Rieke JM , Zhang R , Braun D , Yilmaz Ö , Japp AS , Lopes FM , Pleschka M , Hilger AC , Schneider S , Newman WG , Beaman GM , Nordenskjöld A , Ebert AK , Promm M , Rösch WH , Stein R , Hirsch K , Schäfer FM , Schmiedeke E , Boemers TM , Lacher M , Kluth D , Gosemann JH , Anderberg M , Barker G , Holmdahl G , Läckgren G , Keene D , Cervellione RM , Giorgio E , Di Grazia M , Feitz WFJ , Marcelis CLM , Van Rooij IALM , Bökenkamp A , Beckers GMA , Keegan CE , Sharma A , Dakal TC , Wittler L , Grote P , Zwink N , Jenetzky E , Brusco A , Thiele H , Ludwig M , Schweizer U , Woolf AS , Odermatt B , Reutter H .

MR/L002744/1 Medical Research Council , MR/R006237/1 Medical Research Council

                bladder exstrophy

	FIGURE 1. slc20a1a MO KD. KD was performed by injecting 0.75 ng of ATG-blocking MO into one-cell-staged wt zf eggs. Lateral view, dorsal to top, cranial on the left. (A) Phenotypical grading of MO injected zfl at 2 dpf in four grades increasing in severity and lethality. Scale bar: 500 μm. See main Results text for grading details. In brief, GI were normal in direct inspection; GII had mild defects; GIII had moderate defects, and GIV had severe defects. G V resembles dead zfl. Supplementary Data Sheet S3 shows close-ups for better demonstration of hydrocephalus and eye development in all four grades presented. (B) Examples for cloacal abnormalities in MO KD zfl at 2 dpf. Scale bar: 100 μm. Inspection of cloacal region after previous grading revealed malformations in cloacal region in G II and G III sorted zfl increasing in amount and severity with grading. G I sorted zfl had normal cloacal morphology, that is, a thin and curved organ with a distal opening (arrow). GII and GIII zfl have abnormally shaped cloacae, with dilated and/or apparently blind-ending lumens. Further cloacal close-ups of G II and III sorted zfl are shown in Supplementary Data Sheet S4. (C) WB shows efficacy of MO KD in zebrafish protein lysates from 2 dpf. 70 kDa: slc20a1a, 42 kDa: ß-Actin loading control; WT = uninjected control, Crtl = control MO injected, G II/III/IV = slc20a1a MO KD zfl sorted by grading II–IV. Slc20a1a can be detected in uninjected wt control and injected control MO group. Only a weak slc20a1a signal was seen in MO KD groups G II, III, and IV, which showed phenotypical features as described in (A). WB shows correlation between phenotype and protein expression. Raw data of WB is shown in Supplementary Data Sheet S6. (D) Co-injection of slc20a1a MO with 35 μg in vitro transcribed human SLC20A1 polyA mRNA shows partial rescue of various phenotypes underlining the Morpholino’s specificity. n = 5 (here n = 1 represents the average score in each experimental batch). Error bars show SD. X-axis shows groups at 3 dpf that were compared: zfl showing no or only a very mild phenotype (G I + G II), larvae with a moderate phenotype (G III) and a last group of larvae with a severe and lethal phenotype (G IV) together with those who were already dead at time of comparison (G V). Y-axis shows the percentage of zfl in the corresponding groups described before. A significant difference (∗p = 0.01) was seen within the first group of MO and MO-mRNA rescue group, showing a partial rescue of slc20a1a MO KD phenotype reflected in bigger group of phenotypically not or only mildly affected zfl. Mere overexpression of SLC20A1 wt mRNA in zfl resulted in phenotypical aberrations, which did not fit the grading characteristics and the phenotypes observed in slc20a1a MO KD. Further, pure SLC20A1 overexpression in zfl resulted in higher lethality compared to non-injected control groups (data not shown). These findings suggest that the MO rescue effect of SLC20A1 wt mRNA is weakened and disguised by the mRNA’s general negative overexpression effect. (E) Kaplan-Meyer curve shows significantly reduced survival (p < 0.0001) in slc20a1a MO-injected group compared to uninjected and control MO. Survival rates by day five post-fertilization: WT = 100%, control MO injected group = 93%, MO injected group = 35% (including all grades). Exclusion of embryos dying/not developing by 8 h post-fertilization (hpf) due to failed fertilization or consequences of tissue damage caused by injections with mechanical manipulation. Within these initial 8 hpf intervals, no difference was seen between control MO and slc20a1a MO injected groups.
	FIGURE 2. Pronephric cysts in slc20a1a MO KD resulting from cloacal obstruction due to malformations. (A) Scheme of human abdomen, green: urogenital tract with kidneys, ureter, bladder, urethra; blue: abdominal wall. (B) Pronephros in zfl (green: right side) as equivalent to human urinary tract. Scheme of zfl, dorsal view at (2 dpf, patterning of zebrafish pronephros is similar to human nephron segmentation. Specific segments are color coded (left side) for better identification: G, Glomerulus; N, Neck; PCT, Proximal Convoluted Tubule; PST, Proximal Straight Tubule; DE, Distal Early; DL, Distal Late; CD, Collecting Duct. (C) Whole mount in situ hybridization (WISH) against slc20a1a in zfl at 2 dpf, labeling proximal part of pronephros. Scale bar: 500 μm. (D) Control MO zfl in Tg(wt1b:GFP) (Perner et al., 2007) marking proximal part of pronephros, Scale bar: 100 μm. (E) Glomerular close-ups in Tg(wt1b:GFP) zfl in dorsal view at 2 dpf (Scale bar: 50 μm). On the left, pronephric cysts (arrows) and dilatation of proximal part of pronephros increasing in severity with grading are shown. G I (upper pictures) showing no cystic phenotype, G II (middle pictures) showing a mild cyst formation, and G III (bottom pictures) showing severe cysts and a wide dilatation of the pronephros. Corresponding cloacal close-ups, shown on the right in lateral view (Scale bar: 50 μm), underline correlation between malformations in urinary outflow tract and cysts as well as pronephric dilatations in corresponding groups. (F) Graph shows percentage of zfl (Y-axis) in the following groups: Control MO, slc20a1a MO G I, slc20a1a MO G II, slc20a1a MO G III (X-axis). Each dot stands for one individual experiment (here N = 1 represents the average score in each experimental batch). Whereas control MO and slc20a1a MO G I larvae do not show any cystic phenotype, only an average of 21% of MO G II, and solely 3% of G III MO zfl show normal configuration of the proximal part of pronephros. The graph shows significant differences between control MO and phenotypically normal group MO G I compared to MO G II and G III, N = 3, ****p < 0.0001. Error bars show SD.)
	FIGURE 3. Sulforhodamine 101 excretion assay shows imperforate hindgut in slc20a1a MO KD zfl. (A) Scheme of human abdomen, red = GIT with stomach, duodenum, jejunum, ileum in light red and colon, rectum, anus in dark red; blue = abdominal wall. (B) Scheme of zfl, lateral view at 5 dpf; green = urogenital tract ((pronephros), red = GIT, blue = abdominal wall, beige = yolk sac, yellow = cloaca. (C) Cloaca in zfl at 5 dpf: fusion and opening of pronephros and GIT at cloaca between 4 and 5 dpf. Pseudocolored for identification as above in B: green = urogenital tract (pronephros), red = GIT, yellow = cloaca. Scale bar: 50 μm. (D) Opening of cloaca and excretion of SR101, a red fluorescent dye labeling the content of zfl intestine. Upper panel shows brightfield, middle panel red channel, lower panel shows a merged view of both channels. On the left we show slc20a1a MO KD G I zfl at 5 dpf compared to slc20a1a MO KD G II zfl at 5 dpf on the right side of the panel. In control MO and slc20a1a MO GI, zfl dye uptake is not disturbed; we could detect clear and bright red dye fluorescence in the gut of all animals. Dye excretion and opening of the cloaca was not disturbed. White arrow marks dye excretion from the cloaca. In contrary, we observed cloacal opening and excretion defects in slc20a1a MO KD G II zfl at 5 dpf mimicking an imperforate anus as shown on the right side of the panel. Black arrow marks opening defect and therefore resulting dilatation of intestine due to bag log. No changes in peristalsis of the GIT was observed; hence, expansion of distal part of intestine as shown here is solely caused by lack of cloacal opening. Scale bar: 50 μm. (E) Significant differences in opening of cloaca at 5 dpf in zfl between phenotypically affected and control MO. Cloacal opening was monitored for several minutes up to 1 h. Only 13.25% of G II and 5.5% of G III zfl showed cloacal opening and therefore excretion of SR101 from the GIT, whereas 82.5% of G II and 74% of G III zfl did not show any excretion. In the remaining 4.25% of G II and 20.5% of G III, zfl cloacal opening could not be assessed resulting from failure of SR101 uptake in the first place or misshape and tissue malformations not allowing to assess the cloacal region in the respective zfl. N = 4, ****p < 0.0001. Error bars show SD.)
	FIGURE 4. Exome sequencing and targeted Resequencing in families with BEEC phenotype identified disease variants in SLC20A1. (A) IHC of a transverse section of a 6-week-old human embryo. SLC20A1 was immunodetected (brown) in the urogenital sinus (UGS), which develops into the urinary bladder. Furthermore, (textitSLC20A1 was also detected in the spinal cord (SC), dorsal root ganglia (DRG), peripheric nerve trunk (PNT), hind limb (HL), hindgut (HG), migrating neural crest cells (MNCC), umbilical cord (UC), Wharton’s jelly (WG), midgut (MG), and umbilical artery (UA). Scale bars = 100 μm. (B) Histology section of the human 10-week-gestation metanephric metanephros. Note prominent SLC20A1 immunostaining (brown) in the proximal tubule (PT) and the collecting duct (CD). The glomerulus (G) shows a fainter signal. Scale bars: 100 μm. (C) Exome sequencing of eight CE case-parent-trios revealed de novo variant c.709G > A (p.Gly237Arg) in SLC20A1 in family 1 (Reutter et al., 2016). Resequencing of 690 individuals with BEEC led to identification of two more variants in individuals with CBE: c.893T > C (p.Val298Ala) in family 2 as de novo change c.1321A > C (p.Lys441Gln) with maternal inheritance (maternal phenotype: fusion defect of pelvic bone, mild phenotype) in family 3. Pedigrees of all three families are shown with genotypes of all individuals indicated. In family 3, the maternal grandfather (Figure 1B, family 3, person III.1) was not available for testing. For haplotype analysis of all available family members (Figure 1B, family 3, III.2, II.1, II.2, I.1), we used the synonymous marker rs4849091 at chromosomal position chr2:113404708 A > G (p.Leu101=) of the canonical transcript ENST00000272542.7. Genotypes of rs4849091 are shown in brackets. All variants are heterozygote changes and result in missense variants as shown in Sanger sequences including amino acid sequences below each pedigree.)
	FIGURE 5. SLC20A1 as transmembrane phosphate transporter and in vitro characterization of its variants found in BEEC individuals. (A) In silico 2D model of SLC20A1, a multi-pass integral membrane protein, indicating localization of variants found in affected individuals, c.709G > A (p.Gly237Arg) in TMD 7, c.893T > C (p.Val298Ala) and c.1321A > C (p.Lys441Gln) are located in an intracellular loop. Only four (6–9) of a total of twelve TMDs are shown in this simplified model. The model was generated based on the data of Beck et al. (2009). (B,C) SLC20A1 mediated 32PO4 transport in transiently transfected HEK293 cells. (B) WB analysis of 100 μg whole cell homogenates obtained from HEK293 cells transfected with 500 ng plasmid DNA per cm2 well surface. Plasmid DNA was FLAG tagged and transfection efficiency was detected using anti-FLAG antibody (70 kDa) and anti-ß-ACTIN antibody (42 kDa), which served as loading control. As expected, no FLAG signal could be detected in negative control (untransfected HEK293 cells, marked as HEK). Transfection worked for wt SLC20A1 overexpression as well as the variants c.893T > C (p.Val298Ala) and c.1321A > C (p.Lys441Gln). No FLAG signal could be detected for c.709G > A (p.Gly237Arg) transfected cells. Even when transfected with higher plasmid concentrations, p.Gly237Arg was not detectable in HEK293 cells (Supplementary Data Sheet S12). (C) Endpoint assay of transient transfected HEK293 cells. Cells were incubated with 1 μCi 32PO43– and 200 μM K3PO4 for 15 min. For better comparison, a highest number of counts per minute in wt SLC20A1 overexpression group was set as 100% in each experiment (N = 6 with two datasets each), values of HEK293 and variants were calculated correspondingly, and resulting values in percentage are shown on the y-axis (Error bars show SD). A two-way ANOVA of the grouped analysis was significant ((p < 0.0001). p.Gly237Arg did not show any differences of phosphate uptake to negative HEK293 control. This is in line with the expression deficiency of p.Gly237Arg described before. Wt SLC20A1 overexpression showed a significant increase of phosphate uptake compared to untransfected HEK293 (Tukey’s multiple comparison: ****p < 0.0001). Amin acid change p.Lys441Gln and p.Val298Ala overexpression resulted in an even higher phosphate uptake than wt SLC20A1 overexpression with a significant difference between p.Val298Ala and wt SLC20A1 (Tukey’s multiple comparison: *p < 0.05). Therefore, variant overexpression does not impair phosphate uptake capability in vitro. (D,E) Densitometric analysis of WBs (N = 6) from whole cell homogenates was obtained from transfected HEK293 cells. WBs are provided in Supplementary Data Sheet S13. Y-axis shows normalized values against wt SLC20A1 overexpression. Error bars show SD. (D) Expression of CC3 as apoptosis marker was measured in six WBs of corresponding independent transfection experiments. The one-way ANOVA was significant (p = 0.0002), Tukey’s multiple comparison test was significant for HEK293 vs. WT (***p = 0.0005), and WT vs. p.Val298Ala (**p = 0.0038). Wt SLC20A1 overexpression in HEK293 cells increased apoptosis when compared to untransfected negative control (HEK). There is no induction of apoptosis inc.893T > C (p.Val298Ala) transfected cells, comparable to untransfected negative control (HEK). c.1321A > C (p.Lys441Gln) does not result in significant reduction of CC3 expression. However, a trend of reduced CC3 expression in comparison to wt SLC20A1 overexpression can be seen. (E) Same analysis was used to study expression of PCNA as a proliferation marker. A one-way ANOVA did not show significant results (p = 0.1903). Nevertheless, wt SLC20A1 overexpression seems to reduce PCNA expression when compared to negative control (untransfected HEK). Variants analyzed [c.893T > C (p.Val298Ala) and c.1321A > C (p.Lys441Gln)] tend to reduce PCNA less than WT overexpression (red dotted line for better comparison).)
	S1 Fig. Whole Mount In Situ Hybridization (WISH) against slc20a1a and slc20a1b WISH was performed at two different time points (16 hpf: A-D; 24 hpf E-K). Expression of slc20a1a (A, B, E, G – K) is restricted to pronephric tissue, intermediate mesoderm and posterior cardinal vein. slc20a1b (C, D, F) WISH shows expression in CNS and caudal somites. There is no slc20a1b expression in relevant tissues for urinary tract development. Scale Bars: A-D: 100 μm; E-F: 200 μm; G + I: 50 μm; H + J + K: 50 μm
	S2 Fig. Whole Mount In Situ Hybridization (WISH) against slc20a1a, transversal sections WISH was performed at 24 hpf. A shows zfl in lateral view, cranial to left. B-D show transversal sections, dorsal top (*), ventral bottom (**). Dotted line in A show area of section shown in close ups B-D. Compared to S1 WISH staining reaction was stopped much earlier to demonstrate strong expression of slc20a1a around cloacal region at this time point of development. In B we see restricted staining to both proximal pronephric ducts (arrows). C and D show expression further caudal. Scale Bars: A: 250 μm; B: 150 μm; C+D: 100 μm
	S3 Fig. slc20a1a MO KD Zfl show hydrocephalus and eye abnormalities increasing in severity with grading Phenotypical grading was performed at two dpf (detailed information in Figure 1). G I (no phenotype) on the left, G II second from left, G III second from right, G IV on the right. All zfl are shown cranial to the left. Scale bar: 250 μm (A) shows hydrocephalus (arrowhead), which was partially seen in G II zfl and always present in G III and G IV zfl at two dpf. (B) focuses on the left eye (arrow) and shows examples of phenotypic abnormalities seen in slc20a1a MO KD. G II zfl show a reduction in eye size proportional to reduction in total head and body size but no structural difference. Both, G III and G IV show severe developmental defects of the eyes of different severity. Lower panel shows eye and lens circled with white dotted lines in G I zfl. Outlines were copied from G I to G II – IV to better demonstrate reduction in size and structure (loss of lens).
	S4 Fig. Additional examples for cloacal phenotypes in slc20a1a MO KD zfl at zwo dpf resembling urinary tract obstructions (A) shows three examples of slc20a1a MO KD G II zfl with high variability in shape of the urinary outflow tract. In (B) three additional phenotypes of urinary outflow tract in slc20a1a MO KD G III zfl are shown resembling urinary obstruction due to morphological abnormalities. Scale bar: 50 μm.
	S5 Fig. WISH of marker probes in slc20a1a MO KD zfl compared to control MO zfl shows disorganization of cloacal tissue Left side: WISH against pax2a as distal pronephric marker probe at two timepoints (A = 24 hpf, B = 28 hpf); Right side: WISH against evx1 as cloacal marker probe at two timepoints (E = 48 hpf, F = 60 hpf). (A, B, E, F) control MO in upper panel and slc20a1a MO injected zfl below as well as cloacal close-ups of framed region on the right side each. Differences of expression of marker probes were seen and quantified using different approaches: In pax2a WISH (D) a free line was drawn in the middle of the detected region in the proctodeum resembling the urinary outflow tract at this developmental stage. The longest orthogonal to this line was measured using ImageJ. For evx1 (H) a threshold area of expression was measured in pixels. Graphs (C, G) show mean +/- SD (Error bars) of both approaches comparing uninjected, control MO and slc20a1a MO zfl at both analyzed timepoints (pax2a in C, evx1 in G). We see significant differences comparing slc20a1a MO KD with uninjected and control MO. Control MO zfl do not show any differences to uninjected group. p < 0.05 = *, p < 0.0001 = ****, hpf = hours post fertilization, Scale bars: A-B + E-F: 100 μm, D+H: 50 μm
	S6 Fig. Western Blot_Raw data Slc20a1a ß - Actin Raw data of Western Blot confirming Slc20a1a knockdown in slc20a1a MO zfl. Protein was obtained at 2 dpf. Upper panel shows WB against Slc20a1a. While there is a strong signal at 70 kDa for Slc20a1a in uninjected (uninj.) and Control MO (Crtl MO) group there is only a very weak detection of Slc20a1a at 70 kDa in slc20a1a MO KD groups (G II = mild phenotype, G III = moderate phenotype, G IV = severe phenotype; For detailed information on grading criteria see manuscript main text in the results section). Lower panel shows loading control ß – Actin and therefore confirms equal protein loading.
	S11 Fig. Localization of variants in putative three-dimensional structure of SLC20A1 using iterative structural modeling
	S12 Fig. Expression deficiency of c.709G>A (p.Gly237Arg) in HEK293 cells is independent of transfected plasmid concentration compared to overexpression of wt SLC20A1 Endpoint assays and WB analysis was performed as described in Figure 5B. (A) Four different plasmid concentrations (250 ng, 500 ng, 750 ng and 1000 ng plasmid DNA per cm2 well surface) were monitored and phosphate uptake was measured. Counts per minute (cpm) were calculated in percentage with highest amount of cpm in wt SLC20A1 overexpression in highest concentration as 100 %. While phosphate uptake capability increased for wt SCL20A1 overexpression with increase of plasmid concentration transfected, no change was seen for c.709G>A (p.Gly237Arg) overexpression. Transfected plasmid amount correlates with phosphate transport in wt SLC20A1 overexpression. No phosphate transport was seen in c.709G>A (p.Gly237Arg) transfected HEK293. Error bars show SD. (B) 100 μg whole cell homogenates obtained from HEK293 cells transfected with 500 ng and 1000 ng plasmid DNA per cm2 well surface were probed with anti- FLAG and anti-ß-ACTIN antibody. c.709G>A (p.Gly237Arg) was not expressed in HEK293 cells independent of transfected plasmid concentration. Correlation of transfected plasmid concentration and expression of wt SLC20A1 in WB was shown. For better understanding endpoint assay from A was modified showing only the corresponding values to WB. Phosphate uptake in wt SLC20A1 overexpression was significantly higher than in c.709G>A (p.Gly237Arg) overexpression for both plasmid concentrations shown (1000 ng/cm2: p = 0.0007; 500 ng/cm2: p = 0.1). No increase in phosphate transport was seen in c.709G>A (p.Gly237Arg) transfected HEK293 independent of plasmid concentration transfected. Error bars show SD.
	S13 Fig. Western Blot Densitometric analysis of Western Blots (N = 6) from whole cell homogenates obtained from transfected HEK293 cells. This figure shows examples of WBs used for densitometric analysis (data presented in Figure 5D and 5E. A shows WB against apoptosis marker CC3 and loading control beta-actin. Since CC3 ist expressed at 17 and 19 kDa both bands were used for densitometry. B shows WB against proliferation marker PCNA and loading control beta-actin.
	S14 Fig. Bladder Exstrophy Epispadias Complex – phenotypical spectrum Pictures of BEEC spectrum, taken from Ebert et. al (Ebert et al., 2009). Left picture shows a male individual with cloacal exstrophy (CE) comparable to our index person (family 1) with fusion defect of pubic bone, split penis, small omphalocele and analatresia with protrusion of rectal part of intestine between unfused anteriorly exposed hemibladders. Middle picture shows classic bladder exstrophy (CBE) in an affected female individual (comparable to family 2) showing fusion defect of pubic bone, anterior exposure of dorsal wall of urinary bladder, split clitoris and forward displaced anal opening. CBE in an affected male individual (comparable to family 3) can be seen in right picture, including all features mentioned for CBE before including split penis instead of split clitoris.

Beck, Identification of a novel function of PiT1 critical for cell proliferation and independent of its phosphate transport activity. 2009, Pubmed

Beck, Identification of a novel function of PiT1 critical for cell proliferation and independent of its phosphate transport activity. 2009, Pubmed
Dakal, Structural modeling of human organic cation transporters. 2017, Pubmed
Diamond, Cloacal exstrophy: a 22-year experience. 1985, Pubmed
Ebert, The exstrophy-epispadias complex. 2009, Pubmed
El Husseini, High expression of the Pi-transporter SLC20A1/Pit1 in calcific aortic valve disease promotes mineralization through regulation of Akt-1. 2013, Pubmed
Farrell, New structural arrangement of the extracellular regions of the phosphate transporter SLC20A1, the receptor for gibbon ape leukemia virus. 2009, Pubmed
Festing, Generation of mouse conditional and null alleles of the type III sodium-dependent phosphate cotransporter PiT-1. 2009, Pubmed
Howe, ZFIN, the Zebrafish Model Organism Database: increased support for mutants and transgenics. 2013, Pubmed
Inouye, Modern management of the exstrophy-epispadias complex. 2014, Pubmed
Javadpour, Helix packing in polytopic membrane proteins: role of glycine in transmembrane helix association. 1999, Pubmed
Kimmel, Stages of embryonic development of the zebrafish. 1995, Pubmed , Xenbase
Kolvenbach, Rare Variants in BNC2 Are Implicated in Autosomal-Dominant Congenital Lower Urinary-Tract Obstruction. 2019, Pubmed
Li, Glycine and beta-branched residues support and modulate peptide helicity in membrane environments. 1992, Pubmed
Lund, Cloacal exstrophy: experience with 20 cases. 1993, Pubmed
Maruf, Variant Presentations of the Exstrophy-Epispadias Complex: A 40-Year Experience. 2019, Pubmed
Nichane, The Na+/PO4 cotransporter SLC20A1 gene labels distinct restricted subdomains of the developing pronephros in Xenopus and zebrafish embryos. 2006, Pubmed , Xenbase
O'Hara, Characterization of a human gene conferring sensitivity to infection by gibbon ape leukemia virus. 1990, Pubmed
Perner, The Wilms tumor genes wt1a and wt1b control different steps during formation of the zebrafish pronephros. 2007, Pubmed
Raciti, Organization of the pronephric kidney revealed by large-scale gene expression mapping. 2008, Pubmed , Xenbase
Reutter, Genetics of Bladder-Exstrophy-Epispadias Complex (BEEC): Systematic Elucidation of Mendelian and Multifactorial Phenotypes. 2016, Pubmed
Roy, I-TASSER: a unified platform for automated protein structure and function prediction. 2010, Pubmed
Salaün, Identification of a novel transport-independent function of PiT1/SLC20A1 in the regulation of TNF-induced apoptosis. 2010, Pubmed
Stec, Congenital renal anomalies in patients with classic bladder exstrophy. 2012, Pubmed
Thisse, High-resolution in situ hybridization to whole-mount zebrafish embryos. 2008, Pubmed
Wallingford, Loss of PiT-1 results in abnormal endocytosis in the yolk sac visceral endoderm. 2014, Pubmed
Zhang, ISL1 is a major susceptibility gene for classic bladder exstrophy and a regulator of urinary tract development. 2017, Pubmed
van Rooij, AGORA, a data- and biobank for birth defects and childhood cancer. 2016, Pubmed