key: cord-0303854-uvln1ieo authors: Vanslambrouck, Jessica M.; Wilson, Sean B.; Tan, Ker Sin; Groenewegen, Ella; Rudraraju, Rajeev; Neil, Jessica; Scurr, Michelle; Howden, Sara E.; Subbarao, Kanta; Little, Melissa H. title: Enhanced metanephric specification to functional proximal tubule enables toxicity screening and infectious disease modelling in kidney organoids date: 2021-10-15 journal: bioRxiv DOI: 10.1101/2021.10.14.464320 sha: 85232d2ac2574959e1358bca9a35e00545a0716c doc_id: 303854 cord_uid: uvln1ieo While pluripotent stem cell-derived kidney organoids represent a promising approach for the study of renal disease, renal physiology and drug screening, the proximal nephron remains immature with limited evidence for key functional solute channels. This may reflect early mispatterning of the nephrogenic mesenchyme or insufficient maturation. In this study, prolonged differentiation and modification of media conditions to enhance metanephric nephron progenitor specification resulted in the induction of nephrons containing elongated and aligned proximal nephron segments together with SLC12A1+ loops of Henle. Nephron proximal segments showed superior HNF4A gene and protein expression, as well as upregulation of key functional transporters, including SLC3A1/2, SLC47A1, and SLC22A2. The striking proximo-distal orientation of nephrons was shown to result from localised WNT antagonism originating from the centre of the organoid. Functionality of such transporters was evidenced by albumin and organic cation uptake, as well as appropriate KIM-1 upregulation in response to the nephrotoxicant, cisplatin. PT-enhanced organoids also possessed improved expression of receptors associated with SARS-CoV2 entry, rendering these organoids susceptible to infection and able to support viral replication without co-location of ACE2 and TMPRSS2. These PT-enhanced organoids provide an accurate model with which to study human proximal tubule maturation, inherited and acquired proximal tubular disease, and drug and viral responses. Prolonged monolayer culture and delayed nephron induction supports nephron 154 progenitors 155 As noted previously, optimisation of nephron progenitor maintenance in vitro has been 156 investigated by a range of studies using murine and human pluripotent stem cell-derived 157 nephron progenitors (Brown, et al., 2015; Li, et al., 2016; Tanigawa, et al., 2016) . While all 158 studies reported maintenance of nephron progenitors, variations were evident with respect to 159 the final patterning of resulting nephrons following induction. Given the clear influence that 160 initial differentiation conditions and timing can have on nephron progenitor survival and 161 subsequent nephron patterning, we hypothesised that expanding our nephron progenitor 162 population whilst delaying nephron initiation may create a more metanephric population 163 leading to organoids with improved patterning and PT maturation. We have previously shown 164 that SIX2 expression is not detected until day 10 of pluripotent stem cell differentiation 165 (Howden, et al., 2019) . Hence, the initial monolayer differentiation phase was prolonged to 166 between 12 -14 days, along with culture in either of two previously defined NP maintenance 167 media, NPSR (Li, et al., 2016) and CDBLY (Tanigawa, et al., 2016) The prevention of spontaneous differentiation while preserving the nephrogenic capacity of the 176 NP cells was found to be primarily a response to the presence of CDB (CHIR, DAPT, BMP7), 177 with omission of LIF, Y27632, as well as the basal media component TGFα, found to produce 178 a similar result with respect to growth, morphology and nephron segmentation compared to 179 CDBLY ( Figure 1C ). The inhibition of monolayer epithelialisation with preserved nephrogenic 180 capacity was found to be consistent at monolayer differentiation lengths tested (10, 12, 13 and 181 14 days) (Supplementary Fig 1A) . However, a monolayer differentiation length of 12 -13 days 182 produced more consistent nephrogenesis between experiments, with 14 days leading to 183 frequent detachment of the differentiating monolayer. Subsequent studies proceeded using 184 prolonged culture in CDBLY noting the inclusion of an increased concentration of BMP7 185 (10ng/mL; CDBLY2) which improved reproducibility of organoid nephrogenesis between 186 organoids compared to standard CDBLY (5ng/mL BMP7) (Supplementary Figure 1B) . This 187 modified differentiation protocol is detailed in Figure 1A . 188 Quantitative RT-PCR (qRT-PCR) of the extended monolayer differentiations in CDBLY2 189 confirmed an improved metanephric gene expression profile compared to standard 190 differentiations performed in parallel (7 day protocol in E6 (Takasato, et al., 2016; Howden, et 191 al., 2019)) ( Figure 1D ). Extended CDBLY2 monolayers showed a significant increase in 192 SIX1/SIX2 (self-renewing to committed NPs) and WNT4 (primed to committed NPs), while 193 DAPL1 (self-renewing and primed NPs) was increased without significance and no change was 194 observed in TMEM100 (self-renewing NPs). This suggested that the extended protocol 195 promotes a primed, rather than self-renewing, NPC population (Hochane, et al., 2019; 196 Lindstrom, et al., 2018; Lindstrom, et al., 2018) . Extended differentiation in CDBLY2 was not 197 found to alter mediolateral patterning, with no change in paraxial mesodermal marker 198 PARAXIS and unchanged or increased expression of intermediate mesoderm markers 199 HOXD11, LHX1, and GATA3 (Mugford, et al., 2008) . Extended monolayer culture induces SIX2-derived proximalised nephrons 201 Lineage tracing studies in mouse have shown that nephrons are derived entirely from Six2+ 202 nephron progenitors (Kobayashi, et al., 2008) , with histological studies suggesting a similar 203 developmental process in human (Lindstrom, et al., 2018; Lindstrom, et al., 2018 ) ( 204 (Kobayashi, et al., 2008 . Using a SIX2 Cre/Cre :GAPDH dual lineage tracing line, in which SIX2 205 expression induces a permanent GFP/mCherry switch, we have previously shown that kidney 206 organoid nephrons contain cells derived from SIX2 + , at also SIX2 -, progenitor cells, resulting 207 in a chimeric appearance (Howden, et al., 2019) . To confirm and compare the competence of 208 the metanephric progenitor-enriched monolayer differentiation to contribute to nephron 209 formation, organoids were generated from our extended and the standard differentiation to EPCAM + nephrons was significantly higher in organoids derived from the metanephric 215 progenitor-enriched monolayers compared to standard organoids, suggesting improved 216 metanephric identity of prolonged monolayers exposed to CDBLY2 ( Figure 2B ). 217 The patterning of these increasingly SIX2-progenitor nephrons was examined using a range of 218 markers for podocytes, proximal, and distal tubules, indicating clear proximo-distal 219 segmentation and a large proportion of proximal tubule ( Figure 2C) , with little to no GATA3 220 expression marking ureteric epithelialisation ( Figure 2D ). Organoids also displayed aligned derived from extended monolayer culture with CDBLY2 was compared to those derived from 226 the standard differentiation protocol (7 days differentiation, cultured in E6 (Howden, et al., 227 2019)). Organoids were generated using the HNF4A YFP iPSC reporter line which reports the 228 formation of proximal tubule (Vanslambrouck, et al., 2019) . This revealed up to 6.2 times 229 higher average proportions of HNF4A YFP+ proximal tubule cells in organoids derived from the 230 extended monolayer protocol compared to the standard protocol ( Figure 2D ). These results To gain deeper insight into the complexity and maturity of cells within this extended protocol, 237 both as the stage of monolayer (day 13) and within the resulting PT-enhanced organoids, 238 transcriptional profiling was performed using multiplexed single cell RNA sequencing 239 (scRNAseq) and antibody-based cell barcoding. To account for variation, libraries were 240 generated from 4 separate differentiated monolayers representing distinct starting pools of 241 iPSCs (CRL1502.C32) that were used to generate 4 separate batches of organoids ( Figure 3A ). Cells from the 4 replicates (both at day 13 [D13] monolayer stage, prior to organoid formation, 243 and day 14 of organoid culture [D13+14]) were barcoded using hashing antibodies before being 244 pooled. This approach produced a single library for each timepoint (sample) which could be 245 later deconvoluted to retrieve replicate information. The resulting D13 and D13+14 pooled replicate libraries resolved 19,956 and 15,852 individual 247 cell transcriptomes per timepoint, respectively. UMAP plots showed the resolution of distinct 248 clusters for both D13 monolayers and resulting PT-enhanced (D13+14) organoids ( Figure 3B ). (Subramanian, et al., 275 2019; Wu, et al., 2018; Low, et al., 2019; Tran, et al., 2019) that better emulated the mixed 276 reference dataset of week 11, 13, 16, and 18 human fetal kidneys (Hochane, et al., 2019; Tran, 277 et al., 2019; Holloway, et al., 2020) . PT-enhanced organoids derived from these D13 278 monolayer differentiations possessed high and abundant expression of a range of proximal 279 nephron markers in their EPT population ( Figure 3E ). These included genes encoding several 280 membrane proteins critical for proximal tubular transport of proteins and amino acids (CUBN, 281 LRP2, SLC3A1, and SLC3A2), as well as auxiliary proteins and transcription factors required 282 for transporter regulation and functionality, such as AMN, AGT, and HNF4A. This gene 283 signature showed remarkable congruence to reference human fetal kidney and improved PT 284 identity compared to existing published kidney organoid datasets (Czerniecki, et al., 2018; 285 Harder, et al., 2019; Kumar, et al., 2019) (Figure 3E ). An important anatomical feature of the mature PT is its segmentation into functionally and 287 morphologically distinct regions defined as the S1/S2 convoluted tubule segments and the S3 288 straight segment. In addition to differences in proliferation characteristics and protein synthesis 289 (Zhuo and Li, 2013; Avissar, et al., 1994) , the convoluted and straight segments display distinct 290 differences in solute handling to accommodate the declining concentration of solutes as the 291 ultrafiltrate passes through the nephron. As such, early S1 -S2 convoluted segments express 292 low-affinity/high-capacity transporters, with a gradual transition to high-affinity/low-capacity 293 transporters in the later S3 straight segment (Palacin, et al., 2001; Schuh, et al., 2018; Verrey, 294 et al., 2005) . To determine whether the PTs of enhanced organoids show evidence of this 295 segmentation, PT clusters from the 4 integrated D13+14 replicate datasets were isolated and 296 re-clustered, resolving 4740 PT cells across 6 distinct clusters (Supplementary Figure 3B ). The PT population was analysed for the expression of segment-specific PT markers with critical 298 functional roles, including solute carriers for ions (SLC34A1/NPT2 et al., 299 2015) expressed in S1>S2), glucose (SLC2A2/GLUT2 and SLC5A2/SGLT2 expressed in 300 S1>S2; SLC2A1/GLUT1 and SLC5A1/SGLT1 expressed in S2 S3 (Nagamori, et al., 303 2016)), and cationic drugs/toxins (SLC47A1/MATE1 expressed in S1/S2 > S3 (Otsuka, et al., 304 2005)), as well as AKAP12 (involved in cell cycle regulation, expressed in S2S3 306 (Avissar, et al., 1994) ). UMAP plots revealed the largely opposing distributions of cells 307 expressing S1>S2 and S2>S3 gene signatures (Supplementary Figure 3C ). Cells expressing 308 S1>S2 convoluted PT markers (SLC34A1/MATE1, SLC2A2/GLUT2, and SLC5A2/SGLT2) were predominantly located in clusters 0, 3, and the lower portion of cluster 4, whereas cells 310 expressing S2S3), UMAP plots 313 for each gene revealed a similar graded expression pattern, with a higher concentration of 314 positive cells within the S1>S2 cluster (0) and decreasing in prevalence within S2