Histone deacetylase gene variants predict brain volume changes in multiple sclerosis Becky Inkster a,b,1, *, Eva M.M. Strijbis c,d,1 , Maria Vounou b , Ludwig Kappos e , Ernst-Wilhelm Radue f , Paul M. Matthews a,g , Bernard M.J. Uitdehaag c,h , Frederik Barkhof i , Chris H. Polman c , Giovanni Montana b,2 , Jeroen J.G. Geurts d,2 a Department of Medicine, Centre for Neuroscience, Hammersmith Hospital, Imperial College London, London, UK b Department of Mathematics, Statistics Section, Imperial College London, London, UK c Department of Neurology, VU University Medical Center, Amsterdam, The Netherlands d Department of Anatomy and Neurosciences, Section of Clinical Neuroscience, VU University Medical Center, Amsterdam, The Netherlands e Department of Biomedicine, University Hospital, Basel, Switzerland f Institute of Radiology, Department of Neuroradiology, University Hospital, Basel, Switzerland g GlaxoSmithKline Clinical Imaging Centre, Hammersmith Hospital, London, UK h Department of Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands i Department of Radiology, VU University Medical Center, Amsterdam, The Netherlands Received 13 September 2011; received in revised form 5 July 2012; accepted 11 July 2012 Abstract Neuroimaging measures hold promise for enhancing the detection of disease-related genetic variants. In this study, we use advanced multivariate regression methods to assess the predictive value of single nucleotide polymorphisms (SNPs) on several brain volumetric- and lesion-related neuroimaging measures in a well-characterized cohort of 326 patients with multiple sclerosis (MS). SNP selection was constrained to key epigenetic regulatory genes to further explore the emerging role of epigenetics in MS. Regression models consistently identified rs2522129, rs2675231, and rs2389963 as having among the highest predictive values for explaining differences related to brain volume measures. These SNPs are all contained in genes from the same superfamily, histone deacetylases, which have biological functions that are relevant to MS, neurodegeneration, and aging. Our preliminary findings generate hypotheses for testing in future independent MS data sets as well as other neurodegenerative conditions. © 2013 Published by Elsevier Inc. Keywords: Multiple sclerosis; Brain volume; Multivariate regression; Histone deacetylase; Mitochondria; Sirtuin 1. Introduction Multiple sclerosis (MS) is an inflammatory disease in- volving demyelination and neurodegeneration (Hauser and Oksenberg, 2006). The disease is characterized by neuro- degenerative events and autoimmune attacks against myelin in the central nervous system leading to varying degrees of relapsing or progressive neurological impairments (Vukusic and Confavreux, 2007). The clinical course and disease progression of MS is highly variable, and is likely to depend on complex heritable (genetic and epigenetic) and environ- mental factors (Lauer, 2010; Oksenberg and Baranzini, 2010; Urdinguio et al., 2009). Studies are starting to unravel some of the genetic factors that influence MS disease susceptibility, of which the ma- jority relate to immune system functions (Disanto et al., 2011). Additional genetic factors influencing specific fea- * Corresponding author at: Department of Psychiatry, University of Cambridge, Douglas House, 18b Trumpington Road, Cambridge CB2 8AH, UK. Tel.: +44 (0) 1223 746040; fax: +44-(0)1223 746033. E-mail address: bi212@medschl.cam.ac.uk (B. Inkster). 1 These joint first authors contributed equally to this report. 2 These joint last authors contributed equally to this report. Neurobiology of Aging 34 (2013) 238 –247 www.elsevier.com/locate/neuaging 0197-4580/$ – see front matter © 2013 Published by Elsevier Inc. http://dx.doi.org/10.1016/j.neurobiolaging.2012.07.007