Chromosome 25 Signal Tied to Bacterial Kidney Disease Resistance in Atlantic Salmon

A large controlled challenge study of Atlantic salmon has uncovered a region on chromosome 25 that is consistently associated with resistance to bacterial kidney disease, one of the most damaging bacterial infections in salmon aquaculture. The study, published in Frontiers in Genetics, analyzed two generations of the same breeding line and found that fish carrying a specific two-SNP haplotype in that region showed, on average, a 17.45 percent reduction in mortality and roughly three additional days of survival under experimental challenge conditions. The researchers also estimated that the trait is moderately heritable, meaning that selective breeding could realistically improve resistance over successive generations.

Bacterial kidney disease, or BKD, is caused by the Gram-positive bacterium Renibacterium salmoninarum and is a chronic problem for salmon farming worldwide, with particular prevalence in cold northern waters of the Atlantic and Pacific Oceans. The disease produces nodular kidney lesions, lethargy, skin darkening, and hemorrhages, and it spreads both horizontally between fish sharing an environment and vertically from parents to eggs. In Chile, mortality associated with BKD accounted for 4.5 percent of total secondary mortality in 2024, and in the Magallanes region specifically the disease represented 26 percent of diagnoses reported for marine farms. Because losses accumulate through increased mortality, reduced fish health, and slower growth, breeding programs have long sought ways to build resistance into farmed populations.

The new work was carried out by researchers at Benchmark Genetics in Chile and Norway using fish from the SAGA strain, a breeding line originally established in Iceland and transferred to Chile in 2017. Two year classes from the same genetic line were evaluated: YC18, comprising 2,954 fish from 180 families, and YC22, comprising 2,210 fish from 190 families. All fish were pre-smolts. They were challenged at a specialized fish-pathogen facility by intraperitoneal inoculation with a single R. salmoninarum isolate, using a dose selected in a previous dose-finding study. The inoculum concentration was higher in the YC22 trial than in YC18 by roughly 56 percent, a difference the authors note may have influenced infection dynamics between the two cohorts. Fish were monitored for 60 days in YC18 and 56 days in YC22, with mortalities recorded daily and diagnostic assessment confirming findings consistent with BKD.

Across both challenges, mean mortality reached 72.4 percent and the mean time to death was 49.2 days, providing substantial phenotypic variation for genetic analysis. Macroscopic findings in the mortalities were compatible with BKD and included pale gills, splenomegaly, hepatic nodules, hepatomegaly, renomegaly, and pseudomembranes on internal organs. The researchers emphasize that because no uninfected control group was included, mortality and time-to-death should be interpreted as measures of relative resistance among fish exposed to the same standardized challenge. They also note that intraperitoneal inoculation bypasses natural mucosal exposure, so the results reflect systemic resistance under controlled conditions rather than the full complexity of natural infection routes.

After genotyping with a 70K SNP array and merging quality-controlled datasets, the analysis covered 47,816 SNPs across 4,846 individuals. Using a linear mixed model that accounted for population structure and relatedness, the genome-wide association study identified 54 significant SNPs associated with BKD resistance traits, of which 36 were located on chromosome 25. The strongest signals were concentrated there, with p-values as low as 5.94 × 10⁻³⁶ for the time-to-death trait. Thirteen SNPs were significant for both binary mortality and time to death, and the genetic correlation between the two traits was close to unity, suggesting that both are largely controlled by the same underlying genetic mechanisms. Most associations were detected for time to death, which the authors suggest provided greater power to capture genomic variation related to resistance.

Heritability estimates pointed to meaningful additive genetic variation. Using the full dataset, SNP-based heritability was 0.14 ± 0.01 for mortality and 0.19 ± 0.01 for time to death, figures close to those previously reported for a Canadian Atlantic salmon population using genomic information. When generations were analyzed separately, estimates in YC22 were higher than in YC18, which the authors attribute partly to differences in challenge conditions between the trials. Time to death consistently showed higher heritability than binary mortality, suggesting that this trait captures additional genetic variation related to BKD resistance.

Organizing the significant chromosome 25 SNPs into genomic blocks revealed that the association signal was not driven by a single isolated marker but by clusters of SNPs within defined regions. Thirteen genomic blocks were identified, and Block 9, spanning roughly 31.6 to 32.0 megabases, showed the strongest evidence of association, explaining 3.3 percent of the phenotypic variance and 16.8 percent of the genetic variance for time to death, and 2.3 percent and 15.65 percent respectively for mortality. Within this block, a haplotype defined by the A-A and C-C allele combination at two SNPs, AX-88110791 and AX-88118415, was associated with an average decrease in mortality of 17.45 percent and an increase of approximately three days in survival time, equivalent to a 5.6 percent increase in survival time. The favorable pattern held within each year class, with mortality reductions of 17.8 percent in YC18 and 16.2 percent in YC22, and remained significant in a mixed model that included family as a random effect.

Seven candidate genes were proposed within the prioritized regions: fam46c, man1a2, trim45, and vtcn1 in Block 9, pofut2 in Block 4, stam2 in Block 3, and fstl1b in Block 10. Several of these have documented immune-related functions. The gene vtcn1 encodes a known negative regulator of T-cell responses, and its deficiency has been shown to enhance innate immunity against bacterial pathogens in mouse models. The gene stam2 has been reported as a negative regulator of innate immunity in fish, while trim45 has previously been proposed as a candidate for sea lice resistance in Atlantic salmon and is involved in inflammatory regulation. The authors stress that these genes were selected for their location within regions explaining relevant phenotypic variance and their reported immune roles, and that further molecular validation, such as transcriptomic analyses, is required to confirm their functional relevance to BKD resistance.

The findings contrast with the only previous genome-wide study of BKD resistance in Atlantic salmon, which used a Canadian breeding population and found no evidence of major-effect loci. The authors caution that genetic architecture conclusions depend on study design, including population, SNP array density, and population structure, and they note that the chromosome 25 region identified here had not been reported before. They interpret the overall architecture as polygenic, influenced by loci of moderate effect rather than a few major genes, which carries direct implications for breeding strategy. Under a polygenic model, marker-assisted selection based on a small number of SNPs would capture only a limited proportion of the genetic variation, so the authors recommend that genomic selection remain the primary approach, with marker-assisted selection targeting the chromosome 25 region as a complementary tool, potentially through a low-density SNP chip for downstream pre-selection in populations chronically affected by the bacterium.

The authors conclude that validation in additional independent populations and generations will be required before the haplotype or the chromosome 25 region is implemented in breeding decisions. They also flag limitations of the challenge design, including the absence of an uninfected control group, the difficulty of reproducing identical infection dynamics across trials, and the fact that disease tolerance was not directly evaluated because pathogen burden was not measured. Even so, the consistency of the chromosome 25 signal across two generations, despite differences in fish age, body size, and inoculum concentration, suggests that the region represents a reproducible component of the genetic architecture of BKD resistance, one that could help breeders gradually reduce the toll of a disease that continues to burden salmon farms in Chile and other cold-water producing regions.

Subject of Research: Genetic architecture of bacterial kidney disease resistance in Atlantic salmon across two breeding generations

Article Title: Genome-wide association study and haplotype characterization of bacterial kidney disease (BKD) resistance in Atlantic salmon (Salmo salar) across two generations

Article References: Astudillo, A., Houston, R., Lhorente, J. P., & Garcia, B. (2026). Genome-wide association study and haplotype characterization of bacterial kidney disease (BKD) resistance in Atlantic salmon (Salmo salar) across two generations. Frontiers in Genetics, 17, Article 1985591. https://doi.org/10.3389/fgene.2026.1985591

Image Credits: AI Generated

DOI: 10.3389/fgene.2026.1985591

Keywords: Atlantic salmon, bacterial kidney disease, Renibacterium salmoninarum, GWAS, chromosome 25, haplotype, heritability, genomic selection, marker-assisted selection, aquaculture, disease resistance, candidate genes