Parkinson’s disease (PD) is a degenerative disease of the central nervous system (CNS) characterized by progressive motor dysfunction. This condition results from insufficient activation of the nigrostriatal dopaminergic pathway and overactivation of the corticostriatal glutamatergic pathway, leading to abnormal excitatory postsynaptic potentials (EPSPs) (Zhao et al., 2022; Hou et al., 2017). In PD, the degeneration of dopaminergic neurons in the substantia nigra disrupts dopamine-mediated regulation of glutamate release from corticostriatal neurons, causing excessive glutamate release and overactivation of N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, which lead to neurodegeneration and motor dysfunction (Wang et al., 2022; Dong et al., 2009; Meredith et al., 2009). Current treatments for neurodegenerative diseases often have side effects, while exercise training offers a non-invasive, non-pharmacological alternative that has been shown to alleviate motor and cognitive symptoms without adverse effects. It may also prevent extracellular concentrations of glutamate (Glu) and preserve glutamatergic excitatory imbalance, thereby exerting neuroprotective effects on nigrostriatal neurons in PD (Shi et al., 2019; Petzinger et al., 2011). Furthermore, it has been reported that exercise training improves PD symptoms, balance, and independence in daily activities, along with enhancing walking abilities such as walking speed, stride length, stride frequency, and total walking distance (Palasz et al., 2019; Goodwin et al., 2008; Chen et al., 2018).

In glutamatergic regulation, glutamate (Glu), the most abundant neurotransmitter in the central nervous system, acts on ionotropic receptors such as N-methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate (KA) and also on metabotropic G protein-coupled receptors, all of which are crucial for normal brain function. In contrast, glutamatergic overactivation triggers neuronal damage in PD (Dietrich et al., 2005; Garcia et al., 2010). Neurodegeneration in PD animals increases the density of glutamatergic immunolabeling in nerve terminals of striatal medium spiny neurons (MSNs) (Fisher et al., 2004). AMPA receptors mainly manage Na⁺ flux and some Ca²⁺, while NMDA receptors mainly manage large quantities of Ca²⁺. Glutamate-induced elevation of cytosolic Ca²⁺ levels activates protein kinases, such as calcium/calmodulin-dependent protein kinase II (CaMKII) and Cav1.3, which is abundant in medium spiny neurons of the striatum. This elevation also triggers transcription factors such as cyclic AMP response element-binding protein (CREB), leading to increased glutamatergic hyperactivity and neurodegeneration in PD animal models (Chen et al., 2017; Mattson, 2008; Garcia et al., 2010). Metabotropic glutamate receptors (mGluRs), particularly striatal mGluR5 and mGluR2/3, are promising targets for PD treatment due to their roles in modulating presynaptic glutamate release, intracellular Ca²⁺ levels, and neuroprotection. Interventions targeting these receptors have shown potential in alleviating PD symptoms and reducing dendritic spine loss in the striatum of PD animal models (Shi et al., 2019; Garcia et al., 2010).

Damage to dopaminergic neurons in the substantia nigra reduces dopamine levels in the striatum, which, in turn, leads to overactivation of the glutamatergic pathway, causing an excitatory imbalance that exacerbates neural dysfunction (Chen et al., 2017). The glutamate transporter GLT-1, the ion channel affinity protein GLAST, and EAAC1 play a crucial role in maintaining homeostasis by removing excess Glu from the synaptic cleft in the striatum of PD animal models (Feng et al., 2021; Arnold and Salvatore, 2016).

Exercise training exerts neuroprotective effects by modulating the molecular mechanism and is beneficial for symptomatic recovery. The literature supports the effects of exercise training on motor recovery but presents conflicting evidence regarding the recovery of the glutamatergic pathway within the nigrostriatal pathway in animals with PD (Sconce et al., 2015). This systematic review is the first to summarize the effects of exercise training on regulating the glutamatergic pathway and its receptor interactions in the nigrostriatal pathway in animal models of Parkinson’s disease (PD).

The search yielded a total of 96 articles from PubMed (30), the Web of Science (1), and Embase (65). After removing 12 duplicates, 51 ineligible studies—those not related to Parkinson’s disease or with irrelevant titles—were excluded during the title screening. The abstracts of 33 studies were then screened, and 19 studies were removed due to irrelevant study designs or a lack of focus on glutamatergic pathways and exercise training. A total of 14 studies underwent full-text screening, with 6 studies excluded for the following reasons: full text not accessible (n = 1), lack of a well-defined exercise training regimen (n = 2), and failure to measure relevant outcomes (n = 3). Eight studies met the criteria and were included. After screening the references of these eight studies, one additional article was retrieved, bringing the total to nine studies (Figure 1).

All studies included rodent models with male animals aged between 6 and 12 weeks. The studies employed pharmacologically induced PD models, with 55.6% using the 6-OHDA model in rats and 45.4% using the MPTP model in mice. Regarding the exercise interventions, 89% of the studies implemented treadmill training, while 11% employed voluntary wheel running. The treadmill training regimen typically involved running at a speed greater than 10 m per min for 30–60 min per session, 3–5 days per week, for 4 weeks (Table 1).

The regulation of protein and gene expression involved in the nigrostriatal glutamatergic pathway and motor function following exercise training is summarized in Table 1. Motor function assessments revealed that the Parkinson’s disease (PD) group exhibited impairments such as decreased latency and total time on the balance bar, reduced coordination on the suspension test, and poorer performance on the body swing test. In contrast, the PD + Exercise (Ex) group showed improvements in these metrics, with increased latency and total time on the balance bar, enhanced coordination on the suspension test, and better performance on the body swing test (Feng et al., 2021). In addition, the PD group showed reduced total distance traveled, walking time, and velocity, whereas these metrics were significantly improved in the PD + Ex group (Shi et al., 2019; Sconce et al., 2015; VanLeeuwen et al., 2010). In the open field test (OFT), the PD group exhibited decreased total distance traveled and walking time, both of which improved following exercise training (Chen and Li, 2019). The cylinder test indicated that exercise training reduced walking asymmetry in the PD + Ex group compared to the PD group (Chen et al., 2017). Furthermore, the PD + Ex group showed increased stride frequency and stride length, both of which were found to be reduced in the PD group (Churchill et al., 2017).

In terms of nigrostriatal glutamatergic regulation, the extracellular Glu levels were elevated in the PD group but significantly reduced in the PD + Ex group (Chen and Li, 2019; Feng et al., 2021; Shi et al., 2019). The astrocytic proteins, including GLT-1 and GS, were reduced in the PD group, limiting the conversion of Glu to glutamine (Gln), whereas their expression was upregulated in the PD + Ex group, thereby increasing the Gln levels (Feng et al., 2021; Churchill et al., 2017; Sconce et al., 2015; VanLeeuwen et al., 2010). In the presynaptic terminals of the glutamatergic neurons, the PD group exhibited increased VGULT1 expression, which was reversed in the PD + Ex group (Sconce et al., 2015). The PD group showed decreased expression of mGluR2 and mGluR3, whereas the PD + Ex group showed increased expression of these proteins in the presynaptic terminals of the glutamatergic neurons, thereby limiting Glu release into the synaptic cleft (Chen and Li, 2019; Shi et al., 2019).

The PD group showed increased mGluR5 expression in the postsynaptic terminal, enhancing Glu influx into the postsynaptic membrane and causing excitotoxicity to the glutamatergic neurons, while the PD + Ex group showed decreased mGluR5 expression (Shi et al., 2019). The AMPA receptor subunit modulation also revealed key differences: while GluA1 expression increased and GluA2 and GluA3 decreased in the PD group, the PD + Ex group exhibited reduced expression of GluA1 and increased GluA2 and GluA3 (VanLeeuwen et al., 2010; Kintz et al., 2013; Garcia et al., 2017). Concurrently, NMDA receptors, Cav1.3, p-CaMKII (Thr286), and Arc, which were elevated in the PD group and contributed to hyperactivation and excitotoxicity, were reduced in the PD + Ex group (Chen et al., 2017; Garcia et al., 2010). However, one study presented contradictory findings, reporting increased expression of GLT-1, EAAC1, and GLAST in the PD group, with reductions observed in the PD + Ex group (VanLeeuwen et al., 2010).

The methodological quality of the included studies, as assessed using the CAMARADES checklist, ranged from 4/10 to 6/10. All studies used PD animal models, adhered to compliance standards, and were published in peer-reviewed journals. A total of five of the nine studies reported temperature control and randomization, while seven of the nine studies addressed conflicts of interest. However, none of the studies provided information on allocation concealment, blinded assessments, anesthetic compliance, or sample size calculations (Table 2).

Our findings demonstrated that exercise training exerted a neuroprotective effect on the glutamatergic pathway by modulating the activity of the glutamatergic neurons within the nigrostriatal pathway of Parkinson’s disease (PD) animal models, potentially reducing excitatory imbalances associated with dopaminergic depletion. Exercise training increased glutamate (Glu) reuptake from the synaptic cleft by enhancing the expression of the astrocyte protein glutamate transporter (GLT-1). This was followed by an increase in the expression of glutamate synthase (GS), which converted Glu to glutamine (Gln), subsequently releasing it into the extracellular space, where it was reabsorbed into the presynaptic terminal of glutamatergic neurons in the nigrostriatal pathway in PD animal models. Exercise training also limited Glu release into the synaptic cleft via VGULT1 and increased the expression of inhibitory mGluR2/3 receptors on the presynaptic membrane within the nigrostriatal pathway in PD animal models. It also reduced the excitotoxic effects on the glutamatergic neurons by decreasing the expression of the excitatory mGluR5 receptor in the nigrostriatal pathway in PD animal models. Exercise training effectively modulated ionotropic glutamate receptors, including AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) subunits and NMDA (N-methyl-D-aspartate), which limited long-term potentiation and synaptic hyperactivity. It restored normal synaptic transmission in glutamatergic neurons within the nigrostriatal pathway in PD animal models (Figure 2). Exercise training increased PD latency; improved motor function, balance, and coordination; enhanced total distance, walking time, walking velocity, stride frequency, stride length, and reduced walking asymmetry in PD animal models compared to the non-exercised PD animal models. CAMARADES scores of 4–6/10 indicated moderate methodological quality in the included studies.

Our study consistently demonstrated that exercise training increased the reuptake of Glu from the synaptic cleft in the nigrostriatal region of the MPTP- and 6-OHDA-induced PD animal models and improved motor function (Feng et al., 2021; Chen and Li, 2019; Shi et al., 2019). Exercise training facilitated the reduction of extracellular Glu levels, limited abnormal burst firing, and restored normal synaptic transmission in the nigrostriatal glutamatergic pathway in PD animal models (Ferreira et al., 2021). A previous study reported that treadmill training for more than 4 weeks decreased the nerve terminal density of striatal medium spiny neurons, possibly reflecting a reduction in Glu levels in the synaptic cleft, and improved the running velocity of animals in MPTP-induced PD models (Fisher et al., 2004). Increased Glu levels in the extracellular space in PD lead to an increase in spontaneous excitatory postsynaptic currents (sEPSCs) in striatal medium spiny neurons (MSNs), causing hyperexcitation and damage to glutamatergic neurons. However, exercise training reduces sEPSCs in MSNs, protecting against damage and improving motor behavior (Zhao et al., 2022; Wang et al., 2022). Exercise training was found to improve the transmission of dopaminergic neurons and simultaneously normalize glutamate-driven spikes in the GABAA-mediated current through GABAA receptors and calcium channel blockers in the striatum, ameliorating spatiotemporal gait disturbances in the PD animal model (Chen et al., 2018; Hou et al., 2017).

Extracellular fluid in the central nervous system (CNS) does not contain Glu-metabolizing enzymes, so the removal of extracellular Glu relies primarily on inward uptake by transporters such as GLT-1, which terminates the synaptic effects of Glu and prevents the potential excitotoxic accumulation of extracellular Glu. Elimination of GLT-1 has been linked to dyskinesia and motor dysfunction in the striatum of animal models of PD (Farrand et al., 2016). In our study, exercise training facilitated the removal of Glu from the synaptic cleft by astrocytic GLT-1. After being converted to glutamine by the GS protein within the astrocytes, glutamine was subsequently transported to the presynaptic membrane. Along with the action of VGULT, this process limited Glu release into the synaptic cleft within the nigrostriatal pathway in PD animal models, thereby reducing MSN damage, exerting neuroprotective effects, and improving motor behavior (Feng et al., 2021; Churchill et al., 2017; Sconce et al., 2015).

According to our study findings, voluntary exercise training for 4 weeks downregulated the astroglial proteins EAAC1 and GLAST, which have an affinity for the Na + channels to reuptake Glu from the synaptic cleft within the nigrostriatal region of the MPTP-induced PD models. Although the decreased expression of these proteins raises questions about the neuroprotective effects, the study still demonstrated improvements in motor function (Sconce et al., 2015). In contrast, another study indicated that treadmill training for 4 weeks increased the expression of EAAC1 and GLAST in the substantia nigra, although not in the striatum, and improved motor function by increasing the total distance traveled and walking time in 18-month-old male Brown–Norway/Fischer 344 F1 hybrid rats (Arnold and Salvatore, 2016).

Our results also reported that 4 weeks of treadmill training significantly modulated the AMPA receptors. Specifically, the decreased expression of GluA1 limited long-term potentiation, while the increased expression of GluA2/3 enhanced ionic interactions with Na⁺ and Ca²⁺, impaired Mg²⁺ influx, and restored normal synaptic transmission in the nigrostriatal pathway in PD animal models. These molecular changes were associated with improved motor function, as reflected by increased total distance traveled, walking time, stride length, and stride frequency (VanLeeuwen et al., 2010; Kintz et al., 2013). However, one study reported similar results after 4 weeks of treadmill training, while conflicting results were observed after 16 weeks in the nigrostriatal region of PD models. The authors attributed this to the maintenance of homeostasis (Garcia et al., 2017). Another study supported the modulation of AMPA receptors following 4 weeks of voluntary exercise training but found no significant effects on NMDA receptors in the cortex of animals with neurodegenerative diseases (Dietrich et al., 2005).

Cav1.3 subunits, abundant in medium spiny neurons (MSNs), are crucial for excitation-transcription coupling with NMDA receptors in the striatum of animals with PD (Picconi et al., 2004). Upon Cav1.3 activation, intracellular Ca²⁺ levels increase, leading to auto-phosphorylation of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) at Thr286. This process is associated with hyperactivity in the corticostriatal glutamatergic pathway in the dopamine-depleted striatum (Mohanan et al., 2022). This study demonstrated that 4 weeks of treadmill training decreased the Cav1.3 subunits, inhibited NMDA receptors and auto-phosphorylation of CAMKII at Thr 286, decreased intracellular Ca2+ in the striatum, and improved walking asymmetry in the 6-OHDA-induced PD rat models (Chen et al., 2017).

This systematic review is limited by the relatively small number of studies that specifically examined the impact of exercise training on the glutamatergic pathway and its receptor interactions in animal models of Parkinson’s disease. The studies only included male animal models and primarily focused on gene and protein expressions within the nigrostriatal pathway, limiting the scope of the findings. The review did not address interactions with medium spiny GABAergic neurons. Due to a lack of relevant studies, it did not explore the glutamatergic pathway in regions such as the cortex-basal ganglia or other brain areas. In addition, this review did not include postsynaptic excitatory potentials, which could have provided a more comprehensive understanding of the glutamatergic system’s role in neuroprotection and motor function recovery. The included studies did not report on cognitive function in animals with PD.

Exercise training modulated the glutamatergic pathway and exerted a neuroprotective effect in the nigrostriatal pathway in Parkinson’s disease (PD). It enhanced extracellular glutamate reuptake through astrocyte proteins such as GLT-1 and supported the conversion of glutamate to glutamine (Gln) via glutamine synthetase (GS). Exercise training also reduced glutamate release through VGluT1 and increased the expression of inhibitory mGluR2/3 receptors. In addition, at the postsynaptic terminal, it down-regulated excitatory mGluR5 receptors and modulated ionotropic AMPA and NMDA receptors, preventing calcium overload and abnormal glutamatergic excitation. Exercise training is a promising therapeutic approach for improving motor deficits and mitigating glutamatergic dysregulation in PD.