BackgroundDevelopmental biology relies to a large extent on the observation and comparison of phenotypic traits through time using high resolution microscopes. In this context, transparent model organisms such as the zebrafish Danio rerio in which developing tissues and organs can be easily observed and imaged using fluorescent proteins have become very popular. One limiting factor however is the acquisition of a sufficient amount of data, in standardized and reproducible conditions, to allow robust quantitative analysis. One way to improve this is by developing mounting methods to increase the number of embryos that can be imaged simultaneously in near-to-identical orientation.ResultsHere we present an improved mounting method allowing semi-automated and high-content imaging of zebrafish embryos. It is based on a 3D-printed stamp which is used to create a 2D coordinate system of multiple μ-wells in an agarose cast. Each μ-well models a negative of the average zebrafish embryo morphology between 22 and 96 h-post-fertilization. Due to this standardized and reproducible arrangement, it is possible to define a custom well plate in the respective imaging software that allows for a semi-automated imaging process. Furthermore, the improvement in Z-orientation significantly reduces post-processing and improves comparability of volumetric data while reducing light exposure and thus photo-bleaching and photo-toxicity, and improving signal-to-noise ratio (SNR).ConclusionsWe present here a new method that allows to standardize and improve mounting and imaging of embryos. The 3D-printed stamp creates a 2D coordinate system of μ-wells in an agarose cast thus standardizing specimen mounting and allowing high-content imaging of up to 44 live or mounted zebrafish embryos simultaneously in a semi-automated, well-plate like manner on inverted confocal microscopes. In summary, image data quality and acquisition efficiency (amount of data per time) are significantly improved. The latter might also be crucial when using the services of a microscopy facility.

In order to form an organ, cells need to take up specialized functions and tasks. Cellular specialization is guided by an interplay of chemical signals and physical forces, where one influences the other. One aspect in cellular identity is its shape, which e.g. defines how susceptible the cell may be to intercellular signaling or in which section of the cell cycle it is and therefore can tell us about its current state. Shape changes are introduced by motor proteins that are controlled and activated in a locally confined manner. For my thesis, I was interested to understand better how cellular shape and geometry impacts downstream cell and organ development. What happens if a cell cant transition to a specific shape? How does it affect tissue structure? How does it affect further development? One regulator of motor proteins like non-muscle myosin is Shroom3, which recently has been been shown to be expressed and involved in the development of the zebrafish lateral line organ (1 ). Development of the lateral line occurs through a migrating cluster of initially about 150 cells, the posterior lateral line primordium (pLLP), which migrates from the anterior (head) to the posterior (tail) while depositing cell clusters in a regular pattern. Literature on development of the lateral line suggests that in order for a cell cluster to be deposited from the pLLP, rosette formation is a key requirement. Therefore our expectation from the shroom3 mutant was that the number of clusters deposited was significantly reduced. To our surprise, when we first inspected the end of migration lateral line phenotype we found many individuals with a significant increase in cell clusters deposited. This made us re-think the role of Shroom3 during rosette assembly and the processes its involved in. To study the effects of Shroom3 on lateral line development, a mutant line was generated and crossed with various transgenic lines which express fluorescently labeled proteins that locate to organelles such as the plasmamembrane or the nucleus. Following, the mutant with its fluorescent labels was microscopically imaged under different conditions to quantify and analyze various cell-morphometric features. Even though the zebrafish is a popular model organism and its perfectly suited for developmental biology and advanced microscopy, there were no methods that would allow for a standardized and more automated pipeline of data acquisition and processing. Therefore, in order to accurately quantify the morphogenic processes Shroom3 is involved in, I developed a new toolset that significantly improved and facilitated my research. The toolset consists of (1) a new sample mounting method that is based on a 3D agarose gel that increases the number of embryos that can be mounted and imaged at once and speeds up the imaging process significantly (2) for subseqent image analysis I developed four programs that automate the process and therefore make the results much more reproducible and the analysis much more efficient. The first program is used for end of migration analyses, to deduce the pattern, count and size of Lateral Line cell clusters. The second is used not for end of migration, but for migration analyses (on timelapse recordings). Besides this it also prepares the images for more advanced downstream migration analyses and allows to analyse fluorescence signal on a second channel. The third program is used to analyse the pLLP only at high spatial resolution and to deduce the cell count, 3D cell morphometrics (like the volume) and cell orientation. The fourth program finally is used downstream of the second and third program and is capable of detecting and comparing them with the look of wildtype rosettes. Here I show that in absence of Shroom3 rosette formation in the migrating pLLP is destabilized leading to facilitated cell cluster deposition and I show how this might be related to traction forces due to a possible interdependence of pLLP acceleration and speed of migration. Furthermore I show that apical constriction and rosette formation is not blocked in Shroom3 deficient embryos, but that larger rosettes are fragmented into many smaller ones. Finally, I give an outlook on how the absense of Shroom3 and hence the absense of morphological changes may deregulate gene transcription by elevating the levels Atoh1a, a transcription factor necessary for hair cell development. My results and methodology demonstrate the importance of morphology in guiding developmental processes and how rather small morphological changes on the cellular level can impact further development significantly. My work also shows how powerful modern genetics, imaging and image analysis are and how diverse they are in terms of range of questions they are capable of answering. The methods and tools I developed prepare the ground for at least three quarters of the analyses I carried out and together with the documentation and data I provide, they are highly reproducible. In that regard I am especially happy that one of my developments, an improved sample preparation method, is already used by many different labs all over the world helping them to make their results more reproducible.