Primary Progressive versus Bout Onset Multiple Sclerosis: GWAS and pathway-network analysis. G. Giacalone1, F. Clarelli1, P. Brambilla1, A. Osiceanu1, M. Sorosina1, V. Martinelli1, M. Leone2, S. D'Alfonso2, G. Comi1, F. Martinelli Boneschi1 1) San Raffaele Scientific Institute, Milan, Italy; 2) University of Eastern Piedmont, Novara, Italy.



Background: Genetic contribution to the two different Multiple Sclerosis (MS) courses, Primary Progressive MS (PPMS) and Bout Onset MS (BOMS), has previously been hypothesized by epidemiological studies, but so far no single genetic marker association has been found. Aim: to identify single nucleotide polymorphisms (SNPs) associated to MS course by comparing PPMS and BOMS through a genome-wide association study (GWAS) and a pathway-network analysis approach. Methods: a total of 444 PPMS and 541 BOMS patients genotyped on two different platforms were merged on common autosomal SNPs (296,589). After applying standard quality controls, logistic regression analysis was carried out including age, age at onset, gender and the first principal component as covariates. A pipeline of multiple-pathway and network analysis was developed to move from SNPs, through gene-wise p-values (p), to biological enriched terms/networks and potential disease-modifier genes. In the pipeline we used Gene Ontology and KEGG databases and the following bioinformatic tools: proxyGeneLD, MetaCoreTM, GOSTAT, Genecodis, Pathway-Express. Results: ProxyGeneLD program mapped 296,589 SNPs to 16,583 genes according to linkage-disequilibrium structure (HapMap CEU II 23a/hg18) and computed gene-wise adjusted p. Genes with adjusted p <0.05 (n=958) were prioritized for those directly interacting with each other (MetacoreTM protein interactions database) and the final list (n=218) submitted to multiple-pathway and network analysis. The first 3 hubs of the network composed of 218 genes were HIF1A, ETS1, NOTCH1, known to be involved in angiogenesis, immunity and myelination. Multiple-pathway analysis produced 32 significant terms, the majority of which related to immune functions (“B cell mediated immunity”, “T cell receptor signaling”, “Leukocyte transendothelial migration”, “Cytokine production”), “Canonical Wnt receptor signaling”, “L-glutamate transport”. Network analysis produced 11 significant networks, some of them enriched also in immune functions and Wnt signaling.



Conclusions: in the comparison of PPMS vs BOMS patients, pathway and network analyses approach suggest a possible role of genes related to immune functions, Wnt signaling and myelination processes, which should be further investigated in additional studies.