performed to evaluate motor coordination. A rough-surfaced pole (diameters 10 mm; height 60 cm) was placed in a cage covered with bedding to protect the mice from injury. The animals were placed on the top of the pole facing the head-upward direction, and 120 s of latency was assigned. The time the mice took to turn their heads down (T-Turn) and descend to the floor (T-Descend) was recorded in 3 trials per mouse. There was a 60-second interval between each trial to recover the condition of the mice. The mean values of the 3 trials were calculated and used for statistical analysis. When the mice were not able to turn downward or when they were dropped from the pole, T-Turn and T-Descend were taken as 120 s (maximum value). At the end of the test, the mice were released back into cages in which water and food were provided. The results of the pole tests after treatment are shown in Fig. 2. Inverted screen tests Unlike the grip test, inverted screen tests measure the muscle strength of all four limbs and significantly decrease muscle strength⁶². When the muscle strength of the mice weakened, they could not hold on to the inverted screen. Inverted screen tests were performed to assess endurance. The mice were placed in the center of the wire screen (45 × 30 cm, 1 mm diameter metal strings). The screen was gently rotated 180° above a transparent cage covered with bedding, and the timing started. The times when the mice persisted on the inverted screen were recorded in at least 5 continuous measurements, and 120 s of latency was assigned in each measurement. The results of 5 consecutive measurements were statistically calculated and plotted as a time-varying curve. The trend of the time-varying curve was of concern since it reflects changes in muscular endurance during the experiment. The decline ratio (1^st measurement/5^th measurement) was also calculated for each mouse. The greater the decline ratio is, the worse the endurance of the mouse limbs. Following the test, the mice returned to their cages to rest. The results of the inverted screen tests after treatment are shown in Fig. 2.

PET/CT imaging acquisition

Neuroinflammation was quantified via [¹⁸F]GSK PET scans, and DA neurons were quantified via [¹⁸F]L-DOPA PET scans. Following MK treatment, we performed [¹⁸F]GSK PET scans (n = 6 each group) and [¹⁸F]L-DOPA PET scans (n = 3–4 mice per group).

Prior to the intravenous administration of [¹⁸F]GSK (238.9-336.4 MBq/kg) or [¹⁸F]L-DOPA (231.2-343.3 MBq/kg), the animals were anesthetized with 1.5% isoflurane. Following the injection, the mice were allowed to recover from anesthesia and were maintained in a wakeful state under observation for a period of 30 minutes. Subsequently, the mice were re-anesthetized using 1.5% isoflurane, and PET imaging was conducted utilizing a nanoPET scanner (Mediso Inc., Hungary). At the time of the initial experimental design, we planned to perform dynamic scans (including 30 minutes of dynamic PET scans and 10 minutes of whole-body CT scans). Animals were anesthetized using 1.5% isoflurane to maintenance for a 40-minute period. However, most of the A53T mice died during the dynamic scans.

A PET study of an A53T mouse model indicated that static scanning may be more appropriate than dynamic scanning for A53T mice⁶³. Therefore, we performed a static scan (including 15 minutes of PET scans and 10 minutes of whole-body CT scans), and all mice survived after the static scans. Thirty minutes prior to the injection of [¹⁸F]L-DOPA, all animals were pretreated via intraperitoneal injection with benserazide (10 mg/kg, B801922; Macklin, China), an aromatic L-amino acid decarboxylase (AADC) inhibitor⁶⁴, and entacapone (10 mg/kg, 134750; BIOBERRY, USA), a catechol-O-methyltransferase (COMT) inhibitor⁶⁴, to increase the bioavailability of [¹⁸F]L-DOPA. All the PET images were reconstructed by Fourier rebinning using InterView FUSION software (Mediso, Hungary).

MRI data collection

MRI scans were performed to obtain the anatomical locations of the brain regions (including the lateral ventricles) in the PET images. The brain MRI scans of the mice were collected through a 9.4 T micro-MRI scanner (BioSpec 94/30 USER, Bruker BioSpin MRI Gmbh, Germany). The mice underwent T2-weighted MRI scans with the following parameters: repetition time (TR) = 6200 ms, echo time (TE) = 22.5 ms, slice thickness = 0.3 mm, average = 5, scan time = 9 m 18 s, number of slices = 55, field of view = 15 × 15, and image size = 150 × 150.

PET-CT-MRI data analysis

PET images and CT images were aligned via PMOD software (PMOD Technologies LLC, Switzerland), and then the standardized uptake values (SUVs) of the brain regions were obtained with the PMOD Volumes of Interest (VOIs) template (Ma-Benveniste-Mirrione)⁶⁵. The SUV was corrected by the initial injection dose and body weight of the mice. Then, the fused PET/CT images were further fused with MRI images via Carimas 2.10 software (Turku PET Center, Finland) to ensure alignment and obtain PET-MRI images. Since most of the mice underwent static scanning, normalization of the [¹⁸F]L-DOPA PET imaging data was conducted by SUV_R^Cere (SUV of the brain regions/SUV of the cerebellum) according to previous methods⁶⁶. The normalization of [¹⁸F]GSK PET imaging analysis was conducted by SUV_R^LV (SUV of the brain regions/SUV of the lateral ventricle).

Ex vivo autoradiography of [¹⁸F]GSK

The mice were sacrificed 45 minutes after the injection of [¹⁸F]GSK (1057 MBq/kg). The brains of the mice were quickly removed and cut into 30 μm-thick slices at −20 °C. The brain slices were subsequently exposed to the Beaver micropattern gas detector system (Ai4R Inc., France) for 3 hours. Autoradiography images were quantified by BeaQuant software 3.3 (Ai4R Inc., France). The radioactivity (cp/min/mm²) of the cortex, basal forebrain septum, striatum, thalamus, hypothalamus, amygdala, hippocampus, midbrain, brainstem, and cerebrallum was quantified. In the autoradiography experiments, each experimental group comprised two mice. To ensure comprehensive data acquisition, at least ten serial sections were systematically collected from specific brain regions of each mouse. A representative region of interest (ROI) was delineated within each section, resulting in a minimum of 10 independent ROI data points per group.

Immunofluorescence

The mice were deeply anesthetized with a 3.0% isoflurane/oxygen gas mixture and transcardially perfused with NS. The brains were removed and cut into 10 μm pieces. Fluorescence staining was performed using primary antibodies against P2X7R (rat, 1:200, ab195356, Abcam, UK) and ionized calcium-binding adaptor molecule 1 (IBA1) (mouse, 1:200, ab283319, Abcam, UK). The slides were incubated with primary antibodies at 4 °C overnight. After incubation with primary antibodies, the sections were labeled with the corresponding species-specific secondary antibodies for 1 hour at room temperature under light-protected conditions: Alexa Fluor 488-conjugated goat anti-mouse IgG (1:500, Bs-0296G-AF488, Bioss, China) and Alexa Fluor 647-conjugated goat anti-rat IgG (1:500, 112-605-143, Jackson, USA). Then, the slides were incubated with DAPI (1:1000, D1306, Thermo Fisher, USA). Images were taken using an Axio Observer 7 microscope (ZEISS, Germany). The immunofluorescence intensity was quantified with ImageJ (version 1.53k, NIH, USA).

Statistics and reproducibility

Statistical analysis was carried out via GraphPad Prism 8 (GraphPad Software, USA). All the data are presented as the mean ± standard deviation (SD). Since the experiment was divided into four groups involving two intervention factors (genotype and MK treatment), a two-way analysis of variance (ANOVA) was employed. Correlation analysis and simple linear regression were performed. The results were statistically significant when p was <0.05. Each p value was adjusted to account for Tukey’s multiple comparisons test in two-way ANOVA. All hypothesis tests conducted as two-tailed analyses.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

H.J. contributed to the study’s conception and design. H.J. and P.Y. provided an experimental scheme. P.Y., L.B., and Y.Q. participated in radiochemistry and HPLC quality control. P.Y., L.B., and M.Y. carried out the PET analysis. P.Y. and Y.Q. supported the MRI scans. P.Y. participated in the immunofluorescence staining. H.J. and P.Y. participated in the writing of the article. Y.L., Y.H., P.Z., X.C. and J.S. participated in the suggestion of an experimental scheme and modification of the article. All the authors commented on previous versions of the manuscript. All the authors read and approved the final manuscript.

Peer review

Peer review information

Communications Biology thanks Cassis Varlow and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Claudia Kathe and Benjamin Bessieres. A peer review file is available.

Data availability

The PET and MRI dataset assessed in this analysis is available from the corresponding author upon reasonable request. The source data behind the graphs in the paper can be found in Supplementary Data. All data generated or analyzed during this study are included in this published article and its Supplementary materials. Original data are available at Sun Yat-sen University Research Data Deposit (#RDDYJ101499, https://rdd.sysu.edu.cn/UserHome/ProjectAuthorConfirm.aspx?ProjectID=28760&ProjectAuthorID=6589332B388235DC (up to dated)).

Code availability

The source code used for analyzing PET and MRI data is publicly available at PMOD and Carimas (https://carimas.fi/).

Competing interests

The authors declare no competing interests.

Ethics approval

All experimental procedures and animal care were under the supervision of the Institutional Animal Care and Use Committee (IACUC) in the Fifth Affiliated Hospital of Sun Yat-sen University (animal protocol No. #00264).

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-025-08602-x.