Kidney tubulogenesis in human renal organoids: an investigation into the contribution of the microtubule-associated protein DCDC2

Abstract

During early human kidney development, the cells of the metanephric mesenchyme repeatedly undergo a mesenchymal to epithelial transition (MET). This forms an initial tube which subsequently extends into the nephrons, which are the functional unit of the kidney. When kidney nephrons are lost due to disease or damage, kidney function is reduced because the total number of nephrons in an individual is formed pre-natally and nephrons cannot be added after birth. Therefore, understanding the process of kidney nephron formation is critically important to human kidney health. Kidney organoids present an exciting system for investigation of human kidney MET events. These are inaccessible in vivo and require a 3D signalling environment, limiting the suitability of classical cell culture. A key step during renal MET is the rearrangement of cytoskeleton components, however the contribution of these components to driving renal MET is not well defined. In this thesis I use human kidney organoids derived from induced pluripotent stem cells (iPSCs) to study the role of the cytoskeleton in MET and kidney tubulogenesis. I initially assessed and demonstrated that the human renal organoids match the key mesenchymal and epithelial stages seen in the human foetal kidney and published organoid data. Using analysis of both epithelial and mesenchymal markers, I conclude that the organoid protocol used here was suitable for modelling the metanephric mesenchyme to early nephron stages of development. I then demonstrate that both actin filaments and the microtubules, two major cytoskeletal systems, undergo drastic rearrangements across the MET process in kidney organoids. Both cytoskeletal systems are controlled by a large variety of regulators. Such as proteins affecting nucleation, polymerisation and depolymerisation as well as crosslinking or bundling, and so my further focus was on such factors. In particular for the regulation of the microtubule rearrangement in the polarising nephron precursors. One form of microtubule regulation is by microtubule associated proteins (MAPs) and interrogation of a single cell RNA sequencing data set of human renal organoids for upregulated MAPs across the time course of MET, was completed. This highlighted doublecortin domain containing 2 (DCDC2) as a promising candidate. Based on a literature search DCDC2 is interesting because it acts as a MAP, it is highly upregulated in the epithelial portions of the human kidney and mutations in DCDC2 produce human kidney disease. In this thesis I show that DCDC2 localised to the microtubules and cilia of epithelial cells in renal organoids, and that this matches with published RNA sequencing data sets for both organoid and human foetal kidney. I have then utilised DCDC2 knockdown and overexpression experiments in organoids to investigate DCDC2 function during MET. Knockdown of DCDC2, using a CRISPR/dead Cas9 -based system, caused a failure in the MET process, with a loss of all epithelial cell types despite preservation of the metanephric mesenchyme fate. Furthermore, immunofluorescence staining in these organoids showed that the cytoskeleton was not rearranged and retained the mesenchymal patterning. Overexpression of DCDC2 in organoids significantly increased the volume of epithelial tubes present on day 14, suggesting that DCDC2 could act as a driver of renal MET. However, overexpression of DCDC2 in mesenchymal cells was not sufficient alone to drive MET in these cells and DCDC2 over expressing cells with mesenchymal morphology were present in organoids. I am proposing two mechanisms for DCDC2 involvement in MET: firstly, that DCDC2 is important for cilia-based modulation of the wnt pathway and secondly that DCDC2 stabilisation and bundling of the microtubules is required for MET. In DCDC2 knockdown organoids epithelial cilia length remained unchanged, however changes to cilia in overexpression organoids have been observed. Published studies about DCDC2 in mouse spheroids show DCDC2 knockdown phenotypes can be rescued by chemical inhibition of Wnt signalling (iCRT14) and data from our group has shown that continued high Wnt pathway activation prevents tubule maturation. Thus, a link for DCDC2 function to cilia-mediated Wnt modulation is plausible. If DCDC2’s function in renal MET is to bundle and stabilise microtubules, then the overexpression of DCDC2 should present with a clear increase in microtubule bundles, but this was not observed, indicating that microtubule budling might not be at the core of DCDC2’s function in renal MET. However, more evidence is needed to conclude which of these mechanisms is responsible for DCDC2 control of MET. In summary, in this work I show that actin and microtubules cytoskeletal systems become reorganised in MET, demonstrate the importance of the MAP DCDC2, and test two hypotheses through loss- and gain-of-function approaches. Finally, I propose a set of further experiments to investigate DCDC2 function during MET in more detail.