Skip to main content
Premium Trial:

Request an Annual Quote

Drugs to Treat Smoking Addiction Suggested by Multi-Ancestry Transcriptome-Wide Association Study

NEW YORK — Researchers have developed a new approach for transcriptome-wide association analyses that incorporates multi-ancestry data, which they applied to uncover additional genes associated with smoking-related behaviors and suggest new drugs to treat smoking addiction.

Transcriptome-wide association studies (TWAS) generally rely on genome-wide association studies and expression quantitative trait locus (eQTL) data from cohorts with matched ancestries. This has limited the power of TWAS, though, when they have incorporated non-European ancestry data, a team led by researchers at Penn State College of Medicine and the University of Minnesota noted.

The researchers instead developed an approach they dubbed TESLA, a multi-ancestry integrative study using an optimal linear combination of association statistics. As they reported in Nature Genetics on Thursday, they applied their new method to examine tobacco-use phenotypes and uncovered hundreds of new genes linked to smoking behaviors, which they further used to identify drugs that could be repurposed to treat smoking addiction.

TESLA relies on three key steps: modeling phenotypic effects across ancestries using meta-regression, estimating phenotypic effects in the ancestry matching the eQTL dataset, and then combining TWAS results using multiple meta-regression models. In simulations comparing TESLA to other approaches like fixed-effect TWAS, random effect-TWAS, and a European TWAS using METASOFT data, the researchers found their approach outperformed the others, or scored similar to them, in various scenarios.

They then applied TESLA to data from 61 cohorts from the GWAS and Sequencing Consortium of Alcohol and Nicotine use (GSCAN) and Trans-Omics Precision Medicine (TOPMed) studies that examined four smoking traits: smoking initiation, cigarettes per day, smoking cessation, and age of smoking initiation. In all, these cohorts represented 1.2 million people from GSCAN and 150,000 participants of diverse ancestries from TOPMed. They further folded in data from the Genotype-Tissue Expression (GTEx) project on 48 tissue types from samples of European ancestry.

Using TESLA, they identified 4,475 gene-by-trait associations, more than FE-TWAS, RE-TWAS, and EURO-TWAS did. Within these, they found 783 gene-by-trait associations representing 384 unique genes within a dozen brain tissues. In particular, they uncovered 15 new genes associated with age of smoking initiation, 193 with links to smoking initiation, 19 linked to smoking cessation, and 46 associated with cigarettes per day. This is 23.5 percent more and 55.1 percent more than those identified by FE-TWAS and EURO-TWAS, respectively, they reported.

"Although TESLA optimizes the power for TWASs using existing eQTL datasets, it does not take away the need to generate eQTL datasets from non-European populations," the researchers noted. They added that the "ancestry of the eQTL dataset strongly influences the interpretation of TESLA results" and that the addition of non-European eQTL datasets will improve those interpretations.

Using pathway analyses, the researchers found that the genes they tied to smoking behavior were often involved in pathways with known links to addiction, such as the dopaminergic synapse pathway.

They also discovered shared pathways between smoking behaviors and other substance abuse or psychiatric phenotypes, suggesting that existing drugs affecting these pathways could be repurposed for use in smoking cessation efforts. In addition to drugs with known efficacy in smoking cessation, their analysis highlighted medications like dextromethorphan and galantamine for which there is already some preliminary clinical evidence, as well as new drug categories to explore such as muscle relaxants.

"Given the tremendous public health burden that continues to be incurred by smoking, repurposing drugs for smoking cessation is extremely valuable, because it offers a potentially quicker and more cost-effective route to treatment than the development of new therapeutic targets," the researchers wrote.