Neurodegenerative diseases pose a significant global health challenge, characterised by the gradual deterioration of neuronal structure and function. The most common age-related neurodegenerative disease is Alzheimer’s disease (AD), followed by Parkinson’s disease (PD) [1]. Current treatments do not significantly alter the progression of these conditions; therefore, effective therapeutic and preventive interventions are urgently needed [2]. AD is primarily associated with cognitive decline driven by the accumulation of amyloid-beta plaques and neurofibrillary tangles, whereas PD is characterised by motor dysfunction resulting from the degeneration of dopaminergic neurons in the substantia nigra and the accumulation of α-synuclein proteins [3, 4]. Although AD and PD have distinct clinical manifestations and disease-associated proteins, recent evidence indicates that they may coexist in some individuals and share several pathological mechanisms, including oxidative stress, mitochondrial dysfunction, neuroinflammation, and abnormal protein aggregation. These overlapping mechanisms highlight the complexity of neurodegenerative disease progression and suggest the potential value of therapeutic strategies capable of simultaneously modulating multiple biological processes [5].

The biological overlap between AD and PD extends beyond the accumulation of pathological proteins and involves interconnected molecular processes that contribute to neuronal dysfunction and degeneration. Oxidative stress and impaired mitochondrial function can promote neuronal injury and amplify neuroinflammatory responses [6], while chronic activation of microglia and astrocytes can further increase oxidative and inflammatory damage [7, 8]. Furthermore, dysregulation of protein homeostasis and abnormal protein aggregation may interact with mitochondrial and inflammatory pathways, creating a self-reinforcing cycle of neuronal damage [9]. Thus, although AD and PD are distinct disorders, their convergent molecular mechanisms suggest that some biological pathways may contribute to neurodegeneration across both conditions. Understanding these shared mechanisms may provide opportunities for identifying therapeutic candidates capable of modulating multiple disease-relevant pathways simultaneously. This systems-level perspective provides a rationale for applying network pharmacology to investigate potential multi-target interventions for AD and PD.

Plants have consistently remained an important foundation of conventional therapeutic practices and have long served as sources of biologically active compounds. Medicinal herbs contain diverse compounds with potential therapeutic properties, including antitumor, anti-inflammatory, antiviral, and antimalarial activities [10]. Thus, scientific interest in plants as sources of new therapeutic substances for neurodegenerative diseases has increased significantly [10]. Various phytochemicals obtained from plants may influence different stages of neurodegenerative processes, potentially through their roles in maintaining the chemical environment of the brain and modulating the functions of several receptors and signalling pathways [11]. The diverse secondary metabolites present in medicinal plants therefore provide an important source for the discovery of novel therapeutic candidates for specific disorders [10].

The medicinal plant P. amarus is well-known for its many therapeutic uses. Modern medicine is interested in it because of its hepatoprotective, antioxidant, anti-inflammatory, and neuroprotective properties, as well as the presence of several bioactive substances such as saponins, lignans, flavonoids, tannins, and alkaloids [12, 13]. According to research, P. amarus has strong antioxidant activity, with DPPH inhibition reaching 74.4% at optimal doses [14]. Its potential as a natural antioxidant source is further indicated by its significantly high total phenolic content (TPC) and total flavonoid content (TFC) [14]. By potentially reducing oxidative stress, neuroinflammation, and other processes associated with neurodegeneration, P. amarus may contribute to the mitigation of neurodegenerative processes [15]. These multi-target properties suggest its potential relevance to the shared molecular mechanisms underlying AD and PD. Despite these encouraging findings, little is known about the precise molecular pathways underlying its potential benefits, particularly in relation to the shared molecular mechanisms of these neurodegenerative disorders. Therefore, this study explored the potential mechanisms of action of P. amarus in mitigating neurodegenerative processes by investigating shared genes and molecular pathways associated with AD and PD.

The present study provides integrative insights into the shared molecular landscape underlying the comorbidity of AD and PD, with a particular emphasis on the network pharmacology of P. amarus and its potential multi-target therapeutic relevance. The findings reinforce the growing consensus that neurodegenerative disorders are not isolated pathologies but rather interconnected conditions driven by overlapping molecular cascades, including oxidative stress, mitochondrial dysfunction, protein misfolding, and chronic neuroinflammation. These converging pathways have been extensively documented as central to both amyloid-beta and tau pathology in AD, as well as α-synuclein aggregation and dopaminergic neuronal loss in PD, thereby validating the rationale for exploring shared therapeutic targets [19–21].

The identification of bioactive constituents in P. amarus via GC-MS revealed a chemically diverse profile, with fifty-five unique compounds retrieved after deduplication. The GC-MS chromatogram (Figure 1) showed thirteen different peaks, each correlated with five (5) different compounds (Table S2 and S3, Supplementary material). While the P. amarus contains a broad spectrum of bioactive compounds, the application of physicochemical and medicinal chemistry properties, and ADMET screening was essential to narrow these candidates down to those with genuine therapeutic potential for the central nervous system (CNS). Many phytochemicals identified, such as those with low bioavailability (F20), low potential to cross the intestinal wall as indicated by Caco-2, and those that failed to meet the stringent criteria required for crossing the BBB permeability. Specifically, the distribution analysis showed that only six compounds possessed the requisite BBB- and plasma protein-binding properties to potentially reach neurodegenerative target sites. Ultimately, bicyclo[3.3.1]non-6-en-3-ol emerged as the most promising lead candidate, being the only compound to satisfy all metabolism, excretion, and toxicity parameters, including a favorable LD₅₀ of 2,830 mg/kg and no predicted neurotoxicity. This study supports and extends existing CNS drug discovery rules rather than replacing them. Traditional rules emphasize that brain drugs must combine adequate exposure in the CNS (oral absorption, BBB penetration, suitable plasma/brain unbound fractions) with acceptable safety and ADMET profiles [22, 23].

The selection of bicyclo[3.3.1]non-6-en-3-ol was based on systematic computational prioritization rather than previously established neuroprotective activity. Among the 56 curated phytochemicals identified from P. amarus, bicyclo[3.3.1]non-6-en-3-ol demonstrated the most favourable integrated pharmacokinetic profile, satisfying the predefined criteria for oral bioavailability, BBB permeability, intestinal absorption, acceptable metabolic behaviour, clearance, and toxicity. These characteristics were considered particularly important because AD and PD primarily affect the CNS, requiring candidate molecules capable of reaching brain tissue following systemic administration. Accordingly, bicyclo[3.3.1]non-6-en-3-ol was prioritised as the most suitable compound for subsequent network pharmacology analysis. The ADMET criteria were selected to prioritise compounds with favourable oral pharmacokinetic characteristics and predicted CNS exposure, with BBB permeability considered essential for therapeutic candidates targeting neurodegenerative disorders. Human intestinal absorption, CYP-mediated metabolism, systemic clearance, and toxicity predictions were incorporated to reduce the likelihood of selecting compounds with poor drug-like properties (Tables 2–6). Importantly, these criteria served as a prioritization strategy rather than definitive evidence of biological efficacy. Furthermore, the therapeutic activity of P. amarus is unlikely to be attributable to a single constituent, as herbal extracts typically exert their pharmacological effects through multiple bioactive compounds acting on interconnected molecular targets. Therefore, the prioritisation of bicyclo[3.3.1]non-6-en-3-ol should not be interpreted as excluding potential contributions from other phytochemicals. Rather, it represents a computationally prioritised candidate for further investigation, while synergistic interactions among the remaining constituents warrant future experimental evaluation.

Bicyclo[3.3.1]non-6-en-3-ol has a total of 187 target genes that code for specific target proteins, and 7,735 common genes shared between AD and PD were obtained from the GeneCards database. From the list of genes, we created a Venn diagram (Figure 2) using the Venny 2.0 application. The intersection corresponds to the genes that code for proteins potentially targeted by bicyclo[3.3.1]non-6-en-3-ol. These 137 genes were matched with bicyclo[3.3.1]non-6-en-3-ol, and a compound-target network was constructed using Cytoscape as shown in Figure 3. We also constructed the PPI network to determine important proteins in AD & PD that can be targeted by the compound. The PPI network (Figure 4) was constructed from the 137 targets using the STRING database. The MCODE plug-in of Cytoscape was also utilized to determine protein clusters. The results of the ranking of each node based on centrality measures are summarized in Table 8. The top genes that are pivotal to the comorbidity of AD and PD targeted by bicyclo[3.3.1]non-6-en-3-ol include STAT3, PPARG, HSP90AB1, HIF1A, ESR1, GRB2, PTGS2, FYN, NFKB1, and TLR4. The centrality gives an estimate of how important a particular node is and is commonly measured using degree, betweenness, and closeness centrality. The degree gives a rough estimate of the centrality, but is only a local measure and does not take into consideration the rest of the network. Hence, betweenness and closeness centralities were also included in the ranking of the pivotal proteins. The closeness centrality measures the average of all the shortest distances of the other nodes from a given node, whereas the betweenness centrality measures the number of shortest paths passing through a node. The higher the value of these centrality measures, the more important the node is [24, 25].

Functional enrichment of the 137 shared AD-PD genes targeted by bicyclo[3.3.1]non-6-en-3-ol revealed a convergent molecular profile involving inflammation, oxidative/nitrosative stress, intracellular signaling, neuronal function, and metabolic regulation (Figure 5A–D). The biological-process analysis was dominated by cellular responses to nitrogen- and oxygen-containing compounds, inflammatory responses, chemical stimuli, and regulation of cell communication and signal transduction (Figure 5A). This pattern is biologically relevant because oxidative stress and chronic neuroinflammation are interconnected mechanisms implicated in both AD and PD [26], where persistent inflammatory signaling can exacerbate oxidative injury, synaptic dysfunction, and neuronal loss [24, 27]. The enrichment of receptor complexes, membrane microdomains, neuronal cell bodies, axons, somatodendritic compartments, synapses, and vesicular structures in the cellular-component analysis (Figure 5B) further indicates that the predicted targets are distributed across neuronal signaling and membrane-associated systems, rather than being confined to inflammatory pathways alone.

The molecular-function profile further supports this interpretation, with strong enrichment of estrogen response element binding, nuclear receptor activity, ligand-activated transcription factor activity, protein/kinase binding, oxidoreductase activity, and signaling receptor functions (Figure 5C). These findings suggest that the shared targets participate in transcriptional regulation, redox homeostasis, and receptor-mediated signaling, providing potential links between cellular stress and neuronal responses. Of particular interest is the enrichment of oxidoreductase-related functions and responses to oxygen- and nitrogen-containing compounds, which collectively point toward redox dysregulation as an important component of the shared AD-PD molecular landscape [28].

The KEGG analysis provided additional evidence for a metabolic-inflammatory-neuronal axis, with insulin resistance, prolactin signaling, sphingolipid signaling, Th17-cell differentiation, NOD-like receptor signaling, phospholipase D signaling, calcium signaling, neuroactive ligand-receptor interaction, and metabolic pathways among the most enriched pathways (Figure 5D). The particularly high enrichment of insulin resistance and sphingolipid signaling is noteworthy because impaired brain insulin signaling and altered sphingolipid metabolism have been implicated in both AD and PD, where they may influence glucose utilization, mitochondrial function, inflammation, synaptic integrity, and neuronal survival [29, 30]. Thus, these findings extend the potential mechanism of bicyclo[3.3.1]non-6-en-3-ol beyond conventional neuroinflammation and suggest that metabolic and lipid dysregulation may constitute an important component of the shared molecular pathology targeted by the compound. The enrichment of calcium signaling and neuroactive ligand-receptor interaction additionally supports possible involvement of neuronal communication and excitability [31].

The inflammatory component of the enrichment profile is reinforced by the representation of NOD-like receptor, Th17-related, and other immune pathways, together with the high connectivity of inflammatory hub proteins in the PPI network. The ten most highly connected proteins were STAT3, PPARG, HSP90AB1, HIF1A, ESR1, GRB2, PTGS2, FYN, NFKB1, and TLR4, with STAT3 showing the highest degree of 50 (Figure 6). These hubs collectively connect inflammatory transcription, metabolism, cellular stress, receptor signaling, prostaglandin production and innate immune recognition. STAT3 is particularly relevant because JAK2/STAT3 signaling integrates inflammatory cytokine signaling with cellular survival and has been implicated in neurodegenerative disease mechanisms [32]. Similarly, PPARG provides a potential connection between metabolic regulation and inflammation, while HIF1A is consistent with the enrichment of oxygen- and metabolic-stress responses [33, 34]. PTGS2, NFKB1 and TLR4 further reinforce the inflammatory component, with TLR4-mediated signaling capable of converging on NF-κB and other inflammatory pathways implicated in neurodegeneration [35, 36]. The presence of FYN and GRB2 additionally links the hub network to receptor-associated and tyrosine-kinase signaling [37, 38], whereas ESR1 and HSP90AB1 suggest involvement of transcriptional regulation and cellular protein-stress responses [39].

Integratively, the enrichment and PPI analyses suggest that the potential relevance of bicyclo[3.3.1]non-6-en-3-ol in AD-PD comorbidity may arise from its predicted intersection with interconnected inflammatory, oxidative, metabolic, lipid, receptor-signaling, and neuronal processes, rather than from modulation of a single target. The convergence of insulin resistance and sphingolipid signaling with inflammatory hubs such as STAT3, PTGS2, NFKB1, and TLR4 is particularly compelling because metabolic dysfunction, lipid abnormalities and neuroinflammation can interact to promote neuronal dysfunction in both diseases. Nevertheless, these findings remain hypothesis-generating: enrichment and PPI analyses demonstrate statistical and network associations but do not establish that bicyclo[3.3.1]non-6-en-3-ol directly activates or inhibits the identified pathways or hub proteins. Experimental target-engagement, cellular, and in vivo studies will therefore be required to validate the proposed multi-target neuroprotective mechanism.

Although direct experimental evidence for the neuroprotective activity of bicyclo[3.3.1]non-6-en-3-ol is currently unavailable, the compound belongs to the bicyclo[3.3.1]nonane family, a privileged structural scaffold found in numerous bioactive natural products and synthetic molecules with broad pharmacological activities, including anti-inflammatory, antioxidant, antimicrobial, and anticancer effects [40]. Importantly, several bicyclo[3.3.1]nonane derivatives have demonstrated neuroprotective efficacy in experimental models of AD, cerebral ischemia, and oxidative stress through mechanisms involving activation of the Nrf2/HO-1 pathway, suppression of inflammatory signalling, modulation of cholinergic neurotransmission, and AMPA receptor potentiation [41, 42]. These observations support the biological plausibility of investigating bicyclo[3.3.1]non-6-en-3-ol as a novel candidate while recognising that its specific neuroprotective activity remains to be established experimentally.

Despite providing valuable insights into the shared molecular mechanisms underlying AD and PD and identifying bicyclo[3.3.1]non-6-en-3-ol as a promising multi-target candidate, this study has several limitations. First, the findings are based entirely on computational approaches, including network pharmacology, target prediction, and ADMET analyses, and therefore require validation through molecular docking, molecular dynamics simulations, in vitro assays, and in vivo studies to confirm the predicted target interactions and therapeutic effects. Second, the identified protein targets and disease-associated genes were obtained from publicly available databases, which are dependent on existing annotations and predictive algorithms and may contain incomplete or inaccurate information. Third, the pharmacokinetic and toxicity profiles, including BBB permeability, oral bioavailability, and metabolic properties, were derived from in silico predictions and should be experimentally verified before therapeutic translation. Finally, although multiple phytochemicals were identified in P. amarus, the present study focused primarily on bicyclo[3.3.1]non-6-en-3-ol and did not evaluate potential synergistic interactions among the plant’s bioactive constituents, which may contribute to its overall neuroprotective activity. Accordingly, further experimental and preclinical investigations are warranted to validate these findings and establish their translational relevance.