Candidate genes for nicotine dependence via linkage, epistasis, and bioinformatics

Many smoking‐related phenotypes are substantially heritable. One genome scan of nicotine dependence (ND) has been published and several others are in progress and should be completed in the next 5 years. The goal of this hypothesis‐generating study was two‐fold. First, we present further analyses of our genome scan data for ND published by Straub et al. [ 1999 : Mol Psychiatry 4:129–144] (PMID: 10208445). Second, we used the method described by Cox et al. [ 1999 : Nat Genet 21:213–215] (PMID: 9988276) to search for epistatic loci across the markers used in the genome scan. The overall results of the genome scan nearly reached the rigorous Lander and Kruglyak [ 1995 : Nat Genet 11:241–247] criteria for “significant” linkage with the best findings on chromosomes 10 and 2. We then looked for correspondence between genes located in the 10 regions implicated in affected sibling pair (ASP) and epistatic linkage analyses with a list of genes suggested by microarray studies of experimental nicotine exposure and candidate genes from the literature. We found correspondence between linkage and microarray/candidate gene studies for genes involved with the mitogen‐activated protein kinase (MAPK) signaling system, nuclear factor kappa B (NFKB) complex, neuropeptide Y (NPY) neurotransmission, a nicotinic receptor subunit (CHRNA2), the vesicular monoamine transporter (SLC18A2), genes in pathways implicated in human anxiety (HTR7, TDO2, and the endozepine‐related protein precursor, DKFZP434A2417), and the µ1‐opioid receptor (OPRM1). Although the hypotheses resulting from these linkage and bioinformatic analyses are plausible and intriguing, their ultimate worth depends on replication in additional linkage samples and in future experimental studies. © 2003 Wiley‐Liss, Inc.