🧬 genetics

Genome-wide association study in Heterogeneous Stock rats identifies genetic loci associated with aversion-based learning and cocaine aversion

This genome-wide association study in Heterogeneous Stock rats identifies multiple heritable genetic loci and candidate genes associated with aversion-based learning and cocaine aversion, suggesting these traits are complex and may pleiotropically influence broader addiction biology.

Original authors: Tatom, Z., Eid, M., Missfeldt Sanches, T., Chitre, A. S., Ang, G., Ziegler, K. S., Peng, B., Keung, E., Nguyen, K.-M., Cohen, K., Wang, Y., Cheng, R., Chen, D., Johnson, B., Polesskaya, O., Jhou, T.

Published 2026-08-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Tatom, Z., Eid, M., Missfeldt Sanches, T., Chitre, A. S., Ang, G., Ziegler, K. S., Peng, B., Keung, E., Nguyen, K.-M., Cohen, K., Wang, Y., Cheng, R., Chen, D., Johnson, B., Polesskaya, O., Jhou, T., Palmer, A. A.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine your brain as a massive, bustling city where billions of tiny messengers (neurons) are constantly shouting instructions to one another. Sometimes, these messengers get the signal right, helping you learn that a hot stove burns or that a delicious cookie tastes great. But sometimes, the city's wiring gets a little glitchy, leading to bad decisions or habits that are hard to break. This is the world of addiction science. Researchers have long known that addiction isn't just a bad choice; it's a complex mix of biology and environment. A key part of the puzzle is "learning." We know that drugs like cocaine can feel amazing at first (the "reward"), but later, they can make you feel terrible (the "aversion"). The big question is: why do some people's brains learn to avoid those terrible feelings, while others keep chasing the high despite the crash? To solve this, scientists need to look at the "blueprints" of the brain—the genes—to see which ones might be responsible for these learning habits.

In this study, a team of researchers decided to play detective with a very special group of rats. They didn't use just any rats; they used "Heterogeneous Stock" rats, which are like a genetic melting pot. Imagine a giant family reunion where every cousin is slightly different from the next. This diversity is perfect for finding the specific genetic "typos" that might cause certain behaviors. The team put 1,074 of these rats through a series of challenging tests. Some rats ran down a hallway to get a hit of cocaine, while others had to work hard for a food treat, sometimes even enduring a tiny, harmless foot shock to get it. The goal was to see which rats were brave enough to keep going for the reward despite the pain (resistance to punishment) and which rats were smart enough to avoid the cocaine when it started feeling bad.

Once the tests were done, the scientists took a closer look at the rats' DNA. They were looking for specific spots on the genetic code that matched up with the rats' behavior. It's like finding that every rat who was super brave in the shock test had a tiny, unique mark on Chromosome 5, while every rat who avoided cocaine had a different mark on Chromosome X. And they found them! The study identified eight specific locations in the rats' DNA that were significantly linked to these learning behaviors.

The most exciting discovery was a cluster of genes on Chromosome 2 that seemed to control how the rats handled the foot shocks. These genes, named Cdh10, Cdh12, and Cdh18, act like the "glue" that holds brain cells together. Interestingly, humans with similar versions of these genes have been linked to starting to smoke or drink alcohol. This suggests that the same genetic "glue" that helps our brains learn to avoid pain might also influence whether we get hooked on substances. The researchers also found genes related to tiny, hair-like structures on brain cells called "cilia" (think of them as the cell's antennae). These antennae seem to play a role in how the brain processes rewards and punishments.

However, the story isn't a simple "we found the cure." The paper suggests that these behaviors are incredibly complex, influenced by many different genes working together, not just one magic switch. While the rats' behavior was strongly linked to their genes (with some traits being up to 30% heritable), the specific genes found for avoiding cocaine were different from the ones found for avoiding food shocks, except for a shared area on the X chromosome. This tells us that while the brain's ability to learn from bad experiences is a shared trait, the specific genetic recipes for avoiding drugs versus avoiding pain might be slightly different. The study doesn't prove exactly how these genes work in humans yet, but it provides a fantastic map of where to look next, pointing scientists toward the specific parts of the brain's wiring that might need fixing to help people break free from addiction.

Technical Summary: Genome-wide association study in Heterogeneous Stock rats identifies genetic loci associated with aversion-based learning and cocaine aversion

Problem Statement
Addiction is a complex, heritable trait progressing through developmental stages influenced by partially overlapping genetic factors. While much research focuses on aberrant reward learning (the initial euphoric effects of drugs), the progression to Cocaine Use Disorder (CUD) is also driven by aberrant avoidance-based learning, where repeated exposure desensitizes rewarding effects and sensitizes aversive effects (anxiety, craving, withdrawal). Human Genome-Wide Association Studies (GWAS) for CUD have yielded limited replicable findings, often due to small sample sizes of cocaine-exposed individuals and uncontrolled environmental variables. Consequently, there is a need for animal model GWAS that can control environmental factors to better elucidate the genetic architecture of avoidance-based learning and cocaine aversion.

Methodology
The study utilized an outbred Heterogeneous Stock (HS) rat population, which offers high genetic diversity and recombination events suitable for fine-mapping complex traits.

Subjects: A total of 1,174 HS rats were genotyped, with 1,074 animals (35.3% male) passing sample-level quality control metrics for phenotypic analysis.
Behavioral Assays:
- Runway Operant Cocaine-Seeking: Rats were trained to traverse a corridor to receive an intravenous cocaine infusion (0.75 mg/kg). Metrics included latency to reach the goal box ("Run Start") and the number of direction reversals ("Run Reversals").
- Food-Based Aversion Tasks: Rats were food-restricted and trained to lever press for food pellets. They underwent:
  - Progressive Ratio (PR): Measuring the effort (lever presses) required to obtain a reward.
  - Punishment Resistance: Measuring the maximum footshock intensity endured to obtain a reward.
  - Delay Punishment: Comparing shock tolerance at 0 vs. 40-second delays.
- Locomotion: Exploratory activity in a novel environment was recorded over two sessions.
Genotyping and Analysis: Spleen samples were sequenced using Illumina NovaSeq 6000. Reads were aligned to the mRatBN7.2 reference genome, and haplotypes were constructed to impute 7,347,327 SNPs. After quality control (removing low imputation quality, high missingness, low MAF, and HWE deviations), 5,416,497 SNPs remained.
Statistical Framework:
- Heritability ( $h^2$ ) and genetic correlations ( $r_g$ ) were estimated using GCTA-GREML.
- GWAS was performed using a Mixed Linear Model (MLM) with a leave-one-chromosome-out approach to account for relatedness.
- Significance thresholds were determined via permutation analysis ( $n=1000$ ), setting genome-wide significance at $-\log_{10}(p) = 5.54$ ( $p < 0.05$ ) and suggestive significance at $5.16 $($ p < 0.10$).
- Candidate genes were prioritized using SnpEff for variant effects, RatGTEx for eQTL/sQTL, and the GWAS Catalog for human ortholog associations.
- A Phenome-Wide Association Study (PheWAS) was conducted using the Center for GWAS in Outbred Rats Database (C-GORD) to link identified loci to other HS rat phenotypes.

Key Results

Heritability: Most behavioral phenotypes were significantly heritable, with estimates ranging up to $h^2 = 0.307$ for Run Reversals (combined sexes) and $h^2 = 0.347$ for Body Mass. The binary "High Cocaine Avoid" trait was not significantly heritable.
Genetic Correlations: Significant genetic correlations were found between traits across different assays, such as between Progressive Ratio and Punishment, and between Delay Punishment Ratio and Run Start. Hierarchical clustering suggested overlapping genetic architectures for aversion-based learning traits.
Quantitative Trait Loci (QTL): The study identified eight significant ( $p < 0.05$ $p < 0.05$ ) and five suggestive ( $p < 0.10$ $p < 0.10$ ) loci.
- Punishment Task: Four significant loci were identified on Chromosomes 2, 3, 5, and 6. The Chromosome 5 locus showed the highest significance among autosomal QTLs ( $-\log_{10}(p) = 7.189$ ).
- Cocaine Avoidance: Three significant loci were identified on the X chromosome, overlapping for "High Cocaine Avoid" (both sexes and females only) and "Run Start" (males only).
- Locomotion: A significant locus was found on Chromosome 5 for Session 2 locomotor activity.
Candidate Genes: 172 positional candidate genes were identified. Notable findings include:
- Cadherin Cluster (Chr 2): The Punishment locus on Chromosome 2 contained Cdh10, Cdh12, and Cdh18. These genes are previously associated with smoking initiation, nicotine dependence, schizophrenia, and cognitive function in human GWAS.
- Neuronal Cilia Genes: The Punishment locus on Chromosome 6 contained Adcy3, Drc1, and Cfap206. Adcy3 is associated with smoking initiation and body mass index; Drc1 is linked to ciliary dyskinesia; and Cfap206 is involved in brain development. These suggest a role for primary neuronal cilia in addiction-related behaviors.
- X Chromosome: The overlapping loci on the X chromosome included Pcdh11x, a protocadherin regulating cell-cell adhesion and cerebral asymmetry.
PheWAS Findings: Variants associated with the Punishment task were also linked to reaction time, novelty-seeking behavior, body mass, and social interaction phenotypes in independent HS rat samples.

Significance and Claims
The authors claim that this study successfully identifies specific genetic loci associated with aversion-based learning and cocaine aversion in an outbred rat model. The results demonstrate that these traits are complex, heritable, and influenced by multiple genetic loci. A key contribution is the identification of a shared genetic architecture between food-based punishment resistance and cocaine avoidance, particularly on the X chromosome. Furthermore, the study highlights that genes associated with these aversion phenotypes in rats (specifically cadherins and neuronal cilia genes) have established associations with substance use (smoking, alcohol) and other behavioral phenotypes in human GWAS. The authors conclude that these aversion learning phenotypes are not only heritable but may pleiotropically affect other aspects of addiction biology, such as novelty-seeking and externalizing behaviors, providing a genetic basis for the transition from drug use to disorder. The study acknowledges limitations, including an imbalanced sex ratio and the potential influence of food reward value on specific loci (e.g., Adcy3 and adiposity), but maintains that the findings support the utility of avoidance-based learning models in understanding the genetic etiology of Substance Use Disorders (SUD).

More like this