Exome sequencing of healthy phenotypic extremes links TROVE2 to emotional memory and PTSD

Angela Heck, Annette Milnik, Vanja Vukojevic, Jana Petrovska, Tobias Egli, Jochen Singer, Pablo Escobar, Thierry Sengstag, David Coynel, Virginie Freytag, Matthias Fastenrath, Philippe Demougin (+13 others)
37 38 Many mental disorders represent the extremes of normal distributions of traits 39 related to multiple cognitive and emotional dimensions. By performing whole-40 exome sequencing in healthy young subjects with extremely high versus extremely 41 low aversive memory performance we identified TROVE2 as a gene implicated in 42 emotional memory in health and disease. TROVE2 encodes Ro60, a broadly-43 expressed RNA-binding protein centrally implicated in the regulation of 44 inflammatory gene
more » ... ression and autoimmunity. A regulatory TROVE2 variant was 45 linked to higher emotional memory capacity and higher emotional memory-related 46 brain activation in healthy subjects. In addition, TROVE2 was associated with 47 traumatic memory and the frequency of posttraumatic stress disorder in genocide 48 survivors. 49 Results 84 85 Exome sequencing in phenotypic extremes 86 Whole-exome sequencing (WES) was performed in 88 healthy young participants 87 with extreme high or extreme low aversive memory performance carefully 88 matched for sex (1-to-1 matching), genetic background, age, and smoking behavior 89 (see Methods, Fig. 1, Fig. S1 , Table S1 ). Aversive memory was quantified by means 90 of a picture delayed free recall task. High-and low extremes were defined based on 91 the distribution of aversive memory performance in N = 3418 healthy young 92 subjects (see Methods). 93 WES was performed with the SureSelectXT Human All Exon V5+UTR target 94 enrichment kit (Agilent), which allows for sequencing exonic and near-gene 95 regulatory variants. To avoid discarding variants enriched to high frequency in the 96 extremes 21 , empirical minor allele frequency (MAF) in the extreme data set of N = 97 88 subjects was set to ≤ 0.125 (see Methods). Given that no prior information is 98 available regarding putative differences in effect sizes of variants associated with 99
doi:10.5451/unibas-ep54776 fatcat:25abekhts5g6vgywa4yh3hnbbe