Polyomic analyses of dopaminergic neurons isolated from human substantia 1 nigra in Parkinson’s disease: An exploratory study.

Dopaminergic (DA) neurons of the substantia nigra pars compacta (SNpc) selectively and progressively 29 degenerate in Parkinson’s disease (PD). Until now, molecular analyses of DA neurons in PD have been 30 limited to genomic and transcriptomic approaches, whereas, to the best of our knowledge, no 31 proteomic or combined polyomic study examining the protein profile of these neurons, is currently 32 available.


88
Over the last decade, the progressive improvements of laser microdissection (LMD) technology 9 in 89 automation, velocity and precision offers the opportunity to dissect frozen DA neurons in conditions 90 that are more suitable for relevant molecular analyses. The increased sensitivity of mass spectrometers 91 and RNA sequencers enables comparative and quantitative polyomic approaches using low to very low 92 amounts of biological material.

93
In this exploratory study, we used LMD to dissect DA-neurons from control and PD post-mortem SNpc 94 specimens. In the first part, we used both qualitative transcriptomic and proteomic approaches, to 95 confirm the integrity and validity of our samples, and the LMD-provided access to the specific protein 96 content of DA neurons. This important quality control step led to the second part of this study, where 97 a quantitative comparison of protein and gene expression by label free approach and RNA sequencing 98 (RNAseq), respectively, was performed in DA neurons from control and PD samples. Importantly, the 99 same specimens were used for both analyses. RNAseq analysis revealed 52 differentially expressed 100 genes, and label-free proteomics highlighted 33 differentially expressed proteins in PD samples 101 compared to matched controls. Transcriptomics and proteomics results were compared to identify the 102 mRNA-protein couples for which the expression changes followed the same direction. This work is the 103 first attempt to propose a polyomic analysis of DA neurons in the PD brain.

105
Human brain tissues 106 Ten frozen human midbrains, 5 from age-matched control patients and 5 from PD patients were 107 collected by the Department of Clinical Pathology and Psychiatry of the Geneva University Hospitals 108 under a procedure approved by the Geneva ethical committee (Table 1) and in accordance with the 109 relevant guidelines and regulations. Written informed consent for brain autopsy and use for research 110 was obtained from close family relatives. PD diagnosis was confirmed neuropathologically and 111 controls, with no previous history of neurological or psychiatric disorders, were confirmed to be free 112 of nigral abnormalities. Samples were cryopreserved at -80°C until further analysis.

210
In order to check whether the protein product of the differentially expressed genes has been already 211 detected by MS, we generated a list of brain proteins with Nextprot 13 using the Advanced search 212 (SPARQL) tool and querying for human proteins identified in the brain by MS with 2 distinct peptides 213 7 or more aminoacids long.

215
Integrity and quality of samples by transcriptomics 216 Before proceeding to the quantitative comparisons between PD and control samples, we controlled 217 that our sample preparation protocol preserved extracted molecules in sufficient quality for omics 218 analyses. As RNAs are known to be more vulnerable entities than proteins, we used transcriptomic 219 approaches to analyze RNA quality of our samples, by different ways, at different steps of the 220 workflow.

221
To this purpose, tissue slices from SNpc of controls and PD patients (Table 1)

239
In order to validate the capacity of our protocol to specifically highlight the molecular content of DA 240 neurons, we performed a proteomic analysis of the DA neurons collected from 5 control samples and 241 5 PD samples (Table 1)

246
The total amount of proteins extracted from these neurons ranged from 18 to 24 µg. 6 µg proteins of 247 each sample were trypsin digested and injected in triplicates for three GPF runs with nano-lc-ms/ms.     for Cystatin-C to 3.5-fold change for Vimentin, while downregulated proteins showed differences 294 ranging from 2.5-fold change for Aldehyde dehydrogenase 1 to 3.7-fold change for Alpha-1-antitrypsin. Proteins marked with an asterisk (*) have been already described in PD as dysregulated

306
In fact, only 7 gene-related proteins (13.5%) were identified by nano-lc-ms/ms among the potential 52 307 gene products, whereas no corresponding protein for the 45 remaining deregulated genes could be 308 found, making correlation between transcriptomic and proteomic data impossible. Among these 52 309 genes, 19 had never seen their corresponding protein identified by mass spectrometry from brain 310 samples according to Nextprot database 13 (Table 2).