CORDIS - EU research results

Rules of self-organization and reengineering of liver tissue

Periodic Reporting for period 3 - RULLIVER (Rules of self-organization and reengineering of liver tissue)

Reporting period: 2019-10-01 to 2021-03-31

Understanding the principles of tissue organization and their underlying molecular mechanisms remains an ambitious goal. For this, it is necessary to identify the quantitative rules that govern tissue organization and to link such rules to their underlying molecular mechanisms to provide the means for reengineering the tissue. This requires monitoring of morphogenetic processes in space and time across different levels of complexity, from the molecular to the cellular, and from the cellular to the tissue scale. The pioneering work of Hans Elias in the 1950’s (Elias & Bengelsdorf, 1952) has established what remains today the textbook model of liver tissue organization. Since then, very little progress has been made towards improving the model. In addition, the model describes the tissue but does not make predictions for how the tissue can respond to perturbations.
We address this problem using the mouse liver as model system. The liver is a pertinent example of an organ with a complex 3D tissue organization. It consists of functional units, the liver lobule, containing two intertwined networks, the sinusoids for blood flow and the bile canaliculi (BC) for bile secretion and flux. Sinusoids and BC run antiparallel along the central vein (CV)-portal vein (PV) axis. The hepatocytes are the major parenchymal cells and display a peculiar and unique type of cell polarity distinct from that of simple epithelia. Whereas in epithelia all cells share the same orientation with their apical surface facing the lumen of the organ, hepatocytes are sandwiched between the sinusoidal endothelial cells and share the apical surface with multiple neighbouring hepatocytes to form a 3D BC network. Such an architecture makes it difficult to grasp the 3D organization of cells and tissue from 2D histological sections.
Our main aim is to understand the rules of self-organization of liver tissue and their implementation at the molecular level. Specifically, we plan to address the following questions: First, which are the rules whereby different cell types self-assemble to generate liver tissue? How are cell-cell interaction mechanics integrated with molecular signalling pathways to determine the tissue structure? Second, how can cell polarity be modified to generate the specific organization of hepatocytes? Hepatocytes share many of the molecular machineries expressed in polarized epithelial cells, including liver cholangiocytes, yet have a very different polarity. Therefore, which are the molecular principles responsible for shaping their topologically complex three-dimensional (3D) structure? Are these due to qualitative differences (e.g. hepatocyte-specific genes) or simply the result of quantitative tuning of the same machinery? Third, how are the basic cellular mechanisms, e.g. cell adhesion and cell-cell contacts, modified and/or tuned to generate the specific structure of liver tissue?
The ultimate goal will be to reengineer liver tissue structure as a means of validating our understanding. We will take advantage of a combination of properties of liver tissue to address fundamental questions concerning tissue organization. One of these properties is the tissue dynamics, whereby the liver constantly renews its cells and is capable of regeneration. Another interesting property of tissue organization is that not all hepatocytes are functionally identical within a lobule as different metabolic pathways show gradients of activity between PV and CV, a phenomenon termed metabolic zonation (Gebhardt & Hovhannisyan, 2010). Finally, we will apply newly available technologies to analyse and manipulate it both in vitro and in vivo make it an appropriate model of choice.
Aim 1. In Aim 1 of this project, we worked on the development of a geometrical model of mouse liver tissue. This serves to generate an accurate 3D digital representation of the cells and their essential sub-cellular components in the developing, adult and regenerating mouse liver. We further developed our previously established technology (Morales-Navarrete et al. 2015; Morales-Navarrete et al., 2016; Wiegert et al., 2018), by establishing a pipeline consisting of an appropriate sample preparation, staining for ~20 organelle and pathway markers, microscopy imaging and new algorithms for image processing and analysis. We applied this pipeline to mouse liver tissue sections and used the images to develop and test the different algorithms required for the generation of the geometrical models. Using this pipeline, we investigated the formation of bile canaliculi (BC) in the developing liver through imaging of corresponding tissue samples with sub-cellular resolution. We could follow the polarization of hepatocytes, the expansion of the lumens and their connection leading to BC network formation. For investigations on the regenerating liver, we have been using the partial hepatectomy (70%) model (PH model). We focused on the role of cell division orientation in shaping the liver tissue. We found that, contrary to previous reports, hepatocytes divide in an oriented fashion and it is very likely that oriented cell division is crucial for the formation and orientation of the multi-cellular structural units (see below, Aim 4).
Aim 2. In Aim 2, we investigated the molecular mechanisms whereby hepatocytes are structured, and modify them to re-engineer cell polarity between the simple epithelial and hepatocyte organization. To this end, we needed to develop a variety of techniques to characterize epithelial vs. hepatocyte polarity, culture cells in vitro under conditions that recapitulate tissue features in vivo, and identify candidate genes and pathways accounting for cell fate decisions and cell polarization. We developed super-resolution Correlative Light Electron Microscopy (super-CLEM) method to map different fluorescent signals to the ultrastructure of cellular organelles (Franke et al., submitted). We developed new computational approaches to cluster genes on the basis of their multi-parametric signature. We established in vitro systems suitable to screen numerous genes and test them in corresponding assays. Thus, animal experiments are only necessary for the verification of promising candidates that we have identified in in vitro screens.
As a prerequisite for our investigations, we optimized our novel and unique in vitro cell culture system of primary foetal mouse hepatoblasts. Primary hepatoblasts in this cell culture system differentiate into hepatocytes and recapitulate formation of the bile canaliculi, whereas common epithelial cells, e.g. primary bile duct cells, establish 3D cyst structures with a shared hollow lumen. Primary hepatocytes are furthermore known to lose cell polarity and hepatocellular function within the first 24 hours after isolation and when plated under conventional monolayer culture conditions. However, in our system, these cells recapitulate basic aspects of tissue organization. We performed a focused screen of a set of ~120 genes in total. We were able to identify two genes, previously unrelated to hepatocyte polarity, whose silencing resulted in a switch from the BC towards the common epithelial cyst-like lumen morphology. Preliminary data indicate that the two candidates operate via a joint pathway, but validation experiments are still ongoing.
Aim 3. Based on the results of Aim 1 and Aim 2, we introduce genetic or pharmacological perturbations in vivo to reengineer the structure and function of liver tissue. We first sought to establish a method for silencing genes in the mouse embryonic liver using a mouse strain ubiquitously expressing Green Fluorescent Protein GFP. Through the application of innovative techniques, we are now able to knockdown the expression of GFP in hepatocytes in the embryonic liver.
We began to investigate the candidate genes identified in the in vitro assays (Aim 2). Transiently silencing of the candidate genes in embryonic liver led to encouraging results, which are similar to what we observed in vitro, i.e. the formation of tubular epithelial structures instead of the formation of the BC network.
Aim 4. We used the quantitative image analysis approach to 1) develop a predictive 3D multi-scale model of fluid dynamics (Meyer et al., 2017) and 2) gain insights into liver size control during regeneration (Meyer et al. submitted). We developed a model that simulates fluid dynamic properties from the subcellular to the tissue level. The model integrates the structure of the BC network with measurements of bile transport by intravital microscopy (IVM). The model predicts gradients of bile velocity and pressure in the liver lobule and can be applied to functionally characterize liver diseases and quantitatively estimate biliary transport upon drug-induced liver injury. As proof of principle, we applied and further extended the model to human liver. We identified structural alterations of liver tissue that are predictive for disease progression and provide new insights into NAFLD pathophysiology (Segovia-Miranda et al., submitted).
Using the PH model, we found that the apical surface of hepatocytes forming the BC network expands concomitant with an increase of F-actin and phospho-Myosin, to compensate an overload of bile acids. Interestingly, these changes are sensed by the Hippo transcriptional co-activator YAP, which localizes to apical F-actin-rich regions and translocates to the nucleus in dependence of the acto-myosin system. Using an integrated biophysical-biochemical model of bile pressure and Hippo signalling, we could explain this as a mechano-sensory mechanism that tolerates moderate bile acid fluctuations under tissue homeostasis, but activates YAP in response to sustained bile acid overload (Meyer et al., submitted).
We next analysed the self-organization properties underlying liver tissue. The analysis of the reconstructed 3D tissue geometry using soft-condensed-matter-physics concepts (in collaboration with Frank Jülicher, MPI-PKS Dresden) yielded a breakthrough in the understanding of the organizational principles underlying liver tissue. First, we found that hepatocytes have biaxial cell polarity, in contracts with vectorial polarity found in other epithelial cells (Morales-Navarrete H et al., submitted; Scholich A, et al., submitted). This analysis revealed that hepatocyte cell polarity is not randomly oriented but follow a long-range liquid-crystal order. Second, our quantitative analysis of the reconstructed structures also unravelled the existence of multi-cellular structures as basic building blocks of liver tissue. These stereotypic structural units also follow specific patterns within the liver lobule, consistence with liquid-crystal order of cell polarity.

Elias H, Bengelsdorf H. The structure of the liver of vertebrates. Acta Anat (Basel). 1952;14(4):297-337

Gebhardt R, Hovhannisyan A. Organ patterning in the adult stage: the role of Wnt/beta-catenin signaling in liver zonation and beyond. Dev Dyn. 2010 Jan;239(1):45-55

Morales-Navarrete H, Segovia-Miranda F, Klukowski P, Meyer K, Nonaka H, Marsico G, Chernykh M, Kalaidzidis A, Zerial M, Kalaidzidis Y. A versatile pipeline for the multi-scale digital reconstruction and quantitative analysis of 3D tissue architecture. Elife. 2015 Dec 27;4. pii: e11214

Morales-Navarrete H, Nonaka H, Segovia-Miranda F, Zerial M, Kalaidzidis Y. Automatic recognition and characterization of different non-parenchymal cells in liver tissue. In: 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI) (2016), Piscataway, N.J. IEEE (2016), 536-540

Weigert M, Schmidt U, Boothe T, Müller A, Dibrov A, Jain A, Wilhelm B, Schmidt D, Broaddus C, Culley S, Rocha-Martins M, Segovia-Miranda F, Norden C, Henriques R, Zerial M, Solimena M, Rink J, Tomancak P, Royer L, Jug F, Myers EW. Content-aware image restoration: pushing the limits of fluorescence microscopy. Nat Methods. 2018 Dec;15(12):1090-1097

Christian Franke, Urska Repnik, Sandra Segeletz, Nicolas Brouilly, Yannis Kalaidzidis, Jean-Marc Verbavatz, Marino Zerial. Correlative SMLM and electron tomography reveals endosome nanoscale domains. bioRxiv, May. 6, 2019; doi:

Meyer K, Ostrenko O, Bourantas G, Morales-Navarrete H, Porat-Shliom N, Segovia-Miranda F, Nonaka H, Ghaemi A, Verbavatz JM, Brusch L, Sbalzarini I, Kalaidzidis Y, Weigert R, Zerial M. A Predictive 3D Multi-Scale Model of Biliary Fluid Dynamics in the Liver Lobule. Cell Syst. 2017 Mar 22;4(3):277-290.e9

Kirstin Meyer, Hernan Morales-Navarrete, Sarah Seifert, Michaela Wilsch-Braeuninger, Uta Dahmen, Elly M. Tanaka, Lutz Brusch, Yannis Kalaidzidis, Marino Zerial. Metabolic control of YAP via the acto-myosin system during liver regeneration. bioRxiv 617878; doi:

Fabián Segovia-Miranda, Hernán Morales-Navarrete, Michael Kücken, Vincent Moser, Sarah Seifert, Urska Repnik, Fabian Rost, Alexander Hendriks, Sebastian Hinz, Christoph Röcken, Dieter Lüthjohann, Yannis Kalaidzidis, Clemens Schafmayer, Lutz Brusch, Jochen Hampe and Marino Zerial. 3D spatially-resolved geometrical and functional models of human liver tissue reveal new aspects of NAFLD progression.bioRxiv Mar. 9, 2019; doi:

Hernán Morales-Navarrete, Hidenori Nonaka, André Scholich, Fabián Segovia-4 Miranda, Walter de Back, Kirstin Meyer, Roman L. Bogorad, Victor Koteliansky, Lutz Brusch, Yannis Kalaidzidis, Frank Jülicher, Benjamin M. Friedrich, Marino Zerial. Liquid-crystal organization of liver tissue. bioRxiv, Dec. 13, 2018; doi:

André Scholich, Simon Syga, Hernán Morales-Navarrete, Fabián Segovia-Miranda, Hidenori Nonaka, Kirstin Meyer, Walter de Back, Lutz Brusch, Yannis Kalaidzidis, Marino Zerial, Frank Jülicher, Benjamin M. Friedrich. Quantification of Nematic Cell Polarity in Three-dimensional Tissues. arXiv:1904.08886v1
Through the use of quantitative image analysis and multi-scale phenotypic analysis, we succeeded in developing a state-of-the art 3D model that can make predictions not only how liver tissue is built but also modified in response to genetic perturbations. The conceptual and technological advances obtained will increase our understanding of human liver diseases, currently a major unmet biomedical need. As proof-of-principle we succeeded in re-engineering liver tissue as one of the main aims of this proposal. Using different approaches, we were able to identify candidate genes, previously unrelated to hepatocyte polarity, and to modify cell polarity between the simple epithelial and hepatocyte organization.
Three cutting edge technologies enable this project: First, we have developed a unique cell culture system that enables us to cultivate primary liver cells (hepatoblasts and hepatocytes) in vitro in such a way that they can recapitulate basic aspects of tissue organization. Second, the vertebrate liver is uniquely suited to functional genomics: we are able to silence genes potently and specifically both in primary cells and in the embryonic and adult liver in vivo. Third, we have been developing new image processing and analysis algorithms to extract quantitative information from liver tissue and integrate omics.
Multi-resolution imaging of the mouse liver lobule