• PANCREATIC SECRETION
• PANCREATIC DISEASE
• MOLECULAR GENETICS

Comment on: gutjnl-2016-312423 ‘Human pluripotent stem cell-derived acinar/ductal organoids generate human pancreas upon orthotopic transplantation and allow disease modelling’

Much of our understanding of human diseases comes from the study of model systems such as cell lines. Cell lines are derived from cells that have obtained the property to proliferate indefinitely, often by immortalisation or isolation from cancerous tissues. They have the great advantage that they are easy to work with and can be kept in culture almost endlessly. The disadvantage is that they have lost the genetic signature of healthy primary cells, thwarting the interpretation of test results. This has led to the search for human cell systems that accurately recapitulate healthy or diseased human primary tissues. Advances in stem cell technology have made it possible to create, maintain and expand induced or adult stem cells while they retain multilineage potential. This has led to the development of the so-called organoids, which are defined as a three-dimensional (3D) cellular cluster derived exclusively from primary tissue, embryonic stem cells or induced pluripotent stem cells (iPSCs). Organoids are capable of self-renewal and self-organisation into multiple types of differentiated cells, while maintaining the phenotype of the original tissue. In view of these advantageous properties, organoids have been generated for many of the GI tissues, particularly colon and liver, and also stomach.1 ,2 Pancreatic organoids (POs) have gone through a long developmental phase and have recently matured to their first practical application. Starting with the discovery that pinpointed the lack of insulin as a cause for diabetes in 1920s, scientists attempted to use artificial pancreatic tissue as a path towards a cure. As of 2005, scientists turned to iPSCs to develop in vitro pancreatic cells. Taking a long route from definite endoderm specification,3 towards foregut endoderm and pancreatic endoderm specification,4 and finally ending with the development of pancreatic β cells,5 ,6 we can model pancreatic development from start to finish. One man's particular crusade against the type I diabetes of his children culminated in the production of artificial human glucose-sensitive, insulin-secreting β cells in a petri dish.6 Testament to this scientific tour de force is the addition of α and δ cells to the system. Despite this remarkable progress, development of iPSC-derived pancreatic cells of acinar and ductal lineage has long remained elusive. In 2015, scientists from the group of Hans Clevers in the Netherlands established conditions for the culture of human adult pancreatic stem cells,7 following their culturing recipe of human adult intestinal stem cells.1 Subsequent experimentation led to the mastering of the conditions for 3D culture of pancreatic ductal and acinar cells from iPSCs, opening up a wealth of possibilities to model hereditary diseases of the (exocrine) pancreas.8 One of the most common inherited diseases of the pancreas is cystic fibrosis (CF). An important phenotypical expression of CF is exocrine pancreatic failure which follows the more common pulmonary symptoms. CF is an autosomal recessive disease caused by mutations in the CF transmembrane conductance regulator (CFTR). This gene encodes a chloride channel that is vital for electrolyte and fluid secretion of broncheolar and ductal epithelia. Defective CFTR results in decreased secretion of chloride and increased reabsorption of sodium and water across epithelial cells. This increases viscosity and reduces clearance of secretions from the respiratory tract, pancreas, GI tract and sweat glands. The central pathology of CF is CFTR dysfunction, and fortunately there are many CFTR modifiers in clinical development. These agents activate or correct CFTR channel function. In view of the large number and variable effects of underlying CFTR mutations, the prediction of a clinical response to new therapeutics in a particular patient represents a significant hurdle. This is where organoids can really make the difference. Organoids recapitulate the genotype and phenotype to the level that they are reminiscent of the tissue under study in both architecture and composition. Indeed, primary intestinal organoids from patients with CF show strongly decreased swelling in comparison with those of healthy subjects when induced by forskolin. CFTR chemical corrector VX-809 (lumacaftor) and potentiator VX-770 (ivacaftor) have been shown to restore function of these CF organoids.9 On subsequent testing of rectum-derived organoids, in vitro drug responses did correlate with the outcome data from the clinical trials of these two drugs.10 In this issue of Gut, Hohwieler et al11 further advance the technology and now report a method to produce pancreatic exocrine cells from human pluripotent stem cells using a (mostly) small-molecule based protocol. They use iPSC cells generated from patients with CF to model a CFTR-like phenotype in organoids. This approach has the advantage that it better mirrors the phenotypical, ultrastructural and functional features of mature pancreatic tissue.

The authors started out by achieving high yields of PDX1-positive pancreatic endoderm through stimulation of their set of small molecules. Several iPSC lines were then coaxed into PDX1/NKX6.1-positive pancreatic progenitors (PPs) using growth factors leading to sustained exocrine and ductal marker expression. Pancreatic maturation into cyst-like POs was aided by placement of these PPs in Matrigel facilitating 3D development. These POs expressed specific exocrine markers such as amylase, chymotrypsin C, SOX9 and keratin-19. In addition, organoids displayed carbonic anhydrase activity and CFTR expression, while ultrastructural analysis indicated presence of microvilli, tight junctions and secretory granula, characteristic of the exocrine pancreas. The authors then moved on to disease modelling. By reprogramming keratinocytes derived from the hair of two patients with CF, the authors developed clonal iPSC lines carrying bona fide CF mutations that lead to loss of CFTR function. They convincingly showed that POs with mutant CFTR developed similar to those with wild-type CFTR. Upon transplantation into mice, both type of POs developed into cells bearing markers of ductal, acinar and β cells. The CF phenotype became apparent, because CFTR channel activators forskolin and IBMX elicited luminal swelling of POs with wild-type CFTR but not with mutant CFTR channels. As such, iPSC-derived POs respond to forskolin and CFTR correctors, similar to the rectal organoids originally developed by Clevers. The apparent difference in organoid swelling may serve as a fine read-out for the identification of CFTR enhancers. Indeed, this model can act as a humanised platform for (organ-specific and patient-specific) drug screening and allow testing of the wealth of human therapeutic options, including gene therapy. As such, it embodies true precision medicine, as we may expect that results from drug interventions from a patient-specific PO can be translated faithfully in that patient. More importantly, their model is a significant step closer to the relevant target tissue, while their source materials, keratinocytes, have the advantage that they can be obtained non-invasively. The organoid revolution has expanded to areas outside the laboratory space and has begun to enter the arena of clinical gastroenterology. The shrewd use of organoids by Hohwieler et al to study pancreatic disease is just one example of the wealth of possibilities that are just around the corner. It is easy to predict that organoids will help us to navigate our way through drug development in a time-efficient and cost-effective way.