Semb Group: Human Stem Cell Biology


Through the use of human and mouse cellular systems our lab aims to identify the signals and cell-cell interactions that underlie the integration of morphogenesis and cell differentiation during organogenesis. Based on the success of our multifaceted approach to research, we are establishing an international program for up-scaled manufacturing of human pluripotent stem cell-derived pancreatic islet-like clusters for clinical testing in type-1 diabetes patientsProf. H. Semb


Research Concept

Basic science forms the foundations of most, if not all medical breakthroughs.  To appreciate what goes awry in disease we must fully understand how cells function in periods of homeostasis.  In many epithelial organs, including the pancreas, tissue morphogenesis is inextricably tied to cell fate decisions.  In our lab we aim to understand the connection between epithelial architecture and the commitment of pancreatic progenitor cells to mature endocrine cells. We believe that to understand this process in a physiologically relevant manner, we must integrate findings from mouse and human pancreatic epithelia. Discoveries from our research are incredibly important for the improvement of differentiation protocols that aim to provide functional, glucose responsive beta cells for use in diabetes cellular therapies.

Research Themes and Questions

The developing pancreas undergoes complex and intriguing epithelial cell rearrangements. The gross morphology of the organ changes from an epithelial sheet to an epithelial bud and finally to a branched/tubular epithelial tree. In parallel, pancreatic progenitors differentiate into multiple cell types and elicit major changes in cell organization including; microlumen and tube formation, acinus formation at the tips of the tubes and islet formation via delamination and clustering of new-born endocrine cells. Whilst published descriptive studies provide a good basis for understanding the pancreatic niche, we aim to refine this by studying architecture and cellular behaviors in a more dynamic manner;

  • In human and mouse pancreatic epithelia – the latter provides a physiological reference point whilst the former allows for a more direct analysis of the cells,
  • With a temporal dimension (i.e. live cell imaging),
  • In a quantitative manner that will enable in silico modeling and deep machine learning to identify novel cell parameters.

Whilst it is apparent that apical polarity links gross morphology, cellular rearrangements and cell differentiation as the pancreas forms, the signals coordinating each step are yet to be elucidated. Our scientists are implementing projects using the techniques stated above in order to address the gaps in our knowledge.

Pancreatic tubes and islets provide niches for alpha and beta cell maturation but are in themselves a culmination of multiple dynamic cellular events. What cellular behaviours coordinate the formation of tubes and islets? Our group has published an important observation that apical polarity cell-autonomously affects the differentiation capacity of endocrine progenitors and is recapitulated in the differentiation of hESCs to endocrine cells in vitro.  Yet, we still do not understand how apical proteins, in a previously non-polarised epithelium, are recruited to the membrane of a pancreatic epithelial cell, extend to neighboring cells to form a plexus that is ultimately refined to a tubular network. A number of projects in the lab are focusing on addressing these issues temporally both in vitro and ex vivo. In addition, we want to determine whether cell movement within the plane and from the epithelium is associated with changes in cellular phenotype and how these behaviours are initiated and coordinated. To fully appreciate how all these events are coordinated requires them to be visualized over time at single cell resolution. Thus we are developing unique reporter lines in human (hESCs) and mouse model systems, including those that visualize apical membrane proteins and endocrine progenitors in parallel. In order to quantify our complex imaging datasets we are developing novel tools to segment and track cells and tubular structures. We strengthen our work in this field through our collaborations with experts at the computer science department of University of Copenhagen (DIKU). It is also our aim to simulate and manipulate 4D cell behaviours in silico. Such computer models will enable us to ‘test’ many variables that could be attributed to specific cell behaviors, such as polarity expansion, cell movement, cell shape changes, before testing our hypotheses empirically.

Does a cell’s niche provide distinct mechano-biological properties that instruct commitment to a specific endocrine fate? Our lab recently found that signaling by the Egfr ligand, betacellulin, diminishes apico-basal polarity leading to upregulation of Ngn3, delamination, and beta cell differentiation, (Lof-Ohlin et al MCB 2017).  Yet, delamination and beta cell commitment are not common to all Ngn3 expressing cells. Our scientists are exploring the hypothesis that unique epithelial architectures instruct endocrine progenitors toward an alpha or beta cell lineage. Indeed, we have already discovered that progenitor cells can be directed towards the endocrine lineage by integrin-mediated mechanosignalling (Mamidi et al Nature 2018). Single cell RNA-SEQ data will distinguish whether polarity status can identify transcriptionally diverse subsets of bi-potent progenitor cells. We are performing screens (micro-pattern chips, hydrogels, phospho/proteomic screens and CRISPR/Cas9 KO screens) to identify unique conditions that will promote expansion or differentiation of bi-potent pancreatic progenitors by inducing changes in the shape of individual or multicellular 3D structures.  We will combine these findings with data from in vivo studies determining the extracellular cues and gene dosage effects that also coordinate a cells transition from endocrine precursor to a beta cell.

Does endocrinogenesis impact on epithelial morphogenesis? In a separate project related to the theme that differentiation and morphogenesis are coordinated, we turn to an observation first published a decade ago that the disruption of endocrine differentiation also affects the patterning of the pancreas. This is a surprising result because in many other organs these processes are either independent or sequential. To understand exactly how differentiation can impact upon epithelial morphogenesis, we will perform a 4D analysis of luminal architecture in endocrine mouse mutants using time-lapse imaging. Taking advantage of the high yield of endocrine cells produced in hESC differentiation cultures we will further use machine learning to characterize novel differences between alpha and beta cells in these heterogeneous cultures. 

Can we improve the yield and functional efficiency of insulin producing beta cells from hESCs? The overarching aim of our lab is to utilize the knowledge we gain from basic research projects to facilitate the generation of high yields of beta cells for diabetes focused clinical applications. The translation of the findings to GMP grade differentiation cultures requires optimization of many aspects of the current protocols in a cost-effective manner, including;

  • Develop protocols for up-scaled manufacturing of hESC-derived pancreatic beta cells.
  • Improving the 3D culture system to parallel the 3D niche of an islet in which functional beta cells respond to glucose stimulation in vivo.
  • Development of robust characterization methods that can routinely give an accurate prediction, in a quantitative manner, of the transplantation outcome of hESC derived islet-like clusters.

Only once the above steps have been optimised will we be able to progress phase 1 clinical trials that would assess the safety and efficacy of transplanting hESC-derived pancreatic islet-like clusters in humans.

Lab Awards and Grants


Ulf Tiemann - was awarded a 2-year postdoctoral research fellowship from the German Research Foundation (Deutsche Forschungsgemeinschaft) to investigate “Epithelial polarity as a fate determinant of multipotent endocrine progenitors during human pancreas development”. In this project, a multitude of fluorescent reporter cell lines from human ESCs are being generated to be able to live-image cell morphology and tissue architecture during the directed differentiation into hormone-producing pancreatic cells.

Silja Heilmann – Was awarded “Det Frie Forskningsråd Sundhed og Sygdom” to develop segmentation and cell tracking algorithms for static and time-lapse 3D datasets of polarity in pancreata. The extracted information will be quantifiable and among other things, used to determine in silco model parameters. The subsequent Lundbeckfonden Postdoc in Denmark grant was awarded to develop mechano-chemical feedback functionality parameters in 3D pancreatic tissue models. These models will allow us to model cell phenotype/mechanical properties depending on local environmental constraints and on ‘memory’ of recent cellular deformations.

Pia Nyeng - Has been the recipient of a number of grants and awards during her research career.  Most recently she has been awarded a travel grant from Lundbeck Fonden, a 1st prize in the Danish Bioimaging Image Competition 2017 and a Novo Nordisk Foundation project grant, co-investigator (2.1 mill). to quantify the topological events connecting tubulogenesis and beta-cell differentiation.


The lab has filed for a number of innovative patent based on research from the Semb lab, including:

A total of 8 patents have been applied for and/or granted based on innovative research from the Semb lab. Of note are the following patents;

  1. technologies allowing isolation of pancreatic beta cell progenitors from hPSCs (Ameri et al 2017),
  2. new targets for controlling expansion of such progenitors (Ameri et al 2017),
  3. a novel approach to turn these progenitors into functional beta cells (Löf-Öhlin et al 2017; Mamidi, Prawiro et al, 2018).

Recent Publications

Total number of peer reviewed publications and reviews: 98, H-factor: 39

The study has identified specific factors in the matrix surrounding naive pancreatic cells (progenitors) that act to maintain the ‘gatekeeper’ protein YAP. When the progenitor cells are exposed to an alternative matrix factor, they lose YAP and commit toward a mature cell identity with the potential to become cells that produce insulin in response to glucose.

An integral discovery uncovering a link between endocrine cell differentiation and the size of the endocrine progenitor’s apical domain.

This study has identified a cell surface marker that selects pancreatic endoderm cells from heterogeneous cultures of hESC derived pancreatic epithelia. This pure population of pancreatic progenitors is able to give rise to functional B-cells and paves the way for cellular therapies to treat diabetes.

This study provides mechanistic insight into how Wnt signalling controls early cell lineage decisions, namely that of neural crest cells

Beta-cell delamination and Beta-cell differentiation both require tightly regulated levels of the RhoGTPase Cdc42.

  • Wolfhagen Sand, F., Hörnblad, A., Johansson, J.K., Lorén, C., Edsbagge, J., Ståhlberg, A., Magenheim, J., Ilovich, O., Mishani, E., Dor,Y., Proia, R.L., Ahlgren, U. and Semb, H. (2011) Growth-limiting role of endothelial cells in endoderm development.Dev Biol, 15, 267-77.

The discovery that endothelial cells limit progenitor cell expansion in several foregut-derived organs.    

  • Fischer, Y., Ganic, E., Ameri, J., Xian, X., Johannesson, M. & Semb, H. (2010). NANOG reporter cell lines generated by gene targeting in human embryonic stem cells.PLoS One 5(9),doi:10.1371/journal.pone.0012533.                                                                                                            
  • Ameri, J., Ståhlberg, A., Pedersen, J., Johansson, J.K., Johannesson, M., Artner, I., and Semb, H. FGF2 Specifies hESC-Derived Definitive Endoderm into Foregut/Midgut Cell Lineages in a Concentration Dependent Manner. Stem Cells, 28, 45-56, 2010.

The identification of how FGF signalling in a concentration-dependent manner controls specification of hESC-derived definitive endoderm into liver, pancreas and lung endoderm.

  • Kesavan, G., Wolfhagen Sand, F., Greiner, T., Johansson, J., Kobberup, S., Wu, X., Brakebusch, C., and Semb, H. (2009).Cdc42-mediated tubulogenesis controls cell specification. Cell, 139, 791-801.

The identification of the molecular machinery that controls tube formation and how tubes controls cell specification.

The Semb Group