Systems genomics evaluation of the SH-SY5Y neuroblastoma cell line as a model for Parkinson's disease

Abstract

Background: The human neuroblastoma cell line, SH-SY5Y, is a commonly used cell line in studies related to neurotoxicity, oxidative stress, and neurodegenerative diseases. Although this cell line is often used as a cellular model for Parkinson's disease, the relevance of this cellular model in the context of Parkinson's disease (PD) and other neurodegenerative diseases has not yet been systematically evaluated.

Results: We have used a systems genomics approach to characterize the SH-SY5Y cell line using whole-genome sequencing to determine the genetic content of the cell line and used transcriptomics and proteomics data to determine molecular correlations. Further, we integrated genomic variants using a network analysis approach to evaluate the suitability of the SH-SY5Y cell line for perturbation experiments in the context of neurodegenerative diseases, including PD.

Conclusions: The systems genomics approach showed consistency across different biological levels (DNA, RNA and protein concentrations). Most of the genes belonging to the major Parkinson's disease pathways and modules were intact in the SH-SY5Y genome. Specifically, each analysed gene related to PD has at least one intact copy in SH-SY5Y. The disease-specific network analysis approach ranked the genetic integrity of SH-SY5Y as higher for PD than for Alzheimer's disease but lower than for Huntington's disease and Amyotrophic Lateral Sclerosis for loss of function perturbation experiments.

Figure 1

Systems genomics approach to assess SH-SY5Y as a disease model.

Figure 2

CNVs detected by whole genome sequencing and array-based CGH by Do et al. [36]. The results from whole genome sequencing were from Complete Genomics and are given in the left half of the chromosomes. The results from the array-based CGH are on the right half of the chromosomes. Regions are highlighted for copy number gain (red) and loss (blue). The major events partial trisomy of chromosome 1 and 2, complete trisomy of chromosome 7, gain in 17q and loss in 22q were confirmed. (Generated using http://db.systemsbiology.net/gestalt/cgi-pub/genomeMapBlocks.pl).

Figure 3

Protein-protein interaction network for two diseases – neuroblastoma (A) and for ALS (B). Red nodes refer to OMIM-derived genes mutated in the cell line. Orange nodes refer to the OMIM-derived genes which are intact in the cell line. Dark green nodes represent genes coming from the network expansion, which are mutated in the cell line. Light green nodes represent genes coming from the network expansion, which are intact in the cell line. Nodes are scaled to the magnitude of their betweenness centralities. Blue edges show connections between pairs of nodes in which at least one was damaged in the cell line. For neuroblastoma, mutations in the central genes – NME1 and ALK contribute to the high BC-ratio. For the ALS network, mutations occur in genes with lower betweenness centrality which results in the lower BC-ratio.

Figure 4

Protein-protein interaction network for two PD map modules - glycolysis (A) and ROS metabolism (B). Red nodes refer to genes mutated in the cell line. Orange nodes refer to genes which are intact in the cell line. Nodes are scaled to the magnitude of their betweenness centralities. Blue edges show connections between pairs of nodes in which at least one was damaged in the cell line. For glycolysis, mutations in the central genes lead to the high BC-ratio. Mutated genes lie on the periphery of the ROS metabolism network resulting in a low BC-ratio.

Figure 5

Dopamine biosynthesis and degradation and the genes coding key enzymes. This metabolic pathway is a modified version of a figure from Meiser et al. [60]. In addition to the enzymes involved, the genes coding for those enzymes have been added using the HumanCyc [61]. The mutations in SH-SY5Y affecting enzyme-encoding genes have been annotated only if they were rare protein-altering SNVs or indels, CNVs, or SVs.