Despite the fact that some aspects of single studies differ, there's a generally accepted view that the whole cartilaginous viscerocranium of vertebrates is neural crest derived. By the series of isotopic transplantation experiments of presumptive neural crest on the model organism Ambystoma mexicanum I partly specify this oppinion and prove that the most ventro-caudal cartilage, the second basibranchial, is of a different origin. Furher I mention the level of the presumptive neural crest where the single parts of cartilaginous viscerocranium arise from. Moreover there is one element, the first basibranchial, which has double origin. I discuss also some other neural crest derivatives such as head and outer gills mesenchyme, the trabeculae cranii, part of the cartilaginous otic capsule or the connective tissue in the head. I have performed 179 transplantations between transgenic and normal axolotl embryos. My final analysis is composed of 65 embryos of stage 40 - 42 and 7 larvae of lenght of 15 - 17 mm.

The study of regeneration has inspired centuries of scientific research. Among the many model organisms used in present days, the axolotl has become the gold standard for studying limb regeneration in vertebrates. Limb regeneration is an intricated multi-step process that involves timely regulated events such as an immune response, dedifferentiation and migration of progenitor cells, and the re-establishment of the missing structures. Although axolotl limb regeneration has long been considered a case of perfect regeneration, the mechanisms whereby a regenerated limb is able to properly integrate into the mature tissue seamlessly have long been unstudied. In this work, I have investigated how the skeletal tissue is primed to successfully regenerate and restore its functionality. To properly achieve my research, I have first sought to understand the basic appendicular skeleton biology, particularly, how the skeletal elements change with growth and age. In the first part of my thesis, I show the changes associated with growth in the zeugopodial elements, radius and ulna. A cartilaginous skeleton defines the limb during the early life stages, and ossification starts around the time the animals reach sexual maturity, a process that extends throughout their lives. In the second part of my work, I have extensively described a regeneration-induced skeletal resorption. This process is carried out by osteoclasts and it is absolutely necessary for a successful integration of the skeletal tissue. Interestingly, a direct correlation between the resorption rate and the integration efficiency could be observed. Moreover, my work provides strong evidence linking the formation of the wound epithelium with the induction of skeletal resorption and the position of resorption with blastema formation. Altogether, this work provides a comprehensive study of axolotl skeleton biology in homeostasis and regeneration, with an emphasis on how a histolytic process primes the skeletal tissue for efficient regeneration. From a comparative perspective, the understanding of the events involved in axolotl limb regeneration provides an excellent platform for evaluating how regeneration could be enhanced in non-regenerative animals, such as humans.

The vertebrate tail is unique for each species and fulfils a broad spectrum of functions. In the axolotl (Ambystoma mexicanum), a tailed amphibian, the tail constitutes one-third of the full body length and is necessary for swimming. Despite its size, most of the tail's tissues are derived from the posterior neural plate of the neurula. Although giving rise to neuronal structures of the central nervous system along most of its length, the most posterior part of the neural plate develops preponderantly into presomitic mesoderm (PSM) which forms muscle, bone and cartilage of the tail and posterior trunk. During development, the posterior neural plate reverses its orientation during an anterior turn movement (Taniguchi et al., 2017). Cells of the most posterior plate region become now localised in an anterior position while previously more anterior neural plate cells land at a more posterior site. Simultaneously, the axial neural tube and notochord extend themselves posteriorly. The PSM, developing bilaterally to the central axis, is integrated into posterior tail expansion while forming new somites at its anterior end. It is still elusive which morphological changes the PSM undergoes to facilitate tail formation and posterior elongation of the embryo. Furthermore, it remains enigmatic in what way PSM cells change their shape, orientation, migration behaviour and distribution to meet the requirements needed for adjusting PSM and somite morphology. With homotopic tissue transplantations of posterior neural plate cells from a gfp-expressing donor to a white (d/d) recipient, enabled specific labelling of all mesodermal cells of the tail. Otherwise, mesodermal cells of the trunk and tail can not be distinguished, neither genetically nor morphologically. With this cell labelling approach, the entire tail mesoderm could be imaged in toto. Thus, measurements of the morphological changes of the PSM and cell tracking in 3D was possible during development. With this technique, posterior neural plate cells could be shown to form parts of the posterior neural tube, the entire posterior PSM and the somites of the tail. During this course of development, the PSM becomes longer but does not increase its volume. Only when forming the somites, an increase in volume could be measured in the mesoderm. Single-cell labelling showed an anterior shift of cell movement led by medial PSM cells and followed by more laterally located cells. The anterior displacement happens simultaneously to the posterior elongation of the embryo. A hypothetical push by newly generated cells at the tail tip could be ruled out. Mitotic cells were evenly distributed in all tissues of the tail with a low proliferation rate. The morphological changes and anterior relocations of the tail mesoderm could, therefore, mainly be explained by cell migration. Therefore, further analyses focussed on cell migration, particularly on cellular characteristics displayed during migration such as shape, orientation, volume, distribution and filopodia organisation to obtain more profound information about how PSM cells migrate and contribute to somite formation. The net movement of tail elongation is directed posteriorly regardless of anteriorly relocating PSM cells. That is only feasible if a lateral expansion of the PSM by laterally migrating PSM cells is counteracted. There have been no studies on the lateral boundary so far. In the axolotl, the PSM is covered laterally by a two-layered epidermis and a fibronectin-rich extracellular matrix. After removing the tail epidermis, operated embryos showed missing or malformed tails, especially with lateral and dorsal curvatures and shortenings. Tail mesoderm examined in these cases showed an increased PSM volume and a lateral expansion of the tissue. A nearly normal tail developed when, after removing the epidermis, the embryos developed in 1% agarose supplemented with fibronectin. In contrast, a simple covering of the PSM with a nitrocellulose membrane, incubation in the softer methylcellulose or in agarose without fibronectin did not rescue tail formation. The lateral pressure on the PSM and a fibronectin-rich extracellular matrix seem necessary to preserve the tissue architecture of the PSM during tail formation. This study unravels the behaviour of individual PSM cells during their morphogenesis from single cells in the posterior plate of the neurula until somite formation in the tail bud. Overall, with specific labelling of tail mesodermal cells, their contribution to PSM morphology could be elucidated, and a more detailed model of tail elongation could be proposed: The posterior expansion of the neural tube and notochord pushes the posterior neural plate tissue posteriorly and squeezes the cells into an elongated mediolaterally oriented form. Labelling experiments of small individual cell groups showed that the ventral posteriormost cells are the first to escape this pressure by relocating anteriorly. Then, more anteriorly located cells follow, as well as dorsally located cells. These movements explain the anterior turn. Thereby, mesodermal cells start to migrate randomly, become elongated and change their orientation from mediolateral to anterior-posterior. Random cell migration leads to homogeneous cell mixing, which results in an aligned uniform tissue of trunk and tail PSM. The lateral constriction by the epidermis channels the undirected migration movements in an anterior direction. In this way, cells are directed towards the site of somite formation, the PSM narrows, and the embryo elongates posteriorly. This extension model includes the individual cell behaviour, which on the whole shapes PSM morphology. The analysed dynamic morphological changes of the PSM can be linked to the developmental processes of the tail and the posterior elongation of the axis.:1 Introduction 1.1 Embryonic tail formation . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 Mechanism of tail formation . . . . . . . . . . . . . . . . . . . . . 1 1.1.2 Molecular determination of cell populations in the tail bud . . . . . 5 1.2 Axial elongation of the vertebrate body plan . . . . . . . . . . . . . . . . . 8 1.2.1 Anterior body elongation (elongation of the trunk) . . . . . . . . . 8 1.2.2 Posterior body elongation (tail elongation) . . . . . . . . . . . . . . 9 1.3 Studying tissue morphology during development . . . . . . . . . . . . 11 1.4 Aim of the project . . . . . . . . . . . . . . . . . . . . . . .. . . . . . 12 2 Materials 2.1 Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.2 Antibodies and dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3 Techniqual equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.4 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 3 Methods 3.1 Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1.1 Breeding of axolotls and embryo collection . . . . . . . . . . . . 19 3.1.2 Injections with the vital dye DiI . . . . . . . . . . . . . . . . . . . 19 3.1.3 Tissue transplantation techniques . . . . . . . . . . . . . . . . . . . 19 3.2 Immunohistochemical staining . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1 Vibratome sections . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.2 Whole-mount staining . . . . . . . . . . . . . . . . . . . . . . . . . 21 3.3 Optical tissue clearing protocols . . . . . . . . . . . . . . . . . . 21 3.3.1 Ethyl cinnamate based optical tissue clearing protocol . . . . . . . 21 3.3.2 SeeDB optical clearing protocol . . . . . . . . . . . . . . . . . . . . 22 3.4 Image analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.4.1 3D image generation and processing . . . . . . . . . . .. . . . . . 22 3.4.2 Length measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.4.3 Manual segmentation . . . . . . . . . . . . . . . . . . . . . . . . . 24 3.4.4 Automatic segmentation . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5 Determination of cellular parameters . . . . . . . . . . . . . . .. . . . . 25 3.5.1 Cell shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.2 Cell and tissue volume . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.3 Cellular distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.4 Closest neighbour analysis . . . . . . . . . . . . . . . . . . . . . . . 26 3.5.5 Cell orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.5.6 Length and orientation of filopodia . . . . . . . . . . . . . . . . . . 31 3.5.7 Distance of cells to a plane . . . . . . . . . . . . . . . . . . . . . . . 31 3.5.8 Mitotic rate and spindle orientation . . . . . . . . . . . . . . . . . 32 4 Results 4.1 The presomitic mesoderm is associated with axial elongation. . . . . . 33 4.1.1 Elongation of the body axis . . . . . . . . . . . . . . . . . . . . . . 33 4.1.2 Contribution of different tissues . . . . . . . . . . . . . . . . . . . . 34 4.1.3 Differential contribution of mesoderm and epidermis . . . . . . . . . 40 4.1.4 Dual potential of mesodermal progenitors . . . . . . . . . . . . . . . 42 4.1.5 Mesodermal tissue expansion . . . . . . . . . . . . . . . . . . . . . 46 4.2 Cellular behaviour influences mesodermal morphology . . . . . . . . . 50 4.2.1 Cell division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.2.2 Positional changes . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.2.3 Cellular characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 59 Cell shape changes . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Change of cell orientation . . . . . . . . . . . . . . . . . . . . . . . 61 Orientation of filopodia . . . . . . . . . . . . . . . . . . . . . . . . . 63 Cell distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.3 The epidermis fascilitates mesodermal tissue integrity . . . . . . .. . . . 67 4.3.1 Mesodermal tissue integrity . . . . . . . . . . . . . . . . . . . . . . 68 4.3.2 Malformed tails after epidermis removal . . . . . . . . . . . . . . . 70 4.3.3 Alteration in mesodermal tissue dimensions . . . . . . . . . . . . . 73 4.3.4 Alteration of cell density after epidermis removal . . . . . . . . . . 77 4.3.5 Rescue of tail formation . . . . . . . . . . . . . . . . . . . . . . . . 80 5 Discussion 5.1 Cell migration of the presomitic mesodermal cells . . . . . . . . .. . . . 85 5.1.1 Continuity of gastrulation movements . . . . . . . . . . . . . . . . . 85 5.1.2 Directed migration . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.1.3 Random cell migration . . . . . . . . . . . . . . . . . . . . . . . . . 88 5.1.4 Lateral mechanical constriction . . . . . . . . . . . . . . . . . . . . 90 5.2 Non-volumetric growth of the presomitic mesoderm . . . . . . . . . . . . . 91 5.3 Models of tail presomitic mesoderm formation . . . . . . . . . . . . . . . . 93 / Der Schwanz der Wirbeltiere ist bei jeder Art einzigartig und erfüllt ein breites Spektrum an Funktionen. Beim Salamander Axolotl (Ambystoma mexicanum), macht der Schwanz ein Drittel der gesamten Körperlänge aus und ist zum Schwimmen notwendig. Trotz seiner Größe stammen die meisten Gewebe des Schwanzes von der posterioren Neuralplatte der Neurula ab. Obwohl der größte Teil der Neuralplatte neuronale Strukturen des Zentralnervensystems hervorbringt, entwickelt sich der posteriore Teil der Neuralplatte überwiegend zu präsomitischem Mesoderm (PSM), das Muskeln, Knochen und Knorpel des Schwanzes und des hinteren Rumpfes bildet. Während der Entwicklung kehrt die posteriore Neuralplatte ihre Orientierung in einer anterioren Drehbewegung um (Taniguchi et al., 2017). Zellen der hintersten Plattenregion werden in eine anteriore Position verschoben, während zuvor anteriorere Neuralplattenzellen an einer posterioren Stelle landen. Gleichzeitig verlängert sich das axiale Neuralrohr und das Notochord nach posterior. Das PSM, das sich bilateral zur Zentralachse entwickelt, ist im Prozess der Schwanzverlängerung involviert, während es gleichzeitig an seinem vorderen Ende neue Somiten bildet. Es ist immer noch unklar, welche morphologischen Veränderungen das PSM durchläuft, um die Schwanzbildung und die posteriore Ausdehnung des Embryos zu ermöglichen. Darüber hinaus ist unbekannt, auf welche Weise PSM-Zellen ihre Form, Orientierung, ihr Migrationsverhalten und ihre Verteilung ändern, die für eine Veränderung der PSM- und Somitenmorphologie erforderlich sind. Mit homotopen Gewebetransplantationen von posterioren Neuralplattenzellen von einem gfp-exprimierenden Spender auf einen weißen (d/d) Empfänger, konnte eine spezifische Markierung aller mesodermalen Zellen des Schwanzes erreicht werden. Andernfalls können mesodermale Zellen des Rumpfes und des Schwanzes weder genetisch noch morphologisch unterschieden werden. Mit diesem Zellmarkierungsansatz konnte das gesamte Schwanzmesoderm in toto abgebildet werden. So waren Messungen der morphologischen Veränderungen des PSM und Zellverfolgung in 3D während der Entwicklung möglich. Mit dieser Technik konnte gezeigt werden, dass die Zellen der posterioren Neuralplatte Teile des posterioren Neuralrohrs, das gesamte posteriore PSM und die Somiten des Schwanzes bilden. Dabei wird das PSM länger, ohne sein Volumen zu vergrößern. Erst während der Bildung von Somiten wurde eine Volumenzunahme gemessen Einzelzellmarkierungen zeigten eine anteriore Verschiebung der Zellen, angeführt von medialen PSM-Zellen und gefolgt von lateral gelegenen Zellen. Diese anteriore Verschiebung geschieht gleichzeitig mit der posterioren Streckung des Embryos. Ein hypothetischer Schub durch neugebildete Zellen an der Schwanzspitze konnte ausgeschlossen werden. Mitotischen Zellen waren gleichmäßig in allen Geweben des Schwanzes verteilt und wiesen eine geringe Proliferationsrate auf. Die morphologischen Veränderungen und anterioren Verlagerungen des Schwanzmesoderms können daher hauptsächlich durch Zellmigration erklärt werden. Die Analysen konzentrierten sich daher auf die Zellmigration, insbesondere auf die zellulären Charakteristika, die sich während der Migration zeigen, wie z.B. Form, Orientierung, Volumen, Verteilung und Filopodienorganisation. So konnten neue Informationen darüber gewonnen werden, wie PSM-Zellen wandern und zur Somitenbildung beitragen. Die Nettobewegung der Schwanzverlängerung ist, unabhängig von nach anterior wandernden PSM-Zellen, nach posterior gerichtet. Das ist nur möglich, wenn einer lateralen Ausdehnung des PSM durch ungerichtet migrierenden Zellen entgegengewirkt wird. Über die Rolle einer laterale Begrenzung bei diesem Prozess gibt es bisher keine Untersuchungen. Beim Axolotl ist das PSM seitlich von einer zweischichtigen Epidermis und einer Fibronektin-reichen extrazellulären Matrix bedeckt. Nach Entfernung der Schwanzepidermis zeigten operierte Embryonen fehlende oder missgebildete Schwänze, insbesondere mit einer lateralen und dorsalen Krümmung und einer Verkürzung. Untersuchungen des Schwanzmesoderms zeigten ein erhöhtes PSM-Volumen und eine laterale Ausdehnung des Gewebes. Ein nahezu normaler Schwanz entwickelte sich, wenn die Embryonen nach Entfernung der Epidermis mit 1% Agarose, ergänzt mit Fibronektin, bedeckt wurden. Im Gegensatz dazu konnte eine einfache Abdeckung des PSM mit einer Nitrozellulosemembran, die Inkubation in der weicheren Methylzellulose oder in Agarose ohne Fibronektin die Schwanzbildung nicht normalisieren. Der seitliche Druck auf das PSM und eine Fibronektin-reiche extrazelluläre Matrix scheinen notwendig zu sein, um die Gewebearchitektur des PSM während der Schwanzbildung zu erhalten. Diese Studie zeigt das Verhalten einzelner PSM-Zellen während der Morphogenese der hinteren Neuralplatte bis zur Somitenbildung. Insgesamt konnte durch die spezifische Markierung von mesodermalen Zellen des Schwanzes deren Beitrag zur PSM-Morphologie aufgeklärt und ein detaillierteres Modell der Schwanzverlängerung vorgeschlagen werden: Die posteriore Ausdehnung des Neuralrohrs und des Notochords schiebt das posteriore Neuralplattengewebe nach hinten und quetscht die Zellen in eine verlängerte, mediolateral orientierte Form. Markierungsexperimente einzelner Zellgruppen zeigten, dass die ventralen, posterior gelegenen Zellen diesem Druck als erste entkommen, indem sie sich nach anterior verschieben. Ihnen folgen weiter anterior gelegene Zellen sowie dorsal gelegene Zellen. Diese Bewegungen erklären die anteriore Drehung. Dabei beginnen mesodermale Zellen ungerichtet zu wandern, verlängern sich und ändern ihre Orientierung von mediolateral nach anterior-posterior. Die ungerichtete Zellwanderung führt zu einer homogenen Zelldurchmischung, so dass zusammen mit dem PSM des Rumpfes ein einheitliches Gewebe gebildet wird. Die laterale Begrenzung durch die Epidermis kanalisiert die ungerichteten Migrationsbewegungen in anteriore Richtung. Auf diese Weise werden die Zellen in Richtung der Somitenbildungsstelle gelenkt, das PSM verengt sich, und der Embryo streckt sich nach hinten. Dieses Ausdehnungsmodell beinhaltet das individuelle Zellverhalten, das insgesamt die Morphologie des PSM prägt. Die analysierten dynamischen morphologischen Veränderungen des PSM können mit Schwanzentwicklungsprozessen und der posterioren Elongation der Achse in Verbindung gebracht werden.:1 Introduction 1.1 Embryonic tail formation . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 Mechanism of tail formation . . . . . . . . . . . . . . . . . . . . . 1 1.1.2 Molecular determination of cell populations in the tail bud . . . . . 5 1.2 Axial elongation of the vertebrate body plan . . . . . . . . . . . . . . . . . 8 1.2.1 Anterior body elongation (elongation of the trunk) . . . . . . . . . 8 1.2.2 Posterior body elongation (tail elongation) . . . . . . . . . . . . . . 9 1.3 Studying tissue morphology during development . . . . . . . . . . . . 11 1.4 Aim of the project . . . . . . . . . . . . . . . . . . . . . . .. . . . . . 12 2 Materials 2.1 Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.2 Antibodies and dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3 Techniqual equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.4 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 3 Methods 3.1 Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1.1 Breeding of axolotls and embryo collection . . . . . . . . . . . . 19 3.1.2 Injections with the vital dye DiI . . . . . . . . . . . . . . . . . . . 19 3.1.3 Tissue transplantation techniques . . . . . . . . . . . . . . . . . . . 19 3.2 Immunohistochemical staining . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1 Vibratome sections . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.2 Whole-mount staining . . . . . . . . . . . . . . . . . . . . . . . . . 21 3.3 Optical tissue clearing protocols . . . . . . . . . . . . . . . . . . 21 3.3.1 Ethyl cinnamate based optical tissue clearing protocol . . . . . . . 21 3.3.2 SeeDB optical clearing protocol . . . . . . . . . . . . . . . . . . . . 22 3.4 Image analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.4.1 3D image generation and processing . . . . . . . . . . .. . . . . . 22 3.4.2 Length measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.4.3 Manual segmentation . . . . . . . . . . . . . . . . . . . . . . . . . 24 3.4.4 Automatic segmentation . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5 Determination of cellular parameters . . . . . . . . . . . . . . .. . . . . 25 3.5.1 Cell shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.2 Cell and tissue volume . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.3 Cellular distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5.4 Closest neighbour analysis . . . . . . . . . . . . . . . . . . . . . . . 26 3.5.5 Cell orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.5.6 Length and orientation of filopodia . . . . . . . . . . . . . . . . . . 31 3.5.7 Distance of cells to a plane . . . . . . . . . . . . . . . . . . . . . . . 31 3.5.8 Mitotic rate and spindle orientation . . . . . . . . . . . . . . . . . 32 4 Results 4.1 The presomitic mesoderm is associated with axial elongation. . . . . . 33 4.1.1 Elongation of the body axis . . . . . . . . . . . . . . . . . . . . . . 33 4.1.2 Contribution of different tissues . . . . . . . . . . . . . . . . . . . . 34 4.1.3 Differential contribution of mesoderm and epidermis . . . . . . . . . 40 4.1.4 Dual potential of mesodermal progenitors . . . . . . . . . . . . . . . 42 4.1.5 Mesodermal tissue expansion . . . . . . . . . . . . . . . . . . . . . 46 4.2 Cellular behaviour influences mesodermal morphology . . . . . . . . . 50 4.2.1 Cell division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.2.2 Positional changes . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.2.3 Cellular characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 59 Cell shape changes . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Change of cell orientation . . . . . . . . . . . . . . . . . . . . . . . 61 Orientation of filopodia . . . . . . . . . . . . . . . . . . . . . . . . . 63 Cell distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.3 The epidermis fascilitates mesodermal tissue integrity . . . . . . .. . . . 67 4.3.1 Mesodermal tissue integrity . . . . . . . . . . . . . . . . . . . . . . 68 4.3.2 Malformed tails after epidermis removal . . . . . . . . . . . . . . . 70 4.3.3 Alteration in mesodermal tissue dimensions . . . . . . . . . . . . . 73 4.3.4 Alteration of cell density after epidermis removal . . . . . . . . . . 77 4.3.5 Rescue of tail formation . . . . . . . . . . . . . . . . . . . . . . . . 80 5 Discussion 5.1 Cell migration of the presomitic mesodermal cells . . . . . . . . .. . . . 85 5.1.1 Continuity of gastrulation movements . . . . . . . . . . . . . . . . . 85 5.1.2 Directed migration . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.1.3 Random cell migration . . . . . . . . . . . . . . . . . . . . . . . . . 88 5.1.4 Lateral mechanical constriction . . . . . . . . . . . . . . . . . . . . 90 5.2 Non-volumetric growth of the presomitic mesoderm . . . . . . . . . . . . . 91 5.3 Models of tail presomitic mesoderm formation . . . . . . . . . . . . . . . . 93

In contrast to mammals, some fish and amphibians have retained the ability to regenerate complex body structures or organs, such as the limb, the tail, the eye lens or even parts of the heart. One major difference in the response to injury is the appearance of a mesenchymal growth zone or blastema in these regenerative species instead of the scarring seen in mammals. This blastema is thought to largely derive from the dedifferentiation of various functional cell types, such as skeletal muscle, skin and cartilage. In the case of multinucleated skeletal muscle fibres, cell cycle re-entry into S-phase as well as fragmentation into mononucleated progenitors is observed both in vitro and in vivo. In order to identify molecules that initiate dedifferentiation of cells at the wound site in amphibians we have established a cellular assay with a cultured newt myogenic cell line. Using this assay we have found a serum activity that stimulates cell cycle re-entry in differentiated multinucleated newt myotubes. The activity is present in serum of all mammalian species tested so far and, interestingly, thrombin proteolysis amplifies the activity from both serum and plasma. We think this serum factor provides a link between wounding and regeneration and its identification will be a key step in understanding the remarkable differences in wound healing between mammals and amphibians. In the course of this PhD thesis we have characterized the serum factor as a thermo-labile, pH- and proteinase K-sensitive, high molecular weight protein that is resistant to denaturing conditions such as SDS, urea or organic solvents. Surprisingly, under denaturing conditions the activity behaves as a low molecular weight protein that displays charge heterogeneity on isoelectric focusing. Using these characteristics of the serum factor we have performed a systematic investigation of commonly used protein chromatography modes and separation techniques to develop a successful purification procedure. After four column chromatography steps -- cation exchange, hydrophobic interaction, heparin affinity and size exclusion chromatography under denaturing conditions -- we have achieved a 2,000-fold purification starting from a commercially available Crude Bovine Thrombin preparation. This represents about 40,000-fold purification over bovine serum. Silver stained gels of the most purified fractions revealed ten major protein bands. In order to finally identify the cell cycle re-entry factor, we are currently analyzing the purification by quantitative mass spectrometry by correlating the abundance of tryptic peptides with activity in sequential fractions across a chromatography run.

Les amphibiens urodèles (e.g. les axolotls) possèdent la remarquable capacité de régénérer plusieurs parties de leur corps. Ils peuvent, entre autres, régénérer parfaitement un membre amputé par épimorphose, un processus biphasique comprenant une phase de préparation, spécifique à la régénération, et une phase de redéveloppement, commune à l’épimorphose et au développement embryonnaire. Durant la phase de préparation, les cellules du moignon se dédifférencient en cellules pseudo-embryonnaires, prolifèrent et migrent distalement au plan d’amputation pour former un blastème de régénération. Parmi les vertébrés, la dédifférenciation est unique aux urodèles. Afin de mieux comprendre le contrôle moléculaire de la régénération chez les urodèles, nous avons choisi d’étudier BMP-2, un facteur de croissance, en raison de son implication dans la régénération des phalanges distales chez les mammifères. Le facteur de transcription MSX-1 a également été sélectionné en raison de sa capacité à induire la dédifférenciation cellulaire in vitro et de son interaction potentielle avec la signalisation des BMPs. Les résultats présentés dans cette thèse démontrent que BMP-2 et MSX-1 sont exprimés lors des phases de préparation et de redéveloppement de l’épimorphose, et que leur profil d'expression spatio-temporel est très semblable, ce qui suggère une interaction de leurs signaux. En outre, chez les tétrapodes amniotes, l’expression de Shh est restreinte au mésenchyme postérieur des membres en développement et chevauche l’expression de BMP-2. Toutefois, l’expression de BMP-2 n’est pas restreinte à la région postérieure mais forme un gradient postéro-antérieur. Shh est le principal régulateur de la formation du patron de développement antéro-postérieur du ii membre. Étant donné les domaines d’expression chevauchants de BMP-2 et Shh et la restriction postérieure d’expression de Shh, on croit que Shh régule la formation du patron de développement de postérieur à antérieur par l’activation de l’expression de BMP-2. Fait intéressant, l’axolotl exprime également Shh dans la région postérieure, mais le développement des pattes se fait de la région antérieure à la région postérieure au lieu de postérieur à antérieur comme chez les autres tétrapodes, et ceci durant le développement et la régénération. Nous avons utilisé cette caractéristique de l’axolotl pour démontrer que la signalisation Shh ne structure pas l’autopode via BMP-2. En effet, l’expression de BMP-2 n'est pas régulée par l'inhibition de la signalisation Shh, et son expression est du côté opposé à celle de Shh durant le développement et la régénération des pattes de l’axolotl. Il a été observé durant le développement du membre chez la souris que MSX-1 est régulé par la signalisation Shh. Nos résultats ont démontrés que chez l’axolotl, MSX-1 ne semble pas régulé par l'inhibition de la signalisation Shh au cours de la régénération du membre. De plus, nous avons démontré que contrairement à l’expression de Shh, l’expression de BMP-2 est corrélée avec l’ordre de formation des phalanges, est impliquée dans la condensation cellulaire et dans l'apoptose précédant la chondrogenèse. L’ensemble de ces résultats suggère un rôle de BMP-2 dans l’initiation de l’ossification endochondrale. Enfin, nous avons démontré que la signalisation BMP est indispensable pour l’épimorphose du membre durant la phase de redéveloppement. / Urodele amphibians (e.g. the axolotls) have a remarkable ability to regenerate parts of their body. They will, among other things, fully regenerate an amputated limb by epimorphosis, a biphasic process comprising a preparation phase, specific to the regeneration, and a redevelopment phase, common to epimorphosis and embryonic development. During the preparation phase, the cells of the stump dedifferentiate into embryonic-like cells, proliferate and migrate distally from the level of amputation to form a regeneration blastema. Among vertebrates, the process of dedifferentiation is unique to urodeles. To better understand the molecular control of regeneration in urodeles, we chose to study BMP-2, a growth factor, because of its involvement in mammalian digit tip regeneration. The transcription factor MSX-1 has also been selected because of its ability to induce cellular dedifferentiation in vitro and its potential interaction with BMPs signaling. The results presented in this thesis show that BMP-2 and MSX-1 are expressed during phases of preparation and redevelopment of epimorphosis, and their spatio-temporal expression profiles are very similar at each stage of epimorphosis, suggesting an interaction of their signals during regeneration. In addition, in tetrapod amniotes, the expression of Shh is restricted to the posterior mesenchyme of developing limbs and overlaps with the expression of BMP-2. However, the expression of BMP-2 is not restricted to the posterior region but forms a posterior-anterior gradient. Shh is the main regulator of the anterior-posterior pattern formation of developing limbs. Given the overlapping expression domains of Shh and BMP-2, and the expression restriction of Shh in posterior, Shh is believed to iv regulate the pattern formation of developing limbs by the activation of BMP-2 expression. Interestingly, the axolotl also expresses Shh in the posterior region, but the limb develops from anterior to posterior rather than posterior to anterior as in other tetrapods, and this, during development and epimorphosis. We used this feature of the axolotl to demonstrate that Shh signaling does not regulate pattern formation through BMP-2. Indeed, the expression of BMP-2 is not regulated by the inhibition of hh signaling, and its expression is opposite to that of Shh during development and regeneration of the axolotl limb. It was observed, during limb development in mice that MSX-1 is regulated by Shh signaling. Our results suggest that in the axolotl, MSX-1 is not regulated by the inhibition of Shh signaling during limb regeneration. Furthermore, we demonstrated that unlike the expression of Shh, the expression of BMP-2 is correlated with the order of formation of the phalanges, is involved in cell condensation and apoptosis preceding chondrogenesis. Taken together, these results suggest a role for BMP-2 in the initiation of endochondral ossification. Finally, we demonstrated that BMP signaling is essential for the redevelopment phase of limb epimorphosis.

We studied patterning mechanisms and cell fates during limb regeneration in the axolotl. 1) It is crucial to understand the earliest events of patterning. Since it is technically challenging to study early events, we established single cell PCR. This new tool will allow us to obtain novel insight into the initial steps of limb patterning. 2)We have examined the roles of different tissues regarding their fates and features of proximo- distal patterning. Our strategy was to transplant GFP+ skin, skeleton, muscle and Schwann cells from transgenic donors to limbs of wild type hosts, amputate through the graft and analyze fluorescent progeny combined with the use of molecular markers. Our results revealed that different subpopulations of blastema cells exist regarding two aspects. First, we found that progeny of skin and skeleton have some tissue specific memory since they did not give rise to muscle lineages. However, cells of the skin contributed to other mesenchymal tissues like cartilage or tendons, while the majority of skeleton- derived cells undergoes self- renewal. Second, we performed one cellular and two molecular assays to investigate what tissues generate cells that exhibit features of proximo- distal patterning. Both assays revealed that Schwann cell- derived progeny do not display such features while progeny of skin, skeleton and muscle did. Therefore, we conclude that the blastema is a heterogeneous mix of cells regarding tissue lineages and features of proximo- distal patterning.

Les amphibiens urodèles (e.g. les axolotls) possèdent la remarquable capacité de régénérer plusieurs parties de leur corps. Ils peuvent, entre autres, régénérer parfaitement un membre amputé par épimorphose, un processus biphasique comprenant une phase de préparation, spécifique à la régénération, et une phase de redéveloppement, commune à l’épimorphose et au développement embryonnaire. Durant la phase de préparation, les cellules du moignon se dédifférencient en cellules pseudo-embryonnaires, prolifèrent et migrent distalement au plan d’amputation pour former un blastème de régénération. Parmi les vertébrés, la dédifférenciation est unique aux urodèles. Afin de mieux comprendre le contrôle moléculaire de la régénération chez les urodèles, nous avons choisi d’étudier BMP-2, un facteur de croissance, en raison de son implication dans la régénération des phalanges distales chez les mammifères. Le facteur de transcription MSX-1 a également été sélectionné en raison de sa capacité à induire la dédifférenciation cellulaire in vitro et de son interaction potentielle avec la signalisation des BMPs. Les résultats présentés dans cette thèse démontrent que BMP-2 et MSX-1 sont exprimés lors des phases de préparation et de redéveloppement de l’épimorphose, et que leur profil d'expression spatio-temporel est très semblable, ce qui suggère une interaction de leurs signaux. En outre, chez les tétrapodes amniotes, l’expression de Shh est restreinte au mésenchyme postérieur des membres en développement et chevauche l’expression de BMP-2. Toutefois, l’expression de BMP-2 n’est pas restreinte à la région postérieure mais forme un gradient postéro-antérieur. Shh est le principal régulateur de la formation du patron de développement antéro-postérieur du ii membre. Étant donné les domaines d’expression chevauchants de BMP-2 et Shh et la restriction postérieure d’expression de Shh, on croit que Shh régule la formation du patron de développement de postérieur à antérieur par l’activation de l’expression de BMP-2. Fait intéressant, l’axolotl exprime également Shh dans la région postérieure, mais le développement des pattes se fait de la région antérieure à la région postérieure au lieu de postérieur à antérieur comme chez les autres tétrapodes, et ceci durant le développement et la régénération. Nous avons utilisé cette caractéristique de l’axolotl pour démontrer que la signalisation Shh ne structure pas l’autopode via BMP-2. En effet, l’expression de BMP-2 n'est pas régulée par l'inhibition de la signalisation Shh, et son expression est du côté opposé à celle de Shh durant le développement et la régénération des pattes de l’axolotl. Il a été observé durant le développement du membre chez la souris que MSX-1 est régulé par la signalisation Shh. Nos résultats ont démontrés que chez l’axolotl, MSX-1 ne semble pas régulé par l'inhibition de la signalisation Shh au cours de la régénération du membre. De plus, nous avons démontré que contrairement à l’expression de Shh, l’expression de BMP-2 est corrélée avec l’ordre de formation des phalanges, est impliquée dans la condensation cellulaire et dans l'apoptose précédant la chondrogenèse. L’ensemble de ces résultats suggère un rôle de BMP-2 dans l’initiation de l’ossification endochondrale. Enfin, nous avons démontré que la signalisation BMP est indispensable pour l’épimorphose du membre durant la phase de redéveloppement. / Urodele amphibians (e.g. the axolotls) have a remarkable ability to regenerate parts of their body. They will, among other things, fully regenerate an amputated limb by epimorphosis, a biphasic process comprising a preparation phase, specific to the regeneration, and a redevelopment phase, common to epimorphosis and embryonic development. During the preparation phase, the cells of the stump dedifferentiate into embryonic-like cells, proliferate and migrate distally from the level of amputation to form a regeneration blastema. Among vertebrates, the process of dedifferentiation is unique to urodeles. To better understand the molecular control of regeneration in urodeles, we chose to study BMP-2, a growth factor, because of its involvement in mammalian digit tip regeneration. The transcription factor MSX-1 has also been selected because of its ability to induce cellular dedifferentiation in vitro and its potential interaction with BMPs signaling. The results presented in this thesis show that BMP-2 and MSX-1 are expressed during phases of preparation and redevelopment of epimorphosis, and their spatio-temporal expression profiles are very similar at each stage of epimorphosis, suggesting an interaction of their signals during regeneration. In addition, in tetrapod amniotes, the expression of Shh is restricted to the posterior mesenchyme of developing limbs and overlaps with the expression of BMP-2. However, the expression of BMP-2 is not restricted to the posterior region but forms a posterior-anterior gradient. Shh is the main regulator of the anterior-posterior pattern formation of developing limbs. Given the overlapping expression domains of Shh and BMP-2, and the expression restriction of Shh in posterior, Shh is believed to iv regulate the pattern formation of developing limbs by the activation of BMP-2 expression. Interestingly, the axolotl also expresses Shh in the posterior region, but the limb develops from anterior to posterior rather than posterior to anterior as in other tetrapods, and this, during development and epimorphosis. We used this feature of the axolotl to demonstrate that Shh signaling does not regulate pattern formation through BMP-2. Indeed, the expression of BMP-2 is not regulated by the inhibition of hh signaling, and its expression is opposite to that of Shh during development and regeneration of the axolotl limb. It was observed, during limb development in mice that MSX-1 is regulated by Shh signaling. Our results suggest that in the axolotl, MSX-1 is not regulated by the inhibition of Shh signaling during limb regeneration. Furthermore, we demonstrated that unlike the expression of Shh, the expression of BMP-2 is correlated with the order of formation of the phalanges, is involved in cell condensation and apoptosis preceding chondrogenesis. Taken together, these results suggest a role for BMP-2 in the initiation of endochondral ossification. Finally, we demonstrated that BMP signaling is essential for the redevelopment phase of limb epimorphosis.

Depuis plusieurs années, la sénescence cellulaire a été majoritairement étudiée comme un processus causé par le vieillissement des cellules et un mécanisme pour limiter la propagation des cellules précancéreuses. Cette perspective a changé suite aux publications des groupes Serrano et Keyes, qui ont démontré la présence des cellules sénescentes très tôt durant le développement embryonnaire de la souris. Dernièrement, le laboratoire de Dr Roy a identifié le pronéphros, un organe transitoire qui est remplacé par le mésonéphros (homologue du rein chez les humains) à la maturité, les fascicules du nerf olfactif ainsi que les gencives comme des zones riches en cellules sénescentes durant le développement de l’axolotl. Ces évidences suggèrent que la sénescence est une réponse biologique survenant non seulement lors du vieillissement, mais est également une réponse physiologique pouvant être activée par différents signaux présents dans son environnement. À ce jour, bien que plusieurs chercheurs ciblent l’identification des points de contrôle de la sénescence cellulaire, on connaît peu de chose sur le rôle des cellules sénescentes durant le développement embryonnaire et la manière dont celui-ci pourrait influencer la croissance et la morphogenèse des organismes. Dans la présente étude, on cherche à identifier les gènes qui sont importants dans la sénescence développementale. Les résultats de séquençage d’ARN dans le pronéphros de l’axolotl ont démontré un enrichissement significatif du gène PEBP1, qui code pour une protéine surtout connue pour son rôle inhibiteur sur la kinase Raf. De plus, nos résultats semblent démontrer une diminution de l’activité bêta galactosidase dans le pronéphros de l’axolotl en développement lorsque celui-ci est traité à la Locostatin, un inhibiteur pharmacologique qui bloque l’interaction de PEBP1 avec RAF. Nous suggérons que PEBP1 pourrait être important pour soit l’activation ou le maintien du caractère sénescent dans les organes en développement de l’axolotl. Un phénomène qui semble être important pour conserver la fonctionnalité de l’organe en transition. / For several years, cellular senescence has been mainly studied as a process caused by the aging of cells and a mechanism to limit the propagation of precancerous cells. This perspective changed following the publications of Serrano and Keyes, who demonstrated the presence of senescent cells very early in the embryonic development of mice. Recently, the laboratory of Dr. Roy identified the pronephros, a transient organ which is replaced by the mesonephros (kidney counterpart in humans) at maturity, the olfactory nerve fascicles as well as the gums as areas rich in senescent cells during the development of axolotl. The evidence suggests that senescence is not only related to the aging process, but rather a physiological response which is activated by various signals present in its environment. To date, although several researchers are targeting the identification of control points for the activation of senescence, little is known about the role of senescent cells during embryonic development and how it could influence growth and morphogenesis. In the present study, we seek to identify the genes that are important in developmental senescence. The results of RNA sequencing in the axolotl pronephros have, among other things, demonstrated a significant enrichment of the PEBP1 gene which codes for a protein that acts as an inhibitor of the Raf kinase. Our findings support the idea that cellular senescence that occurs during the embryonic development of the axolotl is dependent on PEBP1 since the pharmacological inhibitor of PEBP1 (Locostatin), which blocks the interaction of PEBP1 with RAF, seems to affect the activity of senescent beta galactosidase. We suggest that PEBP1 is necessary for either the activation or the maintenance of senescence in the pronephros during embryogenesis. We further suggest that embryonic senescent is crucial for the morphogenesis of the developing organ perhaps by keeping the organ functional during the transition.

We studied patterning mechanisms and cell fates during limb regeneration in the axolotl. 1) It is crucial to understand the earliest events of patterning. Since it is technically challenging to study early events, we established single cell PCR. This new tool will allow us to obtain novel insight into the initial steps of limb patterning. 2)We have examined the roles of different tissues regarding their fates and features of proximo- distal patterning. Our strategy was to transplant GFP+ skin, skeleton, muscle and Schwann cells from transgenic donors to limbs of wild type hosts, amputate through the graft and analyze fluorescent progeny combined with the use of molecular markers. Our results revealed that different subpopulations of blastema cells exist regarding two aspects. First, we found that progeny of skin and skeleton have some tissue specific memory since they did not give rise to muscle lineages. However, cells of the skin contributed to other mesenchymal tissues like cartilage or tendons, while the majority of skeleton- derived cells undergoes self- renewal. Second, we performed one cellular and two molecular assays to investigate what tissues generate cells that exhibit features of proximo- distal patterning. Both assays revealed that Schwann cell- derived progeny do not display such features while progeny of skin, skeleton and muscle did. Therefore, we conclude that the blastema is a heterogeneous mix of cells regarding tissue lineages and features of proximo- distal patterning.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.