Genetic analysis of a development rate QTL in backcrosses of clonal rainbow trout, Oncorhynchus mykiss

Abstract

Backcrosses of clonal rainbow trout, Oncorhynchus mykiss, were used in an effort for fine-mapping of a quantitative trait locus (QTL) for rapid embryonic development rate (tth-1). A previously identified QTL was backcrossed for two successive generations from the Clearwater (fast developing) line into the Oregon State University (OSU) (slow developing) background. Microsatellites tightly linked to the QTL and each other were used to genotype individuals in both backcrosses. An association was found between the microsatellites and gamete production in the first backcross and precocious gamete production in the second backcross generation. These results suggest that a region near the QTL-linked microsatellites may influence early sexual maturity, as well as development rate. Two gynogenetic families were produced using individuals from the first backcross (BC₁). These crosses were screened with the tth-1 linked microsatellites and time to hatch was quantified. The Clearwater genotype was associated with early time to hatch, while the OSU genotype was associated with late hatching. Genotypic variation at these microsatellite loci explained 28% and 26% of the variation in time to hatch. Heterozygote frequencies for the microsatellites in the two gynogenetic families were 0.0088 and 0.01, suggesting that the QTL is near the centromere. Fine-mapping was limited due to the low level of recombination observed on this linkage group in the gynogenetic families.

Introduction

The identification of a quantitative trait locus (QTL) is one step in determining the genes associated with variation in a particular trait. Loci influencing quantitative traits have been identified in numerous organisms (Paterson et al., 1988, Jacob et al., 1991, Edwards et al., 1992, Stuber et al., 1992, Georges et al., 1993, Andersson et al., 1994, Sakamoto et al., 1999, Huang et al., 2003) and knowledge of such linkage relationships has been incorporated into breeding programs (Saito et al., 1991, Bumstead and Palyga, 1992, Georges et al., 1993, Bouchez et al., 2002, Kwon et al., 2002). QTLs for traits such as spawning time (Sakamoto et al., 1999, O'Malley et al., 2003), embryonic development rate (Robison et al., 2001), upper thermal tolerance (Jackson et al., 1998, Perry et al., 2001), disease resistance (Ozaki et al., 2001, Nichols et al., 2003), and counts for meristic characters (Nichols et al., 2004) have been identified in rainbow trout (Oncorhynchus mykiss).

Clonal lines of rainbow trout have been used to develop framework genetic maps (Young et al., 1998, Robison et al., 2001, Nichols et al., 2003) and a QTL for development rate has been identified in two different crosses (Robison et al., 2001, Nichols, 2002). The linkage maps from these crosses are useful but have limitations. One limitation is that the current levels of resolution at the QTLs are not adequate for fine mapping. A second related limitation is that the genetic map is derived from male meioses. Male meioses in salmonids exhibit a high degree of crossover interference relative to female meioses (Sakamoto et al., 2000). The unusual recombination patterns in males result in expanded map distances in chromosome arms and reduced map distances in centromeric regions relative to females (Sakamoto et al., 2000). To effectively refine resolution of the QTLs, high levels of recombination are necessary. Sakamoto et al. (2000) found that recombination rates in centromeric regions were as much as ten times higher in females than in males. The use of female meioses will also enable gene-centromere mapping (Thorgaard et al., 1983), which estimates the distance between a genetic locus and the centromere. Therefore the use of female meioses should allow refinement of the QTL and increased resolution for fine mapping.

QTL studies have shown that the majority of quantitative traits tend to be controlled by a few QTLs with major effects with additional minor QTLs with lesser effects (Tanksley, 1993). Heritability of development rate has been estimated as high as 0.23 in rainbow trout (McIntyre and Blanc, 1973). Through QTL analysis a major locus for development rate has been identified in two different crosses (OSU X Swanson, OSU X Clearwater) accounting for as much as 54% of the phenotypic variation in the trait (Robison et al., 2001, Nichols, 2002). Based on isozyme studies in the OSU X Swanson cross (Robison et al., 2001), this QTL does not appear to be related to tissue differences in expression of the phosphoglucomutase (PGM) locus that have previously been shown to affect development in other studies in rainbow trout (Allendorf et al., 1983). However, this locus has not yet been mapped to further confirm non-linkage between PGM and the development rate locus.

Advanced backcross QTL analysis (AB-QTL) has been used for crop improvement through introgression of important alleles into different genetic backgrounds (Bernacchi et al., 1998, Pillen et al., 2003). The AB-QTL strategy along with marker-assisted selection will allow the development of congenic lines. Congenic lines have been used in many species for dissection of the molecular basis of quantitative traits (Zamir and Eshed, 1998, Hill, 1988, Silver, 1995, Fridman et al., 2000, Fridman et al., 2002). Development of an OSU congenic line carrying the major development rate QTL should remove other QTL loci influencing development and allow better localization and determination of the phenotypic effects for the major development rate QTL.

In this study, we address the prospects for fine mapping of the major development rate QTL through the development of a congenic line with the major development rate QTL (tth-1) through advanced backcrosses and marker-assisted selection. We explored the potential for fine-mapping of this development rate QTL using gynogenetic families derived from female meiosis, which might provide greater numbers of recombinants and thus greater map resolution in the region of this centromeric QTL to ultimately identify the genes underlying variations in development rate.