Centromere-linkage in the turbot (Scophthalmus maximus) through half-tetrad analysis in diploid meiogynogenetics

Aquaculture 280 (2008) 81–88

Contents lists available at ScienceDirect

Aquaculture j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a q u a - o n l i n e

Centromere-linkage in the turbot (Scophthalmus maximus) through half-tetrad analysis in diploid meiogynogenetics Paulino Martínez a,⁎, Miguel Hermida a, Belén G. Pardo a, Carlos Fernández a, Jaime Castro a, Rosa M. Cal b, José A. Álvarez-Dios c, Antonio Gómez-Tato d, Carmen Bouza a a

Departamento de Genética, Universidad de Santiago de Compostela (USC), Facultad de Veterinaria, Campus de Lugo, 27002 Lugo, Spain Instituto Español de Oceanografía (IEO), Centro Oceanográfico de Vigo, 36280 Vigo, Spain Departamento de Matemática Aplicada, (USC), Facultad de Matemáticas, 15782 Santiago de Compostela, Spain d Departamento de Geometría (USC), Facultad de Matemáticas, 15782 Santiago de Compostela, Spain b c

A R T I C L E

I N F O

Article history: Received 28 January 2008 Received in revised form 2 May 2008 Accepted 9 May 2008 Keywords: Turbot Scophthalmus maximus Genetic map Centromere position Segregation distortion Deleterious genes Mapping function

A B S T R A C T Seventy nine microsatellite markers selected across all linkage groups (LG) from a previous turbot genetic map were studied in a diploid meiogynogenetic family for centromere mapping using half-tetrad analysis. Significant deviations from Mendelian segregation were observed at 25% loci analyzed. The clustering of distorted loci at specific LGs, suggested the existence of genes of different deleterious effects. The lack of Mendelian segregation distortion at 1 day and 10 days post-hatching larvae at these loci precluded an explanation based on aberrant meiotic segregation. Heterozygote frequency distribution in gynogenetic offspring showed close to 50% values above 0.667, which suggested high chiasma interference in turbot. Complete interference appeared as the best fitting function when estimating centromere position. However, Kosambi and Haldane functions performed better at specific LGs as a consequence of the variable crossover pattern of centromere-distant markers among LG. Great concordance between half-tetrad data and the positions previously reported in the turbot map was observed. Most centromeres were localized with an error around or below 5 cM and closely linked markers exist now in 8 LGs. Centromere location was mostly in accordance with previous karyotypic information. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Genetic maps constitute essential organizational tools for genomic research (Sewell et al., 1999). They are being applied to identify QTL or genomic regions related with evolutionary or productive characters, and eventually for positional cloning or candidate gene strategies (Donovan et al., 2000; Mackay, 2001; Blott et al., 2003; Colosimo et al., 2005). Positioning centromeres constitutes an important goal during map construction (Danzmann and Gharbi, 2001; Nichols et al., 2003; Guyomard et al., 2006). The heterogeneity of recombination frequency along chromosome arms disturbs the correspondence between genetic and physical maps (Kauffman et al., 1995). The lower recombination frequency in the vicinity of centromeres influences positional cloning or marker assisted selection strategies, since markers are embracing larger genomic regions at these areas. Markers close to centromeres are particularly useful to detect linkage of mutations to specific linkage groups (LG), especially in fish where high chiasma interference limits the number of multiple crossovers (Johnson et al., 1995; Kauffman et al., 1995; Mohideen et al., 2000). ⁎ Corresponding author. Tel./fax: +34 982254681. E-mail address: [email protected] (P. Martínez). 0044-8486/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2008.05.011

Half-tetrad analysis using diploid meiogynogenetics has been generally used to locate centromeres in fish (Thorgaard et al., 1983; Allendorf et al.,1986; Kauffman et al.,1995; Sakamoto et al., 2000; Guyomard et al., 2006; Nomura et al., 2006). Crossovers along chromosome arms during meiosis determine that genetic markers close to centromeres segregate mostly during first meiotic division, while distal ones during meiosis II (Johnson et al., 1996). So, heterozygotes in gynogenetic offspring can be used to obtain genetic distance between markers and centromeres (Danzmann and Gharbi, 2001). Different mapping functions can be applied to estimate distances between genetic markers and centromeres (Kauffman et al., 1995). Complete interference, which assumes that one recombinational exchange inhibits the formation of additional crossovers, has been generally applied in fish (Sakamoto et al., 2000; Morishima et al., 2001; O'Malley et al., 2003; Matsuoka et al., 2004; Guyomard et al., 2006). The turbot (Scophthalmus maximus; Scophthalmidae; Pleuronectiformes) is one of the most promising aquaculture species in Europe. Genetic information has increased in the last 15 years in response to the demand of turbot industry for evaluating genetic resources and for parentage analysis (Bouza et al., 2002; Castro et al., 2003, 2004). Recently, a first genetic map of 242 microsatellites distributed across 26 LGs has been published in this species (Bouza et al., 2007).

82

P. Martínez et al. / Aquaculture 280 (2008) 81–88

In this study, we have analyzed a large sample of microsatellites distributed across all LGs for locating centromeres in the reported turbot map (Bouza et al., 2007) using diploid meiogynogenetic offspring. We were specifically concerned with: i) contrasting previous map information with gene centromere distances and searching for congruence between both data sets; ii) analyzing crossover pattern across linkage groups; iii) locating centromeres at linkage groups as a further step towards consolidating turbot map. 2. Materials and methods

Table 1 Diploid meiogynogenetic segregation for the 79 microsatellite loci used to locate centromeres in turbot genetic map Linkage group

Locus

LG1

Sma-USC271 Sma-USC13 Sma-USC218 Sma-USC268 Sma-USC90 Sma-USC36 Sma-USC185 Sma-USC219 Sma-USC64 Sma-USC30 Sma-USC200 Sma-USC77 Sma-USC205 B12-I GT14 Sma-USC10 Sma-USC202 Sma-USC270 Sma-USC65 Sma-USC12 3/3GT Sma-USC28 Sma-USC132 Sma-USC227 Sma-USC37 Sma4-14INRA Sma-USC238 Sma-USC154 Sma-USC272 Sma-USC174 Sma-USC194 Sma-USC18 Sma-USC208 Sma-USC150 Sma-USC21 Sma-USC226 Sma-USC113 Sma-USC281 Sma-USC116 Sma-USC152 Sma-USC275 Sma-USC8 Sma-USC258 Sma-USC183 Sma-USC19 Sma1-125INRA Sma-USC16 Sma-USC27 Sma-USC220 Sma-USC213 Sma-USC253 Sma-USC232 Sma-USC149 Sma-USC45 Sma.USC111 Smax-01 Sma-USC50 Sma3-8INRA Sma-USC250 Sma-USC31 Sma3-129INRA Smax-02 Sma-USC52 Sma-USC160 Sma-USC193 Sma-USC23 F1-OCA19 3/20CA17 Sma-USC24 Sma-USC95 Sma-USC284 Sma-USC29 Sma-USC231 Sma-USC117

LG2

2.1. Diploid gynogenetic offspring A diploid meiogynogenetic family was obtained at the facilities of the Instituto Español de Oceanografía (IEO, Vigo) using a female and a donor-sperm male coming from a natural population following the procedure by Piferrer et al. (2004). Three different samples from this family were considered attending to the objectives of this study: i) forty eight diploid gynogenetic individuals of 90 days post-hatching (dph) were used for estimating gene centromere (G-C) distances; ii) two additional samples at 1 dph and 10 dph (48 larvae each) from the same female, respectively, were obtained for evaluating Mendelian segregation distortion. Both the mother and the sperm-donor male were genotyped for the same microsatellite set to confirm the exclusive maternal inheritance in all offspring (Castro et al., 2003).

LG3

LG4 LG5

LG6

LG7

2.2. Microsatellite markers A specific set of markers were selected at each turbot linkage group to locate centromeres starting from the turbot genetic map previously reported by Bouza et al. (2007). Seventy nine microsatellite markers were finally chosen attending to their distribution across LGs and polymorphism criteria. The markers were amplified following the protocols described by Pardo et al. (2006, 2007).

LG8

LG9

LG10

2.3. Segregation distortion in gynogenetic offspring All markers were tested for deviation from Mendelian expectations using a chi-square test. The sequential Bonferroni correction (Rice, 1989) was considered for multiple tests. G-C distances at those loci with significant deviation after Bonferroni correction were calculated by counting twice the commonest homozygote class according to Thorgaard et al. (1983). The detection of significant deviations from Mendelian segregation in the 90 dph progeny moved us to study two additional samples of 1 dph and 10 dph (48 offspring each). This permitted to confirm the implication of deleterious alleles in such distortion and to find out the time at which this condition could be operating.

LG11

LG12 LG13

LG14

LG15

2.4. Location of centromeres in turbot map: evaluating mapping function Taking into account the majority of acro-subtelocentric chromosomes in turbot karyotype (Bouza et al., 1994; Cuñado et al., 2001), two segregating microsatellite markers located at both extremes of each linkage group were initially selected to ascertain centromere orientation along chromosome axis. Only one locus could be used at LGs 22, 23, 24 and 26 due to their small size and/or availability of segregant markers in the mother. This approach permitted us to localize the ends where the centromere was positioned at uniarmed chromosomes. When large G-C distances were observed with both terminal markers, the centromere was considered internal (biarmed chromosomes). One or more available markers, the closest as possible to the region where the centromere was located, were then selected for a more precise location. The relative position of markers close to centromeres was obtained considering the minimum number of multiple recombinational events. Seventy nine microsatellites were finally analyzed, constituting an average of three markers per LG.

LG16

LG17

LG18 LG19

LG20

LG21

Progenya 11

12

22

2 20 20 12 2 0 37 0 15 7 9 18 31 27 6 2 16 0 0 17 17 10 1 5 8 7 12 20 22 6 27 7 8 8 13 1 3 8 6 16 10 2 21 5 24 3 4 10 6 23 9 6 5 5 5 9 5 19 4 0 6 9 10 8 8 2 12 5 21 3 6 3 21

43 11 0 10 40 17 1 25 31 38 25 5 0 8 28 45 31 34 46 0 19 35 42 37 32 37 31 22 17 37 10 17 33 25 14 42 32 27 40 10 32 38 0 33 2 42 37 34 30 3 33 32 42 27 33 32 35 16 31 44 25 37 27 26 36 44 29 39 0 40 41 40 0

2 16 28 9 4 28 0 19 1 3 10 22 16 12 10 1 0 13 2 30 9 2 3 6 6 4 5 6 7 3 10 23 6 12 19 2 9 11 2 21 6 7 26 10 22 2 7 2 12 22 6 10 0 16 9 5 6 12 13 4 17 2 11 7 4 2 7 2 27 5 1 4 27

yb

Pc

0.915 0.234 0.000 0.323 0.870 0.233 0.013 0.397 0.508 0.792 0.568 0.111 0.000 0.170 0.636 0.938 0.492 0.567 0.958 0.000 0.422 0.745 0.913 0.771 0.696 0.771 0.646 0.458 0.370 0.804 0.213 0.362 0.702 0.556 0.304 0.933 0.727 0.587 0.833 0.213 0.667 0.809 0.000 0.688 0.042 0.894 0.771 0.739 0.625 0.063 0.688 0.667 0.894 0.478 0.702 0.696 0.761 0.340 0.646 0.917 0.521 0.771 0.563 0.634 0.750 0.917 0.604 0.848 0.000 0.833 0.854 0.851 0.000

1.000 0.505 0.248 0.513 0.414 0.000 0.000 0.000 0.000 0.206 0.819 0.527 0.029 0.016 0.317 0.564 0.000 0.000 0.157 0.058 0.117 0.021 0.317 0.763 0.593 0.366 0.090 0.006 0.005 0.317 0.005 0.003 0.593 0.371 0.289 0.564 0.083 0.491 0.157 0.411 0.317 0.096 0.466 0.197 0.768 0.655 0.366 0.021 0.157 0.881 0.439 0.317 0.025 0.016 0.285 0.285 0.763 0.209 0.029 0.046 0.022 0.035 0.827 0.796 0.248 1.000 0.251 0.257 0.386 0.480 0.059 0.705 0.386

P. Martínez et al. / Aquaculture 280 (2008) 81–88

83

Table 1 (continued) Linkage group

Locus

LG22 LG23 LG24 LG25

Sma-USC80 Sma-USC2 Sma-USC229 Sma-USC100 Sma-USC177 Sma-USC175

LG26

Progenya 11

12

22

0 9 10 1 1 9

44 28 22 42 40 28

3 11 6 4 4 10

yb

Pc

0.936 0.583 0.579 0.894 0.889 0.596

0.083 0.655 0.317 0.180 0.180 0.819

Underlined those microsatellites apparently not congruent with previous genetic map information (Bouza et al., 2007). a 1 and 2 represent the two alleles at all loci, so 11 and 22 are the two homozygote genotypes and 12 the heterozygote one. b Heterozygote frequency. Those loci with significant segregation distortion after Bonferroni correction have been corrected according to Thorgaard et al. (1983). c Probability of conformance to Mendelian segregation. In bold characters P values b 0.05.

Once established the centromere orientation along the chromosome axis, the markers were analyzed for consistency between recombination frequencies and their position in turbot map (Bouza et al., 2007). This analysis enabled us to identify a very small fraction of markers, whose position in the map was apparently not congruent with diploid gynogenetic segregation. These markers were excluded from further analysis for centromere location. As a result of our approach, several estimations of the centromere positions were available at each LG (one per microsatellite used) starting from the turbot map (Bouza et al., 2007). Three different mapping functions were applied to estimate G-C distances starting from half recombination frequency (y/2), where y is the frequency of heterozygotes in the diploid gynogenetic progeny: i) complete interference (CI), where one recombinational exchange precludes additional crossovers; ii) Kosambi function (K), which assumes a reduction of interference with distance (Kosambi, 1944); iii) Haldane function (H), which assumes no interference (Haldane, 1919). The mean and standard error of centromere position was then obtained for each LG and mapping function applied, providing a first estimation of centromere position. The standard error was also used to evaluate the adjustment of the different functions (CI, K, H) to the crossover pattern at each LG in turbot meiosis. 3. Results 3.1. Segregation distortion An important proportion of loci (25.6%) showed departure from Mendelian expectations among the 90 dph diploid gynogenetic offspring used to locate centromeres (Table 1). Six of them showed a Table 2 Diploid meiogynogenetic segregation at different times post-hatching in linkage groups with Mendelian segregation distortion

1 dph

10 dph

90 dph

Genotype

SmaUSC219 (LG2)

SmaUSC205 (LG4)

SmaUSC65 (LG5)

SmaUSC174 (LG7)

SmaUSC208 (LG8)

Smax02 (LG17)

Hom1 Het Hom2 P Hom1 Het Hom2 P Hom1 Het Hom2 P

15 23 10 0.480 12 30 6 0.317 0 19 25 0

32 0 16 0.102 25 0 23 0.838 31 0 16 0.029

10 32 6 0.371 8 37 3 0.132 0 34 13 0

18 16 14 0.617 14 18 16 0.796 22 17 7 0.005

15 16 17 0.803 14 13 21 0.403 7 17 23 0.003

11 30 7 0.505 14 24 10 0.563 6 25 17 0.022

Microsatellites analyzed are the closest to the centromeres at each linkage group. dph: days post-hatching. In parentheses linkage groups (LG) with segregation distortion. P: chi-square probability. Hom: homozygote; Het: heterozygote.

Fig. 1. Histogram of recombination frequencies (y) between the 79 turbot microsatellite and their centromeres.

single homozygote class (P = 0), other four loci showed a strong departure (0.01 N P N 0.001), though not significant after Bonferroni correction, and finally other ten revealed a moderate deviation from Mendelian proportions (0.05 N P N 0.01). Remarkably, most of these loci appeared clustered at specific LGs and the degree of the deviation was also similar for all distorted loci within each LG. Furthermore, most loci analyzed within these LGs showed departure from Mendelian segregation, excluding some far apart from centromeres. Four distorted loci (0.05 N P N 0.01) were not in accordance with these observations and could represent random departures from the null hypothesis: Sma-USC132 and Sma-USC220, which were the only loci deviated within their LGs; and Sma-USC45 and Sma-USC111 at LG15, which showed deviations not congruent with the Mendelian segregation of the remaining loci at this LG.

Table 3 Centromere location (cM) estimated from several microsatellite loci at each linkage group using different mapping functions Linkage group

Complete Interference

S.E.a

Kosambib

S.E.a

Haldanec

S.E.a

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8 LG9 LG10 LG11 LG12 LG13 LG14 LG15 LG16 LG17 LG18 LG19 LG20 LG21 LG25 Mean S.E. Range

109.9 48.0 72.0 76.5 69.6 −8.8 84.2 44.9 65.2 91.6 59.7 −1.5 −12.6 59.0 65.0 4.4 −10.0 45.4 55.1 −16.7 −8.3 46.2

2.9 3.2 14.1 2.0 4.1 3.4 2.7 2.4 13.7 4.3 7.3 1.5 11.6 6.8 5.2 4.0 3.6 13.7 2.4 11.1 16.9 1.5 6.4 1.0 1.5–16.9

120.7 52.9 78.0 76.6 88.2 −19.7 89.5 49.7 69.3 115.3 68.9 − 5.4 −26.2 64.5 76.4 − 2.6 −23.5 50.1 72.2 −29.7 −18.5 73.1

8.3 4.8 10.2 1.9 12.4 9.8 1.1 7.7 11.6 9.6 4.7 5.4 25.2 5.4 1.3 1.0 3.1 12.7 5.2 16.9 27.1 1.1 8.7 1.6 1–27.1

136.3 60.7 89.6 76.9 113.1 −37.2 100.9 56.8 78.9 150.3 85.5 − 13.4 − 46.3 75.6 96.5 − 16.3 − 46.0 61.3 99.6 − 50.5 − 34.6 112.6

23.1 11.4 5.4 1.6 26.1 19.7 4.6 12.4 8.3 25.8 4.9 13.4 45.2 7.7 7.3 5.0 11.5 11.0 12.3 26.8 43.2 0.6 15.1 2.7 1.6–45.2

LG: linkage group. A single polymorphic marker could be analyzed at LGs 22, 23, 24 and 26 and therefore centromere position could not be estimated. The mean, standard error and range of SE across all LGs is presented in the last rows. Centromeres were located using as reference the previous genetic turbot map (Bouza et al., 2007). a Standard error. b Kosambi mapping function. c Haldane mapping function.

84 P. Martínez et al. / Aquaculture 280 (2008) 81–88 Fig. 2. Location of centromeres in turbot map after half-tetrad segregation analysis. In bold characters framework microsatellite loci. Accessory markers in parentheses beside the closest framework marker. Marker positions appear modified regarding turbot map (Bouza et al., 2007) due to location of some centromeres at position 0.0 and to the inversion of some LGs to place the short arm upwards. The centromeres of the shortest LGs (22–26) are not represented because a single marker was available or the centromere was too long from the available markers (LG25).

85

Fig. 2 (continued ).

P. Martínez et al. / Aquaculture 280 (2008) 81–88

86

P. Martínez et al. / Aquaculture 280 (2008) 81–88

These results supported the existence of alleles of different deleterious effect at specific loci, which would be determining a lower or null viability of one homozygote class. To confirm this hypothesis and to check the time at which the deleterious condition could be acting, segregation of the locus closest to the centromere within each of these LGs was analyzed at 1 and 10 dph progenies from the same female (Table 2). At all cases no significant deviation from Mendelian segregation was observed neither at 1 nor at 10 dph, discarding an explanation based on aberrant meiotic segregation, and suggesting the action of deleterious or lethal alleles beyond 10 dph. 3.2. Gene-centromere (G-C) distances: evaluating mapping function Heterozygote frequencies (y) for the 79 microsatellite loci appeared mostly distributed at both extremes of the frequency histogram (y mean = 0.575; range: 0–0.958), but with a higher representation above 0.50 (68.4% of the data; Fig. 1). A large number of loci showed heterozygote frequency above 0.667 (48.1%), the expected value for independent segregation between one locus and its centromere under the assumption of no interference. Furthermore, 10.1% loci revealed y values above 0.9. The shape of the distribution and the scarcity of intermediate y values could be partially explained by the initial sampling scheme of loci at both extremes of each LG to find out centromere orientation along chromosome axis. Anyway, these results showed high interference in turbot meiosis, so the existence of a chiasma would be precluding or reducing the probability of new chiasmata along the same chromosome axis. The comparison between the turbot map (Bouza et al., 2007) and the diploid gynogenetic segregation data evidenced some incongruities between the position of specific markers within particular LGs and the recombination frequency observed (Table 1). So, Sma-USC268 (LG1: 115.1 cM) showed a y value similar or below to that observed at other loci apparently much closer to the centromere. The remaining loci at this LG showed congruent information between their y values and the position in the map. In this way, we could identify 5 microsatellites, which showed apparent incongruities between both data sets. Most of them, excluding Smax-01, were not framework markers (linked at LOD b 3.0) in the turbot map and appeared located at the telomeres far away from centromeres. These markers were excluded for estimating centromere position. The remaining microsatellites, even those ordered at LOD ≤ 2 (Sma-USC218, Sma-USC116, Sma-USC270, Sma-USC65 and Sma-USC194), showed consistency between both data sets, constituting an additional support for their location in the turbot map. A first approach to locate centromeres was to use all markers from each LG applying different mapping functions: complete interference, Kosambi and Haldane. The mean and standard error (SE) of centromere position was then obtained at each LG (Table 3). The centromeres appeared accurately located at most LGs, especially when complete interference was applied (mean G-C SE: 6.4 cM), being standard errors mostly around or below 5 cM. This precision diminished when applying Kosambi function (mean G-C SE: 8.7 cM), and was really poor with Haldane function (mean G-C SE: 15.1 cM). However, large errors (above 10 cM) were evidenced at some LGs after applying complete interference (LGs 3, 9, 13, 18, 20 and 21), in most cases attributable to distant markers to the centromere, which rendered aberrant centromere estimations (3, 9, 20 and 21). However, this was not a general rule, and some distant markers did not increase the error when applying complete interference (i.e.: Sma-USC271 at LG1; Sma-USC90 at LG2; Sma-USC194 at LG8). Also, A few LGs performed better with Kosambi and Haldane functions (3, 9, 11). These results evidenced variable chiasma interference across turbot chromosomes. Finally, a specific distortion was observed at LG13, 18 and 21, probably as a consequence of the different recombination frequency observed between males and females in turbot map (Bouza et al., 2007). The position of some of the markers used in these LGs was obtained only via male segregation by Bouza et al. (2007), while

in this study the segregation analysis was via female in the gynogenetic family. This could increase the distance with regard to the original map and therefore magnify the error in centromere location. 3.3. Centromere location: consolidating turbot map Centromeres were finally located at all LGs where two or more markers were available (Fig. 2; LGs 22, 23, 24 and 26 excluded). The following criteria were applied: i) when fully linked markers existed (y=0), the centromere was considered at the same position of that marker (LGs 1, 4, 6, 12, 20 and 21); ii) when genetic markers around or below 10 cM were available, these were used to estimate centromere position using complete interference (LGs 2, 3, 8, 11, 13, 14); iii) at the remaining cases (LGs 5, 7, 9, 10, 15, 16, 17, 18, 19 and 25), the mean G-C distance of all available markers at each LG applying complete interference was used. Following these criteria all these centromeres, excluding those at LG9 and LG18, were positioned with an error ≤5.2 cM in the turbot map. At LG9 a distant marker (N63 cM) strongly distorted centromere location, so other functions which diminished interference with distance (SE: 11.6) or considered no interference (SE: 8.3) performed better than complete interference (SE: 13.7). As outlined before, a strong distortion due to unequal male/female recombination frequencies could exist at LG18, hence the high error observed with all functions (N11.0 cM). Marker positions in turbot map were corrected according to centromere position. To say, the shortest chromosome arm was oriented upwards and the centromere was located at 0.0 cM when appeared at the end of the chromosome (Fig. 2). Via male and female recombination frequencies were compared regarding the position of markers along the chromosome axis using centromere position and previous segregation data (Bouza et al., 2007). The correlation performed between the absolute differences between male and female recombination at 24 common pairs of loci in 17 LGs and their distance to centromere showed a moderate, but significant inverse relationship (r = −0.414; P = 0.044). The closer the markers to the centromere, the larger the difference between male and female recombination frequency. 4. Discussion 4.1. Deviation from Mendelian expectations Lethal or deleterious genes should cause distortion in Mendelian segregation in diploid meiogynogenetics (Allendorf et al., 1986; Nichols et al., 2003). The number of lethal-equivalents has been estimated between 1.5 and 5 per zygote in humans (Jorde, 2001), 3.6 in zebrafish (McCune et al., 2004) and it has demonstrated to be quite variable among vertebrates (Keller et al., 2002). In spite of this, many studies in fish using diploid gynogenetics did not find segregation distortion (Allendorf et al., 1986; Liu et al., 1992; Lindner et al., 2000; Matsuoka et al., 2004; Li and Kijima, 2005, 2006). In our work, near 25% loci showed Mendelian distortion and 8% were still significant after correction for multiple tests. Furthermore, most loci with weak or moderate deviations co-mapped at specific LGs, as in the study by Nichols et al. (2003), suggesting the existence of deleterious alleles of variable effect in different LGs. The lack of distortion for the same microsatellites observed in the 1 dph and 10 dph samples supported the deleterious-based deviation, and that these alleles would be operating beyond 10 days after hatching. This result is in agreement with the conformance of haploid meiogynogenetic embryos to Mendelian segregation (Bouza et al., 2007) and with the low viability observed in turbot diploid gynogenetic (Piferrer et al., 2004). The use of a limited number of markers in other studies, together with the mechanical application of correction for multiple tests and the use of recently hatched larvae (Li and Kijima, 2006) could be preventing the detection of the expected deleterious segregation distortion. Also, the existence of duplicated loci could account for this observation, especially in species of tetraploid origin, like salmonids (Allendorf et al., 1986).

P. Martínez et al. / Aquaculture 280 (2008) 81–88

A strong mortality has been observed in turbot culture around 10 days post-hatching, much higher than in other flatfish, even of the same genus (Purdom et al., 1972). Bouza et al. (1997) invoked inbreeding depression problems associated to a historical bottleneck in this species (Blanquer et al., 1992) to explain this observation. The number of lethal-equivalents observed in the mother of meiogynogenetics in our study would be around 4, considering two lethals at LGs 2 and 5, and four semilethals at LGs 4, 7, 8 and 17. This value would be in the range reported in other species (Keller et al., 2002; McCune et al., 2004). Although a single female was evaluated, the results obtained do not support a particularly high genetic load in turbot, which could explain the low larval viability observed.

87

In this study, seventy nine microsatellite markers selected among all LGs have been used to position centromeres in turbot map. Segregation distortion was detected at several loci suggesting the presence of genes of variable deleterious genes. Complete interference was the function that best fitted to the recombinant pattern observed, though other functions performed better at specific LGs. The information obtained will be valuable for applying the turbot map for QTL identification or eventually, for positional cloning strategies in this species. This study represents a further step towards consolidating the genetic map of turbot, a species of great relevance for European aquaculture. Acknowledgements

4.2. Gene centromere (G-C) distances in turbot Heterozygote frequencies (y) in turbot gynogenetics ranged between 0 and 0.938 in our work with a mean of 0.575. Near 50% of genes showed y values above 0.667, the expected value under independent segregation without chiasma interference (Thorgaard et al., 1983), and more than 10% above 0.9. Recombination frequencies appeared skewed toward high values, as reported in other fish species (Kauffman et al., 1995; Lindner et al., 2000; Matsuoka et al., 2004; Nomura et al., 2006). All these data suggest high chiasma interference in turbot, as reported in fish (Thorgaard et al., 1983; Allendorf et al., 1986; Kauffman et al., 1995; Lindner et al., 2000; Matsuoka et al., 2004; Nomura et al., 2006). These observations have led to the assumption of complete interference as the best mapping function to explain meiotic events in fish, and therefore, to estimate G-C distances as half of recombination frequency (Thorgaard et al., 1983; Johnson et al., 1995). Our data supported that complete interference was globally the best fitted function to turbot data. However, other functions like Kosambi or even Haldane function, which assumes no interference, showed a lower error than complete interference at specific LGs, mostly due to genetic markers far from centromeres. The heterogeneity of LGs regarding chiasma distribution pattern has been reported in other species (Nachman and Churchill, 1996; Danzmann and Gharbi, 2001), and suggests caution when applying complete interference for estimating G-C distances in fish. 4.3. Centromere mapping Recombination frequencies between microsatellite markers and the centromeres obtained in our study were mostly congruent with their positions in the map previously reported by Bouza et al. (2007). Only five out of 79 markers showed striking discordances, which could be related with an incorrect position in the turbot map. The fact that most of them were not framework markers would support this observation. However, if we look at their telomeric and remote position regarding centromeres, it is possible that double crossovers could be determining lower recombination frequencies than interstitial markers, as suggested by Danzmann and Gharbi (2001). Positioning centromeres is relevant for studies on meiotic function or comparative mapping. It is also necessary for applying genetic map to marker assisted selection programs, to assign mutants or markers to LGs, and for positional cloning strategies to identify genes related with productive characters (Thorgaard et al., 1983; Johnson et al., 1995, 1996; Mohideen et al., 2000; Danzmann and Gharbi, 2001). Centromeres were positioned at most LGs in turbot with a SE ≤ 5 cM and markers closely linked to centromeres (≤5 cM) are now available at 8 LGs in turbot map. Centromere location was mostly in accordance with previous cytogenetic data in turbot (Bouza et al., 1994; Cuñado et al., 2001). Most LGs showed an acrocentric structure and two large metacentric LG were identified. As in other studies (Sakamoto et al., 2000; O'Malley et al., 2003), recombination frequency differences between males and females were evident close to centromeres.

This study was supported by a project from the Xunta de Galicia local Government (PGIDIT04RMA261014PR). We are indebted to Lucía Insua, María Portela, Susana Sánchez, María López and Sonia Gómez for the technical assistance. References Allendorf, F.W., Seeb, K.E., Knudsen, K.L., Thorgaard, G.H., Leary, R.F., 1986. Genecentromere mapping of 25 loci in rainbow trout. J. Hered. 77, 307–321. Blanquer, A., Alayse, J.P., Berrada-Rkhami, O., Berrebi, P., 1992. Allozyme variation in turbot (Psetta maxima) and brill (Scophthalmus rhombus) (Osteichthyes, Pleuronectiformes, Scophthalmidae) throughout their range in Europe. J. Fish Biol. 41, 725–736. Blott, S., Kim, J.J., Moisio, S., Schmidt-Kuntzel, A., Cornet, A., Berzi, P., Cambisano, N., Ford, C., Grisart, B., Johnson, D., Karim, L., Simon, P., Snell, R., Spelman, R., Wong, J., Vilkki, J., Georges, M., Farnir, F., Coppieters, W., 2003. Molecular dissection of a quantitative trait locus: A phenylalanine-to-tyrosine substitution in the transmembrane domain of the bovine growth hormone receptor is associated with a major effect on milk yield and composition. Genetics 163, 253–266. Bouza, C., Sánchez, L., Martínez, P., 1994. Karyotypic characterization of turbot (Scophthalmus maximus) with conventional, fluorochrome, and restriction endonuclease banding techniques. Mar. Biol. 120, 609–613. Bouza, C., Sánchez, L., Martínez, P., 1997. Gene diversity analysis in natural populations and cultured stocks of turbot (Scophthalmus maximus L.). Anim. Genet. 28, 28–36. Bouza, C., Castro, J., Presa, P., Sánchez, L., Martínez, P., 2002. Allozyme and microsatellite diversity in natural and domestic populations of turbot (Scophthalmus maximus) in comparison with other Pleuronectiformes. Can. J. Fish. Aquat. Sci. 59, 1460–1473. Bouza, C., Hermida, M., Pardo, B.G., Fernández, C., Castro, J., Fortes, G.G., Sánchez, L., Presa, P., Pérez, M., Sanjuán, A., Comesaña, S., Álvarez-Dios, J.A., Calaza, M., Cal, R., Piferrer, F., Martínez, P., 2007. A microsatellite genetic map in the turbot (Scophthalmus maximus). Genetics 177, 2457–2467. Castro, J., Bouza, C., Sánchez, L., Cal, R.M., Piferrer, F., Martínez, P., 2003. Gynogenesis assessment by using microsatellite genetic markers in turbot (Scophthalmus maximus). Mar. Biotechnol. 5, 584–592. Castro, J., Bouza, C., Presa, P., Pino-Querido, A., Riaza, A., Ferreiro, I., Sánchez, L., Martínez, P., 2004. Potential sources of error in parentage assessment of turbot (Scophthalmus maximus) using microsatellite loci. Aquaculture 242, 119–135. Colosimo, P.F., Hosemann, K.E., Balabhadra, S., Villarreal, G., Dickson, M., Grimwood, J., Schmutz, J., Myers, R.M., Schluter, D., Kingsley, D.M., 2005. Widespread pararell evolution in sticklebacks by repeated fixation of ectodysplasin alleles. Science 307, 1928–1933. Cuñado, N., Terrones, J., Sánchez, L., Martínez, P., Santos, J.L., 2001. Synaptonemal complex analysis in spermatocytes and oocytes of turbot, Scophthalmus maximus (Pisces, Scophthalmidae). Genome 44, 1143–1147. Danzmann, R.G., Gharbi, K., 2001. Gene mapping in fishes: a means to an end. Genetica 111, 3–23. Donovan, A., Brownlie, A., Zhou, Y., Shepard, J., Pratt, S.J., Moynihan, J., Paw, B.H., Drejer, A., Barut, B., Zapata, A., Law, T.C., Brugnara, C., Kingsley, P.D., Palis, J., Fleming, M.D., Andrews, N.C., Zon, L.I., 2000. Positional cloning of zebrafish ferroportin1 identifies a conserved vertebrate iron exporter. Nature 403, 776–781. Guyomard, R., Mauger, S., Tabet-Canale, K., Martineau, S., Genet, C., Krieg, F., Quillet, E., 2006. A type I and type II microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) with presumptive coverage of all chromosome arms. BMC Genomics 7, 1–13. Haldane, J.B.S., 1919. The mapping function. J. Genet. 8, 299–309. Johnson, S.L., Africa, D., Horne, S., Postlethwait, J.H., 1995. Half-tetrad analysis in zebrafish: mapping the ros mutation and the centromere of linkage group I. Genetics 139, 1727–1735. Johnson, S.L., Gates, M.A., Johnson, M., Talbot, W.S., Horne, S., Baik, K., Rude, S., Wong, J.R., Postlewait, J.H., 1996. Centromere-linkage analysis and consolidation of the zebrafish genetic map. Genetics 142, 1277–1288. Jorde, L.B., 2001. Consanguinity and prereproductive mortality in the Utah Mormon population. Hum. Hered. 52, 61–65. Kauffman, E.J., Gestl, E.E., Kim, D.J., Walker, C., Hite, J.M., Yan, G., Rogan, P.K., Johnson, S.L., Cheng, K.C., 1995. Microsatellite-centromere mapping in the zebrafish (Danio rerio). Genomics 30, 337–341.

88

P. Martínez et al. / Aquaculture 280 (2008) 81–88

Keller, L.F., Grant, P.R., Grant, B.R., Petren, K., 2002. Environmental conditions affect the magnitude of inbreeding depression in survival of Darwin's finches. Evolution 56, 1229–1239. Kosambi, D.D., 1944. The estimation of map distances from recombination values. Ann. Eugenics 12, 172–175. Li, Q., Kijima, A., 2005. Segregation of microsatellite alleles in gynogenetic diploid Pacific abalone (Haliotis discus hannai). Mar. Biotechnol. 7, 669–676. Li, Q., Kijima, A., 2006. Microsatellite analysis of gynogenetic families in the Pacific oyster. J. Exp. Mar. Biol. Ecol. 331, 1–8. Lindner, K.R., Seeb, J.E., Habicht, C., Knudsen, K.L., Kretschmer, E., Reedy, D.J., Spruell, P., Allendorf, F.W., 2000. Gene-centromere mapping of 312 loci in pink salmon by halftetrad analysis. Genome 45, 538–549. Liu, Q., Goudie, C.A., Simco, B.A., Davis, K.B., Morizot, D.C., 1992. Gene-centromere mapping of six enzyme loci in gynogenetic channel catfish. J. Hered. 83, 245–248. Mackay, T.F.C., 2001. The genetic architecture of quantitative traits. Ann. Rev. Genet. 35, 303–339. Matsuoka, M.P., Gharrett, A.J., Wilmot, R.L., Smoker, W.W., 2004. Gene-centromere distances of allozyme loci in even- and odd-year pink salmon, (Oncorhynchus gorbuscha). Genetica 121, 1–11. McCune, A.R., Houle, D., McMillan, K., Annable, R., Kondrashov, A.S., 2004. Two classes of deleterious recessive alleles in a natural population of zebrafish, Danio rerio. Proc. R. Soc. Lond. B 271, 2025–2033. Mohideen, M.A.P.K., Moore, J.L., Cheng, K.C., 2000. Centromere-linked microsatellite markers for linkage groups 3, 4, 6, 7, 13, and 20 of zebrafish (Danio rerio). Genomics 67, 102–106. Morishima, K., Nakayama, I., Arai, K., 2001. Microsatellite-centromere mapping in the loach, Misgurnus anguillicaudatus. Genetica 111, 59–69. Nachman, M.W., Churchill, G.A., 1996. Heterogeneity in rates of recombination across the mouse genome. Genetics 142, 537–548. Nichols, K.M., Young, W.P., Danzmann, R.G., Robison, B.D., Rexroad, C., Noakes, M., Phillips, R.B., Bentzen, P., Spies, I., Knudsen, K., Allendorf, F.W., Cunningham, B.M., Brunelli, J., Zhang, H., Ristow, S., Drew, R., Brown, K.H., Wheeler, P.A., Thorgaard, G.H., 2003. A consolidated linkage map for rainbow trout (Oncorhynchus mykiss). Anim. Genet. 34, 102–115.

Nomura, K., Morishima, K., Tanaka, H., Unuma, T., Okuzawa, K., Ohta, H., Arai, K., 2006. Microsatellite-centromere mapping in the Japanese eel (Anguilla japonica) by halftetrad analysis using induced triploid families. Aquaculture 257, 53–67. O'Malley, K.G., Sakamoto, T., Danzmann, R.G., Ferguson, M.M., 2003. Quantitative trait loci for spawning date and body weight in rainbow trout: testing for conserved effects across ancestrally duplicated chromosomes. J. Hered. 94, 273–284. Pardo, B.G., Hermida, M., Fernández, C., Bouza, C., Pérez, M., Llavona, A., Sánchez, L., Martínez, P., 2006. A set of highly polymorphic microsatellites useful for kinship and population analysis in turbot (Scophthalmus maximus L.). Aquac. Res. 37, 1578–1582. Pardo, B.G., Fernández, C., Hermida, M., Vázquez, A., Pérez, M., Presa, P., Calaza, M., Alvarez-Dios, J.A., Comesaña, A.S., Raposo-Guillán, J., Bouza, C., Martínez, P., 2007. Development and characterization of 248 novel microsatellite markers in turbot (Scophthalmus maximus). Genome 50, 329–332. Piferrer, F., Cal, R.M., Gómez, C., Alvarez-Blázquez, B., Castro, J., Martínez, P., 2004. Induction of gynogenesis in the turbot (Scophthalmus maximus): effects of UV irradiation on sperm motility, the Hertwig effect and viability during the first 6 months of age. Aquaculture 238, 403–419. Purdom, C.E., Jones, A., Lincoln, R.F., 1972. Cultivation trials with turbot (Scophthalmus maximus). Aquaculture 1, 213–230. Rice, W.R., 1989. Analyzing tables of statistical tests. Evolution 43, 223–225. Sakamoto, T., Danzmann, R.G., Gharbi, K., Howard, P., Ozaki, A., Khoo, S.K., Woram, R.A., Okamoto, N., Ferguson, M.M., Holm, L.E., Guyomard, R., Hoyheim, B., 2000. A microsatellite linkage map of rainbow trour (Oncorhynchus mykiss) characterized by large sex-specific differences in recombination rates. Genetics 155, 1331–1345. Sewell, M.M., Sherman, B.K., Neale, D.B., 1999. A consensus map for loblolly pine (Pinus taeda L.). I. Construction and integration of individual linkage maps from two outbred three-generation pedigrees. Genetics 151, 321–330. Thorgaard, G.H., Allendorf, F.W., Knudsen, K.L., 1983. Gene-centromere mapping in rainbow trout: high interference over long map distances. Genetics 103, 771–783.