Embryonic Stem Cells Derived Kidney Organoids as Faithful Models to Target Programmed Nephrogenesis

The kidney is a complex organ that is comprised of thousands of nephrons developing through reciprocal inductive interactions between metanephric mesenchyme (MM) and ureteric bud (UB). The MM undergoes mesenchymal to epithelial transition (MET) in response to the signaling from the UB. The secreted protein Wnt4, one of the Wnt family members, is critical for nephrogenesis as mouse Wnt4−/− mutants fail to form pretubular aggregates (PTA) and therefore lack functional nephrons. Here, we generated mouse embryonic stem cell (mESC) line lacking Wnt4 by applying the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated systems 9 (Cas9). We describe here, differentiation of the wild type and Wnt4 knockout mESCs into kidney progenitors, and such cells induced to undergo nephrogenesis by the mouse E11.5 UB mediated induction. The wild type three-dimensional (3D) self-organized organoids depict appropriately segmented nephron structures, while the Wnt4-deficient organoids fail to undergo the MET, as is the case in the phenotype of the Wnt4 knockout mouse model in vivo. In summary, we have established a platform that combine CRISPR/Cas9 and kidney organoid technologies to model kidney development in vitro and confirmed that mutant organoids are able to present similar actions as in the in vivo studies.

collecting duct structures [13][14][15][16][17] . The 3D kidney culture technology, allow human and mouse PSCs to exhibit their remarkable self-organizing properties depicted by appropriately segmented structures of nephrons 14,16 . The genome engineering technique, the CRISPR/Cas9 gene editing, provides an unprecedented opportunity for studying kidney disease and development with hPSCs ex vivo. These techniques provide new resources for modelling and studying human kidney development and disease. The induction of renal lineage has been conducted using mESCs [18][19][20][21][22][23][24] and these mESCs-derived nephron progenitor cells can be induced to nephron structures by spinal cord 17 . These findings show that mESCs have important potential for modelling the development as well as regeneration.
Here, we report a novel setting to be able to combine CRISPR/Cas9 with kidney organoid technologies to model kidney development. By using the CRISPR/Cas9 technique, we generated the Wnt4 deficient mESCs. We programmed differentiation of wild type and mutant mESCs into kidney progenitors and through the interaction with UB were able to induce nephrogenesis and generate kidney organoids ex vivo. We demonstrate that the Wnt4 CRISPR-knock out cells, generate kidney organoids which fail to advance the MET and lead to failure in nephrogenesis. Taken together, these results depict an innovative platform for mouse kidney development modelling and regenerative medicine application for detailed molecular genetic studies.

Results
Generation of Wnt4 deficient mESCs with double nicking by RNA-guided CRIPSR/Cas9. The mouse Wnt4 gene consists of five exons; previously reported conventional Wnt4 knockout mouse model, generated a probable null allele by replacing the whole exon 3 with a neo selection cassette 10 . To analyze the role of Wnt4 during kidney organoid development in vitro, we generated a Wnt4 deficient mESC line using the CRISPR/ Cas9 genome editing technology. We used a pair of small guided RNAs (sgRNAs) guiding paired Cas9 nickases to knockout genes in mESCs, which have been shown to reduce the off-target activity and facilitate gene knockout efficiency in cell lines 25 . We designed the sgRNAs to target Wnt4 exon 2 (Fig. 1A), and constructs encoding GFP or mCherry-tagged Cas9 and sgRNAs were electroporated into the wild type mESCs. GFP and mCherry co-expressing cells were FACS sorted and positive clones were picked and expanded (Fig. 1B). Sanger-sequencing results revealed the knockout mESC line with one allele 10 bp and another allele 17 bp deletion in the Wnt4 exon 2 (Fig. 1C).
We observed that the Wnt4 −/− mESCs colonies were indistinguishable in size and shape from unmodified mESCs (Fig. 1D), and presented similar expression level of mESC markers such as Sox2, Oct4 and Nanog ( Fig. 1E and Supplementary S1A), indicating that the Wnt4 −/− cells maintained pluripotency and self-renewal properties.

Induction of intermediate mesoderm differentiation in the mouse ES cells.
In order to use the CRISPR Wnt4 knockout mESCs to model kidney development in vitro we have optimized protocol to generate kidney organoids from wild type mESCs. mESCs were isolated from mouse blastocysts at E3.5 26,27 . Activation of LIF-Stat3 or Wnt/β-catenin signaling promotes mESCs self-renewal [28][29][30] and CHIR99021, an inhibitor of the GSK-3, induces non-neural differentiation 28 . We have therefore treated the mESCs with the 5 h pulse of CHIR99021 in monolayer cultures. Cells presented expression of the epiblast markers such as Fgf5 and T (Brachyury) but not the extraembryonic endoderm marker Afp ( Fig. 2A, step A1 and Fig. 3A,C), suggesting direct epiblast differentiation. At 24 hours, Fgf5 was downregulated while the T marker was upregulated depicting early primitive streak differentiation ( Fig. 2A, step A1 and Fig. 3A,C).
The canonical Wnt signaling pathway induces primitive streak development in mESCs and hESCs 31,32 . The intermediate mesoderm (IM) arises from the primitive streak. Fibroblast growth factor 9 (Fgf9) is expressed in IM 33 and FGF9 signaling supports MM differentiation in vitro 34 . Previous studies demonstrated that differentiation of IM from hPSCs requires CHIR99021-to-FGF9 13,14 , we therefore analyzed whether mESCs-derived epiblast cells reacted similarly to the Wnt and FGF signaling. We treated these cells with CHIR99021 for 48 h and found expression of primitive streak markers such as Mixl1, T and Tbx6 (Fig. 2, step A1 and Fig. 3A,D), implicating early mesendoderm differentiation. Indeed, FGF signaling was required for generation of the IM cells. In the presence of FGF9 and heparin, these cells advanced their differentiation and their growth dynamics changed from a monolayer-type towards cellular clusters (Fig. 3B). The cells isolated at day 8 of differentiation expressed Osr1, Pax2, Lhx1, Wt1, Ctied2, Eya1, Hoxd11 and Gata3 genes (Fig. 3E). These genes are typically expressed in the cells of IM associated with kidney lineages 19 . In summary, we have established a novel protocol to induce the mESCs towards the IM stage in a monolayer culture setting with CHIR99021-FGF9/heparin supplementation (Fig. 3A).
Generation of kidney organoids from mESCs-derived nephron progenitors with embryonic UB. Nephron Progenitor Cells (NPCs, Six2+ cells) purified from mouse embryonic kidney present long-term self-renewal properties when cultured in the medium enabling NPC self-renewal (NPSR) 35 . The NPSR medium mimics the in vivo nephron progenitor niche by delivering important small molecules and necessary growth factors to maintain undifferentiated progenitor cell proliferation and self-renewal 35 . To test, whether NPSR can indeed induce the NPCs from the mESCs-derived IM cells, we harvested these cells at day 8 of differentiation and cultured them as 3D pellets in the NPSR medium overnight (Fig. 2, step A2 and Supplementary Fig. S2A). This process enhanced expression of the NPC markers, such as Cited2, Wt1, Hoxd11, and Six2 (Fig. 4A,B). Incubation of the mESCs-derived IM cells in the NPSR medium overnight was a necessary step in priming these IM cells for renal differentiation. If this step was omitted, the nephrogenesis in the IM cells failed to be induced by the UB (Supplementary Fig. S2C).
We reported recently that the UB separated from mouse embryos (E11.5) induces MET and nephrogenesis in an intact and dissociated to single cells and re-aggregated E11.5 MM 36 . Hence, we have used the mouse E11.5 UB tissue as a potent nephrogenesis inducer. We aggregated "primed" 3D pellets with the UB and cultured in 3D Trowell culture system. This led to successful nephrogenesis induction (Fig. 2, steps B1-AB5 and Fig. 4C To verify, that the kidney organoid structures were generated via the interaction between the UB and the mESCs-derived kidney cells, but not by contaminated UB tip cells with the primary MM cells, we cultured the E11.5 UB tissue in isolation in the 3D culture. The UB cells underwent apoptosis already at the second day of culture and died at day 3 ( Supplementary Fig. S2B). In addition, there is lack of nephron structures formation when "un-primed" IM cells were aggregated with UB or the IM cells "primed" in NPSR medium were transferred to 3D  culture without integration with the UB (Supplementary Fig. S2C,D). These data suggest that the mESCs-derived IM cells can differentiate into nephron progenitors via a "priming step". Such cells are also competent to undergo the MET and generate 3D kidney organoid when induced with the embryonic UB.
Wnt4 regulates nephrogenesis in kidney organoids. Wnt-signaling play multiple roles in different tissues during development. It regulates the pattern formation, cell fate choices, cell renewal, proliferation and migration 28,[37][38][39][40] . Wnt4 is important for kidney development; it is the mesenchymal signal for epithelial transformation of MM in the developing kidney 10 and is expressed in the kidney mesenchyme and its derivatives, namely the nephrons [41][42][43][44] . Mouse mutants with loss of Wnt4 fail to form PTAs and therefore lack functional nephrons 10 .
To investigate whether Wnt4 would play the same role during kidney organoid development, we generated the Wnt4 −/− organoids with the CRISPR/Cas9 knockout Wnt4 mESC line. The differentiation conditions of Wnt4 −/− organoids were the same as the wild type mESCs organoids ( Fig. 2A). There is no visible morphology difference between wild type and Wnt4 −/− mESCs when differentiated to IM population ( Fig. 5A and Supplementary S2B); 3D pellets show no distinction before and after priming step (Supplementary S2A and S3D). In addition, Wnt4 −/− mESCs differentiated to kidney lineage present expression of the same markers as the wild type cells during all differentiation stages (Supplementary S3A-C, E,F). These data show that Wnt4 −/− mESCs were able to differentiate into kidney lineage.
We used these Wnt4 −/− kidney lineage cells to aggregate with the wild type UB to make the organoids ( Fig. 2A, step CB5). In contrast to the wild type organoids (Fig. 4C), the 3D morphology of the Wnt4 −/− organoids appeared to be flatter (Fig. 5B). These organoids failed to undergo nephrogenesis and did not generated any kidney structures, as depicted on Fig. 5C with immunostaining for nephron specific markers ( Fig. 5C and Supplementary Fig. S4A). The, Wnt4 −/− organoid failed to undergo the MET and subsequently kidney development failed, as is the case in the in vivo model (Fig. 2A,B and Supplementary Fig. S4B). In conclusion, Wnt4 signal is crucial to regulate epithelial transformation of nephron progenitors in the developing kidney organoids. Moreover, organoids are a good tool to study and model development.

Discussion
Recent advances in genome editing and stem cell-derived kidney organoid technologies provide the possibility to perform sophisticated genetic studies in PSC-derived kidney lineage. Several groups reported using gene-editing settings to study kidney development and disease in the organoids. Recapitulated features of glomerular and tubular diseases by transiently transfecting undifferentiated hPSCs with plasmids expressing wild type Cas9 and sgRNAs targeting disease-relevant genes PODXL (podocalyxin like), polycystic kidney disease (PKD) genes PKD1 and PKD2 45,46 and TALEN knockout PAX2 to study UB development in vitro 16,47 . Here, we report for the first time using the double nicking by RNA-guided CRISPR/Cas9 technique to knockout of Wnt4 in the mESCs and study kidney development in vitro.
Mouse ESC lines possess the ability to differentiate into a variety of cell types, and therefore are a source of cells for functional studies 48,49 . Usually, mESCs differentiation starts with embryoid body (EB) formation as a way of following normal developmental events that take place in the embryo 50   epiblast stage cells into IM cell population using CHIR99021-FGF9/heparin treatment. Our studies revealed that using Wnt and Fgf signaling we could efficiently differentiate mESCs into kidney precursor cells in 2D cultures. With previously reported high efficiency of maintaining nephron progenitor's stemness in NPSR medium 35 , we managed to direct the differentiation of IM into nephron progenitors in a 3D pellet in this medium. This incubation -"priming" step, appeared to be crucial in derivation of NPC from IM that would be competent to undergo nephrogenesis. Our method illustrates that the nephron progenitors derived from mESCs gain the potential to interact with the functional UB and are competent to advance the nephrogenesis. Human WNT4 mutation causes kidneys dysgenesis syndrome and mouse knockout of Wnt4 leads to a failure in kidney development (kidney agenesis). Our studies of Wnt4 knockout in mESCs illustrate that combining of CRISPR/Cas9 genome editing technology to determine the function of specific genes in different mouse cell types, with 3D generation of organoids leads to similar effects as an in vivo study. Moreover, we believe that the mESC Wnt4 −/−derived NPCs behave similarly to MM. These cells express many markers of competent MM before combining them with freshly isolated UB ( Supplementary Fig. S3E), however, after 8 days of culture as an organoid, the Wt1 and Pax2 are lost (Fig. 5) and there are no nephrons present. Normally, MM stays competent for induction signal from the UB for 24-36 h 51 , and lack of it leads to apoptosis of these cells. Also UB undergoes apoptosis if no signals reach it from the MM 52 (Supplementary Fig. S2B). Given that, in these mutant organoids the Wnt9b from the UB has no cells to act upon (all being Wnt4 −/− ) 10 , it is no surprising that the nephrogenesis fails and all MM markers are lost. The CRISPR/Cas9 technology has enabled efficient creation of various disease models, in our work and of others 45,53 proving that generation of renal organoids from gene-modified PSCs provide an excellent tool and endless possibilities to model kidney development and disease. These are important breakthroughs, which will promote the development of regenerative medicine.
In summary, we showed here wild type mESCs-derived nephron progenitors aggregated with primary UB formed kidney organoids with full nephron structures. While genome-modified nephron progenitors (Wnt4 −/− ) aggregated with UB formed mutant organoids with failed kidney development, which functionally recapitulate kidney development phenotypes in vivo. The described methodologies ( Fig. 2A) are broadly relevant for functional studies of factors involved in development and their potential in regenerative medicine. In the long term, this system may provide a useful setting that will benefit personalized medicine and gene therapy.

Materials and Methods
The animal care and experimental procedures in this study were in accordance with Finnish national legislation on the use of laboratory animals, the European Convention for the protection of vertebrate animal used for experimental and other scientific purposes (ETS 123), and the EU Directive 86/609/EEC. The animal experimentation was also authorized by the Finnish National Animal Experiment Board (ELLA) as being compliant with the EU guidelines for animal research and welfare.

Mouse ESCs culture. All experiments used the wild type mouse embryonic stem cells derived from
Taconic's W4/129S6 inbred mouse strain. Undifferentiated wild type and the CRISPR/Cas9 knockout mESCs were maintained on the mouse embryonic fibroblasts (MEFs) as a feeder layer with mESCs medium as previously reported 54 . Paired oligoes corresponding to Wnt4 gRNA2 (5′-GCGCTGGAACTGTTCCACAC-3′) were cloned into pSpCas9n (BB)-2A-mCherry vector. Paired GFP and mCherry constructs were co-electroporated into mESCs. GFP and mCherry double positive cells were isolated by flow cytometry sorting (FACS) two days after electroporation, and immediately plated onto the 10 cm MEF coated plate. One to two weeks later, there were colonies growing in the culture plate. Using 100 μl pipette tips we picked up the colonies and placed them individually in the 96 well culture plate filled with trypsin. After dissociation into single cells, the colonies were transferred onto MEF-coated 24-well plates and then expanded in 10 cm plates separately. Further, the genotyping by PCR/TA-cloning and chromatogram sequencing were used to analyze the mutations and select the positive clones for analysis and differentiation experiments. The primers (5′-3′) used for genotyping: Wnt4 forward: GTATCACATCCAACCACTG, reverse: AGAAGCCTGATGCCAAGGGA.
Cell differentiation. Mouse ESCs were cultured in Matrigel-coated 6 cm culture dishes, in mESCs medium until reaching 70-90% confluency. mESCs were passaged on Matrigel-coated 6 cm plates at 30,000 cells/cm 2 . Next day, cells reached 80-90% of confluency and were treated with 8 μM CHIR99021 in APEL basal medium (STEMCELL Technologies) for 4 days, followed by FGF9 (200 ng/ml) and heparin (1 μg/ml) treatment for another 4 days; medium was changed every other day. Following the differentiation, there were some floating cells; these were apoptotic cells and were removed during changing medium.
3D kidney organoids formation. At day 8 of differentiation, cells were collected and dissociated into single cell suspension using TrypLE select (Life Technologies). Cells (3 × 10 5 ) were centrifuged at 1000 rpm for 4 min to form a pellet and were incubated overnight (ON) with NPSR medium in U-bottom low-attachment 96-well plates (Thermo, Cat. No. 174929) at 37 °C and 5% CO 2 . After ON incubation, the cellular pellet was aggregated with freshly dissected and 30 mins hrGDNF (PeproTech) treated E11.5 UBs as described previously 36 . The aggregated pellets were centrifuged at 1000 rpm 4 min and kept in the 1.5 ml Eppendorf LoBind tubes with DMEM and 10% fetal bovine serum (FBS) medium ON. Next day, the pellets formed 3D aggregates at the bottom of the tubes and were transferred into a Trowell-type culture onto 0.1 μm or 1 μm pore polyester membrane and cultured for around 8 days in DMEM supplemented with 10% FBS medium at 37 °C and 5% CO 2 ; medium was changed every other day.

RT-PCR.
An RNeasy kit (Qiagen) was used according to the manufacturer's recommendations to extract the total RNA. cDNA synthesis (First Strand cDNA Synthesis Kit, ThermoFisher) was performed using standard protocols. qRT-PCR analyses were displayed with SYBR Green (Agilent) by an CFX96 Real-Time PCR machine.
The Brilliant III SYBR ® Green QPCR Master Mix (Agilent Technologies) was used according to the manufacturer's instructions. The GAPDH probe served as a control to normalize the data. The gene expression experiments were performed in triplicates on three independent experiments. All the Primers sequences are given in Table S1.

Data Availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on request.