Kidney adysplasia and variable hydronephrosis, a new mutation affecting the odd-skipped related 1 gene in the mouse, causes variable defects in kidney development and hydronephrosis

Abstract

Many genes, including odd-skipped related 1 (Osr1), are involved in regulation of mammalian kidney development. We describe here a new recessive mutation (kidney adysplasia and variable hydronephrosis, kavh) in the mouse that leads to downregulation of Osr1 transcript, causing several kidney defects: agenesis, hypoplasia, and hydronephrosis with variable age of onset. The mutation is closely associated with a reciprocal translocation, T(12;17)4Rk, whose Chromosome 12 breakpoint is upstream from Osr1. The kavh/kavh mutant provides a model to study kidney development and test therapies for hydronephrosis.

mammalian kidney development is a complex process regulated by several transcription factors and regulatory proteins, including EYA1, GATA3, LHX1, RET, SIX2, PAX2, SALL1, LIM1, GDNF, WT1, and OSR1 (4). The kidney develops from mesenchymal progenitor cells in the intermediate mesoderm, from which the metanephric mesenchyme and the nephric duct both arise. By embryonic day 10.5 (E10.5), the metanephric mesenchyme stimulates the adjacent nephric duct to invade the nephric tissue, initiating reciprocal signaling interactions that direct growth and branching of the embryonic kidney.

The odd-skipped related 1 (Osr1) gene, one of two mammalian orthologues of the Drosophila odd-skipped (odd) gene (18), encodes one of the earliest transcription factors expressed during kidney development and is required for metanephric mesenchyme formation (11, 22). Its expression in the intermediate mesoderm activates the formation and expansion of mesenchymal cells, initiating a cascade of other transcription factors. Deletion of Osr1 in the mouse completely blocks expression of key transcription factors in the metanephric mesenchyme, halting development of the embryonic kidney (11, 21, 22). Here, we report a new recessive mutation (kidney adysplasia and variable hydronephrosis, kavh) in the mouse that leads to downregulation of Osr1 transcript, causing variable kidney agenesis, hypoplasia, and hydronephrosis. The mutation is closely linked with a reciprocal translocation, T(12;17)4Rk (hereafter T4Rk), whose Chromosome (Chr) 12 breakpoint is upstream from Osr1. The kidney anomaly has not segregated from the translocation during inbreeding of the strain, suggesting that the translocation breakpoint on Chr 12 may disrupt a regulatory region that is needed for normal Osr1 expression. The kavh/kavh mutant provides a model for better understanding the earliest stages of kidney development and testing potential therapies for hydronephrosis.

MATERIALS AND METHODS

Mice and husbandry.

The origin and genetic background of the inbred strain carrying kavh and T4Rk [STOCK T(12;17)4Rk, Stock No. 001189] is described in results. Embryos from the strain were cryopreserved both as homozygotes (Stock No. 001189) and by crossing homozygous males to C57BL/6J females at generation F6 (Stock No. 001488). All mice were maintained in conventional caging in the Mouse Mutant Resource in The Jackson Laboratory's Research Animal Facility in a room with HEPA-filtered air and a 14:10 light:dark cycle. They were fed initially NIH31 (6% fat) diet, currently 5K52 (6% fat), and acidified water ad libitum. All studies were performed in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) guidelines, and all animal procedures were approved by The Jackson Laboratory (JAX) Institutional Animal Care and Use Committee (Comprehensive protocol No. 99066, latest annual approval date June 27, 2013).

G-banding.

Metaphase chromosomes were prepared from cultured whole blood (5) (http://www.jax.org/cyto/proto.html). Chromosomes in air-dried slide preparations were G-banded by staining 4–6 min in a trypsin:Giemsa mix (2.2% Gurr Giemsa with 4 drops 0.0125 trypsin: 1 ml Giemsa/45 ml Gurr buffer), after aging in a dehydrator for 7 days at room temperature.

FISH with bacterial artificial chromosomes.

Fluorescent in situ hybridization (FISH) was performed on chromosomes of cultured blood cells from T4Rk heterozygotes by a modification of a method published previously (1). Briefly, air-dried slide preparations of chromosomes were hybridized with DIG-labeled bacterial artificial chromosomes (BACs) (digoxygenin-11-dUTP 250–400 ng/slide) and counterstained with 4,6-diamidino-2-phenylindole (DAPI). The DIG-labeled probe was detected with antidigoxygenin Fluorescein Fab Fragments (10 μg/ml), with one round of amplification with Affini Pure Rabbit anti-Sheep IgG (10 μg/ml), followed by incubation with Alexafluor 488 Conjugate Goat anti-Rabbit IgG (5 μg/ml). Antibody dilution buffer contained 0.05% Tween 20, 5% goat serum, and 0.5% Tris-NaCl-blocking buffer. BAC clones in the Chrs 12 and 17 regions of the cytologically identified translocation breakpoints were selected with the Ensembl website and were obtained from the Roswell Park Cancer Institute (RPCI)-23 and -24 libraries through JAX Scientific Services.

Linkage cross.

The kavh mutation was genetically mapped with a backcross and an intercross with CAST/EiJ. Linkage cross progeny DNAs were genotyped for DNA (MIT) markers on Chrs 12 and 17 because these two chromosomes were cytogenetically identified in the translocation, T(12;17)4Rk (hereafter T4Rk). Progeny were scored for the translocation by G-banded chromosome preparations from bone marrow and for kidney defects by necropsy. Data were analyzed with MapManager (16).

Pathology and histology.

Necropsies were carried out on mice euthanized by CO₂ asphyxiation in accordance with AAALAC guidelines. Kidneys were removed from these mice, fixed in Bouin's fixative, sectioned at 4–5 μm, and stained with hematoxylin and eosin.

Ultrasound imaging and analysis.

Vevo 770 High-Frequency Ultrasound (Visualsonics, Toronto, Ontario, Canada) was used to image kidneys in anesthetized mice. A 30 or 40 MHz real-time microvisualization scan head was used to yield ultrasonic images with infiltration depths of 12.7 or 6 mm, respectively. At multiple time points, ∼1.5, 2.5, 4, 6, 9.5, 19.5, 21.5, and 26 wk of age, long-axis and short-axis measurements were collected, as well as total kidney volumes with appearance and progression of cysts documented.

CT scans and analysis.

A MicroCAT II scanner (Siemens Medical Solutions, Melvern, PA) was used to acquire two-dimensional (2D) X-ray images of an anesthetized mouse. The MicroCAT II provides very fine spatial resolution of 24 to 96 microns. This system, combined with BioVet gating hardware and v2.0 software (Siemens Medical Solutions), allows for imaging to only occur during a user-defined segment of the respiratory or cardiac cycle. Fifteen to thirty minutes prior to imaging, mice were given an iodine-based contrast agent (Isovue) via intraperitoneal injection. They were then anesthetized with 5% isoflurane in oxygen at 0.8 liters/min and maintained at 1–1.5% isoflurane throughout imaging. The 2D images were reconstructed with a real-time reconstruction program, RVA (Siemens Medical Solutions), and further analyzed with reconstruction program AMIRA v3.10 (Zuse Institute Berlin). Volumes were calculated for both left and right kidneys.

Molecular analysis.

Long-range PCR was used to narrow the Chr 17 breakpoint region, and the resulting data were used to design a long-range PCR assay across the Chr 12:17 translocation breakpoint to genotype homozygous and heterozygous kavh/kavh mutants. This easy and reliable genotyping assay and nonrecombinant MIT markers were crucial for the BAC rescue experiments (below) because T4Rk/T4Rk kavh/kavh mutants could be genotypically identified. Exons from the candidate Osr1 gene on Chr 12 in mutants were sequenced by the standard Sanger sequencing method.

BAC rescue experiments.

Chr 12 and 17 BACs overlapping the translocation breakpoints identified by FISH were selected with Ensembl and obtained from the RPCI-23 and -24 BAC libraries. Transgenic stocks carrying the selected BACs were generated by pronuclear injection. Mice hemizygous for individual BAC transgenes (C57BL/6J genetic background) were mated to homozygous T4Rk/T4rk kavh/kavh mice from the inbred T4Rk stock generating F1 progeny that were obligate heterozygotes for kavh. F1 mice were genotyped for the BACS and Tg/0 kavh/+ mice were backcrossed to homozygotes from the T4Rk stock. Backcross progeny were genotyped for the BACs and for the kavh mutation with nonrecombinant DNA markers or by the long-range PCR method. All backcross mice were necropsied for kidney phenotype. Missing, hypoplastic, or hydronephrotic kidneys in BAC hemizygotes showed failure of a BAC to rescue.

In situ hybridization.

Whole mount in situ hybridization was performed on E9.5 mouse embryos as described (14). kavh/kavh embryos were obtained from crosses of homozygous T4Rk mice from the T4Rk maintenance colony. Control embryos were taken from C57BL/6J mice, which is the predominant genetic background of the T4Rk strain. A total of nine mutant and nine control embryos were analyzed. The Osr1 riboprobe, ∼600 bp, was synthesized from the 3′ end of the Osr1 cDNA (gift of R. Jiang, University of Cincinnati). Detectable probe development was noted in the mesonephros about 30 min after the addition of 5-bromo-4-chloro-3-indoyl phosphate/nitro blue tetrazolium chloride (BCIP/NBT). Embryos were post fixed in 4% paraformaldehyde in PBS and placed in 50% glycerol in PBS for imaging.

Test for allelism.

A test for allelism or complementation test with Osr1 was carried out by mating kavh/kavh mice from the inbred T4Rk strain to mice heterozygous for a null allele of Osr1 (22) (B6.129S1-Osr1^tm1Jian/J, Stock No. 009387) obtained from The Jackson Laboratory Repository.

RESULTS

Origin of mutation and translocation, genetic background.

T4Rk was induced in the gametes of a DBA/2J male with triethylenemelamine (0.4 mg/kg body wt). The treated DBA/2J (D2) male was mated to a C57BL/6J (B6) female, and T4Rk was identified in a B6D2F1 son (6). Mice carrying T4Rk were backcrossed several generations to B6, AEJ/GnRk, and B6C3HF1 mice to increase litter size sufficiently to obtain translocation homozygotes from heterozygote × heterozygote matings. The line was then maintained by sibling matings and is now an inbred strain derived from these genetic backgrounds. During cryopreservation of the T4Rk strain, it was discovered that T4Rk/T4Rk homozygotes were frequently missing one kidney. Subsequent breeding revealed that the kidney phenotype was inherited in an autosomal recessive manner (see Genetic analysis).

Phenotype of kavh/kavh mice.

The primary defects in kavh/kavh mutant mice are renal: sporadic kidney agenesis, hypoplasia and variable uni- or bilateral hydronephrosis (Fig. 1). Because affected mice do not show abdominal swelling as with cystic kidney disease, live mutants are only detectable by microCT or ultrasound. In some mice, the affected kidney(s) was small, dysmorphic, or missing, which suggested developmental etiology. Small or missing [195] kidneys have been observed at P0. Small kidneys appeared hypoplastic with no obvious scarring characteristic of atrophy following damage. Mice of varying ages exhibited hydronephrosis in one or both kidneys and some with unilateral agenesis showed contralateral progressive hydronephrosis. Necropsies of mice from the inbred T4Rk strain showed the kavh/kavh kidney phenotype was ∼50% penetrant.

Fig. 1.

Gross appearance and histopathology of kavh/kavh kidneys. A: hydronephrotic kidneys in a 10-day-old kavh/kavh male (right) and sibling male control (left). Both the bladder and hydronephrotic kidney on the right are filled with renal fluid, grossly enlarged, and distorted (black arrows). The mutant kidney on the left has pale translucent areas that indicate the early stages of disease (white arrow). B: an adult kidney showing an intermediate stage of hydronephrosis. Dilation of the pelvis is present, the medulla is reduced (arrow), but the cortex is intact. C: an adult kidney with more advanced hydronephrosis. Dilation of the pelvis has caused most of the medulla to atrophy and only a thin crescent layer remains (arrow). The cortex is also thin. Both B and C are hematoxylin and eosin. Intestines were excised for visibility. Scale bar = 1 mm.

Additional phenotypes that have not yet been pursued include possible auditory defects and low-penetrance hydrocephaly. A small sample of mice from the T4Rk strain exhibited hearing loss as early as 3 wk of age, and 15 mice from the backcross tested by auditory brain stem response (ABR) showed complete concordance between the translocation and the hearing loss (T/T hearing loss = 7, T/+ normal hearing = 8). Although embryos homozygous for the Osr1 targeted mutation died of cardiac anomalies (22), no obvious cardiac anomalies were seen in the kavh/kavh embryos.

Histological analysis of early-stage hydronephrotic kidneys revealed dilated kidney pelvis and compressed medulla; the cortex was intact (Fig. 1). With more advanced hydronephrosis, dilation of the pelvis causes extreme compression of the medulla with only a thin crescent layer remaining and now thin cortex. The hydronephrosis was probably not caused by lithiasis because 1) neither in vivo imaging nor histological sections from the ureteropelvic junction of several hydronephrotic mutant mice revealed obvious structural blockage, 2) no stones were seen with CT or ultrasound, and 3) no precipitates were found in the urine of older mice.

To assess nephropathy in affected animals, we measured plasma urea nitrogen (BUN) levels, which are known to positively correlate with severe nephropathy. BUN from severely hydronephrotic mice of all ages ranged from 35 to 54 mg/dl with mean ± SE = 42.29 ± 8.63 (n = 21) compared with control values in mice of a B6C3H related stock (mean ± SE = 23.30 ± 4.79). BUN values in mutant mice with minimal hydronephrosis (pale spots but no kidney enlargement, as in mutant in Fig. 1A) were not statistically different from controls (data not shown).

Longitudinal analysis of kidney hydronephrosis.

Longitudinal ultrasound of 13 kavh/kavh mutants from three litters confirmed the variable nature of the hydronephrotic phenotype (Table 1). Kidney volume to body weight was roughly correlated with degree of hydronephrosis. We observed that two mice were missing one kidney and three mice had one very small kidney and one normal-sized kidney consistent with prior observations (2715-Lt1m, 2717-2R, 2717-4B). Mice that showed hydronephrosis varied in age of onset and rate of progression. For example, the right kidney of mouse 2715-Lt1f was hydronephrotic from 3 wk of age on, whereas in 2717-3L the right kidney became progressively hydronephrotic from 3 to 38 wk of age. Mouse 2715-Rt3 had an extremely hydronephrotic right kidney at 3 wk of age and died at 4 wk of age. Two other mice died prior to 38 wk of age. In these cases, the kidney volume to body weight was not elevated, but they had early stages of hydronephrosis in both kidneys. Five mice appeared unaffected, showing a normal kidney volume to body weight ratio that remained stable over time (2715-Lt2f, 2715-Rt2m, 2715-Rt3m, 2771-2R, 2771-3L).

Table 1.

Abbreviated summary of ratio of kidney volumes to body weights to show variation in expression of hydronephrosis in 12 kavh/kavh, imaged with ultrasound

Mouse ID	Kidney	Sex	3 wk	24 wk	38 wk
2715-Lt1f	Left	F	missing	—	—
	Right		15.1	nd	14.9
2715-Lt2f	Left	F	5.5	4.9	6.7
	Right		6.3	5.9	6.6
2715-Rt3	Left	F	10.3	died	—
	Right		55.6	—	—
2715-Lt1m	Left	M	3.8	3.3	4.7
	Right		6.1	3.0	5.3
2715-Rt2m	Left	M	5.0	3.4	5.7
	Right		6.3	4.1	6.2
2715-Rt3m	Left	M	4.1	4.2	5.0
	Right		4.3	4.6	6.6
2717-2R	Left	M	3.0	5.9	died
	Right		6.3	6.1	—
2717-3L	Left	F	missing	—	—
	Right		5.6	11.4	20.5
2717-4B	Left	M	3.5	nd	4.9
	Right		11.4	7.7	12.9
2771-1N	Left	M	5.8	died	—
	Right		5.8	—	—
2771-2R	Left	M	3.6	5.2	nd
	Right		4.6	4.7	5.0
2771-3L	Left	M	4.6	4.8	7.8
	Right		5.3	4.4	6.6

Values are the ratio kidney volume (mm³)/body weight (g). Missing, clear evidence kidney was absent; nd, not done; died, animal died prior to ultrasound at that age.

Genetic analysis.

The mutation is recessively inherited as no affected mice were found among 65 T/+ translocation heterozygotes in outcrosses. Although we cannot confirm that the kavh mutation has been present in the T4Rk strain from when the first translocation mouse was discovered, the kidney defect did not segregate from the translocation during inbreeding and maintenance of the strain when we were selecting only for the translocation. This tight linkage suggested that one of the translocation breakpoints causes either a sequence disruption or deletion, or alternatively disrupts the regulatory context of the nearby genes leading to the observed phenotype. Although the genetic background is a mix of four different strains (DBA/2J, C57BL/6J, AEJ/GnRk, and C3H/HeJ), we determined that the regions spanning the translocation breakpoints are ancestral DBA/2J, which was the strain background of the male in which the translocation was induced. During inbreeding of the strain, mice from every sixth generation were karyotyped as insurance against translocation loss. After kavh was discovered, colony mice were typed for DNA markers D12Mit219 (12:26441005-26441113), D12Mit46 (12:35041005-35041138), and D17Mit205 (17:68629147-68629369) as well. All genome coordinates are based on NCBI Build m38.

Despite the fact that the reciprocal translocation breakpoints inhibited recombination in their vicinity (segments of ∼12 cM on Chr 12 and ∼16 cM on Chr 17), the backcross (n = 48) and intercross (n = 36 chromosomes) linkage data confirmed that the translocation breakpoints are in proximal Chr 12 and distal Chr 17, placing the Chr 12 translocation breakpoint between DNA markers D12Mit136 (12:30173738-30173884) and D12Mit153 (12:35235497-35235638) and the Chr 17 breakpoint between D17Mit152 (17:65340484-65340609), and D17Mit186 (17:73557074-73557184). Although there appeared to be recombinants between the translocation and the kidney phenotype in the genetic crosses, the variability of the phenotype made it impossible to confirm recombinants. In the backcross, eight individuals in which the T4Rk breakpoint and the phenotype appeared to segregate (T/T, normal kidneys) could be the result of the partially penetrant phenotype because there were 24 T/T and 24 T/+ (50:50), as expected in a backcross, but only 16 of 48 mice (33%) displayed a kidney phenotype. In the intercross, two T/+ mice with affected kidneys could be recombinant or the result of heterozygotes occasionally being affected on a hybrid background. Because T4Rk suppressed recombination in the breakpoint regions on both chromosomes and litter sizes were small, it was clear that we could not obtain a high-resolution genetic map with linkage crosses. Therefore, our strategy for narrowing the region and identifying the underlying kavh mutation was to identify more precisely the T4Rk translocation breakpoint junctions by chromosomal FISH with BAC probes to the chromosomal regions of the translocation breakpoints on Chrs 12 and 17.

Translocation breakpoint analysis.

Analysis of G-banded chromosomes identified the translocation chromosomes (6) and localized the breakpoints to cytological G-bands Chr 12B and 17E3 (7). For FISH analysis, BACs were selected initially based on the translocation breakpoint positions relative to cytological G-bands and genetic mapping data (see below). The small translocation chromosome designated T12¹⁷ has the centromere end of Chr 12 with the distal telomere end of Chr 17; the large translocation chromosome designated T17¹² has the centromere end of Chr 17 with the distal telomere end of Chr 12. We determined which side of the breakpoints each BAC mapped to by visualization of the signal on heterozygous translocation chromosomes. Chr 12 BACs that hybridize to T12¹⁷ are proximal to the Chr 12 breakpoint; those that hybridize to T17¹² are distal to the breakpoint. Chr 17 BACs that hybridize to T17¹² are proximal to the breakpoint on Chr 17, and those that hybridize to the T12¹⁷ are distal. By starting on each side of the breakpoints (one BAC hybridizing to one translocation chromosome and one hybridizing to the other) and walking along the chromosome with BACs in between, we determined which BACs flank the breakpoints (Fig. 2, A–C). These FISH results narrowed the candidate gene interval(s) to <0.5 Mb on both Chr 12 and Chr 17.

Fig. 2.

Localization of the T4Rk translocation breakpoints by FISH. A: diagram illustrating location of BAC probes, Chromosome (Chr) 12 probes in black, Chr 17 probes in red. B and C: examples of hybridization of Chr 12 probe to Chr 12 and translocation chromosome 17¹² (B) and Chr 17 probe to Chr 17 and translocation chromosome 12¹⁷ (C) (DAPI false colored red to indicate Chr 17 BACs). D and E: molecular location of cytological breakpoint intervals as defined by flanking BAC probes on translocation chromosomes 17¹² (D) and 12¹⁷ (E). See also Tables 2 and 3.

Chr 12.

Initially, BACs mapping to cytological bands B1–B3 were selected for FISH because the cytological breakpoint data and the genetic data positioned the translocation breakpoint within this region (Table 2). The breakpoint interval was defined by BAC RP23-296J15 (at 12:8962294-9159908bp) on the proximal side and BAC RP24-289C19 (at 12:9239596-9417247bp) on the distal side (Fig. 2, D and E). These data place the Chr 12 breakpoint proximal to both the genetic and cytological (G-band) position (Ensembl).

Table 2.

Positioning of Chr 12 BACs with respect to the Chr 12 breakpoint by FISH analysis

BAC Library	BAC ID	Base Pair Map on Genome Sequence	Cytological Band	FISH Signal on Translocation Product
RP23	296J15	8962294-9159908	dist A1.3	T1217
RP24	289C19	9239596-9417247	prox A2	T1712
RP23	138M19	9384958-9564695	prox A2	T1712
RP23	398M9	9498256-9691061	prox A2	T1712
RP23	36H21	10097957-10320003	prox A2	T1712
RP23	139H6	10647692-10837885	prox-mid A2	T1712
RP23	335D14	13764950-13958080	prox A3	T1712
RP23	440I16	23974392-24153886	prox-mid B1	T1712
RP23	378B3	26248548-26469096	mid-dist B1	T1712
RP23	66A14	26649794-26863717	dist B1	T1712
RP23	139L8	29965347-30167750	mid B2	T1712
RP23	74N6	51340153-51457989	dist C1	T1712

The bold line indicates the breakpoint on Chr 12. Chr, chromosome; BAC, bacterial artificial chromosome; FISH, fluorescent in situ hybridization.

Chr 17.

Initially, BACs mapping to cytological band E3 were selected for FISH because the cytological breakpoint data positioned the translocation breakpoint within this band (Table 3). The Chr 17 breakpoint interval encompasses a segment flanked by sequences recognized by BACs RP23-20F13 on the proximal side (at 17: 73622361-73832587bp) and RP23-100K11 on the distal side (at 17: 73879550-74086730bp) (Fig. 2, D and E). Both BACs map to cytological band E2, which is proximal to the cytological (G-band) location (Ensembl).

Table 3.

Positioning of Chr 17 BACs with respect to the Chr 17 breakpoint by FISH analysis

BAC Library	Chr 17 BAC	Base Pair Map on Genome Sequence	Cytological Band	FISH Signal on Translocation Product
RP23	319G18	70263418-70437776	dist E1.3	T1712
RP23	294D3	73122678-73292416	dist E2	T1712
RP23	20F13	73622361-73832587	dist E2	T1712
RP23	100K11	73879550-74086730	dist E2	T1217
RP23	304O17	73979515-74171161	dist E2	T1217
RP23	456J9	83261478-83450649	prox E4	T1217
RP23	439E12	84823659-85024348	mid E4	T1217

Candidate gene analysis on Chr 17.

Analysis of the genomic region surrounding the Chr 17 translocation breakpoint ruled out the only three genes located between the flanking BAC sequences on Chr 17 (Fig. 2, D and E). Ehd3 (EH-domain containing 3), near the breakpoint on the proximal side [17: 73455501-73482753 (+)] is a member of a family of four EH-domain-containing proteins, which are ATP- and calcium-binding components of endocytic vesicles and appear to be involved in endocytic trafficking (17). In mouse, Ehd3 is highly expressed in kidney and brain (8), two organs that are involved in the phenotype we observe in kavh/kavh mutants. Transgenic mice carrying the BAC RP23-314C23 (73209539-73399250bp) that contains all exons and introns of Ehd3 failed to rescue the phenotype. Thirteen of forty-seven transgenic mice from 10 different mated pairs showed a kidney defect and thus were presumably kavh/kavh: seven with only one kidney, four with tiny kidneys, and two with hydronephrosis. In these and other BAC rescue experiments, we did not test for BAC expression because it was not clear which tissues and at what age to test; therefore, we made crosses with mice from multiple founder lines.

The second candidate, Xdh (xanthine dehydrogenase), is near the breakpoint on the distal side [17: 73534568-73600842bp (-)]. XDH deficiency is known to be responsible for a human disease called type I xanthinuria (10); several human case studies have reported kidney abnormalities, including hydronephrosis (16a). Deficient XDH leads to a strikingly reduced level of uric acid in blood and urine, diagnostic for type I xanthinuria. Blood uric acid levels of 12 kavh/kavh mutants of various ages and nonsymptomatic littermate controls from the T4Rk stock showed normal levels of uric acid. Thus XDH was eliminated from further study.

A potential novel gene covering a genomic region of about 30 kb, with three exons predicted from yeast cDNA and a large second intron, is located between Edh3 and Xdh. Sanger sequencing of the exons in this gene showed 90% sequence homology with the Chr 8 Gpsn2 (glycoprotein, synaptic 2) gene, which encodes an integral membrane protein with unknown functions. For transgenic rescue, we modified BAC RP24-226L7 (73233184-73499682bp), which contains both Ehd3 and the novel gene, to truncate Ehd3 and leave the novel gene intact. Two of the first three transgenic kavh/kavh mice recovered were missing one kidney. Failure to rescue the phenotype ruled out the novel gene.

Identification of Osr1 on Chr 12 as the affected gene.

While T(12;17)4Rk breakpoint analyses and the transgenic rescue experiments were in progress, an update of mouse reference genome build placed the Osr1 gene closer to the breakpoint interval, revealing it as a potential candidate. Concurrently, a targeted null allele of the Osr1 gene was published with a recessive embryonic lethal phenotype accompanied by kidney defects (11, 22). Sanger sequencing of all three exons of Osr1 did not reveal any differences between kavh/kavh and controls or reference sequence, suggesting the lesion is in a noncoding or regulatory sequence. The 5′ untranslated region (UTR) is in exon 1, the coding region begins in exon 2, and the 3′ UTR is at the end of exon 3; all were sequenced without finding any putative lesions. In addition, whole exome sequencing of kavh/kavh DNA did not reveal any coding mutations within the mapped interval. Although other mutant genes on Chr 12 gave kidney phenotypes, they were not as close to the breakpoint interval. A kidney phenotype is not reported for Wdr35 alleles, the only other gene near the breakpoint with published mutant mouse alleles.

Analysis of Osr1 expression.

Because the translocation does not disrupt any coding sequence, we hypothesized that misregulation of the Osr1 gene through modification of distant regulatory elements could explain the kavh phenotype. We performed in situ hybridization to examine the expression of the Osr1 gene in the intermediate mesoderm in E9.5 embryos. As shown in Fig. 3, we observed a consistent but incomplete reduction of Osr1 expression in the intermediate mesoderm in kavh/kavh embryos vs. wild-type C57BL/6J controls. These data suggest kavh is a hypomorphic allele of Osr1, with residual expression at levels permissible for survival and for kidney development, but frequently resulting in both agenesis and hydronephrosis.

Fig. 3.

Expression of Osr1 is reduced in kavh/kavh embryos. Whole mount in situ hybridization for Osr1 of E9.5 control (A and C) and kavh/kavh (B and D) embryos. Blue/purple staining indicates Osr1 mRNA expression. Reduced expression is seen in the developing metanephric mesenchyme in mutant embryos vs. control. H, heart.

Test for allelism.

To further test our hypothesis that kavh is an allele of Osr1 we mated heterozygotes for the targeted mutation of Osr1 (Osr1^tm1Jian) (22) to kavh/kavh mice from the homozygous T4Rk stock for a complementation test; all F1 progeny were obligate heterozygotes for kavh. All 12 F1 pups that were doubly heterozygous (i.e., kavh/Osr1^tm1Jian) failed to complement and died either prenatally or within the first 24 h after birth. In contrast, 13 wild-type mice for the Osr1 targeted mutation and kavh/+ survived to adulthood without apparent abnormalities. Histological analysis revealed that all kavh/Osr1^tm1Jian P0 pups completely lacked both kidneys (Fig. 4), further demonstrating that kavh is a hypomorphic allele of Osr1. In addition, the increased severity (100% penetrant kidney agenesis) supports a stochastic gene dosage threshold model for Osr1 expression in kidney development.

Fig. 4.

Complementation test with an Osr1 targeted mutation. Transverse sections of P0 pups heterozygous (control) for the kavh mutant allele (A and C) and compound heterozygotes for kavh and Osr1^tm1Jian alleles (B and D) born from a cross of kavh/kavh mutant animals to Osr1^tm1Jian heterozygotes. Compound mutant animals display complete agenesis of both kidneys and die shortly after birth (B and D). Note the normal development of both adrenal glands in the absence of the kidney. K, kidney; A, adrenal; L, liver.

DISCUSSION

We report a new, recessive mutation in the mouse (kidney adysplasia and variable hydronephrosis, kavh) with a primary phenotype of abnormal kidney development. The kidney phenotype shows incomplete penetrance and variable expressivity, ranging from small or missing kidney(s) to uni- or bilateral hydronephrosis or a combination of these. In addition, some homozygotes have no kidney phenotype. The mutation is tightly linked to a chromosomal translocation (T4Rk) whose Chr 12 breakpoint appears to be 300–400 kb proximal to the odd-skipped related 1 (Osr1) gene on Chr 12. Analysis of Osr1 transcript in embryos and a test for complementation with a targeted mutation that ablates OSR1 expression confirmed that expression of the Osr1 gene is disrupted by the T4Rk translocation breakpoint.

Nature of the kavh mutation.

Although DNA sequence analysis of T4Rk/T4Rk kavh/kavh mice does not reveal disruption of the coding sequence of the Osr1 gene, our expression and complementation tests confirm that kavh is indeed an allele of Osr1. The fact that mutants do not entirely lack both kidneys suggests it is a hypomorphic allele. It is likely that the T4Rk translocation disrupts one or more key distant enhancer or regulatory elements, thus altering its expression in the metanephric mesenchyme. Compound Osr1 KO/kavh heterozygotes display defects specific to the kidney, without any evidence of the cardiac phenotypes present in the Osr1 knockout (22), supporting the hypothesis that kidney-specific expression regulatory sequences or enhancers are affected by the T4Rk translocation. This is consistent with the genomic and regulatory architecture of transcriptional regulators, which can frequently lie quite distant from the proximal promoter. Further studies to identify the precise translocation breakpoint, and thus the putative enhancers involved, are currently underway.

Role of OSR1 in kidney development.

OSR1 plays multiple roles in mouse development, including a central role in fate determination of the metanephric mesenchyme. Its expression precedes and is required for the subsequent expression of key determinants of kidney development. Loss of Osr1 in the mouse results in lethality at E11.5 in most homozygous mice because of cardiac abnormalities (22). Ureteric bud and metanephric mesenchyme condensation also are absent in homozygous null embryos at 11.5 when the ureteric bud has already invaded the metanephric mesenchyme in wild-type embryos (11). Mice lacking Osr1 also completely lack several other factors required for metanephric kidney formation, including EYA1, SIX2, PAX2, SALL1, and GDNF, in the metanephric region. In addition, expression of Lim1, which is required for formation of the nephric duct (12) and is a specific marker for the nephric duct primordium, is reduced, and defects in the nephric duct are present from as early as the initiation of nephric duct formation (11).

In our kavh/kavh mutant, variably reduced Osr1 expression leads to small or missing kidneys. In human newborns, variant alleles of OSR1 that reduce expression also lead to reduced kidney size and function, and cultured kidney cells with some of these variants have no OSR1 expression (24). Gain-of-function experiments in the chick suggest short-term ectopic expression results in a fate shift toward tubule formation with increased expression of specific tubule determinants, while longer-term expression inhibits tubule and kidney development (11). Thus it appears that tight and dynamic regulation of Osr1 in time and space is essential for proper formation of the kidney at various stages of development. The variable hypomorphic nature of the kavh allele could provide a particularly useful tool for studying events during intermediate mesoderm development, as it does not always result in complete agenesis.

Renal agenesis and hypoplasia.

Mouse mutations in several different transcription factor or regulatory protein genes lead to renal agenesis or dysplasia, revealing key pathways controlling kidney morphogenesis (Mouse Genome Database, informatics.jax.org). For example, ablation of RET or PAX2 leads to complete agenesis and pre- or perinatal death (3, 20). Complete kidney agenesis also results from failure of induction of the ureteric bud in Gdnf-null mutants (19) and in SLIT1 or its receptor ROBO2 (9). Like our kavh Osr1 mutant, mice with knockouts of the eyeless (Eya1) gene lack kidneys and ear defects, showing common pathways in the two organs (23). Because OSR1 is one of the earliest markers of metanephric mesoderm, it provides an earlier entry point into the kidney development cascade. Unlike homozygotes for the Osr1 knockout, which die early in embryogenesis from cardiac abnormalities, the kavh/kavh mutants survive to birth and often adulthood despite unilateral kidney agenesis or hypoplasia. This hypomorphic allele enables comparison of severity and types of defects in pro-, meso-, and metanephric with levels of Osr1 expression.

In human beings renal agenesis or adysplasia, uni- or bilateral absence or hypoplasia of the kidneys (OMIM No. 191830), occurs in 1 in 4,000 to 1 in 6,400 human births. In some cases, these congenital defects appear to be associated with two candidate genes, RET and UPK3A. Mutations in a third, PAX2, cause papillorenal syndrome including isolated renal hypoplasia (OMIM No. 120330). Haploinsufficiency for EYA1 in human beings results in Branchio-oto-renal BOR syndrome with missing kidneys and ear abnormalities (OMIM No. 601653). Renal adysplasia also is associated with several pleiotropic syndromes. So far none of these syndromes have been mapped to human Chr 2p24 where OSR1 is located, despite the fact that some variants of human OSR1 are associated with kidney adysplasia (24).

Hydronephrosis.

Mouse mutations causing hydronephrosis, accumulation, or retention of urine in the kidneys leading to destruction of kidney tissues, are much less common. Deletion of β-catenin (Ctnn1b) by a Tcf21/Pod1-Cre driver strain that expresses Cre recombinase throughout the condensing and stromal mesenchyme of developing kidneys results in mice born with hypoplastic kidneys, hydroureters, and hydronephrosis (15). Ablation of uroplakin 1 or 2 (Upk1 and Upk2) genes alters uroplakin plaques in the apical surface of mouse urothelium and leads to reflux and hydronephrosis at later stages of development and into adulthood (13). Mouse mutants with null mutations in both Adamts1 and Adamts4 usually die perinatally because of thinning of the renal medulla expressed at birth but not during embryogenesis (2). Again, because of the early expression of OSR1, analysis of kavh/kavh mutants may reveal early embryogenic defects that can lead to subsequent hydronephrosis.

Hydronephrosis is one of the most common urological disorders seen in pediatric clinics. Estimates of incidence range from 1 to 2% of children born (http://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0001535/). This developmental disorder is frequently detected in fetuses by ultrasound. Mild and moderate forms may resolve as children grow, but severe hydronephrosis can be life threatening, leading to kidney infections, loss of nephrons, and, ultimately, renal failure. Hydronephrosis is caused by abnormalities of the urinary tract. The most common cause results from narrowing or blockage at the ureteropelvic junction where the ureter joins the kidney. Genetic forms are usually associated in humans with pleiotropic syndromes (OMIM, multiple records). A second cause is reflux or the backward flow of urine from the bladder back into the ureters and kidneys usually due to functional or anatomic defect at the ureterovesical junction where the lower end of the ureter enters the bladder. This condition may be hereditary and seven mapped genes (VUR1–VUR7) have been associated with vesicoureteral reflux (OMIM No. 193000). VUR2 (OMIM No. 610878) is a mutation in the ROBO2 (homolog of the Drosophila roundabout) gene, which is involved in early kidney development. So far hydronephrosis in human beings has not been associated with the OSR1 gene. In adults, hydronephrosis can be caused de novo by mechanical obstructions such as kidney stones, neoplasias, scar tissue, etc. Since human beings can function with one kidney, the hydronephrosis in kavh/kavh mice is unlikely to result from stress due to compensation for the hypoplastic or missing kidney. It seems likely that a developmental defect we did not detect causes obstruction in the urinary tract leading to hydronephrosis.

Value of new mutant.

Defects in kidney development seen in kavh/kavh mice are consistent with the established key role of OSR1 in kidney development. Additionally, the hypomorphic nature of the kavh allele reveals novel phenotypes, including variability in kidney agenesis and postnatal development of hydronephrosis, that are not present in complete loss of Osr1 function (11). The former suggests a stochastic failure of initiation of kidney development below a minimum threshold of Osr1 transcript level, providing a useful model to study intermediate mesoderm development not possible with complete null alleles where the penetrance of agenesis is 100%. The latter provides evidence that misregulation of Osr1 might also contribute to disease pathogenesis, although it is not clear if the etiology of the hydronephrosis is developmental in origin or an emergent property in the postnatal period. Further studies including detailed expression analysis at multiple points of development would provide insight into this question. If the apparent hearing defect is confirmed, further analysis could reveal a role for Osr1 in the ear and a common pathway in the ear and kidney. Thus this mutant mouse provides a useful new model for studying the etiology and genetics of human kidney development and hydronephrosis.

GRANTS

This work was supported by National Institutes of Health Grants P40 RR01183 (to the Mouse Mutant Resource), R01 DK064674 (to M. Davisson), and Cancer Center Core Grant P30 CA034196.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.T.D., S.A.C., and S.A.M. conception and design of research; M.T.D., S.A.C., E.C.A., D.L., C.H., P.G., H.F., and S.A.M. analyzed data; M.T.D., S.A.C., E.C.A., D.L., C.H., P.G., H.F., and S.A.M. interpreted results of experiments; M.T.D., E.C.A., and S.A.M. prepared figures; M.T.D. and S.A.M. drafted manuscript; M.T.D., S.A.C., E.C.A., and S.A.M. edited and revised manuscript; M.T.D. and S.A.M. approved final version of manuscript; S.A.C., E.C.A., D.L., C.H., P.G., and H.F. performed experiments.