Non-classical Splice-site Splicing-disrupt Variants And Associated Gene Expression Patterns In Parkinson’s Disease

Background: Non-classical splice-site splicing-disrupt variants contribute significantly to the genetic risk of numerous genetic diseases by affecting precursor messenger RNA(pre-m RNA)splicing.However,up to now,there are no large-scale and systematic studies to investigate the impact of non-classical splice-site splicing-disrupt variants in Parkinson’s disease(PD)causing genes.Meanwhile,the expression pattern characteristics of genes represent a phenotype to a certain extent.However,the expression patterns of PD-associated genes and their relationship with age at onset(AAO)remain unclear.Object: To comprehensively assess the genetic contribution of non-classical splice-site splicing-disrupt variants in 23 PD-causing genes to PD in a large cohort and to describe 107 PD-associated gene expression pattern characteristics using public databases.Methods:(1)Based on the whole-exome sequencing data of 1676 PD patients,we used the Splice AI and SPIDES to screen for non-canonical splice-site variants that may lead to splicing abnormalities in PD-causing genes.Then,minigene RNA splicing experiments or patient blood RNA analyses were used to verify the reliability of the software’s predictions.(2)To describe the expression pattern characteristics of PD-associated genes,we included 107 PD-associated genes.The expression data from the GTEx,Allen Brain Map,and Brain Span databases were extracted to characterize the tissue specificity,inhibitory-excitatory neuron expression profile,and spatio-temporal expression pattern of PD-associated genes,respectively.The correlation between PD-causing genes and AAO was analyzed using MDSgene and Gene4 PD database data.Results:(1)12 non-canonical splice-site variants that may affect splicing were screened out by bioinformatics software prediction.GIGYF2 and EIF4G1 were the most common PD-causing genes with putative non-canonical splice sites splicing-disrupt variants.Among the14 PD patients,three patients carried two novel GIGYF2 missense variants GIGYF2 c.1905A>G(n=2)and GIGYF2 c.1079A>G(n=1),and three patients harbored two novel EIF4G1 mutations,of which two took EIF4G1 c.2067+5G>A intron variation and the other one carried EIF4G1 c.64 G>C variation.The GIGYF2 c.1079A>G(p.D360G)missense variant was further verified by minigene RNA splicing experiments to cause abnormal splicing.While the remaining 11 mutations were demonstrated to have no effect on splicing.(2)107 PD-associated genes were significantly expressed in 12 tissues,all of which belong to the brain,including the substantia nigra,hippocampus,amygdala,spinal cord,frontal cortex,hypothalamus,putamen,anterior cingulate cortex,caudate,nucleus accumbens,and cerebellum.Among the 57 PD-associated genes with significantly different expression levels in excitatory and inhibitory neurons,34 genes showed higher expression levels in excitatory neurons than in inhibitory neurons.Furthermore,two independent co-expression modules(M1 and M2)comprising 69 PD-associated genes were detected.M1 integrated genes showed low expression in the human embryonic phase,which gradually increased in expression starting in the prenatal period(16–18 post-conceptual weeks),and reached a highly stable expression level after birth(both in the adolescent and adult period).M2 integrated genes were highly expressed during the embryonic and early-to-middle fetal periods(8–24 post-conceptual weeks)but showed decreased expression starting in the prenatal and adult periods.Finally,the median AAO of patients with mutations in PD-causing genes in the first module was lower than that of the second module.Conclusions:(1)Non-classical splice-site splicing-disrupt variants in PD-causing genes are rare.(2)The possibility of PD-associated gene expression in brain tissue is higher than in non-brain tissues.The expression level of PD-associated genes in excitatory neurons is higher than in inhibitory neurons.Two spatio-temporal expression modules of PD-associated genes in the human brain: the first module showed a higher expression level in the adult period than in the prenatal period,and the second module showed the opposite features.12 figures,32 tables,and 129 references...