Parkinson’s Disease (PD) is a progressive neurodegenerative disorder characterized by the loss of dopaminergic neurons in the substantia nigra and α-synuclein aggregation forming Lewy bodies [1,2]. Patients present with motor symptoms (bradykinesia, rigidity, tremor, postural instability) and non-motor symptoms including depression, anxiety, and autonomic dysfunction [3].

Deep brain stimulation (DBS) is an established surgical treatment for advanced PD [4], while spinal cord stimulation (SCS) has been investigated for specific symptoms, particularly gait impairment [3,5]. Vibroacoustic therapy has also been explored as an adjunctive intervention for improving motor symptoms [6]. More recently, focused ultrasound (FUS) has emerged as a clinically approved non-invasive treatment option for selected patients with advanced PD [7].

Although levodopa remains the first-line therapy, its efficacy over time does not halt disease progression, highlighting the need for novel therapeutic approaches that target underlying neurodegenerative mechanisms [6,8,9,10]. PD has a multifactorial etiology involving complex genetic, environmental, and biochemical interactions [11,12,13]. Systems biology approaches, particularly those incorporating nutrigenomics, are essential for understanding these interactions and identifying novel biomarkers and therapeutic targets by integrating multi-omics data [11,14,15,16,17,18].

In addition, the utilization of sophisticated computational models provides deeper insight into the complex processes involved in PD, paving the way for new therapeutic approaches that go beyond traditional paradigms [11,13]. Traditional Chinese Medicine (TCM) adopts a holistic, multi-targeted paradigm consistent with systems biology principles [19,20]. TCM constituents can simultaneously modulate multiple signaling cascades and gene networks relevant to PD [21,22]. Network pharmacology approaches enable exploration of TCM active constituents and their biological targets, providing mechanistic insights for therapeutic development [23,24,25].

However, despite the established strengths of a systems-level viewpoint and the multitarget capabilities of TCM, a systematic, integrative nutrigenomic systems biology analysis of how PD-related gene expression patterns directly correspond to predicted TCM treatments is a significant literature gap [26,27]. Published Gene Expression Omnibus (GEO) sets offer a powerful tool for discovering robust and reproducible gene signatures across diverse patient groups, yielding a set of differentially expressed genes with high confidence in PD pathology. Future integration of this genomic information with enrichment software such as DAVID, GeneMANIA, and specific TCM databases, e.g., BATMAN-TCM 2.0, can help close the knowledge gap between molecular processes and therapeutic interventions. The research design was thus to conduct an integrative systems-biological study, with the first step being to identify shared DEGs across PD patient samples obtained from GEO databases. These genes were then characterised by functional enrichment and network analyses. Lastly, a nutrigenomic tool, employing TCM network pharmacology, was used to forecast possible specific herbs and active constituents with the potential to adjust the identified core PD gene network, which provided new evidence-based insights into TCM-based therapeutic interventions in PD. While individual epigenetic or miRNA alterations have been implicated in PD, an integrated model combining repressive histone modifications with specific miRNA-mediated repression remains underexplored. This study addresses this gap by suggesting a candidate dual regulatory axis that may contribute to dopaminergic neurodegeneration, with TCM predictions providing complementary hypothesis-generating leads.

The overall workflow consisted of: (1) data acquisition and DEG identification using GEO2R from four GEO datasets, (2) identification of core overlapping DEGs via Venn analysis, (3) functional enrichment with DAVID and Enrichr, (4) network analysis with GeneMANIA, and (5) TCM network pharmacology using BATMAN-TCM 2.0. A schematic summary of the workflow will be provided as Figure S1. Full lists of DEGs from each dataset and GEO2R analysis parameters are provided in Supplementary Materials. All analyses were performed using the publicly available web-based tools described; no custom scripts were generated.

This paper used a systematic bioinformatics methodology that combined various publicly available microarray datasets of post-mortem brains of PD patients to identify robustly core differentially expressed genes (DEGs) that relate to the pathogenesis of PD. Our meta-analysis converged on a signature of six highly significant, consistently downregulated genes: TH, SLC18A2, EN1, RET, PCSK1, and PCDH8. Follow-up functional enrichment and consistently implicated them in the dopaminergic synapse and the translocation of synaptic vesicles, establishing their significant role in the neurochemical phenotype of PD. Furthermore, we identified candidate epigenetic regulators and predicted TCM compounds that may interact with these core genes, providing additional insights into PD molecular mechanisms and generating hypotheses for future therapeutic investigation. Downregulation of the vesicular monoamine transporter 2 (VMAT2), which is encoded by the gene SLC18A2, and of tyrosine hydroxylase (TH) is a critical event in the neurobiological processes that drive Parkinson disease, and specifically in the substantia nigra. The storage and release of dopamine in the synaptic vesicles depends on the transporter called SLC18A2 (VMAT2). A significant decrease in VMAT2 levels in PD results in a decrease in the ability to store dopamine in vesicles. This decreases dopamine availability and exacerbates its cytosolic toxicity, leading to increased oxidative stress and neurodegeneration in the substantia nigra [39]. Moreover, animal models with VMAT2 deficiency have motor impairment and pathology similar to human PD [40]. The accumulation of misfolded proteins, including α-synuclein, adds to the toxicity of reduced VMAT2 levels, causing progressive neuronal loss [41]. Enhancement of VMAT2 expression has been proposed as a potential neuroprotective strategy [42]. The primary contribution of this study is the identification of a candidate dual regulatory axis-combining repressive histone modifications (H3K27me3 and H3K9me3) with specific miRNA targeting- that may cooperatively contribute to silencing key neuroprotective and dopaminergic genes in PD. This integrated model extends beyond individual regulatory layers reported previously. It suggests that concurrent repressive marks may establish a closed chromatin state while miRNAs (e.g., hsa-miR-431-3p on EN1; mmu-miR-362-5p on SLC18A2 and PCDH8) provide additional post-transcriptional repression. Such synergy offers a plausible mechanistic hypothesis for the robust downregulation observed across independent cohorts and highlights potential high-order control points for therapeutic intervention. This dual axis aligns with emerging evidence of epigenetic-miRNA crosstalk in neurodegeneration but adds to the growing systems-level evidence in PD substantia nigra linking these layers to a concise set of core dopaminergic targets.

The core signature identified here shows strong overlap with prior PD transcriptomic studies, particularly for TH and SLC18A2, which are among the most consistently downregulated genes in the substantia nigra across multiple meta-analyses [43,44,45,46,47]. EN1 has been implicated in dopaminergic neuron maintenance and vulnerability in both human and mouse studies [48]. RET and PCSK1 appear in several meta-analyses as downregulated genes, while PCDH8 was recently proposed as a potential biomarker [46]. The consistent co-downregulation of all six genes across our four independent datasets suggests they may function as part of a coordinated module in PD pathogenesis, even if some members are less frequently reported as top hits individually.

TH is the rate-limiting enzyme of dopamine production; the downregulation of TH plays a role in the depletion of dopamine. It has been shown that the expression of TH in the substantia nigra of PD patients is vastly reduced, which is associated with the loss of dopaminergic cells, as well as the classic motor symptoms of the condition. This under-regulation has been attributed to symptomatic manifestations such as bradykinesia and rigidity [49,50]. Degradation of both TH and VMAT2 would suggest more extensive dysregulation of dopaminergic signaling pathways [51,52].

Engrailed-1 (EN1) is an essential transcription factor used in survival and maintenance of dopaminergic neurons. EN1 defects have been associated with the gradual degeneration of these neurons and are a prerequisite for the continued survival of dopaminergic neurons in adult mice. EN1 heterozygous mouse model has a severe and progressive degeneration of dopaminergic cells in the substantia nigra [53,54,55]. EN1 preserves dopaminergic neurons against apoptosis and increased expression correlates with protective mechanisms against mitochondrial apoptosis [56,57]. EN1 has also been implicated in pathways associated with α-synuclein biology [58]. RET enhances the effects of GDNF on RET neurotrophic signaling and may enhance the resilience of dopaminergic neurons in PD [59,60].

Proprotein convertase Subtilisin/Kexin Type 1 (PCSK1) is involved in the processing of pro-neurotransmitters and pro-neuropeptides [61] and has been identified as a candidate gene in PD [26]. Protocadherins, including PCDH8, are involved in synaptic adhesion and stability [62]; PCDH8 may contribute to impaired synaptic stability and neuronal vulnerability in PD [26,46].

H3K27me3 and H3K9me3 are known as repressive histone marks in the substantia nigra that contribute to the establishment of closed chromatin and inactive transcription states [63]. Studies have demonstrated cooperative repressive functions of H3K27me3 and H3K9me3 during chromatin silencing [63]. These marks have been found to increase in PD neurons and are associated with decreased expression of neuroprotective genes and activation of apoptotic pathways [64,65].

Bioinformatic analysis predicts that hsa-miR-431-3p targets the engrailed homeobox 1 (EN1) gene. More broadly, altered miRNA regulation has been widely implicated in PD [66], and hsa-miR-431-3p has been reported among dysregulated miRNAs in PD cerebrospinal fluid [67]. Altered expression of circulating miRNAs, including potential dysregulation of hsa-miR-431-3p, has been observed in PD and may serve as a source of diagnostic biomarkers [67,68].

mmu-miR-362-5p is predicted to target both PCDH8 and SLC18A2 (VMAT2), which is consistent with the critical role of SLC18A2 in dopaminergic function [12,69]. Altered miRNA expression, including potential dysregulation of hsa-miR-431-3p, has been observed in PD and may serve as a source of diagnostic biomarkers [69].

The core signature shows strong enrichment in dopaminergic synapse function, synaptic vesicle loading, and neuronal structure, consistent with the central pathophysiology of impaired dopamine neurotransmission in PD.

The core gene signature is central to dopaminergic neuron dysfunction in the substantia nigra. The reduced dopamine machinery involves Tyrosine Hydroxylase (TH) and Vesicular Monoamine Transporter 2 (SLC18A2/VMAT2), which are strongly downregulated. Their downregulation leads to reduced dopamine availability, impaired vesicular storage, and increased cytosolic dopamine toxicity, contributing to oxidative stress and neurodegeneration [69,70]. EN1, a critical transcription factor for dopaminergic neuron survival, is also consistently downregulated, with its depletion worsening cell loss and apoptosis [12]. RET, PCSK1, and PCDH8 further contribute to neurotrophic signaling, neuropeptide processing, and synaptic stability.

The enrichment analysis supports the conclusion that the identified gene changes are focused on synaptic transmission. The term with the highest enrichment level (Fold Enrichment: 2164.22) in the biological process category is amine-neurotransmitter loading into synaptic vesicle. The cellular-location analysis concentrates the genes in neuronal structures, especially the dendrite (Fold Enrichment: 37.27), axon, and synaptic vesicle. In addition, KEGG pathway analysis reveals that the Dopaminergic synapse and Parkinson disease pathways are significantly enriched. TCM network analysis provides both graphical and statistical support for the high degree of co-expression and direct functional connections among the core genes TH, SLC18A2, and EN1.

The main pathway that may contribute to the repression of neuroprotective and dopamine-related genes involves chromatin remodeling and post-transcriptional regulation. Consistent with previous studies demonstrating cooperative repressive functions of H3K27me3 and H3K9me3 [63,64], the concurrent enrichment of these marks may facilitate the establishment of a closed chromatin state in PD neurons. The concurrent enrichment of these marks may facilitate the development of closed chromatin and non-transcriptional states in PD neurons. In particular, H3K27me3 can target RET, PCSK1, PCDH8, SLC18A2, and EN1. Besides changes in the chromatin, certain microRNAs (miRNAs) are also predicted to have a direct action on the core DEGs; that is, hsa-miR-431-3p is predicted to target the neuroprotective transcription factor EN1 (Combined Score: 2580.41), which also leads to the loss of dopaminergic neurons by disrupting the balance in the expression of neurons to survive. On its own, mmu-miR-362-5p is predicted to target both the dopamine storage gene, SLC18A2, and the adhesion protein PCDH8.

Importantly, the BATMAN-TCM tool predicts potential binding interactions based on structural similarity but does not indicate the functional direction of the effect (i.e., whether a compound activates, inhibits, or has no net effect on the target). This is particularly relevant because several top-enriched compounds, such as reserpine and rotenone, are known VMAT2 inhibitors that can induce parkinsonism. Therefore, these results should be regarded strictly as hypothesis-generating.

To prioritize promising candidates for future experimental testing, we propose the following criteria: (1) exclusion of known VMAT2 inhibitors and neurotoxins; (2) presence of prior literature demonstrating neuroprotective effects in PD cellular or animal models; (3) favorable toxicity and safety profiles; and (4) potential to modulate multiple components of the dopaminergic-epigenetic-miRNA network.

Collectively, these TCM predictions serve as hypothesis-generating leads for future multi-target nutrigenomic strategies in PD, rather than direct therapeutic recommendations. Rigorous experimental validation is essential before any translational development.

A key consideration is that BATMAN-TCM predictions reflect target interaction potential rather than directional effects. Several top-enriched compounds (e.g., reserpine from Rauwolfia species, rotenone) are established VMAT2 inhibitors and dopaminergic toxins, which validate the tool’s sensitivity for SLC18A2/TH but warrant caution in direct therapeutic interpretation. Future studies should prioritize experimental screening to identify activators or neuroprotective modulators among lower-ranked or herb-derived mixtures.

This integrative nutrigenomic systems-biology study identifies a highly robust core set of six differentially expressed genes central to Parkinson’s disease pathophysiology and, most importantly, suggests a candidate dual epigenetic-miRNA regulatory axis that may coordinately contribute to their silencing. The concomitant enrichment of repressive histone marks (H3K27me3 and H3K9me3) with specific post-transcriptional repression predicted for hsa-miR-431-3p (predicted targeting of EN1) and mmu-miR-362-5p (predicted targeting of SLC18A2 and PCDH8) provides an integrated systems-level mechanistic model for the sustained downregulation of neuroprotective and dopaminergic genes in the substantia nigra.

Functional enrichment analysis showed a strong correlation between these genes and impairment of the dopaminergic system, particularly aminergic neurotransmitter packaging into synaptic vesicles and dopaminergic synapse function. The significant downregulation of tyrosine hydroxylase (TH) and the vesicular monoamine transporter 2 (SLC18A2/VMAT2) appears to be a leading factor in reduced dopamine storage and increased cytosolic toxicity in the substantia nigra.

A key mechanistic insight of this study is the identification of a candidate dual regulatory network involving overlapping repressive histone marks (H3K27me3 and H3K9me3) and specific microRNAs. This joint epigenetic and microRNA dysregulation likely contributes to the transcriptional silencing observed in PD neurons.

In addition, network pharmacology analyses using BATMAN-TCM identified TCM-derived candidates predicted to interact with core targets (TH and SLC18A2). These findings provide hypothesis-generating leads for future multi-targeted nutrigenomic approaches. Overall, this integrative framework illustrates how nutrigenomics can reveal nutrient-gene interactions that influence epigenetic and microRNA-mediated regulation in neurological disorders.