THÈME DE RECHERCHE : Le phénotype callipyge est une hypertrophie musculaire généralisée post-natale décrite chez le mouton (introduit dans le Chapitre 1). Son mode de transmission non mendélien, qualifié de surdominance polaire, est unique : seuls les hétérozygotes ayant reçu la mutation callipyge (CLPG) de leur père expriment le phénotype. La mutation CLPG, localisée dans un domaine soumis à l'empreinte parentale, est une mutation ponctuelle (SNP) détruisant un élément qui contrôle, en cis, le taux dexpression musculaire des gènes voisins. L'hypertrophie musculaire, sans doute causée par l'expression ectopique de la protéine DLK1 chez les animaux +mat/Cpat, n'est pas observée chez les animaux Cmat/Cpat dont l'allèle maternel apparaît trans-inactiver la traduction du messager DLK1. Cette thèse a pour objectif d'approfondir notre compréhension de la surdominance polaire, aussi bien au travers de l'étude des mécanismes impliqués dans l'effet de contrôle à longue distance en cis de la mutation que de la caractérisation des inhibitions en trans induites par les transcrits maternels sur les messagers paternels. RÉSULTATS : L'étude de l'effet en cis (Chapitre 2) a démontré que la mutation CLPG se différenciait de l'allèle sauvage par au moins trois marques épigénétiques distinctes : (i) l'hypométhylation de l'ADN à proximité du SNPCLPG, (ii) la création d'un site d'hypersensibilité à la DNase, et (iii) l'activation de la transcription bidirectionnelle de la région autour de la mutation. En démontrant que la mutation inactivait bel et bien un élément de contrôle à longue distance, ces données nous ont permis d'élaborer un modèle plus précis de la surdominance polaire. L'étude de la trans-inhibition des transcrits protéiques à expression paternelle par les transcrits non codants maternels (Chapitres 3 et 4) se fondait sur l'hypothèse d'une implication des nombreux microARNs (miRNAs) du domaine DLK1-GTL2. Nous avons ainsi pu montrer que les cinq miRNAs abrités par anti-PEG11 étaient capables, par leur parfaite complémentarité de séquence, de cliver le messager PEG11 (Chapitre 3). Bien que le rôle de la protéine PEG11 dans le phénotype callipyge soit inconnu, cette étude a néanmoins démontré l'existence d'une trans-inhibition entre les deux allèles du locus, tout en identifiant le premier cas de dégradation miRNA/cible impliquant des gènes soumis à l'empreinte chez les mammifères. L'étude de la trans-inhibition appliquée au messager DLK1 (Chapitre 4) a quant à elle nécessité la génération par séquençage haut débit d'un catalogue exhaustif des miRNAs du muscle squelettique ovin. Ce travail nous a permis de démontrer que les miRNAs du domaine DLK1-GTL2 étaient bien exprimés maternellement et eux aussi soumis à l'effet en cis de la mutation. Si aucun miRNA capable d'inhiber le messager DLK1 n'a été identifié sans ambiguïté, nos analyses d'affinité ont toutefois révélé un effet significatif de miR-376c sur sa 3'UTR, ainsi que de l'ensemble des miRNAs du locus DLK1-GTL2 sur sa région codante. Notons que nous avons également considéré comme potentiels candidats à la trans-inhibition les snoRNAs du locus, de même que les miRNAs du locus soumis à l'édition ARN. CONCLUSIONS ET PERSPECTIVES : À l'issue de cette thèse, nous avons donc contribué à la compréhension du phénomène de surdominance polaire au locus DLK1-GTL2 du mouton callipyge. Si de nombreuses questions restent en suspend, elles devraient néanmoins trouver réponse grâce aux nouveaux outils d'analyse en cours de développement. Ainsi, deux lignées de souris transgéniques développées dans notre laboratoire, respectivement porteuses de la mutation CLPG ou surexprimant PEG11 dans le muscle, éclaireront les mécanismes moléculaires de l'effet en cis de la mutation ainsi que le rôle de PEG11 dans l'établissement du phénotype callipyge. Par ailleurs, pour confirmer in vivo l'effet inhibiteur sur DLK1 des miRNAs identifiés dans notre étude, les progrès apportés par le séquençage haut débit (notamment le HITS-CLIP) et la transfection en modèle cellulaire se révéleront sans doute d'une grande aide. / RESEARCH OVERVIEW: The callipyge phenotype is a generalized muscular hypertrophy described in sheep (see Chapter 1). It features a non-mendelian mode of inheritance termed polar overdominance: only heterozygous animals having inherited the mutation from their father display the phenotype. The callipyge mutation (CLPG), which falls in an imprinted domain, is a single-nucleotide polymorphism (SNP) that disrupts a putative long-range control element affecting, in cis, the expression level of neighboring imprinted genes in skeletal muscle. The callipyge phenotype is thought to be caused by ectopic expression of DLK1 protein in +mat/Cpat animals. In contrast, Cmat/Cpat animals exhibit a wild-type phenotype that is certainly due to a trans-inhibition from the maternal allele on DLK1 translation. The objective of our thesis was to improve our knowledge of polar overdominance, both by studying mechanisms of long-range cis regulation and by characterizing the trans effect of maternal noncoding transcripts on the paternally-expressed messenger RNAs. RESULTS: Our study of the cis effect (Chapter 2) showed that the CLPG mutation differs from the wild-type allele by at least three distinct epigenetic marks: (i) DNA hypomethylation in the vicinity of the SNPCLPG, (ii) creation of a DNase-hypersensitive site, and (iii) activation of a bidirectional transcription start site centered on the mutation. Altogether, our data provided strong evidence for the SNPCLPG inactivating a long-range control element and allowed us to refine our model of polar overdominance. Our working hypothesis for the trans-inhibition of paternally-expressed genes by maternal noncoding transcripts (Chapters 3 and 4) involved microRNAs (miRNAs) from the DLK1-GTL2 domain. In this respect, we showed that five miRNAs from anti-PEG11 were able to cleave PEG11 transcript, owing to perfect sequence complementarity (Chapter 3). This study was the first demonstration of miRNA-mediated RNAi involving imprinted genes in mammals. Furthermore, it allowed us to confirm the existence of a trans-inhibition between both alleles in the domain, albeit the role of PEG11 protein in the callipyge phenotype is still unknown. To study the trans-inhibition of DLK1 messenger (Chapter 4), we used high-throughput sequencing to build an exhaustive catalogue of skeletal muscle-specific miRNAs in sheep. Our analyses showed that miRNAs from the DLK1-GTL2 domain are maternally expressed and affected by the cis effect of the CLPG mutation. Even if we could not find miRNAs unambiguously able to repress DLK1 transcript, affinity analyses revealed a significant effect on its 3' UTR for miR-376c, as well as on its coding region for all miRNAs considered as team. Of note, we also investigated snoRNAs and miRNAs subjected to RNA editing in the DLK1-GTL2 locus. CONCLUSIONS AND PERSPECTIVES: During the course of this research, we contributed to improve the understanding of polar overdominance in the sheep DLK1-GTL2 locus. Although many questions remain, most will eventually be answered thanks to upcoming analysis tools. Hence, our lab has already generated two transgenic mouse lines, either carrying the CLPG mutation or over-expressing PEG11 in skeletal muscle. These mice should respectively shed light on the molecular mechanisms underlying the cis effect of the CLPG mutation and on the role of PEG11 in the callipyge phenotype. Finally, to confirm inhibiting effects of miRNAs identified in our study, technological improvements granted by high-throughput sequencing (such as HITS-CLIP) and miRNAs transfection in cell model systems will both prove to be very useful.

An inherited muscle hypertrophy in sheep is caused by the callipyge gene (CLPG) located on ovine chromosome 18. It has been suggested that this gene is a single autosomal dominant gene. Animals expressing the callipyge gene show increased muscling in the pelvic and torso regions of the body and, thus, have been given the phenotype- descriptive name callipyge (from Greek calli-beautiful; pyge-buttocks). In this study 21 wether lambs, the offspring of callipyge rams (genotype CLPG /clpg) and normal Rambouillet ewes (genotype clpg/clpg), were used to determine the difference s in muscle fiber type percentages, composition, and tenderness between normal and callipyge lambs . Eleven of these lambs showed muscle hypertrophy typical of the callipyge phenotype; 10 were classified as normal . Several histochemical, biochemical , and physical measures were examined in order to study changes in the physiology and biochemistry of some economically important muscles. When compared to normal lambs, the callipyge lambs have a larger (P < .05) average percentage of fast-twitch glycolytic (FG) muscle fibers and smaller average percentages of fast-twitch oxidative and glycolytic (FOG) and slow-twitch oxidative (SO) muscle fibers in both the longissimus and gluteus medius muscles. The diameter of the fast-twitch muscle fibers was larger in the callipyge group, but slow-twitch fibers were smaller than those of normal lambs. No differences were observed in the supraspinatus muscle of the normal and callipyge groups. Thus there is an indication of a differential effect of the callipyge gene among muscles of the callipyge lambs. The semitendinosis muscles of both the callipyge and normal groups were dissected from the carcasses. This muscle was larger (P < .01) and contained more protein in the callipyge lambs than in the normal lambs. However, the callipyge semitendinosis muscle did not have a significantly higher content of DNA than the normal lamb semitendinosis, suggesting that the muscle hypertrophy is not associated with an increase in muscle nuclei. The protein-to-DNA ratio was larger (P < .05) in the semitendinosis muscle of callipyge lambs than in the normal lambs. Protein-to-RNA and RNA-to-DNA ratios were similar; this suggests that the semitendinosis muscle was enlarged without increased translational or transcriptional activity. Samples from the callipyge longissimus and gluteus medius muscles had RNA, DNA, and protein ratios similar to those of the semitendinosis muscle, suggesting a similar mode of action for muscle enlargement in other muscles affected by this gene. Loin chops from the callipyge lambs had lower tenderness scores (P < .01) as measured by the Wamer-Bratzler shear force and myofibril fragmentation index (MFI). However, aging increased MFI scores and decreased shear scores (P < .0 l) of the callipyge lamb chops . The normal lamb chops also had decreased shear and increased MFI scores following the aging period. The loin chops from the callipyge lambs also tended to be less red (P < .1) than chops from normal animals as measured by Hunter 'L,' 'a,' and 'b' colorimeter scores.

The muscle hypertrophy of lambs expressing the Callipyge phenotype is possibly linked to characteristics of their muscle satellite cells. Therefore, characteristics (proliferation, fusion %, and protein accretion) of cultured satellite cells isolated from the longissimus muscle of Callipyge (n = 3) and normal (n = 3) lambs were compared in this study. In the first experiment, we tested whether or not the lll proliferation rates differ for satellite cells isolated from Callipyge or normal sheep when cultured in the presence of different serum types (horse, normal lamb, or Callipyge lamb). The average population doubling time (PDT, h) during log phase growth was calculated for cells from each animal grown in each serum type. Population doubling time was not affected (P > .1) by the interaction of satellite cell type with serum type, or by satellite cell type. Unexpectedly, PDT was longer (P < .05) for satellite cells grown in Callipyge serum (22 h) than for cells grown in normal sheep serum (20 h) or horse serum (18 h). These results suggest that muscle hypertrophy of Callipyge lambs is not linked to intrinsic differences in satellite cell proliferation, although hypertrophy may be associated with a decreased proliferation induced by a factor in Callipyge serum. In the second experiment, we tested whether cell fusion, or protein accretion differ for cultured satellite cells isolated from Callipyge or normal sheep. DNA and protein were determined at 24, 48, 72, and 96 h after satellite cell cultures were induced to differentiate. Fusion percentage was determined in a Giemsa stained plate after 72 h in differentiation medium (Dulbecco's Modified Eagle Medium containing 1.5% of horse serum). Callipyge cultures tended (P = .14) to have higher fusion% than normal cultures exhibited, suggesting that muscle hypertrophy of Callipyge lambs may be linked to an increased tendency of satellite cells to fuse. Protein content (μg/well) and protein:DNA ratio (ng of protein/ng of DNA) were not affected by satellite cell type (P = .80 and P = .79, respectively). Thus, there was no evidence for a link between increased protein accretion and Callipyge hypertrophy.