Zied Landoulsi, Ashwin Ashok Kumar Sreelatha, Nicole Kuznetsov, Claudia Schulte, Dheeraj Reddy Bobbili, Ludovica Montanucci, Costin Leu, Lisa‐Marie Niestroj, Emadeldin Hassanin, Cloé Domenighetti, Pierre‐Emmanuel Sugier, Milena Radivojkov‐Blagojevic, Peter Lichtner, Berta Portugal, Connor Edsall, Jens Krüger, Dena Hernández, Cornelis Blauwendraat, George D. Mellick, Alexander Zimprich, Walter Pirker, Manuela Tan, Ekaterina Rogaeva, Anthony E. Lang, Sulev Kõks, Pille Taba, Suzanne Lesage, Alexis Brice, Jean‐Christophe Corvol, Marie‐Christine Chartier‐Harlin, Eugénie Mutez, Kathrin Brockmann, Angela Deutschländer, G. Hadjigeorgiou, Efthimos Dardiotis, Leonidas Stefanis, Athina‐Maria Simitsi, Enza Maria Valente, Simona Petrucci, Letizia Straniero, Anna Zecchinelli, Gianni Pezzoli, Laura Brighina, Carlo Ferrarese, Grazia Annesi, Andrea Quattrone, Monica Gagliardi, Lena F. Burbulla, Hirotaka Matsuo, Akiyoshi Nakayama, Nobutaka Hattori, Kenya Nishioka, Sun Ju Chung, Yun Joong Kim, Lukas Pavelka, Pierre Kolber, Bart PC van de Warrenburg, Bastiaan R. Bloem, Andrew Singleton, Dan Vitale, Mathias Toft, Lasse Pihlstrøm, Leonor Correia Guedes, Joaquim J. Ferreira, Soraya Bardien, Jonathan Carr, E. Tolosa, Mario Ezquerra, Pau Pástor, Karin Wirdefeldt, Nancy L. Pedersen, Caroline Ran, Andrea Carmine Belin, Andreas Puschmann, Carl E Clarke, Karen Morrison, Dimitri Krainc, Matthew J. Farrer, Dennis Lal, the Global Parkinson Genetics Program (GP2), Yasser Mecheri, Emilia M. Gatto, Marcelo Kauffman, Federico Capparelli, Samson Khachatryan, Zaruhi Tavadyan, Mariam Isayan, Claire E. Shepherd, Kishore Kumar, Melina Ellis, Miguel E. Rentería, Simon Rowe, Dennis Yeow, Carolyn Sue, Victor Flores Ocampo, Laura Rudaks, Kanan Jafarov, David Crosiers, Artur F. Schumacher-Schuh, Carlos Rieder
Abstract Objective To investigate the impact of copy number variations (CNVs) on Parkinson’s disease (PD) pathogenesis using genome-wide data and explore their role in sporadic PD. Methods We analyzed CNV data from 11,035 PD patients (including 2,731 early-onset PD (EOPD)) and 8,901 controls from the COURAGE-PD consortium using a sliding window CNV-GWAS and genome-wide burden analysis. The independent dataset from the Global Parkinson Genetics Program (GP2) consisted of 23,089 cases and 18,824 controls were used to validate our initial findings. Results The exploratory dataset identifies multiple CNV regions associated with PD risk. The nominated CNV loci were not confirmed in an independent dataset, except that only a deletion in the PRKN gene, a well-established EOPD locus, remained genome-wide significant and robustly supported. CNV burden analysis showed a higher prevalence of CNVs in PD-related genes in patients compared to controls (OR=1.56 [1.18-2.09], p=0.0013), with PRKN showing the highest burden (OR=1.47 [1.10-1.98], p=0.026). Patients with CNVs in PRKN had an earlier disease onset. Burden analysis with controls and EOPD patients showed similar results. Interpretation The largest CNV-based GWAS on PD highlights both the promise and pitfalls of array-based CNV detection in PD and underscores the relevance of whole-genome sequencing approaches in resolving the role of CNV in PD. The array-based findings are prone towards false positive findings that might arise either from platform limitations and/or cohort biases. Future studies require improved genotyping resolution and rigorous cross-cohort validation to reliably assess CNV contributions to PD risk.