Identification of candidate genes for Parkinson's disease through blood transcriptome analysis in LRRK2-G2019S carriers, idiopathic cases, and controls

Abstract

The commonest known cause of Parkinson's disease (PD) is the G2019S mutation of the LRRK2 gene, but this mutation is not sufficient for causing PD, and many carriers of the mutation never develop PD symptoms during life. Differences at the expression level of certain genes, resulting from either genetic variations or environmental interactions, might be one of the mechanisms underlying differential risks for developing both idiopathic and genetic PD. To identify the genes involved in PD pathogenesis, we compared genome-wide gene expression (RNA-seq) in peripheral blood of 20 PD patients carrying the G2019S mutation of the LRRK2 gene, 20 asymptomatic carriers of the mutation, 20 subjects with idiopathic PD, 20 controls and 7 PD patients before and after initiating dopaminergic therapy. We identified 13 common genes (ADARB2, CEACAM6, CNTNAP2, COL19A1, DEF4, DRAXIN, FCER2, HBG1, NCAPG2, PVRL2, SLC2A14, SNCA, and TCL1B) showing significant differential expression between G2019S-associated PD and asymptomatic carriers and also between idiopathic PD and controls but not between untreated and treated patients. Some of these genes are functionally involved in the processes known to be involved in PD pathogenesis, such as Akt signaling, glucose metabolism, or immunity. We consider that these genes merit further attention in future studies as potential candidate genes involved in both idiopathic and LRRK2-G2019S–associated forms of PD.

Introduction

The etiology of Parkinson's disease (PD) is considered to be complex and to involve both genetic and environmental factors. Most of these factors remain elusive, and the molecular pathways leading to the disease are not fully understood (Valente et al., 2012). The heritability of PD is estimated in ∼40% to 60%; however, the rare monogenic forms of the disease and the common genetic variants through the genome recently identified through genome-wide association studies (GWAS) explain just a minor proportion of the genetic component of the disease (Hamza and Payami, 2010). In this scenario, the identification of genes involved in the pathogenesis of the disease and the understanding of the biological consequences of gene-environmental interactions becomes a priority for identifying targets for disease modification.

What we know as PD is indeed a heterogeneous disorder, in terms of genetics, phenotypes, and also possibly in the underlying molecular mechanisms (Thenganatt and Jankovic, 2014). It has been proposed that different molecular pathways converging in common pathogenic routes might lead to a common PD phenotype. This heterogeneity, by adding complexity, might hamper the identification of genes and pathways involved in PD pathogenesis when studying populations of idiopathic PD patients.

The G2019S mutation of the LRRK2 gene (OMIM no. 609007) is the commonest known cause of PD, representing ∼1% to 5% of both familial and sporadic cases of the disease and showing an increased prevalence in certain populations (Healy et al., 2008). The disease is clinically and pathologically indistinguishable from idiopathic late-onset PD in most of the cases; therefore, it is considered an ideal scenario for investigating molecular mechanisms involved in PD pathogenesis in a more homogeneous setting. One of the features of the LRRK2-G2019S mutation is its reduced age-dependent penetrance, 47% at 80 years in our population (Sierra et al., 2011). The fact that only a subset of LRRK2-G2019S carriers will develop PD symptoms during life implies that other factors, either genetic or environmental, are contributing to the disease pathogenesis. Whether these unknown factors are also involved in idiopathic PD is plausible but unproven.

We aimed here to search for candidate genes and molecular pathways involved in PD pathogenesis starting from the analysis of differential expression of blood transcriptome between manifesting and nonmanifesting carriers of the LRRK2-G2019S mutation, in an homogeneous cohort of cases coming from a limited number of families in a restricted geographical area in Cantabria (Northern Spain).

Genome-wide study of gene expression, or transcriptomic profiling, is an emerging and promising tool for deciphering pathways involved in the pathogenesis of medical diseases with potential diagnostic or prognostic applications. Transcriptome analysis using next-generation sequencing (RNA-seq) allows analysis of the genomes at a higher resolution than microarray-based methods allowing the quantification of transcript levels and the sequence information. Although variations in the transcriptome are known to be tissue specific, recent findings indicate that blood and brain share significant gene expression similarities (Liew et al., 2006). Indeed, transcriptional profile in blood from PD patients has shown altered expression of genes involved in the pathobiological pathways known to be altered in PD (Cooper-Knock et al., 2012, Mutez et al., 2014, Scherzer et al., 2007). Also, a cross-talk between the immune and nervous systems exists and immune mechanisms, including leukocyte infiltration, have been reported to participate in PD neurodegeneration (Kannarkat et al., 2013). It is noteworthy that peripheral blood mononuclear cells also express protein markers of dopaminergic metabolism. For all these reasons, blood expression analysis has potential advantages over brain expression studies: brain tissue is not accessible during life, and postmortem brain areas directly affected by the pathology are likely to reveal downstream unspecific processes rather than pointing to the pathogenic mechanisms.

Here, we first compared blood transcriptome between PD patients carrying the G2019S mutation of the LRRK2 gene and nonmanifesting carriers of the mutation. Then, to limit and validate the identified genes as convincingly related with PD but not with medication, we conducted 2 additional comparisons between idiopathic PD and controls and between PD patients tested before and after treatment. The proposed candidate genes arisen from these analyses would be those showing significant differential expression in both initial comparisons (manifesting carriers of LRRK2-G2019S mutation vs. asymptomatic carriers of the mutation and idiopathic PD cases vs. controls) excluding those genes deregulated after medication.