University of Groningen Development of MAPC derived induced endodermal progenitors Sambathkumar, Rangarajan

(1)

University of Groningen

Development of MAPC derived induced endodermal progenitors Sambathkumar, Rangarajan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2017

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Sambathkumar, R. (2017). Development of MAPC derived induced endodermal progenitors: Generation of pancreatic beta cells and hepatocytes. University of Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

177

Chapter 6

General Conclusions and Future Perspectives

(3)

Chapter 6

6. General Conclusions and Future Perspectives

In the field of regenerative medicine, numerous groups are in search of novel and renewable cell sources to generate mature functional insulin producing β-‐cells to treat diabetes. Likewise, renewable sources of hepatocytes are also highly sought after for pharmaceutical studies as well as regenerative medicine.

(6.1) In the first part of this thesis, I tested if human MAPCs, that can be easily isolated from bone marrow and expanded for up to 70 population doublings could be used to create these cell types.

In contrast to rodent MAPCs, that generate hepatocyte-‐like cells and β-‐cell-‐like cells, human MAPCs are less potent. Therefore, I addressed the question by developing a method to transdifferentiate hMAPCs into a population of expandable endodermal progenitors that can subsequently be differentiated to β-‐cells and / or hepatocytes. Initially I selected 16 TFs based on their role in the endodermal lineage cell fate specification. Upon transduction in MAPC and culture in supportive medium, epithelial clusters of cuboidal cells could be found, accompanied by expression of endodermal but also more mature hepatocyte markers. Based on insight from studies of direct reprogramming of fibroblasts into induced hepatocyte like cells (iHEPs) we concluded that transduction of FOXA2 and HNF1A as part of the 16TFs might be responsible for the generation of cells already fated to the hepatocyte lineage. Therefore, I removed FOXA2 and HNF1A from the pool of 16 TFs, and transduced hMAPCs with 14TFs. This caused the generation of cuboidal epithelial cells that expressed definitive endoderm markers but not express late mature hepatic and pancreatic endoderm markers. The resultant cells could be expanded up to 20 passages without significant changes in the expressed gene profile, and were named iENDO cells.

We tested if the removal of the four pluripotency factors, OCT4, SOX2, KLF4 and CMYC would allow generation of iENDO cells, but found none of the morphological and gene expression changes that were seen when 14 or 16TFs were used. However,

(4)

Chapter 6 General conclusion and Future perspectives

179

factors and cultured in the same induction medium, cuboidal epithelial cells could be seen, but the cells did not express endodermal markers and could not be expanded. Finally, we found that the endodermal induction medium containing Activin-‐A and Wnt3A, low concentration BMP4 was essential for the generation of iENDO cells in combination with the 14TFs. It should be noted that the endodermal medium alone did not induce an iENDO phenotype in hMAPCs.

Thus, a combination of pluripotency factors and 10 other endodermal TFs as well as the endoderm supportive medium are required for creation of iENDO cells. Future studies should address if the 14TF combination can be further reduced, which might be possible based on studies wherein a single TF at the time was removed, demonstrating that only removal of HNF6 prevented creation of cuboidal cells (Supplementary figure-‐1 A-‐C). To perform such studies I hypothesized that creating a MAPC line using CRISPR-‐Cas9 nucleases mediated knock-‐in reporter, knock-‐in of a fluorochrome in the SOX17 or FOXA2 gene locus (without affecting the transcription of the gene) might enable the identification of the importance and the role of the 9 other endodermal TFs in the creation of iENDO cells with a minimal complement of TFs.

I demonstrated that 14TF iENDO cells grafted in Immunodeficient mice form highly vascularized tumors that contain only endodermal lineage restricted cell types, including hepatocytes or pancreatic endocrine cells, and less frequently intestinal cell types. However, the in vivo differentiation varied between grafts. For example in one mouse hepatic endoderm was dominant while in another mouse pancreatic endoderm and intestinal cell types were dominant, but less hepatic endoderm. Blood of grafted mice contained human albumin 1 and 3 months after grafting but negligible amounts of C-‐peptide. For a more reliable comparison with other studies it will be necessary to measure human albumin serum levels weekly. The differences in differentiation in vitro could be due to the heterogeneity of iENDO cells. In addition, it is not known yet which factors drive in vivo differentiation of iENDO cells into different endodermal cell types. Future research will be required wherein more mice are grafted to gain insight in the mechanisms underlying spontaneous

(5)

differentiation, and using FACS purified CXCR4 positive iENDO cells, to decrease the heterogeneity of the cells grafted.

It is highly likely that the tumor formation is caused by the persistent expression of the OCT4 transgene in the grafted iENDO cells. To overcome this problem, future research should test if iENDO cells can be generated using doxycycline-‐controlled expression of (OCT4, SOX2 KLF4 and CMYC) allowing eliminating expression of OCT4 in grafted cells by omitting docycycline. I hypothesize that this would prevent tumor formation. However, unknown is whether this would also prevent endodermal differentiation and maturation.

Based on the in vivo differentiation potential of iENDO cells, I investigated the in vitro differentiation of iENDO cells in 3D organoids to hepatocyte-‐like cells and pancreatic endocrine cells using modifications of established differentiation protocols. Differentiated cells expressed hepatic and pancreatic endocrine markers at levels relatively similar to those of hESC differentiated into the same cells types. However, functional mature hepatocyte and β-‐cell markers were not expressed at the same level in iENDO-‐differentiated cells compared to the primary hepatocytes

and primary human islets. The immature phenotype of iENDO cells could be caused

by the persistent expression of OCT4. In addition, at least for stem cell derived hepatocytes, differentiated human PSCs also do not yet generate fully functional hepatocytes. Therefore, further improved differentiation protocols will be required, for example by adding small molecules and missing growth factors, metabolic regulators, overexpression of missing TFs and non-‐coding microRNAs, epigenetic modifiers, and / or changing the extracellular matrix. Furthermore, co-‐culture of iENDO cells with other cell types (such as organ specific mesenchymal cells and endothelial cells) present during in vivo liver and pancreas development might have a positive influence on maturation. Once more mature cells can be generated, functional maturity should be shown by C-‐peptide or insulin secretion following glucose stimulated for β-‐cells; and phase I and II detoxification assays, glycogen, urea production, coagulation factor production, should be done to determine functional hepatocyte generation. Finally, transplantation of predifferentiated iENDO

(6)

181

hepatocyte-‐like and β-‐cell like cells will be needed to demonstrate their functionality.

Preliminary studies suggested that hepatocyte committed iENDO progeny did not result in the formation of big tumors, suggesting that the in vitro pre-‐differentiation reduced the uncontrolled cell proliferation. Immunostaining revealed that the grafts contained a high number of ALB expressing cells and lower levels of AFP, suggesting that in vivo maturation did take place. However, OCT4 positive cells were still found in the grafts. However, additional mice will be required for full enumeration of the different cells present in the grafts

Once we can demonstrate that switching off OCT4 does not interfere with the final maturation of iENDO hepatocytes, it will be of interest to test of iENDO-‐hepatocyte like cells can repopulate an injured mouse liver, using for instance Alb-‐uPA mice

(Urokinase plasminogen activator)/SCID mice or Fah−/−_/Rag2−/−_/Il2rg−/−_{mice
(FRG}

mice) or chemically induced liver injury mice (Retrosine, CCL4, or other chemical

xenobiotics mouse models).

For iENDO cells differentiated to β-‐cell like cells in vitro, transplantation of organoids with or without alginate encapsulation, in diabetic or non-‐diabetic mice is ongoing. Gene expression and immunostaining of the grafts should be conducted to define if further maturation is seen in vivo, as we found for iENDO cells committed to the hepatocyte lineage. In addition, glucose responsiveness of the mice and rescue from diabetes should be analyzed. In addition, future studies should also test the ability of pancreas committed iENDO cells wherein the OCT4 gene can be switched off can mature to functional β-‐cells in vivo.

Overall, I can conclude that hMAPC cells can be transdifferentiated into expandable iENDO cells. In vivo, iENDO cells can differentiate towards more mature endodermal cell types, but with the formation of an endodermal tumor. iENDO cells can also differentiate towards pancreatic endocrine-‐like cells and hepatocyte-‐like cells in 3D-‐ spheroid cultures in vitro, although the cells remain immature.

(7)

(6.2) Epigenetic induction of definitive and pancreatic endoderm cells fate (linked to chapter 4)

A number of studies have shown that pre-‐treatment of fibroblasts or other somatic cells with epigenetic modifiers prior to TF overexpression or addition of small molecules after TF overexpression can promote iPSCs generation and direct reprogramming of one somatic cell into another. In mice and rats pancreatic and non-‐pancreatic endodermal cells types can be reprogrammed into cells with some insulin producing β-‐cell features by overexpression of three key pancreatic TFs, PDX1, NGN3 and MAFA. However, overexpression of these TFs in mouse and human fibroblasts have not succeeded in the creation of β-‐cells till now. To overcome this problem several studies have tested if mouse and human fibroblasts can be reprogrammed into pancreatic β cells by a combination of epigenetic modifiers and specific culture media alone and/or in combination with transient expression of pluripotency TFs. However, these studies were performed using embryonic origin of fibroblasts and it is not known yet whether this might also work for adult human fibroblasts.

In this thesis, I demonstrated that the histone deacetylase inhibitor (HDACi), Trichostatin A (TSA) combined with the chromatin remodeling medium (CRM) can induce transiently the expression of definitive endoderm and pancreatic endoderm markers in adult human fibroblasts and this without inducing pluripotency, hepatocyte and endothelial markers, even if the skeletal muscle master gene, MYOD1, was also induced. Combining TSA with the DNA methyltransferase inhibitor (DNMTi), 5’Azacytidine (5’AZA), did not increase the expression levels of the endodermal and pancreatic genes compared to TSA treatment alone. However, removal of TSA and CRM caused gene expression to return to baseline.

Therefore, I hypothesize that to convert adult fibroblasts permanently into insulin producing functional β cells, overexpression of key pancreatic β-‐cell TFs such as PDX1, NGN3, MAFA and PAX4, NKX6.1 and / or NEUROD1 before or after addition of the epigenetic modifier, TSA, and culture in β-‐cell specific conditions may promote generation of insulin producing β-‐cells from adult fibroblasts. Likewise it will be of

(8)

183

with the TFs used to create iENDO cells in Chapter 3 would enhance the ability to transdifferentiate hMAPCs in iENDO cells and subsequently in mature endodermal progeny.