Mdm4 controls ureteric bud branching via regulation of p53 activity

Ub^Mdm4−/− mice have renal hypoplasia despite the presence of functional Mdm2

RNA analysis by quantitative RT-PCR revealed that Mdm4 gene expression is high in embryonic kidneys and declines postnatally (Fig. 1A). Attempts to examine the spatial expression of Mdm4 protein immunohistochemically were not successful due to lack of suitable antibodies. To examine the role of Mdm4 in Ub branching, we crossed mice with conditional alleles of Mdm4 (Mdm4^f//f) with Hoxb7Cre-eGFP driver mice (Fig. 1B). The progeny developed to term and were recovered at expected Mendelian ratios. Upon gross and histological examination of newborn kidneys, we found that the Ub^Mdm4−/− kidneys were smaller in size (Fig. 1 C–F) and exhibited hypoplasia of the nephrogenic zone (defined as the outermost zone of the kidney cortex containing nascent Ub tips and nephrons but excluding maturing glomeruli or proximal tubules) (Fig. 1 G,H) and the renal medulla (Fig. 1 I,J).

Mdm4 function is required for normal ureteric bud branching morphogenesis

To better characterize the hypoplastic phenotype, we followed the branching potential of Ub^Mdm4−/− kidneys ex vivo. At E12.5, the Ub had branched twice in both Ub^Mdm4+/+ and Ub^Mdm4−/− kidneys (Fig. 2A). We cultured E12.5 Ub^Mdm4+/+ and Ub^Mdm4−/− kidney explants for 72 hours. At the end of culture, the kidneys were fixed and stained for pan cytokeratin (Ub-specific marker) and WT1 (cap mesenchyme and glomerular podocyte marker). The Ub^Mdm4−/− kidneys had an obvious reduction in Ub branching by 24 hrs (Fig. 2A) and cytokeratin/WT1⁺ structures at 72 hrs compared to Ub^Mdm4+/+ control kidneys (Fig. 2 B, C). Quantitation of Ub tips in E12.5 kidney explants at the end of 72 hr culture revealed a 57% (2.3-fold) reduction in Ub^Mdm4−/− kidneys (number of Ub tips: 24 ± 4.6, n=3) compared to the wild type controls (56 ± 2.2, n=3) (Fig. 2D). Examination of GFP fluorescence from E13.5 to P0 revealed attenuation of branching morphogenesis in Ub^Mdm4−/− as compared to Ub^Mdm4+/− kidneys as early as E13.5 in vivo (Fig. 2 E). Thus, the Ub of mutant kidneys was able to undergo branching past the T-stage but showed deficits in branching as subsequent development progressed.

Cell proliferation and survival are important determinants of branching morphogenesis (Dziarmaga et al., 2003; Maretto et al., 2003; Michael and Davies, 2004). Therefore, we examined the cellular basis for the impaired branching morphogenesis seen in the Ub^Mdm4−/− mutants. Examination of apoptosis levels using cleaved Caspase 3 immunostaining revealed significant cell death in the Ub lineage (cytokeratin-positive) of the E13.5 mutant kidneys (7.8-fold elevation, p=0.0009) relative to wild type littermate control kidneys (Fig. 3 A–E). Some cells positive for active caspase 3 showed a luminal distribution similar to Ub^Mdm2−/− mutant kidneys (Hilliard et al., 2011) (Fig. 3D). Co-immunostaining for phosphorylated histone H3-positive G2/M phase cells and pan cytokeratin at E13.5 did not show a significant change in UB cell proliferation (Fig. 3 F–J). Thus, the elevation in apoptosis contributes to impaired Ub branching morphogenesis seen in the mutants.

Mdm4 loss in the ureteric lineage causes secondary nephrogenesis defects

Since nephron induction depends on reciprocal interactions between the metanephric mesenchyme and Ub compartments, we assessed the effects on Ub-specific deletion of Mdm4 on nephrogenesis. Deficits in Ub branching yielded fewer nascent nephrons (Lef1/Jagged1-positive) in Ub^Mdm4−/− mutants at E14.5-E15.5 (Fig. 4 A–F). Nascent nephron counts were 40% lower in Ub^Mdm4−/− than Ub^Mdm4+/+ kidneys (Fig. 4 E). At this time, when Ub branching is superseded by elongation processes essential for establishing the renal collecting duct system we begin to see cyst-like dilation of the ureteric tips (Fig. 4 F, I, J). The cystic tubules were derived from the ureteric bud lineage (Cytokeratin/E-cadherin positive) (Fig. 4 J, M).

Genome-wide profiling of Ub^Mdm4−/− kidneys

In order to determine the differences in gene expression between Ub^Mdm4+/+ and Ub^Mdm4−/− kidneys, we performed a global analysis of gene expression on E14.5 whole kidneys using mouse 4X 44K array from Agilent Technologies. We chose E14.5 for our analyses because of the timing of the onset of the phenotype. A two-color dye-swap strategy was used to compare four wild type and four mutant pools (100ng of total RNA per sample) to minimize bias from signal detection with the Cy3 and Cy5 color channels. Of the 2547 significantly altered genes in our array (p ≤ 0.05), 1627 were upregulated (range +1.2 to +18.3) while 920 were downregulated (range −1.2 to −4.3) (Fig. 5A). Ingenuity Pathway Analysis-aided analysis of our microarray dataset revealed that the biological functions predicted to be impacted could be resolved into the following categories: (1) disease and disorders 48%; (2) organismal injury and abnormality 23%; (3) metabolism and molecular functions 8%; (4) cellular functions 5% and (5) inflammation 5% (Fig. 5 B). Among the highly up-regulated genes are those encoding the translation initiation factor Eif2s3y, RNA helicase Ddx3y, the cell cycle inhibitor p21 (Cdkn1a), the histone H3K4 demethylase Kdm5d, and Growth and differentiation factor 15 (GDF15) (Fig. 5 C). Interestingly, all of these upregulated genes are known p53-target genes. On the other hand, among the highly downregulated genes are immune and RNA binding proteins, the relevance of which to kidney development is unknown (Fig. 5 C).

Loss of Mdm4 impacts Ub tip and stalk networks.

Ub-specific deletion of Mdm4 activates p53-dependent pathways.

Given that Mdm4 can regulate p53 transactivation function, we looked for transcript level changes in p53 target genes in the whole-mouse genome following Mdm4 deletion from the Ub. Previously, the widely recognized role of p53 as a tumor suppressor largely overshadowed its contribution to development, differentiation, metabolism, autophagy, aging etc. (Kastenhuber and Lowe, 2017). The nuanced nature of developmental phenotypes encountered in p53^−/− mice became evident later on against different mouse strain backgrounds and includes reduced survival, exencephaly, craniofacial abnormalities, congenital anomalies of the kidney and urinary tract, lung defects, and/or reduced fertility of female mice (Armstrong et al., 1995; Donehower et al., 1992; Hu et al., 2007; Kaufman et al., 1997; Rinon et al., 2011; Saifudeen et al., 2009; Tateossian et al., 2015). The corollary, where p53 levels and activity are above what is physiologically normal, similarly poses developmental challenges in multiple tissues and organs including the kidney (Hilliard et al., 2011; Hilliard et al., 2014; Moyer et al., 2017). We found that the abrogation of Mdm4 function from the ureteric epithelia increases p53 transactivation. The most significant changes observed were in p53 target genes modulating the cell cycle (Cdkn1a, Ccng1), cellular apoptosis (Fas, Apaf1, Perp, p53INP1), EMT and invasion (snai2 or SLUG, Pai1 or serpine 2), or inflammation (Thbs1, Adgrb1) (Table 2). Absence of Mdm4 in the Ub affects cell survival and cell fate and ultimately branching morphogenesis.

In order to delineate the p53-mediated gene changes specific to the kidney, we compared our dataset to that derived from our previously published p53-Chip sequencing of E15.5 kidneys (Li et al., 2013). We found 464 gene targets shared by the two datasets (Table 3). Of them, the genes pertinent to Ub morphogenesis and differentiation and bound by p53 are Wnt11, Axin2, Ctnnb1, Gja1, Tcf4, Lgr4 Etv4, Ptch1, Calb1, Crlf1 and Pde5a.

Elimination of p53 function rescues Ub^Mdm4−/− kidney phenotype

Given the strong involvement of p53 regulated genes in the Ub^Mdm4−/− kidney phenotype, we bred Ub^Mdm4−/− mice into a p53 null background. We found that the Ub^Mdm4−/−; Trp53^+/+ mice have small kidneys (Fig. 8 A, B, D, E). However, the size and histology of Ub^Mdm4−/−; Trp53^−/− kidneys resembles those of control kidneys (Fig. 8 A, C, D, F). The Ub^Mdm4+/−; p53^+/+ kidneys show a 15% reduction while the Ub^Mdm4−/− p53^+/+ kidneys show a 41% reduction in surface area when compared to control kidneys (Fig. 8 G). The loss of a single allele of Trp53 (Ub^Mdm4−/− p53^+/−) brings about a partial rescue (27% reduction in kidney area) while loss of both Trp53 alleles narrows it further to 8% reduction in kidney area relative to control kidneys (Fig. 8 G). Comparison of neonatal Ub^Mdm4+/+; p53^+/+, Ub^Mdm4−/−; p53^+/+; and Ub ^Mdm4−/−; p53^−/− kidneys by co-immunostaining for Six2 and pan cytokeratin confirms restoration of gross morphology and histological patterning of the medulla and nephron progenitors (Fig. 8 H–J).

To confirm if Wnt-β-catenin signaling downstream of p53 regulation is restored in the kidneys of rescued mice, we performed in situ hybridization for Axin2 a target of canonical Wnt signaling. In the absence of functional p53, the levels of Axin2 are restored in the Ub^Mdm4−/−; p53^−/− medulla compared to Ub^Mdm4−/−; p53^+/+ kidneys (Fig. 9 A–F). Similarly, in newborn mice the expression of the Wnt-β-catenin nuclear effector Lef1 shows severe downregulation in the Ub and associated nascent nephrons in Ub^Mdm4−/− kidneys but is restored to wild type levels in the p53 null rescued kidneys (Fig. 9 G–I). This is illustrated schematically in Fig. 9 J–L.

The extent of rescue on a p53 null background was evident in histological analysis of P20 kidneys from Ub^Mdm4+/+; p53^+/+, Ub^Mdm4−/−; p53^+/+; and Ub ^Mdm4−/−; p53^−/− mice. The Ub^Mdm4−/− kidneys are severely dysmorphic in gross appearance and have prominent deficits in medullary collecting ducts formation resulting in severe hydronephrosis (Suppl. Fig. S-2). The histology of Ub^Mdm4−/− kidneys in a p53 null background shows restoration of kidney architecture including the outer medulla (Suppl. Fig. S-2).

Animals

Animal protocols utilized in this study were in adherence to the guidelines outlined by the Institutional Animal Care and Use Committee in accordance with NIH. The Mdm4 ^floxed mice (Grier et al., 2006) originally from Dr. G. Lozano, were a gift from Dr. Hua Lu. The Mdm2 floxed mice (01XH9, Dr. Mary Ellen Perry) were initially obtained from the NCI repository (Frederick MD) while the Hoxb7-Egfp-Cre mice (Zhao et al., 2004) were a kind gift from Dr. Carlton Bates. The p53^+/− mice (Jacks et al., 1994) obtained from The Jackson Laboratory (Bar Harbor, Maine, USA) were used in our rescue experiments. The specifications for genotyping were outlined by the respective vendors for Mdm2-floxed mice and p53 conventional mutant mice. To distinguish between Mdm4 wild-type and conditional alleles the following primer set was used (Grier et al., 2006):

• F3, 5’-GGTGTCCTTGAACTTGCTGTGTAGAA-3’,

• E2re, 5’-CTGGGCCGAGGTGGAATGTGATGT-3’

• The Cre transgene was genotyped using the following primer pair:

• Forward primer: 5’ -ACCAGCCAGCTATCAACTC-3’

• Reverse primer: 5’ -TATACGCGTGCTAGCGAAGATCTCCATCTTCCAGCAG-3’

The breeding strategy used involved crossing Hoxb7-Egfp-Cre ^tg/−; Mdm4 ^flox/+ male mice to Mdm4 ^flox/flox (Mdm4^fl/fl) female mice. The p53 rescue experiments involved crossing Hoxb7-Egfp-Cre^tg/−; Mdm4 flox/+; p53^+/− males to Mdm4 flox/flox; p53^+/− females. For all timed pregnancies, noon of the day on which the vaginal plug was detected was regarded as embryonic day (E) 0.5.

Gross Morphology

All bright field images were captured using the Nikon SMZ100 stereo microscope mounted with a DS-Fi1 camera. The images were processed using NIS Elements software (Nikon, NY, USA).

Histology

Freshly isolated kidneys were rinsed in ice-cold PBS and fixed overnight in 10% buffered formalin (for histology or IF) or 4% PFA/PBS (for section ISH) at 4C. Following serial alcohol dehydration and clearing in Xylene, the samples were embedded in paraffin. Paraffin embedded kidneys were sectioned either at 4 um (for immunofluorescence staining) or 10um (for in situ hybridization detection).

Hematoxylin and Eosin Staining

Routine Hematoxylin and Eosin staining (Richard-Allan Scientific, ThermoFisher Scientific) was performed on age matched wildtype control and mutant kidney sections (4um). The protocol used was the one outlined by the manufacturer. Slides were mounted in Permount (Electron Microscopy Sciences, VWR) mounting media. Images were captured using a Nikon Digital-Sight DS-U3 camera mounted on a Nikon Eclipse Ni Fluorescent Scope. Images were processed using NIS elements (version 4.4) software.

Organ Culture

Ex vivo kidney explants were grown in 6-well tissue culture treated Transwell plates with 24mm membrane inserts (Cat#3450 Corning Inc.) with a pore size of 0.4um. The kidneys were cultured in Advanced DMEM/F-12 medium (Invitrogen) supplemented with 10% fetal calf serum and 1% penicillin streptomycin (10,000 U/ml, ThermoFisher Scientific) at the air-medium interface in 5% CO₂ atmosphere at 37°C. To examine ureteric branching morphogenesis the kidney explants were cultured for up to 72 h. Images were captured using an Olympus BX51 fluorescence microscope.

Immunostaining

Clearing and rehydration of paraffin sections (4um) was followed with antigen retrieval (10mM Sodium Citrate, pH 6.0) in a steamer for 45 min. Endogenous peroxidase was quenched by incubating slides in 3% hydrogen peroxide at room temperature. Sections were blocked for 90 min at room temperature in 0.5% blocking reagent (Perkin-Elmer FP1012) in Tris-buffered saline supplemented with 10% normal donkey serum together with unconjugated, monovalent donkey anti-Rabbit Fab (711-007-003 Jackson Immunoresearch laboratories, Inc., PA) and donkey anti-Mouse Fab (715-007-003) fragments at 15 μl/ml each. Primary antibodies used were goat anti-AQP2 (C17, sc-9882, Santa cruz), anti-rabbit Calbindin (ab25085, Abcam), rabbit anti cleaved caspase 3 (Asp 175, Cat# 9661, Cell Signaling), mouse pan Cytokeratin (C2562, Sigma), mouse anti-E-cadherin (610181, BD Biosciences), rabbit phospho-histone H3 (Ser 10) antibody (9701,Cell Signaling), rabbit anti-Jagged1 (H114, Cat# sc-8303, Santa Cruz), rabbit anti-LEF1 (C12A5, Cat #2230, Cell Signaling), mouse anti-p63 (SFI-6, DCS Immunoline), rabbit anti-Six2 (11562-1-AP, Proteintech), goat anti-V-ATPase (B1) (N-20, cat# sc-21206), rabbit anti-V-ATPase (B1/B2) (H-180, sc-20943), rabbit anti-WT1 (C-19:sc-192, Santa Cruz), and DAPI (D1306, Invitrogen).

For secondary detection we used donkey anti-rabbit, donkey anti-mouse, or donkey anti goat antibodies as was required, conjugated to one of the following Alexafluor dyes: Alexa-Fluor555, AlexaFluor488 or AlexaFluor 647 (Molecular Probes, Invitrogen). The Tyramide Signal Amplification (TSA) fluorescence kit (NEL760001KT, Perkin Elmer) was used in instances where the secondary antibody was conjugated to horse radish peroxidase enzyme instead of a fluorophore. The immunofluorescent images were captured with a Zyla/Andor camera mounted on a Nikon Eclipse Ni fluorescent microscope. The confocal images were captured with the aid of the Confocal Nikon eclipse Ti2 scope.

Quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR)

Quantitative qRT-PCR was performed on total RNA isolated from E14.5 kidneys using the RNeasy Mini Kit (Qiagen). Real-time primer-probe mixes were ordered from Applied Biosystems. The assay was done using the Taqman RNA-to-CT 1step kit (4392938; Applied Biosystems, New York, U.S.A). The thermal profile used was as follows: 48°C for 15min, 95°C for 10min and 40 cycles of 95°C for 15s, 55°C for 1 min and 72°C for 1 min. The reactions were performed in triplicate. The relative transcript levels were calculated using the ΔΔCt method with GAPDH as the endogenous normalizing gene. Three independent experiments using total RNA from kidneys of representative genotypes from age-matched litters were used in our assays. The scale bars represent the standard error of mean of average values. P value was calculated using the two-tailed Student’s T test with a probability of 0.05 considered as significant.

In situ Hybridization

In situ hybridization protocol used has been previously described (Hilliard et al., 2011). The samples were fixed overnight in 4%PFA/PBS and prepared for paraffin embedding. The samples were sectioned at 10um. Digoxigenin labeled RNA probes were used to bind the endogenous transcripts in situ. Probe binding was detected using anti-Digoxigenin Fab fragments coupled to Alkaline Phosphatase (Roche Diagnostics) and the chromogenic substrate BM purple.

Genome-wide Microarray Analysis

Microarray analysis was performed using Agilent’s ‘Two color Microarray Based Gene Expression Analysis (Agilent G4140–90050). The assay used 100ng of total RNA isolated from E14.5 whole mouse kidney explants (four pools of wild type matched with four pools of mutant samples) using the RNeasy Mini Kit (Qiagen). The quality of RNA samples was assessed on a Bio Analyzer (Agilent 2100, Palo Alto, CA). Fluorescently labeled (Cy3-CTP or Cy5-CTP) cRNA was generated from 100 ng of total RNA using the Low input quick amp labeling protocol from Agilent. The labeled cRNA was hybridized to Agilent’s 4X 44K whole mouse genome microarray slide. The arrays were scanned using a dual-laser DNA microarray scanner (Agilent). To prevent bias from signal detection with different color dyes (Cy5, Cy3), we applied dye-swap strategy to this experiment, in which identical sample pairs were reverse labeled. The data were then extracted by the Feature Extraction 6.1 software (Agilent). Raw data were processed using GeneSpring GX (version 13.0) software. Only genes, showing a significant (P≤0.05) differential change in expression after Benjamini and Hochberg false discovery rate (FDR) correction in the four datasets, were used for further analyses. Additional analysis of the microarray data was completed using the Ingenuity Pathway Analysis (IPA) software (Qiagen, CA). The raw data is deposited in GEO under GSE137994.

Statistical Analysis.

All experiments were performed on 3–4 biological replicates as indicated in the text. Statistical differences were performed using t-test or one-way ANOVA. A p-value <0.05 was considered statistically significant.