Journal of Heredity 2003:94(4)
© 2003 The American Genetic Association 94:273-284
Quantitative Trait Loci for Spawning Date and Body Weight in Rainbow Trout: Testing for Conserved Effects Across Ancestrally Duplicated Chromosomes
From the Department of Zoology, University of Guelph, Guelph, Ontario, Canada N1G 2W1. K. G. O'Malley is currently at the Coastal Oregon Marine Experiment Station, Hatfield Marine Science Center, 2030 SE Marine Science Dr., Newport, OR 97365, and T. Sakamoto is currently at the Department of Aquatic Biosciences, Tokyo University of Fisheries, Minato, Tokyo 108, Japan.
Address correspondence to M. M. Ferguson at the address above, or e-mail: mmfergus{at}uoguelph.ca.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
We incorporated 69 microsatellite loci into an existing data set of 132 markers to test for quantitative trait loci (QTLs) affecting spawning date and body weight in a backcross between two outbred strains of rainbow trout (Oncorhynchus mykiss). Twenty-six linkage groups were identified and synteny of duplicated microsatellite markers was used to confirm 13 homeologous chromosome pairs. Gene-centromere data were used to localize the centromeres for 13 linkage groups whose orientations were previously unknown. We applied a combination of interval mapping and single marker analysis to the segregating maternal and paternal alleles at 201 microsatellite loci. Four spawning date QTLs with suggestive evidence for an additional two QTLs were detected in female trout spawning at 3 and 4 years of age. Similarly we detected three QTLs for body weight in females at 2 years of age plus four suggestive QTLs for this trait. We found marginal evidence that three pairs of ancestral homeologues contained detectable QTLs for the same trait. In one of the three pairs of homeologues, the duplicated QTL regions mapped to the same relative chromosomal location, while the exact localization of the QTL position in one of the other pairs was difficult to infer since it was based on data from a male-derived map. The existing data were unable to refute a hypothesis that duplicated functional genes will be maintained within the telomeric regions of salmonids due to preferential male-mediated crossing over in this region. Two of the four spawning date QTLs were detected on linkage groups with unknown homeologous relationships. QTLs with possible pleiotropic effects on both spawning date and body size were localized to two linkage groups.
Gene duplication is recognized as an integral component of genome evolution by providing opportunities for the evolution of new gene functions (Ohno 1970). Of the many processes that generate gene duplications, polyploidization is the only one in which the entire genome is duplicated (Otto and Whitton 2000). Considerable evidence suggests that two genome-wide duplications occurred early in vertebrate evolution, the most recent approximately 250 million years ago (Meyer and Schart 1999). Studies with protein coding loci (Bailey et al. 1978) and Hox gene clusters (Malaga-Trillo and Meyer 2001) suggest that a third genome duplication event occurred in teleosts after their divergence from the lineages leading to tetrapods.
Polyploidy is the primary mechanism for generating genomic redundancy, as no other process can produce a comparable increase of genetic material on which selection may act. Wendel (2000) suggested that duplicated genes generated from a polyploidization event may diversify in function, experience a loss of expression at one of the two duplicated copies, or may retain the original function. Alternatively the interaction among duplicate genes may occur as genetic material is exchanged between homeologues, resulting in a loss of independence between pairs. As a consequence of such exchanges, alternative splicing of functional gene products may occur (Force et al. 1999), which would suggest that duplicated homeologues may retain similar gene functions through integrated translation processes. Duplicate gene expression is common, as rates of gene silencing are much lower than predicted by traditional models (Nadeau and Sankoff 1997). Genetic redundancy may offer a slight fitness advantage that might only be evident in certain life stages or environmental conditions (Cooke et al. 1997).
Salmonid fishes evolved from a single autotetraploid event approximately 25–100 million years ago (Allendorf and Thorgaard 1984). Immediately following genome duplication, an autotetraploid lineage is expected to demonstrate tetrasomic chromosome segregation. Over time, however, diploidization of the genome occurs and disomic segregation becomes prevalent (Allendorf and Danzmann 1997). Many homeologous chromosome arms still exchange chromatid segments as a result of multivalent formations in salmonid fishes (Allendorf and Thorgaard 1984; Wright et al. 1983). Of interest is that this meiotic event appears to be almost exclusive to males (Allendorf and Danzmann 1997). Differential crossovers between homeologous chromosomes result in pseudolinkage (the aberrant joint segregation of duplicated loci) and thus may regulate the duplication of some genes and diploidization of others (Sakamoto et al. 2000). These forms of residual tetrasomy suggest that diploidization may still be in progress.
Large sex-specific differences in recombination rates (female:male ratio = 3.25:1) have been observed in rainbow trout (Oncorhynchus mykiss) based on a genetic linkage map primarily constructed with microsatellite loci (Sakamoto et al. 2000). Multivalent formations likely constrain crossover events in males, resulting in the repressed rates. However, the observed differences may be conditional on the chromosomal location of the chiasmata. If chiasmata are localized to the telomeric regions (Allendorf and Danzmann 1997; Wright et al. 1983), then regions proximal to the centromere may experience no crossing over facilitating the diploidization of loci, while telomeric regions may experience an exchange of genetic material with homeologous regions. This would tend to inflate the recombination levels in the telomeric regions of males compared to females, which may in turn lead to the increased conservation of duplicated genes due to increased interhomeologue meiotic recombination (Allendorf and Danzmann 1997; Allendorf and Thorgaard 1984; Sakamoto et al. 2000; Wright et al. 1983).
Most traits related to fitness have continuous phenotypic distributions and are influenced by multiple chromosomal segments with quantitative trait loci (QTLs) (Mackay 2001). Testing the allelic effects on quantitative characters facilitates an understanding of the underlying genetics of these traits, primarily the number of polygenes and the magnitude of effect (Barton and Turelli 1989). Two alternate models attempt to elucidate the nature of quantitative variation. Fisher's infinitesimal model proposes that quantitative traits are controlled by a very large number of loci, each with a small phenotypic effect, while the oligogenic model describes continuous phenotypic variation as the result of a few loci with very large effects [reviewed by Tanksley (1993)]. Mapping studies have shown that at least some traits can be explained by the segregation of a few major QTLs, perhaps modified by QTLs of minor effect (Lin 2000; Mackay 2001). However, it is not always clear whether this outcome is a true reflection of the underlying genetics or a statistical artifact caused by sampling bias (Beavis 1998).
The recent construction of a genetic linkage map based on three backcross families in rainbow trout (Sakamoto et al. 2000) provides the opportunity to examine duplicate gene function (i.e., QTL expression) in a polyploid derivative species. Synteny of duplicated microsatellite markers was used to identify 10 homeologous chromosome pairs among the 29 linkage groups detected (Sakamoto et al. 2000). We would expect that 25–26 homeologous chromosome sets could exist in rainbow trout, given that most salmonids have 104 chromosome arms (Hartley 1987). The evolutionary fate of homeologous loci has been investigated in a very limited number of cases [e.g., allopolyploid cotton (Gossypium hirsutum)] (Cronn et al. 1999) and hexaploid derivative Brassica species (Axelsson et al. 2001). In Brassica sp., retention of duplicate gene function for flowering time QTLs has been observed, which suggests that the retention of homeologous gene function may be the norm rather than the exception within polyploid derivative lineages (Axelsson et al. 2001).
Fitness traits such as spawning date and body weight are major factors in the life history of salmonid fishes. Each genetic stock of rainbow trout spawns during a defined time of the year (Bromage et al. 1992). Narrow sense heritabilities for spawning date and body weight are relatively large in rainbow trout with the most variable estimates reported for body weight (Su et al. 1996). While the genetic architecture underlying these traits is largely unknown, an initial analysis using the same backcross family as the one used in this study has identified spawning date QTLs on seven linkage groups [A, C, G, I, J, K, and Oi of Sakamoto et al. (2000)] using single-marker analysis (Sakamoto et al. 1999). Robison et al. (2001) has also detected QTLs influencing embryonic length and weight in hybrids between clonal lines (androgenetically derived doubled haploid lines) on the linkage groups they designated as R6, R11, and R13. Linkage group R13 is homologous to linkage group 9 of Young et al. (1998) and Oi of Sakamoto et al. (2000), while the homeologies of R11 and R13 are unknown.
We searched for QTLs affecting spawning date and body weight in the rainbow trout backcross family used by Sakamoto et al. (1999) using segregating maternal and paternal alleles at 201 microsatellite loci spanning 26 linkage groups. We updated the genetic linkage map for this family in order to increase marker density for the first application of interval mapping to an outbred pedigree in salmonid fishes and to establish the homeologous relationships among linkage groups. The use of microsatellite loci in an interstrain backcross allowed us to map QTLs inherited from both parents, to determine if QTLs map to pairs of ancestral homeologues, and to determine if pairs of conserved QTLs are preferentially located in telomeric regions of homeologous chromosomes. This would be expected if the increased rates of telomeric recombination observed in male salmonids delay functional gene divergence (or silencing) (Allendorf and Thorgaard 1984). We also assessed the effects of individual linkage groups on both traits to study the covariation between spawning date and body weight.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Source of Fish
A backcross (lot 44) family containing 90 progeny made between two strains of rainbow trout that spawn in different seasons was the source material used for this study. Generally females designated as fall spawners ovulate between September and December, while spring spawners ovulate between February and April (Ferguson et al. 1993). In 1990 a male from a spring-spawning strain (government hatchery strain derived recently from a feral population) was crossed to a female from a fall-spawning strain (cultured strain) to produce a hybrid family. The male was induced to spawn in the fall through photoperiod manipulation. In 1992 an F1 hybrid (F x S) male was backcrossed to a fall-spawning female to produce the backcross family [details in Sakamoto et al. (1999)]. Neither the spring-spawning nor the fall-spawning strain was inbred, so the genotypes of the backcross parents were not expected to be completely homozygous. Likewise the F1 (F x S) parent was not completely heterozygous for the marker loci examined.
Phenotypic Data
All progeny were reared until they were 2 years old (1994). Each fish was then weighed (wet weight to the nearest gram) and marked with an individual brand. The average weight of the females was 776.1 g (SD 159.6 g; range 441.4–1147.6 g). The spawning dates (ovulation) for 45 female progeny from the backcross family were collected in 1995 and 1996 (3 and 4 years old postfertilization). Spawning date was calculated as the number of days from August 1. Females were examined weekly to identify those individuals that had ovulated between August and December in both years. Some females produced batches of eggs in successive weeks. For those females, the spawning date was designated according to the week when the larger volume of eggs was collected. A midweek spawning date was assigned when egg volumes were approximately equal. In 1995 the recorded spawning dates ranged from 16 to 128 days from August 1 (mean 74.3 days; SD 30.2), whereas in 1996 the spawning dates ranged from 30 to 114 days from August 1 (mean 65.7 days; SD 22.2). The body weight analysis was not repeated with the females when they were 3 and 4 years old because sexual maturation has major effects on growth rate. Analysis of male weight was not performed because many males mature at 2 years of age.
Microsatellite Analysis
Genomic DNA was extracted from muscle, liver, or gill tissue from the experimental fish using a standard phenol/chloroform method as described by Bardakci and Skibinski (1994). Microsatellite loci were amplified by polymerase chain reaction (PCR) as described by Sakamoto et al. (1999). The amplified DNA fragments were separated in a 6% polyacrylamide-7-M urea gel and visualized with a Hitachi FMBIOII fluorescence imaging system. Allele base pair size was determined using a 350 bp Tamra lane standard.
Genetic Map
The naming of microsatellite markers follows the standard proposed by Jackson et al. (1998). The label begins with a three-letter acronym usually specifying the species (e.g., Omy = O. mykiss) followed by a laboratory-specific term and a suffix acronym indicating the laboratory of origin (Table 1). Sex-specific linkage maps were generated because of the large differences in recombination rates between the sexes. A Visual Basic program, LINKMFEX (available at http://www.uoguelph.ca/
rdanzman/software/ ) (Danzmann and Gharbi 2001) was used to perform the linkage analysis through a series of pairwise comparisons between loci. The analysis was executed using segregation data for all 201 microsatellite markers in all 90 progeny (both sexes). To overcome the complicating factor of pseudolinkage in measuring classical linkage, the program assumes that the least abundant pairs of genotypes are the recombinants. A log of odds ratio (LOD) score 
3 was accepted as demonstrating linkage between markers (Botstein et al. 1980). The program MAPORD (available at http://www.uoguelph.ca/rdanzman/software/ ) was used to determine linear assignments of markers within a linkage group. Estimates of the differences between the sexes in recombination rates along chromosomal intervals were calculated using the program RECOMDIF (available at http://www.uoguelph.ca/rdanzman/software/ ). Map configurations were plotted with MAPCHART (Voorrips 2002) using the observed recombination distances between adjacent markers.
|
Gene centromere data were collected on markers localized to linkage groups from which no previous data on gene centromere orientations were available from the Sakamoto et al. (2000) study. Progeny from the same four females (i.e., lot 44 siblings) as those utilized by Sakamoto et al. (2000) were scored for their segregation products, and gene centromere distances were inferred as y/2, where y is the number of heterozygous recombinants scored in a given family (Thorgaard et al. 1983). At least two markers per linkage group were scored to orient the markers along a centromere-telomere axis.
QTL Analysis
QTL analysis was performed separately for segregating maternal and paternal alleles at each of the 26 linkage groups analyzed and for each of the three traits (spawning date at 3 and 4 years, body size at 2 years). Linear probability plots were used to confirm the normality of both spawning date and growth distributions in the backcross family prior to the analyses. Interval mapping with the program MultiQTL (version 2.1.2) (available at http://esti.haifa.ac.il/poptheor/) was used for linkage groups with more than one marker or for tests with single markers that were unlinked within the mapping parent (Korol et al. 2001). Linkage group assignments for unlinked markers within a mapping parent were inferred based on linkage information for the same marker in the opposite mapping parent or from information obtained from additional mapping families. Marker allele segregation data were first converted into a phase-corrected format (for each mapping parent) and then analyzed with MultiQTL. This was accomplished separately for each linkage group using the program GENOVECT (available at http://www.uoguelph.ca/rdanzman/software/) contained within the LINKMFEX package once a linear map order was established within that linkage group using MAPORD.
We began with a reduced model in MultiQTL that assumed similar trait variance (spawning date in 1995 or 1996; body size at 2 years) between those individuals receiving different marker alleles from either parent. A chromosome-wide permutation test (10,000 replications) was run in MultiQTL to determine the significance of the maximum LOD value over the various intervals analyzed for each chromosome. This permutation test is much more conservative compared to a single marker analysis [general linear model with permutation testing; Churchill and Doerge (1994)] on the same data set [details in O'Malley (2001)]. A P value of.01 in the MultiQTL analysis was equivalent to a P of.001 generated from the single-marker analysis [10,000 permutations in PROC MULTITEST; SAS Institute (1996)]. A P value of.001 in the single-marker analysis would pass a Bonferonni corrected threshold for an alpha level of 0.05 given the number of linkage groups analyzed (0.05/26 = 0.002). Therefore we considered a P value equivalent to.01 or less in the interval analysis permutation test as indicative of a significant QTL effect, values between.01 and.05 as suggestive, and values between.1 and.05 as marginal. A second model (general model) that assumed unequal variances was then compared to the reduced model and accepted if the LOD values generated by each were significantly different (10,000 permutations). Single marker tests were also run with the marker option in MultiQTL. This latter option was necessary in certain instances for two reasons. First, the current linkage map of rainbow trout is incomplete. Markers assigned to a male linkage group may be unassigned to the equivalent female linkage group (at a LOD threshold of 3.0) due to the large differences in recombination rate detected between the sexes (Sakamoto et al. 2000). Markers unassigned to a linkage group on the female map were tested independently. Also, multiple markers in a male linkage group may assign to a single location due to a lack of recombination between them. Small male linkage groups (e.g., linkage group U) were tested as single-marker positions. An arbitrary criterion of 25% variation was used to define a major QTL (Bradshaw et al. 1998).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Genetic Map
The incorporation of 69 microsatellite markers into the data set of Sakamoto et al. (2000) resulted in the construction of 26 linkage groups in the lot 44 family; 30 linkage groups have been detected using combined data from additional families (unpublished data). Fifteen markers remain unassigned at a LOD threshold of 3.0 in both sexes. With the combined data from one additional family (lot 25) (Sakamoto et al. 2000 and unpublished data) we were able to infer the homeology of 13 pairs of linkage groups (Table 2). In combination with the data from lot 25, we were able to map centromeres to linkage groups A, B, C, D, E, Fi, Fii, G, H, I, J, K, L, M, N, Oi, Oii, P, Q, R, S, T, U, 2, 5, 8, 15, and 18 with new data (Table 3) not described in Sakamoto et al. (2000).
|
|
The distribution of duplicated loci varied among linkage groups. For instance, duplicated loci appeared to be clustered in the intercalary region of several linkage groups [K, Oi, and N (i.e., OmyFGT28TUF), Fi, 15, and 5] but mapped primarily to the distal region in others [Oi (OmyFGT32TUF; OmyRGT42TUF), L, and R]. There were several female linkage groups where the relative position of duplicated loci could not be determined because of a lack of gene-centromere information or a lack of recombination information from the female map (Oii, U, 8, Q) (Figure 1).
|
The sexes differed significantly in the recombination rate between pairs of adjacent markers (analysis not shown). Female map distances were generally larger than those in the male in regions of the chromosome proximal to the centromere (Figure 1). However, male recombination rates were significantly higher in the telomeric region of several linkage groups (>30 cM).
QTLs
Four significant and two suggestive spawning date QTLs were detected on the 26 linkage groups analyzed (Oii, A, I, J, U, and G) (Figure 1 and Table 4). The alleles on linkage group A inherited from the sire showed associations with spawning date in both 3- and 4-year-old females. The observation of a marginal effect of dam alleles in year 4 adds further support to the location of a spawning date QTL on linkage group A. The existence of a spawning date QTL on linkage group I is supported by significant effects of sire alleles in year 4, suggestive effects of dam alleles on spawning date at 4 years, as well as marginal effects (both sire and dam alleles) for spawning date at 3 years. Linkage group J provided the strongest evidence for a spawning date QTL where significant effects were detected for both parents in both years (variation ranged from 16.5% to 64.2%). Information on the primers used that are localized to linkage groups with the QTL regions shown in Figure 1 can be obtained as an appendix to the manuscript at the following Web site: http://www.uoguelph.ca/rdanzman/appendices/.
|
There was marginal evidence that 2 of 13 pairs of ancestral homeologues contained spawning date QTLs (Oi and Oii; A and K) (Figure 1 and Table 4). In both cases the QTL effect on one homeologue was significant (Oii, A), while the effect was marginal for the other member of the pair (Oi, K). The amount of variance attributed to homeologous QTLs was similar across linkage groups A and K (±3%). The parental source (sire or dam) contributing to the allelic effect was the same for both homeologous pairs (Oi and Oii; A and K). The spawning date QTLs with the greatest effects were identified on linkage groups for which the homeologous relationships are unknown ( J and I ) (Figure 1). It should be noted that we could only directly compare 8 of 13 homeologous regions identified by the duplicated microsatellite markers since the other 5 duplicated microsatellite regions were not variable in either of the lot 44 mapping parents (see Table 2).
Significant QTLs for body weight were detected on three linkage groups (G, C, and 15) (Figure 1 and Table 5). The effects on G were detected between sire alleles, while those on C and 15 were detected between dam alleles. The QTLs on linkage group C accounted for 24.8% of the variation in body weight. Suggestive QTLs were detected on linkage groups H, P, N, and S, and marginal effects on linkage groups A and B. There was marginal evidence that body weight QTLs mapped to one homeologous group (linkage groups 5 and 15) in the dam. A significant effect was detected on linkage group 15, while the effect on the homeologue (linkage group 5) was marginal. The homeologies of linkage groups N and P are largely unknown (although a proximal segment to the centromere containing the duplicated locus OmyFGT28TUF appears homeologous to linkage group Oi).
|
Covariation Among Traits
There was no association between spawning date and body weight ( year 3 R 2 = 0.01; year 4 R 2 = 0.03). Spawning date QTLs and body weight QTLs were identified on the same linkage group in four instances (A, G, C, and P) if suggestive and marginal QTLs are considered in addition to significant associations (Tables 3 and 4). In most cases the allelic effects for one trait were not from the same parent as for the other trait. For example, progeny inheriting different G chromosomes from the sire had significantly different body weights but not spawning dates (the reverse was true for maternal chromosomes for the same linkage group). In two cases (linkage groups A and P), variations in chromosomes inherited from one parent or both parents had a marginal influence on one trait and a significant or suggestive influence on the other trait. For example, variation at Ogo1UW inherited from the sire had a pronounced influence on spawning date and a marginal influence on growth. Variation at Ots515NWFSC inherited from the dam had a suggestive influence on body weight, while alleles inherited from the sire had a marginal influence on spawning date.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Two of the four significant QTLs for spawning date identified in rainbow trout could be classified as having major effects. Similarly, one of the three significant QTLs for body weight identified here could be considered a major QTL. For the remaining significant QTL, 15–18% of the variation in spawning date and 12–19.1% of the variation in body size could be explained by the segregation of the parental alleles within the respective QTL regions. The average difference in spawning date between segregating alleles at the two major spawning date QTLs was 35.1 days, whereas the difference in spawning date between segregating alleles at a minor spawning date QTL was an average of 24.1 days. Likewise, the difference between segregating alleles at the major body weight locus was approximately 148.1 g (19% of average body weight of all females), whereas the average difference in weight between segregating alleles at the two minor body weight loci was 124.5 g. Our data concur with findings from recent QTL studies (Bradshaw et al. 1998; Hurme et al. 2000; Lin and Ritland 1997) indicating that a large proportion of quantitative variation can be explained by the segregation of a few major QTLs (reviewed by Tanksley 1993). These findings are counter to Fisher's infinitesimal model in which quantitative traits are controlled by a very large number of loci, each with a small phenotypic effect. However, the manifestation of such large phenotypic differences between segregating QTL alleles may be contingent on the genetic background of the test families (Whitlock et al. 1995). Larger differences are typically observed in hybrid and F1 backcross families such as those used here compared to intrastrain families.
A statistical bias toward the detection of genes of larger phenotypic effects may result, however, in an underestimation of the total number of genes affecting a trait. Power to detect QTLs depends on the average allelic substitution effect of the alleles involved, the recombination distance between the QTL and associated markers, and the sample size of the progeny used to detect QTLs (Falconer and Mackay 1996). In our case, the limited number of female progeny (N = 45) will lead to a tendency to detect QTL regions of larger phenotypic effect. Also, distinguishing between single-gene versus multigene composition of individual QTLs is difficult when QTL mapping resolution is limited to 10–20 cM (reviewed by Tanksley 1993). Additional QTL analysis will further elucidate the true number and magnitude of genes affecting these two fitness-related traits in rainbow trout.
The use of interval mapping has confirmed the existence of the two QTLs with strong effects (linkage groups I and J) for spawning date in rainbow trout first reported by Sakamoto et al. (1999). The QTL on linkage group I is located in the interval surrounding OmyFGT34TUF and Ssa103NVH. In linkage group J, the QTL detected in the female appears to be near the centromere and spans the region from One5ASC to OmyFGT12TUF. This region would appear to include One112ADFG (strongest QTL effect from the male parent), although the exact synteny of the QTLs in the male cannot be established because of reduced recombination levels in males. The report of multiple QTLs for spawning date on linkage group G (Sakamoto et al. 1999) is supported with the current analysis in that the interval spanning OmyRGT36TUF and One2ASC may contain one or more QTLs. Support for spawning date QTLs on linkage groups C and K [note that marker OmyFGT22TUF in linkage group K of Sakamoto et al. (1999) equals OmyFGT21/iiTUF in Sakamoto et al. (2000)], and Oi [called M in Sakamoto et al. (1999)] is only marginal, while the presence of a strong QTL effect on linkage group A is supported. In addition, we provide evidence for a suggestive spawning date QTL residing on linkage group U and a strong QTL effect on linkage group Oii. Significant interaction between segregating female and male alleles at OmyFGT32/iTUF (O'Malley 2001) on linkage group Oi suggests the presence of a weaker QTL effect in that region at the same relative chromosomal location as detected on Oii.
It should be noted that the marginal QTL effect for spawning date detected with One14ASC on linkage group Oi in the female parent does not correspond to the same QTL region marked by OmyFGT32/iTUF. Marker One14ASC is unlinked to OmyFGT32/iTUF on the female map. Thus one possibility is that at least two different QTLs influence spawning date within linkage group Oi. The region marked by OmyFGT32/iTUF and not One14ASC is homologous to the one containing a QTL effect on Oii. A second possibility is that One14ASC may not be part of linkage group Oi and any associations with Oi on the male map may be because of a pseudolinkage. The true location of the marginal QTL marked by One14ASC will become evident with additional mapping.
Spawning date QTLs were conserved across two pairs of homeologues (A/K and Oi/Oii) and body weight QTLs were conserved across a single homeologous pair (5/15) (although, see the discussion below with respect to body weight QTLs) out of 13 putative homeologous pairings in rainbow trout (A/K, C/L, D/E, Fi/Fii, G/Q, H/U, N/Oi, Oi/Oii, 8/R, 5/15, 2/Oii, D/T, T/Oi). Eight of these homeologous regions could be directly tested because duplicated microsatellite markers within these regions were polymorphic in the mapping parents used in this study. Caution must be exercised in the interpretation of the data thus far, as many of these assignments are based on information from a single pair of duplicated markers. For example, the duplicated marker Omy296UoG maps to D and T and the duplicated type I gene TRCARR(INRA) (trout red cell arrestin gene) maps to linkage groups Oi and T. These regions may only be representative of small translocation events and not reflect larger-scale chromosomal homeologies. Multiple homeologous assignments (i.e., Oi/Oii and 2/Oii) reflect that one arm of a metacentric chromosome such as Oii may be homeologous to two different linkage groups following Robertsonian translocations.
The duplication and mapping of other functional genes in the rainbow trout genome allows us to make inferences on additional homeologies among chromosomes. The localization of duplicate copies of the Wilm's tumor (WT-t1) gene to linkage groups 6 and 27 of Young et al. (1998) (Brunelli et al. 2001), which are homologous to linkage groups S and 15 of Sakamoto et al. (2000), respectively (unpublished data), also suggests the conservation of growth QTLs across a possible homeologous pair (S/15). Duplicates of the WT-t1 gene map syntenically with zero recombination to OmyFGT20TUF on S and OmyRGT31TUF on l5. Also, the recent report of a significant growth QTL effect near OmyRGT1TUF in a hatchery strain of rainbow trout (Martyniuk 2001) strongly supports the marginal associations we detected at this position on linkage group 5.
The distribution of duplicated loci along chromosome arms was used to identify corresponding homeologous chromosomal segments and make inferences on relative QTL locations. The inferred location of the spawning date QTL may be similar across two homeologous pairs (A/K and Oi/Oii). However, the data for A/K is based on the male map and thus any inferences in relative location must be considered tentative. The QTL markers for linkage group K (OmyRGT7TUF) and linkage group A (Ogo1UW) both localize to a large central cluster of markers in the male map. Most markers in central clusters of male salmonid maps tend be centromeric or intercalary in location when verified with the order of markers in females (Sakamoto et al. 2000). This observation plus the close proximity of Ogo2/iiUW with the cluster on linkage group K and Ogo2/iUW to Ogo1UW on linkage group A (see Sakamoto et al. 2000) suggest that the pair of QTLs on linkage groups A and K are intercalary (OmyRGT7TUF on K has a gene-centromere distance of 22 cM; Table 3). However, this conclusion might be counter to the observation of a marginal QTL effect at Ssa38NVH on the telomeric region of linkage group A. If the region marked by Ssa38NVH is homologous to the spawning date QTL region detected in the male parent discussed above, then the QTL is telomeric rather than intercalary.
Both the QTL linked to One14ASC (linkage group Oi) and OmyRGT42/iiTUF (linkage group Oii) appear to be in telomeric regions (gene-centromere distance for One14ASC = 48.9) (see Table 3). However, as mentioned above, the spawning date QTL marked by One14ASC is not on the same chromosome arm as that marked by OmyRGT42/iTUF. Thus the two spawning date QTLs may share the same relative location to the centromere, but may be on opposite telomeric regions of linkage group Oi.
The body size QTL on the 5/15 homeologues do not appear to map to similar regions. The QTL marked by OmyRGT31TUF localizes to a small chromosomal segment on linkage group 15. However, that segment is unlinked to the segment of 15 (marked by Omy272/iUoG) that is homologous to the QTL-containing region on linkage group 5 (marked by RGT1TUF); RGT1TUF shows zero recombination with Omy272/iiUoG in the female map (see Figure 1 and Sakamoto et al. 2000). This suggests that two different QTL regions for body size may be present on the 5/15 homeologues in rainbow trout. We were unable to test for the presence of two QTLs on linkage group 15 for the female parent of lot 44 using MultiQTL, since the fragment containing OmyRGT31TUF is unlinked in the female (but linked in the male).
Although we have provided some evidence for the possible conservation of QTL effects across homeologues in rainbow trout, it remains difficult to estimate the propensity for such conservation at the genome level overall. Those cases where QTL effects were detected in only one of the two homeologues might be due to the inadequate density of markers on those linkage groups leading to limited statistical power. Furthermore, only one family was examined in this study, resulting in a lack of tested variation in the species. In addition, tests were not possible for those linkage groups where homeologous relationships are unknown. Ultimately a high-density map will reveal additional homeologous relationships and provide a robust estimate of QTL conservation.
Consideration of meiotic configurations in rainbow trout suggest that QTLs in telomeric regions of homeologues should have a greater probability of being conserved than those located close to the centromere. In female rainbow trout, duplicated loci exhibit random assortment, whereas in males, an aberrant pattern of nonrandom segregation results in an excess of nonparental progeny types relative to the parental types. Wright et al. (1983) developed a meiotic segregation model to explain this apparent pseudolinkage. Essentially, two nonhomeologous acrocentric chromosomes pair to form a metacentric chromosome. Homeologous acrocentric pairs of chromosomes may then randomly pair with their homeologous arms in the metacentrics to form multivalents. Multivalent formation in male salmonids likely constrains crossovers proximal to the centromere, thus facilitating the diploidization of loci located in this region (Allendorf and Thorgaard 1984). If crossovers between homeologous chromosome segments in the telomeric region preserve duplicated gene regions, then homeologous QTLs located in the telomeric region should be conserved relative to QTLs proximal to the centromere. Only one duplicated type I gene marker (TRCARR) has been placed on the microsatellite map (linkage groups Oi and T) of Sakamoto et al. (2000). We suggest that TRCARR/ii(INRA) is located telomerically in linkage group T because it maps distal to Omy8DIAS and SsaLEEN82. Since Omy8DIAS is proximal to the centromere (Table 3), it would appear that this single type I duplicated gene marker also maps telomerically in the rainbow trout genome.
There is suggestive support for the hypothesis that functionally duplicated QTLs may be preferentially located telomerically. The telomeric location of one of the proposed spawning date QTLs on Oi (based on the significant male/female allelic interaction effects) and the homeologous region on Oii supports the retention of duplicate gene expression on the telomeres. Similarly, the pair of spawning date QTLs on linkage groups A and K may be located telomerically, although this is difficult to ascertain given that these effects were detected in males. The localization of the two putative body size QTLs on linkage groups 5 and 15 suggests that one region (OmyRGT31TUF on 15) is telomeric and the other is intercalary (OmyRGT1TUF/Ssa125/iiNVH interval on 5). However, this pair is not an appropriate test of the hypothesis since the two regions are not homologous in lot 44. The region on 5 and 15 may have undergone an inversion event (see Sakamoto et al. 2000), which will confound the ability to localize the homeologous segments.
Quantitative variation is based on the overlapping effects of many genes on many characters, and selection on one character will strongly influence that on any other (Charlesworth 1994). We have mapped QTLs for two different traits to the same four linkage groups (A, C, G, and P). In one of these linkage groups (P) the QTLs for each trait were mapped in the opposite sex but to the same marker location (Ots515NWFSC), suggesting a possible pleiotropic function for the QTLs on these traits. Similarly, the localization of a QTL effect for spawning date and body weight to Ogo1UW in the male parent suggests a pleiotropic role for the QTL on linkage group A. However, on linkage group G, the QTL effect on spawning date detected in the female parent appears to be on the chromosome arm opposite that for the growth QTL detected in the sire. Similarly, on linkage group C, the marginal evidence for a spawning date QTL appears localized close to the centromere (based on female gene-centromere data; see Sakamoto et al. 2000), while the growth QTL region appears telomeric.
In conclusion, we have found evidence that QTLs affecting spawning date but not body weight may have been conserved across homeologous chromosome pairs in rainbow trout. However, the propensity for such conservation at the genome level remains difficult to measure. Furthermore, the number and magnitude of QTLs detected suggest that both spawning date and body weight may be controlled by a few loci with large effects. Additional research will elucidate the nature of the covariation between these two traits to determine if antagonistic pleiotropy is an important factor maintaining polymorphism.
|
|
Acknowledgments |
|---|
We thank Chris Martyniuk for technical assistance. This work was supported by the Natural Sciences and Engineering Research Council of Canada (Research Grants Program and Genomics Projects).
|
Footnotes |
|---|
Corresponding Editor: Lisa Seeb
Received March 28, 2002
Accepted April 23, 2003
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
-
Allendorf FW, Danzmann RG, 1997. Secondary tetrasomic segregation of MDH-B and preferential pairing of homeologues in rainbow trout. Genetics. 145:1083-1092.[Abstract]
Allendorf FW, Thorgaard GH, 1984. Tetraploidy and the evolution of salmonid fishes. In: Evolutionary genetics of fishes (Turner B, ed). New York: Plenum; 1–53.
Axelsson T, Shavorskaya O, Lagercrantz U, 2001. Multiple flowering time QTLs within several Brassica species could be the result of duplicated copies of one ancestral gene. Genome. 44:856-864.[Medline]
Bailey GS, Poulter RT, Stockwell PA, 1978. Gene duplication in tetraploid fish: model for gene silencing at unlinked duplicated loci. Proc Natl Acad Sci USA. 75:5575-5579.
Bardakci F, Skibinski DOF, 1994. Application of the RAPD technique in tilapia fish: species and subspecies identification. Heredity. 73:117-123.
Barton NH, Turelli M, 1989. Evolutionary quantitative genetics: how little do we know? Annu Rev Genet. 23:337-370.[Web of Science][Medline]
Beavis WD, 1998. QTL analyses: power, precision, and accuracy. In: Molecular dissection of complex traits (Paterson AH, ed). Boca Raton, FL: CRC Press; 145–162.
Botstein D, White RL, Skolnick M, Davis RW, 1980. Construction of a genetic map in man using restriction fragment length polymorphisms. Am J Hum Genet. 32:314-331.[Web of Science][Medline]
Bradshaw H, Otto K, Frewen B, McKay J, Schemske DW, 1998. Quantitative trait loci affecting differences in floral morphology between two species of monkeyflower (Mimulus). Genetics. 149:367-382.
Bromage N, Jones J, Randell C, Thrush M, Davies J, Springate J, Duston J, Barker G, 1992. Broodstock management, fecundity, egg quality and timing of egg production in the rainbow trout (Oncorhynchus mykiss). Aquaculture. 100:141-166.[CrossRef]
Brunelli JP, Robison BD, Thorgaard GH, 2001. Ancient and recent duplications of the rainbow trout Wilms' tumor gene. Genome. 44:455-462.[Medline]
Charlesworth B, 1994. Evolution in age-structured populations. Cambridge: Cambridge University Press.
Churchill GA, Doerge RW, 1994. Empirical threshold values for quantitative trait mapping. Genetics. 138:963-971.[Abstract]
Cooke JMA, Nowak MA, Boerlijst A, Maynard-Smith J, 1997. Evolutionary origins and maintenance of redundant genes expression during metazoan development. Trends Genet. 13:360-364.[CrossRef][Web of Science][Medline]
Cronn RC, Small RL, Wende, JF, 1999. Duplicated genes evolve independently after polyploidy formation in cotton. Proc Natl Acad Sci USA. 96:14406-14411.
Danzmann RG, Gharbi K, 2001. Gene mapping in fishes: a means to an end. Genetica. 111:3-23.[CrossRef][Web of Science][Medline]
Falconer DS, MacKay TFC, 1996. Introduction to quantitative genetics. Harlow, UK: Longman.
Ferguson MM, Danzmann RG, Arndt SKA, 1993. Mitochondrial DNA and allozyme variation in Ontario culture rainbow trout spawning in different seasons. Aquaculture. 117:237-259.[CrossRef]
Force A, Lynch M, Pickett BF, Amores A, Yan Y, Postlethwait J, 1999. Preservation of duplicate genes by complementary, degenerative mutations. Genetics. 151:1531-1545.
Hartley SE, 1987. Chromosomes of salmonid fishes. Biol Rev Camb Philos Soc. 62:197-214.
Hurme P, Sillanpaa MJ, Arjas E, Repo T, Savolainen O, 2000. Genetic basis of climatic adaptation in Scots pine by Bayesian quantitative trait locus analysis. Genetics. 156:1309-1322.
Jackson TR, Ferguson MM, Danzmann RG, Fishback AG, Ihssen PE, O'Connell M, Crease TJ, 1998. Identification of two QTL influencing upper temperature tolerance in three rainbow trout (Oncorhynchus mykiss) half-sib families. Heredity. 80:143-151.[CrossRef][Web of Science]
Korol A, Ronin Y, Itzcovich A, Peng J, Nevo E, 2001. Enhanced efficiency of QTL mapping analysis based on multivariate complexes of quantitative traits. Genetics. 157:1789-1803.
Lin J-Z, 2000. The relationship between loci for mating system and fitness-related traits in Mimulus (Scrophulariaceae): a test for deleterious pleiotropy of QTL with large effects. Genome. 43:628-633.[Medline]
Lin J-Z, Ritland K, 1997. Quantitative trait loci differentiating the outbreeding Mimulus guttatus from the inbreeding Mimulus platycalyx. Genetics. 146:1115-1121.[Abstract]
Mackay TFC, 2001. The genetic architecture of quantitative traits. Annu Rev Genet. 35:303-339.[CrossRef][Web of Science][Medline]
Malaga-Trillo E, Meyer A, 2001. Genome duplications and accelerated evolution of Hox genes and cluster architecture in teleost fishes. Am Zool. 41:676-686.[CrossRef]
Martyniuk CJ, 2001. A microsatellite DNA analysis examining the relationship between growth and maturation rate genes in rainbow trout (master's thesis). Guelph, Ontario: University of Guelph.
May B, Johnson KR, 1990. Composite linkage map of salmonid fishes (Salvelinus, Salmo, Oncorhynchus). In: Genetic maps: locus maps of complex genomes (O'Brien SJ, ed). Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 151–159.
Meyer A, Schart M, 1999. Gene and genome duplications in vertebrates: the one-to-four (-to-eight in fish) rule and the evolution of novel gene functions. Curr Opin Cell Biol. 11:699-704.[CrossRef][Web of Science][Medline]
Nadeau JH, Sankoff D, 1997. Comparable rates of gene loss and functional divergence after genome duplications early in vertebrate evolution. Genetics. 147:1259-1266.[Abstract]
Ohno S, 1970. The enormous diversity in genome sizes of fish as a reflection of nature's extensive experiments with gene duplication. Trans Am Fish Soc. 99:120-130.
Olsen JB, Wilson SL, Kretschmer EJ, Jones KC, Seeb JE, 2000. Characterization of 14 tetranucleotide microsatellite loci derived from sockeye salmon. Mol Ecol. 9:2155-2234.[CrossRef][Medline]
O'Malley KG, 2001. Quantitative trait loci for spawning time and body weight in rainbow trout: testing for conserved effects across ancestral homeologues (master's thesis). Guelph, Ontario: University of Guelph.
Otto SP, Whitton J, 2000. Polyploid incidence and evolution. Annu Rev Genet. 34:401-437.[CrossRef][Web of Science][Medline]
Robison BD, Wheeler PA, Sundin K, Sikka P, Thorgaard GH, 2001. Composite interval mapping reveals a major locus influencing embryonic development rate in rainbow trout (Oncorhynchus mykiss). J Hered. 92:6-22.
Sakamoto T, Danzmann RG, Gharbi K, Howard P, Ozaki A, Sokkean K, Woram RA, Okamoto N, Ferguson MM, Holm L-E, et al., 2000. A microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) characterized by large sex-specific differences in recombination rates. Genetics. 155:1331-1345.
Sakamoto T, Danzmann RG, Okamoto N, Ferguson MM, Ihssen PE, 1999. Linkage analysis of quantitative trait loci associated with spawning time in rainbow trout (Oncorhynchus mykiss). Aquaculture. 173:33-43.[CrossRef][Web of Science]
SAS Institute,, 1996. SAS version 6.12. Cary, NC: SAS Institute.
Su G-S, Liljedahl L-E, Gall GAE, 1996. Genetic and environmental variation in body weight in rainbow trout. Aquaculture. 143:71-80.
Tanksley SD, 1993. Mapping polygenes. Annu Rev Genet. 27:205-233.[CrossRef][Web of Science][Medline]
Thorgaard GH, Allendorf FW, Knudsen KL, 1983. Gene-centromere mapping in rainbow trout: high interference over long map distances. Genetics. 103:771-783.
Voorrips RE, 2002. MapChart: software for the graphical presentation of linkage maps and QTLs. J Hered. 93:77-78.
Wendel JF, 2000. Genome evolution in polyploids. Plant Mol Biol. 42:225-249.[CrossRef][Web of Science][Medline]
Whitlock MC, Phillips PC, Moore FBG, Tonsor SJ, 1995. Multiple fitness peaks and epistasis. Annu Rev Ecol Syst. 26:601-629.[CrossRef][Web of Science]
Wright JE, Johnson K, Hollister A, May B, 1983. Meiotic models to explain classical linkage, pseudolinkage, and chromosome pairing in tetraploid derivative salmonid genomes. Isozymes Curr Top Biol Med Res. 10:239-260.[Web of Science][Medline]
Young WP, Wheeler PA, Coryell VH, Keim P, Thorgaard GH, 1998. A detailed linkage map of rainbow trout produced using doubled haploids. Genetics. 148:839-850.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Zhang and N. A. Rosenberg On the Genealogy of a Duplicated Microsatellite Genetics, December 1, 2007; 177(4): 2109 - 2122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Nikinmaa and W. Waser Molecular and cellular studies in evolutionary physiology of natural vertebrate populations: influences of individual variation and genetic components on sampling and measurements J. Exp. Biol., June 1, 2007; 210(11): 1847 - 1857. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Rogers, N. Isabel, and L. Bernatchez Linkage Maps of the dwarf and Normal Lake Whitefish (Coregonus clupeaformis) Species Complex and Their Hybrids Reveal the Genetic Architecture of Population Divergence Genetics, January 1, 2007; 175(1): 375 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. B. Phillips, K. M. Nichols, J. J. DeKoning, M. R. Morasch, K. A. Keatley, C. Rexroad III, S. A. Gahr, R. G. Danzmann, R. E. Drew, and G. H. Thorgaard Assignment of Rainbow Trout Linkage Groups to Specific Chromosomes Genetics, November 1, 2006; 174(3): 1661 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Gharbi, A. Gautier, R. G. Danzmann, S. Gharbi, T. Sakamoto, B. Hoyheim, J. B. Taggart, M. Cairney, R. Powell, F. Krieg, et al. A Linkage Map for Brown Trout (Salmo trutta): Chromosome Homeologies and Comparative Genome Organization With Other Salmonid Fish Genetics, April 1, 2006; 172(4): 2405 - 2419. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. H. Leder, R. G. Danzmann, and M. M. Ferguson The Candidate Gene, Clock, Localizes to a Strong Spawning Time Quantitative Trait Locus Region in Rainbow Trout J. Hered., January 1, 2006; 97(1): 74 - 80. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. L. Perry, M. M. Ferguson, T. Sakamoto, and R. G. Danzmann Sex-linked Quantitative Trait Loci for Thermotolerance and Length in the Rainbow Trout J. Hered., March 1, 2005; 96(2): 97 - 107. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. M. L. Somorjai, R. G. Danzmann, and M. M. Ferguson Distribution of Temperature Tolerance Quantitative Trait Loci in Arctic Charr (Salvelinus alpinus) and Inferred Homologies in Rainbow Trout (Oncorhynchus mykiss) Genetics, November 1, 2003; 165(3): 1443 - 1456. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||