Detección de QTLs de deformidades esqueléticas en dorada (Sparus aurata L.): herramientas moleculares y análisis de segregación

ANEXO II UNIVERSIDAD DE LAS PALMAS DE GRAN CANARIA Instituto Universitario de Sanidad Animal y Seguridad Alimentaria PROGRAMA DE DOCTORADO TITULADO “ACUICULTURA: PRODUCCIÓN CONTROLADA DE ANIMALES ACUÁTICOS” Tesis Doctoral presentada por Dña. Davinia Negrín Báez Dirigida por la Dra. María Jesús Zamorano Serrano y la Dra. Ana Navarro y Guerra del Río La Directora La Directora La Doctoranda Mª Jesús Zamorano Ana Navarro y Davinia Negrín Serrano Guerra del Río Báez Las Palmas de Gran Canaria, 2014 UNIVERSIDAD DE LAS PALMAS DE GRAN CANARIA Instituto Universitario de Sanidad Animal y Seguridad Alimentaria TESIS DOCTORAL DETECCIÓN DE QTLS DE DEFORMIDADES ESQUELÉTICAS EN DORADA ( Sparus aurata L.): HERRAMIENTAS MOLECULARES Y ANÁLISIS DE SEGREGACIÓN Davinia Negrín Báez Las Palmas de Gran Canaria, 2014 María Jesús Zamorano Serrano, Profesora Contratada Doctora de la Universidad de Las Palmas de Gran Canaria, perteneciente al Departamento de Patología Animal, Producción Animal, Bromatología y Ciencia y Tecnología de los Alimentos de la Universidad de Las Palmas de Gran Canaria, Ana Navarro y Guerra del Río, Investigadora Contratada por la Universidad de Las Palmas de Gran Canaria, INFORMAN: Que Davinia Negrín Báez, Licenciada en Veterinaria por la Universidad de Las Palmas de Gran Canaria, ha realizado bajo nuestra dirección y asesoramiento, el presente trabajo titulado “DETECCIÓN DE QTLS DE DEFORMIDADES ESQUELÉTICAS EN DORADA ( Sparus aurata L): HERRAMIENTAS MOLECULARES Y ANÁLISIS DE SEGREGACIÓN, el cual consideramos reúne las condiciones y la calidad científica para optar al grado de Doctora. En Las Palmas de Gran Canaria, a 18 de Septiembre de 2014. Fdo.: Mª Jesús Zamorano Serrano Fdo.: Ana Navarro y Guerra del Río AGRADECIMIENTOS Llevo diez minutos mirando la pantalla, así que voy a empezar a escribir ya a ver si me viene la inspiración por el camino. Han pasado ya ocho años desde que empecé la primera tesis, y por lo tanto se han ido acumulando personas a las que, por una cosa o por otra, les tengo que agradecer que haya llegado hasta aquí, y que, por fin, después de tres intentos, haya terminado la “bendita” tesis. En primer lugar, para empezar por algún sitio, quiero agradecer a mi co-directora de tesis, María Jesús Zamorano, haberme dado la oportunidad de llevar a cabo este trabajo. Agradezco su confianza en mí y su apoyo, así como ese toque maternal y femenino que tanta falta le hace a un lugar de trabajo. Aunque no lo es de esta tesis, sí que lo fue de los otros dos intentos, así que para mí es como si lo fuera. Quiero agradecer a Juan Manuel Afonso, primero, que me haya permitido formar parte de este grupo de trabajo, y segundo… mil cosas más: gracias por transmitirme parte de tu formación en biología molecular (tanto en el laboratorio, como por los pasillos, en el despacho, en un avión y en cualquier sitio y momento donde me surgiera una duda); gracias por recordarme con tu actitud, que ante todo, hay que hacer las cosas con ilusión y fe en uno mismo; gracias por apoyarme, a tu manera, en los momentos más complicados de estos años y por esa concienzuda diplomacia que te ha hecho mantener la paz cuando yo había desplegado el arsenal de guerra. GRACIAS a aquellas personas que han pasado por el laboratorio durante algún tiempo y con los que he compartido momentos inolvidables. GRACIAS: Sara, cuyo trabajo está reflejado en gran parte de esta tesis, Dulce (Dulcelina), Miguel (culito pelota), Andrea, Fatma,… GRACIAS también a todo el personal del GIA, a los técnicos, investigadores y becarios que en mayor o menor medida han colaborado en la realización de este documento, ya sea por la ayuda en un muestreo, en el análisis de las muestras, en el asesoramiento científico o, por hacer que venir a trabajar merezca la pena hasta en los días más difíciles. Especial mención para Rafa Ginés, Ada, Carmen (Carmelita), Silvia González, Nati, Silvia Torrecillas, Alex Makol, Fran Otero, Juan (el pulpo), Tibi, Mónica, Fefy, Ivonne. ¡Gracias chic@s! Son los mejores compañeros del mundo. También quiero agradecer su trabajo y apoyo al equipo de trabajo de la facultad de veterinaria y del IUSA, que han contribuido a que las largas horas de trabajo en el laboratorio fueran más amenas. GRACIAS a Manola, María, Jimena, Lorena, Sergio (Maruco) y Yeray Parásitos, … Sin duda, lo único por lo que volvería hacer una tesis si retrocediera en el tiempo, sería porque estos años he conocido a los mejores amigos que podría encontrar. Gracias a la Superpandi: GRACIAS a Guacimara Alejandro (no sin mi Guaci), por ser mi mejor alumna, por tu ilusión ante el trabajo, pero sobre todo, por tu amistad incondicional y tu férreo apoyo. GRACIAS a Laura (Lauris), por tus constantes subidones de autoestima, tu fidelidad, por las horas interminables de confesiones y por hacer de todas las situaciones un motivo para la risa. GRACIAS a Silvia Hildebrant (la Dra. Sildebrant), nuestra alemana “cuadriculada” (jiji), porque siempre se puede contar contigo, por muy ocupada que estés, ya sea para corregir nuestro inglés, para asesoramiento tecnológico o para un relajante Agradecimientos spa . GRACIAS a Mohamed Soula, que aunque ahora estés lejos y los últimos momentos no fueran los mejores, serás siempre para mí, mi Moji. Gracias al resto, a los sufridores cónyuges, por esas horas de diversión, comilonas y risas: mil gracias a Borja, Oscar, Jacob, Jose Ángel y Cristina. No creo que nadie me pregunte por qué le dedico un espacio solamente a ella: GRACIAS a Ana Navarro. No sé cuántas vidas tendría que vivir para agradecerte todo lo que has hecho por mí estos años. Tengo muy claro que, si voy a hacer Doctora, es gracias a ti. Y no sólo porque en la recta final te hayas convertido en co-directora (ya te lo merecías hacía tiempo), sino porque has trabajado en esta tesis casi tanto como en la tuya. Has luchado conmigo codo con codo para sacar este trabajo adelante, levantándome cuando me caía o cuando me dejaba caer, sin dejarme desfallecer ni rendirme, confiando en mí hasta el final. GRACIAS y mil veces GRACIAS. No me olvido, claro que no, de aquellos amigos que, aunque no están metidos en este mundillo, me han apoyado incondicionalmente y han sufrido conmigo las dificultades de hacer una tesis. GRACIAS a Yure y Rubén (mis “canarios en Londres” ) por esas horas delante de un café hablando de la vida, por confiar en que podía llegar hasta aquí y recordármelo cuando yo lo olvidaba. GRACIAS a Chedey por su apoyo, cariño y paciencia los años que estuvimos juntos (y también después); y por intentar enseñarme a valorar las cosas que realmente importan. Si alguien me pidiera que definiera lo que es la amistad, yo sólo diría: los amigos son aquellas personas que creen en ti cuando tú has dejado de hacerlo. Todas estas personas lo hacen por mí, por eso son mis amigos y los quiero. DOCTORAL THESIS DETECTION OF QTLs FOR SKELETAL DEFORMITIES IN GILTHEAD SEABREAM ( Sparus aurata L.): MOLECULAR TOOLS AND SEGREGATION ANALYSIS Davinia Negrín Báez Las Palmas de Gran Canaria, 2014 i I ndex Abstract .................................................................................................................................................. 1 1. General Introduction ..................................................................................................................... 5 1.1. Species of Study: gilthead seabream ........................................................................................ 7 1.1.1. Taxonomy and Nomenclature .......................................................................................... 7 1.1.2. Distribution and Habitat ................................................................................................... 7 1.1.3. Anatomy ........................................................................................................................... 8 1.1.4. Feeding and Reproduction ............................................................................................... 8 1.2. Aquaculture Production of Gilthead Seabream ....................................................................... 9 1.2.1. Aquaculture Production at European and Global Level .................................................. 9 1.2.2. Aquaculture Production in Spain ................................................................................... 10 1.3. Justification of the Study ....................................................................................................... 11 1.4. Objectives of the Study .......................................................................................................... 14 2. A set of 13 multiplex PCRs of specific microsatellite markers as a tool for QTL detection in gilthead seabream (Sparus aurata L.)................................................................................................. 17 2.1. Abstract .................................................................................................................................. 19 2.2. Introduction ............................................................................................................................ 20 2.3. Material and Methods ............................................................................................................ 23 2.3.1. Samples and DNA extraction ......................................................................................... 23 2.3.2. Multiplex PCRs .............................................................................................................. 23 2.3.3. Multiplex PCRs validation and genotyping reliability revaluation................................ 28 2.4. Results .................................................................................................................................... 29 2.4.1. Multiplex PCRs .............................................................................................................. 29 2.4.2. Multiplex PCRs validation and genotyping reliability revaluation................................ 32 2.5. Discussion .............................................................................................................................. 34 3. Segregation Analysis of Skeletal Deformities in the Gilthead Seabream (Sparus aurata, L); Lack of Operculum, Lordosis, Vertebral Fusion and LSK.............................................................. 39 3.1. Abstract .................................................................................................................................. 41 3.2. Introduction ............................................................................................................................ 43 3.3. Material and Methods ............................................................................................................ 44 3.3.1. Experiment 1 (mass spawning with sorting) .................................................................. 44 3.3.2. Experiment 2 (mass spawning without sorting) ............................................................. 45 3.3.3. Experiment 3 (designed mating) .................................................................................... 46 3.3.4. Statistical analysis .......................................................................................................... 47 ii 3.4. Results .................................................................................................................................... 49 3.4.1. Experiment 1 .................................................................................................................. 49 3.4.2. Experiment 2 .................................................................................................................. 50 3.4.3. Experiment 3 .................................................................................................................. 51 3.5. Discussion .............................................................................................................................. 53 4. Quantitative Trait Loci for Skeletal Deformities in Gilthead Seabream (Sparus aurata, L); Lack of Operculum, Lordosis, Vertebral Fusion and LSK complex ............................................... 61 4.1. Abstract .................................................................................................................................. 63 4.2. Introduction ............................................................................................................................ 65 4.3. Material and Methods ............................................................................................................ 69 4.3.1. Fish, Trait Measurement and Parental Assignment ........................................................ 69 4.3.2. Genotyping ..................................................................................................................... 70 4.3.3. QTL mapping ................................................................................................................. 70 4.3.4. Genotypic Association Analysis .................................................................................... 72 4.4. Results .................................................................................................................................... 74 4.4.1. Genotyping ..................................................................................................................... 74 4.4.2. Study 1 ........................................................................................................................... 74 4.4.3. Study 2 ........................................................................................................................... 76 4.4.4. Study 3 ........................................................................................................................... 78 4.5. Discussion .............................................................................................................................. 81 4.5.1. Study 1 ........................................................................................................................... 82 4.5.2. Study 2 ........................................................................................................................... 85 4.5.3. Study 3 ........................................................................................................................... 87 5. Conclusions ................................................................................................................................... 91 6. References ..................................................................................................................................... 95 Anexo. Resumen en español ................................................................................................................ 105 iii Figures Index Figure 1. Side view of the external appearance of an adult gilthead seabream ...................................... 8 Figure 2. Progression of gilthead seabream aquaculture production in the Mediterranean area and in the rest of the World from 1983 to 2013 ............................................................................................... 10 Figure 3. Distribution of the gilthead seabream productions in Spain for Autonomous regions in 2013 (%) ......................................................................................................................................................... 10 Figure 4. Aquaculture production of gilthead seabream in the Canary Islands.................................... 11 Figure 5. Genetic distances between markers used in this study based in 26 linkage groups from gilthead seabream genetic map ............................................................................................................. 28 iv Tables Index Table 1. Description of microsatellite marker used in this study. ......................................................... 24 Table 2. Genetic information of each microsatellite marker of 13 multiplex PCRs (ReMsa) and alternative microsatellite markers .......................................................................................................... 30 Table 3. Number of loci and genetic variability for each multiplex PCR (ReMsa) developed in this study on gilthead seabream.................................................................................................................... 37 Table 4. Deformities prevalence in total offpring in each experiment ................................................. 49 Table 5. Deformities prevalence in descendants of breeders responsible of LSK deformed fish (lordosis/scoliosis/kyphosis complex) with respect to each sire, dam and family of the Experiment 1 50 Table 6. Deformities prevalence in offspring from mw♂1 of the Experiment 2 .................................. 51 Table 7. Deformities prevalence in offpring with respect to different matings of the Experiment 3 ... 52 Table 8. Family structure, number of offspring analyzed in each study and number of fish showing each deformity ....................................................................................................................................... 72 Table 9. QTL mapping information. ..................................................................................................... 74 Table 10. Description of QTL detected in Study 1 for LSK complex in gilthead seabream ................ 77 Table 11. Description of QTLs detected in Study 2 for lack of operculum deformity in gilthead seabream ................................................................................................................................................ 78 Table 12.Description of QTL detected in Study 3 for vertebral fusion, lordosis and jaw deformity in gilthead seabream .................................................................................................................................. 79 Table 13. Percentage of variance explained by the QTL expressed in PVE and R2 and allelic association analysis for microsatellite markers close to each solid QTL detected in the three studies . 80 1 Abstract The aim of the present study was to analyze the existence of QTLs (Quantitative Trait Loci) affecting the most important skeletal deformities in gilthead seabream (Sparus aurata L.), which could be used as a tool to minimize the prevalence of these deformities in this species under a genetic breeding program. The incidence of these anomalies in this species supposes important economic losses in the industry, both hatchery and on-growing companies, so abnormal fish have to be removed at different points of the production line. In a first step, a set of multiplex PCR reactions containing 138 microsatellite markers of gilthead seabream genetic map, which were redesigned to be amplified under the same PCR conditions, was developed. The final set was conformed by 13 new multiplex PCR reactions (named ReMsa1 to ReMsa13) containing 106 of these microsatellite markers. This covered 100% of the linkage groups, and, after being successfully validated, this set constitutes a powerful tool to search QTLs in this species. In order to find a family structure to facilitate the detection of QTLs, the prevalence of important and severe skeletal deformities was analyzed in different populations of gilthead seabream in three large-scale experiments. These deformities were lack of operculum, lordosis, vertebral fusion and LSK complex (lordosis-scoliosis-kyphosis). In Experiment 1, fish were obtained from a mass spawning and, at 111 days, they were sorted in deformity terms. At 509 days post-hatching, 900 fish were analyzed: 846 fish on-grown in a farm and 54 LSK fish selected in the initial sorting process and reared separately in a tank. All the individuals with this deformity were in six of 89 families represented, which showed a significant association with the deformity. So, five of these six families were selected for QTL mapping. In Experiment 2, fish were obtained from a mass spawning but no fingerling sorting process was carried out. Although at small size, the prevalence of lack of operculum in the Abstract 2 offspring was independent of family or breeder, at 539 days post-hatching, significant relationships between a sire, a dam, and two of the families, and prevalence of this deformity in the offspring were found. In fact, 48.94% of individuals showing this deformity were descendants of this sire. Therefore, all the descendants of this sire (six half-sib families) were selected for QTL mapping. In Experiment 3, designed matings were proposed: sires suffering from lordosis or operculum or vertebral fusion deformities were mated with non-deformed dams. A mass-spawning was considered as a control mating. At 129 days post-hatching, a total of 11503 offspring were visually analyzed for deformities. A significant relationship between the prevalence of each deformity and the mating of breeders suffering from the same deformity was observed. However, a significant prevalence of lack of operculum was also observed in mating with lordotic sire, thus, revealing an important influence of environmental factors on this deformity presence, while lordosis and vertebral fusion deformities are related to family structure. Finally, a search for QTLs for five skeletal deformities in 12 gilthead seabream families was performed. This search was divided into three studies. All the fish were genotyped by using the set of multiplex PCRs ReMsa1-13. In Study 1, five families selected in Experiment 1 were analyzed for LSK complex deformity. 14 LSK-related significant QTLs were found and four of them, located in linkage groups (LG) 5, 8, 17 and 20, were the most solid with an extremely large effect. A strongly association between the genotype of closely located markers of these QTLs and the phenotype was observed. In Study 2, the paternal halfsib family selected in Experiment 2 was analyzed for lack of operculum deformity. Two of the four QTLs detected for this deformity were significant QTLs, and they were located in LG9 and LG10. Both QTLs showed a large effect, and a significant association between lack of operculum deformity and male allelic segregation was observed in one of them. In Study 3, a full-sib family from a mating of deformed breeders was analyzed for different deformities. Abstract 3 Three of the seven QTLs found were significant for vertebral fusion in LG21, for lordosis in LG9 and for jaw deformity in LG13, but they should be confirmed in other families. The results of the present work confirm the genetic origin of these skeletal deformities in gilthead seabream, as well as their relationship with family structure. In addition, QTLs for these deformities were detected, as well as a significant association between allelic segregation of their closely located molecular markers and phenotype. This represents a major step towards the location of genes that determine the presence of skeletal deformities in this species. The above reveals the potential of applying molecular geneticsbased knowledge from identificating specific QTLs, to reduce the prevalence of these deformities in gilthead seabream industry within a genetic breeding program. General Introduction 10 Figure 2. Progression of gilthead seabream aquaculture production in the Mediterranean area and in the rest of the World from 1983 to 2013 (APROMAR, 2014) 1.2.2. Aquaculture Production in Spain The production of gilthead seabream is the highest fish production in Spain in 2013 (16,795 t), followed by the production of rainbow (Oncorhynchus mykiss) (16,732 t) and of European seabass (Dicentrarchus labrax) (14,707 t). Figure 3. Distribution of the gilthead seabream productions in Spain for Autonomous regions in 2013 (%) (APROMAR, 2014) In 2013, the aquaculture production of gilthead seabream in Spain was led by Autonomous region of Valencia (6,974 t (42% of the total production)), followed by Murcia (3,730 t (22%)), the Canary Islands (3,016 t (18%)), Andalusia (1,786 t (11%)) and Catalonia (1,292 t (8%)) (APROMAR, 2014). Production of gilthead seabream in Autonomous regions General Introduction 11 of Spain is shown in Fig. 3; as well as, production in Canary islands from 1998 to 2012 is shown in Fig. 4. Figure 4. Aquaculture production of gilthead seabream in the Canary Islands. Information provided by Ministerio de Agricultura, Alimentación y Medio Ambiente, Secretaría General de Pesca, Junta Nacional Asesora de Cultivos Marinos (JACUMAR, 2014) With respect to juvenile fish production in Spain, it reached 51.4 millions units in 2013, which is 6.5% lower than those produced in 2012. This production of juvenile fish in Spain concentrates on Comunidad Valenciana, the Balearic Islands, Cantabria and Andalusia. Nevertheless, to produce gilthead seabream at commercial size in Spain, it is necessary to import juvenile fish. The first sale value of commercial size gilthead seabream in Spain was 72,165,279 € in 2012 (JACUMAR, 2014). 1.3. Justification of the Study At present, 95.3% of the total amount of gilthead seabream consumed in Spain comes from aquaculture. This notable development of gilthead seabream aquaculture production entailed an increase in their high market competence and, in consequence, a decrease of their final product sale price. So, companies had to minimize costs and to increase their production General Introduction 12 in order to maintain their benefits. In this regard, genetic improvement has been recognized as an important factor to develop a profitable, efficient and sustainable aquaculture (Gjedrem, 2005; Rye et al., 2010). Recently, the project PROGENSA® has developed at national scale a pilot program to improve gilthead seabream aquaculture production in collaboration with on-growing companies and research centers from four autonomous Spanish regions. It has achieved the articulation of the genetic tools and protocol that were needed to establish a genetic breeding program without interfering with the industry performance (Afonso et al., 2012). Within a genetic improvement program, the identification of Quantitative Trait Loci (QTL) is a way to improve the selection efficiency. QTL is a cromosomic region (one or more gens) that determines a quantitative trait. The identification of QTLs also allows the Markers Assisted Selection (MAS), which is based on the linkage disequilibrium of these molecular markers and the phenotype of a specific studied trait in the population (Pérez-Enciso and Toro, 2007). The advantages of MAS selection are notable as compared with the traditional selective breeding, so it is an increasingly important tool in genetic animal breeding (Yue, 2014). To get an efficient QTL detection, it is necessary to have the following: a genetic linkage map with a high level of molecular markers saturation, proper phenotypic data and an adequate genetic statistic model (Lynch and Walsh, 1998). In gilthead seabream, a linkage map based on 204 microsatellite markers and 26 linkage groups is available (Franch et al., 2006). Microsatellite markers are probably the most useful molecular markers for mediumhigh density maps, as they are highly polymorphic and codominant, they are distributed throughout the genome and they can be easily analyzable by the Polymerase Chain Reaction (PCR) (Bouza et al., 2007). Currently, MAS has not played a major role in most of genetic improvement programs in aquaculture industry. However, a wide number of studies have revealed the existence of QTLs for the most interesting traits in different species (reviewed by Yue, 2014). To search General Introduction 13 and detect QTLs, it is necessary a high number of markers, in order to cover as many loci as possible, and a high number of genotyped animals. An efficient way to reduce reaction costs and to minimize genotyping errors is to use multiplex PCR (Navarro et al., 2008). In gilthead seabream, it has also been reported a number of QTLs that are related with commercially important traits, such as growth, sex determination, several morphometric traits, resistance to pasteurellosis and confinement stress; these studies were conducted by microsatellite genotyping (Massault et al., 2010; Boulton et al., 2011; Loukovitis et al., 2011, 2012 and 2013). Nevertheless, none of these QTLs has been determined by multiplex PCR, as there is not any gilthead-seabream-specific multiplex PCR considering a high number of microsatellite markers located into the genetic map and adequate to identify QTLs. On the other hand, growth and deformities are the most economically important trait for the industrial production of gilthead seabream (Georgakopoulou et al., 2010). In gilthead seabream industry, in which products are mainly commercialized as whole fish, deformed fish affect negatively to the turnover of hatcheries and on-growing companies. In fact, this industry invests heavily in reducing the prevalence of deformities, and they discard deformed fish at different steps of the production cycle (Afonso and Roo, 2007). In aquaculture, deformities altering the fish appearance are considered the most important, since they affect directly to their production traits (Gjerde et al., 2005; Kause et al., 2005; Afonso and Roo, 2007; Fjelldal et al., 2009). Among them, skeletal deformities, such as neurocranium or head, vertebral column and appendicular skeleton, are the most relevant deformities. Head deformities include those affecting jaw and operculum complex (lack of operculum). Scoliosis, kyphosis, lordosis and vertebral fusion are the most frequent vertebral column anomalies (reviewed by Boglione et al., 2013). A high number of studies has determined that physiological, environmental, xenobiotic and nutritional factors are linked to fish deformities appearance (reviewed by Bardon et al., 2009), but just a few of them contributed to their genetic determination. In gilthead seabream, Afonso et al. (2000) found a General Introduction 14 significant family association for a triple column deformity (LSK complex, lordosis-scoliosiskyphosis), Astorga et al. (2004) estimated the heritability of any type of deformity, and LeeMontero et al. (2014) estimated the heritabilities of lack of operculum and of vertebral deformity. These studies suggest that skeleton deformities variability are influenced by the environment but could be also explained by a genetic origin. In this regard, a MAS selection would be an efficient tool to optimize the genetic improvement, being especially profitable for traits that are difficult to measure on offspring, that exhibit low heritability and/or that are expressed later in development (reviewed by Yue, 2014), such as skeletal deformities. However, there is no QTL described for any type of deformity in gilthead seabream. 1.4. Objectives of the Study In this context, the main objective of this study was to analyze and to determine the existence of QTLs affecting the most frequent and important skeletal deformities in gilthead seabream, which could be useful to minimize these deformities prevalence when being considered within a specific genetic breeding program. To get that objective, the following secondary objectives were established: 1To develop a set of multiplex PCRs containing the highest number of gilthead seabream specific microsatellite markers, to cover the highest proportion of linkage groups of the genetic map of this species, so it could be used as an efficient tool to identify QTLs. As well as to evaluate and to validate its proper performance. 2To analyze the prevalence of four skeletal deformities (lack of operculum, lordosis, vertebral fussion and LSK complex), and to determine if they show a correlation with the family structure, in different gilthead seabream populations. In order to confirm the genetic origin of these deformities and to identify a family structure that could facilitate the search for QTLs. General Introduction 15 3To genotype, by using the set of multiplex PCRs developed, the families showing a significant association with the analyzed deformities to identify and localize QTLs associated to these deformities. 2. A set of 13 multiplex PCRs of specific microsatellite markers as a tool for QTL detection in gilthead seabream ( Sparus aurata L.) Negrín-Báez et al. (2014a). Aquaculture Research,1-14. doi:10.1111/are.12378 Oral comunication. 4º Congreso Nacional de Acuicultura. Enero, 2013. Puerto Montt. Chile Poster. Aquaculture International Conference. Noviembre de 2013. Gran Canaria. España. 19 2.1. Abstract The growth and consolidation of the gilthead seabream (Sparus aurata L.) industry require improvement based on permanent and cumulative aspects, such as those derived from genetic breeding programs. Marker Assisted Selection (MAS) by Quantitative Trait Loci (QTL) can be usefully implemented with the appropriate tools. In this study, 138 microsatellite markers from the genetic map of gilthead seabream were redesigned to be amplified under the same PCR conditions. A final set of 13 multiplex PCRs (named ReMsa) with 106 of these markers was developed to cover 100% of the linkage groups. These effective multiplex PCRs enable to optimize QTL searching with a critical reduction in costs and errors. Results showed that the mean value of the number of alleles for 106 markers was 6.9. The mean observed heterozygosity ranged from 0.53 to 0.86, and 74.5% of markers were highly informative according to their polymorphic information content. The correct inheritance and segregation of alleles of each locus was confirmed after genotyping 62 individuals of a full-sib family by these multiplex PCRs. Additionally, genetic features of the 20 microsatellite markers that worked correctly but were not included in any multiplex PCR are also reported to provide geneticists with the possibility of including them in more comprehensive screening studies. Material and Methods 26 The chosen amplicon length ranged from 70 to 200 base pairs (bp), the existence of flanking DNA sequences free from interfering polymorphisms was considered, GC content stood at 35–60% and hairpin or secondary structures were avoided using the software found on-line at Integrated DNA Technologies (http://www.idtdna.com/Scitools/Applications/ mFold/). Primers were named with the same internal code used in Lee-Montero et al. (2013) in order to facilitate understanding of this work. PCR reaction and microsatellites evaluation Each microsatellite marker was tested in a single PCR using the control samples in order to check its correct amplification, allele size range and banding pattern. The PCR conditions consisted of an initial denaturation at 95°C for 10 min, followed by 28 cycles of 94°C for 30s, 60°C for 1min and 65°C for 1 min, with a final extension of 65°C for 60 min. Reactions were carried out in a final volume of 12.5μl with the following component concentrations: 1X GeneAmp PCR Buffer II (100mM Tris–HCl pH 8.3, 500mM KCl) (Applied Biosystem®, USA), 3mM MgCl2, 0.2mM of each dNTP, 0.04 units μl−1 AmpliTaq Gold DNA polymerase (Applied Biosystem®), 2.4-6.4 ng·μl−1 of DNA template, and 0.2μM of each primer. Before running PCR fragments in the automatic sequencer, an aliquot was checked on 2% agarose gel for 30 min (8V·cm−1) to test the appropriate amplification of each marker. Later, 1μl of each reaction product (diluted at 75% with Milli-Q water) was mixed with 9.75 μl of Hi-Di formamide and 0.25 μl of GeneScan LIZ 500 (Applied Biosystem®, USA) molecular weight marker, and loaded in an ABI Prism® 3130XL automatic sequencer (Applied Biosystem®, USA) with 16 capillaries using POP-7 polymer and run conditions of 60°C, 3000 V, 1500 s. The reaction products of each single PCR assay were initially loaded in groups of three markers taking into account their fluorochromes, allelic ranges and the intensity of the band on agarose gel. Electropherograms were analyzed using the GeneMapper (v.3.7) software (Applied Biosystem®, USA). The genotyping reliability of each microsatellite was evaluated according to the GremmProtocol (Lee-Montero et al., 2013). This method Development of a set of 13 multiplex PCRs 27 classifies genotyping errors or potential errors into four types, according to their peak scoring (inadequate amplification, long allele dropout and unclear banding pattern) and allele calling (intermediate alleles). Since, the reliability of each microsatellite genotyping is evaluated by quantification of these errors or potential errors in samples of each study and scored as 1 (ambiguous genotypes in more than 30% of the samples), 2 (ambiguous genotypes in 30% or less of the samples) or 3 (unambiguous genotypes in 100% of the samples). Multiplex PCRs optimization Monomorphic markers or markers without amplification were discarded from the multiplex reactions. For the remaining markers, the maximum number of multiplex PCRs were designed including the higher possible number of markers in each one taking into account their allele size range and assigned labelling fluorochrome (in order to avoid overlapping). These multiplex PCRs were named ReMsa (Redesigned Multiplex Sparus aurata) Design and optimization were performed for each multiplex reaction sequentially one by one in such a way that when the amplification of a marker failed in one multiplex PCR, this marker was included in the next one. The initial concentration of each primer was 0.2 μM for each multiplex PCR and it was subsequently modified to obtain peak heights between 600 and 3000 RFU for each microsatellite marker as described Navarro et al. (2008). Each set-up was performed on the testing samples for each multiplex reaction. PCR conditions were the same as those used in the initial single PCR. Again, an aliquot was checked on 2% agarose gel for 30 min (8V·cm−1), to verify the appropriate amplification of the multiplex amplicons. The automatic sequencer conditions were also the same as those used in the single PCR. The reaction product was not diluted, except in ReMsa1, ReMsa6 and ReMsa8 in which 50% of PCR product dilution was needed in order to improve the peak scoring. Electropherograms were analyzed using the GeneMapper (v.3.7) software and a kit of bin set was created for each multiplex PCR. Material and Methods 28 The observed heterozygosity (Ho), expected heterozygosity (He) and the Polymorphic Information Content (PIC) of each microsatellite marker, and the Combined Non-Exclusion Probability for the parent pair and second parent options (Jamieson and Taylor 1997) of each multiplex assay were estimated using Cervus (v 3.0.3) (Kalinowski et al., 2007). Markers were sorted according to their PIC value as highly informative (PIC≥ 0.5), reasonably informative (0.5>PIC>0.25) and slightly informative (PIC≤0.25) (Botstein et al., 1980). 2.3.3. Multiplex PCRs validation and genotyping reliability revaluation Validation samples were genotyped with each multiplex PCR optimized in this study. These genotypes were used to analyze the inheritance and segregation of alleles. Offspring genotypes were checked to verify that one allele matched one breeder and another allele matched with another breeder. False homozygotes in one parent and several descendents allowed to identify null alleles. Moreover, these genotypes were also used to check the values of the genotyping reliability of each microsatellite according to the GremmProtocol (LeeMontero et al., 2013). Figure 5. Genetic distances between markers used in this study based in 26 linkage groups from gilthead seabream genetic map described by Franch et al., (2006) 29 2.4. Results 2.4.1. Multiplex PCRs After initial evaluation, 10 of the 138 microsatellite markers (A3, B9, B11, C5, D8, E7, H3, J4, J12, K12) did not work individually and two were monomorphic (D3, M1). Of the remaining 126 markers, 106 markers were combined in 13 final multiplex PCRs named ReMsa1ReMsa 13. When a marker failed to amplify in a multiplex PCR reaction, it was included in another one, thus the multiplex PCR design and optimization were performed sequentially. Only two markers, C3 and D10, did not work in the first multiplex PCR in which they were tested (ReMsa3 and ReMsa5, respectively), but they worked in another multiplex PCR (ReMsa13). The remaining 20 markers, named alternative microsatellite markers, were dismissed from multiplex PCR assays because of their allele size range or labelling fluorochrome (B1, B5, D6, E5, E8, F5, H1, H5, H6, I3, I4, I6, K3, K8 and L3), or because they did not work in any of the multiplex PCR in which they were tested (A6, A9, B10, D2 and I5). The final number of markers of each multiplex PCR and their concentration, observed and expected heterozygosity, polymorphism information content and the genotyping reliability evaluation of all markers are shown in Table 2. The number of markers in the 13 multiplex PCRs ranged from 6 to 11. The mean value of number of alleles for these markers was 6.9 (from 2 to 21). Ho ranged from 0.125 to 0.999, and He ranged from 0.121 to 0.978. Most markers (74.5%) were highly informative according to their PIC value, 21.7% were reasonably informative and 3.8% were slightly informative. Mean values and combined nonexclusion probabilities of each multiplex PCR are shown in Table 3. Out of 1241.9 cM from the first genetic linkage map of gilthead seabream, this set of multiplex covers a total of 990.1 cM. The genetic distances between markers are showed in Fig. 5. Results 30 Table 2. Genetic information of each microsatellite marker of 13 multiplex PCRs (ReMsa) and alternative microsatellite markers on 16 gilthead seabream samples (control samples): Internal code (IC), locus, fluorochrome (Fl), linkage group (LG), type of nucleotide motif, number of alleles (NA), allele size range (bp), primer concentration (µM), observed (Ho) and expected heterozygosity (He), polymorphism information content (PIC) and genotyping reliability (GR) revaluated by GremmProtocol IC Fl LG Nucleotide motif NA bp µM Ho He PIC GR ReMsa1 A1 6-FAM 9 (CA)10 5 68-78 0.1 0.571 0.529 0.483 3 A2 6-FAM 10 (CA)12 2 82-98 0.1 0.25 0.484 0.359 3 H2 6-FAM 18 (GT)28 15 106-190 0.1 0.733 0.947 0.909 3 A4 NED 21 (GT)12 3 87-97 0.1 0.5 0.524 0.428 3 A5 NED 20 (CA)15 4 104-110 0.03 0.813 0.696 0.608 3 M5 NED 24 (TAGA)15 13 134-204 0.1 0.999 0.897 0.855 3 A7 PET 4 (CT)13 3 86-92 0.1 0.2 0.297 0.26 3 A8 PET 8 (AC)30 10 94-118 0.1 0.733 0.89 0.844 3 A10 VIC 16 (CA)16 3 83-93 0.1 0.5 0.599 0.489 3 A12 VIC 5 (TG)15 10 128-172 0.1 0.875 0.851 0.804 3 ReMsa2 B2 6-FAM 9 (AC)12 2 103-105 0.04 0.143 0.138 0.124 3 B3 6-FAM 17 (CT)34 7 116-139 0.04 0.467 0.589 0.544 3 M2 6-FAM 18 (CA)32 14 177-263 0.1 0.8 0.938 0.899 3* B4 NED 7 (GT)15 13 78-111 0.03 0.999 0.94 0.902 2 B6 NED 20 (AGAT)13 12 138-200 0.09 0.999 0.921 0.877 3 B8 PET 4 (CA)27 8 101-137 0.1 0.667 0.8 0.62 3 F7 VIC 25 (TG)15 6 96-108 0.08 0.583 0.66 0.643 3 H9 VIC 22 (GAGG)6 (AC)2 (TG)2 11 140-186 0.1 0.938 0.883 0.84 3* ReMsa3 C1 6-FAM 9 (CA)12 7 75-103 0.1 0.375 0.819 0.763 3 C4 6-FAM 17 (CA)41 8 140-186 0.1 0.714 0.653 0.601 3 C7 NED 11 (CA)30 6 105-117 0.15 0.933 0.805 0.743 1 C8 PET 12 (TG)13 2 88-96 0.05 0.125 0.121 0.11 3 C10 PET 23 (AC)21 (CT)7 9 127-157 0.15 0.8 0.869 0.821 3 C11 VIC 16 (CTG)10 2 84-90 0.04 0.125 0.508 0.371 3 C12 VIC 18 (CT)19 6 119-133 0.08 0.813 0.722 0.652 3 K5 VIC 11 (AC)14 4 97-102 0.06 0.533 0.561 0.495 2 ReMsa4 D1 6-FAM 8 (GCT)7 5 51-63 0.04 0.75 0.627 0.557 2 J2 6-FAM 22 (CA)21 9 89-115 0.07 0.8 0.857 0.807 3 D5 NED 2 (CA)8 5 70-88 0.07 0.313 0.383 0.348 3 D7 NED 4 (GT)35 19 102-171 0.1 0.875 0.966 0.932 2 D9 PET 3 (GT)14 (GA)4 7 115-140 0.07 0.75 0.756 0.687 3 D11 VIC 5 (GT)16 8 73-93 0.03 0.938 0.869 0.823 3 J6 VIC 11 (CA)8 2 132-136 0.03 0.25 0.315 0.258 1 I7 VIC 24 (GT)17 7 97-129 0.09 0.733 0.731 0.686 3 ReMsa5 E1 6-FAM 23 (GT)10 5 77 -85 0.05 0.733 0.731 0.649 3 E2 6-FAM 6 (AC)13 3 90 -101 0.07 0.333 0.301 0.271 3 E3 6-FAM 10 (TG)30 12 129166 0.1 0.867 0.91 0.869 3 E4 NED 11 (AC)13 10 63 -90 0.08 0.867 0.867 0.819 3 E6 NED 7 (AC)41 21 99 -197 0.1 0.929 0.976 0.938 3 H7 PET 1 (GT)23 4 123129 0.08 0.625 0.732 0.654 3 E9 VIC 12 (GT)13 9 84105 0.04 0.733 0.876 0.83 3 E10 VIC 25 (ATA)9 3 114 -119 0.07 0.533 0.522 0.407 3 Development of a set of 13 multiplex PCRs 31 Table 2. Continued M4 VIC 8 (GT)31 5 124134 0.15 0.571 0.741 0.671 3 ReMsa6 F1 6-FAM 5 (TATC)13 12 80-138 0.06 0.833 0.929 0.892 3 D4 6-FAM 1 (TG)25 7 183205 0.15 0.8 0.816 0.76 3 F3 NED 14 (TAGA)10 9 98-134 0.05 0.923 0.887 0.863 3 F4 PET 14 (AC)14 7 72-86 0.08 0.667 0.763 0.701 3 F6 PET 3 (GT)15 6 124-142 0.09 0.938 0.7 0.506 3 B7 PET 10 (GT)11 6 91-113 0.09 0.5 0.534 0.493 3 F8 VIC 11 (AC)30 5 105-147 0.1 0.400 0.667 0.665 1 L5 VIC 6 (CA)13 2 89-92 0.03 0.5 0.508 0.371 3 ReMsa7 L2 6-FAM 12 (TG)10 5 8795 0.07 0.857 0.738 0.667 1* G2 6-FAM 13 (CA)15 8 113-135 0.1 0.7 0.774 0.739 3 G3 6-FAM 26 (TG)19 8 150-172 0.08 0.867 0.807 0.736 3 G4 NED 15 (AC)17 6 83-93 0.06 0.909 0.814 0.768 3 G5 NED 24 (AC)22 8 113-147 0.1 0.800 0.874 0.808 3 G7 PET 14 (CA)11 8 84-112 0.1 0.999 0.848 0.785 3 G8 VIC 9 (CA)14 4 80-92 0.04 0.818 0.749 0.649 2 G9 VIC 1 (CA)15 8 106-124 0.05 0.917 0.790 0.716 3 ReMsa8 J1 6-FAM 5 (GT)14 7 74-98 0.06 0.938 0.806 0.748 3 J3 6-FAM 5 (TG)30 16 112-178 0.1 0.938 0.95 0.914 3 J5 VIC 9 (TG)18 10 107-160 0.07 0.875 0.827 0.782 1 J7 NED 21 (GT)12 3 91-97 0.05 0.188 0.28 0.248 3 J8 NED 15 (CA)21 5 104-120 0.07 0.625 0.613 0.554 3 J9 NED 1 (GT)43 12 126-159 0.1 0.999 0.885 0.841 3 J10 PET 12 (AC)15 5 91-104 0.08 0.625 0.732 0.656 3 J11 PET 2 (TG)18 8 116-141 0.09 0.813 0.796 0.742 3 ReMsa9 K1 6-FAM 3 (CA)12 5 72-104 0.06 0.5 0.468 0.422 2 F2 6-FAM 13 (CA)31 9 108-148 0.09 0.636 0.797 0.734 3 K4 VIC 25 (TG)10 5 79-95 0.08 0.75 0.645 0.57 3 K6 VIC 14 (CA)20 8 100-126 0.06 0.875 0.835 0.787 3 K7 NED 7 (AC)12 8 79-101 0.06 0.688 0.813 0.759 3 K9 NED 3 (CA)25 10 106-134 0.09 0.75 0.887 0.844 1 K10 PET 19 (CA)14 5 89-103 0.08 0.563 0.724 0.654 2 K11 PET 17 (TG)17 5 111-127 0.08 0.5 0.732 0.656 2 ReMsa10 L1 6-FAM 5 (GT)11 3 71-86 0.04 0.563 0.567 0.452 3 C2 6-FAM 1 (CA)13 7 99-131 0.08 0.867 0.749 0.683 2* L7 NED 16 (CA)11 3 74-78 0.04 0.75 0.534 0.412 2 L8 NED 8 (TG)13 4 89-108 0.07 0.563 0.542 0.486 3 L9 NED 20 (GT)15 3 118-127 0.07 0.563 0.575 0.457 3 L10 PET 9 (AC)12 3 76-86 0.07 0.625 0.462 0.371 3 L11 PET 10 (GT)14 5 101-111 0.08 0.75 0.669 0.606 3 L12 PET 25 (TG)21 5 121-138 0.08 0.375 0.339 0.313 3 L4 VIC 12 (AG)13 2 76-86 0.05 0 0.138 0.124 3 L6 VIC 4 (CA)22 2 96-121 0.07 0.25 0.315 0.258 3 ReMsa11 G1 6-FAM 17 (TG)15 13 82152 0,9 0,947 0,89 0.893 3 M6 NED 11 (AC)12 3 9498 0.09 0.4 0.628 0.539 3 M7 NED 18 (CA)22 8 114140 0.09 0.8 0.766 0.703 3 M8 PET 13 (TG)13 8 6898 0.07 0.133 0.798 0.746 3 Results 32 Table 2. Continued M9 PET 24 (AC)20 5 112120 0.08 0.8 0.752 0.678 3 M3 VIC 7 (CA)16 6 8193 0.07 0.667 0.68 0.601 3 A11 VIC 25 (CA)23 (GA)5 11 97135 0.06 0.75 0.913 0.874 3 I9 VIC 19 (TAGA)13 5 153169 0.06 0.5 0.623 0.551 3 ReMsa12 K2 6-FAM 3 (GT)14 8 86-117 0.06 0.875 0.855 0.806 3 I2 6-FAM 20 (GT)18 12 121-161 0.1 0.938 0.923 0.885 3 C6 NED 8 (CA)15 4 97-103 0.04 0.313 0.421 0.375 3 C9 PET 7 (GT)16 5 109-123 0.1 0.4 0.687 0.603 3 D12 VIC 23 (AC)27 10 88-119 0.09 0.999 0.887 0.842 3 I8 VIC 3 (GT)24 7 126-142 0.09 0.938 0.754 0.697 3 ReMsa13 I1 6-FAM 23 (TGG)3 (TG)9 3 96-102 0.05 0.188 0.373 0.327 3 C3 6-FAM 6 (TG)27 13 106-140 0.08 0.933 0.92 0.879 3 H4 NED 17 (CA)15 6 92-104 0.1 0.875 0.758 0.692 3 G6 PET 18 (AC)17 6 84-112 0.15 0.727 0.771 0.694 3 D10 PET 18 (CA)40 8 126-155 0.15 0.563 0.786 0.734 3 H8 VIC 19 (TG)16 3 108-116 0.05 0.875 0.677 0.582 3 D13 VIC 10 (AC)29 11 123-172 0.15 0.6 0.897 0.853 3 Alternative microsatellite markers A 6 - 12 (CA)27 8 107 - 133 - 0.214 0.881 0.832 3 A 9 - 2 (TG)31 8 132 - 152 - 0.667 0.793 0.74 3 B 1 - 2 (CA)16 9 79 - 115 - 0.733 0.816 0.762 3 B 5 - 8 (TCC)12 2 102 - 117 - 0.077 0.471 0.350 3 B 10 - 6 (GT)18 5 97 - 135 - 0.467 0.587 0.388 1 D 6 - 11 (GT)26 11 81 - 112 - 0.999 0.853 0.811 3 D 2 - 24 (CA)19 10 87 - 120 - 0.938 0.897 0.856 3 E 5 - 13 (CA)20 9 90 - 112 - 0.875 0.917 0.844 3 E 8 - 4 (GT)11 6 114 - 139 - 0.364 0.756 0.544 3 F 5 - 6 (GT)13 6 114 - 148 - 0.500 0.742 0.663 3 H 1 - 14 (GT)16 11 81 - 119 - 0.813 0.810 0.767 3 H 5 - 25 (AC)25 7 103 - 121 - 0.600 0.694 0.644 3 H 6 - 26 (GT)17 13 83 - 141 - 0.438 0.915 0.876 3 I 3 - 15 (GT)18 4 101 - 115 - 0.600 0.687 0.597 2 I 4 - 26 (CA)16 6 105 - 119 - 0.867 0.805 0.749 3 I 6 - 7 (CA)24 7 110 - 140 - 0.938 0.859 0.81 3 I 5 - 2 (GT)11 5 102 - 122 - 0.500 0.611 0.551 3 K 3 - 14 (TG)15 7 104 - 134 - 0.400 0.772 0.719 3 K 8 - 3 (CA)26 14 81 - 134 - 0.867 0.931 0.892 3 L 3 - 9 (GT)28 10 73 - 137 - 0.400 0.768 0.714 2 GremmProtocol (Lee-Montero et al., 2013) evaluates the reliability of each microsatellite genotyping by quantification of genotyping errors or potential errors as: 1 (ambiguous genotypes in more than 30% of the samples), 2 (ambiguous genotypes in 30% or less of the samples) and 3 (unambiguous genotypes in 100% of the samples). * New score after validation samples genotyping 2.4.2. Multiplex PCRs validation and genotyping reliability revaluation The correct inheritance and segregation of alleles of each locus between parents and offspring was confirmed after the validation samples were genotyped. The alleles of these samples were within the size range described for the 16 control samples. The microsatellite markers that showed null alleles (allelic frequency in parents: 0.25), detected through familial Development of a set of 13 multiplex PCRs 33 segregation, were C11, E9, I1, K9, M3 and M8. After revaluation by the GremmProtocol, only the scores of the markers C2, H9, L2 and M2 changed. The marker C2 changed from 3 to 2 and the marker L2 from 2 to 1 due to an “intermediate allele” genotyping error found in both markers. The scores of H9 and M2 changed from 2 to 3 because any potential genotyping error was not found in the 62 validation samples. 34 2.5. Discussion Microsatellite-based strategies have been successfully used to identify QTLs of economic interest in many aquaculture species (Wang et al., 2006; Massault et al., 2010; Sánchez-Molano et al., 2011; Norman et al., 2012). Their genome-wide distribution and high levels of allelic polymorphism (Chistiakov et al., 2006) make them greatly used tools in QTL search since, in this case, genotyping many individuals with a high number of loci is required. For this very reason, the use of these markers by multiplex PCR offers a desirable reduction of costs per sample (Navarro et al., 2008). The 13 new multiplex PCRs developed in this study, including a total of 106 microsatellites, constitute the first set of multiplex PCRs to identify QTL in gilthead seabream. Several studies have used microsatellites markers for QTL detection in this species (Massault et al., 2010; Boulton et al., 2011; Loukovitis et al., 2011, 2012, 2103), but none of them have used a set of multiplex PCRs. Two of these studies used a multiplex PCR with 9 microsatellites, but only to determine the family relationship (Boulton et al., 2011; Loukovitis et al., 2011). For QTL searching, these authors used additional markers by single PCR reactions. Other multiplex PCRs have been developed in gilthead seabream for parentage assignment and/or populations genetic studies (Launey et al., 2003; Brown et al., 2005; Navarro et al., 2008; Porta et al., 2010; Borrell et al., 2011; Vogiatzi et al., 2011; LeeMontero et al., 2013), but none have included as many microsatellites as this study. Actually, a genetic linkage map with a high level of saturation of molecular markers is needed for an effective QTL detection (Lynch and Walsh, 1998). The length of the genetic linkage map of gilthead seabream is 1241.9 cM and it presents 26 LG (Franch et al., 2006). Apart from the 22 microsatellites that conform the SMsa1 and SMsa2 (Lee-Montero et al., 2013), the 13 multiplex PCRs developed in this study are the only ones whose microsatellites are all located Development of a set of 13 multiplex PCRs 35 on the genetic map of this species. In fact, these 13 multiplex reactions cover 100% of LG of the genetic map of this species, thus making this a powerful and low-cost tool to identify QTL. Moreover, the 13 multiplex PCRs work under the same conditions, thereby making it possible for several multiplex reactions to be executed in a unique PCR run or exchange markers between assays. Additionally, the 20 alternative microsatellite markers not included in any multiplex PCR offer geneticists the possibility to include them later in further screening, given that their corresponding primers have also been redesigned to be amplified under the same PCR conditions. A further advantage of all the markers included in this study is that the amplicons are small (< 200 bp). These similar sizes should avoid differential amplification of size variants due to the competitive nature of PCR by which short-length alleles often amplify more efficiently than longer ones (Dakin and Avise, 2004), especially when DNA quality is low (Pompanon et al., 2005). On the other hand, QTL searching is more efficient with highly polymorphic markers (Lynch and Walsh, 1998). Most of the multiplex reactions proposed in this study showed high mean values of Ho and He, similar to other multiplex PCRs in gilthead seabream (Navarro et al., 2008; Borrell et al., 2011; Lee-Montero et al., 2013). Likewise, 74.5% of the markers that make up these multiplex reactions are highly informative according to their PIC values. However, in ReMsa3 and ReMsa10, the mean PIC values were somewhat lower. Nevertheless, the objective of this study was to create a set of multiplex PCRs to identify QTL, so the maximum possible number of markers to carry out a comprehensive sweep of the map was included, even though the mean variability values fell. Higher genetic variability values are needed to determine pedigree, but highly efficient multiplexes to this end already exist for this species (Navarro et al., 2008; Borrell et al., 2011; Lee-Montero et al., 2013). In fact, the genetic variability values measured in mean allele numbers found in the microsatellite markers of this study are very similar to others used in QTL search (8.1 for the Abstract 42 family structure, while lack of operculum is related to environmental factors. The results of the present study report that the analyzed deformities prevalence has a genetic origin although it is also influenced by environmental factors. Producers should consider both factors to improve significantly their fish morphological quality and to minimize the incidence of deformities in farmed fish. 43 3.2. Introduction The presence of morphological abnormalities in farmed gilthead seabream (Sparus aurata L.) is currently a major problem in aquaculture as it entails significant economic losses (Afonso and Roo, 2007; reviewed by Bardon et al., 2009). Skeletal deformities are the most relevant deformities and they include head and vertebral column anomalies. Lack of operculum in head, and lordosis, scoliosis, kyphosis and vertebral fusion in column, are the most frequent skeletal anomalies in this species (Galeotti et al., 2000; Beraldo and Canavese, 2011; reviewed by Boglione et al., 2013). They affect the fish appearance and lead to physiological alterations that result in a devaluation of their commercial traits (Gjerde et al., 2005; Karahan et al., 2013). High number of studies has determined that environmental factors are linked to fish deformities (reviewed by Bardon et al., 2009), but only a few of them have contributed to their genetic determination. In gilthead seabream, a significant family association for a triple column deformity (LSK complex, lordosis-scoliosis-kyphosis) was found by Afonso et al. (2000) and a high heritability for presence/absence of any kind of deformities was estimated by Astorga et al. (2004). Lee-Montero et al. (2014) estimated medium value of heritabilities for deformity traits in fish reared in four Spanish regions (PROGENSA®). These results suggest that skeletal deformities variability in a particular population, besides being influenced by the environment, can also be explained by genetic origin. The main objective of the present study was to analyze in gilthead seabream the prevalence of four skeletal deformities (lack of operculum, lordosis, vertebral fusion and LSK complex) associated with the family structure and considering different breeding conditions, phenotype of breeders and offspring handling; in order to study how genetic predisposition and management of breeders can determine the emergence of these deformities in offspring. 44 3.3. Material and Methods Three experiments to assess the prevalence of deformity in the offspring from different broodstocks were carried out at the facilities of the Marine Science and Technology Park of the University of Las Palmas de Gran Canaria (PCTM-ULPGC, Gran Canaria, Spain), which belongs to the Foundation of Science and Technology Park of ULPGC (FCPCT-ULPGC). 3.3.1. Experiment 1 (mass spawning with sorting) Breeders and mating structure. An egg batch was collected by using mass spawning from an industrial hatchery broodstock. This broodstock was composed of 66 non-deformed breeders and the sex ratio of which was 1♀: 2♂. Offspring and rearing conditions. Eggs and larvae were reared at PCTM-ULPGC and under the conditions described by Roo et al. (2009). Fingerlings were sorted by presence vs absence of deformity at 111 days post-hatching (1.9 ± 0.02 g) (mean ± standard error), as performed by companies. Deformed fish were culled with the exception of LSK (lordosis/scoliosis/kyphosis complex) fish that were selected. The percentage of LSK found in the whole batch of fingerlings was 0.22%. All LSK fish were reared at PCTM-ULPGC separately in a 1000 L fiberglass tank. Water temperature ranged from 19.3 ± 0.1 °C in March to 25.0 ± 0.1 °C in September, and values for dissolved oxygen and water flow were 6.0 ± 0.1 ppm and 21 l·min−1, respectively. Commercial feed was provided through self-feeders. With respect to fish considered as normal or without abnormalities, they were taken to the facilities of Playa de Vargas 2001 S.L. Company (PLV2001, Gran Canaria, Spain) at 130 days post-hatching (4.8 ± 0.1 g). The fish at PLV2001 were reared under industrial conditions in a cage and fed with commercial Segregation Analysis of Skeletal Deformities 45 fish feed. Dissolved oxygen in the water had an average value of 7.4 ppm. The on-growing period lasted up to 509 days post-hatching when fish were 419.7 ± 3.1 g. Visual assessment. At the end of the experiment, a sample of 846 fish from PLV2001 was slaughtered and visually analyzed to determine the presence or absence of deformities according to AquaExcel-ATOL (AquaExcel Project, 2013; ATOL: 0000087). Vertebral deformities were assessed, after being filleted, by direct observation of the axial skeleton. Parental assignment. Family relationships between breeders and offspring (846 fish from PLV2001 and 54 LSK fish from PCTM-ULPGC) were determined by the exclusion method using the software VITASSING (v8.2.1) (Vandeputte et al., 2006), after DNA analysis and genetic characterization by using the multiplex PCR SMsa1 (SuperMultiplex Sparus aurata), as described by LeeMontero et al. (2013). 3.3.2. Experiment 2 (mass spawning without sorting) Breeders and mating structure. Egg batches were collected on two consecutive days from a mass spawning of an industrial broodstock of PCTM-ULPGC within the context of the PROGENSA® breeding program (Afonso et al., 2012). This broodstock was composed of 59 non-deformed breeders, the sex ratio of which was 1♀: 1.81♂. Offspring and rearing conditions. Eggs and larvae were cultured at the PCTM-ULPGC facilities as described in Experiment 1. At day 179 post-hatching (17.2 ± 0.2 g), fingerlings were individually tagged with Passive Integrated Transponder (PIT; Trovan Daimler-Benz) by following the tagging protocol described by Navarro et al. (2006) and transported to a cage of CANEXMAR S.L. company (Gran Canaria, Spain). They were reared under intensive conditions and fed with commercial feed. Dissolved oxygen in the water had an average value of 7.4 ppm and water Material and methods 46 temperature ranged from 20.2°C to 24.2°C. No sorting or culling processes were performed during larval rearing or on-growing periods, which lasted up to 689 days post-hatching, when fish were 524.4 ± 12.6 g. Visual assessment. When fish were tagged (179 days post-hatching), they were also observed in order to determine the presence or absence of deformities (initial analysis) (ATOL: 0000087). At the end of the experiment (689 days post-hatching), a sample of 810 fish were slaughtered and visually analyzed (final analysis) to determine the presence or absence of deformities (ATOL: 0000087). Vertebral deformities were directly assessed after being fish filleted like in Experiment 1. Parental assignment. Family relationships between breeders and offspring (810 fish) were determined as described in Experiment 1. 3.3.3. Experiment 3 (designed mating) Breeders and mating structure. Different directed matings were established using gilthead seabream breeders from PCTM-ULPGC facilities, more concretely, 15 deformed sires and 10 normal (N, nondeformed) dams. Deformed fish were classified into three groups: Lack of operculum (O), Lordosis (L) and Vertebral Fusion (VF). Three matings (replica) by each deformity were constituted: NxO, NxL and NxVF. The sex ratio in each tank was one normal dam with two deformed sires (1♀N: 2♂D). Additionally, a broodstock of non-deformed breeders (1♀N: 2♂N) was used to conform a Control mating (C) by mass spawning. Offspring and rearing conditions. Eggs from each mating were separately cultured in PCTM-ULPGC facilities as described in Experiment 1. Fingerlings from each mating were separately reared in 1000 L fiberglass tanks until the end of the experiment (two tanks per type of mating). Rearing conditions were as Segregation Analysis of Skeletal Deformities 47 follow: commercial feed was provided by automatic feeders; water flow was 0.5 l·min−1, dissolved oxygen concentration was 5.9 ± 0.1 ppm and water temperature ranged from 19.3 ± 0.1ºC at the beginning of the experiment to 23 ± 0.1ºC at the end. The experiment ended at 129 days post-hatching (9.6 ± 0.1 g). Visual assessment. At the end of the experiment, all fingerlings from all mating types, including mass spawning, (11503 fish) were visually analyzed to determine the presence or absence of deformities (ATOL: 0000087). 3.3.4. Statistical analysis The identified deformities were: lordosis (L), vertebral fusion (VF), lack of operculum (O) and lordosis/scoliosis/kyphosis complex (LSK). Fish that did not show any type of these deformities were considered as normal (N). The prevalence rate of deformities was calculated as a percentage of deformed descendants with respect to the total of analyzed fish in each group (breeder, family, mating or experiment). The association between any factor (breeder, family, mating or experiment) and deformity, was analyzed by a log linear model by using the statistical software SPSS (PASW Statistics v18). Log linear model gives the significance of any deformity factor (its prevalence [i]) against any biological or functional factor (depending of data [j]), organized under a two-way contingency table, through the normalized values or Z values. Normalized Z values > +1.96 or <-1.96 indicate an excess or defect statistical significance of deformity, respectively, in any family, breeder, mating or experiment. lnƒij= µ + α i+ β j+ αβ ij Where, ln ƒ ij is the expected frequency of each deformity (i) in each family, breeder, mating or experiment (j) considered (ij); µ is the average value of expected frequencies logarithms, α i is Material and methods 48 the effect of the deformity factor (i), β j is the effect of the family, breeder, mating or experiment (j) and αβ ij is the effect due to these factors interaction. In experiments 1 and 2 was analyzed the relationship between each deformity and all families, sires and dams. In experiment 3, was analyzed the relationship between each deformity and all mating, families and tanks. 49 3.4. Results The total deformities prevalence of offspring of the broodstock in the three experiments is showed in Table 4. Table 4. Deformities prevalence (%) in total offpring in each experiment Operculum Lordosis Vertebral fusion LSK Normal N Experiment 1 0.44 5.67 0.44 6 87.44 900 Experiment 2 5.80 1.73 0.12 0 92.35 810 Experiment 3 4.84 0.64 0.63 0.03 93.85 10650 N: Numbers of descendants from each experiment. Experiment 1: Mass spawning with sorting. Experiment 2: mass spawning without sorting. Experiment 3: Only designed mating with deformed breeders. 3.4.1. Experiment 1 Of the total of LSK fish that were selected, only 54 survived at the end of the experiment (75% of mortality). A 100% success was obtained in the parental assignment, i.e. whole offspring (900 individuals) were assigned to 28 breeders (17 dams and 11 sires). A total number of 89 full-sibling families were represented, and only six of them included the total of LSK individuals (54). These families (named msF1-msF6) were composed of six breeders: four dams and two sires (named ms♀1-ms♀4 and ms♂1-ms♂2). Deformities prevalence (with its Z values) in the descendants of these dams and sires compared with the other breeders (mean value of the prevalence), and with respect to each sire, dam and family, are shown in Table 5. Both sires, the three dams and five families showed a statistically significant relationship with deformity due to their high prevalence of LSK fish (Zfamily, LSK> +2; P<0.05). These families had from 28.57 to 66.67 % of LSK deformed descendants. It is remarkable that ms♂1 and ms♂2 were responsible for 75% of vertebral fusion deformed fish, while ms♀2 was responsible for 51.85% of the LSK deformed fish and 25% of the fish with lack of operculum. None of the dams showed any descendant with VF deformity. Results 50 Table 5. Deformities prevalence (%) in descendants of breeders responsible of LSK deformed fish (lordosis/scoliosis/kyphosis complex) with respect to each sire, dam and family of the Experiment 1 Operculum Lordosis Vertebral fusion LSK Normal N Sires ms♂1 0 -0.95 4.52 -1.40 0.65 -0.24 22.58 3.62 72.26-1.03 155 ms♂2 0.94-1.55 4.72 -0.52 1.89 -0.80 17.92 4.72 74.53 -1.85 106 Mean ms♂R 0.34 6.77 0.10 0 92.79 639 Dams ms♀1 0 -0.93 3.15 -0.90 0 -0,59 11.81 3.75 85.04 -1.34 127 ms♀2 1.20 -0.14 3.61 -1.04 0 -1,33 33.73 5.37 61.45 -2.86 83 ms♀3 0 -0.37 7.50 -0.34 0 0,11 5 1.06 87.50 -0.47 40 ms♀4 0 -0.19 0 -0.16 0 -1,47 27.27 3.83 72.73 -2.01 33 Mean ms♀R 0.21 13.29 0.43 0 86.07 617 Families msF1 (♀1♂1) 0 -0.62 0-1.00 0-0.62 48.15 4.42 51.85 -2.18 27 msF2 (♀1♂2) 0 -0.18 0-0.55 0-0.18 28.57 1.78 71.43 -0.87 7 msF3 (♀2♂1) 0 -0.82 3.33-0.41 0-0.82 66.67 5.43 30 -3.38 30 msF4 (♀2♂2) 0 -0.63 6.250.12 0-0.63 50 3.73 43.75 -2.36 16 msF5 (♀3♂1) 0 0.04 0-0.33 0 0.04 66.67 2.10 33.33 -1.85 3 msF6 (♀4♂2) 0 -0.45 0-0.83 0-0.45 56.25 4.01 43.75 -2.28 16 Mean msFR 0.95 6.75 0.64 0 91.66 801 Z values subscript. Significant associations when Z≥+1.96; Z≤-1.96 (Bold characters). N: Number of descendants. R: Mean of prevalence from the rest of males, females and families 3.4.2. Experiment 2 100% success was also obtained in the parental assignment, i.e. whole offspring (810 individuals) were assigned to 45 breeders (19 dams and 26 sires) and 66 full-sibling families were formed. In the initial analysis at 179 days post-hatching, 7% of offspring showed lack of operculum, and neither lordosis nor vertebral fusion deformities were observed. Deformity prevalence was independent of family or breeder (-1.96<Zbreeder/family,deformity<+1.96; P>0.05). At 689 days, at the final analysis, the total prevalence of deformities was 7.65%. Only the operculum deformity prevalence showed a significant statistically association (Z>+2, P<0.05) with sires, dams and the families formed by these breeders. 48.94% of individuals with lack of operculum were descendant of one sire (named mw♂1), which represents a significant Segregation Analysis of Skeletal Deformities 51 prevalence. Only two families, the formed by mw♂1 and two dams (named mw♀1 and mw♀3), showed a significant prevalence of operculum deformity, as they represented 12.77 and 21.28% of all the individuals, respectively. The deformities prevalence in offspring from mw♂1 at final analysis is shown in Table 6. Table 6. Deformities prevalence (%) in offspring from mw♂1 of the Experiment 2 Operculum Lordosis Vertebral fusion Normal N Sires mw♂1 14.84 3.24 0 -1.59 0 -1.27 85.16 -0.37 155 Mean mw♂R 6.15 3.13 0.02 90.70 655 Dams mw♀1 17.54 2.01 1.75 -0.50 0 -0.89 80.70 -0.62 57 mw♀2 3.66 0.47 0 -0.95 0 -0.48 96.34 0.97 82 mw♀3 4.62 0.59 1.54 -0.52 0.31 -1.56 93.54 1.50 325 mw♀4 4.69 0.16 1.56 -0.21 0 -0.69 93.75 0.73 64 mw♀5 2.08 -0.66 2.08 0.18 0 -0.42 95.83 0.90 48 mw♀6 4.59 0.17 2.75 0.44 0 -1.13 92.66 0.52 109 Mean mw♀R 10.97 4.14 0 84.89 125 Families mwF1(♀1♂1) 46.15 2.57 0 -0.66 0 -0.51 53.85 -1.41 13 mwF2(♀2♂1) 20 0.69 0 -0.22 0 -0.06 80 -0.41 5 mwF3(♀3♂1) 11.63 2.34 0 -1.29 0 -1.14 88.37 0.10 86 mwF4(♀4♂1) 7.69 0.38 0 -0.44 0 -0.29 92.31 0.36 13 mwF5(♀5♂1) 12.50 0.54 0 -0.33 0 -0.18 87.50 -0.03 8 mwF6(♀6♂1) 13.33 1.51 0 -0.86 0 -0.71 86.67 0.06 30 Mean mwFR 6.23 2.40 0.01 91.35 655 Z values subscript. Significant associations when Z≥+1.96; Z≤ -1.96 (Bold characters). N: Number of descendants. R: Mean of prevalence from the rest of males, females and families 3.4.3. Experiment 3 After three replicas of each type of mating, the following were viable: the three NxO, two NxL and only one NxVF. High mortality was observed in an NxFV and an NxL matings, so offspring from these matings could not be analyzed. From 11503 analyzed fish, 6.15% showed any of the analyzed deformities. The prevalence (with its Z values) of each type of deformity in offspring from different matings is shown in Table 7. As it can be observed, the Discussion 58 diagnosed at early ages, so that hatcheries can detect and eliminate them before the fish sale to on-growing companies, entailing economic losses. In addition, as it has been observed in this study, mortality of these individuals is high even when they are separated from nondeformed fish, so, it can be deduced that these individuals do not reach their final on-growing stage. In Experiment 1, by using non-deformed breeders and culling deformed individuals, the prevalence of deformities at the final point was 12.55%, more concretely; column deformities prevalences were higher than operculum deformities prevalence. Similarly, Oliva (2008) observed, in an industrial batch of 430 g gilthead seabream, a 9.80% of the fish with vertebral deformity and a 0.8% of head and operculum deformities. Contrastly, in Experiment 2, despite no sorting process was carried out, the column deformity prevalence was lower than in Experiment 1. It could due to differences among batches, which has already been found by several authors (Witten et al., 2009; reviewed by Boglione et al., 2013). This highlights that the sorting process is not an efficient method to cull deformed fish from the production. With respect to Experiment 3, the deformities prevalence of descendants from deformed breeders was significantly higher than in control mating descendants and than in Experiment 2 at the initial size. In fact, in Experiment 3, the presence of fish showing column deformity was easily detected even at an early age, possibly due to a higher intensity or severity of that deformity (reviewed by Bardon et al., 2009). Differently, in Experiments 1 and 2, deformed individuals were not detected until they reached a higher size. Conclusion The results of this study confirm that an important part of the phenotypic variation depends on non-genetic or environmental factors and suggest that producers should control these environmental factors in order to minimize the prevalence of deformities. They also show that the prevalence of the studied deformities (lack of operculum, lordosis, vertebral fusion and LSK complex) significantly correlates with the family structure in the three studied populations. So, to improve significantly the fish morphological quality, it is essential to Segregation Analysis of Skeletal Deformities 59 know the breeders genetic predisposition to produce deformed descendants, as well as to management optimally breeders during their spawning process. 4. Quantitative Trait Loci for Skeletal Deformities in Gilthead Seabream ( Sparus aurata , L.); Lack of Operculum, Lordosis, Vertebral Fusion and LSK complex Negrín-Báez et al. (2014c). Genetics. In preparation. 63 4.1. Abstract Morphological abnormalities, especially skeletal deformities, are some of the most important problems affecting aquaculture industry. In this work, a QTL analysis for five skeletal deformities in gilthead seabream (Sparus aurata L.) is reported. A total of three studies were conducted. All the offspring analyzed and their parents were genotyped by using a set of multiplex PCRs (ReMsa1-13), which includes 106 microsatellite markers, and a linear regression methodology by the GridQTL software was used to perform the QTL analysis. In Study 1, 78 offspring individuals from five families and six breeders (five full-sib families, three maternal and two paternal half-sib families) were analyzed. They had showed a significant association with prevalence of LSK complex (lordosis-scoliosis-kyphosis) deformity in a previous segregation analysis from an industrial mass-spawning. A total of 14 QTLs were identified for this deformity. Four of them (QTLSK2, 5, 10 and 12), located in LG5, 8, 17 and 20, respectively, were considered the most solid ones. They were significant at genome level and showed an extremely large effect (>35%), and the genotype of their closely located markers showed a strongly association with their phenotype. Five of these molecular markers (BId-39-T, P96, Gt57, DId-03-T and Bt-14-F) were considered as potential markers linked to this deformity. In Study 2, 142 individuals of six half-sib families from a unique male, which had showed a high significant association with prevalence of lack of operculum in a previous segregation analysis, were selected for QTL mapping. Two of the four QTLs detected for this deformity were significant QTLs (QTLOP1 and QTLOP2) and they were located in LG9 and LG10, respectively. Both QTLs showed a large effect (about 25%), and a significant association between lack of operculum deformity and male allelic segregation was observed in the QTLOP1. Their confidence intervals were very large, so a finer mapping would be required to more exactly establish the QTL positions. In Study 3, a full-sib family Abstract 64 with 152 descendants from a mating of deformed breeders (lordosis) was analyzed. Seven QTLs were identified for different deformities. Three of them were significant: one for vertebral fusion (QTLFV3) in LG21, one for lordosis (QTLORD1) in LG9 and one for jaw deformity (QTLJW1) in LG13. Close markers analyzed for QTLFV3 (Hd-46-T) and QTLJW1 (CId-26-H and CId-03-F) showed a significant association between a breeder allele and the phenotype. These QTLs were detected in only one family, so they should be confirmed in other gilthead seabream families. This work reports, for the first time in gilthead seabream, the identification of QTLs related with skeletal deformities, which supposes a critical step in Marker Assisted Selection implementation in this species. . 65 4.2. Introduction In aquaculture industry, the presence of morphological abnormalities in fish entails important economic losses. In gilthead seabream (Sparus aurata L.) industry, in which products are mainly commercialized as whole fish, deformed fish affect negatively to the turnover of hatcheries and on-growing companies. Because of this, the incidence of anomalies in juveniles from hatchery companies, which ranges currently from 15 to 50%, has to be reduced to 5% prior to batch commercialization (Afonso and Roo, 2007; Prestinicola et al., 2013). This reduction requires individual manual sorting and visual assessment, which introduces an additional cost of about 10% per sold health fingerling (Divanach et al., 1996). However, many abnormalities are detected later and they persist in fish at harvest size. In this way, on-growing companies have to remove deformed fish prior to its commercialization or to sell them at below their production costs, as customers rarely accept fish showing malformations (reviewed by Bardon et al., 2009). The deformities altering the fish appearance are considered the most important, since they affect directly to their production traits (Afonso and Roo, 2007). Among them, skeletal deformities, such as neurocranium or head, vertebral column and appendicular skeleton, are the most relevant deformities. Head deformities include those affecting jaw and operculum complex (lack of operculum). Jaw abnormalities consist of torsion of the upper or the lower jaw or extension of this in different magnitud, that have been associated with lethal effects (Afonso and Roo, 2007). The lack of operculum is the most common external abnormality in gilthead seabream, affecting up to 80% of the reared population (Galeotti et al., 2000). Although this abnormality does not affect directly to the growth traits, it has been related with lower fish resistance to environment stress and higher fish predisposition to bacterial infections of gills. The lack of operculum damages the final image of the product because of Introduction 66 the exposure of the gills and, in consequence, decreases its commercial value (Beraldo and Canavese, 2011). Lordosis, scoliosis, kyphosis and vertebral fusion are the most frequent vertebral column anomalies (Afonso et al., 2000; reviewed by Boglione et al., 2013), which not only affect the fish appearance, but they lead to physiological alterations that result in a decrease of their commercial traits value: a lower growth rate, a higher mortality during handling and an increased difficulty of filleting. Furthermore, the effect of these anomalies in animal welfare must be also considered (Gjerde et al., 2005; Karahan et al., 2013). A high number of studies has determined that physiological, environmental, xenobiotic and nutritional factors are linked to fish deformities appearance (Bardon et al., 2009), but an increasing number of studies also consider a genetic origin. In gilthead seabream, a significant association was found between a family and the prevalence of LSK complex (lordosis-scoliosis-kyphosis) by Afonso et al. (2000). Other significant associations were also described by Negrín-Báez et al. (2014b) between families or breeders and the prevalence of lordosis, vertebral fusion, lack of operculum and LSK complex under different breeding conditions, phenotype of breeders and offspring handling. In the context of PROGENSA® (a Spanish Breeding Program at national scale), heritabilities from 0.06 to 0.11 for lack of operculum and from 0.16 to 0.41 for any type of vertebral deformity were estimated in fish reared under different production systems in four Spanish regions (LeeMontero et al., 2014). All this suggests that the prevalence of skeletal deformities in this species is determined, not only by environmental factors but also by genetic factors. The development of genetic tools has allowed researchers to analyze and identify genomic regions that are associated with a specific trait of interest (QTL, Quantitative Trait Loci) and variations of this region that are responsible for phenotypic variations (Doerge, 2002). Having an accurate linkage map is crucial for the searching for QTLs, where the density of genetic markers is an important factor to detect them (Rodriguez-Ramilo et al., Quantitative Trait Loci for Skeletal Deformities 67 2014). In gilthead seabream, a linkage map based on 204 microsatellite markers and 26 linkage groups (LG) is available (Franch et al., 2006). Microsatellites are probably the most useful molecular markers for medium-high density maps. Multiplex PCR is a highly effective tool to reduce cost per reaction and minimize genotyping errors by reducing steps and introducing automation during the sample analysis process (Navarro et al., 2008). This is especially important in methodologies where a large number of markers has to be genotyped, such as searching QTLs. In gilthead seabream, a set of 13 multiplex PCR assays formed by 106 specific microsatellite markers that cover all LG in the gilthead seabream genetic map have been developed by Negrín-Báez et al. (2014a). Marker-assisted selection (MAS) may be implemented through identification of QTLs. The advantages of MAS are notable as compared with the traditional selective breeding. MAS selection would be especially profitable for traits that are difficult to measure on offspring, that exhibit low heritability and/or that are expressed later in development (Yue, 2014), such as skeletal deformities. The detection and classification of this type of anomalies is very difficult because of the lack of a standardized classification system for reared fish, mainly, at small size. Moreover, estimated heritability for presence/absence of any kind of vertebral deformities and lack of operculum in gilthead seabream showed medium-low values, and was close to zero in small size (Lee-Montero et al., 2014). Currently, MAS has not played a major role in most of genetic improvement programs in aquaculture industry (Yue, 2014). However, a wide number of studies using microsatellite markers have revealed the existence of QTLs for the most interesting traits in different species. Among these, growth and morphometric parameters (total length, standard length, eyes diameter and eyes cross) were related to QTLs presence in Asian seabass (Lates calcarifer), European seabass (Dicentrarchus labrax), rainbow trout (Oncorhynchus mykiss), turbot (Scophthalmus maximus), Atlantic salmon (Salmo salar), Artic charr (Salvelinus 74 4.4. Results 4.4.1. Genotyping Out of 106 microsatellite marker genotyped, there were several no informative markers in each family because either the breeders were homozygous or null alleles were observed. The number of informative microsatellite markers per family, map length covered, the average distance between microsatellite markers and the average number of microsatellite markers per linkage group are shown in Table 9. Table 9. QTL mapping information. Number of microsatellite markers used per family (NMk), map length covered in cM, the average distance between microsatellite markers (MD) and the average number of microsatellite markers per linkage group (MMk) in two types of QTL analysis Study* Families NMk cM MD MMk Study 1 FAM1 95 936.7 15 3.8 FAM2 96 886.2 13.8 4 FAM3 94 932.5 15 3.8 FAM4 94 952.4 14.9 3.9 FAM5 94 936.9 15.1 3.9 Study 2 FAM♂1O 101 916.9 13.8 4 Study 3 FAM12 93 748.5 14.6 3.9 *Data shown in Study 1 and 3 were obtained from full-sib analysis, while data shown in Study 2 were obtained from half-sib analysis 4.4.2. Study 1 QTL Mapping A full-sib regression analysis was carried out for FAM1-5 and half-sib regression analysis was carried out for families from ♂1 L and ♂2L (paternal half-sib) and families from ♀1L, ♀2 L and ♀3 L (maternal half-sib). Both analysis, moreover, were carried out at chromosome and genome-wide level. In this study, the QTLs that were significant in, at least, two of the analyses related with family structure were considered solid. Fourteen QTLs for LSK complex deformity were detected and their positions, confidence intervals and Quantitative Trait Loci for Skeletal Deformities 75 significances from all the analyses are shown in Table 10. Four of them (named QTLSK2, QTLSK 5, QTLSK10 and QTLSK 12) were considered solid, and showed significance also at genome-wide level in, at least, one analysis. The percentages of variance explained by solid QTLs (PVE) are shown in Table 13. Genotypic Association Analysis All microsatellite markers from LG where a solid QTL for LSK complex deformity was detected were analyzed, although only data from those that were closer to each QTL are shown. Microsatellite markers, their allelic significant association for breeders and the percentage of variance explained by genotype (R2) are shown in Table 13. Markers BId-39-T (28.9 cM) and P96 (40.1 cM) were the closest marker to QTLSK2 and a significant association between the allelic segregation of a dam and LSK deformed offspring was found. All the individuals with LSK complex showed the allele 74 for BId-39-T and the allele 78 for P96, while none of the normal individuals showed these alleles. In LG8 the closet marker for QTLSK5 was DId-22-F, which showed a significant association between the allelic segregation of a dam and LSK deformed individual (P=0.05). Only a microsatellite marker was close to QTLSK10, Gt57 (0 cM). Significant Pearson chi-square test revealed a relationship between the allelic segregation of a sire and a dam and their LSK offspring (P<5·10-4 and P≤4·10-3, respectively). For this marker, LSK individuals inherited the allele 150, while normal individual did no show this allele. The closest microsatellite markers for QTLSK12 position were DId-03-T and Bt-14F (1.3 cM). For DId-03-T, the use of contingence tables and a significant Pearson X2 test showed a high association of breeder’s allelic segregation for this marker with LSK offspring, both sires (P≤4·10-5) and dams (P≤0.01). All the individuals with this deformity inherited the allele 126 from sire or from dam, while none of the normal fish inherited this allele. Similar results were observed for Bt-14-F, with a high significant association between LSK and allelic Results 76 segregation of sires (P≤0.01) or dams (P<5.10 -3). In this case, all the LSK offspring showed allele 110, which did not appear in normal individuals. 4.4.3. Study 2 QTL Mapping Results revealed two significant QTLs in LG9 (QTLOP1) and LG10 (QTLOP2) and two suggestive QTLs in LG12 (QTLOP3) and LG14 (QTLOP4), with respect to the lack of operculum deformity (Table 11). Half-sib regression for descendants from the sire ♂1 O was the most robust analysis because of the structure familiar design, so the significant QTLs from this analysis were considered solid. The percentage of variance explained (PVE) by these solid QTLs are shown in Table 13. Genotypic Association Analysis Closely located microsatellite markers, their allelic significant association for breeders and percentage of variance explained by genotype (R2) for each solid QTL are shown in Table 13. The microsatellite markers closely located to QTLOP1 position were CId-89-H (26.7 cM) and Gd-78-F (40.5 cM), however, CId-89-H was an uninformative marker because breeders were homozygous for this locus. With respect to Gd-78-F, Pearson chi-square test showed a significant association (P value= 0.02) between dam allelic segregation and presence of the deformity in offspring. No significant association with closely located markers was found for QTLOP2. Quantitative Trait Loci for Skeletal Deformities 77 Table 10. Description of QTL detected in Study 1 for LSK complex in gilthead seabream: type of analysis (FS, full-sib; pHS, paternal half-sib; and mHS, maternal half-sib), position in cM, 95% confidence interval in cM (CI) and statistical value for QTL (F) Analysis QTL LG FS pHS mHS QTLSK1 1 Position 0 CI 0-16 F 3.6* QTLSK2 5 Position 39 17 CI 11-45 0-47 F 25.1**+ 6** QTLSK3 6 Position 25 CI 3-45 F 3.6* QTLSK4 7 Position 13 CI 0-56 F 6.1** QTLSK5 8 Position 29 34 26 CI 10-36 17-39 3-39 F 9.1* 8.6**+ 4.9** QTLSK6 10 Position 46 F 3.8* CI 0-46 QTLSK7 11 Position 0 CI 0-51 F 3.5* QTLSK8 12 Position 48 CI 0-48 F 20.4** QTLSK9 15 Position 9 CI 0-9 F 4* QTLSK10 17 Position 0 0 7 CI 0-20 0-17 0-129 F 147.4**++ 25.6**++ 4.3* QTLSK11 18 Position 8 CI 0-28 F 4.8* QTLSK12 20 Position 2 0 2 F 633.6**++ 51**++ 30.3**++ CI 1-4 0-10 1-4 QTLSK13 23 Position 0 CI 0-2 F 6.4** QTLSK14 25 Position 4 CI 1-37 F 5.4* * P≤0.05 at chromosome level (suggestive QTL). **P≤0.01 at chromosome level (significant QTL). +P≤0.05 at genome level (significant QTL). ++P≤0.01 at genome level (highly significant QTL). Results 78 Table 11. Description of QTLs detected in Study 2 for lack of operculum deformity in gilthead seabream: type of analysis (FS, full-sib and paternal half-sib), position in cM, 95% confidence interval in cM (CI)and statistic for QTL (F) Analysis QTL LG SP pHS QTLOP1 9 Position 29 17 CI 21-49 2-84 F 33.1** 10.9** QTLOP2 10 Position 29 CI 13-46 F 10.5** QTLOP3 12 Position 42 CI 7-44 F 8.1* QTLOP4 14 Position 84 CI 3-96 F 6.1* * P≤0.05 at chromosome level (suggestive QTL). **P≤0.01 at chromosome level (significant QTL). 4.4.4. Study 3 QTL Mapping The full-sib regression analysis carried out revealed 7 QTLs (Table 12). Among them, there were a significant QTL for each deformity studied in this analysis. One significant QTL for vertebral fusion was detected in LG21 (QTLFV3), and two suggestive QTLs in LG4 and LG13. With respect to lordosis, one significant QTL was detected in LG9 at chromosome and genome-wide level (QTLORD1) and another suggestive QTL in LG22. Additionally, a significant QTL for jaw deformity was detected (QTLFW1) in LG13 and another suggestive in LG16. The percentage of variance explained (PVE) by all significant QTLs are shown in Table 13. Genotypic Association Analysis Close microsatellite markers, their allelic significant association for breeders and percentage of variance phenotypic explained by genotype (R2) for each significant QTL are shown in Table 13. Two closed mircrosatellite markers were analyzed for QTLFV3, Hd-46-T Quantitative Trait Loci for Skeletal Deformities 79 and DId-12-F (0 cM), however, DId-12-F was an uninformative marker, because breeders were homozygous for this locus. A significant association between sire allelic segregation and vertebral fusion offspring was observed (P≤0.03). For QTLORD1, the closely located microsatellite markers Gd-78-F (40.5 cM) and CId-89-H (26.7 cM) were uninformative (same above). No significant association was found between lordotic offspring and breeder’s allelic segregation for any microsatellite marker located in LG9. Table 12.Description of QTL detected in Study 3 for vertebral fusion, lordosis and jaw deformity in gilthead seabream: type of analysis (FS, full-sib), position in cM, 95% confidence interval in cM (CI) and statistic for QTL (F) Analysis QTL LG SP QTLFV1 4 Position 8 CI 8-16 F 11.8* QTLFV2 13 Position 12 CI 7-12 F 6.3* QTLFV3 21 Position 0 CI 4-4 F 4.3** QTLOR1 9 Position 27 CI 26-27 F 63.6**+ QTLOR2 22 Position 0 CI - F 4.3* QTLJW1 13 Position 11 CI 11-12 F 19.1** QTLJW2 16 Position 4 CI 1-4 F 16.6* * P≤0.05 at chromosome level (suggestive QTL). **P≤0.01 at chromosome level (significant QTL). +P≤0.05 at genome level (significant QTL). Results 80 Table 13. Percentage of variance explained by the QTL expressed in PVE and R2 and allelic association analysis for microsatellite markers close to each solid QTL detected in the three studies Study QTL LG PVE Marker IC X2(♂) X2(♀) R2 HC MH Study 1 QTLSK2 5 7.2 43.6 BId-39-T L1 (-) ♀3L (P< 2·10-3) 42.4±23.5 P96 J1 (-) ♀3L (P< 6·10-3) 30.9±24.7 QTLSK5 8 2.5 60 DId-22-F M4 ♀1L (P=0.05) 30.4±40.4 QTLSK10 17 36.8 100.8 Gt57 B3 ♂2L (P<5·10-4) ♀2L(P<4·10-3) 85.7±40.9 QTLSK12 20 98.7 220.5 DId-03-T L9 ♂2L (P<4·10-5) ♀1L(P<0.01) 89.8±17.7 ♀2L(P<5·10-8) Bt-14-F A5 ♂1L (P<0.01) ♀3L(P<5·10-3) 69.4±41.6 ♂2L (P<8·10-6) Study 2 QTLOP1 9 7.4 26.5 Gd-78-F L10 ♂1o (P<0.03) (-) 3.6 QTLOP2 10 - 25.3 BId-15-F B7 (-) (-) 4.3 Study 3 QTLFV3 21 0.1 - Hd-46-T A4 ♂1 (P<0.03) (-) 3.4 QTLORD1 9 1.4 - DId-14-H C1 (-) (-) 1.6 QTLJW1 13 0.4 - CId-26-H F2 (-) ♀1(P<8·10-5) 11.4 CId-03-F G2 (-) ♀1(P<0.01) 7.8 X2: Pearson chi-square test: significance when P≤0.05); (-): any significance observed For QTLJW1, the position estimated was 11cM, but this position was calculated by considering as 0cM the position of marker Bd-29-H (17.3 cM), which was the first analized in this LG. So, the real position of QTLJW1 was 28.3 cM, and CId-26-H (27.9 cM) and CId-03F (30 cM) were its closest microsatellite markers. A significant association (P ≤ 0.01) for maternal allelic segregation in both markers was obtained by Pearson chi-square test. 81 4.5. Discussion The development of new genetic technologies has significantly increased in the last decades. From the emergence of the molecular-based knowledge as a tool in the aquaculture genetic improvement, this has continued to gain importance and it will become increasingly important as aquaculture further develops (Dunham et al., 2014). In the present study, genotyping of all families analyzed was conducted by using 106 microsatellite markers from 13 multiplex PCRs located in the linkage map for this species (Franch et al., 2006; Senger et al., 2006). The use of multiplex PCR allowed the automation of the genotyping process, which entails a significant reduction of costs and minimization of errors (Negrín-Báez et al., 2014a). The average space between these microsatellite markers was 14.6 cM, what is below the recommended maximum distance at QTL searching (20 cM) (Massault et al., 2008). This distance between markers is larger than in other studies of searching QTLs for this species (Boulton et al., 2011; Loukovitis et al., 2011, 2012 and 2013). However, it must be noted that, in the present study, an average of 72.6% (901.4 cM) of total length of the genetic linkage map (1241.9 cM) of gilthead seabream has been covered; while in the other studies, the covered length was significantly lower: 495.4 cM by Loukovitis et al. (2012) or 472 cM by Boulton et al. (2011). In Aquaculture, QTL mapping studies that are based on detection in F1 (full-.sib or half-sib) families have been demonstrated to be successful methods developed for most of marine fish species, since they are prolific and it is easy to obtain large families. In this type of method, sample size, heritability of trait, average allelic substitution effect of the involved alleles and association with closely located markers determine its power to detect the QTLs that are responsible for a specific target trait phenotypic variance (Wang et al., 2006). In this way, methodology based on linear regression was used in the present study to detect QTLs by Discussion 82 performing GridQTL software, which has been evidenced to be robust for use with discrete characters, including binary traits (Gorman et al., 2011), as it is the case of deformities. More concretely, in this study, six families containing from 5 to 152 individuals per family were analyzed and medium-low heretitabilities for deformities had been estimated for this species (Lee-Montero et al., 2014). 4.5.1. Study 1 LSK complex is a triple column deformity that is easily diagnosed at early ages and that is presented with a low prevalence in gilthead seabream fingerlings. However, it is very severe and produces high mortality, so hatcheries have to detect and discard the fish suffering from this deformity before selling them to on-growing companies. Thus, it has a substantial economic impact (Negrín-Báez et al., 2014b). In this study, for the first time for gilthead seabream, QTLs for this deformity have been identified. A total of five full-sib families (three maternal and two paternal half-sib families) were genotyped and 14 LSK-complex related QTLs were found. The finding of a large number of QTLs suggests that this deformity is influenced by a complex interaction between several regions of the genome, which was expected since this deformity is composed of three different deformities. This is also in concordance with the genetic origin of this complex deformity previously proposed by Afonso et al. (2000) and Negrín-Báez et al. (2014b), who found significant statistical associations between the prevalence of this deformity and certain specific families in gilthead seabream. Four of the QTLs detected in the present study (QTLSK2, 5, 10 and 12) were considered the most solid ones, since they were significant in, at least, two of the analyses conducted and related with family structure (full-sib and half-sib) at chromosome level. It is remarkable that these 4 QTLs also reached a genome-wide level of significance in, at least, one of the analyses; this is a more stringent statistical level (Loukovitis et al., 2012). QTLSK2 (in LG5) and QTLSK5 (in LG8) were significant, and QTLSK10 (in LG17) and QTLSK12 (in LG20) were highly significant. Additionally, according to their percentage of phenotypic Quantitative Trait Loci for Skeletal Deformities 83 variance explained (PVE), these solid QTLs showed an extremely large effect in half-sib analysis, and QTLSK10 and QTLSK12 also in full-sib analysis. Actually, PVE values of these two QTLs in the half-sib analysis were very high and theoretically impossible, since they were over 100% (100.8 and 220.5%, respectively). It is a common result in QTL mapping studies, as PVE tends to be overestimated (Wang et al., 2006). It can be explained by the type of analysis (linear regression versus mixed models) (Rowe et al., 2006) and by the Beavis effect (Xu, 2003). The Beavis effect explains that the detected QTLs effect is overestimated because QTL identification requires a genetic effect estimation and a statistical analysis in a shorten distribution; and the overestimated QTL effect values can be lower as offspring size is larger (Xu, 2003). In this study, families sizes ranged from 6 to 25 individual per full-sib family and from 15 to 44 per half-sib family, so QTL effect values could be closer to real magnitude when analyzing larger families. Anyway, it is clear that the effect of the four solid detected QTLs is extremely large, especially QTLSK10 and QTLSK12. In fact, they are the only QTLs described in gilthead seabream that have shown an extremely large effect (Massault et al., 2010; Boulton et al., 2011; Loukovitis et al., 2011, 2012, 2013). These results indicate that LSK complex deformity might be controlled mainly by these few largeeffect QTLs. In this way, it must be considered that only 78 fish were analyzed, so these QTLs must be solid. This was expected, since the five analyzed families were from a larger experiment in which only six out of 89 families contained the total incidence of LSK individuals, and a significant relationship between breeders and families showing this deformity was obtained (Negrín-Báez et al., 2014b). Moreover, a highly significant association between phenotypic trait and molecular genotype in these solid QTLs for LSK complex was observed. In fact, close microsatellite markers analyzed for these QTLs seem to be linked with LSK complex, so all LSK descendants of, at least, one breeder showed the determined alleles, while normal individuals did not show these alleles, and all LSK descendants of, at least, one full-sib family showed 5. CONCLUSIONS Conclusions 93 1. A set of 13 multiplex PCRs (named from ReMsa1 to ReMsa13) has been developed. They include 106 microsatellite markers of gilthead seabream genetic map that cover 100% of the linkage groups. This battery is a powerful tool for QTLs searching in this species. 2. Twenty additional microsatellite markers have been redesigned and evaluated. Their amplification occurs under the same PCR conditions of set ReMsa, and they can be included in analyses in future. 3. It has been determined that the skeletal deformities LSK complex, lordosis and vertebral fusion have a genetic origin in gilthead seabream, as they showed a significant association with family structure. 4. The high prevalence of lordosis and vertebral fusion deformities in descendants from LSK-complex breeders suggests that a relationship between these three deformities exists in gilthead seabream. 5. Lack of operculum deformity in this species is mainly influenced by environment factors in small size fish, but its genetic determination at commercial size has been confirmed. 6. Both, genetic and phenotypic, controls of gilthead seabream breeders in designed matings affect the quality of spawning in terms of viability and prevalence of deformities. 7. Fourteen QTLs have been related with LSK complex deformity in gilthead seabream. Four of them (those located in linkage groups 5, 8, 17 and 20) were considered the most solid ones, as they showed a very high responsibility in the phenotypic variance for this deformity. 8. Genotype of five microsatellite markers closely located to these four LSK-complex solid QTLs showed a strong association with presence of this deformity in offspring. Conclusions 94 So, they have been proposed as potential candidates to be included in a Marker Assisted Selection for this species. 9. Two significant QTLs related to lack of operculum deformity in gilthead seabream were identified. They were located in linkage groups LG9 and LG10 and showed a medium effect on phenotypic variance. 10. A full-sib family of gilthead seabream showed one significant QTL for each one of the following analyzed deformities: lordosis (located in linkage group 9), vertebral fusion (located in linkage group 21) and jaw deformity (located in linkage group 13). However, they should be confirmed in other families. 11. For the first time in gilthead seabream, the genetic origin of jaw deformity has been demonstrated. 12. These results represent a major step towards the location of genes determining the presence of the analyzed skeletal deformities in gilthead seabream, and, subsequently, they offer a useful tool to reduce the prevalence of these deformities when included within a genetic breeding program in gilthead seabream industry. 6. REFERENCES References 97 Afonso J.M. and Roo F.J (2007) Anomalías morfológicas en peces cultivados: Heredabilidad y selección. In: Genética y genómica en acuicultura (Ed. by J. Espinosa; Coord. by P. Martínez and A. Figueras). Consejo Superior De Investigaciones Científicas, Madrid, pp. 215-240. Afonso J.M., Manchado M., Estévez A., Ramis G., Lee-Montero I., Ponce M., Sánchez J.A., Armero E., Navarro A., Puertas M.A., Borrell Y., García M., Negrín-Báez D., Crespo A.M., Blanco G., Mariadolores E., Sánchez J.J., Martín N., Aguilera C., Roo F.J., Zamorano M.J. and Toro M.A (2012) PROGENSA®: development of a genetic improvement program in gilthead sea bream Sparus aurata L between industry and research centers in Spain. In: Proceedings of AQUA 2012, September 1-5, Prague, Czech Republic. Afonso J.M., Montero D., Robaina L., Astorga N., Izquierdo M.S. and Ginés R (2000) Association of a lordosis-scoliosis-kyphosis deformity in gilthead seabream (Sparus aurata L) with family structure. Fish Physiology and Biochemistry 22,159-163. Antonello J., Massault C., Franch R., Haley C., Pellizzari C., Bovo G., Patarnello T., Koning D.J. and Bargelloni L. (2009) Estimates of heritability and genetic correlation for body length and resistance to fish pasteurellosis in the gilthead seabream (Sparus aurata L.). Aquaculture 298, 29–35. APROMAR (2012; 2014) Fish marine aquaculture in Spain. Available at: http://www.apromar.es. AquaExcel Project, 2013. Deliverable 3.2: Best practices and cross-applicability of methods to measure phenotypes.available at: http://www.aquaexcel.eu/index.php?option=com_ k2&view=item&layout =item&id=7&Itemid=212 Argyrokastritis A., Pasparaki A., Kokkinaki M. and Moschonas N.K. (2002) Gilthead seabream (Sparus aurata L.) genome survey. Available at: http://www.ncbi.nlm.nih.gov. Arias A. (1980) Crecimiento, régimen alimentario y reproducción de la dorada (Sparus aurata L.) y del róbalo (Dicentrarchus labrax L.) en los esteros de Cádiz. Investigación Pesquera 44, 59-83. Astorga N., Zamorano M.J, Toro M.A., García Cortés L.A., Montero D., Izquierdo M. and Afonso J.M. (2007) Body deformities in gilthead seabream (Sparus auratus L.): An estimation of heritability. Proceedings of the Ninth International Symposium, 26–30 June 2006, Montpellier, France. Aquaculture 272, Supplement 1 (Ed. by Vandeputte M., Chatain B. and Hulata G.), pp. S238–S321. Astorga N., Zamorano M.J., Toro M.A., García Cortés L.A., Montero D. and Afonso J.M. (2004) Heredabilidad del carácter presencia o ausencia de deformaciones esqueléticas en dorada (Sparus aurata L.). ITEA 100a., 256–250. Bardon A., Vandeputte M., Dupont-Nivet M., Chavanne H., Haffray P., Vergnet A., and Chatain B. (2009) What is the heritable component of spinal deformities in the European sea bass ( Dicentrarchus labrax)? Aquaculture 294, 194-201. Batargias C., Dermitzakis E., Magoulas A. and Zouros E. (1999) Characterization of six polymorphic microsatellite markers in gilthead seabream, Sparus aurata (Linnaeus 1758). Molecular Ecology 8, 897-898. Ben-Tuvia A. (1979) Studies of the population and fisheries of Sparus aurata L. in the Bardawil Lagoon, eastern Mediterranean. Investigación Pesquera 43, 43-67. Beraldo P. and Canavese B. (2011) Recovery of opercular anomalies in gilthead sea bream, Sparus aurata L.: Morphological and morphometric analysis. Journal of Fish Diseases 34, 21–30. Beraldo P., Pinosa M., Tibaldi E., and Canavese B. (2003) Abnormalities of the operculum in gilthead seabream (Sparus aurata L): Morphological description. Aquaculture 220, 89-99. Boglione C. and Costa C. (2011) Skeletal deformities and juvenile quality. In: Sparidae: Biology and Aquaculture of Gilthead Sea Bream and Other Species (Ed. by M. Pavlidis and C. Mylonas), Blackwell, Oxford, pp. 233–94. References 98 Boglione C., Gagliardi F., Scardi M. and Cataudella S. (2001) Skeletal descriptors and quality assesment in larvae and post-larvae of wild-caugth and hatchery-reared gilthead seabream (Sparus aurata L). Aquaculture 92, 1-22. Boglione C., Gisbert E., Gavaia P., Witten P.E., Moren M., Fontagné S. and Koumoundouros G. (2013) Skeletal anomalies in reared european fish larvae and juveniles. Part 2: main typologies, occurrences and causative factors. Reviews In Aquaculture 5 (Suppl. 1), S121–S167. Borrell Y., Álvarez J., Vázquez E., Fernández C., Mártinez C., Sánchez J.A. and Blanco G. (2004) Applying microsatellites to the management of turbot stocks (Scophthalmus maximus L.) in hatcheries. Aquaculture 241, 133–150. Borrell Y.J., Gallego V., García-Fernández C., Mazzeo I., Pérez L., Asturiano J.F., Carleo C.E., Vázquez E., Sánchez J.A. and Blanco G. (2011) Assessment of parental contributions to fastand slow-growing progenies in the sea bream Sparus aurata L. using a new multiplex PCR. Aquaculture 314, 58–65. Botstein D., White R.L., Skolnick M. and Davis R.W. (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. American Journal of Human Genetics 32, 314-331. Boulton K., Tsigenopoulos C.S., Massault C., Houston R.D., de Koning D.J., Haley C.S., Bovenhuis H., Batargias C., Canario A.V.M. and Kotoulas G. (2011) QTL affecting morphometric traits and stress response in the gilthead seabream (Sparus aurata L). Aquaculture 319, 58-66. Bouza C., Sanjuán A., de Carlos A., Alvárez-Dios J.A., Ezcurra S., Cal R.M., Piferrer F., Martínez P., Hermida M., Pardo B.G., Fernández C., Fortes G.G., Castro J., Sánchez L., Presa P. and Pérez M. (2007) A Microsatellite Genetic Map of the Turbot (Scophthalmus maximus). Genetics 177, 2457-2467. Brown R.C., Tsalavouta M., Terzoglou V., Magoulas A. and Mcandrew J. (2005) Additional microsatellites for Sparus aurata L and cross-species amplification within the Sparidae family. Aquaculture 5, 605-607. Castelló-Orvay F. and Calderer A. (1993) Growth of gilthead sea bream (Sparus aurata L.) under different culture conditions. In: Production, environment and quality (Ed. by Barnabé G. et al.): Proceedings of the International Conference Bordeaux Aquaculture '92, Bordeaux, France, March 25-27, 1992. EAS Special Publication 18. pp. 227-233. Castro J., Pino-Querido A., Hermida M., Chavarrías D., Romero R., García-Cortés L.A., Toro M.A. and Martínez P. (2008) Heritability of skeleton abnormalities (lordosis, lack of operculum) in gilthead seabream (Sparus aurata L) supported by microsatellite family data. Aquaculture 279, 18-22. Cejas J., Villamandos J. y Samper M. (1992) Estudio sobre la reproducción de la dorada (Sparus aurata L) en Canarias: Influencia del peso/edad de las hembras sobre la calidad de la puesta. Informes Técnicos Instituto Español de Oceanografía 126, 1-38. Chistiakov D.A., Hellemans B. and Volckaert F.A.M. (2006) Microsatellites and their genomic distribution, evolution, function and applications: A review with special reference to fish genetics. Aquaculture 255, 1-29. Churchill G.A. and Doerge R.W. (1994) Empirical threshold values for quantitative trait mapping. Genetics 138, 963-971. Dakin E.E., Avise J.C. (2004) Microsatellite null alleles in parentage analysis. Heredity 93, 504-509. De Wolf T., Courtens V., Capiferri U., Pirone A., Lenzi C. and Lenzi F. (2004) The influence of light conditions on the operculum recovery of sea bream (Sparus aurata L.) fry. In: Proceedings of Aquaculture Europe 2004, Biotechnologies for Quality, Barcelona, Spain, pp. 292–29. Divanach P., Boglione C., Menu B., Koumoundourus G., Kentouri M. and Cataudella S. (1996) Abnormalities in finfish mariculture: An overview of the problem, causes and solutions. In: References 99 International workshop on seabass and seabream culture: problems and prospects (Ed. By Chatain, B., Saroglia, M, Sweetan, J., Lavens, P.) European Aquaculture Society, Oostende, pp. 45-66. Doerge R.W. (2002) Mapping and analysis of quantitative trait loci in experimental populations. Nature Reviews Genetics 3, 43-52. Dunham (2014) Introduction to genetics in aquaculture XI: The past, present and future of aquaculture genetics. In: Proceedings of The Eleventh International Symposium on Genetics in Aquaculture. Auburn University, Auburn, Alabama, USA. Aquaculture 420–421, S1–S2. Ebrahimnezhad H., Modarres Mousavi M., Tabatabaei Naeini A. and Bozorgi H. (2009) Coincidence of lordosis, kyphosis and scoliosis syndrome (LSK) in parrot cichlid (Cichlasoma). In: Proceedings of 1st International Congress on Aquatic Animal Health Management and Diseases. Tehran (Iran), abstract 1296. Evans M.L. and Neff B.D. (2009) Non-additive genetic effects contribute to larval spinal deformity in two populations of chinook salmon (Oncorhynchus tshawytscha). Aquaculture. 296, 169-173. Falconer D.S. and Mackay T.F.C. (1996) Introducción a la genética cuantitativa. Acribia, Zaragoza, pp. 469. Fjelldal P.G., Glover K.A., Skaala Ø., Imsland A. and Hansen T.J. (2009) Vertebral body mineralization and deformities in cultured Atlantic salmon (Salmo salar L.): Effects of genetics and off-season smolt production. Aquaculture 296, 36-44. Franch R., Louro B., Tsalavouta M., Chatziplis D., Tsigenopoulos C.S., Sarropoulou E., Antonello J., Magoulas A., Mylonas C.C., Babbucci M., Patarnello T., Power D.M., Kotoulas G. and Bargelloni L. (2006) A genetic linkage map of the hermaphrodite teleost fish Sparus aurata L. Genetics 174, 851-861. Galeotti, M., Beraldo P., de Dominis S., D’Angelo L., Ballestrazzi R., Musetti R., Pizzolito S. and Pinosa M. (2000) A preliminary histological and ultrastructural study of opercular anomalies in gilthead sea bream larvae (Sparus aurata). Fish Physiology and Biochemistry 22, 151–157. Georgakopoulou E., Katharios P., Divanach P. and Koumoundouros G. (2010) Effect of temperature on the development of skeletal deformities in gilthead seabream (Sparus aurata L, 1758). Aquaculture 308, 13-19. Gislason H., Karstensen H., Christiansen D., Hjelde K., Helland S. and Bæverfjord G. (2010) Rib and vertebral deformities in Rainbow trout (Oncorhynchus mykiss) explained by a dominantmutation mechanism. Aquaculture 309, 86-95. Gjedrem T. (2005) Selection and breeding programs in aquaculture. Dordrecht, NLD: Springer. Gjerde B., Pante M.J. and Baeverfjord G. (2005) Genetic variation for a vertebral deformity in Atlantic salmon (Salmo salar). Aquaculture 244, 77–87. Gorman K.F., Christians J.K., Parent J., Ahmadi R., Weigel D., Dreyer C. and Breden F. (2011) A major QTL controls susceptibility to spinal curvature in the curveback guppy. BMC genetics 12, 16. Herbinger C.M., O’Reilly P.T., Doyle R.W., Wright J.M. and O'Flynn F. (1999) Early growth performance of Atlantic salmon full-sib families reared in single family tanks versus in mixed family tanks. Aquaculture 173, 105-116. Hough C. (2009) Improving the sustainability of European fish aquaculture by the control of malformations. Ghent University, Belgium. Resumen español 107 RESUMEN El objetivo del presente trabajo fue analizar la existencia de QTLs (Quantitative Trait Loci) que afecten a las deformidades esqueléticas más relevantes en dorada (Sparus aurata), que pudieran ser utilizados como herramienta para minimizar la prevalencia de estas deformidades dentro de un programa de mejora para esta especie. La incidencia de estas anomalías en esta especie supone importantes pérdidas económicas en la industria, tanto en las empresas de cría como en las de engorde, por lo que los peces deformes se eliminan mediante cribas manuales en diferentes puntos de la línea de producción. En un primer paso, se desarrolló una batería de multiplex PCRs a partir de 138 marcadores microsatélites del mapa genético de la dorada, que se rediseñaron para ser amplificados en las mismas condiciones de PCR. 13 nuevas multiplex PCR (nombradas como ReMsa1 a ReMsa13) formaron el set con 106 de estos microsatélites, cubriendo el 100% de los grupos de ligamiento. Este set se validó satisfactoriamente, constituyendo una poderosa herramienta para buscar QTLs en esta especie. Con el fin de encontrar una estructura familiar que facilitara la localización de QTLs, se analizó, en distintas poblaciones de dorada, la prevalencia de cuatro de las deformidades esqueléticas más importantes y graves: ausencia de opérculo, lordosis, fusión de vértebra y complejo LSK (lordosis-escoliosis-cifosis). Para ello se realizaron 3 experimentos a gran escala. En el Experimento 1, se realizó una puesta masal y la descendencia, a los 111 días, se cribó en términos de deformidades. A los 509 días de edad, se analizaron 900 peces: 846 criados en una granja y 54 peces LSK seleccionados en el proceso inicial de criba y criados por separado en un tanque. Todos los individuos que presentaron esta deformidad estaban en seis de las 89 familias representadas, mostrando éstas una asociación significativa con la prevalencia de la deformidad. Por lo tanto, cinco de estas seis familias se seleccionaron para el mapeo de QTL. En el Experimento 2, los peces procedían también de una puesta Resumen español 108 masal pero no se realizó ningún tipo de criba. Aunque a talla pequeña la prevalencia de ausencia de opérculo en la descendencia fue independiente de la familia o reproductor, a los 539 días, sí se detectaron asociaciones significativas entre un padre, una madre, y dos de las familias, con la prevalencia de esta deformidad. De hecho, el 48,94% de los individuos que presentaron esta deformidad fueron descendientes de este padre. Por lo que todos los descendientes de este macho (seis familias de medios hermanos) se seleccionaron para el mapeo de QTL. En el Experimento 3, se realizaron cruces dirigidos entre machos deformes que presentaban lordosis, fusión de vértebras o ausencia de opérculo con hembras no deformes, además de una puesta masal como control. A los 129 días un total de 11.503 alevines se analizaron visualmente para las deformidades. Los resultados mostraron una asociación significativa entre la prevalencia de cada deformidad y el cruce en el que el reproductor presentaba la misma deformidad. En los descendientes del cruce del macho lordótico también se observó una significativa prevalencia de falta de opérculo, revelando que esta deformidad parece estar más influenciada a esta edad por factores ambientales, mientras que la lordosis y la deformidad de fusión de vértebras están relacionadas con la estructura familiar. Por ultimo, se realizó un análisis de QTL para cinco deformidades esqueléticas en 12 familias de la dorada, dividido en tres estudios. Todos los peces se genotiparon utilizando el set de multiplex PCRs ReMsa1-13. En el Estudio 1, se analizaron para la deformidad LKS cinco familias seleccionadas en el Experimento 1. Cuatro QTLs significativos, de los 14 encontrados para esta deformidad, situados en los grupos de ligamiento (GL) 5, 8, 17 y 20 presentaron un efecto extremadamente grande y se consideraron los más sólidos. Se observó una fuerte asociación entre el genotipo de los marcadores cercanos a estos QTLs con el fenotipo. En el Estudio 2 se analizó, para la ausencia de opérculo, la familia de medios hermanos paternos seleccionada en el Experimento 2. De los cuatro QTLs detectados para esta deformidad, dos fueron significativos, localizados en el GL 9 y el GL 10, y mostraron un Resumen español 109 efecto grande. En uno de ellos se observó una asociación significativa entre la deformidad falta de opérculo y la segregación alélica del macho. En el estudio 3, se analizó para diferentes deformidades esqueléticas, una familia de hermanos carnales procedente de un cruce entre reproductores deformes. Tres de los siete QTLs encontrados fueron significativos, uno para la fusión de vértebras en el GL 21, otro para la lordosis en el GL 9 y otro para la deformidad de mandíbula en el GL 13. No obstante sería recomendable confirmarlos en otras familias de dorada. Los resultados derivados del presente trabajo, confirman el origen genético de estas deformidades esqueléticas en dorada y su relación con la estructura familiar. Además, se han detectado QTLs para estas deformidades, así como una asociación significativa entre la segregación alélica de los marcadores moleculares cercanos a ellos y el fenotipo. Esto supone un gran paso hacia la localización de los genes que determinan el desarrollo de las deformidades esqueléticas en esta especie. Lo expuesto ofrece a la industria de dorada una herramienta con gran potencial para disminuir la prevalencia de las deformidades dentro de un programa de mejora genética. Resumen español 110 1. INTRODUCCIÓN 1.1 Especie de estudio: la dorada 1.1.1 Taxonomía y nomenclatura La dorada es una especie marina, perteneciente a la familia de los espáridos y al género Sparus (Sparus aurata, Linnaeus, 1758). Su clasificación taxonómica es la siguiente: Reino Animal Phylum Chordata Subphylum Gnathostomata Clase Actinopterygii Subclase Teleostei Superorden Neognathi Orden Perciformes Familia Sparidae Género Sparus Especie Sparus aurata Sus nombres vernaculares en los distintos idiomas son los siguientes: Español Dorada Inglés Gilthead seabream Francés Dorade Alemán Goldbrasse Italiano Orata 1.1.2 Distribución y Hábitat El área de distribución de esta especie es a través de las costas orientales del océano Atlántico, desde Gran Bretaña hasta Cabo Verde, así como todo el mar Mediterráneo. Debido a sus hábitos eurihalinos y euritérmicos, habita zonas litorales en ambientes marinos, lagunas Resumen español 111 costeras y zonas de estuarios, en particular durante las primeras fases del ciclo de vida. Las doradas hacen una migración trófica, desovan en el mar en invierno, posteriormente los juveniles migran a principios de primavera en busca de lugares más protegidos, con abundancia de comida y con temperaturas medias; para luego regresar a mar abierto a finales de otoño para reproducirse (Moretti et al., 1999). 1.1.3 Anatomía La dorada es un teleósteo de cuerpo oval, más bien profundo y comprimido. El perfil de la cabeza aparece regularmente curvado y tiene los ojos pequeños. La boca es baja, levemente oblicua y con labios gruesos. Presenta de cuatro a 6 dientes anteriores de tipo caninos en cada mandíbula, seguidos posteriormente por dientes romos, los cuales se transforman progresivamente a molares y se disponen en 2 a 4 filas (los dientes en las 2 filas externas más fuertes). Presenta un color gris plateado; una gran mancha negra, en el origen de la línea lateral, extendiéndose sobre el margen superior del opérculo. Presenta una banda frontal dorada característica entre los ojos, bordeada por dos áreas oscuras (no bien definidas en los individuos jóvenes); líneas longitudinales oscuras a menudo presentes sobre los costados del cuerpo; una banda oscura sobre la aleta dorsal; así como la horquilla y las puntas de la aleta caudal bordeadas con negro (Figura 1). Resumen español 112 1.1.4 Alimentación y Reproducción La dorada es un pez preferentemente carnívoro, depredadora de especies de fondo (bivalvo y gasterópodo), crustáceos y peces pequeños. En general, se considera una especie de crecimiento rápido en la naturaleza, alcanzando los 300 g en el segundo año y aproximadamente los 600 g en el tercero, pudiendo llegar a tener un tamaño de 70 cm y un peso de 5 Kg (Castelló, 1993). Es una especie hermafrodita proterándrica, que primero madura como macho y a partir del segundo o tercer año se convierte en hembra. Su estación reproductiva tiene lugar durante un periodo relativamente largo, cuando los días son cortos y las horas de luz decrecientes (octubre a diciembre) en el Mediterráneo Occidental (Arias, 1980). En el Mediterráneo Oriental se produce un poco más tarde, entre noviembre y febrero (Ben-Tuvia, 1979). Los huevos son planctónicos, esféricos y transparentes, de aproximadamente 1 mm de diámetro y con una o varias gotas de aceite en su interior, siendo las cantidades puestas de entre 500.000 y 6.000.000 de huevos/kg de hembra (Cejas et al., 1992). Para el desove en época invernal, la dorada migra hacia aguas costeras, ya que son aguas más protegidas, en busca de alimento abundante y temperaturas más suaves (Migración Tropical). A final de otoño vuelve a migrar a mar abierto para comenzar la época de cría en peces adultos. Resumen español 113 1.2. Producción Acuícola de dorada 1.2.1 Producción a nivel mundial y europea Según datos publicados en APROMAR (2014), existe producción de dorada de acuicultura en 19 países, siendo los principales productores Grecia, con aproximadamente 75.000 toneladas (t) (41,7% de la producción total), Turquía con 41.700 t (23,2%) y España con 16.795 t (9,3%) durante el año 2013. Su cultivo se realiza también en Italia, Egipto, Francia, Chipre, Portugal, Croacia, Malta, Túnez y Marruecos, y hay producciones incipientes en Albania, República Dominicana, Marruecos, Emiratos Árabes Unidos, Bosnia, Omán, Libia y Kuwait (Figura 2). 1.2.2 Producción en España De la producción de peces en este país, la producción de dorada fue la más elevada en 2013 (16.795 t), seguida por la producción de trucha arcoíris (Oncorhynchus mykiss) (16.732 t) y lubina (Dicentrarchus labrax) (14.707 t). (APROMAR, 2014). En 2013, la producción de dorada de acuicultura en España fue liderada por la Comunidad Valenciana con 6.974 t (el 42% del total), seguida por Murcia (3.730 t, el 22%), Canarias (3.016 t, el 18%), Andalucía (1.786 t, el 11%) y Cataluña (1.292 t, 8%) (APROMAR, 2014). En la Figura 3, se muestra la producción de dorada en diferentes comunidades autónomas de España. En la Figura 4 se muestra la producción acuícola total de dorada en Canarias, cuyos datos fueron facilitados por el Ministerio de Agricultura, Resumen español 114 Alimentación y Medio Ambiente, Secretaría General de Pesca, Junta Nacional Asesora de Cultivos Marinos (JACUMAR, 2014). En cuanto a la producción de juveniles en España, el número de unidades asciende a 51,4 millones, lo que supone una reducción del 6,5% sobre el dato de 2012. Esta producción de alevines de dorada en España se concentra en la Comunidad Valenciana, Islas Baleares, Cantabria y Andalucía. Sin embargo, para la producción española de dorada de talla comercial es necesaria la importación de juveniles adicionales a éstos de producción propia. El valor de primera venta a talla comercial de dorada en España en el año 2012 fue de 72.165.279 € (JACUMAR, 2014). 1.3 Mejora genética en dorada Actualmente, la dorada de crianza supone un 95,3% del total de esta especie consumida en este país. Este marcado desarrollo y expansión de la industria acuícola en torno a la dorada, ha supuesto una alta competencia en el mercado que se ha traducido en un menor precio de venta al consumidor. Por lo tanto, las empresas se han visto obligadas a minimizar costes y aumentar la producción, para así poder mantener los beneficios. En este sentido, la selección genética ha sido reconocida como un importante factor para el desarrollo de una Acuicultura sostenible, eficiente y rentable (Gjedrem, 2005; Rye et al., 2010). Navarro et al. (2009a y 2009b) estimaron heredabilidades bajo condiciones industriales, a partir de doradas cultivadas en Canarias tanto en tanques como en jaulas insulares y desarrollaron las herramientas genéticas y protocolos necesarios para implementar Resumen español 115 un programa de mejora genética en esta especie sin interferir con el funcionamiento de la industria. Este modelo fue recientemente implementado por el proyecto PROGENSA®, donde se ha desarrollado a escala industrial un programa piloto de mejora en dorada con diferentes empresas de engorde y centros de investigación de cuatro comunidades autónomas, en el que se estimaron parámetros genéticos relacionados con los más importantes caracteres comerciales y establecido la interacción genotipo-ambiente entre comunidades (Afonso et al., 2012). 1.4 QTLs Una forma de mejorar la eficiencia de la selección, dentro de un programa de mejora, es a través de la identificación de QTLs. Un QTL (del anglicismo Quantitative Trait Loci) es una región cromosómica (un gen o grupo de genes) que determinan un carácter cuantitativo. La localización de un QTL que afecte a un carácter de marcado interés productivo, permite la identificación del gen o de los genes que afecten a este carácter y con qué intensidad. Esto puede realizarse a través de la Selección Asistida por Marcadores (MAS), aprovechando el desequilibrio de ligamiento entre estos marcadores moleculares y el fenotipo del carácter estudio en la población que se esté analizando (Enciso y Toro, 2007). La detección de QTLs implementada en MAS muestra notables ventajas respecto a un programa de selección tradicional y el uso de esta herramienta se está incrementando cada vez más en la mejora genética animal (Yue, 2014). Para una eficiente detección de QTLs, son necesarios: un mapa Resumen español 122 esquelética tanto a escala industrial como a escala experimental. Además, la selección por MAS sería una estrategia eficaz para optimizar su mejora, ya que es especialmente rentable para los caracteres que son difíciles de medir en la descendencia, que exhiben baja heredabilidad o que aparecen en etapas avanzadas del desarrollo (Yue, 2014), como es el caso de las deformidades esqueléticas. Sin embargo, no existen en dorada ningún QTL descrito para ningún tipo de deformidad. 1.9 Objetivos En este contexto, el objetivo principal de este trabajo fue: analizar y determinar si existen QTLs que afecten a las deformidades esqueléticas más frecuentes y graves en doradas, que puedan ser útiles para minimizar la prevalencia de estas deformidades en un programa de mejora en esta especie. Para ello, se determinaron los siguientes objetivos secundarios: 1Desarrollar un batería de PCRs múltiplex que contuvieran el mayor número de marcadores microsatélites específicos para dorada, con el fin de cubrir la mayor proporción de grupos de ligamiento del mapa genético de esta especie, como una herramienta eficiente para localizar QTLs. Así como evaluar y validar su correcto funcionamiento. 2Analizar la prevalencia de cuatro deformidades esqueléticas (ausencia de opérculo, lordosis, fusión de vértebra y complejo LSK) y determinar si tiene relación con la estructura familiar, en distintas poblaciones de dorada, con el fin de confirmar el Resumen español 123 origen genético de estas deformidades y encontrar una estructura familiar que facilite la búsqueda de QTLs. 3Genotipar con un set de múltiplex PCRs las familias que tuvieran una asociación significativa con las deformidades analizadas para buscar y localizar posibles QTLs asociados a estas deformidades. 124 2. MATERIAL Y MÉTODOS 2.1 Diseño de una batería de 13 PCRs múltiplex de marcadores específicos como una herramienta para la detección de QTLs en dorada (Sparus aurata L.) 2.1.1 Material biológico Se analizaron un total de 16 doradas procedentes de tres lotes de reproductores industriales de diferentes comunidades autónomas españolas pertenecientes al proyecto de investigación PROGENSA® (Muestras control). De estas muestras, cuatro individuos se utilizaron para optimizar las reacciones de PCRs múltiplex (Muestras de prueba). Por otra parte, una familia de hermanos carnales de 60 hijos y sus reproductores (Muestras de validación) fueron utilizados para verificar la correcta segregación de los alelos y revalorar la facilidad de lectura del genotipado de cada marcador microsatélite mediante el GremmProtocol (Lee-Montero et al., 2013). Se extrajo ADN del tejido de la aleta caudal, utilizando el kit de DNAeasy (QIAGEN®) y luego se mantuvo a-20°C siguiendo las instrucciones proporcionadas por el fabricante. La integridad, cantidad y calidad del ADN se determinaron utilizando el espectrofotómetro Nanodrop 1000 v.3.7 (Thermo Fisher Scientific, Wilmington, Estados Unidos) y mediante electroforesis en gel de agarosa al 1% (8V·cm−1). Resumen español 125 2.1.2 Marcadores microsatélites De los microsatélites publicados en el mapa genético de la dorada (Franch et al., 2006), 138 parejas de cebadores se rediseñaron para que amplificaran bajo las mismas condiciones de PCR. A cada pareja de cebadores se le asignó un código interno para facilitar el trabajo de laboratorio. En la Tabla 1 se muestra el nombre de los loci, su código interno, su código de acceso en la base de datos GenBank, el fluorocromo asignado y la secuencia de los cebadores de cada marcador microsatélite usado en este trabajo. Cada marcador microsatélite fue probado mediante PCR simple utilizando las 16 muestras de ADN para confirmar su correcta amplificación, rango alélico, morfología, variabilidad genética y facilidad de lectura del genotipado. La heterocigosidad observada (Ho), esperada (He) y el contenido de información polimórfica (PIC) de cada marcador, así como la probabilidad de exclusión combinada (CEP) de cada multiplex se estimaron usando el programa informático Cervus (v 3.0.3). 2.1.3 PCR simple Las condiciones de PCR consistieron en una desnaturalización inicial a 95° C durante 10min, seguida de 28 ciclos de 94° C durante 30 segundos, 60° C durante 1minuto y 65° C durante 1minuto, con una extensión final de 60 º C durante 60 minutos. La reacciones se llevaron a cabo en un volumen final de 12,5μl con las siguientes concentraciones para cada componente: 1 X GeneAmp PCR Buffer II (100 mM Tris-HClpH 8,3, 500 mMKCl) (Applied Resumen español 126 Biosystem®, USA), 3 mM MgCl2, 0,2 mM de cada dNTP, 0,04 unidades μl−1 AmpliTaq Gold ADN polimerasa (Applied Biosystem®), 2,4-6,4ng•μl−1de ADN y 0,2μM de cada cebador. 2.1.4 Evaluación de la facilidad de lectura del genotipado La facilidad de lectura del genotipado de cada marcador microsatélite se evaluó mediante la identificación de los errores o potenciales errores de lectura y asignando a cada marcador una puntuación entre 1 y 3 según describieron Lee-Montero et al. (2013) mediante el GremmProtocol. Según esta clasificación 1 denota genotipos ambiguos en más del 30% de las muestras; 2, genotipos ambiguos en un 30% o menos de las muestras; y 3, genotipos sin ambigüedades en el 100% de las muestras. 2.1.5 PCRs múltiplex Después de la evaluación de los marcadores microsatélites, aquéllos que no amplificaron o fueron monomórficos se desecharon. Con el resto de marcadores, se diseñaron 13 PCRs múltiplex que se denominaron ReMsa 113 (Redesigned multiplex Sparus aurata). Los marcadores fueron combinados en cada múltiplex según su fluorocromo, su grupo de ligamiento y el rango del tamaño alélico. Aquellos microsatélites que no amplificaron en una múltiplex, se incluían en otra con el fin de maximizar el número de marcadores por múltiplex. Inicialmente, cada cebador fue utilizado con una concentración de 0,2μM en la PCR multiplex, modificando dicha concentración hasta conseguir una altura del pico entre 600 y 3000 RFU (Relative Fluorescente Units) para cada marcador microsatélite como describieron Navarro et al. (2008). Para la puesta a punto de cada PCR múltiplex se utilizaron las muestras Resumen español 127 de prueba. Las condiciones PCR fueron las mismas que las utilizadas en la PCR simple inicial. Para comprobar la adecuada amplificación de los productos de PCR múltiplex se realizó una electroforesis cargando una alícuota del producto de PCR en un gel de agarosa al 2% durante 30 minutos (8v•cm−1). Tras comprobar su correcta amplificación, 1μl de la reacción (diluido al 75% con agua Milli-Q) se mezcló con 9,75μl de Hi-Di formamida y 0,25μl del marcador de peso molecular Gene Scan LIZ 500 (Applied Biosystem®) y se cargó en un secuenciador automático ABI Prism® 3130XL (Applied Biosystem®). Los electroferogramas resultantes se analizaron usando el programa informático GeneMapper (v.3.7) (Applied Biosystem®). 2.2 Análisis de segregación de deformidades esqueléticas en dorada (Sparus aurata L.); ausencia de opérculo, lordosis, fusión de vértebras y complejo LSK 2.2.1 Experimento 1 (puesta masal con criba) Se obtuvo un lote de huevos mediante puesta masal de un stock industrial de reproductores. Dicho stock estaba compuesto por 66 reproductores no deformes, y la proporción de sexos fue 1♀: 2♂. Los huevos y las larvas se cultivaron en las instalaciones del Parque Científico Tecnológico Marino de la Universidad de Las Palmas de Gran Canaria (PCTM-ULPGC, Gran Canaria, España) y bajo las condiciones descritas por Roo et al.(2009). Cuando los alevines contaban con 111 días de edad (1,9 ± 0,02 g) (promedio ± error estándar), fueron cribaros y aquellos que presentaron algún tipo de deformidad, se retiraron del lote, con la excepción de Resumen español 128 los individuos que presentaron complejo LSK (lordosis/escoliosis/cifosis) que se seleccionaron y se cultivaron separadamente en un tanque de fibra de vidrio de 1000 L. El porcentaje de individuos LSK encontrado en todo el lote fue 0,22%. Con respecto a los individuos considerados como normales, se transportaron a las instalaciones de la empresa Playa de Vargas 2001 S.L. (PLV2001, Gran Canaria, España) cuando contaban con 130 días de vida (4,8 ± 0,1 g). Estos peces fueron criados en condiciones industriales en una jaula y alimentado con pienso comercial de peces. El período de crecimiento duró hasta 509 días cuando los peces alcanzaron 419,7 ± 3,1 g. Al final del experimento, una muestra de 846 peces de PLV2001 se sacrificó y se analizó visualmente para determinar la presencia o ausencia de deformidades según AquaExcel-ATOL (AquaExcel Project, 2013; ATOL: 0000087). Dichas deformidades se evaluaron mediante observación directa del esqueleto axial, tras filetearlos. Las relaciones familiares entre los reproductores y la descendencia (846 de PLV2001 y 54 peces LSK del PCTM-ULPGC) se determinaron por el método de exclusión utilizando el software VITASSING (v8.2.1) (Vandeputte et al., 2006), tras el análisis de ADN y caracterización genética con la PCR múltiplex SMsa1 (SuperMultiplex Sparus aurata) descrita por Lee-Montero et al. (2013). 2.2.2 Experimento 2 (puesta masal sin criba) Se recogieron huevos en dos días consecutivos de una puesta masal de un lote industrial de reproductores del PCTM-ULPGC, en el marco del proyecto PROGENSA® Resumen español 129 (Afonso et al., 2012). Este lote de reproductores estaba compuesto por 59 reproductores no deformes, con una proporción de 1♀: 1.81♂. Los huevos y las larvas se cultivaron en las instalaciones de PCTM-ULPGC, tal como se describe en el Experimento 1. 179 días después de la eclosión (17,2 ± 0,2 g), los alevines se marcaron individualmente con Passive Integrated Transponder (PIT) (Trovan Daimler-Benz) siguiendo el protocolo descrito por Navarro et al. (2006) y se transportaron a una jaula de la empresa CANEXMAR S.L. (Gran Canaria, España). Se criaron bajo condiciones intensivas y se alimentaron con alimento comercial. Ningún proceso de criba se llevó a cabo durante el cultivo de los peces, el cual se prolongó hasta 689 días, cuando los peces pesaron 524,4 ± 12,6 g. Se realizó una evaluación visual inicial en el momento del marcaje (179 días), para determinar la presencia o ausencia de deformidades (ATOL: 0000087). Al final del experimento (689 días), una muestra de 810 peces se sacrificaron y se analizaron visualmente (análisis visual final), para determinar la presencia o ausencia de deformidades (ATOL: 0000087). Al igual que en el Experimento 1, se evaluaron las deformidades vertebrales directamente después del fileteado del pescado. Las relaciones familiares entre los reproductores y la descendencia, se determinaron como se describe en el Experimento 1. Resumen español 130 2.2.3 Experimento 3 (cruces dirigidos) En las instalaciones del PCTM-ULPGC se establecieron diferentes cruces dirigidos con reproductores de dorada, más concretamente, 15 machos deformes y 10 hembras normales (N). Los reproductores deformes se clasificaron en tres grupos: ausencia de opérculo (O), lordosis (L) y fusión de vértebra (FV) según la deformidad que presentaban. Se constituyeron tres apareamientos (réplicas) por cada deformidad: NxO, NxL y NxFV. La proporción de sexos en cada tanque fue una hembra normal con dos machos deformes (1♀N: 2♂D). Además, una puesta masal con reproductores no deformes (1♀N: 2♂N) fue utilizada para conformar un cruce control (C). Los huevos de cada apareamiento se cultivaron cada uno separadamente, en las instalaciones de PCTM-ULPGC, tal como se describe en el Experimento 1. Los alevines de cada apareamiento se criaron por separado en tanques de fibra de vidrio de 1000 L hasta los 129 días (9,6 ± 0,1 g de peso). Al final del experimento, se analizaron visualmente todos los peces procedentes de los cuatro cruces (11.503 individuos) para determinar la presencia o ausencia de deformidades esqueléticas (ATOL: 0000087). 2.2.4 Análisis estadístico La asociación entre cualquier factor (reproductor, familia, cruce o experimento) y deformidad, se analizó con un modelo loglineal utilizando el programa estadístico SPSS (v18 PASW Statistics). El modelo loglineal da el significado de cualquier factor de deformidad (su prevalencia [i]) contra cualquier factor biológico o funcional (dependiendo de los datos [j]), organizado en una tabla de contingencia bidireccional, a través de los valores normalizados o Resumen español 131 valores Z. Los valores Z >1,96 ó Z<-1,96indican un exceso o defecto significativo de la deformidad con respecto a la familia, reproductor, cruce o experimento, respectivamente. lnƒij= µ + α i+ β j+ αβ ij Donde lnƒij es la frecuencia esperada de cada deformidad (i) en cada familia, reproductor, cruce o experimento (j) considerado (ij); µ es el valor medio de los logaritmos de las frecuencias esperadas, α i es el efecto del factor de deformidad (i), β j es el efecto de la familia, reproductor, cruce o experimento (j) y αβ ij es el efecto debido a la interacción de estos factores. En los Experimentos 1 y 2 se analizó la relación entre cada deformidad y todas las familias, machos y hembras. En el Experimento 3, se analizó la relación entre cada deformidad y el cruce, las familias y los tanques. 2.3 Identificación de QTLs asociados a deformidades esqueléticas en dorada (Sparus aurata L.): ausencia de opérculo, anomalía mandibular, lordosis, fusión de vértebra y complejo LSK 2.3.1 Material biológico, caracteres analizados y asignación parental Estudio 1 El diseño de los cruces de reproductores, las condiciones de cultivo, el análisis fenotípico y la asignación parental están descritos con detalle en Negrín-Báez et al. (2014b).