Final Report Summary - ANAMORPH (Comparative systems-level studies of animal appendage morphogenesis)
The ANAMORPH project supported by the EU Marie Curie Intra-European Fellowships actions has been a multidisciplinary research program in the Life Sciences, addressing the fundamental cellular and molecular mechanisms that control organogenesis in developing animal embryos. The project made use of recent breakthroughs in developmental biology, functional genetics, microscopy and image informatics to study development in living embryos for several consecutive days, at a high spatial and temporal resolution, and at different levels of biological organization: from gene expression and function, to morphogenetic cell activities, to organ size and shape control. In particular, in this study we focused on development of body outgrowths (appendages, for example limbs) in a new experimental animal model, the amphipod crustacean Parhyale hawaiensis; we established transgenic (genetically modified) experimental lines that allowed us to fluorescently label the nuclei in all cells of developing Parhyale embryos; we used a powerful state-of-the-art light microscopy technique, called multi-view Selective Plane Illumination Microscopy (multi-view SPIM), to image intact developing embryos over long periods of time at single-cell resolution; we developed new open-source computational tools that enabled us to navigate through these terabyte-sized image datasets and to reconstruct the cell lineages (cell pedigrees) of different types of developing appendages; and we developed sophisticated genetic tools to monitor gene expression and perturb gene function in vivo. All these advancements are explained in more detail below.
Crustaceans (barnacles, copepods, ostracods, shrimps, lobsters, crabs and their kin) exhibit an incredible diversity in the structure and function of their body units (segments) and associated appendages. This morphological diversity is evident not only across crustacean species but also within the same species, offering exceptional material to study the genetic, cellular and mechanical basis of organ patterning, growth and differentiation in a single developing embryo. However, until recently, it has not been possible to explore the developmental mechanisms underlying segmental specialization and appendage diversification in any crustacean species due to the lack of suitable experimental approaches. During the last decade, the amphipod Parhyale hawaiensis has emerged as the most powerful available crustacean model for developmental genetic studies. Previous work by us and other groups has shown that Parhyale embryos can be subjected to various embryological manipulations and molecular genetic and genomic approaches to address key processes in animal development, like germ layer specification, cell fate specification, nervous system development, organ morphogenesis and regeneration. In this project, we demonstrated that Parhyale satisfies several appealing biological and technical requirements that make it also ideal to image embryogenesis live at high spatio-temporal resolution and over long periods of time. Transgenic embryos with fluorescently labeled nuclei were imaged on a Zeiss Lightsheet Z.1 SPIM microscope every few minutes for about 900 consecutive time points, resulting in 4.5-day long time-lapse recordings that covered appendage development from early specification up to late differentiation stages. In each time point, the embryo was imaged from 5 different views (ventral side, 2 ventrolateral sides and 2 lateral sides) generating 5 series of two-dimensional images (optical sections) covering the entire depth of the embryo.
The amount and kind of data generated by multi-view SPIM raised several challenges for image processing and image analysis. First, each Parhyale embryo multi-view SPIM recording was composed of more than 2 million images resulting in datasets bigger than 7 terabytes! Second, the acquired views in each time-point were aligned to a reference view, a process known as spatial registration. To correct for small drifts of the embryo during long imaging periods, spatial registration was also accompanied by temporal registration against a reference time point. Third, image registration was followed by image fusion, whereby the registered views within each time-point were combined into a single series of two-dimensional images that covered the entire specimen with an isotropic resolution (unlike the original views that have a superior lateral to axial resolution). Fourth, for visualization purposes, the final step involved rendering of the fused data into a three-dimensional representation of the embryo in each time point. Registration, fusion and rendering were iterated for all 900 time points, resulting in four-dimensional representations (three spatial dimensions + temporal dimension) of the embryo as it developed and grew over time (a sample movie can be viewed at http://www.youtube.com/watch?feature=player_embedded&v=ikt7ZcrzDOQ). All these steps required big investments in computing power and ingenious ways for time-efficient image processing. All used computational tools were developed in our group and are available on the Fiji open-access platform for bioimage analysis.
This 4D microscopy approach allowed us to address one of the big challenges in modern biology; capture cell dynamics in a multicellular organism during development as it becomes organized into increasingly more elaborate and specialized tissues and organs. The next key aim of the ANAMORPH project was to make use of the generated recordings to reconstruct the developmental history (cell lineage), as well as the position, movement, division and death of cells over time (cell tracking). For this reason, we developed the Manual Multi-view Tracker tool (MaMuT) that allowed cell lineaging and tracking simultaneously from all different views available in the multi-view SPIM recordings. MaMuT uses the image registration parameters to detect and display all selected cells and their tracks in all available views. A separate linked window displays the cell lineage tree with cells tracked over time depicted as nodes and divisions depicted as branches. The recorded x, y, z, t coordinates for all marked cells can be also exported for downstream analyses. MaMuT was built on another new tool called Big Data Viewer that facilitates on-the-fly browsing through the terabyte-range SPIM datasets. Like previously described software for image analysis, Big Data Viewer and MaMuT are highly user-friendly and available through the Fiji platform. Using these computational tools, we tracked every single cell and reconstructed for the first time complete cell lineages in developing animal appendages, from their early specification up to late differentiation. Furthermore, by comparing different types of Parhyale head and thoracic appendages that acquire distinct size and form, we were able to understand how the developmental history of cells and their various behaviors (like rates and orientation of cell divisions) contribute to organ morphogenesis and diversification.
Besides characterizing the cellular anatomy of developing Parhyale appendages, this project also aimed to study in live imaged embryos the role of gene expression and function in directing morphogenetic cell behaviors. The development of tissues and organs with particular shapes and sizes depends on the regional activity of developmental regulatory genes. For example, our previous studies in Parhyale have demonstrated the capacity of Hox genes to specify different types of head and thoracic appendages. We extended these studies by establishing in Parhyale the BAC recombineering and the flip-out approaches for labeling and manipulation of gene expression in vivo, respectively, to link gene expression dynamics to cell and tissue dynamics. BAC recombineering was used to generate live gene expression reporters by fluorescently tagging Hox genes in the context of their native regulatory sequence instructing when and where the tagged gene is going to be expressed. The flip-out binary system was employed to perturb Hox gene expression in random clones of cells and study the effects in imaged developing appendages at single cell resolution. All these novel genetic tools have been inserted into the Parhyale genome and are expected to provide further insights into the molecular genetic basis of cell fate specification and organ morphogenesis in the near future.
We are currently experiencing an exciting but also challenging transition in Biology and more broadly in the Life Sciences. The synthesis of concepts and methodologies from many disciplines (biology, physics, mathematics, engineering, computer science and others), and the interplay between theory and experiment are beginning to uncover the fundamental principles governing the self-organization and function of living systems, from the sub-microscopic to the macroscopic level. The resources and knowledge generated during the ANAMORPH project for non-invasive image acquisition and analysis methodologies will be of wide interest to all these disciplines. More generally, this kind of new imaging and molecular technologies have a growing impact on basic and applied research and on medicine for diagnosis, prognosis and treatment of disease; therefore, they will have a profound impact on the excellence and competitiveness of the European Research Area, on European health and quality of life, and on the biomedical industry sector. For example, this project contributed valuable expertise to the host institute MPI-CBG on multi-view SPIM microscopy. MPI-CBG is a partner of the Euro-BioImaging infrastructure project, a pan-European network of imaging infrastructure across Member States, and one of the very few research centers worldwide that provides access, training and standardized pipelines and protocols on this cutting-edge imaging technology. During this project, we also worked closely with Carl Zeiss MicroImaging Gmbh on bringing the powerful SPIM technology to the hands of developmental biologists, by testing their prototype instruments and providing extensive feedback for product optimization. Last but not least, the movies and animations generated during this project captivated general public interest during science open day events, and helped establish a constructive and effective dialogue between researchers and citizens about modern biological and biomedical research. Samples of this material have been posted on YouTube (for example see link above and http://www.youtube.com/watch?v=t6U8yJ_zx0U) and on our group’s website (http://www.mpi-cbg.de/research/research-groups/pavel-tomancak/news.html).