Mapping of quantitative trait loci in Atlantic salmon and tilapia / Kartlegging av Quantitative Trait Loci i Atlantisk laks og tilapia
The number of genes influencing a trait may vary from a few to tens or even hundreds. Traits that are controlled by more than a few genes, or have significant environmental components, are called quantitative traits. For a quantitative trait, genotypes at trait-controlling genes cannot be deduced directly from the phenotype. It is, nevertheless, possible to locate genomic regions containing genes affecting the trait, by observing the co-segregation of trait values and genetic marker genotypes in a resource population. This is the basis for Quantitative Trait Loci (QTL) mapping, the mapping of genes with effect on a quantitative trait. The main topic of this thesis is the mapping of QTL for resistance against Infectious Salmon Anaemia (ISA) in Atlantic salmon (Salmo salar). ISA is a viral disease causing substantial losses for farmers of Atlantic salmon in Norway and several other countries. There is genetic variation in resistance towards the disease, but very little is known about the genetics underlying this variation. Genes affecting resistance could be used to improve the efficiency of selection for resistance, through so-called Marker Assisted Selection (MAS).
A genome scan was performed on two full-sib families, within one paternal half-sib family, of Atlantic salmon from a Norwegian breeding population. The families had been through a challenge test for ISA resistance. Several hundred Amplified Fragment Length Polymorphism (AFLP) markers were genotyped on parents and offspring. For the statistical analysis, a multi-stage testing strategy was developed that incorporated 1) a within-family Transmission Disequilibrium Test (TDT) on affected offspring, 2) a test for non-Mendelian segregation, and 3) a test for correlation between marker genotypes ant the trait “days in test”, based on survival analysis. In the third test stage, both affected and resistant offspring were considered, but only markers that were significant for the TDT and displayed Mendelian segregation were analysed. The benefit of the strategy was that for most markers, only affected offspring needed to be genotyped. Several putative ATL were identified in the genome scan, though only two were experiment-wise significant.
Next, 54 microsatellites were genotyped on the families from the genome scan. These genotypes, together with the AFLP genotypes, were used to construct a genetic linkage map of Atlantic salmon. The initial mapping results showed that there was an unusually large difference beteween the male and female recombination rates. A mapping method was developed that took advantage of this difference, together with the half-sib family structure, to facilitate more efficient mapping of dominant AFLP markers. Separate male and female maps were made. The lengths of the male and female maps were 103 and 901 cM, respectively. The ratio of female to male crossover events was 8.3. This is the largest sex-specific difference in recombination rate observed in any vertebrate.
When the data from the genome scan was combined with the results from the linkage mapping, it became clear that linkage group 1 was particularly likely to contain a true QTL. We therefore wanted to investigate this linkage group in a larger family material, to see if the QTL was replicable. Twenty-five additional full-sib families from the same challenge test were genotyped for 8 microsatellites across linkage group 1. An interval mapping program using the Cox proportional hazard model was developed and put to use. The QTL was found to be experiment-wise significant at P < 0.01. The most likely position of the QTL was in the centre of the linkage group, although the 95 % bootstrap confidence interval for QTL position covered the whole linkage group. Significant non-Mendelian segregation was observed in the same region (P < 0.05), but the data showed that the observed QTL was not an artefact coming from non-Mendelian segregation.
In addition to the Atlantic salmon study described above, a QTL study in tilapia was done, searching for QTL for body weight and cold tolerance in four-way crosses (4WC) between different tilapia species. A genome scan was performed on one 4WC family, using 54 microsatellites and 23 AFLP primer combinations. Several putative QTL were found, although none were experiment-wise significant. The putative QTL were next tested for replication in another 4WC family. The two families were sired by the same male, and the mothers were full sibs. None of the putative QTL from the first family were replicated in the second family. However, one marker showed a strong association to cold tolerance in the second family, the marker being located on linkage group 23. The same marker had been found to be associated with body weight in the first family. Because QTLs for both weight and cold tolerance had been reported on linkage group 23 in another, recent, study, we decided to investigate this linkage group more closely. Interval mapping was performed on the second family, using 10 microsatellites across linkage group 23. The results showed strong indications of a QTL for cold tolerance segregating in the sire, and weaker evidence of the same QTL segregating in the dam. The study therefore supported the existence of a QTL for cold tolerance on linkage group 23.
Pr. desember 2004: Thomas Moen, Akvaforsk AS, Postboks 5010, 1432 Ås
