Pathways and genes differentially expressed in the motor cortex of patients with sporadic amyotrophic lateral sclerosis

BMC Genomics Pathways and genes differentially expressed in the motor cortex of patients with sporadic amyotrophic lateral sclerosis Carsten W Lederer 1 Antonietta Torrisi 0 Maria Pantelidou 1 Niovi Santama 1 Sebastiano Cavallaro 0 0 Functional Genomics Center, Institute of Neurological Sciences, Italian National Research Council , 95123 Catania , Italy 1 Department of Biological Sciences, University of Cyprus and Cyprus Institute of Neurology and Genetics , 1678 Nicosia , Cyprus Background: Amyotrophic lateral sclerosis (ALS) is a fatal disorder caused by the progressive degeneration of motoneurons in brain and spinal cord. Despite identification of disease-linked mutations, the diversity of processes involved and the ambiguity of their relative importance in ALS pathogenesis still represent a major impediment to disease models as a basis for effective therapies. Moreover, the human motor cortex, although critical to ALS pathology and physiologically altered in most forms of the disease, has not been screened systematically for therapeutic targets. Results: By whole-genome expression profiling and stringent significance tests we identify genes and gene groups de-regulated in the motor cortex of patients with sporadic ALS, and interpret the role of individual candidate genes in a framework of differentially expressed pathways. Our findings emphasize the importance of defense responses and cytoskeletal, mitochondrial and proteasomal dysfunction, reflect reduced neuronal maintenance and vesicle trafficking, and implicate impaired ion homeostasis and glycolysis in ALS pathogenesis. Additionally, we compared our dataset with publicly available data for the SALS spinal cord, and show a high correlation of changes linked to the diseased state in the SALS motor cortex. In an analogous comparison with data for the Alzheimer's disease hippocampus we demonstrate a low correlation of global changes and a moderate correlation for changes specifically linked to the SALS diseased state. Conclusion: Gene and sample numbers investigated allow pathway- and gene-based analyses by established error-correction methods, drawing a molecular portrait of the ALS motor cortex that faithfully represents many known disease features and uncovers several novel aspects of ALS pathology. Contrary to expectations for a tissue under oxidative stress, nuclear-encoded mitochondrial genes are uniformly down-regulated. Moreover, the down-regulation of mitochondrial and glycolytic genes implies a combined reduction of mitochondrial and cytoplasmic energy supply, with a possible role in the death of ALS motoneurons. Identifying candidate genes exclusively expressed in non-neuronal cells, we also highlight the importance of these cells in disease development in the motor cortex. Notably, some pathways and candidate genes identified by this study are direct or indirect targets of medication already applied to unrelated illnesses and point the way towards the rapid development of effective symptomatic ALS therapies. - Background Amyotrophic lateral sclerosis (ALS) is a fatal neuromuscular disorder affecting 12 in 100,000 persons. It is caused by the degeneration of motoneurons in brain and spinal cord, leading to muscle atrophy, progressive paralysis, and death, commonly by respiratory failure. Most cases of ALS are sporadic (SALS) and about 10% familial (FALS), with mutant forms of copper-zinc superoxide dismutase (SOD1) causing 20% of FALS cases. Findings in ALS patients and model systems have implicated numerous genes in ALS pathogenesis, and have identified diverse processes, such as oxidative stress, excitotoxicity, mitochondrial dysfunction, protein aggregation, cytoskeletal abnormalities, impaired axonal transport, inflammation, and apoptosis, as contributing factors [1]. As a mainly sporadic disease affecting multiple cellular processes, ALS therefore suggests itself for comprehensive expression profiling and gene- and pathway-based analyses. The few existing genomics studies of ALS [2-4], however, have had limited genome coverage and have been restricted to gene-based analyses. Adding a further layer of complexity, ALS is a highly heterogenous disease, with clinical indicators helping to define ALS subtypes. One such indicator is a differential depletion of motoneurons in motor cortex and spinal cord [5-7], giving significance to the characterization of both tissues. The motor cortex contains upper motoneurons (UMNs), which extend axons traversing the corticospinal tract to signal to the spinal cord, where lower motoneurons (LMNs) relay their signal. The corticospinal tract volume is reduced [8] and UMNs are depleted [9,10] in ALS patients, and UMNs are required for LMN function and muscle control [11]. Most importantly, patients with sporadic, non-SOD1-associated forms of ALS show alterations in the motor cortex, such as increased excitability and reduced inhibitory activity, which are not readily detectable in SOD1-linked FALS patients, thus stressing the particular importance of investigating the motor cortex in SALS subjects [12-14]. Possibly owing to earlier clinical manifestations of defects in LMNs and their easier accessibility to experimenters, however, most investigations of ALS, including previous genomics studies [2-4], focus on the spinal cord and LMNs, leaving ALS-related responses and defects of cellular maintenance in the motor cortex under-investigated. Here we address the importance of UMN abnormalities in SALS pathology and exploit the power of pathway-based significance tests by whole-genome expression profiling of the motor cortex of SALS patients. We identify differentially expressed genes and pathways, interpret the role of candidate genes in ALS pathology using these pathways as a functional outline, and evaluate the implications of our findings for ALS research and the development of ALS therapies. Results Expression profiling of the motor cortex of SALS patients We investigated the motor cortex of eleven SALS and nine control subjects (see Table 1) with whole-genome oligonucleotide microarrays and following microscopic assessment of tissue architecture (see Figure 1 for Nissl staining of six representative samples). Out of over 41,000 genes and expressed sequence tags tested, 19,431 genes passed our quality control criteria (see the Methods section), constituting our complete data set for subsequent analyses. Median and average fold-changes of our complete data set were 1.01 (Diseased vs. Control), establishing that no systematic bias for up- or down-regulation was introduced into subsequent statistical analyses. To exploit the comprehensiveness of our data, we combined two complementary, stringent approaches, investigating changes in the SALS motor cortex for functional gene groups and on the level of individual genes. Pathway-based analysis We first performed a pathway-based analysis of the complete data set by testing changes of gene expression in the contex (...truncated)