Parkinson’s disease (PD) is one of the most rapidly growing neurological disorders and is characterized by movement symptoms, such as rigidity, slowness, and tremor (Emamzadeh and Surguchov, 2018; Armstrong and Okun, 2020; Morelli and Pinna, 2023). These motor symptoms are caused by the pathological conditions of dopaminergic neurons in the substantia nigra (SN), which includes α-synuclein (SNCA) aggregation, oxidative stress, ferroptosis, mitochondrial dysfunction, and neuroinflammation (Surmeier, 2018; Dong-Chen et al., 2023). However, since these symptoms vary widely, and their underlying causes have not been clearly identified, they may be not specific to PD, emphasizing the need for an accurate diagnosis (Madabushi et al., 2023). Recent studies have indicated that genetic factors such as SNCA, parkin (PRKN), PTEN-induced kinase 1 (PINK1), leucine-rich repeat kinase 2 (LRRK2), and vacuolar protein sorter-35 (VPS35) play crucial roles in PD (Funayama et al., 2023). Despite substantial progress in genetic development, body fluid diagnostic biomarkers for PD are still lacking (Adam et al., 2023). Consequently, the identification of potential biomarkers suitable for diagnosis continues to pose challenges.

Given that protein-coding genes constitute only 2% of the human genome, non-coding RNAs (ncRNAs), which are transcripts that do not code for any protein, are considered key regulatory molecules in all cellular processes (Ranga et al., 2023). Among the ncRNAs, microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) are potential biomarkers of PD (Kuo et al., 2021). However, a single biomarker cannot fully elucidate the pathological mechanism of PD and fails to attain the high specificity and sensitivity required for accurate diagnosis (Cocco et al., 2023). Numerous studies on PD have proposed integrated biomarkers based on the competing endogenous RNA (ceRNA) hypothesis, which explains the competition between coding RNAs (mRNAs) and ncRNAs for shared miRNAs (Lin et al., 2021). Dysregulation of ceRNA networks has been observed in PD, influencing apoptosis, SNCA misfolding, mitochondrial dysfunction, autophagy, and neuroinflammation, all implicated in PD pathogenesis (Asadi et al., 2023). For example, LINC-p21 alters apoptosis in SH-SY5Y cells by sponging miR-1277-5p (Xu et al., 2018). LINC00943 alleviates neuronal damage in PD models by sponging miR-7-5p and increasing CXCL12 expression (Lian et al., 2021). Furthermore, two ceRNA axes, LINC09238/miR-30c-5p/LRRK2 and LINC001128/miR-30c-5p/ATP13A2, are involved in PD through the regulation of mitochondrial dysfunction (Yousefi et al., 2022). However, these studies have reported different single ceRNA axes, and diagnostic biomarkers for PD remain questionable.

In this study, we constructed an lncRNA–miRNA–mRNA ceRNA regulatory network to reveal the impact of lncRNA–miRNA competition in the blood on SN mRNA. To propose an integrative perspective on PD, we collected several datasets from the Gene Expression Omnibus (GEO) database and analyzed them using various bioinformatics software tools.

Our aim was to construct lncRNA–miRNA–mRNA ceRNA networks according to the flowchart shown in Figure 1.

Recently, an increasing number of studies have identified PD-specific genes in the brain, including SNCA, PRKN, PINK1, and LRRK2. However, it is problematic to use these genes as body fluid diagnostic biomarkers, and the quest for diagnostic biomarkers using body fluids remains challenging. As ncRNAs, including lncRNAs and miRNAs, have emerged as key factors in PD, we focused on the competing regulatory networks between ncRNAs to propose an integrative perspective on PD.

In this study, we constructed two lncRNA-mediated ceRNA networks based on the expression profiles of patients with PD. We screened out common DE mRNAs from the GEO database to investigate genes expressed in the brain SN, as well as DE lncRNAs and DE miRNAs to include genes expressed in plasma. To check their associations with PD, we also demonstrated the difference between PD patients and healthy controls using PCA plots. Subsequently, we constructed two lncRNA–miRNA–mRNA ceRNA regulatory networks to elucidate the effect of plasma lncRNAs on changes in SN mRNA. We conducted KEGG pathway and GO enrichment analyses to comprehend the biological functions of each ceRNA network. Finally, we performed ROC analysis to evaluate the accuracy of ceRNA results as a diagnostic marker.

In the upregulated lncRNA-mediated ceRNA network, 10 lncRNAs were found to regulate the expression of 5 genes in the SN by sponging 4 miRNAs. Several nodes in this ceRNA network have been shown to play crucial roles in PD associated with neuroinflammation. For example, P2RX7 has been widely reported to play a central role in PD models by promoting neuroinflammation (Calzaferri et al., 2020). HSPA1, also known as Hsp70, has been found to be released from damaged cells, acting as a local danger signal (Giuliano et al., 2011). Both SIRT4 and RAD52 have received attention for their therapeutic effects on DNA damage, which has been found to induce inflammatory responses in several neuronal cell types (Wang et al., 2021, 2023). Although these genes have contradictory effects, neuroinflammation induces damage to nerve tissues while promoting the activation of neuroprotective mechanisms (Gelders et al., 2018). In addition, our KEGG pathway and GO enrichment analyses showed that the genes in the upregulated lncRNA-mediated ceRNA network were significantly enriched in nicotinamide metabolism and prostaglandin transport. These pathways have been reported to regulate neuroinflammation during the development of PD (Corwin et al., 2018; Zhou et al., 2018). Therefore, it is reasonable to hypothesize that this upregulated lncRNA-mediated ceRNA network may explain the neuroinflammatory mechanisms in patients with PD.

In a downregulated lncRNA-mediated ceRNA network, 26 lncRNAs appeared to be downregulated, leading to the downregulation of 9 genes in the SN by miR-671. Several nodes in this ceRNA network have been studied for their crucial roles in PD. For example, the brain expression of SLC6A3 (Habak et al., 2014), SYNGR3 (Ho et al., 2023), DDC (Pereira et al., 2023), and SV2C (Dunn et al., 2017) has been shown to modulate dopamine neurotransmission in PD models. Additionally, based on our KEGG pathway and GO enrichment analysis results, genes in the downregulated lncRNA-mediated ceRNA network were significantly enriched in the dopaminergic synapses. Therefore, we speculate that this downregulated lncRNA-mediated ceRNA network plays an important role in patients with PD by regulating dopamine neurotransmission.

Our study had some limitations. First, only one dataset from PD patient blood was available in the GEO database. Second, the functions of the ncRNAs in each ceRNA network have not been previously reported in association with PD, or they showed different expression trends compared to our study. Some studies have indicated that miR-671 is downregulated to play a role in MPP + −induced SK-N-SH cells (Zhang et al., 2022; Hao et al., 2023). However, these studies were conducted at the cellular level, and it remains unclear how miR-671 in the blood affects the SN in the brain. Similarly, the effects of other RNA species on brain regions might also require further investigation. Thus, it is more reasonable to propose that the ceRNA regulatory network might be implicated in the pathogenesis of PD through the function of mRNAs, and that miRNAs and lncRNAs are significant body fluid biomarkers. Further experiments are needed to reveal the exact functions of blood ncRNAs in PD.

Despite these limitations, our study holds significance as it persevered in the analysis of blood datasets, which is in contrast to previous studies that relied solely on bioinformatics tools offering reported ncRNA data. By incorporating analyses of blood samples from patients with PD, future research has the potential to provide a more meaningful ceRNA network. Such efforts are likely to contribute to a more precise elucidation of the pathological mechanisms of PD, ultimately enhancing treatment strategies.

In conclusion, we constructed a ceRNA regulatory network in patients with PD, including 36 lncRNAs, 5 miRNAs, and 14 mRNAs. Our results suggest that these lncRNAs (from the plasma) are involved in PD pathogenesis by sponging miRNAs and regulating gene expression in the SN of the brain. We propose that the upregulated lncRNA-mediated ceRNA network may explain the mechanism underlying neuroinflammation in PD. In addition, the downregulated lncRNA-mediated ceRNA network appeared to contribute to PD by influencing dopaminergic neurotransmission. As our ceRNA network has been found to be associated with PD, the DE miRNAs and lncRNAs in our ceRNA network could be utilized as body fluid diagnostic biomarkers. Our findings provide an integrated view of the mechanisms underlying gene regulation and interactions in PD.

Publicly available datasets were analyzed in this study. This data can be found here: five datasets from SN studies (GSE7621, GSE8397-GPL96, GSE8397-GPL97, GSE20163, and GSE20164) and one dataset from blood studies (GSE160299).

K-YC: Conceptualization, Formal analysis, Investigation, Visualization, Writing – original draft. S-NK: Conceptualization, Supervision, Writing – review & editing.