top of page

Multi-omics approaches for therapeutic target and biomarker discoveries in Parkinson’s Disease

Updated: Aug 10, 2023



Parkinson’s disease (PD) is the second most common neurodegenerative disorder, after Alzheimer’s disease, and affects an estimated 6.1 million people worldwide (1). Its cause remains unknown, and there is no definitive diagnosis test or cure for PD with only therapies for motor symptom management currently available (2). Such a lack of effective therapies leads to pessimistic predictions for the PD incidence worldwide, which is that by 2040 neurodegenerative diseases may surpass cancer as the leading cause of disease-related death. This urges a thorough understanding of the neuropathology of PD and its progression throughout the nervous system.

Fortunately, we are living in an exciting decade of science that offers new powerful approaches that can tackle ongoing disease challenges. Two outstanding examples are the recent advances in sequencing technologies and mass spectrometry instrumentation that have successfully yielded data on multiple levels including genome, transcriptome, proteome, and metabolome, also known as multi-omics data (3). Particularly, in contrast to the study of only one data type the integration of these multi-omics data generates an information flow from one omics level to the other, hence elucidating a more complete understanding and potential causative changes underlie disease origins (4). In this study, we performed meta-analysis on PD multi-omics data, to obtain a better understanding of PD. These data include transcriptomics, proteomics and metabolomics datasets that were downloaded from public databases. Information on the titles of each dataset, sample numbers, sample collection types and software used are provided in Table 1. We applied a consistent computational workflow to preprocess these datasets and generate differentially expressed genes/proteins (DEGs/Ps) lists comparing between control and PD patients. These lists were then analyzed via the CDIAM platform, developed by Pythia Biosciences, that offers custom tools to accelerate and aggregate insights across omics datasets, i.e. unveil meaningful cell-cell interactions and ligand-receptor pairs, identify important pathways, drug targets, and biomarkers to prioritize.


Notes: This is a summary article. To view the full report, visit the link below.


Table 1: Summary of multi-omics PD datasets

Most overlapped significant pathways are involved in phagocytosis clearance of cell aggregates.


To better understand the biological processes and molecular functions represented by the differentially expressed genes, we performed pathway enrichments via either gene set enrichment analysis (GSEA) or hypergeometric method for all datasets. Queried pathway databases included Reactome (17), Wikipathway (18) and NCPID (19).


To identify pathways consistently relevant across datasets, we examined enriched pathways that were shared within at least two datasets, of which several are involved in phagocytosis clearance (Table 2). Indeed, dysregulation of glial phagocytosis and degradation has been revealed to play a key role in PD pathogenesis (20). Within the central nervous system, phagocytosis is a critical process required for maintaining the homeostasis within the synapses, apoptotic cells and debris, hence ensuring proper neural circuit function (21, 22). Phagocytotic pathways are highly dynamic due to the involvement of several protein components as well the extensive requirement of membrane remodeling and recycling events that intersect with other cellular processes such as autophagy (23).


Table 2: Overlapped intracellular signaling pathways related to phagocytosis of cell aggregates.

GAPDH and HRAS continue to be potential targets for PD treatments.


Next, we processed outputs from GSEA or hypergeometric pathway enrichment into Pathway2Targets (P2T) (Scott et al., 2021), a signaling pathway-driven bioinformatics pipeline to unravel potential cellular targets and therapeutics. To prioritize novel targets, we adjusted the weights of each factor in P2T target weighting and prioritization algorithm accordingly, and then consistently applied these parameters for all datasets (Figure 1).


Figure 1. Pathway2Targets interface with customized factors

Several targets were commonly found in at least two datasets’ P2T outputs such as GAPDH, HRAS, RAC1, CDH1 and PCNA. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and HRas proto-oncogene (HRAS, GTPase) are linked to current Parkinson’s disease treatments. GAPDH modulates the mechanism of action of Deprenyl (Selegiline) and its derivative that function via inhibiting neuronal cell deaths (24-27), while HRAS associates with L-DOPA (most effective treatment of PD so far) debilitating dyskinetic side effects via the activation of ERK and mTOR pathways in the striatum (28, 29). The P2T workflow also recognized Rac family small GTPase 1 (RAC1) as a potential target, which is in agreement with recent findings presented by a research group from the University of Barcelona (30). Interestingly, the functional pathways of two identified targets, cadherin 1 (CDH1) and proliferating cell nuclear antigen (PCNA), are not only involved in PD but also cross-linked with cancer pathways, further supporting the association between PD and cancer (31, 32).


CD44, VEGFA and HK1 as potential biomarkers for prediction of PD

We also used outputs from GSEA or hypergeometric pathway enrichment for Biomarker2Validate (B2V), to predict potential biomarkers for PD (Figure 2). From proteomics and RNA data, CD44 was marked as the most potential biomarker that can be detected in blood, in line with recently published data confirming CD44 as PD-status-dependently-regulated factors (14),(33). Another potential biomarker was VEGFA (ranked 4th), belonging to the VEGF family, whose cerebrospinal fluid level was reportedly upregulated in PD patients (34).


Figure 2. Biomarker2Validate pipeline curated from outputs from GSEA or hypergeometric pathway enrichments of RNA and proteomics data.

Meanwhile, from metabolomics data, most identified genes in B2V are in the glucose pathways. The accumulation of hexokinase 1 (HK1, ranked 1st), initial enzyme of glycolysis pathway, and its mis-location from usual location at the mitochondria was indeed described in a knockout model of familial PD, suggesting HK1 as the interplay factor among the three major genetic cases of early onset PD.


Communications between immune cells with other cells were significantly upregulated in PD patients.


Finally, we studied the interactions between different cell types in PD brains versus healthy brains from two snRNA datasets (Table 1) via CellPhoneDB available on CDIAM platform. Interestingly, both datasets yielded a significant upregulation in cell-cell communication between immune cells and other cell types in PD patients compared to control (Figure 3). Consistently, PPIA-BSG (CyPA-CD147) and SEMA4D-PTPRC (CD100-CD45), ligand-receptor pairs that modulate inflammatory processes, scored as the most significant upregulated pairs in both datasets. Our results supported the previously reported role of inflammation dysregulation in PD (35, 36).


Figure 3. Summary of upregulated cell-cell communication between brain cell types in human PD brains from GSE184950 (A) and GSE202210 (B). Red indicates upregulation.


Advancing PD research

In this study, we provide an outlook on the powerful tools offered by CDIAM platform that facilitates multi-omics data integration. We successfully identified several important drug targets and biomarkers that were consistent with previous PD research findings, signifying that our computational workflow pipelines are reliable in terms of identifying pathways and biomarkers with clinical and pathological relevance. Still, additional experiments would be required to validate our results in the future.


THIS ARTICLE IS FOR RESEARCH USE ONLY AND NOT FOR USE IN DIAGNOSTIC PROCEDURE.

Learn more about our research via the link below.


References

1. Global, regional, and national burden of Parkinson's disease, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 17, 939-953 (2018).

2. B. S. Connolly, A. E. Lang, Pharmacological treatment of Parkinson disease: a review. Jama 311, 1670-1683 (2014).

3. T. M. Santiago-Rodriguez, E. B. Hollister, Multi ‘omic data integration: A review of concepts, considerations, and approaches. Semin. Perinatol. 45, 151456 (2021).

4. M. Bersanelli et al., Methods for the integration of multi-omics data: mathematical aspects. BMC Bioinformatics 17, S15 (2016).

5. W. Qian et al., Single-cell Transcriptomic Atlas of the Human Substantia Nigra in Parkinson’s Disease. bioRxiv, 2022.2003.2025.485846 (2022).

6. F. A. Wolf, P. Angerer, F. J. Theis, SCANPY: large-scale single-cell gene expression data analysis. Genome Biol19, 15 (2018).

7. Z. Biqing et al., Single-cell transcriptomic and proteomic analysis of Parkinson’s disease Brains. bioRxiv, 2022.2002.2014.480397 (2022).

8. C. Lang et al., Single-cell sequencing of iPSC-dopamine neurons rconstructs disease progression and identifies HDAC4 as a regulator of Parkinson cell phenotypes. Cell Stem Cell 24, 93-106.e106 (2019).

9. M. D. Robinson, D. J. McCarthy, G. K. Smyth, edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139-140 (2010).

10. C. Wilks et al., recount3: summaries and queries for large-scale RNA-seq expression and splicing. Genome Biol. 22, 323 (2021).

11. N. Antoniou et al., High content screening and proteomic analysis identify a kinase inhibitor that rescues pathological phenotypes in a patient-derived model of Parkinson's disease. npj Parkinsons Dis 8, 15 (2022).

12. K. Elena, T. S. Maximilian, M. Matthias, AlphaPeptStats: an open-source Python package for automated, scalable and industrial-strength statistical analysis of mass spectrometry-based proteomics. bioRxiv, 2023.2003.2010.532057 (2023).

13. M.-T. Mackmull et al., Global, in situ analysis of the structural proteome in individuals with Parkinson’s disease to identify a new class of biomarker. Nat. Struct. Mol. Biol 29, 978-989 (2022).

14. O. Karayel et al., Proteome profiling of cerebrospinal fluid reveals biomarker candidates for Parkinson’s disease. Cell Rep Med 3, (2022).

15. M. Scholefield et al., Multi-regional alterations in glucose and purine metabolic pathways in the Parkinson’s disease dementia brain. npj Parkinson Dis 9, 66 (2023).

16. C. Kumari Sonal et al., MetENP/MetENPWeb: An R package and web application for metabolomics enrichment and pathway analysis in Metabolomics Workbench. bioRxiv, 2020.2011.2020.391912 (2020).

17. A. Fabregat et al., Reactome pathway analysis: a high-performance in-memory approach. BMC Bioinformatics 18, 142 (2017).

18. M. Martens et al., WikiPathways: connecting communities. Nucleic Acids Res. 49, D613-D621 (2021).

19. C. F. Schaefer et al., PID: the Pathway Interaction Database. Nucleic Acids Res37, D674-679 (2009).

20. M.-E. Tremblay, M. R. Cookson, L. Civiero, Glial phagocytic clearance in Parkinson’s disease. Mol Neurodegeneration 14, 16 (2019).

21. S. Gordon, Phagocytosis: an immunobiologic process. Immunity 44, 463-475 (2016).

22. S. Arandjelovic, K. S. Ravichandran, Phagocytosis of apoptotic cells in homeostasis. Nat Immunol 16, 907-917 (2015).

23. Y. Nash, E. Schmukler, D. Trudler, R. Pinkas-Kramarski, D. Frenkel, DJ-1 deficiency impairs autophagy and reduces alpha-synuclein phagocytosis by microglia. J Neurochem 143, 584-594 (2017).

24. M. D. Berry, Glyceraldehyde-3-phosphate dehydrogenase as a target for small-molecule disease-modifying therapies in human neurodegenerative disorders. J Psychiatry Neurosci 29, 337-345 (2004).

25. J. Gerszon, A. Rodacka, Oxidatively modified glyceraldehyde-3-phosphate dehydrogenase in neurodegenerative processes and the role of low molecular weight compounds in counteracting its aggregation and nuclear translocation. Ageing Res Rev 48, 21-31 (2018).

26. M. R. Hara et al., Neuroprotection by pharmacologic blockade of the GAPDH death cascade. Proc Natl Acad Sci U S A 103, 3887-3889 (2006).

27. P. C. Waldmeier, A. A. Boulton, A. R. Cools, A. C. Kato, W. G. Tatton, Neurorescuing effects of the GAPDH ligand CGP 3466B. J Neural Transm Suppl, 197-214 (2000).

28. N. Martín-Flores et al., mTOR pathway-based discovery of genetic susceptibility to L-DOPA-induced dyskinesia in Parkinson's disease patients. Molecular neurobiology 56, 2092-2100 (2019).

29. M. Eshraghi et al., RasGRP1 is a causal factor in the development of l-DOPA-induced dyskinesia in Parkinson's disease. Sci Adv 6, eaaz7001 (2020).

30. H. Kim et al., The small GTPase RAC1/CED-10 Is essential in maintaining dopaminergic neuron function and survival against α-synuclein-induced toxicity. Molecular neurobiology 55, 7533-7552 (2018).

31. S. B. Lee et al., Parkin regulates mitosis and genomic stability through Cdc20/Cdh1. Mol Cell 60, 21-34 (2015).

32. I. P. Heremans et al., Parkinson's disease protein PARK7 prevents metabolite and protein damage caused by a glycolytic metabolite. Proc Natl Acad Sci U S A 119, (2022).

33. L. Yao, S. Zhang. (Res Sq, 2021).

34. S. Janelidze et al., Increased CSF biomarkers of angiogenesis in Parkinson disease. Neurology 85, 1834-1842 (2015).

35. E. Contaldi, L. Magistrelli, C. Comi, T Lymphocytes in Parkinson's Disease. J Parkinsons Dis 12, S65-s74 (2022).

36. M. G. Tansey et al., Inflammation and immune dysfunction in Parkinson disease. Nat Rev Immuno 22, 657-673 (2022).



325 views0 comments

Comments


bottom of page