Skeletal myogenesis is under the control of a combinatorial network of transcription factors. It has been shown that the homeobox protein Six1 is required for embryonic myogenesis. Using functional genomics approaches, I determined that Six1 is required for myoblasts differentiation through direct binding to a cluster of genes that are related to muscle function and muscle structure during my Master’s studies. However, it was still not fully understood how Six1 selects its genomic targets and whether Six1 regulates the expression of Myod directly. I devoted my PhD work to study three central aspects of Six1 function: through what DNA motif it binds to DNA, how it regulates the expression of the myogenic regulatory factor MyoD, and how it might regulate chromatin structure at the enhancer regions of muscle genes. A more degenerate MEF3-like DNA sequence consensus has been identified from Six1 ChIP-on-chip experiments. This MEF3 motif was further optimized using bioinformatic methods and was proved to discover Six1 binding sites with improved specificity and sensitivity. Myod, a member of myogenic regulatory factors (MRFs), is a master regulator in the myogenic lineage. Multiple MEF3 sites were identified on the regulatory regions of Myod, including two MEF3 sites within its core enhancer region (CER). Six1 was able to bind to the CER directly through these two MEF3 sites and regulated the Myod expression in cultured primary myoblasts. Previous work has suggested that the CER is also bound by Myod in myoblasts. I demonstrated that the binding of Myod to the CER depended on the presence of Six1. Six1 was also involved in maintaining a relatively ‘open’ chromatin structure at the CER, suggesting that Six1 may play a direct or indirect role in chromatin remodeling. During my Master’s studies, I demonstrated a synergistic regulation by the Six and MRF families. This synergistic function gains potential importance by the fact that ~25% of Six1 genomic targets are also bound by Myod. I decided to study whether the co-occupancy of Six1 and Myod was essential to maintain the proper global chromatin structure at these loci. Six1 and Myod co-bound genomic regions correlated with more accessible chromatin, which was detected by the formaldehyde-assisted isolation of regulatory elements (FAIRE) assay followed by DNA deep sequencing (FAIRE-seq). When combined with small interfering RNA-mediated gene knockdown of Six1 or Myod, FAIRE-seq data suggested that Six1, but not Myod, was involved in regulating the chromatin accessibility at these co-bound DNA loci. To shed light on the mechanism by which Six1 functions, proteomics approaches were used and revealed that proteins involved in “regulation of transcription” and “chromatin organization” were enriched among Six1-bound proteins. Cdk9 and its partner cyclin T have been shown to stimulate gene expression by releasing RNA polymerase II from transcriptional pause, but they can also function at gene enhancers. I determined that Six1 and Cdk9 participated in the same protein complex, and that the Cdk9 activity appeared to mediate the effect of Six1 on the chromatin accessibility at the CER to regulate the Myod expression. Taken together, these results demonstrate that Six1 regulates the expression of Myod through its direct binding on the CER which facilitates transcriptional elongation.

Myogenesis is the process of skeletal muscle tissue formation where committed muscle progenitor cells differentiate into skeletal muscle fibres. Depending on the instructive cues the muscle progenitor cells receive they will differentiate into specific fibre types with different properties. The skeletal muscle fibres can be broadly classified as fast-twitch fibres or slow-twitch fibres, based on their contractile speed. However, subgroups of fast- and slow-twitch fibres with different metabolic properties, endurance and different isoforms of sarcomeric components have also been identified, adding complexity to the process of muscle tissue patterning. The skeletal muscle tissue has the capacity to regenerate throughout life. Upon muscle tissue damage muscle satellite cells are recruited to the area of injury where they proliferate and either form new fibres similar to those damaged, or fuse with existing fibres. This thesis aims to investigate the process of muscle progenitor cell proliferation and differentiation, as well as the fast-twitch fibre formation and muscle tissue patterning in the zebrafish embryo. I present results identifying the previously uncharacterised gene myl1, encoding an alkali-like myosin light chain, which is specifically expressed in fast-twitch muscle progenitors before fibre formation. Furthermore, I introduce data showing that the transcription factor six1 is expressed in Pax7+ muscle progenitor cells, which has been reported to contribute to part of the fast-twitch muscle tissue as well as to a pool of quiescent muscle satellite cells. With support from the presented data, I hypothesise that six1 keeps the Pax7+ muscle progenitor cells in a proliferative state and consequently prevents them from differentiating into muscle fibres. In addition, I demonstrate that the zebrafish fast-twitch muscle fibres can be divided into different subgroups that express unique forms of fast myosin heavy chain genes along the anterior-posterior (head-tail) axis, and that this subspecification depends on a balance between RA and Wnt signalling. Collectively I propose a previously unknown role for Six1 in zebrafish Pax7+ muscle progenitor cell proliferation and differentiation. Furthermore, I present novel data suggesting that distinct regions of the zebrafish body musculature are composed of different fast-twitch fibre types, and that this regionalisation is conserved in adult zebrafish.

The transcription factor Six1 of the sine oculis homeobox family has been tied to skeletal muscle formation. Work completed thus far has allowed our research team to identify the precise mechanism by which Six1 regulates the expression of MyoD, a key myogenic gene, in muscle stem cells. Furthermore, loss-of-function of this protein, mediated by RNA interference, has implicated Six1 as essential towards normal myogenic differentiation. However, beyond Six1 and its involvement towards myogenesis, our data also suggests the transcription factor as a potential regulator of the cell cycle. Data from our lab shows that loss of Six1 expression significantly impairs primary myoblast proliferation and appears to impair satellite cell activation in response to muscle injury in vivo. Furthermore, loss of Six1 decreases the expression of key cell cycle genes. Combining functional genomics approaches such as ChIP-Seq and Gene Expression Profiling together with Gene Ontology Term Enrichment shows a significant representation for biological processes regarding the cell cycle and its regulation; these biological clusters contain a large subset of genes that are bound and modulated by Six1. In particular, Ccnd1 was found to display a similar expression pattern as Six1 in growing myoblasts and its expression was found to be directly controlled by Six1. Furthermore, Ccnd1 over-expression was sufficient to rescue the Six1-knockdown associated cell cycle phenotype. Together, these data suggest that in response to injury Six1 enhances the expression of the cell cycle gene Ccnd1 thus modulating myoblast proliferation for muscle regeneration.

京都大学 / 0048 / 新制・論文博士 / 博士(医学) / 乙第13219号 / 論医博第2166号 / 新制||医||1033(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 妹尾 浩, 教授 小川 修, 教授 戸口田 淳也 / 学位規則第4条第2項該当 / Doctor of Medical Science / Kyoto University / DFAM

Les homéoprotéines SIX sont codées par les gènes Sine oculis homeobox related genes Six1 à Six6 chez les vertébrés parmi lesquels Six1, Six2, Six4 et Six5 sont exprimés dans le lignage myogénique. Bien que Six1 et Six4 soient requis pour la myogenèse hypaxiale, les animaux doubles KO pour ces deux gènes (s1s4KO) forment leurs muscles épaxiaux et craniofaciaux. Nous avons caractérisé le phénotype de mutants composites des gènes Six et avons montré que l’absence de Six1 et Six2 empêchait la formation des muscles craniofaciaux et empirait les défauts de formation des muscles des membres observés chez les fœtus mutants pour Six1. Nous avons aussi observé que les fœtus dépourvus d’activité de SIX1, SIX2, SIX4 et SIX5 étaient toujours capables de former leurs muscles épaxiaux, mais que l’expression de Pax7 dans leurs progéniteurs myogéniques était fortement diminuée et mêlée à l’expression de Myogénine. Alors que les fœtus s1s4KO forment des muscles épaxiaux, leurs cellules PAX7+ ont un défaut de nichage entre la membrane plasmique des myofibres et la lame basale qui les entoure. Nos analyses transcriptomiques, nos expériences de transplantation et nos études in vitro nous ont permis de conclure que le nichage des cellules PAX7+ nécessitait un environnement adéquat combinant des propriétés des myofibres et des cellules PAX7+ ; environnement perturbé dans les muscles épaxiaux s1s4KO. Nos expériences de transplantation nous ont aussi permis de conclure que Six1 et Six4 étaient requis pour une bonne ré-innervation des myofibres après blessure et pour la mise en place du phénotype rapide de ces myofibres. De plus, les muscles transplantés avec des cellules PAX7+ fœtales s1s4KO après blessure se reforment d’un grand nombre de petites myofibres. Nous avons pu relier ce phénotype au comportement des cellules s1s4KO in vitro où elles montrent un défaut de fusion. Enfin, les homéoprotéines SIX ont besoin de co-facteurs pour induire l’expression de leurs gènes cibles, tels que les protéines EYA codées par les gènes Eya1 à Eya4 chez les vertébrés. Eya3 et Eya4 sont fortement exprimés dans les cellules satellite au cours de la régénération, cellules qui requièrent aussi Six1 pour une réparation musculaire efficace. Nous avons étudié la régénération musculaire en absence d’expression d’Eya3 et n’avons pas observé de défaut nous menant à la conclusion qu’Eya3 n’est pas requis pour la régénération musculaire adulte, mais que sa perte d’expression était peut-être compensée par un autre gène Eya chez les animaux mutants. Pour conclure, Six1 et Six2 sont indispensables à la formation des muscles craniofaciaux, et Six1 et Six4 sont requis pour la myogenèse hypaxiale, et pour l’établissement d’un environnement propice à la maturation des myofibres fœtales et au nichage des cellules PAX7+ au cours de la myogenèse épaxiale, et permettant la croissance des myofibres et leur ré-innervation après blessure. La collaboration des protéines SIX avec leurs co-facteurs EYA au cours de la myogenèse nécessite d’autres études pour mieux définir leurs fonctions. / SIX homeoproteins are encoded by the Sine oculis homeobox related genes Six1 to Six6 in vertebrates among which Six1, Six2, Six4 and Six5 are expressed in the muscle lineage. Whereas Six1 and Six4 are required for hypaxial myogenesis, double KO for those two genes (s1s4KO) still form their epaxial and craniofacial muscles. We further characterized the phenotype of compound Six mutant embryos and showed that the absence of Six1 and Six2 completely impairs craniofacial myogenesis and worsen muscle limb development observed in single Six1 mutants. We also showed that mouse fetuses devoid of SIX1, SIX2, SIX4 and SIX5 activity are still able to develop epaxial muscles, but that Pax7 expression in myogenic progenitors of these mutants is reduced and intermingled with Myogenin expression. While s1s4KO fetuses still develop epaxial muscles, their PAX7+ cells show a perturbed homing process into their niche, between the plasma membrane of a myofibre and the basal lamina surrounding it. Transcriptomic analysis, transplantation experiments and in vitro studies allowed us to conclude that the homing of PAX7+ cells into their niche during fetal myogenesis requires an adequate environment combining properties of the myofibers and the PAX7+ cells; environment disturbed in s1s4KO epaxial muscles. Transplantation experiments also led us to conclude that Six1 and Six4 are required for proper myofiber re-inervation after injury and for the establishment of the fast phenotype of myofibers. Furthermore, muscles transplanted with s1s4KO fetal PAX7+ cells after injury are formed of numerous and tiny myofibers. We could link this phenotype to the behavior of s1s4KO cells in vitro where they showed perturbed fusion. Finally, SIX homeoproteins require co-factors to induce their target genes expression, as EYA proteins encoded by Eya1 to Eya4 in vertebrates. Eya3 and Eya4 are strongly expressed in satellite cells during regeneration, cells in which Six1 is also required for proper muscle repair. We investigated muscle regeneration in absence of Eya3 expression and observed no obvious phenotype. We concluded that Eya3 is not required for muscle regeneration but that other Eya genes might compensate its function in KO mouse. To conclude, Six1 and Six2 are required for craniofacial myogenesis and Six1 and Six4 for hypaxial myogenesis and for the establishment of a proper environment allowing myofibre maturation and PAX7+ cells homing during fetal epaxial myogenesis and enabling myofibre growth and re-innervation after injury. The role of the collaboration between SIX and EYA proteins during myogenesis still needs more investigation.

La crête neurale (CN) est une structure multipotente transitoire de l'embryon de vertébrés. La CN céphalique (CNC), mais pas la CN troncale (CNT), fournit des tissus mésenchymateux (squelette, derme et tissus adipeux de la face). Cette capacité de la CNC est liée à l'absence d'expression des gènes de type Hox. Cependant, les cellules de la CNT possèdent des potentialités mésenchymateuses à l'état dormant, qui peuvent s'exprimer en culture. Les mécanismes moléculaires qui régulent les potentialités mésenchymateuses de la CN le long de l'axe antéro-postérieur restent incompris. Chez l'embryon d'oiseau, nous avons étudié l'influence des facteurs de transcription Hox et Six sur la formation du mésenchyme par la CN. D'une part, nos analyses in vivo et in vitro montrent que Six1 est présent dans des cellules mésenchymateuses de la CN et du mésoderme, suggérant un rôle dans le développement musculo-squelettique de la tête. D'autre part, nous avons testé l'hypothèse d'un rôle inhibiteur des facteurs Hox. Nos résultats montrent que l'expression ectopique de Hoxa2 dans les cellules de CNC en culture inhibe la production d'ostéoblastes, sans affecter celle des cellules nerveuses et mélanocytaires. Dans la CNT, nous avons trouvé que la différentiation osseuse, cartilagineuse et adipocytaire, est fortement réduite après la surexpression de Hoxa2, sans effet sur les autres phénotypes dérivés de la CN. Ces résultats suggèrent que les potentialités mésenchymateuses de la CN sont régulées, au moins en partie, par un mécanisme commun aux cellules de CNC et CNT, mettant en jeu une inhibition de l'activité du gène Hoxa2. / The neural crest (NC) is a transitory multipotent structure of the vertebrate embryo. The cephalic NC (CNC), not the trunk NC (TNC), gives rise to mesenchymal cell types (contributing to craniofacial skeleton, dermis and adipose tissue). This capacity of the CNC has been linked to the absence of Hox gene expression in the most rostral region of the embryo. However, TNC cells do have mesenchymal potentialities, although in a dormant state in vivo, but which can be disclosed after NC in vitro culture. The molecular mechanisms that regulate mesenchymal potentials of the NC cells along the rostral-caudal axis are still elusive. Here, we have used the avian embryo model to investigate the possible influence on NC mesenchymal fate, of Hox and Six transcription factor genes. On the one hand, in vivo and in vitro culture analyses show that Six1 gene is expressed in mesenchymal cell populations derived from both cranial NC and mesoderm, suggesting a role for Six1 in muscle-skeletal development in the head. On the other hand, we have tested the hypothesis of an inhibitory action of Hox genes on NC cell mesenchymal differentiation using NC in vitro cultures. In CNC cells, we found that ectopic expression of Hoxa2 strongly reduces the production of osteoblasts, while neural and melanocytic phenotypes are unaffected. In the cultured CNT cells, overexpression of Hoxa2 results in largely impaired differentiation into bone cells, chondrocytes and adipocytes, whereas other NC derivatives are unchanged. These results suggest that mesenchymal potentials of the CNC and TNC are controlled, at least in part, via a common mechanism that involves inhibition of Hoxa2 gene activity.

Wilms tumör (WT) är en malign snabbväxande tumör och den vanligaste solida buktumören hos barn under sex år. En kombination av operation, cytostatika och strålbehandling har lett till att 90 % av barnen idag blir helt botade. Enligt ett enhetligt europeiskt morfologiskt klassifikationssystem och behandlingsprotokoll behandlas alla barnpatienter med WT med cytostatika pre-operativt innan nefrektomi. Den histologiska tumörtypen, klassad efter operation, är helt avgörande för vidarebehandling av WT. Tumören består av tre celltyper, stromala epiteliala och blastemala, där de blastemala har en stor likhet med mesenkymala celler i njuren hos embryot. Tumörer med mer än 2/3 blastem klassificeras som högrisktumörer som följs upp med ytterligare, extra intensiv kemoterapi. I dagsläget görs en histologisk bedömning av tumörens riskgrupp endast på hematoxylin-eosin färgning. Proteinet SIX1 har nyligen rapporterats fungera som lämplig biomarkör för blastem. Med hjälp av immunohistokemi undersöktes i detta arbete om SIX1-färgning på blastem kan tillämpas kliniskt för säkrare riskbedömning av WT. Efter omfattande utvecklingsarbete med testning av olika varianter av förbehandling och detektionsmetod kunde ett fullgott protokoll för kliniskt bruk utformas. Detta kommer nu att implementeras inom klinisk patologi i Skåne. / Wilms tumor (WT) is a malign fast growing tumor and the most common solid abdominal tumor amongst children under the age of 6. A combination of surgery, chemotherapy and radiotherapy has led to a 90 % cured rate among children affected by WT. According to a uniform European morphological classification system and treatment protocol, all WTs receive pre-operative chemotherapy prior to nephrectomy. Histological tumor type, determined after surgery, is crucial for further treatment of WT. The tumor consists of three cell types: stromal elements, epithelial elements and blastema, of which the blastema has a great similarity to embryonic mesenchymal cells. Tumors that consist of more than 2/3 blastema are classified as high risk tumors, recieving additional intensive chemotherapy. Today, an assesment of WT histology is performed on hematoxylin-eosin stained specimens only. Recently, the SIX1 protein has been reported to function as a suitable biomarker for blastema. Using immunohistochemistry, it was investigated if SIX1 detection could be applied clinically for classification of WT. After several modifications of pre-treatment and primary antibody detection methodology, a clinically robust immunohistochemical staining protocol was finally developed. This will now be implemented in the clinic.

No description available.

Our understanding of protein lysine methyltransferases and their substrates remains limited despite their importance as regulators of the proteome. The SMYD (SET and MYND domain) methyltransferase family plays pivotal roles in various cellular processes, including transcriptional regulation and embryonic development. Among them, SMYD2 is associated with oesophageal squamous cell carcinoma, bladder cancer and leukemia as well as with embryonic development. Initially identified as a histone methyltransferase, SMYD2 was later reported to methylate p53, the retinoblastoma protein pRb and the estrogen receptor ERalpha and to regulate their activity. Our proteomic and biochemical analyses demonstrated that SMYD2 also methylates the molecular chaperone HSP90 on K209 and K615. We also showed that HSP90 methylation is regulated by HSP90 co-chaperones, pH, and the demethylase LSD1. Further methyltransferase assays demonstrated that SMYD2 methylates lysine K* in proteins which include the sequence [LFM]-₁-K*-[AFYMSHRK]+₁-[LYK]+₂. This motif allowed us to show that SMYD2 methylates the transcriptional co-repressor SIN3B, the RNA helicase DHX15 and the myogenic transcription factors SIX1 and SIX2. Finally, muscle cell models suggest that SMYD2 methyltransferase activity plays a role in preventing premature myogenic differentiation of proliferating myoblasts by repressing muscle-specific genes. Our work thus shows that SMYD2 methyltransferase activity targets a broad array of substrates in vitro and in situ and is regulated by intricate mechanisms.

Thyroid hormone (TH) is a critical signalling molecule for all vertebrate organisms, playing an especially crucial role in postembryonic development. Given its importance, many studies have focused on further elucidating the initial TH signal response and its method of transduction. Although the primary mechanism of TH response is genomic signalling, alternative mechanisms of early TH signal transduction have been relatively poorly studied. The North American bullfrog, Rana catesbeiana, is a useful model to study these early responses as tadpole post-embryonic development, or metamorphosis, can be experimentally induced through exposure to TH. The experimental induction of the TH signalling program leads to similar morphological endpoints as seen in natural metamorphosis in the transition of a tadpole to a juvenile froglet, such as regression of the tail. This TH-induced developmental program can also be manipulated through temperature where, as temperatures lower, developmental rate is delayed and at 5°C metamorphosis is completely stalled. Interestingly, when tadpoles exposed to TH at 5°C are introduced to permissive temperatures (24°C), an accelerated developmental program ensues, even when no more endogenous TH signal remains. Previous research has shown that this phenomenon can also be seen on the molecular level where only a select few transcripts have been shown to be responsive to TH at 5°C. However, the characteristic, if not augmented, TH response program is seen on the transcriptomic level when tadpoles are shifted to 24°C. This indicates that there is a molecular memory where the TH signal is induced in cold temperatures but not executed until more permissive temperatures arise. The extent and regulation of the transcriptomic program involved in this TH-induced molecular memory has yet to be understood. Herein we use the broader probing technique of RNA-seq analysis to identify potential components of the molecular memory. Eighty-one gene transcripts were TH-responsive at 5°C in cultured R. catesbeiana tail fin indicating that the molecular memory is more complex than previously thought. A number of these transcripts encoded regulators of transcription. Closer examination of select transcripts including a novel krüppel-like factor family member, klfX, at 5oC indicated that not all of the candidate molecular memory transcripts are regulated through active transcription and active translation is not required. When moved into 24°C an accelerated transcriptomic response occurred even when no additional TH is added, suggesting that a priming event occurs by TH exposure at 5°C allowing an accelerated metamorphosis at permissive temperatures. The molecular memory may be used as a means to isolate the initiating TH signalling response and the regulation of this program to allow further elucidation of early TH signalling in post-embryonic development. / Graduate