Humans can, if they survived a cardiac injury such as heart infarction, heal this cardiac injury only by scarring and with minimal regeneration of some cardiac cells. Zebrafish, however, can fully regenerate cardiac tissue after surgical resection of up to 20% of the ventricle. Regenerating tissue includes cells of the three cardiac layers, i.e. myocardium, epicardium and endocardium. Thus, zebrafish, with its ability to regenerate damaged heart and as a model enabling genetic manipulations, provides the possibility to study cellular and molecular mechanisms of heart regeneration. Understanding these mechanisms may help develop new therapeutic approaches to improve the situation after a heart injury in humans. Since molecular mechanisms regulating heart regeneration are so far largely unknown, I aimed to identify and analyze molecular signals that are important for cardiac regeneration in zebrafish. Molecular signals that are crucial during heart development have been suggested to be reactivated during cardiac regeneration. Since Wnt/β-catenin signaling is crucial for vertebrate heart development, it is likely to be important for zebrafish cardiac regeneration as well. First, I focused on the functional role of Wnt/β-catenin signaling in cardiac regeneration mainly by using transgenic fish lines that allow inducible activation or inhibition of the pathway. By using in situ hybridization and expression profiling, I tested whether endogenous Wnt/β-catenin target genes are detectable in regenerating hearts and screened for activity of the β-catenin responsive reporter in TOPdGFP transgenic fish (Tg(TOP:GFP)w25) after ventricular resection. I could not identify endogenous Wnt/β-catenin targets during the early phase of regeneration up to 7 days post amputation (dpa) using oligoexpression microarrays or in situ hybridization. An injury specific activation of the β-catenin responsive TOPdGFP reporter was not detectable either, suggesting that Wnt/β-catenin signaling is not active during this early phase of regeneration. The manipulation of Wnt/β-catenin signaling using transgenic fish lines did not influence cell proliferation or the overall extent of zebrafish heart regeneration. These results suggested that Wnt/β-catenin signaling has no functional role during entire zebrafish heart regeneration. Second, I found the transcription factor Sox9a to be upregulated after ventricular resection during the early phase of heart regeneration. Using transgenic reporter fish lines, I detected Sox9a expression in cardiomyocytes and endothelial cells, part of which were proliferative. Furthermore Sox9a was expressed in some cells of the epicardial layer that activated the expression of developmental genes in the entire heart in response to injury. These results indicated that Sox9a is expressed in cells that were actively involved in the regenerative response. To gain insight into the functional role of Sox9a, I generated a transgenic fish line where a repressor construct is inducibly expressed, which then interferes with Sox9a target gene transcription. I detected a significant reduction in myocardial and endothelial regeneration after induction of the repressor. These results suggested that Sox9a function is important for regeneration of endothelial and myocardial cells after heart injury. Third, using oligoexpression microarrays, I performed systematic gene expression profiling of the zebrafish heart regeneration within the first 2 weeks following amputation. I found that known genes, which have previously been shown to be strongly expressed during heart regeneration, as well as novel genes were upregulated after ventricular resection. Some of these genes have been implicated in vertebrate heart development, supporting the idea that cardiac developmental genes are reactivated during heart regeneration. Hence, these results reveal a good starting point for further analysis of the cellular and molecular events occurring within the first days after cardiac injury. Finally, I developed a cryoinjury method that more closely resembled the injured tissue after human heart infarction. I induced tissue death by exposing the ventricle to dry ice and detected that the zebrafish heart can regenerate upon this cardiac injury similarly as in response to a ventricular resection injury. After cryoinjury, the entire epicardium activated the expression of developmental genes and started to proliferate. I detected also proliferating cardiomyocytes, indicating that similar cellular mechanisms are induced in the epicardium and the myocardium after cryoinjury and ventricular resection. Furthermore, activation of Sox9 expression early after cryoinjury suggested that molecular mechanisms of regeneration are also similar in both injury methods. Thus, cryoinjury provides a useful tool for future studies of zebrafish heart regeneration with more relevance to human cardiac infarction. I discuss all results with reference to vertebrate heart development and to the response after mammalian heart infarction. Furthermore, the results were put into the context of cellular mechanisms that are present in the process of zebrafish heart regeneration.

Das Herz eines Zebrafishs ist bemerkenswert, da es sich nach einer Verletzung vollständig regenerieren kann. Der Regenerationsprozess wird von Fibrose begleitet - der Bildung von überschüssigem Gewebe der extrazellulären Matrix (ECM). Anders als bei Säugetieren ist die Fibrose im Zebrafish nur transient. Viele Signalwege wurden identifiziert, die an der Herzregeneration beteiligt sind. Allerdings sind die Zelltypen, insbesondere Nicht-Kardiomyozyten, die für die Regulation des Regenerationsprozesses verantwortlich sind, weitgehend unbekannt. In dieser Arbeit haben wir systematisch alle Zelltypen des gesunden und des verletzten Zebrafischherzens mithilfe einer auf Mikrofluidik basierenden Hoch-Durchsatz- Einzelzell-RNA-Sequenzierung bestimmt. Wir fanden eine große Heterogenität von ECM-produzierenden Zellen, einschließlich einer Reihe neuer Fibroblasten, die nach einer Verletzung mit unterschiedlicher Dynamik auftreten. Wir konnten aktivierte Fibroblasten beschreiben und Fibroblasten-Subtypen mit einer pro-regenerativen Funktion identifizieren. Darüber hinaus haben wir eine Methode entwickelt, um die Transkriptomanalyse und die Rekonstruktion von Zell-Verwandtschaften auf Einzelzellebene zu kombinieren. Unter Verwendung der CRISPR-Cas9-Technologie führten wir zufällige Mutationen in bekannte und ubiquitär transkribierte DNA-Loci während der Embryonalentwicklung von Zebrafischen ein. Diese Mutationen dienten als zellspezifische, permanente und vererbbare “Barcodes”, die zu einem späteren Zeitpunkt erfasst werden konnten. Mit maßgeschneiderten Analysealgorithmen konnten wir dann Stammbäume der sequenzierten Einzelzellen erstellen. Mit dieser neuen Methode haben wir gezeigt, dass im sich regenerierenden Zebrafischherz ECM-produzierende Zellpopulationen entweder mit dem Epi- oder mit dem Endokardium verwandt sind. Zusätzlich entdeckten wir, dass vom Endokardium abgeleitete Zelltypen vom Wnt-Signalweg abhängig sind. / The zebrafish heart has the remarkable capacity to fully regenerate after injury. The regeneration process is accompanied by fibrosis - the formation of excess extracellular matrix (ECM) tissue, at the injury site. Unlike in mammals, the fibrosis of the zebrafish heart is only transient. While many pathways involved in heart regeneration have been identified, the cell types, especially non-myocytes, responsible for the regulation of the regenerative process have largely remained elusive. Here, we systematically determined all different cell types of both the healthy and cryo-injured zebrafish heart in its regeneration process using microfluidics based high-throughput single-cell RNA sequencing. We found a considerable heterogeneity of ECM producing cells, including a number of novel fibroblast cell types which appear with different dynamics after injury. We could describe activated fibroblasts that extensively switch on gene modules for ECM production and identify fibroblast sub- types with a pro-regenerative function. Furthermore, we developed a method that is capable of combining transcriptome analysis with lineage tracing on the single-cell level. Using CRISPR-Cas9 technology, we introduced random mutations into known and ubiquitously transcribed DNA loci during the zebrafish embryonic development. These mutations served as cell-unique, permanent, and heritable barcodes that could be captured at a later stage simultaneously with the transcriptome by high-throughput single-cell RNA sequencing. With custom tailored analysis algorithms, we were then able to build a developmental lineage tree of the sequenced single cells. Using this new method, we revealed that in the regenerating zebrafish heart, ECM contributing cell populations derive either from the epi- or the endocardium. Additionally, we discovered in a functional experiment that endocardial derived cell types are Wnt signaling dependent.