Mapping the porcine RN gene to chromosome 15

Des marqueurs microsatellites ont ete utilises pour localiser le gene RN du porc sur le chromosome 15 dans une etude anterieure. Ce locus influence la qualite technologique de la viande. Les familles etudiees ici ont ete construites par accouplement de 16 verrats croises (Hampshire x Pietrain) avec 65 truies croisees (Large White x Landrace). Des echantillons de viande de 564 descendants ont ete extraits du muscle longissimus dorsi 24 heures post mortem pour mesurer la concentration en glycogene. La determination du genotype RN est basee sur la valeur de la concentration en glycogene. Les verrats etaient heterozygotes (RN - /rn + ) et les truies homozygotes (rn + /rn + ) au locus RN. Une sous-population de 263 animaux a ete typee a l'aide de quatre marqueurs microsatellites du chromosome 15. L'analyse des liaisons genetiques confirme que le locus RN est localise sur le chromosome 15 et montre qu'il est situe a une distance de 4 cM du marqueur Sw936.


INTRODUCTION
A new meat quality trait was defined by Naveau et al (1985), the 'rendement technologique Napole' (RTN). This parameter describes the technological yield of 'Paris ham' processing and is correlated with the glycolytic potential in muscle tissue. Complex segregation analysis (Le Roy et al, 1990) showed that this trait is influenced by a major gene, called the RN gene, with two alleles, the dominant RN-allele and the recessive rn + allele. Studies by Feddern (1994) revealed that the RN-allele is present in the Hampshire line of a German crossbreding programme, but not in the Pi6train, Large White and Landrace lines. In the presence of the RN-allele an economic loss of 1% of the carcass value was estimated when meat is cooked. Furthermore it was revealed that the glycogen concentration in muscle tissue is highly correlated with the glycolytic potential and therefore a good parameter for characterizing the RN locus.
Nevertheless a DNA test would simplify the determination of RN genotypes. The localization of the RN locus within the pig genome is the first step to establish such a test. During our study Milan et al (1995) mapped the RN locus to chromosome 15 at a distance of 18 cM from the marker S0088. To confirm this result we screened resource families with highly polymorphic microsatellite markers from chromosome 15. This paper reports the first results of this investigation.

Animals
Resource families were created by mating 16 crossbred boars (Hampshire x Pi6train) with 65 crossbred sows (Large White x Landrace). Each boar was mated to four to seven sows and some sows were mated to two boars. The resulting 564 offspring were fattened in the experimental pig unit Hohenschulen of the Christian-Albrechts University, Kiel, and slaughtered when they reached approximately 107 kg liveweight. Feddern (1994) demonstrated the presence of the RN-allele in the Hampshire line, while this allele was absent in the Pi6train, Large White and Landrace lines.
Thus we expected the crossbred boars to be heterozygous (RN-/rn + ) and the crossbred sows to be homozygous (rn + /rn + ) at the RN-locus. We expected an equal proportion of both possible genotypes for the offspring.

Glycogen phenotypes and RN genotypes
Meat samples were drawn from 564 offspring at 24 h post mortem out of the longissimus dorsi and stored at -20°C for subsequent analysis of muscle glycogen concentration, because this trait has been shown to be closely related to RTN (Le Roy et al, 1996). Thawed samples were analysed with the iodine-binding method of Dreiling et al (1987). This method consists of three steps: tissue preparation, incubation with an iodine solution, and determination of the optical density.
RN genotypes for linkage analysis were determined from the muscle glycogen concentration of the animals. Sire genotypes were determined from the distribution of the glycogen concentration within full-sib families. A bimodal distribution within families gave us strong evidence that the sires were heterozygous (RN-/rn + ), except for two, at the RN locus. Previous analyses of Feddern (1994) showed that crossbred animals of the dam line had a low glycogen concentration with a unimodal distribution (n = 71). Thus we confirmed that the dams of our experiment had the rn + /rn + genotype. The genotypes of the offspring were determined by their muscle glycogen concentrations. The heterozygous genotype (RN-/rn + ) was attached to animals with a high glycogen concentration (> 80 J1.mol/ g). Animals with a low glycogen concentration (< 15 J1.mol/g) were classified as rn + /rn + . These RNgenotypes were used for linkage analysis and animals with a glycogen concentration between 15 and 80 J1.mol/g were classified as uncertain genotypes and excluded from the study. The thresholds 15 and 80 J1.mol/g were chosen by visual inspection of the glycogen distribution. Of course these thresholds are somewhat arbitrary but analyses with different thresholds showed that results were not greatly modified.

Marker genotypes and linkage analysis
A subsample of 263 animals was genotyped with four microsatellite markers from chromosomes 15, which were selected from the USDA map (Rohrer et al, 1994 and the corresponding database) and a map reported by Ellegren et al (1994). Primers were synthesized and labelled with fluorescein. PCR amplifications were carried out on a 9600 Perkin-Elmer-Cetus thermal cycler in a microtitre format. Four different annealing temperatures and four different numbers of cycles were tried to optimize the reaction conditions for each marker. PCR products were analysed with an automated laser fluorescence detection system (ALF, Pharmacia). PCR fragments with defined lengths were used as internal standards flanking the alleles closely. One animal was analysed on each gel with the corresponding marker and used as external standard to ensure that genotypes were comparable between gels. Offspring were always analysed with their parents on the same gel in order to avoid typing errors.
The size of the amplified fragments was determined with the ALF fragment manager software. In conjunction with the ALF fragment manager, the automated linkage preprocessor (ALP) software (Mansfield et al, 1994) converts raw data into genotypic data. This system was used to size the microsatellite alleles, check Mendelian inheritance of the markers used and create an output file for linkage analysis. Alleles were defined by the size of the amplified fragments. Genotypic data were stored in a database.
Pairwise linkage analyses between the analysed marker and the RN locus and a multipoint linkage analysis were performed with the CRIMAP program (Green et al, 1990). Because of the assumed heterozygosity of the boars and homozygosity of the sows at the RN locus, only male meioses were informative for linkage analysis.

Glycogen distribution
The bimodal distribution of the muscle tissue glycogen concentration (fig 1) supports the hypothesis of Le Roy et al (1994) that the RN locus is a major locus with two alleles and our hypothesis that the boars were heterozygous at the RN locus. From the presence of both alleles in almost all the families it can be concluded that the RN-alleles has a high frequency in the Hampshire population analysed.

Linkage between the markers and the RN locus
In the present study we could confirm the localization of the RN locus on chromosome 15. All chromosome 15 markers were linked to the RN locus (table I).
The number of informative meioses varied between 167 and 336. Lod scores exceed the value of 3.0 for all markers. Genetic linkage between the porcine microsatellite markers Sw936 and the RN locus was demonstrated at a distance of 4 cM with a maximum lod score of 37.51 (table I). The RN locus is also linked to the marker S0088 at a distance of 20 cM, Sw964 at a distance of 21.2 cM and Swr312 at a distance of 31.6 cM.
The calculations are based only on male meioses because dams were obviously homozygous rn + /rn + at the RN locus and consequently not informative for linkage analysis. Thus a larger sex-averaged distance between the marker Sw936 and the RN locus can be assumed, because of a higher recombination rate in females than in males (Ellegren et al, 1994). The distance of 20 cM between the marker S0088 and the RN locus is nearly the same as calculated by Milan et al (1995).
The order of the markers presented in the multipoint linkage map (fig 2) is in agreement with the map published by Rohrer et al (1994), noting however that Swr312 has been mapped to chromosome 15 later. An important difference between the two maps exists only for the interval between markers S0088 and Sw964.
The determination of RN genotypes from glycogen measurements seems to be correct because misclassifications would have led to increased recombination estimates. Thus the measurement of glycogen in the muscle seems to be a good parameter to describe the RN locus. Nevertheless this method is relatively labourintensive and expensive.
A marker flanking the RN locus at a recombination fraction of 0.04 offers the possibility of applying a positional cloning and a comparative mapping strategy. Although the gene product of the RN locus is still unknown, it is known from the glycogen measurements that the gene product influences the glycogen metabolism. Furthermore, comparative gene mapping studies (Johansson et al, 1995;Rettenberger et al, 1995) showed that the porcine chromosome 15 is homologous to the and the development of a PCR-based test would offer the possibility of selecting for the RN locus in breeding programmes. Additionally to the measurement of the glycogen concentration, growth and carcass traits were recorded. Further analysis will reveal if the RN locus influences other economically important traits in pigs.