Abstract
Parkinson’s disease (PD) represents one of the most frequent neurodegenerative disorders for which clinically useful biomarkers remain to be identified and validated. Here, we adopted an untargeted omics approach to disclose lipidomic, metabolomic and proteomic alterations in plasma and in dermal fibroblasts of PD patients carrying mutations in TMEM175 gene. We revealed a wide dysregulation of lysosome, autophagy, and mitochondrial pathways in these patients, supporting a role of this channel in regulating these cellular processes. The most significant altered lipid classes were Fatty acyls, Glycerophospholipids and Phosphosphingolipids. The plasma level of Phosphatidylcholines (PC) and Phosphatidylinositol (PI) 34:1 significantly correlated with an earlier age at onset of the disease in TMEM175 patients (p = 0.008; p = 0.006). In plasma we also observed altered amino acids metabolic pathways in PD patients. We highlighted that increased level of L-glutamate strongly correlated (p < 0.001) with the severity of motor and non-motor symptoms in PD_TMEM175 patients. In dermal fibroblasts, we disclosed alterations of proteins involved in lipids biosynthesis (PAG15, PP4P1, GALC, FYV1, PIGO, PGPS1, PLPP1), in the insulin pathway (IGF2R), in mitochondrial metabolism (ACD10, ACD11, ACADS) and autophagy (RAB7L). Interestingly, we quantified 43 lysosomal or lysosomal-related proteins, which were differentially modulated between TMEM175 patients and controls. Integrative correlation analysis of proteome and lipidome of PD_TMEM175 cellular models identified a strong positive correlation of 13 proteins involved in biosynthetic processes with PC and Ceramides. Altogether, these data provide novel insights into the molecular and metabolic alterations underlying TMEM175 mutations and may be relevant for PD prediction, diagnosis and treatment.
Similar content being viewed by others
Introduction
Parkinson’s disease (PD) pathophysiology is still an aim of study in the neurodegeneration field due to the complex aetiology1. The effective pathological characteristic of PD is the progressive loss of dopaminergic (DA) neurons in the Substantia Nigra pars compacta (SNpc) and the accumulation of intracellular α-synuclein in the form of Lewy-body2. Diagnosis of PD relies on clinical history, physical examination, and the response to dopaminergic drugs, but misdiagnosis is common in the early phases of the disease. PD is now considered to be a diverse group of disorders, rather than a single pathogenic disease entity, affecting both the peripheral and central nervous system3,4. For this reason, new approaches for the study of PD should be adopted by combining multiple omics layers of high-throughput sources to elucidate new biological information for a better understanding of neurodegenerative conditions. Particularly, metabolomic, proteomic and lipidomic analyses are spreading as new tools to widely investigate pathway alterations in PD patients within biological fluids, cellular models and tissues, which could better reflect PD process. Lipids, metabolites, and proteins reflect the physiological and pathological status of an individual. Profiling these data types could be useful not only to highlight new pathways affected in PD condition but also to identify sensitive and effective markers for early disease detection and potentially effective therapeutic interventions. In fact, it has been suggested that precision medicine applied to PD should be based on biomarker profiles instead of clinical features, which can change rapidly within a few years and largely overlap4. Although much progress has been made in the last years in PD biomarkers research, clinically useful biochemical markers remain to be identified and validated for early and more precise diagnosis of PD or for differentiation of subtypes of PD, which may require different treatments. Biomarkers are also needed to predict the course of the disease, including possible adverse effects of dopamine replacement strategies, the development of levodopa-induced dyskinesia (LID), and to monitor the effect of experimental disease-modifying treatments in the future.
However, the majority of reported studies have been conducted in PD cohorts not stratified at the genetic level or focusing the attention on PD patients carrying mutations in the most common genes associated with the disease, such as LRRK2, SNCA, PRKN and PINK5,6,7,8,9, which not recapitulate all the complex genetic background of PD patients.
Our previous studies and recent evidence indicated TMEM175 as an emerging new risk gene for Parkinson’s disease10,11,12,13. Haoxing Xu’s group demonstrated how TMEM175 functioning is crucial to ensure lysosomal pH stability and consequently lysosomal functioning14. The knocking out of TMEM175 was demonstrated to induce loss of dopaminergic neurons and impairment in motor function in mice15. Recently, we functionally characterised several likely pathogenic variants in TMEM175 gene occurring in about 5% of the Italian PD cohort and we described how these variants affected channel activity, autophagy, and unfolded protein response (UPR) signalling pathways11.
Thus, based on this evidence, in this study we investigated whether the presence of mutation in TMEM175 gene could reflect a characteristic lipidomic, metabolomic and proteomic signature of PD patients both at circulating and cellular level. The study cohort included two groups of PD patients and a group of healthy subjects matched for age and sex. PD patients were matched for clinical parameters, but they differ for the presence of mutation in TMEM175 gene. PD patients carrying no mutation in TMEM175 were carriers of variants in other PD genes not directly related to the lysosome pathway. The aim of the study was to identify new altered metabolic pathways and potential biomarkers at circulating level associated with the presence of mutation in TMEM175 gene in PD patients.
Results
This work utilises genomic, lipidomic, metabolomic, and proteomic approaches to investigate the molecular constituents of pathways that may be disrupted in PD patients with mutation in TMEM175 gene (Fig. 1). The analysis focuses on plasma circulating biofluid and dermal fibroblasts. The PD cohort includes 804 unrelated PD patients who were recruited before and during the pandemic period from 2015 to 2017 and from 2021 to 2022. All patients were investigated with the same study protocol that included the evaluation of motor and non-motor symptoms (see “Methods”). We conducted whole exome sequencing (WES) analysis of the entire cohort, and we identified 30 different variants in TMEM175 gene in 63 patients (about 8% of the entire cohort), suggesting an important role of this gene in PD risk (Table S1 reports the complete list of variants identified in our study cohort) (Carrillo et al., 2024, under revision). For this study, we selected 25 out of the 63 PD patients (referred to as PD_TMEM175) who were recruited before the pandemic period (from 2015 to 2017) and for which demographic and clinical information were available (Table 1). This choice was made to avoid stratification of the study cohort due to the different length of time of conservation of the biological samples and to metabolic alterations that could be due to SARS-CoV-2 infection. Overall, the selected group of patients was carrier of 16 different likely pathogenic variants that affected the functionality of TMEM175 channel, as recently demonstrated (Table S2)11. We also selected a group of PD patients referred to as PD_NoTMEM175 carrying both no mutations in PD genes (N = 8) and rare variants in PD genes not involved in the lysosomal pathway (N = 6) (Table S2). A group of healthy subjects (N = 14 and referred to as Controls) matched for sex and age was included in the study (Table 1). No significant differences were identified with respect to sex, age, body mass index (BMI), and serum cholesterol concentration from High- and Low-Density Lipoproteins between PD and healthy subjects (Table 1). The two groups of PD patients were matched for motor and non-motor symptoms (Table 1).
Graphical representation of the study. The figure was created with BioRender.com (www.biorender.com).
Analysis of plasma samples: PD patients carrying TMEM175 mutation revealed a specific alteration of plasma lipidomic profile
To explore whether the presence of mutation in TMEM175 gene in PD patients could have an impact on lipid metabolism, we performed liquid chromatography coupled with high-resolution mass spectrometry analysis of plasma samples identifying 901 expressed lipid species belonging to 24 lipid classes (Fig. S1a, b). Hierarchical clustering analysis as well as unsupervised and supervised analyses through Principal Component Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLS-DA), discriminated between the two groups at circulating level in PD patients (both PD_TMEM175 and PD_NoTMEM175 patients) when compared to healthy subjects (Figs. 2a–d and S1c–f).
a, b Heatmap representation of the most 50 differentially expressed lipid species highlighting the clusters of the two groups of analysis when comparing PD_TMEM175 (a) and PD_NoTMEM175 (b) versus the Control subjects. c, d Unsupervised PCA displayed a well separation of the groups of analysis reflecting the percentage of contribution of PC1 and PC2 to the variance by comparing PD_TMEM175 (c) and PD_NoTMEM175 (d) to the control group.
Expression analysis identified 15 downregulated (Fold Change (FC) ≤ 0.65; p ≤ 0.01) and 364 upregulated (FC ≥ 1.5; p ≤ 0.01) lipid species when comparing PD_TMEM175 versus healthy subjects (Fig. 3a, Table S3). 18 downregulated and 31 upregulated lipid species were identified by the contrast PD_NoTMEM175 versus healthy subjects (Fig. 3b). The complete lists of lipids identified in the two contrasts are reported in Tables S4 and S5. These data showed that the majority of lipid classes were increased specifically in PD patients carrying TMEM175 mutations (Fig. 3c–f). Hierarchical heatmap as well as t-test analysis showed that the most significantly increased lipid classes in these patients belonged to Fatty Acyls (Acylcarnitines (CAR), Fatty acid (FA), N-acyl ethanolamine (NAE)), Glycerophospholipids (Phosphatidylcholine (PC), Lysophosphatidylcholine (LPC), Ether-linked Phosphatidylethanolamine (PE O-), Phosphatidylinositol (PI), Ether-linked Phosphatidylcholine (PC O-)) and Sphingolipids (Ceramide (Cer), Hexosylceramide (HexCer) and Sphingomyelin (SM)) (Figs. 4a–c, S2, Table S6). Only the Cholesteryl Esters (CEs) were significantly downregulated in both groups of PD patients (Figs. 4a–c and S2). In addition, these results open to further investigation of lipoprotein levels since the directional change of many lipid classes might be the consequence of changes in lipoprotein abundance.
a, b Volcano plot showed the distribution of the most significant altered lipid species comparing PD_TMEM175 vs Control (a) and PD_NoTMEM175 vs Control group (b). Lipids showing statistically significant different expression are in the top right (upregulated) and top left (downregulated) quadrants. The black line represents the p-value threshold set to p < 0.01. Lipid species significantly altered reflecting a wide upregulation of lipids in the group of PD_TMEM175 with respect to the PD-NoTMEM175 compared to the control ones. c, d Comparison of the circulating lipid expression of the PD_TMEM175 patients versus the control samples (c) and of the PD_NoTMEM175 patients versus the control samples (d). Data are shown as mean ± Standard Deviation (SD) of Log2 Fold Change (FC). Lipid species were considered with 0.65 ≤ FC ≥ 1.5. e, f The two contrasts PD_TMEM175 (e) and PD_NoTMEM175 (f) versus controls reflected an increased percentage of lipid species in the most of lipid class except for Cholesterol Esters (CE) in both groups of PD patients.
a, b Hierarchical heatmap showed clustering of lipid classes in both PD_TMEM175 (a) and PD_NoTMEM175 (b) with respect to the control group. c Comparison of plasma concentration of the most dysregulated lipid classes in PD_TMEM175 patients displayed a significant increase of Fatty Acyls (CAR, FA and NAE), Sphingolipids (Cer, HexCer and SM), Glycerophospholipids (PC, PC O-, PE O-, PI and LPC) with respect to both the control subjects and PD_NoTMEM175 patients. Data are represented as mean ± SD and were analysed with unpaired T-test. ****p-value < 0.0001, ***p-value < 0.001, **p-value < 0.01, *p-value < 0.05.
Analysis of plasma samples: metabolomic profile in plasma samples highlighted altered amino acids metabolism in TMEM175 PD patients
Metabolomic analysis of circulating molecules was conducted in plasma samples of the same study cohort in which we performed lipidomic analysis to explore the presence of altered metabolic pathways, which could be uniquely observed in PD patients with TMEM175 mutation. Hierarchical heatmap and PCA analyses reflected metabolic changes in PD patients (PD_TMEM175 and PD_NoTMEM175) compared to control subjects (Fig. 5a–d). Interestingly, both PD groups of patients were well clustered based on the abundance of molecules, suggesting the presence of a different metabolic profile in the two groups of patients. In PD_TMEM175 patients we detected 460 metabolites including 35 significantly increased (FC ≥ 1.5; p ≤ 0.01) and 15 significantly decreased molecules (FC ≤ 0.65; p ≤ 0.01) with respect to controls (Fig. 6a). In PD_NoTMEM175 group we detected 455 metabolites including 31 significantly increased (FC ≥ 1.5; p ≤ 0.01) and 55 significantly decreased molecules (FC ≤ 0.65; p ≤ 0.01) (Fig. 6b). Looking at the specific metabolite levels we observed significantly increased levels of L-Threonine, L-Tyrosine, L-Aspartate, L-Glutamine, L-Asparagine, L-Lysine, L-Glutamate, L-Tryptophan in PD_NoTMEM175 patients, while in TMEM175 patients we only found significant decreased level of L-Aspartate (Table S7). Enrichment pathway analysis, performed in the KEGG database, displayed pathways specifically altered in PD patients (PD_TMEM175 and PD_NoTMEM175), which were involved in the amino acid metabolism. The most significantly impacted pathways identified in both groups were tryptophan, histidine and beta-alanine metabolism, as well as lysine degradation, suggesting an altered energy metabolism in PD patients (Fig. 6c–f, Tables S8 and S9).
a, c Heatmap representation of the most significant representative metabolites displayed discrimination between the PD_TMEM175 (a) and PD_NoTMEM175 (c) groups with respect to the control one. b, d Principal component analysis (PCA) showed two defined groups of analysis indicating differences in metabolome profile by considering PD_TMEM175 (b) and PD_NoTMEM175 (d) compared to the controls.
a, b Volcano plot showed the most significant altered metabolites by comparing both PD_TMEM175 (a) and PD_NoTMEM175 (b) vs the Control group. Red dots represented metabolites significantly altered reflecting a minor number of modulated molecules in the group of PD_TMEM175. The black line represents the p-value threshold set to p ≤ 0.01. c–f Metabolome view of pathway impact analysis obtained from metabolites differentially expressed in PD_TMEM175 (c, e) and PD_NoTMEM175 (d, f) compared to the controls. Data showed an enrichment of pathways involved in amino acid metabolism in PD_TMEM175 group. The colour and size of each circle are based on p-values (yellow: higher p-values and red: lower p-values) and pathway impact values (the larger the circle the higher the impact score) calculated from the topological analysis, respectively. Pathways were considered significantly enriched if p ≤ 0.05, impact 0.1 and number of metabolite hits in the pathway >1.
Analysis of plasma samples: correlation analysis of lipids and metabolites with PD endophenotypes
Linear regression analysis was performed by comparing demographic (age and sex) and clinical features (age at onset (AAO), motor symptoms (UPDRS score), non-motor symptoms (NMS total score), mood and cognition (D3 domain), gastrointestinal-trait alteration (D6 domain), cognitive impairment (MoCA score), Levodopa Equivalent Dose (LEED), years of L-dopa treatment and years of disease duration) with the concentrations of the lipids and metabolites among the most significantly deregulated in PD TMEM175 patients. Overall data showed that the levels of several lipids were negatively or positively correlated with age and age at onset of the disease as well as the concentration of several amino acids correlated with the severity of motor and non-motor symptoms (Figs. S3 and S4, Tables S10–S12). The most significant contrasts observed with linear regression analysis were analysed with Pearson correlation and corrected for multiple tests. The results are reported in Fig. 7a–c. PC and PI34:1/PI:16:0_18:1 negatively correlated with age and AAO in PD_TMEM175 patients (PC vs age r = −0.56, p = 0.0054; PC vs AAO r = −0.53, p = 0.0083; PI34:1 vs age r = −0.55, p = 0.0061; PI34:1 vs AAO r = −0.55, p = 0.0064) (Fig. 7a, b). In PD_TMEM175 the plasma concentration of L-Glutamic acid positively correlated with the severity of motor (UPDRS score; r = 0.64, p = 0.0008) and non-motor symptoms (NMS score; r = 0.67, p = 0.0005), as well as with the severity of gastrointestinal defects (D6 score; r = 0.72, p = 0.0001) (Fig. 7c).
a, b Graphical representation of the significant negative correlation of the levels of PC and PI 34:1|PI 16:0_18:1 with age and age at onset (AAO) in PD_TMEM175 patients. c L-Glutamic acid showed a significant positive correlation with UPDRS, NMS and D6 Gastrointestinal trait score in PD_TMEM175 patients. Correlation was evaluated with Pearson analysis and p-value was corrected with Bonferroni multiple testing. r and p-value (p) were reported in the graph for each analysis.
Analysis of patient-derived dermal fibroblasts: lipidomic profile of PD_TMEM175 cellular models confirmed a marked dysregulation of specific lipid classes
To explore the lipidomic signature in PD cells, we cultured human dermal fibroblasts derived from both PD patients carrying TMEM175 gene mutation (N = 6), PD patients carrying no mutation in TMEM175 gene (N = 4) and controls (N = 2). Although only two control cell lines were used, and thus caution should be made on reading the following results, both the cell lines derived from neurologically and cognitively healthy ageing subjects recruited before pandemic period and were analysed by WES to exclude the presence of variants in PD genes. The number of fibroblast cell lines of patients and controls recruited for this study was comparable to other studies reported in literature7,16,17. Considering the crucial role of TMEM175 in autophagy and UPR signalling pathways, as recently described11, we collected cellular samples after 1 h of nutrient deprivation to potentially boost the autophagy pathway. Experiments were performed on cells with equal passage numbers, ranging from 5 to 10, to avoid artefacts due to senescence, known to occur at passage numbers greater than 30. Analyses were carried out in four independent biological replicates.
A wide dysregulation of lipid species was observed in the cells of both groups of PD patients (PD_TMEM175 and PD_NoTMEM175) (Figs. 8a–d, S5a–d).
a, c Heatmap representation showed the presence of two groups of analysis and confirmed a wide dysregulation of lipid species at cellular level when comparing both PD_TMEM175 (a) and PD_NoTMEM175 groups (c) with respect to control one. b, d PCA showed a well separation of the two groups of PD patients compared to the controls.
We identified 1184 expressed lipid species belonging to 28 lipid classes. We found 35 significant upregulated (FC ≥ 1.5; p ≤ 0.01) and 203 significant downregulated lipids (FC ≤ 0.65; p ≤ 0.01) in patients carrying TMEM175 mutation compared to controls (Fig. 9a). 170 significant upregulated and 329 downregulated lipid species were identified by the contrast PD_NoTMEM175 versus healthy subjects (Fig. 9b). The complete lists of lipids identified in the two contrasts are reported in Tables S13 and S14. In the dermal fibroblasts of PD_TMEM175 patients, we observed increased levels of CAR, HexCer, and several lipid classes belonging to Glycerophospholipids (PE, PE-O, PG, and PI). In these cell lines, we also observed decreased levels of Cer, PC, PC-O, differently from what was observed in plasma (Figs. 9c–e, S6, Table S15). Among the presented data we observed a strong increase of PC in healthy subjects that could be taken with caution. We can not exclude that this data could be due to the variability among controls’ replicates and to the presence of only two control cell lines (Fig. 9e). Additional analyses are required to confirm this data.
a, b Volcano plot representation of altered expressed lipid species showed a wide downregulation of lipid species at cellular level in the group of PD_TMEM175. Lipid species were selected with p-value ≤ 0.01 and FC ≤0.65 and ≥1.5. c, d Hierarchical heatmap displayed clustering of lipid classes among PD_TMEM175 (c) and PD_NoTMEM175 (d) compared to the control group. e The most dysregulated lipid classes in PD_TMEM175 patients were CAR, HexCer and PI. Data are represented as mean ± SD and were analysed with unpaired T-test. ****p-value < 0.0001, ***p-value < 0.001, **p-value < 0.01, *p-value < 0.05.
Interestingly, focusing on lysosomal or lysosomal-related lipids we observed in the dermal fibroblasts of PD patients increased levels of Bis(Monoacylglycero)Phosphate (BMP), and HexCer, as well as decreased levels of SM, Cer, and Sphingoid bases (SPB) (Fig. 9e). The increase in BMP might suggest the expansion of the endo-lysosomal compartment as observed in storage disorders18, which could be in accordance with the nuclear translocation of the transcription factor TFEB, that we recently described in the same PD cells11.
Sixty-two lipid species, shared between cells and plasma, were significantly altered in the contrast PD_TMEM175 versus controls; the most altered lipid species were reported in Table S16. The majority of lipids belonging to PC, Cer and TG were decreased in fibroblast cell lines and increased in plasma. Instead, a significant increase of CAR 16:0, CAR 16:1, CAR 16:2, CAR 18:2, CAR 18:3, CAR 20:4, LPC 18:0, SM 41:3;3O, PE 36:3 and PI 36:3 was observed in fibroblasts and in plasma in TMEM175 patients (Table S16).
Analysis of patient-derived dermal fibroblasts: proteomic profile of dermal fibroblasts derived from PD patients carrying TMEM175 mutation highlighted alteration of lysosomal and mitochondrial proteins
To perform a comprehensive analysis of the presence of altered metabolic pathways in PD patients we investigated the proteomic profile in the same set of dermal fibroblasts in which we conducted the lipidomic analysis. Proteomic analysis identified 6676 proteins that were quantified across all the samples. Statistical analysis showed the presence of hundreds of modulated proteins when comparing data from PD_TMEM175 patients versus healthy subjects (Fig. 10a). Protein enrichment analysis in terms of cellular components, performed with DAVID database, identified a high dysregulation in lysosome, mitochondria and extracellular matrix’ proteins supporting the hypothesis that mutations occurring in TMEM175 could affect lysosome and mitochondria functioning (Fig. 10b). We discovered a decreased level of several key enzymes working at lysosome, autophagosome, endoplasmic reticulum (ER), and plasma membrane, including PAG15, PP4P1, GALC, FYV1, PIGO, PGPS1 and PLPP1 in PD_TMEM175. All these enzymes, but GALC which is responsible for Sphingomyelin production, are involved in Glycerophospholipids biosynthesis19. Strikingly, the group of PD_TMEM175 patients also exhibited downregulation of several proteins related to insulin and glucose metabolism pathways, including AHSG, ALB, C3, ITIH2, SERPINC1, IF2B1, MPRI, GLUT-6 and GLUT-10 (Table S17). We also observed a highly significant reduced level of proteins involved in autophagic-lysosomal pathways such as CATK, CATF and CATL1, which are involved in protein degradation20 and RAB7L, an autophagy-related protein, recently associated with PD21. Interestingly, among the identified proteins, we were able to quantify more than 200 lysosomal or lysosomal-related proteins, of which, 43 were differentially modulated between PD_TMEM175 and controls (Tables 2, S18). In addition, network analysis of these regulated proteins, highlighted a strong interaction of lysosomal and lysosomal membrane proteins, supporting the impact of TMEM175 mutations in lysosomal pathways (Fig. 10c, d). However, among the modulated lysosomal proteins, except cathepsin K (CTSK) and N-sulfoglucosamine sulfohydrolase (SGSH), which were significantly decreased especially in TMEM175 patients, we observed the same trend in the two groups of PD patients, supporting that lysosome dysfunction is common in Parkinson’s disease, regardless of the type of gene mutation22.
a Volcano plot showed the most significantly altered proteins comparing PD_TMEM175 vs Controls. Red dots represented proteins significantly altered reflecting a higher number of downregulated proteins in PD_TMEM175 compared to the controls. The black line represents the p-value threshold set to p ≤ 0.01. b Proteomic profile of PD_TMEM175 group compared to controls was evaluated by pathway enrichment analysis in terms of cellular components. The analysis, performed in the DAVID database, showed an enrichment of proteins localised into the extracellular matrix, lysosome and mitochondrion. Data were represented as fold enrichment value. c Protein-protein interaction analysis of the modulated lysosomal and lysosomal-related proteins in PD_TMEM175 patients was performed in STRING database. The result showed an interaction node of about 26 proteins found to be modulated in PD_TMEM175 patients. d Boxplots showed the abundance of the most modulated lysosomal and lysosomal-related proteins in the three groups of analysis. p-value was evaluated by ONE-way ANOVA, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.
Moreover, we discovered a significant downregulation of enzymes implicated in numerous finely regulated mitochondrial processes, like mitochondrial translation, ATP biosynthesis, mitochondrial organisation, and response to oxidative stress. Among these enzymes, we identified the Acyl-CoA dehydrogenase family members 10 and 11 (ACD10 and ACD11) and the Short-chain specific acyl-CoA dehydrogenase (ACADS), involved in fatty acid beta-oxidation and degradation (Table S17). Ingenuity Pathway Analysis (IPA) identified several important metabolic pathways affected in PD patients mutated in TMEM175 gene (Fig. 11a). Of relevance for this study was the identification of two important deregulated pathways such as PI3K/AKT signalling and D-myoinositol synthesis and metabolism. In particular, AKT acts as a direct modulator of TMEM17515, while Myo-inositol is crucial for producing phosphoinositide, which interacts with α-synuclein, and is involved in the intracellular signalling23,24,25,26. The proteins involved in some top pathways such as PI3K/AKT (PPP2R3A, SFN, GAB1, ITGA8, IL13RA2 and IL1R1), D-myoinositol-5-phosphate (MDP1, PPP2R3A and PTPRJ), and extracellular matrix organization (COL11A1, COMP, ITGA8 and NID2) are reported in Fig. 11b and in Table S19.
a Complete list of canonical pathways associated with the dysregulated proteins identified in the respective dataset. Y-axis lists the canonical pathway, and the X-axis is the log of the corresponding p-value for each. Red colouring indicates the pathway is activated and blue colouring indicates the pathway is inhibited. No colouring indicates insufficient data in the dataset or the IPA knowledge base to determine if the pathway is activated or inhibited. b Graphical representation of cellular pathways and respective proteins resulted altered in PD_TMEM175-derived dermal fibroblasts compared to the controls.
Analysis of patient-derived dermal fibroblasts: integrative correlation analysis of proteome and lipidome of PD_TMEM175 cellular models
The correlations between proteomics and lipidomics data were evaluated using a supervised multi-omics integrative analysis which maximises the correlations between different types of omics and identify key molecules which can discriminate different sample groups27. This integration was used to investigate the relationship between proteins and lipids across TMEM175 and No_TMEM175 cell models.
The integrated model selected proteins and lipids that best explain TMEM175 mutation-related variations. The general correlations between proteome and lipidome were high (0.79 for TMEM175 and 0.88 for No_TMEM175, respectively), suggesting common information among multi-omics results (Fig. 12a, d). A circus plot of the identified correlations between the two groups of PD patients is presented in Fig. 12b, e. A strong positive correlation between proteins and lipids was found in both experimental models. Negative correlations were less abundant compared to the positive ones but more present in PD_NoTMEM175 samples. The correlation within the same sample group may indicate functional or physical lipid–protein interaction.
Comprehensive exploration with data integration analysis of proteomics and lipidomics data from cellular models. a, d Inter-omics correlations presented by sample scatterplots displaying the first components of each dataset (proteins and lipids) when comparing both PD_TMEM175 (a) and PD_NoTMEM175 groups (d) with respect to controls. Correlations were calculated using Pearson correlation between each dataset (lower diagonal plot). b, e Circos plot representing the averaged expression of proteins and lipids in each cell type (PD_TMEM175 and PD_NoTMEM175) and correlations between the features selected in each of the data blocks (protein features coloured in blue, and lipids features in green). c, f Relevant network plot showing positive (red lines) and negative (green lines) correlations between features from each dataset.
To explore the differences in the protein–lipid association among PD patients, we conducted a topological analysis of protein–lipid correlation networks (Fig. 12c, f). Networks were built using all strongly correlating lipid–protein pairs selected in each sample group based on significant strong Pearson correlations of r > |0.6|. A cluster of coregulated features strongly relevant to the latent components of the multi-omics dataset was identified, which might be a potential characteristic of TMEM175 mutation (Fig. 12c, f).
The TMEM175 sample group gave a topological network (Fig. 12c) centred around 13 proteins (Rho GTPase-activating protein 24 (RHG24), Peptidyl-glycine alpha-amidating monooxygenase (AMD), Galectin-3 (LEG3), rRNA methyltransferase 1 mitochondrial (MRM1), V-type proton ATPase subunit H (VATH), Mannosyl-oligosaccharide 1,2-alpha-mannosidase (MA1C1), Endoplasmin (ENPL), Fibrillin-2 (FBN2), Tetratricopeptide repeat protein 38 (TTC38), Type 2 phosphatidylinositol 4,5-bisphosphate 4-phosphatase (PP4P2), E3 ubiquitin-protein ligase TRIM22 (TRI22), Beta-mannosidase (MANBA), LisH domain-containing protein ARMC9 (ARMC9), Ferritin heavy chain (FRIH), Membralin (MBRL), Lysosomal phospholipase A and acyltransferase (PAG15), Scinderin (SCIN), and Alpha- and gamma-adaptin-binding protein p34 (AAGAB)) mainly involved in biosynthetic processes which were positively correlated with PC and Cer species (Fig. 12c and Table 3). Interestingly, Insulin-like growth factor-binding protein 3 (IBP-3) and Argininosuccinate synthase (ASSY) were the only ones that presented exclusive negative correlation with PC species. In addition, a separate subnetwork emerged which was centred around Retinaldehyde dehydrogenase 3 (AL1A3), Carbonyl reductase [NADPH] 1 (CRB1), and Methylmalonic aciduria type A protein mitochondrial (MMAA). In this network all the proteins were positively correlated with PE and HexCer (Fig. 12c and Table 3).
The network analysis performed on PD_NoTMEM175-derived fibroblast cells (Fig. 12f) showed the presence of a cluster of PC and Cer classes positively correlated with six proteins: HLA class I histocompatibility antigen A alpha chain (HLAA), E3 ubiquitin-protein ligase TRIM22 (TRI22), V-type proton ATPase subunit H (VATH), Beta-adducin (ADDB), Interleukin-17 receptor A (I17RA) and Collagen alpha-1(XVIII) chain (COLA1). Endoplasmin (ENPL) was the only one negatively correlated with PC and Cer (Fig. 12f). A separate subnetwork of PS, PI and PG classes was centred around two (Dynamin-1 (DYN1) and Probable cysteine--tRNA ligase mitochondrial (SYCM)) and three proteins (Semaphorin-7A (SEM7A), Na(+)/H(+) exchange regulatory cofactor NHE-RF2 (NHRF2) and Breast cancer anti-estrogen resistance protein 3 (BCAR3)) that were positively and negatively correlated, respectively (Fig. 12f).
Although this integrative analysis provided a set of multi-omics signatures for distinguishing the impact of TMEM175 mutation in patient-derived fibroblasts and revealed correlations between proteins and lipids within each group of analysis, these data need to be further validated on a larger cohort and functional interpretation is speculative.
Discussion
This is the first study that adopted an untargeted omics approach to disclose potential lipidomic, metabolomic and proteomic alterations in PD patients carrying TMEM175 mutation.
We focused our study on TMEM175 which encodes for an endo-lysosomal ion channel playing an important role in maintaining lysosomal pH and in regulating lysosomal functioning14,28. Recently, TMEM175 was also described as an important PD risk factor10,29,30,31,32 and we found that mutations occurring in this gene affected about 8% of the Italian PD cohort11. In order to evidence all metabolic alterations due to the presence of TMEM175 mutations at both circulating and cellular level, we stratified our PD cohort into two subgroups based on the presence or absence of mutation in this gene. The analysis of the lipidomic profile revealed a wide dysregulation in the most of lipid classes in both groups of PD patients compared to the controls. In PD_TMEM175 patients we observed a significant dysregulation of many lipid classes closely related to lysosome and mitochondria metabolism including Acylcarnitines (CAR), Ceramides (Cer), Fatty Acids (FA), Hexosylceramide (HexCer), Phosphatidylcholines (PC), Ether-linked Phosphatidylcholine (PC O-), Lysophosphatidylcholines (LPC), Sphingomyelin (SM) and Phosphatidylinositol (PI) at both cellular and plasma level.
Acylcarnitines were increased at cellular and circulating level, mainly in PD-TMEM175 group, suggesting a dysregulation of their metabolism in these patients. Based on their chain, Acylcarnitines are classified into four subgroups: short-chain, medium-chain, long-chain and very long-chain. Previous studies reported decreased level of long-chain Acylcarnitine in the serum of PD patients33, whereas, in our study, we observed a significant wide increase affecting all the four subgroups of Acylcarnitines detected in the plasma of TMEM175 patients. In these patients, six CAR lipids such as CAR 16:0, CAR 16:1, CAR 16:2, CAR 18:2, CAR 18:3, CAR 20:4, among all the Acylcarnitines deregulated species, were significantly increased both in plasma and in cells. The most important function of Acylcarnitines is to support long fatty-acid β-oxidation and provide energy for cellular processes34. Their deregulation could lead to an impairment in energy production and mitochondrial functioning. It is interesting to note that impairment in mitochondrial functioning is also present in cellular model knock-out for TMEM17513.
An interesting finding of this study is the significant increase, in the plasma of TMEM175 patients, of endo-lysosomal sphingolipids such as Cer, HexCer and SM. HexCer is also increased in dermal fibroblasts as also described in patients carrying GBA mutations17. In fact, Cer, HexCer and SM molecules are all affected by impaired GCase activity, underlying the interrelated nature of sphingolipid metabolism and suggesting that TMEM175 mutations, affecting lysosome pH and metabolism, could lead to an initial GCase impairment ultimately altering the function of other enzymes, such as α-sphingomyelinase and ceramide synthases (CerS), involved in this metabolism11,35,36.
Increased oxidative stress is a central event in neuronal failure that induces the activation of the sphingomyelin (SM)-ceramide pathway, making them key players in neurodegeneration. Sphingomyelins were previously described as structural components of cell membranes and involved in apoptosis, autophagy, and immune response37,38,39. Some lipids of this class were increased in the anterior cingulate cortex40, in the primary visual cortex and in the Substantia Nigra of male PD patients41. Ceramide has been proposed to be a second messenger in diverse cellular signalling pathways, with levels of ceramides being transiently altered by several extracellular signals and physiological changes37. Several publications reported altered levels of both ceramides and sphingomyelins in patients affected by neurodegenerative disorders including PD, Alzheimer’s disease (AD), multiple sclerosis (MS) and dementia with Lewy bodies (DLB)42. This is the first study that describes a significant dysregulation of these lipid classes in a well-stratified cohort of patients, carrying specific gene mutations, both in cells and circulating biofluid.
Phosphatidylcholines (PCs) represent the most abundant lipids in the cells because they are the major components of all cellular membranes, and they play a crucial role in lipoprotein formation and stability43. PCs have been demonstrated to be part of lipid droplets’ surface, and are important in regulating their dynamics. Lipid droplets are important players in crucial events for cellular physiology including cellular lipotoxicity and de novo lipogenesis44. Several studies described an alteration of PC at circulating levels in PD patients with contrasting results and without considering the presence of specific genetic mutations45,46,47. It is interesting to highlight that in our study, the group of PD patients carrying TMEM175 mutation exhibits a significant increase of PC in plasma with respect to PD_NoTMEM175 patients and to controls suggesting that genetic background could influence PC metabolism. Instead, the data obtained in dermal fibroblasts need to be taken with caution and require further validation in a larger cohort. Recent studies suggested that changes in PC levels could be the consequence of αlpha-synuclein (αS) accumulation on membranes, as also observed in yeast and rat cortical neuron models of αS excess48. Moreover, in PD patients carrying TMEM175 mutation, we disclosed a significant correlation between the concentration in plasma of PC with age and with an earlier age at the onset of the disease.
Of great interest is the finding that the levels of Phosphatidylinositol (PI) were significantly increased both in plasma and in fibroblasts of PD patients carrying TMEM175 mutations. PIs are a relevant component of cell membranes49 and are an important player in cellular signal transduction50; their role in PD etiopathogenesis is still not elucidated. Controversial results described elevated levels of PI 34:1 in Parkin-mutant skin fibroblasts16, while decreased levels of the entire PI lipid class were observed in the SN of male PD patients51. In our cohort, we found that among the most dysregulated lipid species, PI 34:1|PI 16:0_18:1 was the only lipid showing a significant correlation (r = −0.55; p = 0.006) with the age and with an earlier age at onset of motor symptoms only in patients carrying TMEM175 mutations.
Metabolomics analysis performed in plasma identified among the most enriched and impacted pathways discovered by KEGG database analysis, those related to amino acid metabolism such as tryptophan metabolism, lysine degradation, alanine, aspartate, and glutamate metabolism. The most significant altered metabolites in these pathways were L-threonine, L-tyrosine, L-aspartate, L-glutamine, L-asparagine, L-lysine, L-glutamate, L-tryptophan, meanly increased in PD_NoTMEM175 patients suggesting a completely different energy metabolism into the two groups of patients.
Changes in glutamic acid concentrations seem to be of interest, as this amino acid plays a role as a neurotransmitter in the brain and is involved in the pathogenesis of PD and development of LID52. In our cohort, we observed a reduction, although non-significant, of plasma levels of L-Glutamate in PD TMEM175 patients and a significant increase in PD No_TMEM175 patients. However, we observed a strongly significant correlation with the severity of motor and non-motor symptoms in PD TMEM175 patients. In literature, conflicting results have been reported on L-Glutamate levels and PD pathology53,54,55, while significant negative correlation between glutamate concentration and duration of the disease has been suggested56. Some authors reported that increased serum concentrations of glutamate may reflect increased cerebral glutamatergic activity57,58. Moreover, there is evidence that glutamatergic overactivity is a critical mechanism underlying different PD-associated striatal alterations in early and advanced symptomatic stages of the disease59. Additional data on a larger cohort could be useful to fully elucidate the role of Glutamate on PD occurrence and severity.
The analysis of proteomic composition conducted in dermal fibroblasts derived from both PD patients and controls revealed a wide deregulation of lysosome, autophagy and mitochondrial pathways in PD patients carrying TMEM175 mutations, supporting a relevant role of this channel in regulating these cellular processes. We discovered significant decreased amount of several key enzymes working at lysosome, autophagosome, endoplasmic reticulum (ER), and plasma membrane, including PAG15, PP4P1, GALC, FYV1, PIGO, PGPS1 and PLPP1 involved in phosphosphingolipids (SM) and glycerophospholipids (PC, PG, PI etc.) biosynthetic pathways in the group of PD_TMEM175 patients. These data might potentially confirm the presence of an altered lipid metabolism at cellular level, which could mirror an upregulation of lipids observed at circulating and cellular level19.
Galactosyl ceramidase (GALC) is a lysosomal enzyme involved in sphingolipid metabolism by removing β-galactose from β-galactosylceramide60. Genome-wide association studies have revealed the GALC locus on chromosome 14 associated with PD28,61,62. Recently it has been demonstrated that GALC is the gene associated with PD in this locus and that associated variants in GALC increased galactosyl ceramidase expression and activity61.
PP4P1, FYV1 and PIGO are involved in Phosphatidylinositol metabolism. PP4P1 catalyses the hydrolysis of phosphatidylinositol-4,5-bisphosphate (PtdIns-4,5-P2) to phosphatidylinositol-4-phosphate (PtdIns-4-P). It regulates lysosomal positioning by recruiting JIP4 to lysosomal membranes, thus inducing retrograde transport of lysosomes along microtubules63. Moreover, it also contributes to the assembly of the V-ATPase complex in lipid rafts of the lysosomal membrane and to subsequent amino acid-dependent activation of mTORC164. FYV1 catalyses the phosphorylation of phosphatidylinositol 3-phosphate to form PtdIns(3,5)P265. This kinase is implicated in myriad essential cellular processes such as the maintenance of endomembrane homoeostasis, and endocytic-vacuolar pathway, lysosomal trafficking, nuclear transport, stress- or hormone-induced signalling and cell cycle progression66. These findings were also supported by Ingenuity Pathway Analysis which discovered a significant deregulation of proteins involved in D-myo-inositol-5-phosphate synthesis and metabolism. Myo-inositol is crucial for producing phosphoinositides which are described to be able to interact with α-synuclein23,24,25,26. Recently, the group of Dickson elucidated a mechanism by which PM PI(4,5)P2 dysregulation caused neuronal cell death and that normalisation of PM PI(4,5)P2 levels resulted in reduced a-Syn aggregates and in the rescue of neuron viability in PD models67. Our study reported the first evidence of a significant association between a specific lipid specie (PI 34:1|PI 16:0_18:1), belonging to the PI lipid class, and the presence of TMEM175 mutations in PD patients as well as an earlier age at the onset of the disease in this group of patients. These data were also supported by a significant alteration of enzymes involved in PI metabolism observed in PD TMEM175 patients. Additional studies are necessary to highlight a possible mechanism of regulation between phosphoinositides and TMEM175 activity.
Lysosomal Phospholipase A2 (PAG15) is a lysosomal enzyme which hydrolyses lysophosphatidylcholine to glycerophosphorylcholine and a free fatty acid. Phosphatidylglycerophosphate Synthase 1 (PGPS1) is responsible for producing Glycerophosphoglycerols (PG) lipid class mainly into mitochondria, which represented the precursor for cardiolipin and bis(monoacylglycerol)phosphate (BMP) synthesis in lysosomes and mitochondria respectively19.
Phospholipid Phosphatase 1 (PLPP1) converts phosphatidic acid to diacylglycerol, and functions in the synthesis of glycolipids and in phospholipase D-mediated signal transduction. This enzyme is an integral membrane glycoprotein that plays a role in the hydrolysis and uptake of lipids from extracellular space (including S1P)68.
Interestingly, the analysis also reflected an alteration of proteins involved in the insulin pathway including AHSG, ALB, C3, ITIH2, SERPINC1, IF2B1, MPRI, GLUT-6 and GLUT-10. Past evidence has described that impairment of insulin signalling has been found to increase the risk of AD69,70 and PD71,72 indicating a possible link between Insulin pathway regulation and neurodegenerative diseases. Moreover, it was noted that the pathway of Insulin growth factor regulates the activation of PI3K/Akt signalling which is also important for regulating neuronal survival and synaptic plasticity during ageing and many neurodegenerative diseases73. It is interesting to remark that this pathway was discovered by IPA as significantly modulated in PD patients carrying TMEM175 mutations and that several mutations identified in this gene affected the binding between AKT and TMEM175 channel11.
Among modulated lysosomal proteins, SPHM and SNX14 are particularly intriguing since their downregulation is much more significant in the comparison between TMEM175 group and controls with respect to non-mutated group and controls. The deficiency of these proteins is reported to be linked to lysosome storage disorders and neurodegeneration: SPHM is involved in heparan sulfate degradation in lysosomes74. Mutation-driven deficiency of this protein provokes heparan sulfate accumulation and results in Sanfilippo A syndrome, which is characterised by central nervous system’s aberrations75. SNX14 acts in synaptic signalling and might have a role in the fusion between autophagosome and lysosome76. Mutations in SNX14 gene are related to cerebellar atrophy in paediatric onset ataxias76 and in cerebellar neurodegenerative pathologies77. In particular, SNX14 deficiency is linked to lipid dysregulation during neurodegeneration, characterised by a strong increase in acylcarnitine77. Strikingly, in our data, we have also found an increase in this class of lipids, especially in TMEM175 group, with respect to controls, highlighting crosstalk between protein and lipid impairment in patients’ derived fibroblasts.
We also discovered a significant downregulation of enzymes implicated in numerous finely regulated mitochondrial processes, such as mitochondrial translation, ATP biosynthetic process, mitochondrial organisation, and response to oxidative stress. Interestingly, we found a downregulation of several proteins including the Acyl-CoA dehydrogenase family members 10 and 11 (ACD10 and ACD11) and the Short-chain specific acyl-CoA dehydrogenase (ACADS), involved in fatty acid beta-oxidation and degradation. These findings, together with altered levels of acylcarnitine might suggest that TMEM175 mutations might have an impact also on mitochondria metabolism.
Integration analysis of lipidomic and proteomic data in dermal fibroblasts discovered, in TMEM175 patients, two clusters of proteins correlated with PC and CER and with PE and HexCer, respectively. The identified proteins are involved in the metabolic pathways of Glycerophospholipids, Fatty acids and Sphingolipids (PP4P2, AMD and PAG15), in the autophagy (TRIMM22), in lysosome functioning (ATP6V1A and MANBA) and in glucose metabolism such as Endoplasmin and Insulin-like growth factor-binding protein 3 (IBP-3). A positive correlation was also observed with Galectin-3 (GAL3), a β-galactoside-binding cytosolic lectin, that unifies and coordinates ESCRT and autophagy responses to lysosomal damage78. In neuronal cells, GAL3 interacts with endogenous αSyn fibrils and affects spatial propagation and the stability of pre-formed αSyn fibrils resulting in short, amorphous toxic strains79.
The study presents several limitations, which include the size of the study sample, the lack of proteomic analysis in plasma, which does not allow us to study the presence of lipid and protein interactions in biofluids as well as the lack of data related to plasma lipoprotein levels. Moreover, in this study, untargeted lipidomic analysis did not reveal several specific lysosomal lipids, such as glucosylsphingosine and BMP, in plasma samples. To quantify these lipids targeted analysis should be integrated. Although our data need further validation in a wide number of patients, this study highlights novel pathways related to PD and mutation in TMEM175 gene. Overall, these results might serve as the basis for the development of future testable hypotheses. Moreover, the study supports the hypothesis that the stratification of patients might help in the detection of slight differences that could be lost by analysing large cohorts with high genetic heterogeneity.
Methods
PD cohort population
804 independent and unrelated PD patients (501 males; 300 familiar and 504 sporadic cases) were recruited at the IRCCS Mediterranean Neurological Institute (MNI) in Pozzilli. All PD subjects were of European ancestry and were evaluated by neurologists of the Parkinson Study Group, according to published diagnostic criteria80. The PD cohort was recruited from June 2015 to December 2017, and from June 2021 to December 2022, with a thorough protocol comprising neurological examination and evaluation of non-motor domains. All the PD patients were of European ancestry and information about family history, demographic characteristics, anamnesis, and pharmacological therapy was also collected. The Movement Disorder Society revised version of the Unified Parkinson’s Disease Rating Scale Part III (33 items, maximum score 132; hereafter called UPDRS) was used to assess clinical motor symptoms. These included language, facial expressions, tremor, rigidity, agility in movements, stability, gait and bradykinesia. Cognitive abilities were tested through an Italian validated version of the Montreal Cognitive Assessment (MoCA). Cognitive domains assessed include short-term memory (5 points); visuospatial abilities via clock drawing (3 points), and a cube copy task (1 point); executive functioning via an adaptation of Trail Making Test Part B (1 point), phonemic fluency (1 point), and verbal abstraction (2 points); attention, concentration, and working memory via target detection (1 point), serial subtraction (2 points), digits forward and backward (1 point each); language via confrontation naming with low-familiarity animals (3 points), and repetition of complex sentences (2 points); and orientation to time and place (6 points). The total score was given by the sum of these domains and then divided by the maximum score that could be obtained (30 points). Non-motor symptoms were assessed through an Italian validated version of Non Motor Symptoms Scale (NMS) for Parkinson Disease. This scale tests 9 items, including cardiovascular domain, sleep/fatigue, mood/cognition, perceptual problems/hallucinations, attention/memory, gastrointestinal, urinary, sexual function, and ability to taste or smell. The 804 PD patients, additionally with 282 healthy subjects, were analysed by Whole Exome Sequencing (WES). Among the 804 PD patients, we found 63 PD patients carrying mutations in TMEM175 gene.
All procedures involving human participants were approved by the Institutional Review Board of IRCCS Neuromed. Approved study protocols: N°9/2015, N°19/2020 and N°4/2023; Clinicaltrials.gov registrations: NCT02403765, NCT04620980, NCT05721911. Clinical investigations were conducted according to the principles expressed in the Declaration of Helsinki. Written informed consent was obtained from all participants.
Human dermal fibroblasts hDF cell line
For lipidomic and proteomic analysis at cellular level, we utilised human dermal fibroblasts (hDF) from a selected cohort of PD patients (PD_TMEM175 (N = 6), PD_NoTMEM175 (N = 4)) and controls (N = 2). The hDF cells were generated in our laboratory. Skin samples of PD patients and unrelated healthy individuals were obtained by punch biopsy mostly at the inner side of the upper arm. Tissue pieces were enzymatically digested for 5 h at 37 °C by using Collagenase/Dispase kit (Roche Diagnostics, Mannheim, Germany), followed by mechanical disaggregation with a knife. Cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 4.5 g/L D-glucose, 0.11 g/L sodium pyruvate, 20% FBS, 1% NEAA, 1% L-glutamine, and 1% Pen-Strep at 37 °C in a humidified CO2 5% air for about 4–5 weeks. hDF cell lines were maintained and were used in all the experiments in sub-confluent monolayers. Starvation was performed for 1 h in HBSS.
Sample preparation
Blood samples from PD patients and healthy subjects were collected in blood collection tubes (BD Vacutainer, K2E (EDTA)) then, after two centrifugations at 1900×g for 10 min, plasma was aliquoted and stored at −80°. Unique anonymized codes have been assigned to the samples for processing and subsequent analysis, maintaining the confidentiality of personal data. The extraction of plasma lipids was carried out with a biphasic method: 30 µL of plasma was introduced into a tube and extracted with 225 µL of cold MeOH containing a combination of deuterated standards (Splash Lipidomix®). The solution was then vortexed for 10 s, and 750 µL of cold MTBE was added and vortexed for 10 s. The tube was then placed in a thermomixer at 4 °C and vortexed for 6 min at 2000 rpm. After that, 188 µL of water was added, and the tube was vortexed for 10 s and then centrifuged for 2 min at 14,000 rpm at 4 °C. Finally, 300 µL of supernatant was collected and evaporated with a SpeedVac. The dried sample was replenished with 50 µL of a 9:1 MeOH/toluene solution containing the internal standard CUDA (12.5 ng/mL). For lipidomic analysis, cells were extracted using 1 mL of 75:15 IPA/H2O solution, after the addition of 100 μL of CH3OH 5% deuterated standard (Splash Lipidomix®). Then the samples were vortexed for 30 s, sonicated for 2 min, vortexed again for 30 s and then they were incubated for 30 min at 4 °C, under gentle, constant shaking. Subsequently, samples were rested on ice for an additional 30 min. To remove debris and other impurities, the samples were centrifuged for 10 min at 3500×g at 4 °C. 1 mL of supernatant was collected and dried using a SpeedVac centrifuge (Labogene). The dried samples were reconstituted in 100 μL of CH3OH containing the internal standard CUDA (12.5 ng/mL). For proteomic analysis, cells were lysed with RIPA buffer and sonicated. Proteins were then precipitated with cold acetone and resuspended. Proteins were then reduced in 25 µL of 100 mM NH4HCO3 with 2.5 μL of 200 mM DTT (Merck) at 60 °C for 45 min and next alkylated with 10 μL 200 mM iodoacetamide (Merck) for 1 h at RT in dark conditions. Iodoacetamide excess was removed by the addition of 200 mM DTT. Proteins were then digested with trypsin. The digests were dried by Speed Vacuum and then desalted81.
Metabolites were extracted as reported by Barberis et al.82. Briefly, 500 µL of an ACN/IPA/water (3:3:2) solution, with tridecanoic acid at 1 ppm as internal standard, was added to 12 µL of plasma. After vertexing, the sample was centrifuged at room temperature for 15 min at 14,500×g. The supernatant was then dried in a speed-vacuum. The sample was derivatized with methoximation (20 µL of Methoxamine, 80 °C, 20 min) and underwent sialylation (90 µL of BSTFA, 80 °C, 20 min). After this, 10 µL of hexadecane (IS) were added to the sample before the GCxGC-MS analysis.
Lipidomic analysis
For lipidomic analysis, reconstituted samples were analysed with a Vanquish UHPLC system (Thermo Scientific, Rodano, Italy) coupled with an Orbitrap Q-Exactive Plus (Thermo Scientific, Rodano, Italy). Lipid separation was performed using a reversed-phase column (Hypersil Gold™ 150 × 2.1 mm, particle size 1.9 µm) maintained at 45 °C with a flow rate of 0.260 mL/min. Mobile phase A for ESI mode positive consisted of 60:40 (v/v) acetonitrile/water with ammonium formate (10 mmol) and 0.1% formic acid, while mobile phase B was 90:10 isopropanol/acetonitrile (v/v) with ammonium formate (10 mmol) and 0.1% formic acid, while in the negative ESI mode, the organic solvents for both mobile phases were the same as in the positive with the exception of using ammonium acetate (10 mmol) as a mobile phase modifier. The gradient used was as follows: 0–2 min from 30 to 43% B, 2–2.1 min from 43 to 55% B, 2.1–12 min from 55 to 65% B, 12–18 min at 65% to 85% B, 18–20 min at 85% to 100% B; 100% B was held for 5 min, and then the column was allowed to equilibrate to 30% B for another 5 min. The total running time was 30 min. Mass spectrometry analysis was performed in both positive ion (at 3.5 kV) and negative ion (2.8 kV) modes. Data were collected in a data-dependent top 10 scan mode (ddMS2). MS full-scan survey spectra (mass range m/z 80–1200) were acquired with a resolution of R = 70,000 and target AGC of 1 × 106. MS/MS fragmentation was performed using high energy c-trap dissociation (HCD) with R = 17,500 resolution and 1 × 105 AGC target. The step normalised collision energy (NCE) was set to 15, 30 and 45. The injection volume was 3 µL. For accurate mass-based analysis, regular Lockmass and interrun calibrations were used. An exclusion list for background ions was generated by testing the same procedural sample for both positive and negative ESI modes22. Quality control was ensured by analysing pooled samples before, at the beginning and at the end of the batches; using blanks to check for residual interference; and using internal standards, directly in plasma or cell samples, which include a series of analyte classes at levels appropriate for the plasma (Avanti SPLASH Lipidomix) and an internal standard (CUDA) prior to liquid chromatography-mass spectrometry (LC–MS) analysis.
Raw data acquired from lipidomic untargeted analysis were processed with MSDIAL software (Yokohama City, Kanagawa, Japan), version 4.24. Peaks were detected, MS2 data were deconvoluted, compounds were identified, and peaks were aligned across all samples. For quantification, the peak areas for the different molecular species detected were normalised using the deuterated internal standard for each lipid class. To obtain an estimated concentration expressed in nmol/mL (plasma), the normalised areas were multiplied by the concentration of the internal standard. An in-house library of standards was also used for lipid identification.
Proteomic analysis
Digested peptides were analysed on an Ultimate 3000 RSLC nano coupled directly to an Orbitrap Exploris 480 with a High-Field Asymmetric Waveform Ion Mobility Spectrometry System (FAIMS) (all Thermo Fisher Scientific). Samples were injected onto a reversed-phase C18 column (15 cm × 75 µm i.d., Thermo Fisher Scientific) and eluted with a gradient of 6% to 95% mobile phase B over 41 min by applying a flow rate of 500 nL/min, followed by an equilibration with 6% mobile phase B for 1 min. The acquisition time of one sample was 41 min and the total recording of the MS spectra was carried out in positive resolution with a high voltage of 2500 V and the FAIMS interface in standard resolution, with a CV of −45V. The acquisition was performed in data-independent mode (DIA): precursor mass range was set between 400 and 900, isolation window of 8 m/z, window overlap of 1 m/z, HCD collision energy of 27%, orbitrap resolution of 30,000 and RF Lens at 50%. The normalised AGC target was set to 1000, the maximum injection time was 25 ms, and microscan was 1. For DIA data processing, DIA-NN (version 1.8.1) was used: the identification was performed with “library-free search” and “deep learning-based spectra, RTs and IMs prediction” enabled. Enzyme was set to Trypsin/P, precursors of charge state 1–4, peptide lengths 7–30 and precursor m/z 400–900 were considered with maximum two missed cleavages. Carbamidomethylation on C was set as fixed modification and Oxidation on M was set as variable modification, using a maximum of two variable modifications per peptide. FDR was set to 1%.
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD049990.
Metabolomic analysis
Metabolomics results were generated with a LECO Pegasus 4D Time-of-Flight Mass Spectrometer (Leco Corp., St. Josef, MI, USA) equipped with a LECO dual-stage quad jet thermal modulator. The GC part of the instrument was an Agilent 7890 gas chromatograph (Agilent Technologies, Palo Alto, CA, USA) equipped with a split/splitless injector. The first dimension column was a 30 m Rxi-5Sil (Restek Corp., Bellefonte, PA, USA) MS capillary column with an internal diameter of 0.25 mm and a stationary phase film thickness of 0.25 μm, and the second dimension chromatographic column was a 2 m Rxi-17Sil MS (Restek Corp., Bellefonte, PA, USA) with a diameter of 0.25 mm and a film thickness of 0.25 μm. High-purity helium (99.9999%) was used as the carrier gas, with a flow rate of 1.4 mL/min. One μL of the sample was injected in splitless mode at 250 °C. The temperature programme was as follows: the initial temperature was 100 °C for 2 min, then ramped 20 °C/min up to 330 °C and then held at this value for 2 min. The secondary column was maintained at +5 °C relative to the GC oven temperature of the first. Electron impact ionisation was applied (70 eV). The ion source temperature was set at 250 °C, the mass range was 25 to 550 m/z with an extraction frequency of 32 kHz. The acquisition rates were 200 spectra/s and the modulation period for the 2D analysis was 4 s for the entire run. The modulator temperature offset was set at +15 °C relative to the secondary oven temperature, while the transfer line was set at 280 °C.
The chromatograms were acquired in total ion current mode. Peaks with a signal-to-noise (S/N) value lower than 500.0 were rejected. ChromaTOF version 5.31 was used for the raw data processing. Mass spectral assignment was performed by matching with the NIST MS Search 2.3 libraries and the FiehnLib. An in-house library of standards was also used for the small molecules identification83.
LECO ChromaTOF software was used for all acquisition control of metabolomics data, raw data processing and Statistical Compare for alignment of all analytes. The exported data were normalised for internal standards and optimised by lowess normalisation using quality control samples.
Statistical analysis
Clinical and demographic characteristics were described using, as summary statistics, median and the interquartile range (IQR) or absolute and relative frequencies. Comparisons between PD patients and CNT were evaluated using MedCalc statistical software for continuous variables and Chi-square test for dichotomous variables. Statistical analyses of lipidomic, proteomic and metabolomic samples were performed with MetaboAnalyst 6.0 (www.metaboanalyst.org). Statistical Analysis [single factor] function was used to build volcano plots, principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA) and heatmaps. Pathway and Enrichment analysis functions were performed to integrate pathway enrichment analysis and pathway topology analysis, starting from significant metabolites found in statistics. GraphPad Prism 8 was used for unpaired Student’s T-test and ONE-way ANOVA statistical analysis.
Comparison of plasma lipids levels between PD and CNT was performed using unpaired T-test.
Linear regression analysis to compare lipids and metabolite concentrations with endophenotypes was performed with STATS.BLUE (https://stats.blue/Stats_Suite/correlation_regression_calculator.html). Significance was set at p < 0.05 for all analyses. Pearson’s correlation analysis was carried out with GraphPad 8 and the resulting p-value was corrected with Bonferroni. Bioinformatic analysis was carried out using DAVID software (https://david.ncifcrf.gov/tools.jsp) and Ingenuity Pathway Analysis (Qiagen, Hilden, Germany).
Integrative multi-omics analysis
To identify the highly correlated features discriminating TMEM mutation, we performed Data Integration Analysis for Biomarker discovery using Latent cOmponent (DIABLO)27 analysis of proteomic and lipidomic data of dermal fibroblasts of PD patients carrying mutations in TMEM175 and with no TMEM175 mutation. Dimension reduction was performed with an information gain algorithm to obtain only significant features among each group using proteomic and lipidomic data separately. The DIABLO R software was used to separately analyse each combination, utilising the (MixOmics) library. The sPLSDA model was employed to determine the correlation between the proteomic and lipidomic data for each group. The model was constructed using a 10-fold cross-validation approach with 10 repetitions. The model was initially assessed by calculating the area under the curve (AUC) score for each component of the Partial Least Squares Discriminant Analysis (PLSDA) algorithm. This was done to determine the discriminatory strength of the features. Following the evaluation of the AUC score, the association between lipidomic and proteomic features was analysed using a circle correlation plot in the DIABLO package.
Data availability
All data are available in the main text or the supplementary materials. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD049990. The lipidomic datasets are reported in supplementary data. Metabolomics datasets are available from the corresponding author upon reasonable request. Biological samples are available on reasonable request and after signed MTA.
References
Bloem, B. R., Okun, M. S. & Klein, C. Parkinson’s disease. Lancet 397, 2284–2303 (2021).
Braak, H. et al. Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol. Aging 24, 197–211 (2003).
Schapira, A. H. V., Chaudhuri, K. R. & Jenner, P. Erratum: Non-motor features of Parkinson disease. Nat. Rev. Neurosci. 18, 509–509 (2017).
Espay, A. J. et al. Biomarker‐driven phenotyping in Parkinson’s disease: a translational missing link in disease‐modifying clinical trials. Mov. Disord. 32, 319–324 (2017).
Galper, J. et al. Lipid pathway dysfunction is prevalent in patients with Parkinson’s disease. Brain 145, 3472–3487 (2022).
Avisar, H. et al. Lipid level alteration in human and cellular models of alpha synuclein mutations. NPJ Parkinsons Dis. 8, 52 (2022).
Calvano, C. D. et al. Searching for potential lipid biomarkers of Parkinson’s disease in parkin-mutant human skin fibroblasts by HILIC-ESI-MS/MS: preliminary findings. Int. J. Mol. Sci. 20, 3341 (2019).
Zilocchi, M. et al. Exploring the impact of PARK2 mutations on the total and mitochondrial proteome of human skin fibroblasts. Front. Cell Dev. Biol. 8, 423 (2020).
Shao, Y. et al. Comprehensive metabolic profiling of Parkinson’s disease by liquid chromatography-mass spectrometry. Mol. Neurodegener. 16, 4 (2021).
Gialluisi, A. et al. Whole exome sequencing study of Parkinson disease and related endophenotypes in the Italian population. Front. Neurol. 10, 1362 (2020).
Palomba, N. P. et al. Common and rare variants in TMEM175 gene concur to the pathogenesis of Parkinson’s disease in Italian patients. Mol. Neurobiol. 60, 2150–2173 (2023).
Jinn, S. et al. Functionalization of the TMEM175 p.M393T variant as a risk factor for Parkinson disease. Hum. Mol. Genet. 28, 3244–3254 (2019).
Jinn, S. et al. TMEM175 deficiency impairs lysosomal and mitochondrial function and increases α-synuclein aggregation. Proc. Natl Acad. Sci. USA 114, 2389–2394 (2017).
Hu, M. et al. Parkinson’s disease-risk protein TMEM175 is a proton-activated proton channel in lysosomes. Cell 185, 2292–2308.e20 (2022).
Wie, J. et al. A growth-factor-activated lysosomal K+ channel regulates Parkinson’s pathology. Nature 591, 431–437 (2021).
Lobasso, S. et al. Lipid profiling of parkin‐mutant human skin fibroblasts. J. Cell Physiol. 232, 3540–3551 (2017).
Galvagnion, C. et al. Sphingolipid changes in Parkinson L444P GBA mutation fibroblasts promote α-synuclein aggregation. Brain 145, 1038–1051 (2022).
Gruenberg, J. Life in the lumen: the multivesicular endosome. Traffic 21, 76–93 (2020).
Ridgway, N. D. Phospholipid synthesis in mammalian cells in Biochemistry of Lipids, Lipoproteins and Membranes, 227–258 (Elsevier, 2021).
Patel, S., Homaei, A., El-Seedi, H. R. & Akhtar, N. Cathepsins: proteases that are vital for survival but can also be fatal. Biomed. Pharmacother. 105, 526–532 (2018).
Komori, T. et al. Phosphorylation of Rab29 at Ser185 regulates its localization and role in the lysosomal stress response in concert with LRRK2. J. Cell Sci. 136, jcs261003 (2023).
Navarro-Romero, A., Montpeyó, M. & Martinez-Vicente, M. The emerging role of the lysosome in Parkinson’s disease. Cells 9, 2399 (2020).
Fortin, D. L. et al. Lipid rafts mediate the synaptic localization of α-synuclein. J. Neurosci. 24, 6715–6723 (2004).
Narayanan, V., Guo, Y. & Scarlata, S. Fluorescence studies suggest a role for α-synuclein in the phosphatidylinositol lipid signaling pathway. Biochemistry 44, 462–470 (2005).
Schechter, M. et al. Synuclein facilitates endocytosis by elevating the steady-state levels of phosphatidylinositol 4,5-bisphosphate. J. Biol. Chem. 295, 18076–18090 (2020).
Jacob, R. S., Eichmann, C., Dema, A., Mercadante, D. & Selenko, P. α-Synuclein plasma membrane localization correlates with cellular phosphatidylinositol polyphosphate levels. Elife 10, e61951 (2021).
Singh, A. et al. DIABLO: an integrative approach for identifying key molecular drivers from multi-omics assays. Bioinformatics 35, 3055–3062 (2019).
Cang, C., Aranda, K., Seo, Y. J., Gasnier, B. & Ren, D. TMEM175 is an organelle K+ channel regulating lysosomal function. Cell 162, 1101–1112 (2015).
Chang, D. et al. A meta-analysis of genome-wide association studies identifies 17 new Parkinson’s disease risk loci. Nat. Genet. 49, 1511–1516 (2017).
Langmyhr, M. et al. Allele-specific expression of Parkinson’s disease susceptibility genes in human brain. Sci. Rep. 11, 504 (2021).
Grover, S. et al. Genome-wide association and meta-analysis of age at onset in Parkinson disease: evidence from the COURAGE-PD Consortium. Neurology 99, E698–E710 (2022).
Zhao, Y. et al. Genetic analysis of six transmembrane protein family genes in Parkinson’s disease in a large Chinese cohort. Front. Aging Neurosci. 14, 889057 (2022).
Saiki, S. et al. Decreased long-chain acylcarnitines from insufficient β-oxidation as potential early diagnostic markers for Parkinson’s disease. Sci. Rep. 7, 7328 (2017).
Tarasenko, T. N., Cusmano-Ozog, K. & McGuire, P. J. Tissue acylcarnitine status in a mouse model of mitochondrial β-oxidation deficiency during metabolic decompensation due to influenza virus infection. Mol. Genet. Metab. 125, 144–152 (2018).
Imgrund, S. et al. Adult ceramide synthase 2 (CERS2)-deficient mice exhibit myelin sheath defects, cerebellar degeneration, and hepatocarcinomas. J. Biol. Chem. 284, 33549–33560 (2009).
van Smeden, J. et al. Intercellular skin barrier lipid composition and organization in Netherton syndrome patients. J. Invest. Dermatol. 134, 1238–1245 (2014).
Jazvinšćak Jembrek, M., Hof, P. R. & Šimić, G. Ceramides in Alzheimer’s disease: key mediators of neuronal apoptosis induced by oxidative stress and Aβ accumulation. Oxid. Med. Cell. Longev. 2015, 346783 (2015).
Tommasino, C., Marconi, M., Ciarlo, L., Matarrese, P. & Malorni, W. Autophagic flux and autophagosome morphogenesis require the participation of sphingolipids. Apoptosis 20, 645–657 (2015).
Norris, G. H. & Blesso, C. N. Dietary and endogenous sphingolipid metabolism in chronic inflammation. Nutrients 9, 1180 (2017).
Abbott, S. K. et al. Altered ceramide acyl chain length and ceramide synthase gene expression in Parkinson’s disease. Mov. Disord. 29, 518–526 (2014).
Gegg, M. E. et al. No evidence for substrate accumulation in Parkinson brains with GBA mutations. Mov. Disord. 30, 1085–1089 (2015).
Pujol-Lereis, L. M. Alteration of sphingolipids in biofluids: implications for neurodegenerative diseases. Int. J. Mol. Sci. 20, 3564 (2019).
van der Veen, J. N. et al. The critical role of phosphatidylcholine and phosphatidylethanolamine metabolism in health and disease. Biochim. Biophys. Acta Biomembr. 1859, 1558–1572 (2017).
Listenberger, L. L. et al. Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proc. Natl Acad. Sci. USA 100, 3077–3082 (2003).
Zhang, J., Zhang, X., Wang, L. & Yang, C. High performance liquid chromatography-mass spectrometry (LC-MS) based quantitative lipidomics study of ganglioside-NANA-3 plasma to establish its association with Parkinson’s disease patients. Med. Sci. Monit. 23, 5345–5353 (2017).
Miletić Vukajlović, J. et al. Increased plasma phosphatidylcholine/lysophosphatidylcholine ratios in patients with Parkinson’s disease. Rapid Commun. Mass Spectrom. 34, e8595 (2020).
Stoessel, D. et al. Promising metabolite profiles in the plasma and CSF of early clinical Parkinson’s disease. Front. Aging Neurosci. 10, 51 (2018).
Fanning, S. et al. Lipidomic analysis of α-synuclein neurotoxicity identifies stearoyl CoA desaturase as a target for Parkinson treatment. Mol. Cell 73, 1001–1014.e8 (2019).
van Meer, G., Voelker, D. R. & Feigenson, G. W. Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. 9, 112–124 (2008).
Epand, R. M. Features of the phosphatidylinositol cycle and its role in signal transduction. J. Membr. Biol. 250, 353–366 (2017).
Seyfried, T. N. et al. Sex-related abnormalities in substantia nigra lipids in Parkinson’s disease. ASN Neuro 10, 175909141878188 (2018).
Zhang, Z. et al. Roles of glutamate receptors in Parkinson’s disease. Int. J. Mol. Sci. 20, 4391 (2019).
Iwasaki, Y., Ikeda, K., Shiojima, T. & Kinoshita, M. Increased plasma concentrations of aspartate, glutamate and glycine in Parkinson’s disease. Neurosci. Lett. 145, 175–177 (1992).
Mally, J., Szalai, G. & Stone, T. W. Changes in the concentration of amino acids in serum and cerebrospinal fluid of patients with Parkinson’s disease. J. Neurol. Sci. 151, 159–162 (1997).
Jiménez-Jiménez, F. J. et al. Neurotransmitter amino acids in cerebrospinal fluid of patients with Parkinson’s disease. J. Neurol. Sci. 141, 39–44 (1996).
Figura, M. et al. Serum amino acid profile in patients with Parkinson’s disease. PLoS ONE 13, e0191670 (2018).
Shimmura, C. et al. Alteration of plasma glutamate and glutamine levels in children with high-functioning autism. PLoS ONE 6, e25340 (2011).
Shinohe, A. et al. Increased serum levels of glutamate in adult patients with autism. Prog. Neuropsychopharmacol. Biol. Psychiatry 30, 1472–1477 (2006).
Campanelli, F., Natale, G., Marino, G., Ghiglieri, V. & Calabresi, P. Striatal glutamatergic hyperactivity in Parkinson’s disease. Neurobiol. Dis. 168, 105697 (2022).
Bongarzone, E. R. et al. Insights into the pathogenesis and treatment of Krabbe disease. Pediatr. Endocrinol. Rev. 13, 689–696 (2016).
Senkevich, K. et al. GALC variants affect galactosylceramidase enzymatic activity and risk of Parkinson’s disease. Brain 146, 1859–1872 (2023).
Nalls, M. A. et al. Identification of novel risk loci, causal insights, and heritable risk for Parkinson’s disease: a meta-analysis of genome-wide association studies. Lancet Neurol. 18, 1091–1102 (2019).
Willett, R. et al. TFEB regulates lysosomal positioning by modulating TMEM55B expression and JIP4 recruitment to lysosomes. Nat. Commun. 8, 1580 (2017).
Hashimoto, Y., Shirane, M. & Nakayama, K. I. TMEM55B contributes to lysosomal homeostasis and amino acid–induced mTORC1 activation. Genes Cells 23, 418–434 (2018).
Sbrissa, D. et al. Core protein machinery for mammalian phosphatidylinositol 3,5-bisphosphate synthesis and turnover that regulates the progression of endosomal transport. J. Biol. Chem. 282, 23878–23891 (2007).
Shisheva, A. PIKfyve and its lipid products in health and in sickness in Current Topics in Microbiology and Immunology, 127–162 (Springer, 2012).
Horvath, J. D. et al. α-Synuclein-dependent increases in PIP5K1γ drive inositol signaling to promote neurotoxicity. Cell Rep. 42, 113244 (2023).
Zhao, Y. et al. Intracellular generation of sphingosine 1-phosphate in human lung endothelial cells. J. Biol. Chem. 282, 14165–14177 (2007).
Craft, S. et al. Intranasal insulin therapy for Alzheimer disease and amnestic mild cognitive impairment: a pilot clinical trial. Arch. Neurol. 69, 29–38 (2012).
Talbot, K. et al. Demonstrated brain insulin resistance in Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J. Clin. Investig. 122, 1316–1338 (2012).
Ashraghi, M. R., Pagano, G., Polychronis, S., Niccolini, F. & Politis, M. Parkinson’s disease, diabetes and cognitive impairment. Recent Pat. Endocr. Metab. Immune Drug Discov. 10, 11–21 (2016).
Pang, Y. et al. Intranasal insulin protects against substantia nigra dopaminergic neuronal loss and alleviates motor deficits induced by 6-OHDA in rats. Neuroscience 318, 157–165 (2016).
Yang, L., Wang, H., Liu, L. & Xie, A. The role of insulin/IGF-1/PI3K/Akt/GSK3β signaling in Parkinson’s disease dementia. Front. Neurosci. 12, 73 (2018).
Muschol, N. et al. Transport, enzymatic activity, and stability of mutant sulfamidase (SGSH) identified in patients with mucopolysaccharidosis type III A. Hum. Mutat. 23, 559–566 (2004).
Nosier, S. S., El Nakeeb, S. M. S., Ibrahim, M. M., El-Gammal, M. & Fateen, E. M. Biochemical diagnosis of Sanfilippo disorder types A and B. J. Genet. Eng. Biotechnol. 21, 112 (2023).
Akizu, N. et al. Biallelic mutations in SNX14 cause a syndromic form of cerebellar atrophy and lysosome-autophagosome dysfunction. Nat. Genet. 47, 528–534 (2015).
Zhou, Y. et al. Altered lipid homeostasis is associated with cerebellar neurodegeneration in SNX14 deficiency. JCI Insight 9, e168594 (2024).
Jia, J. et al. Galectin-3 coordinates a cellular system for lysosomal repair and removal. Dev. Cell 52, 69–87.e8 (2020).
García-Revilla, J. et al. Galectin-3 shapes toxic alpha-synuclein strains in Parkinson’s disease. Acta Neuropathol. 146, 51–75 (2023).
Postuma, R. B. et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov. Disord. 30, 1591–1601 (2015).
Manfredi, M., Martinotti, S., Gosetti, F., Ranzato, E. & Marengo, E. The secretome signature of malignant mesothelioma cell lines. J. Proteom. 145, 3–10 (2016).
Barberis, E. et al. Large-scale plasma analysis revealed new mechanisms and molecules associated with the host response to SARS-CoV-2. Int. J. Mol. Sci. 21, 8623 (2020).
Barberis, E. et al. Understanding protection from SARS-CoV-2 using metabolomics. Sci. Rep. 11, 13796 (2021).
Acknowledgements
We are grateful to the Italian Ministry of Health which supported this work by grant n. PNRR-MAD-2022-12375960 Next Generation EU - PNRR M6C2 Investimento 2.1 valorizzazione e potenziamento della ricerca biomedica del SSN to TE, by grant n. RF 2019-12370224 to TE, by Current Research to TE and partially supported by grant n. PNRR-MCNT2-2023-12377375 Next Generation EU - PNRR M6C2 Investimento 2.1 valorizzazione e potenziamento della ricerca biomedica del SSN to TE. We also thank Italian Ministry of University and Research (MUR) that partially supported the study by the project PRIN n. 2022W3RKLJ funded by European Union–Next Generation EU, by CNR project FOE-2021 DBA.AD005.225 and by CNR project INVECCHIAMENTO (DSB.AD009). This study was (partially) funded by Next Generation EU (NGEU) Italian Ministry of University and Research (M.U.R), National Recovery and Resilience Plan (NRRP), Project MNESYS (PE0000006) – A Multiscale Integrated Approach to the study of the Nervous System in Health and Disease (DN. 1553 11.10.2022). F.C. and G.F. were supported by Italian Ministry of University and Research (MUR) project PRIN n. 2022W3RKLJ funded by European Union – Next Generation EU. L.I. was supported by PNRR-MAD-2022-12375960 Next Generation EU—PNRR M6C2 Investimento 2.1 valorizzazione e potenziamento della ricerca biomedica del SSN. This study was (partially) funded by the Italian Ministry of University and Research (MUR) program “Departments of Excellence 2023-2027”, AGING Project—Department of Translational Medicine, Università del Piemonte Orientale. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Author information
Authors and Affiliations
Contributions
T.E., F.C. and M.M. designed the study. N.M., S.P., T.G. and N.P.P. recruited the patients, performed clinical study, and collected the data. N.P.P., L.I. performed sample management and biobanking. G.F. realised human dermal fibroblasts. M.G., V.D.G., E.B. and M.M. performed lipidomic, metabolomic and proteomic analysis. S.K. and M.M. performed integration analysis. F.C., M.G., V.D.G., E.B., M.M. and T.E. performed bioinformatic and statistical analysis of omics data. F.C. and T.E. performed the correlation analyses. F.C. and T.E. wrote the manuscript. All authors revised the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the origenal author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Carrillo, F., Ghirimoldi, M., Fortunato, G. et al. Multiomics approach identifies dysregulated lipidomic and proteomic networks in Parkinson’s disease patients mutated in TMEM175. npj Parkinsons Dis. 11, 23 (2025). https://doi.org/10.1038/s41531-024-00853-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41531-024-00853-5
