Genome-wide identification of the shared genetic basis of cannabis and cigarette smoking and schizophrenia implicates NCAM1 and neuronal abnormality

Objectives: Confirming the existence and composition of the shared genetic basis of Schizophrenia and cannabis and cigarette smoking has critical values for the clinical prevention and intervention of psychosis. Methods: To achieve this goal, we leveraged Genome-Wide summary statistics of Schizophrenia (n = 99,934), cigarette smoking (n = 518,633) and cannabis usage (n = 162,082). We applied Causal Analysis Using Summary Effect Estimates (CAUSE) and genomic structural equation modeling (GenomicSEM) to quantify the contribution of a common genetic factor of cannabis and cigarette smoking and schizophrenia (referred to as SCZ_SMO), then identified genome-wide loci that made up SCZ_SMO. Results: We estimated that SCZ_SMO explained 8.6% of Schizophrenia heritability (Z score <-2.5 in CAUSE, p<10(-20) in Genomic SEM). There were 20 independent loci showing association with SCZ_SMO at the genomewide threshold of p<5 Chi 10(-8). At the top locus on chromosome 11, fine-mapping identified rs7945073 (posterior inclusion probability =0.12, p = 2.24 Chi 10(-32)) as the top risk variants. Gene-level association and fine-mapping highlighted NCAM1, PHC2, and SEMA6D as risk genes of SCZ_SMO. Other risk genes were enriched in cortex, neuron, and dendritic spines (adjusted p<0.05). SCZ_SMO showed significant positive correlation (p<10(-6)) with the genetic risk of attention deficit hyperactivity disorder (r = 0.50), lifestyle problems (r = 0.83), social deprivation (r = 0.58) and all-cause pregnant loss (r = 0.60). Conclusion: Our result provided new evidence on the shared genetic basis model for the association between Schizophrenia and smoking and provided genetic and biological insights into their shared mechanism.

Automated summary

This study confirms the existence of a shared genetic basis between schizophrenia, smoking, and cannabis usage, and identifies specific genes associated with this relationship. These findings have important implications for the clinical prevention and intervention of psychosis.