Involvement of Cyclin Dependent Kinase 5 in M4 Muscarinic Acetylcholine Receptor-Mediated Cholinergic Transmission within the Mouse Dorsal Striatum

Background: An imbalance between dopamine (DA) and acetylcholine (ACh) within the striatum has reemerged as key to the pathophysiology of the neurodegenerative disorder, Parkinson's disease (PD). M 4 is a prominent muscarinic ACh receptor subtype in the striatum and we have previously reported that M 4 controls cyclin-dependent kinase 5 (Cdk5) / dopamine- and cAMP-regulated phosphoprotein of Mr 32 kDa (DARPP-32) activity in cultured medium spiny neurons (MSNs). However, the mechanism of this control remains unclear. Methods: Genetic, electrophysiological, and immunohistochemical approaches were used in conjunction with pharmacological methods to study isolated M 4 -deleted MSNs (M 4 -KD MSNs) and a dorsomedial striatum (DMS) M 4 knockout mouse model. We examined the role of Cdk5 in M 4 -mediated neural cholinergic transmission and related behavior. Results: Oxotremorine M, a nonselective mAchR agonist, promoted Cdk5/P35 signaling activity in DSM MSNs both in vivo and in vitro. Either pharmacological inhibition or genetic knockdown of M 4 decreased the amount of Cdk5 and DARPP-32 phosphorylation at Thr75 in dopamine 1 receptor-expressing MSNs. Furthermore, whole-cell patch-clamp recording conrmed Cdk5 is necessary for M 4 -mediated cholinergic inhibition of excitatory synaptic transmission in MSNs in vivo and in vitro. Concomitantly, deletion of M 4 activity in the DMS caused Oxotremorine M-induced Cdk5 signaling and glutamatergic synaptic input to be altered in parallel with behavioral responses. Conclusions: We characterized a novel regulatory mechanism of Cdk5/DARPP-32 involved in M 4 mediated cholinergic regulation on striatonigral neurons and on motor behavior. The ndings indicate that inhibition of M 4 mAChR could be a novel approach to correct the pathological conditions of PD. ACh, acetylcholine; Oxo-M, oxotremorine M; mAChRs, muscarinic acetylcholine receptors; D 1 , dopamine D 1 receptor; M 4 , muscarinic acetylcholine 4 receptor; DARPP-32, dopamine- and cAMP-regulated phosphoprotein, Mr 32 kDa; KD, knockdown; MSNs, medium spiny neurons; PKA, protein kinase A; PBS, phosphate buffered saline; DMS, dorsomedial striatum; CRISPR-Cas9, clustered regularly interspaced short palindromic repeats-associated protein 9; sgRNA, single-guide RNA; eGFP, enhanced green uorescent protein; i.p., intraperitoneally; i.c.v., intracerebroventricular; WT, wild type; NC, negative control; FST, forced swimming test; mEPSC, miniature excitatory postsynaptic current; ACSF, articial cerebrospinal uid; TTX, tetrodotoxin; PAM, positive allosteric modulator; Cdk5, cyclin-dependent kinase 5; AP, anteroposterior; ML, mediolateral; DV, dorsoventral; SDS, sodium dodecyl sulfate; ANOVA, analysis of variance; LV, lentiviral; PD, Parkinson’s disease

These ndings suggest that cyclin-dependent kinase 5 (Cdk5)-speci c phosphorylation associated with DARPP-32 seems to be controlled by M 4 stimulation. However, the mechanisms by which Cdk5 acts in M 4 -mediated neural cholinergic transmission and related behavioral responses are unclear.
The dorsomedial striatum (DMS) is thought to be involved in early stages of motor processing and thus alterations in DMS signaling are likely to yield global changes in motor behavior [18,19]. The information conveyed to this subcortical structure through glutamatergic projections from the cerebral cortex and thalamus is processed by several striatal neuromodulatory systems, including the cholinergic system [20]. In this study, we used lentiviral-mediated M 4 -shRNA transduction to knockdown the M 4 gene in isolated MSNs in vitro. We also used lentiviral CRISPR-Cas9 technology to delete M 4 in the DMS of mice in vivo. DARPP-32 also acts as an integrator of dopaminergic and glutamatergic inputs [21]; Therefore, using pharmacological and electrophysiological approaches, we analyzed Cdk5 activity and the glutamatergic input to striatonigral MSNs in vitro and in vivo. In parallel with biochemical changes, we also studied behavioral responses in the DMS M 4 knockout (DMS-M 4 -KD) mouse. It is important to understand the functional role of M 4 modulation of region-speci c striatal neural circuitry in neurodegenerative disorders such as Parkinson's disease.

Animals and reagents
Experiments were performed with 8-week-old C57BL/6J mice (RRID: MGI: 5657312) and 12-hour-old Sprague-Dawley rats (RRID: MGI: 5651135) supplied by the Animal Center of the Academy of Military Medical Science (Beijing, China). C57BL/6J mice were group-housed in a temperature-controlled environment (22 ± 2°C) with a 12/12-h light/dark cycle and access to food and water ad libitum. All mice were group-housed for 3 days prior to use and were handled daily throughout the experiment to minimize the effects of handling stress. All tests were performed and recorded between 9:00 and 15:00 during the lights-on cycle. Animal experiments (Ethics approval No. LACUC-DWZX-2020-517) were approved by the Beijing Local Committee on Animal Care and Use and were performed according to the Guide for the Care and Use of Experimental Animals produced by the Beijing Local Committee and the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH publication No. 85-23, revised 1985). Efforts were made to minimize the number of animals used for each experiment.

Microinjection and drug treatment
Oxo-M was dissolved in vehicle and the nal dose used was 0.1 µM. Mice received intracerebroventricular (i.c.v) injection of Oxo-M (2.0 µL) or vehicle (2.0 µL) at the following coordinates: +1.0 mm AP, ±0.2 mm ML, −2.5 mm DV. The solutions were injected bilaterally through a microinjection needle (30 gauge) that extended 1 mm beyond the tip of the guide cannula. Each microinjection needle was attached to a 10-µL Hamilton microsyringe through polyethylene tubing (PE-10). Infusions were controlled by an infusion pump (Model Bi2000e Insight Equipment, Sao Paulo, Brazil), programmed to deliver solution at a constant speed of 0.5 µL per min. The microinjection needle was kept in place for an additional minute to allow for drug diffusion. The mice were allowed to move freely during drug administration.

Behavioral tests
Mice were screened 3 days before lentivirus or drug injection, and those with abnormal motor behavior were excluded. Behavioral tests were performed 14 days after lentivirus injection. Mice were transferred to a behavioral testing room (22±1°C) and habituated for at least 5 min before behavioral testing. All behavioral experiments were performed at xed times (autonomic movement test: 9:00-12:00; rotarod and forced swim test: 12:00-14:00). The testing apparatus was cleaned with a hypochlorous acid solution.

Locomotor activity measurements
Mice were individually transferred from their home cage to a plastic open eld apparatus (60 cm × 60 cm × 60 cm; Xingruan Information Technology Co. Ltd., Beijing, China). The apparatus was virtually divided into peripheral, intermediate, and central zones. The test was started by placing the animal in the center of the open eld illuminated by a dim light (5 lx). Each mouse was placed in the locomotor activity box for 60 min and was recorded by a video camera. The total distance and the distance traveled in the central zone (referred to hereafter as central distance) was recorded and analyzed by ANY-Maze automated video tracking software (Stoelting Co., Wood Dale, IL, USA; RRID: SCR_014289).

Rotarod test
The rotarod system (Xingruan Information Technology Co. Ltd, China) for assessing locomotor skills measures the time that an animal maintains balance on a moving rod. Animals were rst conditioned on a stationary rod for 30 s and during this time any animal that fell was placed back on the rotarod. Animals were next conditioned at a constant speed of 5 rpm for a period of 5 min. Twenty-four hours after conditioning, animals were placed on the rod and timed for 30 min to assess their locomotor skill. The rod speed started at 5 rpm and was increased at 0.1 rev/s; the time before falling off the rod (fall off duration) was recorded by ANY-Maze automated video tracking software (Stoelting Co. Ltd., Sheboygan, IL, USA).

Forced swimming test
The test procedure was previously standardized and validated [22]. Mice were individually forced to swim in an open cylindrical container (diameter 10 cm, height 25 cm) containing 19 cm of water (depth) at 25 ± 1°C. Each mouse was considered to be immobile when it ceased struggling and remained oating motionless in the water. The total duration of immobility was recorded during a 10-min period by ANY-Maze automated video tracking software (Stoelting Co. Ltd, China).

Immunocytochemistry and confocal microscopy analysis
For immunocytochemistry, cells were xed in methanol for 20 min and blocked with 5% BSA (ZSGB-Bio, Wuhan, China) in PBS for 1 h, then incubated with primary antibodies overnight at 4°C. The cells were then washed before incubation with the corresponding secondary antibody for 1 h at room temperature.
For immunohistochemistry, tissues were dissected and post xed overnight at 4°C in 4% paraformaldehyde and then cryoprotected with 30% sucrose in 0.1 M PBS for 2 days. Sections were cut at 30 µm on a cryostat microtome (CM1950, Leica, Wetzlar, Germany). Free-oating sections were permeabilized with 0.3% Triton X-100 in PBS for 30 min at room temperature. After blocking with 10% normal goat serum for 2 h, sections were incubated with primary antibodies overnight at 4°C. Sections were then incubated with secondary antibodies for 2 h at room temperature.
The primary and secondary antibodies are shown in Additional le: supplementary Table 1. The nuclear layers were stained with DAPI (Sigma-Aldrich). Images were obtained with a laser confocal microscope (LSM 8000, Carl Zeiss, Oberkochen, Germany). ImageJ imaging software (National Institutes of Health, MD, USA) was used to quantify the uorescence intensity.

Western blotting
Primary cell cultures or dorsal striatum tissue were lysed in RIPA buffer (50 mM Tris, 150 mM NaCl, 5 mM NaF, 0.2% SDS, and 1% NP-40, pH 8.0) containing a mix of protease and phosphatase inhibitors (Google Biological Technology Co., Ltd., Wuhan, China) before being centrifuged at 4°C, 12,000 ×g, for 30 min. The supernatants were collected, and protein concentration measured using a commercial BCA kit (Google Biological Technology Co., Ltd.). Equal amounts of protein were separated on 10% SDS-PAGE gradient gels, transferred to nitrocellulose membranes, and incubated with primary antibodies diluted in TBST (150 mM NaCl, 40 mM Tris-HCl, pH 7.4, 0.1% Tween-20) overnight at 4°C. After washing, membranes were incubated with secondary antibodies coupled to horseradish peroxidase. The primary and secondary antibodies are shown in Additional le: supplementary Table 2. Immunocomplexes were detected by chemiluminescence (ECL, Pierce, USA), imaged using an imaging instrument (GENE Co. Ltd., Beijing, China) and analyzed using Image J. Analyses were performed in duplicate and the mean value was calculated.
Whole-cell patch clamp electrophysiology in the DMS.
Mice were anesthetized by injection with pentobarbital sodium (60 mg/kg, i.p.), then decapitated and the brains removed. Coronal slices (300 µm thick) containing the striatum (0.2-1.3 mm posterior to bregma) were prepared in ice-cold ACSF (as follows: 124 mM NaCl, 2.8 mM KCl, 1.25 mM NaH 2 PO 4 , 2 mM CaCl 2 , 1.25 mM MgSO 4 , 26 mM NaHCO 3 , 10 mM glucose, pH 7.5, bubbled with 95% O 2 /5% CO 2 ) using a vibrating tissue slicer (MA752, Campden Instruments, USA) as previously described [23]. Slices were submerged for 30 min at 32°C in protective media containing 2.5 mM KCl, 1.2 mM NaH 2 PO 4 , 30 mM Patch pipets were prepared from borosilicate glass (Sutter Instrument Company, Novato, CA, USA) using a P-97 Flaming/Brown micropipette puller (Sutter Instrument) and had resistance of 6-8 MΩ when lled with the following intracellular solution: 130 mM CsCl, 10 mM NaCl, 0.25 mM CaCl 2 , 2 mM MgCl 2 , 5 mM EGTA, 10 mM Hepes, 10 mM glucose, 2 mM Mg-ATP, and 0.3 mM Na 2 -GTP. The pH of the pipette solution was adjusted to 7.3 with 1 mM CsOH and osmolarity was adjusted to 285-290 mOsm/L. A low-power objective (4×) was used to identify the striatum and a 40× water immersion objective (NIR Apo, Nikon, Tokyo, Japan), coupled with infrared differential interference contrast microscopy and a charge coupled device camera, was used to visualize individual neurons. Cells in the dorsolateral striatum up to approximately 50 µm beneath the slice surface were patched and monitored. Recordings in normal current-clamp or voltage-clamp mode were acquired at room temperature using the Digidata 1440 A interface (Molecular Devices, Sunnyvale, CA, USA) with an Axon 200B ampli er (Molecular Devices) and Clampex 10.2 software (Molecular Devices). After formation of a tight seal (>1 GΩ), fast and slow capacitance compensation was performed. During the whole-cell recording, series resistance was compensated (80%-90%) and monitored periodically. Neurons were excluded from the analysis when their series resistance was above 50 GΩ or changed by more than 25% during the experiment. Data were ltered at 2 kHz and acquired at a sampling rate of 10 kHz. MSNs were identi ed by previously determined membrane characteristics and ring patterns [24].
For miniature excitatory postsynaptic current (mEPSC) recordings, oxygenated ACSF containing both the GABA receptor antagonist bicuculline (10 µM; Sigma) and the voltage-gated sodium channel blocker tetrodotoxin (TTX; 1 µM; Abcam Biochemicals, UK) was applied to the bath to abolish inhibitory postsynaptic current events and action potentials, respectively. Slices were perfused with this solution at 25°C for at least 15 min following establishment of electrical access. Access resistances were <15 MΩ. mEPSCs were recorded from MSNs held at 70 mV in gap-free mode. After 5 min of stable baseline recordings, Oxo-M was bath applied to slices for 5 min and recording performed for at least 3 min.

Data analysis
For behavioral experiments and biochemistry assays, including immunocytochemistry and western blotting, statistical analysis was performed by one-way ANOVA or two-way ANOVA (with or without repeated measures) followed by Tukey's multiple comparisons test using GraphPad Prism 7.0 (GraphPad, San Diego, CA). For electrophysiology, statistical analysis was performed using Clampex 10.2 software (Molecular Devices). The two-sample Kolmogorov-Smirnov test was used to compare the cumulative distributions of frequency of the two groups. Data are expressed as the mean ± S.E.M. "n=8" is the number of animals used, "n=10" is the total number of ROIs in immunocytochemistry staining and patch clamp electrophysiology (cells or tissue sections), and "n=3" the number of independent western blotting experiments performed, unless otherwise stated. P values below 0.05 were considered signi cant.  Figure S1). Confocal microscopy analysis revealed Cdk5 distribution in M 4 -positive MSNs (Fig. 2a). We used Pearson's correlation coe cients (PCC) and Mander's colocalization coe cients (MCC) to examine the correlation between the M 4 and Cdk5 uorophores.

Cdk5 is necessary for M 4 -mediated cholinergic inhibition of excitatory synaptic transmission in MSNsin vitro
ACh potently modulates glutamate signaling in the striatum via activation of mAChRs, and M 4 is preferentially expressed in D 1 -MSNs where it is clustered near axospinous glutamatergic synapses [2,28].
We investigated whether M 4 and Cdk5 modulate local glutamatergic synaptic input on MSNs by recording miniature postsynaptic potential (mEPSCs), which re ect the presynaptic release of excitatory neurotransmitter vesicles. mEPSCs were recorded from primary cultured MSNs in the presence of bicuculline and TTX. Bath application of Oxo-M (2, 5, 10 µM) evoked a rightward shift in cumulative probability of glutamate release from MSNs, indicating a concentration-dependent inhibition of excitatory synaptic transmission by Oxo-M (Fig. 3a, b). There was a 65.71 ± 4.30% reduction in mEPSC frequency with 10 µM Oxo-M compared with that with vehicle [p < 0.001, F(3,36) = 51.91)]. Considering a potential contribution of Cdk5 in cholinergic transmission, we then tested the effect of Cdk5 inhibitors on Oxo-Minduced cholinergic inhibition of glutamatergic transmission. When we coapplied ROSC (100, 300, 500 µM) with Oxo-M (10 µM), the inhibitory effect of Oxo-M on mEPSC frequency was partially antagonized (Fig. 3c, d), and 500 µM ROSC signi cantly counteracted the action of Oxo-M (10 µM) on mEPSC frequency; the inhibition of mEPSC frequency was decreased to 44.71 ± 12.73% [vs Oxo-M, p < 0.05, F (3,36) = 4.08, Fig. 3d]. We also performed identical studies in M 4 -KD MSNs (Fig. 3e). Compared with WT MSNs, Oxo-M-induced inhibition in mEPSC frequency was abolished in M 4 -KD MSNs (Fig. 3f). The inhibition of mEPSC frequency was decreased to 7.27 ± 3.19% (vs WT: Oxo-M treatment, p < 0.001, Fig. 3g). Taken together, these data indicate that M 4 mediates cholinergic inhibition at glutamate terminals and that Cdk5 plays a role in the cholinergic modulation of MSN synaptic activity.

M 4 deletion in the DMS alters Oxo-M-induced behavioral responses in mice
We next tested M 4 function in vivo using CRISPR-Cas9 gene targeting technology to create mice that lack functional M 4 receptors in the DMS (DMS-M 4 -KD) (see Additional le: Figure S3-4 for details). We directly  Fig. 5c, d). These results con rmed that Cdk5 is an endogenous cholinergic M4 regulator in the DMS that impacts motor behavior.
We then performed whole-cell recordings in brain slices prepared from the DMS of mice. mEPSCs were isolated as described above for isolated MSNs, and we recorded cells from the eGFP-positive areas in WT and DMS-M 4 -KD mice. Characteristics exhibited by MSNs are shown in Fig. 5e. Bath application of Oxo-M (10 µM) evoked similar inhibition of mEPSCs compared with that in isolated MSNs. There was an approximately 89.07 ± 0.62% reduction in cumulative release probability in mEPSC frequency with 10 µM Oxo-M (p < 0.001, Fig. 5f). The Oxo-M-induced suppression of mEPSC frequency was impaired in slices from DMS-M 4 -KD mice compared with that in WT mice. We also tested Cdk5 function on cholinergic inhibition in WT mice (Fig. 5g). ROSC alone did not alter mEPSCs. When we combined ROSC (100, 300,

Discussion
The present results provide several lines of evidence for M 4 being a major mAChR subtype mediating endogenous cholinergic inhibition of glutamatergic transmission of MSNs in the mouse DMS, and that Cdk5 is necessary for this modulation at the molecular and behavioral levels. Our ndings provide new insight into the mechanism by which Cdk5 modulates M 4 -mediated cholinergic regulation of the striatal direct pathway.
In the present study, by using electrophysiological and pharmacological approaches in conjunction with genetic deletion of M 4 from cultured MSNs and intact brain preparations, we showed that M 4 mainly mediates the cholinergic inhibition of local striatal glutamatergic input. In isolated MSNs in which the M 4 gene was deleted, the inhibitory effect of Oxo-M on mEPSCs was completely absent. It is also worth noting that M 4 deletion completed prevented the Oxo-M-mediated effect on mEPSCs in DMS brain tissue, indicating that M 4 is a key mAChR subtype involved in cholinergic inhibition of glutamate release at the synapse throughout the DMS [28]. This is in line with previous reports and our ndings support M 4 being a major mAChR subtype mediating the cholinergic inhibition of corticostriatal glutamatergic input at corticostriatal synapses [29,30].
We also showed that Oxo-M stimulated Cdk5, p25/p35 immunoreactivity and DARPP-32 Thr75 phosphorylation. The Cdk5 inhibitor, ROSC, suppressed both the Oxo-M-induced increase in Cdk5 and the Oxo-M-induced DARPP-32 Thr75 phosphorylation. Cdk5 is usually associated with its regulatory partners, p35 and p25. p25, a cleavage product of p35, results in hyperactivity of Cdk5 [31,32]. We found that acute Oxo-M treatment concomitantly and signi cantly enhanced p35 and p25 expression in a dosedependent manner in MSNs. More importantly, PCC and MCC quanti cation con rmed a high degree of correlation between M 4 and Cdk5 uorophores in isolated MSNs. Furthermore, pharmacological block of M 4 receptors with MT3, or genetic deletion of M4 in MSNs, abolished Oxo-M-enhancement of Cdk5 and DARPP-32 Thr75 expression. These ndings are consistent with our recent report showing that M 4 modulates DARPP-32 phosphorylation at Thr75 [17]. Interestingly, M 4 receptors are expressed postsynaptically on striatonigral projection neurons (i.e., D 1 -MSNs) and interneurons [33,34]. Thus, our ndings indicate that Cdk5/P35 and DARPP-32 signaling mediate endogenous cholinergic modulation through M 4 in striatonigral neurons. We therefore speculated that Cdk5/p35 activity is required for M 4related physiological processes, directly or indirectly. As predicted, a Cdk5 inhibitor partly abolished the inhibitory effect of Oxo-M on mEPSCs recorded from isolated MSNs, indicating that Cdk5 also contributes to Oxo-M-induced glutamatergic transmission of striatal neurons. These data provide strong evidence that Cdk5 is necessary in cholinergic circuit-level mechanisms of the striatum.
In the present study, we also generated mutant mice lacking mice. This led us to conclude that Cdk5 is the key intracellular signaling mechanism underlying M 4 mAChR modulation in the striatum and various behavioral responses. In addition to this novel modulation of M 4 mAChR in normal physiological functions, pathological roles for Cdk5 have been claimed in the aging and degenerating brain. Cdk5 activity is dysregulated in various brain disorders resulting in profound remodeling of the neuronal cytoskeleton, loss of synapses, and ultimately neurodegeneration [31,32]. Thus, further research is needed to elucidate the complexity of striatal network regulation by M 4 and the Cdk5/P25 signaling pathway, and to dissect its effects on neuronal cytoskeleton remodeling, which might be involved in Parkinson's disease pathogenesis.

Conclusion
Our ndings reveal a novel regulatory mechanism that is critical for Cdk5/P35 and DARPP-32 signaling to mediate endogenous cholinergic M 4 modulation on striatonigral neurons and to affect motor behavior.
Furthermore, we suggest that striatal circuits may, in part, be controlled by M 4 through its modulation of