Search

International audience

International audience

I investigated the genetic architecture underlying variation in embryonic developmental rate in rainbow trout ('Oncorhynchus mykiss ') and tested whether there was genetic covariation with age at maturation. Both dams and sires influence the hatching time of their progeny indicating that developmental rate is under significant genetic control. Genome scans with 127 microsatellite loci from all 29 linkage groups in two families detected four significant quantitative trait loci (QTL) for developmental rate. A QTL on linkage group RT-8 explained 23.8% of the phenotypic variation and supports results from previous studies. I found inconsistent support for the hypothesis that embryonic developmental rate is genetically coupled to age of maturation. Maturation background and the maturation schedule of sires did not have a consistent effect on progeny hatching times. However, alleles associated with faster developmental rate were found significantly more often in early maturing rather than typical/later maturing male ancestors from one family.

Clock family genes encode transcription factors that regulate clock-controlled genes and thus regulate many physiological mechanisms/processes in a circadian fashion. 'Clock1' duplicates and a copy of 'NPAS2'-like gene were partially characterized by genomic sequencing and mapped using family-based indels/SNPs in rainbow trout (RT)('Oncorhynchus mykiss'), Arctic charr (AC)('Salvelinus alpinus'), and Atlantic salmon (AS)(' Salmo salar') mapping panels. 'Clock1' duplicates mapped to linkage groups RT-8/-24, AC-16/-13, and AS-2/-18. 'Clock3/NPAS2 '-like genes mapped to RT-9/-20, AC-20/-43, and AS-5. Most of these linkage group regions containing the 'Clock' gene duplicates were derived from the most recent 4R whole genome duplication event specific to the salmonids. These linkage groups contain quantitative trait loci for life history and growth traits (i.e., reproduction and cell cycling). Comparative synteny analyses with other model teleost species reveal a high degree of conservation for genes in these chromosomal regions, suggesting that functionally related or co-regulated genes are clustered in syntenic blocks.

To identify quantitative trait loci (QTL) influencing early maturation (EM), body weight (BW) and condition factor (K) in rainbow trout (' Oncorhynchus mykiss'), a genome scan approach with 96 microsatellite loci across 29 linkage groups was utilized. The six inter-strain paternal half sib families used were derived from two strains that differ in the propensity for EM. Several significant QTL for EM localized to different regions in males and females, suggesting some sex-specific control. Some QTL for BW and K localized to regions identified in previous studies, as well as where candidate genes for these traits have been identified. QTL for multiple traits often co-localized, suggesting a level of genetic correlation. The highest prevalence of co-localization occurred between EM and BW. Linkage group RT-8, which has strong connections with developmental and early life history traits, had detected effects for all traits studied.

A genome scan, using microsatellite markers, was performed to detect quantitative trait loci for age at sexual maturation in two full-sib families of North American (NA) Atlantic salmon. Furthermore, linkage arrangements were compared between these NA fish and those found in the current European (EU) Atlantic salmon map. I found eleven QTL for age at maturity, as well as a rearrangement of the sex linkage region compared to EU fish. Four of the QTL were localized to two homeologous chromosomal segments. In addition, QTL were localized to linkage groups where similar effects were detected on the homologous linkage group regions in Arctic charr ('Salvelinus alpinus ') and rainbow trout ('Oncorhynchus mykiss'). The conservation of QTL effects across two other species of Salmonidae provide support for the identification of the QTL detected in Atlantic salmon, and suggest that certain regions of the salmonid genome retain ancestrally related syntenies with constrained functions.

Knowledge from previous studies directed at locating quantitative trait loci (QTL) for body weight (BW), condition factor (CF) and maturation timing (MT) in rainbow trout was used to investigate these putative QTL associations across 12 linkage groups in a diallel maternal x paternal half-sib family design. Three F1 inter-strain brothers (Spring Valley x Rainbow Springs cross), all descendants of the same granddam known to transmit an extremely high incidence of precocious maturation to her offspring, were crossed with a pure Spring Valley female and a hybrid F, female. Significant (p < 0.05) BW QTL were identified on RT-7, RT-8, RT-9, RT-13, RT-15, RT-17, RT-23, RT-24 and RT-31. Significant CF QTL were identified on RT-8, RT-13, RT-15, RT-19, RT-21, RT-23 and RT-24. Three and two of the five parental fish demonstrated MT QTL on RT-8 and RT-23, respectively, independent of BW QTL indicating that these regions may possess a true and major genetic influence on MT that is largely unconfounded by the fish's body mass growth trajectory. Two separate regions on RT-24 appear to be associated with either BW or MT and may confer a pleiotropic effect on these traits. Three of the five parental fish showed evidence of MT QTL on RT-24, with two of these three QTL effects unrelated to BW. Increased fine mapping of these regions would be beneficial to the understanding of their independent effects.

I constructed genetic linkage maps for North American (NA: New Brunswick) Atlantic salmon ('Salmo salar') and compared them to those of European fish (EU: Scotland). I used 91 microsatellite loci on 16 linkage groups to compare marker arrangement in three NA full-sib families. The NA fish showed greater recombination rates between markers compared to the EU fish. Two potential rearrangements on linkage groups AS-8 and AS-17 were identified in the NA fish. Three significant quantitative trait loci (QTL) for body weight (AS-8, AS-10, AS-11) and two significant QTL for condition factor (AS-4, AS-14) were identified and accounted for between 7.5% and 28.6% of the trait variation. Two of the significant QTL for body weight occur on linkage groups where similar effects have been detected on the homologous regions in rainbow trout ('Oncorhynchus mykiss') and Arctic charr ('Salvelinus alpinus').

A multiplex genotyping system was applied to Atlantic salmon (' Salmo salar') being reared by the Atlantic Salmon Broodstock Development Program in Chamcook, New Brunswick. Nine microsatellite loci were used to determine the parents of 98% of 305 progeny from the 1998 year-class reared in a sea cage. The genotyping system validated family identity of 299 fish, which had been initially identified by branding. The genotyping system was also able to discriminate among all but 5 of 79 families in the 1998 year-class when progeny genotypes were generated from genotypes of known parental pairs. In a second experiment, two quantitative trait loci for body weight were identified, using progeny from two full-sib families of the 1998 year-class. Out of 25 microsatellite loci tested, association between variation and body weight was detected at two of these microsatellite loci. The QTL explained between 9.6 and 14% of the variation in body weight.

The localisation of variance for fitness-related traits to specific chromosomal regions may elucidate the underlying system of the genetic control of quantitative traits, potentially permitting selection on these traits. General linear modeling of allele inheritance at molecular marker loci was used to detect the presence and nature of quantitative trait loci (QTL) in two commercial strains of rainbow trout ('Oncorhynchus mykiss ') controlling variation in three traits: upper thermal tolerance (UTT), length (fork length, FL) and condition factor (residual fatness). Genetic variation was surveyed on three autosomes (linkage groups B, D and S) and the sex chromosomes (linkage group 18). One autosomal and one sex-linked QTL with major-moderate effects (7-10%) on UTT (on linkage groups S and 18, respectively) were detected, although QTL effects accounted for 1.5-10.0% of phenotypic variance. The existence of these QTL was determined both by testing associations between alleles inherited by progeny (i) from parents and (ii) from their grandsires. A sex-linked QTL for FL was also detected, suggesting pleiotropy or linkage disequilibrium among QTL for FL and UTT on the sex chromosomes. Physiological epistasis, the interactive effect of specific combinations of loci on phenotype, was highly locus-specific, with epistasis between genes on linkage groups B and 18 accounting for the majority of the cases in which epistatic effects were detected for all three traits, but particularly for UTT. In tests for physiological epistasis, females were more phenotypically divergent from their full and half sib family mean phenotype than males. Parental effects on UTT and size traits were detected in embryonic rainbow trout, indicating heritable variance for these traits. While autosomal UTT QTL on linkage groups B, S and 18 did not appear to affect UTT in embryonic rainbow trout, QTL affecting body length and weight at this stage were detected on linkage groups S and B. Physiological epistasis between genes on linkage group S and 18 on embryo length was extremely strong in one half sib family (r2 = 0.18; p < 0.0001). My results were suggestive of polygenic control of these traits (by numerous QTL), operating within a Wrightian framework, with physiological epistasis strongly influenced by sex-linked genes.

I investigated the genetic architecture and relationship between growth and age of maturation in two hatchery strains of rainbow trout ('Oncorhyncus mykiss'). Microsatellite loci known to be associated with body weight in an experimental backcross family were used to detect and verify quantitative trait loci (QTL) influencing these two traits. Two males from each strain were used to sire four paternal half-sib families for QTL detection. Markers on linkage groups 5, B, G, N, and Oi showed significant associations (p < 0.05) with body weight measured at three different ages (days post fertilization). This supports previous observations on the location of putative growth QTL. Associations between allelic segregation and age at sexual maturation were also detected on linkage groups 5, G, and N in the Rainbow Springs strain. These associations were weaker than those detected for growth and are most likely due to a genetic correlation (pleiotropy or close linkage of separate QTL) between the traits of growth and precocious maturation.

Two multiplex polymerase chain reaction (PCR) genotyping systems were developed for rainbow trout (Oncorhynchus mykiss) using 14 fluorescently labeled dinucleotide repeat microsatellite loci. These genotyping systems, plus two additional microsatellite loci, revealed significant allele frequency heterogeneity across the spawning distribution of a hatchery broodstock. Distinct patterns in allele frequency variation for 10 of the loci made it possible to predict female spawning time at the hatchery. In a second experiment, 93% of a group of progeny could be assigned to single parental pairs based on microsatellite genotypes determined from the two genotyping systems. Through the use of these genetic pedigrees, it was determined that genetic correlations (multiple trait REML estimates) between three character sets (early progeny growth; later progeny growth at 8.5$\sp\circ$C; later progeny growth at 15$\sp\circ$C) were high, suggesting that early progeny growth is a good predictor of later progeny growth and that the magnitude of GxE is low.