The contribution of Notch1 to nephron segmentation in the developing kidney is revealed in a sensitized Notch2 background and can be augmented by reducing Mint dosage

Abstract

We previously determined that Notch2, and not Notch1, was required for forming proximal nephron segments. The dominance of Notch2 may be conserved in humans, since Notch2 mutations occur in Alagille syndrome (ALGS) 2 patients, which includes renal complications. To test whether mutations in Notch1 could increase the severity of renal complications in ALGS, we inactivated conditional Notch1 and Notch2 alleles in mice using a Six2-GFP::Cre. This BAC transgene is expressed mosaically in renal epithelial progenitors but uniformly in cells exiting the progenitor pool to undergo mesenchymal-to-epithelial transition. Although delaying Notch2 inactivation had a marginal effect on nephron numbers, it created a sensitized background in which the inactivation of Notch1 severely compromised nephron formation, function, and survival. These and additional observations indicate that Notch1 in concert with Notch2 contributes to the morphogenesis of renal vesicles into S-shaped bodies in a RBP-J-dependent manner. A significant implication is that elevating Notch1 activity could improve renal functions in ALGS2 patients. As proof of principle, we determined that conditional inactivation of Mint, an inhibitor of Notch-RBP-J interaction, resulted in a moderate rescue of Notch2 null kidneys, implying that temporal blockage of Notch signaling inhibitors downstream of receptor activation may have therapeutic benefits for ALGS patients.

Figure 1. The Six2-GFP∷Cre transgene is expressed mosaically in renal epithelial progenitor population, with uniform high level of expression in the induced mesenchyme and pre-tubular aggregates prior to RV formation

(A-A") Six2-GFP∷Cre transgene (green) is expressed at variable levels within renal epithelial progenitors. Not all Six2⁺ cells (red) express GFP∷Cre at E13.5 (arrowheads). GFP∷Cre is expressed at higher levels in cells exiting or no longer in the definitive renal epithelial progenitor pool (arrows) and in pretubular aggregates (*) and continues to be expressed in RV (**). A' is a higher magnification of the boxed area in A. (B and B') Co-staining for Pax2 (red) and the Six2-GFP∷Cre transgene (green) reveals that not all mesenchymal cap condensates that are Pax2⁺ express Cre recombinase (blue arrowheads) at P1. The white arrowheads are placed as positional references and point at cells expressing both Pax2 and GFP. (C) Importantly, due to strong expression of Cre in pre-tubular aggregates, all cells achieved recombination of floxed alleles by the S-shaped stage as demonstrated by β-galactosidase expression in every S-shaped body cell of E13.5 Six2-GFP∷Cre^tg;Rosa26R kidneys. All cells of the S-shaped bodies (arrowheads) have experienced cre mediated recombination and have turned on β-galactosidase expression in the E13.5. All scale bars are 100um.

Figure 2. Delayed inactivation of Notch2 in the renal epithelial progenitors results in sufficient nephron segmentation and proximal tubule formation for survival

Compared with N2^f/f littermate kidneys (A), approximately 30% of post-natal day one Six2-GFP∷ Cre^tg; N2^f/f kidneys (B) are visibly smaller. (C) Graph illustrating the average (height of the bar) number of glomeruli in N2^f/f, (n=3) is 4950±488 versus 1902±1038 in the visibly smaller Six2-GFP∷Cre^tg; N2^f/f, (n=3, p=0.001) P1 kidneys. (D) All the kidneys of newborn Six2-GFP∷Cre^tg; N2^f/f mice have many mature nephrons containing podocytes within glomeruli (red) and LTL⁺ proximal tubules (green). (E) P1 Six2-GFP∷Cre^tg; N2^f/f kidneys contain NCAM⁺ (green) and Pax2⁺ (red) S-shaped bodies confirming that nephron segmentation occurred. (F) The kidneys of newborn Six2-GFP∷Cre^tg; N2^f/f mice have no detectable Notch2 protein in the nascent renal epithelial structures, whereas in (G) N2^f/f kidneys, Notch2 protein (red) co-localizes with NCAM (green) in the cell membrane of a subset of cap condensate cells and all nascent renal epithelial structures. (H) By FACS we isolated the GFP+ population (boxed by dashed line) from E14.5 Six2-GFP∷ Cre^tg; N2^f/f kidneys. (I) PCR analysis of genomic DNA extracted from GFP⁺ FACS sorted cells reveals that the Notch2 floxed alleles are inactivated in the GFP⁺ cells. We determined the lower limit of detection of the Notch2 floxed allele to be one copy per 100 cells. Lanes 1 and 13 contain 100bp DNA ladder (Invitrogen), lanes 2 to 12 contain PCR product, in which the starting template for lanes 2 to 6 contain 10-fold serial dilutions of 30ng of genomic DNA extracted from tail of N2^+/f mouse diluted in wild type genomic DNA. The starting template for lane 7 is wild type genomic DNA, for lane 8 is water, for lane 9 is 30ng of genomic N2^f/f DNA and for lanes 10, 11 and 12 the starting template was 30ng of genomic DNA extracted from GFP⁺ kidney cells of three different Six2-GFP∷ Cre^tg; N2^f/f mice. The arrow at the left hand side of gel points at the expected size of the Notch2 floxed allele. We are able to detect a strong PCR product even when only one Notch2 floxed allele is present per 100 genomes (lane 4) but not in the GFP⁺ cells (lanes 10,11 and 12). Scale bars in A and B are 1mm, scale bar in D is 100um, and scale bars in E,F and G are 20um. UB: ureteric bud, S: S-shaped body, V: blood vessel.

Figure 3. Inactivation of Notch1 in the Notch2-deficient kidneys reveals that Notch1 is required for nephrogenesis, normal renal function and survival

WT1 and LTL staining of post-natal day one (A) wild type (N1^+/f; N2^+/f), (B) Six2-GFP∷Cre^tg;N1^f/f and (C) Six2-GFP∷Cre^tg;N1^+/f;N2^f/f kidneys reveals reduced number of proximal tubules with the inactivation of one Notch1 allele in the Notch2-deficient background. (D) The graph depicts the percentage of mice of a genotype surviving as they age up to six months. Whereas the Six2-GFP∷Cre^tg; N2^f/f mice (n=10) have a normal lifespan (red dashes), the Six2-GFP∷Cre^tg;N1^+/f; N2^f/f mice (n=16) have a variably reduced lifespan (blue line) and Six2-GFP∷Cre^tg;N1^f/f;N2^f/f mice (n=4) die soon after birth (green dashes). (E) Bisected kidneys from one Six2-GFP∷Cre^tg;N1^+/f; N2^f/f mouse that died at six months revealed one hydronephrotic obstructed kidney (top panel) and a second hypoplastic kidney (bottom panel). (F) The blood urea nitrogen (BUN) levels are increased in Six2-Cre^tg; N1^+/f;N2^f/f mice within the first week of birth when compared with wild-type littermates (49.8±35.2 mg/dL, n=14, compared with 25± 3.1 mg/dL in controls, n=19, *p=0.0045). The average BUN levels are represented by the height of each bar and one standard deviation by the error bars. (G) In post-natal day one kidneys of N2^f/f mice NCAM and Pax2 staining revealed comma and S-shaped structures in the nephrogenic zone. (H) Abnormal nephron segmentation was evident by the presence of abnormal NCAM+, Pax2+ structures (arrowhead) in the nephrogenic zone of Six2-GFP∷Cre^tg;N1^+/f;N2^f/f kidneys. All scale bars are 100um.

Figure 4. Inactivation of RBP-J resembles the combined inactivation of Notch1 and Notch2 with a severe reduction in S-shaped body formation and development of proximal tubules and glomeruli

Compared with (A) wild type, RBP-J^f/f mice (D) the Six2-GFP∷ Cre^tg ; RBP-J^f/f mice die by post-natal day two and though their kidneys are small, managed to produce a clear filtrate present in the bladder (black arrow). (B) In RBP-J^f/f mice, the RBP-J protein (red) is present in the nucleus of all kidney cells at P1 including S-shaped bodies (arrows) which results in (C) mature LTL+ proximal tubules containing RBP-J. (E) In Six2-GFP∷ Cre^tg; RBP-J^f/f P1 kidneys only a few mesenchymal cap condensate cells still retain RBP-J expression (white arrowhead) and (F) the absence of RBP-J protein in the proximal tubules. (G and H) The Six2-GFP∷ Cre^tg; RBP-J^f/f kidneys develop very few glomeruli (yellow arrowhead) and proximal tubules (blue arrowheads or green signal). SGC is an abbreviation for Six2-GFP∷ Cre^tg. Scale bars are 100um.

Figure 5. The specification of proximal nephron cell fates is acutely sensitive to the dosage of canonical Notch1 and Notch2 signaling activity, which allows the removal of repressors of RBP-J-dependent transcription to partially rescue nephrogenesis

(A) The frequency of animals with normal kidney size, y-axis, inversely correlates with the increasing loss of canonical Notch signaling activity, x-axis. The height of each blue bar represents the frequency of animals of a particular genotype having normal kidney size at P1. The number of mice analyzed per genotype at P1 is indicated next to the genotype. All floxed alleles (f) presented in the graph were inactivated using Six2-GFP∷ Cre^tg. Wt = wild type littermates. (B) A model of the individual contributions of Notch1 and Notch2 to nephron segmentation based on current and previous observations. Notch2 signaling (orange bar) is critical for reaching the threshold of Notch signaling activity (dashed line) required for sufficient PT and podocyte (POD) development to ensure survival. Inactivation of Notch2 in the Six2-GFP∷Cre^tg; N2^f/f mice creates a sensitized Notch signaling background in which a further reduction in Notch signaling by inactivating Notch1 to create Six2-GFP∷Cre^tg;N1^+/f; N2^f/f mice, reveals that both Notch1 and Notch2 are required for PT and POD, in the absence of which only distal tubule (DT) form. (C, D) Progressive loss of Mint on the Pax3-Cre^tg; N2^f/f background results in the reemergence of proximal cell fates. (C) Kidneys from Pax3-Cre^tg; N2^f/f, Pax3-Cre^tg; Mint^+/f; N2^f/f, and Pax3-Cre^tg;Mint^f/f; N2^f/f were scored on a rank-order scale (see figure S3) for degree of rescue based on the presence of LTL+ PT structures. Decreasing doses of Mint consistently corresponded to a greater degree of PT reemergence. * - P= 0.001. ** - P= 0.0008. (D) Representative (average rank-score) images of sagittal, center cut P0 kidney sections from each of the above genotypes stained with WT1 (red; MM and podocytes) and LTL (green; PT). In some cases we observed WT1+ and LTL+ cells polarized in the same structure in Pax3-Cre^tg; Mint^f/f; N2^f/f mice (inset), which we have never observed in Pax3-Cre^tg; N2^f/f animals.