Pathogenic PHIP Variants are Variably Associated With CAKUT

Introduction Congenital anomalies of the kidney and urinary tract (CAKUT) represent the most common cause of chronic kidney disease in children. Although only 20% of cases can be genetically explained, the majority remain without an identified underlying etiology. The neurodevelopmental disorder Chung-Jansen syndrome (CHUJANS) is caused by haploinsufficiency of Pleckstrin homology domain-interacting protein (PHIP) and was previously associated with genital malformations. Anecdotal coincidence of CHUJANS and CAKUT prompted us to investigate whether urorenal malformations are part of the phenotypic spectrum of CHUJANS. Methods Analysis of existing CHUJANS and CAKUT cohorts, consulting matchmaking platforms, and systematic literature review to look for additional patients with both CHUJANS and CAKUT. Prenatal expression studies in murine and human renal tissues to investigate the role for PHIP in kidney development. Results We identified 4 novel and 8 published cases, indicating variable expressivity with a urorenogenital trait frequency of 5% to 35%. The prenatal expression studies supported a role for PHIP in normal kidney and urinary tract development. Conclusion Pathogenic PHIP gene variants should be considered as causative in patients with syndromal CAKUT. Conversely, patients with CHUJANS should be clinically evaluated for urorenogenital manifestations. Because neurodevelopmental disorders are often associated with kidney phenotypes, an interdisciplinary re-evaluation offers promise in identifying incompletely penetrant kidney associations and uncovering novel molecular mechanisms of disturbed nephrogenesis.

D isturbances during kidney and urinary tract development result in a wide spectrum of malformations collectively referred to as CAKUT. 1,2These malformations include kidney agenesis, hypoplasia, multiplex/duplex kidneys, multicystic dysplasia, ureteropelvic junction obstruction, ureterovesical junction obstruction, vesicoureteral reflux, posterior urethral valve disorders, and various other malformations. 35][6] Monogenic causes account for w20% of CAKUT and an additional 5% to 10% of cases are attributed to larger copy number variations affecting functions of multiple genes. 7To date, >50 genes have been associated with CAKUT, a number that is steadily growing. 8CAKUT may present as an isolated, nonsyndromic condition, but is often part of a syndromic disease with multi-organ involvement. 1,2In syndromic diseases, such as neurodevelopmental disorders, urorenogenital phenotypes may only occur in a subset of patients as previously demonstrated for ATN1, 9 ROBO1, 10 KIF4A, 11 and other associated genes.The exploration of a possible urorenogenital involvement in established congenital anomaly syndromes, notably neurodevelopmental disorders, may provide insights into molecular mechanisms of nephrogenesis and enable the discovery of novel CAKUT candidate genes.
CHUJANS (MIM# 617991) is a distinct neurodevelopmental disorders caused by heterozygous lossof-function variants in PHIP (MIM *612870). 12PHIP encodes for 3 functional proteins through alternative splicing: the 902-amino acid C-terminus called PHIP1, 13 the 1019-amino acid N-terminus called neuronal differentiation-related protein, 14 and the full-length protein called DNA damage binding protein 1 and CUL4 (Cullin4)-associated factor 14 (DCAF14). 15The latter belongs to the DCAF protein family and acts as a substrate receptor in the ubiquitin ligase pathway, 16,17 regulating and controlling the essential ubiquitination reaction. 180][21][22][23] Until recently, there was no established link between urorenal anomalies and this disorder.Prompted by the genetic diagnosis of CHUJANS in an adult patient with CAKUT-related KF (index patient), we hypothesized that pathogenic variants in PHIP might interfere with normal kidney development.In this study, we aimed to systematically investigate the frequency of CAKUT as a variable trait in patients with CHUJANS.

Patients
The index patient (ID1) and her family were recruited from the transplant unit at the University of Leipzig Medical Center.The legal guardian provided written informed consent, and the study was approved by the Institutional Review Board (IRB) at the University of Leipzig, Germany (IRB00001750; #402/16-ek) and by the IRBs at Columbia University New York and Boston Children`s Hospital, Boston, USA.To search for more cases, we used the GeneMatcher platform (https:// genematcher.org) 24and systematically reviewed the literature for published cases of CHUJANS and CAKUT or complex genital malformations.Furthermore, we screened established cohorts of patients with CHU-JANS for CAKUT spectrum at the Department of Medicine at Columbia University, New York (n ¼ 61), and at the institute of Human Genetics at University Leipzig Medical Center, Leipzig (n ¼ 4).Conversely, exome data from a cohort of unresolved CAKUT cases were screened for CHUJANS at Boston Children's Hospital, Boston (848 families).For reverse phenotyping, we collected detailed clinical data of the newly identified patients using structured clinical questionnaires.All newly reported patients provided written informed consent for publication.

Molecular Genetics
ID1 underwent whole-exome sequencing conducted from blood-derived DNA samples.Virtual panel analysis excluded competing variants in known genes associated with CAKUT and other hereditary nephropathies.Additionally, parental segregation analysis was done through direct Sanger sequencing to confirm de novo status.Whole-exome sequencing was also performed in ID2.1 and ID2.2, and the parenteral segregation was done by direct Sanger sequencing.The family of ID3 was studied using trio whole-exome sequencing.Copy number variation analysis was performed on the basis of high coverage next-generation sequencing data to the best of our knowledge.For ID4-11 next-generation sequencing-based clinical exome or gene panel testing were conducted as previously published. 19,21,22,25All variants are reported using the reference sequence NM_017934.7 and checked for existing entries in main population and patient databases (gnomADv4 [http://gnomad.broadinstitute.org/], 26 ClinVar (last accessed on January 31, 2024, and HGMD version 2023.4).

Immunostaining in Human Fetal Kidneys
To demonstrate the presence of PHIP in the developing human kidney, we performed immunostaining targeting PHIP with polyclonal rabbit anti-PHIP antibody (Cat# HPA019140, RRID:AB_2670073, Sigma-Aldrich, Darmstadt, Germany). 27We stained fetal human kidney sections from weeks 14 to 22 gestational age.
Murine RNA In Situ Hybridization Expression of Phip mRNA was assayed in the developing mouse kidney by in situ hybridization using RNAScope multiplex V2 Reagent Kit (ACD a biotechne brand, Newark, USA) according to the manufacturer's instruction.In detail, embryonal mouse kidneys at Theiler stage (TS) TS20, TS22, TS24, and postnatal at day 3, 7, and 14 and after 11 weeks (adult) were stained for Phip mRNA using specific probe 1221641-C1 (ACD a biotechne brand, Newark, USA).To highlight the metanephric mesenchyme and developing glomeruli, mRNA of the podocyte marker Wilms Tumor 1 (Wt1) was assayed with probe 432711-C2 (ACD a biotechne brand, Newark, USA).Counterstaining on nuclei was performed with 4',6-diamidino-2-phenylindole.Tissue samples of embryos and postnatal animals were collected from C57Bl6 mice.The procedures were performed under the license T0063/20 issued by the local authorities (Landesamt für Gesundheit und Soziales Berlin, LAGeSo, Germany).After excision, kidneys were immersion fixed using Roti Histofix (Carl Roth, Karlsruhe, Germany) for at least 24 hours.After fixation, samples were paraffinized using ethanol solutions of increasing concentration followed by infiltration with paraffin.Quantification of PHIP signal was performed on 5 to 10 nonoverlapping images (depending on the size of the kidney), each representing 0.019 mm 2 of the assayed kidney.To count Phip punctuae, the 8bit image of the channel was subjected to automated thresholding, and the resultant particles were analyzed using ImageJ version 1.53.

Clinical Characteristics and Genetic Analysis of the Index Patient
The ID1, a 42-year-old woman, underwent genetic testing before being placed on the kidney transplant waitlist.At the age of 40, the patient had developed kidney failure because of congenital bilateral hydronephrosis based on ureterovesical junction obstruction and recurrent urinary tract infections.In addition, the patient presented with dysmorphic facial features, obesity (Body mass index ¼ 35.6 kg/m 2 ), and infancyonset developmental delay later evolving to intellectual disability (Figure 1a-e, Table 1).Next-generation sequencing-based gene-panel diagnostics did not detect pathogenic variants in established CAKUT genes or genes that were already associated with hereditary nephropathies.By exome sequencing, however, we detected a heterozygous nonsense variant (c.241C>T, p.Arg81*) in PHIP leading to an abrogation of the eight b-propellerforming WD40 repeat domain (Figure 2a, b, and d).This variant was found absent from population databases (gnomADv4) and is listed as pathogenic in ClinVar (Variation ID: 1710707). 28Parental segregation analysis revealed it to be a de novo variant (Table 1; Figure 1b).
For systematic evaluation of an association of PHIP haploinsufficiency and CAKUT, we clinically reevaluated a large CHUJANS cohort (Columbia cohort), genetically assessed a large CAKUT cohort (Boston cohort), and used the GeneMatcher platform by searching for PHIP cases with CAKUT presentation. 24s a result, we identified 8 additional cases.Of note, ID4-8 were recently published as part of an aggregate description from the CHUJANS cohort at Columbia University. 25[22][23]29,30 As a result, the frequency of urorenogenital phenotypes in patients with diagnostic PHIP variants and a clinical CHUJANS-diagnosis was found to range between 5% and 35% (n ¼ 37/115) (Figure 2j and k), however, in patients with genetically unexplained CAKUT this was only the reason in 0.12% (n ¼ 1/848).

CAKUT-Cohort Evaluation
By trio exome sequencing of 848 families with genetically unresolved cases of CAKUT, we identified only 1 heterozygous nonsense variant in PHIP (c.637_638del, p.Asp213*) in 2 siblings (ID2.1 and ID2.2).While the older sibling displayed a Prader-Willi-like syndrome with obesity and unilateral kidney hypoplasia, his younger brother showed severe unilateral ureteropelvic reflux resulting in recurrent pyelonephritis.Both presented with 2 of 3 toe syndactyly.The identified PHIP variant was inherited from the mother, who did not show any kidney or urogenital anomalies but did have syndactyly in common with ID2.1 and ID2.2 (Table 1; Figure 1f-k).The variant leads to a truncation within the 8 b-propeller-forming WD40 repeat domains (Figure 2a, b, and e).

Matchmaker Initiatives
Another individual (ID3) was detected through Gene-Matcher. 24The 4-year-old boy showed complex genitourethral malformations, specifically, a deep penoscrotal hypospadia, penis deviation, and bilateral cryptorchidism.He had corrective surgeries after birth, was growth retarded, and developmentally delayed with an anxiety disorder.Genetic testing revealed an ultrarare heterozygous missense variant in PHIP (c.76G>A, p.Gly26Arg) predicted to be deleterious by multiple in silico scoring systems (CADD-PHRED 29.4).The deduced amino acid change is located N-terminal of the 8 b-propeller-forming WD40 repeat domains.
Parental segregation analysis confirmed a de novo status (Table 1; Figure 2a-c).
On the protein level, all 4 missense variants are located in the highly conserved N-terminus (Figure 2c).Similarly, 3 out of 7 truncating alterations are located N-terminal or within the 8 b-propellerforming WD40 repeat domains.We applied alphafold protein structure database to predict the impact on 3D structure (3D model alphafold (Q8WWQ0-F1) (Figure 2b-i). 31

PHIP Exhibits Strong Expression in Normal Human Embryonic Kidneys
To address the question whether PHIP-defects may be causally related to CAKUT, we investigated PHIP expression in human fetal kidneys from unaffected controls.Staining of human kidney sections from the 14 th and 22 nd week of gestation using a polyclonal PHIP antibody revealed widespread expression in the developing kidney.Strong expression was observed in the collecting tubules and the ureteric bud tips, glomerular cells, and the cortical mesenchyme.Moderate expression was found in the cap mesenchyme and the distal tubules, with very low expression in metanephric vesicles, s-shaped bodies, the medullary mesenchyme, and proximal tubules (Figure 3a-d).

Robust Phip Expression was Observed in Murine Embryonic Kidneys
To further investigate PHIP organ expression patterns during embryonic and fetal development, we investigated murine tissues from various developmental stages.Embryonic and postnatal in situ hybridization using RNAscope provided an overview of Phip expression during organogenesis (Figure 4).In early embryogenesis (TS 20) Phip RNA was expressed in most tissues, including gonad and kidney (Figure 4a).The expression increased at TS 22 in tissues with high mitotic activity, such as precursor thalamus or diencephalon (Figure 4b).During later stages of embryogenesis (TS 24), PHIP expression became more pronounced, with strong emphasis on selected organs, including the developing kidney (Figure 4c).In the early embryonic mouse kidney (TS 20), PHIP expression was localized to both the mesenchyme and the ureteric bud (Figure 4d).At TS 24 Phip expression was evident in all structures of the cortical region (Figure 4f).However, glomerular Phip expression was low.

Renal Phip Expression Decreases Postnatally
To investigate Phip expression postnatally, we performed RNA in situ hybridization in murine kidney sections.At developmental stage P3 Phip expression was most prominent in the outer cortical regions and in the collecting ducts (Figure 5b-d).As kidney development progressed, expression in the outer cortical activity regions declined at P7 (Figure 5f-h), followed by a further decrease at P14 (Figure 5j-l) and disappeared in the kidneys of adult mice (Figure 5n, and  o).Overall, the expression of Phip persisted in the  medullary region at P7 and P14 (Figure 5g and k) and was lost in the kidneys of adult mice.Quantification of renal Phip expression (Figure 5a, e, and i) indicated equal expression in the cortical and medullar regions at P3 and P7, with a significant decline in cortical PHIP expression at P14 and a generally low Phip expression in the adult kidneys (Figure 5m and p).

DISCUSSION
In this study, we report on congenital urorenogenital anomalies as part of the CHUJANS-spectrum by identification and characterization of 4 individuals (ID1-ID3) with (likely) pathogenic PHIP variants.Additionally, we found 5 individuals from a large PHIP cohort (ID4-8) and 3 individuals with CHUJANS and syndromic CAKUT through systematic literature review (ID9-ID11). 19,21,22,25When interrogating CHU-JANS cohorts retrospectively, we found kidney and urogenital phenotyping to be underrepresented, leaving the possibility of undiagnosed CAKUT in CHU-JANS.Therefore, we believe that w5% to 35% of individuals with CHUJANS exhibit CAKUT and/or genital anomalies as a variable feature of the phenotypic spectrum.
CAKUT represents one of the most common malformations at birth 32 and is a common cause for kidney failure over lifetime. 2,4In the majority of cases, the underlying cause remains unknown, 7,33,34 probably because of genetic heterogeneity, variable expressivity, incomplete penetrance, and environmental contributors. 2][11] The full-length protein complex CUL4-DDB1 (DCAF14) encoded by PHIP 13,14 forms the active substrate receptor of E3 ubiquitin ligases. 16][42] Furthermore, DCAF14 enhances replication fork stability by preventing fork collapse into toxic double-strand breaks protecting DNA from degeneration and thus protecting genome integrity. 15,40hese crucial regulatory functions of PHIP in the cell cycle make it very likely that haploinsufficiency has systemic effects.Phenotypic changes can, therefore, be expected in many different tissues, including the urorenogenital tract.Previously, only cryptorchidism was part of the known phenotypic spectrum of CHU-JANS, 19,21,23 indicating a role in urogenital development in humans.Of note, a large number of DCAFs, which are conserved WD40 genes, are highly expressed during testicular development and spermatogenesis in mice and humans. 41he murine RNA in situ hybridization data show that PHIP localizes to diverse developing organs, such as the central nervous system, lungs, and the kidneys.These high levels of physiological PHIP expression in the kidneys increase the possibility of causing CAKUT when falling below a certain threshold.The postnatal decline in PHIP expression levels provides additional evidence supporting this assumption and is almost undetectable in the fully developed kidneys of adult mice.This is in accordance to other findings of expression pattern of DCAFs at different developmental stages of the testis. 41In a transgenic mouse model of PHIP deficiency, decreased length and decreased overall survival were observed, supporting a role in postnatal growth and lifespan. 43uman single-cell transcriptomic databases underscore the role of PHIP during nephrogenesis with high expression in nephron progenitor, interstitial proliferating, and tubular precursor cells at the 17 th week of gestation (http://humphreyslab.com/SingleCell/display charts.php). 44,45Differences in cortical and medullar PHIP expression at different development stages might illustrate the variable tissue differentiation at 1 time point.In the medulla, there are regions with enhanced PHIP signaling, which could correspond to yet undifferentiated mesenchymal progenitor cells.
Interestingly, urorenal malformations were extremely heterogeneous, ranging from a horseshoe kidney to kidney hypo/dysplasia, to vesicoureteral reflux.Although ID3 did not exhibit classical CAKUT features, it is worth noting that the lower urethra develops from the same cloacal endoderm-derived tissue as most parts of the bladder.This is in contrast to the mesoderm-driven metanephros, which likely explains the presence of a shortened urethra and penoscrotal hypospadias. 46Disruption of the interaction between endoderm and mesoderm during urogenital development or deficient ureter budding might result in a defective connection of the ureter to the bladder leading to vesicoureteral reflux. 47mmunohistochemistry targeting PHIP in human embryonic kidneys indicated high levels of DCAF14 in the nephrogenic zone of the cortex, which has high proliferative activity (Figure 3).This finding supports the regulatory function of DCAF14 in sustaining genome integrity of newly synthesized DNA in tissues with high mitotic activity. 40urthermore, PHIP shows high expression in tubular progenitor structures (Figures 3 and 4), aligning with the E3 ubiquitin ligases regulated expression of tubular transporters, such as Na þ /K þ / 2Cl À cotransporter or thiazide-sensitive sodium chloride cotransporter. 48,49ll reported missense variants and the half of the truncating variants are located N-terminal or within the 8 b-propeller-forming WD40 repeat domains, which mediate binding of ubiquitin ligase component CUL4 to chromatin before DNA replication.N-terminal or active domain changes result in CUL4/DDB1 ubiquitin ligase dysfunction and substrate accumulation.This leads to genomic instability with double-strand breaks and dysregulation of the cell cycle. 423,25 In summary, this study reports on patients with CHUJANS and concomitant CAKUT carrying deleterious PHIP variants.Based on high embryonic expression shown on both RNA and protein level, a role of PHIP during embryonic kidney development seems plausible, however we cannot demonstrate causality.This assumption supports the function of PHIP as an important regulator of the cell cycle and replication, especially during organogenesis.However, CAKUT does not occur in most patients with CHU-JANS, suggesting variable expressivity that may involve additional genetic or epigenetic factors triggering kidney development impairment.
Conclusively, for patients who were undiagnosed with syndromic CAKUT, which includes neurodevelopmental delay or intellectual disability and obesity, PHIP defects should be considered as part of the differential diagnosis.Conversely, patients diagnosed with CHUJANS should undergo assessment for kidney and urogenital anomalies.
One of the major challenges in complex congenital anomaly syndromes lies in comprehensive phenotyping across various medical disciplines.Because phenotypic characterization often prioritizes the initially affected organ system, there is an urgent need for more comprehensive and systematic organ assessment.Adequate description of multiorgan involvement will be facilitated by digital phenotyping tools that enable systematic and structured data capture in the future. 50

Figure 1 .
Figure 1.Selected clinical images and pedigrees.(a-e) Illustrates index patient (ID1) aged 42 years: (a) Stature and facial appearance, (b) pedigree, indicating de novo status, (c) MRI of the kidneys with significant narrowing of the parenchyma and enlarged renal calyces, probably expression of chronic urinary stasis grade III-IV (*).Cortical kidney cyst left up to 2 cm in diameter (blue arrows), ultrasound imaging of the left (c) and the right (e) kidney.(f-k) illustrates ID 2.1: 99mTc-DMSA scintigraphy of the kidneys (f), ultrasound imaging of the left (g) and right (h) kidney with unilateral kidney hypoplasia, (i-j) syndactyly, (k) and pedigree of the family indicates autosomal inheritance.

Figure 5 .
Figure 5. RNAScope in situ hybridization and quantification of Phip expression of postnatal mouse kidneys at different developmental times (a-d) 3 days postnatal, (e-h) 7 days postnatal, (i-l) 14 days postnatal, (m-o) adult (11 weeks), (p) all stages-bar graph summary stastics.Wt1 (green) was used as a marker for glomeruli.Dapi (blue) was used for visualization of nuclei.Dotted squares represent the area from which images c, g, k, and n were acquired respectively.Scale bars represent 200 mm.

Table 1 .
Patient characteristics