Recessive variants in the intergenic NOS1AP-C1orf226 locus cause monogenic kidney disease responsive to anti-proteinuric treatment

Buerger, Florian; Salmanullah, Daanya; Liang, Lorrin; Gauntner, Victoria; Krueger, Kavita; Qi, Jiansong; Normand, Josee; Sharma, Vineeta; Ranga, Arathi; Rubin, Alexander; Ball, David; Hong, Sunwoo; Lemberg, Katharina; Saida, Ken; Merz, Lea Maria; Sever, Sanja; Issac, Biju; Ma, Qianyi; Sun, Liang; Billing, Anja M.; Demir, Fatih; Rinschen, Markus M.; Reusch, Björn; Beck, Bodo B.; Guerrero-Castillo, Sergio; Gomez, Alexis C.; McNulty, Michelle T.; Sampson, Matthew G.; Al-Hamed, Mohamed H.; Saleh, Mohammed M.; Shalaby, Mohamed A.; Kari, Jameela A.; Fawcett, James P.; Hildebrandt, Friedhelm; Majmundar, Amar J.

doi:10.1038/s41467-025-65663-6

Download PDF

Article
Open access
Published: 27 November 2025

Recessive variants in the intergenic NOS1AP-C1orf226 locus cause monogenic kidney disease responsive to anti-proteinuric treatment

Nature Communications volume 16, Article number: 10654 (2025) Cite this article

2562 Accesses
1 Citations
22 Altmetric
Metrics details

Subjects

Abstract

In genetic disease, an accurate expression landscape of disease genes and faithful animal models can facilitate genetic diagnoses and therapeutic advances respectively. Previously, we found that variants in NOS1AP, the gene that encodes nitric oxide synthase 1 adaptor protein, cause monogenic nephrotic syndrome. Here, we determine that an intergenic splice product of NOS1AP/Nos1ap and neighboring C1orf226/Gm7694, which prevents NOS1AP from binding to nitric oxide synthase 1, is the predominant isoform in mammalian kidney transcriptional and proteomic data. Gm7694^−/− mice, whose allele exclusively disrupts the intergenic product, develop nephrotic syndrome phenotypes. In two male human subjects with nephrotic syndrome, we identify causative NOS1AP splice variants, including one predicted to abrogate intergenic splicing but initially misclassified as benign based on the canonical transcript. Finally, by modifying genetic background, we generate a faithful mouse model of NOS1AP-associated monogenic nephrotic syndrome that responds to anti-proteinuric treatment.

Congenital and infantile nephrotic syndrome: genotype-phenotype associations

Article 03 December 2025

A guide to gene–disease relationships in nephrology

Article 23 October 2024

Multi-population genome-wide association study implicates immune and non-immune factors in pediatric steroid-sensitive nephrotic syndrome

Article Open access 29 April 2023

Introduction

Nephrotic syndrome (NS) is a leading cause of chronic kidney disease in children¹. NS presents with edema, hypoalbuminemia, and proteinuria, arising from disruption of the glomerular filtration barrier comprised primarily of interconnected podocytes². Treatment-resistant NS (steroid-resistant NS or SRNS) frequently progresses to end-stage kidney disease concomitant with podocyte loss^1,2,3. Mendelian genetic forms represent the most severe subset of NS^4,5,6,7.

Monogenic NS genes are predominantly expressed in glomerular podocytes⁸, encoding proteins essential for podocyte development or homeostasis^2,9. Patient variants in disease proteins impair podocyte structure and function^2,8, causing podocytopathies. For example, variants in >10 genes encoding actin cytoskeleton regulators have been shown to cause nephrotic syndrome in humans and mice^{10,11,12,13,14,15,16,17,18,19}. Despite the interconnected nature of these genetic etiologies, specific treatments are lacking.

Causative genetic variants are detected in 11–30% of SRNS cases^4,5,6,7. For families lacking a genetic diagnosis, variant classification can be hindered by transcript annotation based on non-kidney tissues. Understanding kidney-specific isoform expression will facilitate the discovery of additional disease variants and genes, broadening our mechanistic understanding of NS and impacting clinical care for kidney disease patients with an established genetic diagnosis.

We previously discovered that recessive variants in NOS1AP/Nos1ap, affecting N-terminal domains, caused early-onset NS in human subjects and a podocytopathy in C57BL/6-Nos1ap^Ex3-/Ex3- mice²⁰. This locus, which is in synteny between humans and mice, encodes nitric oxide synthase 1 adaptor protein (NOS1AP) (Figs. S1–2)^21,22,23. NOS1AP contains an N-terminal phosphotyrosine binding (PTB) domain and a C-terminal PDZ binding domain (PDZ-BD), through which it interacts with NOS1 (Figs. S1–4)^{24,25,26,27,28,29}. While NOS1AP is most highly expressed in the central nervous system in humans and rats (Fig. S3)^26,27,29, querying of kidney scRNA sequencing data revealed it is highly expressed in podocyte clusters in the kidney^20,30,31. We determined that NOS1AP co-localized with actin-based filopodia in podocytes and with podocyte marker nephrin in kidney sections. NS patient variants, which were all within the PTB domain, impaired filopodia formation and podocyte migration²⁰. NOS1AP acted upstream of the NS-associated CDC42/formin pathway, supporting its key role in the actin regulatory network²⁰.

While the C-terminal NOS1AP-NOS1 interaction is important for neuronal NO signaling and required for hippocampal neuron-mediated anxiolysis^32,33, impairing NOS1 function did not disrupt NOS1AP-dependent actin remodeling in podocytes in our study²⁰, suggesting an NOS1-independent role for NOS1AP in NS.

We previously described an additional intergenic Nos1ap transcript in rat tissues generated from splicing between the 5’ region of canonical exon 10 and two coding exons of the neighboring open reading frame LOC100361087 (C1orf226 in humans and Gm7694 in mice) that was most highly expressed in olfactory bulb tissue^26,27,29. This alternative transcript encodes a distinct C-terminal domain lacking the PDZ-BD that mediates NOS1 interactions^26,27,28,29. The human and murine NOS1AP/Nos1ap orthologs are, similarly, in a common synteny block with C1orf226/Gm7694 (Fig. S1) and encode multiple independent transcripts with varying tissue-specific expression (Figs. S2 and S3). It is intriguing that deletion of Nos1ap in mouse cardiac tissue resulted in downregulation of Gm7694 transcripts³⁴, suggesting co-regulation of these neighboring loci in mice. However, it remains unclear if contiguous intergenic NOS1AP-C1orf226 transcripts are expressed in these species and if the intergenic isoform plays a role in kidney physiology and disease.

In the current study, we aimed to further delineate the role of NOS1AP in podocyte disorders by (i) dissecting the biologically relevant Nos1ap isoforms using transcriptional, protein, and genetic studies, (ii) faithfully modeling NOS1AP-associated podocytopathies using multiple mouse alleles and backgrounds, and (iii) validating treatments of this podocytopathy. We first characterized an intergenic splice isoform of NOS1AP/Nos1ap and neighboring locus C1orf226/Gm7694 by multiple transcriptional and protein approaches in cells and kidney tissue from humans and rodents. The intergenic form predominates over the canonical transcript in adult kidney tissue. We next hypothesized that additional variants in NOS1AP/Nos1ap cause NS and expand the genotype-phenotype relationship between this locus and kidney disease. Gm7694^−/− mice, predicted to exclusively disrupt the intergenic NOS1AP splice product, developed a podocytopathy. These findings were mirrored in Nos1ap^Ex4-/Ex4- mice with an early out-of-frame deletion and consistent with our published Nos1ap^Ex3-/Ex3- in-frame deletion mouse model. Supporting the role of the NOS1AP C-terminus in humans, we identified two additional recessive variants in NOS1AP, predicted to impair splicing of the penultimate coding exon and the intergenic product, in two cases of pediatric-onset NS. Collectively, this suggests that variants, which impact more C-terminal regions and splice isoforms of NOS1AP/Nos1ap, lead to kidney disease. To evaluate the impact of genetic modifying factors, the Nos1ap^Ex3- allele was bred from a C57BL/6 background onto the FVB/N background. FVB/N-Nos1ap^Ex3-/Ex3- mice exhibited 10-fold higher albuminuria than C57BL/6-Nos1ap^Ex3-/Ex3- mice and developed kidney dysfunction, supporting the relevance of genetic background for Nos1ap-variant associated disease. Finally, FVB/N-Nos1ap^Ex3-/Ex3- mice were treated with Renin-Angiotensin-Aldosterone-System (RAAS) inhibitor lisinopril, which reduced albuminuria and prevented premature death. Overall, our findings demonstrate that variants in the intergenic NOS1AP-C1orf226 locus cause a podocytopathy responsive to pharmacologic treatment.

Results

Intergenic splice products of NOS1AP/Nos1ap are predominant in human and mouse kidneys

We previously described an intergenic splice product of the rat ortholog Nos1ap across multiple tissues, with the highest levels in olfactory bulb tissue and low expression in brain cortex and hippocampus^26,27,29. This alternative transcript encodes a distinct C-terminal domain lacking the PDZ-BD that mediates experimentally confirmed NOS1 interactions^26,27,28,29 (Fig. 1A; S2 and S4). We, here, sought to further study this splice isoform in mammalian kidneys, as (i) its product may mediate distinct protein interaction partners and biological roles than the canonical protein, and (ii) interpretation of genetic variants in individuals could be dramatically altered in light of tissue-specific isoforms. In mouse and human tissues, we used (a) qualitative targeted RT-PCR and long-read RNA-sequencing data to assess the existence of contiguous intergenic NOS1AP/Nos1ap isoforms, (b) quantitative RT-PCR to compare expression of canonical and intergenic isoforms between tissues, and (c) bulk RNA-sequencing data to compare the expression of canonical and intergenic isoforms within tissues.

**Fig. 1: Detection/Identification of alternative *NOS1AP/Nos1ap* transcripts in mice and humans/ murine and human kidney tissue.**

First, by qualitative RT-PCR of total RNA from 8-week-old mouse C57BL/6 kidneys, both contiguous canonical (blue only) and intergenic Nos1ap transcripts (blue and green) were detected (Fig. 1B, C; S5A–D). The intergenic splice product arises from joining 153 nucleotides from exon 10 of Nos1ap to the two coding exons of predicted gene Gm7694 (Fig. 1A–C; S2–4).

To quantify the relative abundance of these isoforms by qPCR, three different Nos1ap amplicons were evaluated: 1. Nos1ap exon 2 to 5 present in both canonical and intergenic isoforms; 2. Nos1ap exon 9 to late exon 10 present in canonical isoforms only; 3. Nos1ap exon 9 to Gm7694 coding exon 1 (“exon 11” of the intergenic transcript) exclusively represents intergenic isoforms (Fig. S5E). Consistent with studies in human and rat tissue^26,27,29 (Fig. S3), expression levels of all Nos1ap isoforms (exon 2–5) were relatively higher in brain compared to kidney tissue (median 2.4-fold) (Fig. 1D). Interestingly, this expression level difference was more pronounced for canonical transcripts (exon 9-10) in brain compared to kidney tissue (median 32.3-fold) (Fig. 1D). In contrast, evaluation of the intergenic isoforms (exon 9–11) showed higher levels in kidney compared to brain tissue (median 9.3-fold) (Fig. 1D). Of note, qPCR studies in kidney tissue from neonates, adult 2-month-old mice, and aged 11–12 month old mice did not detect any significant changes in canonical or intergenic transcript levels across age (Fig. S5E, F).

Next, bulk RNA-sequencing data from four adult 8-week-old C57BL/6 mouse kidney was interrogated (Data S1). Sashimi plots were generated to visualize Nos1ap splice junctions and demonstrate intergenic splicing between Nos1ap exon 10 and the first coding exon of Gm7694 (Fig. S6). The intergenic splice junction in Nos1ap exon 10 was further evaluated to determine the ratio of intergenic to canonical reads by their alignment to the canonical or intergenic transcript within a 5 bp sliding window around the splice junction (GRCm38/mm10 chr1:170,318,737-170,318,738) (S7A–C). Intergenic reads were more prevalent relative to canonical reads across the four pooled datasets (94.5%) (Fig. 1E), suggesting intergenic isoforms predominate in adult mouse kidney.

In an independent dataset from neonatal to aged C57BL/6 mouse kidneys (Data S1), sashimi plots of the pooled data, similarly, demonstrated intergenic splicing (Fig. S6) although low read-depth limited quantitative analyses from this dataset (Fig. S7D).

We next evaluated intergenic splicing in the human NOS1AP locus. Qualitative RT-PCR studies, similarly, observed contiguous intergenic transcripts at the identical orthogonal splice junction (GRCh37 chr1:162336994-162336995) in adult human kidney tissue (Fig. 1F, G). This splicing, again, links the first 153 nucleotides of NOS1AP exon 10 to the first coding exon of C1orf226 (NM_001085375) (Fig. 1G; S2, S3). This observation was confirmed in human brain and kidney cell lines (SH-SY5Y, immortalized podocytes, HEK cells) (Fig. S8B–F). As a side note, inclusion of an additional 15 nucleotides in NOS1AP exon 4 (encoding LLLLQ), which we previously found in rat adult brain²⁹, was detected in RT-PCR products from human cell lines but not a human kidney tissue sample (Fig. S8A, S8F).

To examine contiguous NOS1AP transcripts in adult kidney or other tissues (age of acquisition 40 to >90 years), long-read sequencing from the ENCODE project was interrogated and demonstrated contiguous intergenic transcripts in adult heart (n = 11), brain (n = 9), colon (n = 2) and kidney (n = 1) specimens (Fig. 1H; S9, Data S2) to varying degrees relative to contiguous canonical transcripts. For example, the canonical NOS1AP transcript predominated across brain tissues (96%), while the ratio of canonical to intergenic transcript was more variable in heart tissues (48.4%) (Fig. 1H; S9).

As low read count in the single kidney sample from long-read ENCODE data only allowed for a qualitative analysis, we next performed a quantitative evaluation of short-read RNA-sequencing data from (i) human kidney samples in the NEPTUNE cohort (living donor biopsy, nephrectomy, and NS patient biopsy) covering 244 samples of individuals from 2 to >90 years (ii) human fetal kidney samples from ENCODE (20- and 24-weeks of gestation). Sashimi plots of the NOS1AP locus generated from these datasets qualitatively confirmed the presence of intergenic splicing between NOS1AP exon 10 and the first coding exon of C1orf226 (NM_001085375) (Fig. S10A, B and S11). Quantifying reads that crossed the human splice junction, we found that qualifying reads corresponding to the intergenic splice product predominated relative to canonical reads (69.6–93.8%) in all sample groups within RNA isolated from either glomerular or tubular compartments of pediatric and adult human kidney samples (Fig. 1I, S10C–E). Similarly, the intergenic transcript predominated over the canonical in fetal kidney samples (86.20–90%) (Fig. 1J; S11).

Overall, the findings indicated that intergenic NOS1AP/Nos1ap transcripts predominate over canonical isoforms in human and mouse kidney.

Proteins corresponding to intergenic splice products of NOS1AP/Nos1ap were identified in cells and kidney tissue from humans and rodents

We next assessed whether the detected transcripts are successfully translated into detectable protein products. For this, we used immunoprecipitation (IP), immunoblotting, immunofluorescence staining, and proteomics in immortalized cell lines, primary cells, tissues from humans and rodents.

While immunoblotting for NOS1AP in mouse brain tissue lysates has been previously successful³⁴, we have found that immunoblotting with multiple commercially available antibodies could not successfully detect NOS1AP isoforms in kidney cell and tissue lysates. We, therefore, generated multiple custom antibodies against rat NOS1AP (Fig. 2A). As these also did not sensitively detect Nos1ap isoforms in non-neuronal cells and tissues from humans and mice by immunoblotting, we used these antibodies to perform immunoprecipitation (IP) of Nos1ap. We first interrogated lysates of primary mouse embryonic fibroblasts (MEFs) from wild-type and transgenic Nos1ap mouse models to identify select isoforms. Performing the IP and immunoblotting of the eluate with an antibody that detects C-terminal regions of both canonical and intergenic Nos1ap isoforms (“2093”), two prominent protein bands were detected in the wildtype (WT) eluate. The upper 95–100 kDa species was consistent with the intergenic isoform (I), and a lower 65–70 kDa species was at the expected size of the canonical isoform (Fig. 2B; S12A). In eluates from homozygous Nos1ap^Ex4-/Ex4- MEFs, whose allele is predicted to cause early truncation of all Nos1ap isoforms (Fig. S14), both protein bands were absent as expected (Fig. 2B; S12A). In contrast, in the homozygous Gm7694^−/− MEF eluates, whose allele would selectively lead to loss of intergenic isoforms (Fig. S14), only the upper intergenic band (I) was absent, supporting this band is comprised of the intergenic isoform. The lower band (C + L) persisted, suggesting this residual band contains only canonical protein in these Gm7694^−/− knockout eluates and canonical protein must contribute to the lower band in wildtype MEF eluates (Fig. 2B; S12A).

**Fig. 2: Detection of canonical and intergenic NOS1AP/Nos1ap protein isoforms in cells and tissue from humans and rodents.**

Next, the IP in MEFs was performed with 2093, but immunoblotting was conducted with an antibody specific to the intergenic isoforms (“GST Long”) (Fig. 2A). We observed a prominent upper 95–100 kDa band, consistent with the intergenic isoform (I), and a 65–70 kDa lower band (L) in WT MEF eluates while both were absent in Gm7694^−/− MEF eluates (Fig. 2C; S12B). This supports that the upper 95–100 kDa species reflects the full-length intergenic isoform, while the lower 65–70 kDa species in MEFs is more complex and contains both the canonical isoform and, additionally, a protein of similar molecular weight derived from the intergenic locus.

Subsequently, the IP and immunoblotting were performed in MEFs with GST Long (Fig. 2D), again revealing a strong 95–100 kDa band in WT MEF and a lower band at 65–70 kDa in WT MEFs. Both were absent in Nos1ap^Ex4-/Ex4- and Gm7694^−/− MEF eluates (Fig. 2D). These observations were confirmed by performing the IP and immunoblotting with another antibody specific to intergenic Nos1ap (“PPIT”) (Fig. S12C). These findings, similarly, support that the upper 95–100 kDa species is consistent with the full-length intergenic isoform and that protein derived from the intergenic locus also contributes to the 65–70 kDa species. This lower species was previously observed in rat lysates as well and is similar in molecular weight to the canonical isoform²⁹. We interpret this lower band to potentially represent either protein generated from an additional alternative intergenic splice product (e.g., lacking exons prior to or after exon 4) or a cleavage product of the full-length intergenic isoform (e.g., with cleavage occurring before the region encoded by exon 4).

We next employed these IP approaches in other kidney cell lines. Immunoprecipitation and immunoblotting with the pan-NOS1AP antibody (2093) detected the 95–100 kDa intergenic band and the 65–70 kDa band in lysates from both human embryonic kidney (HEK293T) and human immortalized podocyte (HsPd) cell lines (Fig. S12D). Immunoprecipitation with the same pan-NOS1AP antibody and immunoblotting with a NOS1AP intergenic-specific antibody (GSTLong) consistently identified the 95–100 kDa intergenic band in both HEK293T and HsPd cell lysates (Fig. S12E). Using only the NOS1AP intergenic-specific antibody for IP and immunoblotting revealed the same band pattern in HEK293T and HsPd cell lysates (Fig. S12F), and similar observations were made in HEK293T cell lysates using a second intergenic-specific antibody (PPIT) (Fig. S12C). Overall, this indicated that the multiple NOS1AP isoforms, including the full-length intergenic isoform, are present in human kidney cells. We, additionally, evaluated these species within MCF7 human breast cancer cells, where NOS1AP has a reported role³⁵, and observed multiple NOS1AP isoforms as well (Fig. 12D–F).

Given the lack of sensitivity of NOS1AP-specific antibodies in western blotting, we evaluated a commercial antibody raised against C1ORF226 for use in immunoblotting. Immunoblotting alone (without prior IP) of WT MEF cell lysates showed a prominent 95–100 kDa band, consistent with the intergenic isoform, that was absent in both Nos1ap^Ex4-/Ex4- and Gm7694^−/− MEF lysates (Fig. 2E). In HEK293T, HsPd and MFC7 cell lysates, a 95–100 kDa band, consistent with the intergenic isoform, was similarly observed (Fig. 2E; S12G). Overall, these protein-level findings provided qualitative evidence of Nos1ap intergenic protein in mouse fibroblasts and multiple human kidney cell lines.

To assess kidney tissue for the existence of the intergenic protein, we, next, performed Co-IP using two different antibodies: 1) The intergenic-specific GST Long antibody and 2) R300, an antibody, that was raised against a C-terminal fragment of canonical Nos1ap isoform (~ 199 aa) which is partially present in the intergenic Nos1ap isoform (111 aa) (Fig. 2A, F, G). Using R300 for Co-IP and immunoblotting, an upper 95–100 kDa intergenic band and a lower band at 65–70 kDa were detected in heterozygous Nos1ap^Ex4-/+ eluates but were absent in Nos1ap^Ex4-/Ex4- eluates (Fig. 2F). The lower band may represent a combination of canonical and intergenic locus-derived protein as observed in Fig. 2B and lower band seen in Fig. 2C, D. Using GST Long for Co-IP and immunoblotting, an upper 95–100 kDa intergenic band and a lower band at 65–70 kDa were detected in WT and heterozygous Gm7694^+/- eluates but not in homozygous Gm7694^−/− MEF eluates (Fig. 2G). Overall, this supports that the intergenic protein and, potentially, the canonical isoform are qualitatively present in mouse kidney tissue.

To evaluate the localization of the intergenic Nos1ap isoform within kidney tissue, we next used the GST Long antibody to perform immunofluorescence staining and confocal microscopy imaging in rat kidney sections. Co-staining with an antibody against slit diaphragm marker protein nephrin revealed substantial co-localization within podocytes of rat kidney glomeruli (Fig. 2H).

In a last step, we sought to assess whether peptides mapping to regions specific to the different NOS1AP/Nos1ap isoforms can be identified using proteomics approaches. For this, we both re-analyzed publicly available datasets and performs additional proteomics studies using mass spectrometry. We first re-analyzed proteomics data from FACS-sorted primary adult mouse glomerular cells³⁶. After standard trypsin digestion and mass spectrometry, 21 unique peptides specific to either Nos1ap or Gm7694 were identified (Fig. 2I row 4, 2J; Data S3). Interestingly, no peptides aligning to the unique C-terminus of canonical Nos1ap were identified, even though theoretical trypsin cleavage sites could result in at least five detectable peptides of 7 to 25 amino acids in length (Note S2). Similarly, no peptides spanning the junction from Nos1ap to Gm7694 could be detected either (Note S2). However, this was not expected as the only predicted digestion product, including the Nos1ap-Gm7694 junction site, has a peptide length of 54 amino acids and would thus be precluded from detection by the employed mass spectrometry method (Note S2). Since proteomics for adult mouse kidney glomeruli was conducted after separation of podocytes from other glomerular cells, we could further perform semi-quantitative comparison of protein abundance for Nos1ap, Gm7694, and other glomerular marker proteins. In three independent mouse samples, Nos1ap and Gm7694 were predominantly detected in podocytes, similar to podocytic markers nephrin, podocin, synaptopodin, and WT1 (Fig. 2J).In contrast to mouse kidney glomeruli, proteomics data from adult mouse brain revealed peptides mapping throughout canonical Nos1ap but not Gm7694 (Fig. 2I row 5; Data S4)³⁷.

To complete these studies in adult mouse tissue, we performed liquid chromatography-mass spectrometry and analysis for proteome and phosphoproteome in the kidneys of newborn mice (n = 5). This analysis essentially confirmed findings from adult mouse glomeruli with 16 peptides mapped to the region of Nos1ap common to both isoforms, 5 peptides mapped to Gm7694, and none mapped to the canonical-specific region in Nos1ap (Fig. 2I row 6).

Finally, we re-analyzed proteomics data from HEK293 cells with exceptionally high coverage of protein species. This analysis revealed peptides mapping to both NOS1AP prior to the canonical-specific region as well as to C1ORF226, and interestingly, also included a peptide spanning the intergenic transition from NOS1AP to C1ORF226. Of note, this junctional peptide resulted from a cleavage event consistent with trypsin/P specificity. This supported the existence of an intergenic protein species containing this specific peptide. In contrast, no peptides mapping to the specific C-terminal region of canonical NOS1AP were detected.

These proteomics data, in sum, provided orthogonal evidence that the intergenic isoform is present in kidney cells and tissues. Intriguingly, unlike brain tissue, no canonical isoform-specific peptides were detected in proteomic data from kidney cells or tissue, which mirrors the differential transcript expression we observed in mouse tissues (Fig. 1D).

It remained unclear if this intergenic protein was functional. In our previous study, we had shown that NOS1AP induces filopodia formation in podocytes in vitro, a function impaired by patient variants²⁰. To determine whether the intergenic isoform has a similar effect, we expressed GFP-labeled constructs for both isoforms in immortalized human podocytes and performed live cell imaging (Fig. S13A, B). We employed automated quantification of cell circularity and convexity (quotient of total over convex cell area), as these metrics would decrease as filopodia-forming cells become less round and less convex. Interestingly, expression of both tagged isoforms caused significantly reduced circularity and convexity when compared to GFP-only expressing cells, indicating both isoforms induce filopodia (Fig. S13B, D).

Taken together, these observations indicate that the intergenic NOS1AP/Nos1ap isoform is a functional protein, which is present in kidney cells and tissues across humans and rodents, with expected localization to the podocyte.

Variants impacting the Nos1ap-Gm7694 locus cause murine podocytopathy

We previously described the kidney phenotype of Nos1ap^Ex3-/Ex3- mice on a C57BL/6 background²⁰. These mice bear a homozygous deletion of Nos1ap exon 3, causing an in-frame deletion within the N-terminal PTB domain (Fig. 2A; S14A–D). Nos1ap^Ex3-/Ex3- mice developed moderate albuminuria (1-3 g albumin/g creatinine) accompanied by podocyte foot process effacement at an ultrastructural level. Here, we characterized two additional Nos1ap mouse models to (i) evaluate the biological relevance of the intergenic product and (ii) further establish the association between this locus and NS.

We first studied a mouse model affecting both canonical and intergenic transcripts. Nos1ap^Ex4-/Ex4- mice bear a homozygous out-of-frame deletion in Nos1ap exon 4, predicted to cause early truncation within the PTB domain (c.271_329del; p.91_110delfs*7) (Fig. 3A; S14A–D), thereby preventing successful translation of both canonical and intergenic Nos1ap isoforms (Fig. 2B, D–F; S12).

**Fig. 3: *Gm7694*^−/− and *Nos1ap*^Ex4-/Ex4- mice develop podocytopathies.**

Albuminuria was assessed in Nos1ap^Ex4-/Ex4- mice between 4–12 months of age. Urine albumin-to-creatinine ratios (ACR) were significantly elevated in homozygous mice relative to wildtype controls at ages 7–9 months and 10–12 months (median ACR 1227 mg/g creatinine versus 492 mg/g in controls at 10–12 months) (Fig. 3B). These differences were observed for both males and female homozygotes at 10–12 months (S15A, B). Analysis of urine samples from 10 to 12-month-old mice by SDS-PAGE and Coomassie staining revealed albuminuria that was significantly elevated above the basal physiologic levels observed in control mice using band densitometry and normalization to urine creatinine across all homozygotes, males and females (Fig. 3C, D; S15C–F). Kidneys from approximately 10 to 12-month-old mice were evaluated for ultrastructural changes at the glomerular filtration barrier by electron microscopy. Nos1ap^Ex4-/Ex4- mouse kidney sections exhibited reduced podocyte foot process density (median 1.36 foot processes/µm versus 2.34 in controls) and glomerular basement membrane thickening (median 335 nm versus 257 nm in controls) (Fig. 3E–G). Overall, these observations supported that this recessive variant causes a murine podocytopathy, consistent with our published observations in Nos1ap^Ex3-/Ex3- mice²⁰.

Next, we investigated an allele that exclusively affects the intergenic splice isoform but not the canonical Nos1ap transcript. Specifically, we generated Gm7694^−/− mice, which bear a complete deletion of predicted gene Gm7694 (Fig. 3H; S14A–D) and prevent the successful translation of the intergenic while not affecting the canonical isoform of Nos1ap (Fig. 2C–F, G).

Albuminuria was assessed, showing significantly increased ACR levels in Gm7694^−/− mice relative to controls between 4–6 months and 10–12 months (Fig. 3I) (median ACR 647 mg/g creatinine in Gm7694^−/− mice versus 397 mg/g in controls at 10–12 months). These differences were observed for both male homozygotes, but not female homozygotes, at 10–12 months (S15A, B). Analysis of urine samples from 10 to 12-month-old mice by SDS-PAGE and Coomassie staining similarly demonstrated albuminuria, which showed modest but significantly elevated above basal levels observed in control mice by densitometry and normalization to urine creatinine across all homozygotes (Fig. 3J, K, S15A, B) as well as when separately comparing male (Fig. S15G, H) and female animals (Fig. S15I, J). Kidneys from approximately 10- to 12-month-old mice were evaluated for ultrastructural changes by electron microscopy. Similar to Nos1ap^Ex4-/Ex4- mice, Gm7694^−/− mouse kidney sections exhibited reduced podocyte foot process density (median 1.47 foot processes/µm versus 2.16 in controls) and glomerular basement membrane thickening (median 300 nm versus 234 nm in controls) (Fig. 3L–N). These findings suggest that Gm7694 and the intergenic Nos1ap splice product play an important role in murine podocyte homeostasis and disease.

C-terminal recessive variants in NOS1AP are associated with human SRNS

We previously discovered that two recessive variants in NOS1AP cause very early-onset NS in humans²⁰. These variants were predicted to impact the N-terminal PTB domain of NOS1AP (Fig. 4A–C and Table 1). To discover additional disease-causing recessive variants in NOS1AP or the adjacent C1orf226 loci in humans, exome data generated after completion of our initial study were interrogated from an additional 985 families with NS.

**Fig. 4: Recessive C-terminal variants in *NOS1AP* associated with pediatric SRNS.**

Table 1 Recessive mutations in NOS1AP (NM_014697) in families with nephrotic syndrome

Full size table

This revealed a rare, homozygous splice site variant (NM_014697:c.1105+5 G > C) in NOS1AP in family B4606 (Table 1; Fig. 4A–D). No competing variants were detected in an additional 83 nephrotic syndrome disease genes or phenocopy genes (e.g., monogenic nephritis disease genes) (Data S5). B4606_21 is a male subject, born from a consanguineous union. He developed edema at age 3 years. Consistent with the clinical diagnosis of nephrotic syndrome, his initial laboratory evaluation revealed microscopic hematuria, nephrotic-range proteinuria (19 g protein/day), hypoalbuminemia (2.2 g/dL), reduced serum total protein levels (4 g/dL), and elevated serum triglyceride levels (9.98 mmol/L). His proteinuria was resistant to corticosteroids but partially responsive to the calcineurin inhibitor cyclosporine. Sanger sequencing confirmed the variant was present homozygously in the proband and heterozygously in unaffected parents and siblings, consistent with a recessive mode of inheritance (Fig. 4D). The variant was predicted to strongly impair splicing of the penultimate exon 9 by four independent algorithms (Table 1 and Fig. 4A–C). Skipping of exon 9 is predicted to create an out-of-frame deletion and loss of NOS1AP or NOS1AP-C1orf226 C-terminal domains. Furthermore, the variant was absent from control genome databases ExAC and gnomAD (Table 1). To further assess how this variant affects splicing, we performed a midigene assay³⁸. For this, we performed PCR amplification on genomic DNA of the subject and a healthy parent to generate amplicons containing exon 9 and adjacent intronic sequence (~ 500 bp) with the wildtype sequence or the single nucleotide variant (c.1105+5 G > C). Each amplicon was then subcloned into the pCI vector in between Rho exon 3 and 5, and the resulting constructs were then expressed in immortalized human podocytes to perform RT-PCR on extracted RNA with confirmation of product identity by Sanger sequencing. Transfection with the empty vector plasmid yielded a PCR product consistent with a transcript joining exon 3 and 5 (Fig. 4E, H), while the vector containing the wildtype NOS1AP exon 9 sequence yielded products including an additional 166 bp from exon 9 (Fig. 4E, G). Interestingly, expression of the vector containing the splice variant in family B4606 yielded a shorter than the expected band size (Fig. 4E). When interrogating exon 9 sequence, an alternative splice site (score 0.54) 40 bp upstream of annotated splice site (score 0.85) was found³⁹ (Fig. 4F). Sanger sequencing confirmed the alternative splice junction in transcript generated by the vector containing the subject’s variant (Fig. 4G). The resulting variant is predicted to cause a frameshift and an early truncation after 21 amino acids (p.Gln357Glyfs*21). We, thus, concluded the recessive NOS1AP variant identified in family B4606 is pathogenic by ACMG criteria⁴⁰ and the likely cause of SRNS in this subject (Table 1), highlighting the importance of the NOS1AP C-terminus in human podocyte homeostasis.

In addition, we analyzed all NOS1AP or C1orf226 variants in the ClinVar genomic variation database⁴¹ for deleteriousness under the hypotheses that (i) recessive variants in these loci are associated with nephrotic syndrome and (ii) variants were misclassified based on the canonical transcript alone. Focusing on non-structural variants (< 50 bp in size), we observed 67 reported variants, of which 14 variants were deemed deleterious based on rare prevalence and predicted impact on protein coding or splice effect. Of these 14 variants, we previously published two (Fig. 4A–C and Table 1)²⁰. We were unable to obtain further clinical genetic data for three variants. 8 additional variants were found in heterozygous states. A final variant was observed in a subject (VCV001333195), where, upon request, detailed genotype and phenotype data were available. Exome sequencing had been performed, identifying a homozygous variant in NOS1AP exon 10 that was absent from the gnomAD database (Table 1). This single-nucleotide variant was predicted to cause a missense in the canonical product (NM_014697: c.1259 G > C; p.G420A) with weak in silico prediction scores and, therefore, was not deemed deleterious. However, in the context of the alternative NOS1AP-C1orf226 transcript, the variant alters the +1 position of the intergenic splice site within NOS1AP exon 10 (c.1258+1 G > C). Multiple prediction scores strongly indicate impaired splicing and thus deleteriousness. Of note, no competing variants were detected, including in >71 genes associated with Mendelian genetic nephrotic syndrome and/or nephritis⁷. Interestingly, reverse phenotyping of the subject revealed the most severe NS phenotype, congenital nephrotic syndrome, which had progressed to end-stage kidney disease at age 3 years, requiring dialysis. The subject had passed away at the age of 5 years. He also had a unilateral cystic kidney and a healthy appearing contralateral kidney—anomalies that are unlikely to be sufficient for his early-onset kidney failure. Collectively, this data supports that this essential splice site variant is pathogenic by ACMG criteria (Table 1), supporting that intergenic splicing of NOS1AP and C1orf226 is critical in human podocyte homeostasis and corroborates our findings in mouse model studies (Fig. 3).

Recessive Nos1ap variants cause severe albuminuria and kidney dysfunction on FVB/N genetic background

Our current and published data established that bi-allelic variants in the NOS1AP-C1orf226 locus, affecting N-terminal or C-terminal regions of the encoded protein, cause a severe podocytopathy in humans (Table 1 and Fig. 4). In contrast, Nos1ap-deficient mice exhibited a more moderate phenotype. Specifically, we previously showed that Nos1ap^Ex3-/Ex3- mice on a C57BL/6 background developed albuminuria and ultrastructural changes in the glomerular filtration barrier, but not persistent hypoalbuminemia, kidney dysfunction, or reduced survival²⁰—features frequently observed in human SRNS. The additional mouse models described here (Nos1ap^Ex4-/Ex4- and Gm7694^−/− mice) similarly had modest albuminuria and normal survival into late adulthood on a C57BL/6 background (Fig. 3). This suggested that additional loci may modify the phenotype caused by Nos1ap deficiency.

Because genetic background can modify the severity of murine podocytopathies^{42,43,44,45,46,47,48,49,50,51}, we posited that the kidney phenotype in C57BL/6-Nos1ap^Ex3-/Ex3- could be similarly altered by genetic background to enable more faithful modeling of human SRNS and the establishment of an optimal model for evaluating potential therapies. Therefore, C57BL/6-Nos1ap^Ex3-/+ mice were bred with FVB/N or 129/sv mice for more than six generations. This would yield Nos1ap^Ex3-/+ mice with a genetic background reflecting >98% of the outcrossed strain (Fig. 5A). Of note, transcriptional levels of all long and intergenic Nos1ap isoforms were not significantly different in adult kidney samples across C57BL/6, FVB/N, and 129/sv adult mice, while the canonical Nos1ap isoforms was significantly higher expressed in C57BL/6 compared to FVB/N mice (3-fold) (Fig. S5G). Resulting FVB/N-Nos1ap^Ex3-/+ or 129/sv-Nos1ap^Ex3-/+ were in-crossed to yield wildtype, heterozygote, and homozygote mice of the same background.

**Fig. 5: FVB/N-*Nos1ap*^Ex3-/Ex3- mice develop early-onset progressive proteinuric kidney disease.**

We previously established that the Nos1ap^Ex3- variant causes an in-frame deletion at the transcriptional level, but wanted to confirm this occurs at the protein level using newer reagents not previously available at the time of our initial study. For immunoblotting, we used the pan-NOS1AP 2093 antibody and SDS-PAGE on a high percentage (15% Bis-Tris) gel to assess neonatal brain lysates from wild-type, heterozygous, and homozygous Nos1ap^Ex3- mice, as has been described in global Nos1ap knockout mice³⁴. In WT and heterozygous animals, a ~ 60–65 kDa band was detected, which was absent in homozygotes, and consistent with the wildtype protein. A lower band at ~55-60 kDa was detected in heterozygous and homozygous animals, which was absent in WT animals (Fig. S16A). This likely indicates that a stable, shorter Nos1ap protein is produced by the in-frame deletion Nos1ap^Ex3- allele.

FVB/N-Nos1ap^Ex3-/Ex3- mice developed albuminuria at weaning age with a median ACR of 1.9 g/g, which increased to 13.8-15.7 g/g at 4–6 months of life (Fig. 5B) and consistently for both male and female animals (Fig. S17A, B). In contrast, ACR levels in 129/sv-Nos1ap^Ex3-/Ex3- mice were more modestly elevated relative to wildtype or heterozygote littermate controls (1–3 g/g during the first six months of life) (Fig. 5C; S16B). Overall, FVB/N-Nos1ap^Ex3-/Ex3- mice exhibited markedly elevated albuminuria relative to 129/sv-Nos1ap^Ex3-/Ex3- and C57BL/6-Nos1ap^Ex3-/Ex3- mice between 3–6 months of life (Fig. 5C), indicating the FVB/N background uniquely modifies this kidney phenotype.

FVB/N-Nos1ap^Ex3-/Ex3- mice were further analyzed for serum markers of nephrotic syndrome. In association with marked albuminuria, male and female FVB/N homozygous mice developed significantly reduced serum albumin levels in serial measurements at 3–6 months of life relative to littermate controls (median albumin levels 3.0–3.5 g/dL versus 3.7-4.1 g/dL) (Fig. 5D; S17C). Serum BUN levels, a marker of declining kidney function, at 3–6 months were significantly elevated in FVB/N-Nos1ap^Ex3-/Ex3- mice relative to littermate controls (median BUN levels 30-37 mg/dL versus 20-25 mg/dL), more frequently in males than females (Fig. 5E; S16C and S17D). Serum creatinine was not significantly changed (Fig. S16D), but it is not a sensitive marker of renal function in mice⁵². In correlation with reduced kidney function, homozygotes also exhibited reduced survival between 6–8 months of life relative to control mice (Fig. 5F). Thus, in contrast to C57BL/6-Nos1ap^Ex3-/Ex3- mice²⁰, FVB/N-Nos1ap^Ex3-/Ex3- mice develop persistent hypoalbuminemia, kidney dysfunction, and increased mortality, faithfully recapitulating features of human SRNS.

FVB/N-Nos1ap ^Ex3-/Ex3- mice exhibit histologic and ultrastructural features of a severe podocytopathy

Given the urinary and serum abnormalities in FVB/N-Nos1ap^Ex3-/Ex3- mice, we hypothesized that homozygous mice develop glomerular changes consistent with a podocytopathy. Kidney tissue sections from 6–8 months old mice were Periodic Acid Schiff (PAS) stained and analyzed by light microscopy. Measurement of PAS-positive matrix within glomerular tufts revealed significantly increased mesangial matrix deposition, indicative of chronic glomerular injury, in FVB/N-Nos1ap^Ex3-/Ex3- mice relative to heterozygote controls (Fig. 6A). Kidney tissue sections from FVB/N homozygotes, furthermore, showed tubular dilation, reminiscent of human congenital nephrotic syndrome⁵³, in contrast to heterozygote controls (Fig. 6B).

**Fig. 6: FVB/N-*Nos1ap*^Ex3-/Ex3- mice develop glomerular matrix deposition, tubular dilation, and ultrastructural changes to the glomerular filtration barrier.**

Kidney sections were also examined by electron microscopy at weaning age (3–4 weeks of life) to assess early ultrastructural changes when albuminuria was first detected in FVB/N-Nos1ap^Ex3-/Ex3- mice (Fig. 5B). Homozygotes exhibited podocyte foot process effacement and thickening of the glomerular basement membrane (GBM) (Fig. 6C), relative to heterozygote glomeruli exhibiting appropriate rhythmicity of foot process formation and GBM thickness. We, then, also assessed ultrastructure in weanling C57BL/6-Nos1ap^Ex3-/Ex3- mice, who did not yet show markedly elevated albuminuria at this early age. Interestingly, C57BL/6 homozygotes exhibited reduced foot process density, but no thickening of the GBM was observed compared to heterozygous control animals (Fig. S18).

Finally, to determine if these phenotypic differences between genetic backgrounds may be secondary to transcriptional differences in isoform expression, we assessed canonical and intergenic Nos1ap transcript levels in kidney tissue from C57BL/6-Nos1ap^Ex3-/Ex3- and FVB/N-Nos1ap^Ex3-/Ex3- mice but found no significant differences (Fig. S5H).

Overall, FVB/N-Nos1ap^Ex3-/Ex3- mice exhibited biochemical and structural kidney abnormalities consistent with human SRNS and modified by genetic background.

FVB/N-Nos1ap ^Ex3-/Ex3- mice respond to RAAS blockade

Having established a more faithful model of NOS1AP-associated podocytopathy in FVB/N- Nos1ap^Ex3-/Ex3- mice, we next sought to evaluate the impact of potential therapies given the lack of effective options in SRNS. RAAS blockade is widely employed in acquired proteinuric diseases to reduce proteinuria and preserve renal function^54,55,56. In children with genetic or primary SRNS/FSGS, RAAS blockade similarly reduces proteinuria^{57,58,59,60,61}. However, it remains unclear whether this therapeutic approach can preserve GFR or promote survival in monogenic glomerular disorders outside of COL4A3/4/5-variant associated Alport syndrome^62,63. This includes those caused by variants in podocyte-specific actin regulatory genes such as NOS1AP^{10,19,20,64,65,66,67}. Therefore, we evaluated the impact of RAAS blockade in Nos1ap^Ex3-/Ex3- mice using an ACE inhibitor, lisinopril.

FVB/N-Nos1ap^Ex3-/Ex3- mice were treated at weaning age with oral lisinopril. To assess the impact on proteinuria, urine ACRs were measured every two weeks for 14 weeks of treatment (Fig. 7A). FVB/N-Nos1ap^Ex3-/Ex3- mice treated with 100 mg/L and 200 mg/L lisinopril exhibited significantly reduced albuminuria relative to vehicle-treated animals (Fig. 7A), indicating RAAS ameliorates the proteinuric phenotype in this model.

**Fig. 7: Anti-proteinuric treatment reduces albuminuria and prevents mortality in FVB/N-*Nos1ap*^Ex3-/Ex3- mice.**

Serum markers were measured at early (4–8 weeks of treatment) and late (12–16 weeks of treatment) time-points. Serum albumin and total protein were significantly elevated in homozygous mice treated with 100 mg/L lisinopril relative to those receiving vehicle at the early timepoint (Fig. S19A, B), correlating with the improvement in albuminuria (Fig. 7A). FVB/N-Nos1ap^Ex3-/Ex3- mice treated with 100 mg/L also showed reduced BUN levels relative to vehicle-treated animals at the late timepoint (Fig. S19C), indicating kidney dysfunction was blunted by this ACE inhibitor dose. In contrast, FVB/N-Nos1ap^Ex3-/Ex3- mice treated with 200 mg/L did not have significantly different serum albumin, total protein, or BUN levels than vehicle-treated mice (Fig. S19).

Survival of FVB/N-Nos1ap^Ex3-/Ex3- mice was monitored over 140 days of treatment. While only 20% (1/5) of vehicle-treated mice survived this time frame, all (11/11) of the homozygotes treated with lisinopril survived (Fig. 7B). These results indicate that RAAS blockade using lisinopril not only reduces proteinuria but also prevents mortality in a mouse model of human SRNS caused by defective actin remodeling.

FVB/N-Nos1ap ^Ex3-/Ex3- mice do not respond to Dynamin Activator Bis-T-23

Dynamin stimulates actin bundling and, like NOS1AP, promotes filopodia formation^{20,26,68,69,70}. Dynamin activating compound Bis-T-23 reduced albuminuria and normalized podocyte ultrastructural defects in multiple murine podocytopathies⁷¹. We hypothesized that Bis-T-23 would, similarly, improve proteinuria in FVB/N-Nos1ap^Ex3-/Ex3- mice. However, we did not observe altered albuminuria with daily Bis-T-23 treatment in 4-week-old FVB/N-Nos1ap^Ex3-/Ex3- mice, including when collecting urine 1–3 h after treatment to assess for a transient reduction in albuminuria (Fig. S20).

Discussion

In this study, we aimed to further delineate the role of the NS disease gene NOS1AP by (i) dissecting its biologically relevant isoforms using multi-omic approaches, (ii) establishing additional mouse models that strengthen the association between this locus and podocytopathies and faithfully recapitulate human disease, and (iii) validating treatment options for this podocytopathy.

In summary, we demonstrate that variants impacting the NOS1AP-C1orf226/Nos1ap-Gm7694 locus can cause podocytopathy in humans and mice (Figs. 1–4), supporting an important role for the, here described, NOS1AP intergenic splice product in kidney disease. We, furthermore, demonstrate that genetic background of Nos1ap mouse models can modify the severity of this genetic podocytopathy (Figs. 5, 6), which resulted in the generation of a more faithful model of human NS and suggests that variants in additional loci can modify the nephrotic syndrome trait in mammals. Lastly, we provide evidence that in a genetic mouse model of NOS1AP-associated podocytopathy RAAS blockade is effective in reducing proteinuria and preventing early mortality in this genetic form of nephrotic syndrome (Fig. 7).

The NOS1AP alleles in humans and mice identified in this study (Figs. 3, 4, S14; Table 1) not only expand the genotype-phenotype correlation between this locus and NS but also underscore the impact of C-terminal domains and alternative splice isoforms of NOS1AP/Nos1ap in monogenic nephrotic syndrome. We present in vitro experiments on morphological changes of podocytes caused by both isoforms, suggesting they may play similar roles in regulating the cytoskeleton. However, much more detailed analyses are warranted to determine whether canonical and intergenic isoforms of NOS1AP play redundant or independent roles in podocyte biology. Future studies will have to functionally compare both isoforms, including assessing whether the different C-termini of these isoforms engage distinct interaction partners (e.g. through comparative interactome analyses). It will also be important to contrast these interaction partners with those disrupted by N-terminal PTB domain NS variants, which we previously determined impair NOS1AP-dependent actin remodeling and podocyte homeostasis in humans and mice²⁰.

Our previous studies indicate that NOS1AP-dependent actin remodeling in podocytes is independent of its neuronal binding partner NOS1²⁰. It remains possible that the canonical NOS1AP protein may mediate its actin remodeling effects through a key interaction partner other than NOS1. In addition, our current findings lead us to posit that NOS1AP may regulate podocyte homeostasis through interactions mediated by the altered C-terminal domain created through the intergenic splice product. In any case, the previously experimentally established NOS1 interaction domain resides within the C-terminal part of NOS1AP that is only included in the canonical but not the intergenic isoform. Future studies should, therefore, compare isoform-specific interaction partners and determine if an in-frame deletion allele, only affecting the NOS1-binding domain, causes a podocytopathy in mice.

Human subject VCV001333195 received CLIA-certified genetic testing to explain his diagnosis of congenital nephrotic syndrome. However, due to interpretation based on the canonical transcript, the identified homozygous variant in NOS1AP was misclassified as benign. This case underscores the importance of tissue- and condition-specific isoform expression for correct variant interpretation, especially as recent studies indicate that >10 alternative transcripts exist for human coding genes while only an average of 4 transcripts per gene are annotated^72,73,74.

Transcriptional studies suggest predominance of an intergenic over the canonical transcript in human and mouse kidneys of different ages (Figs. 1 and 2). One limitation of our protein analyses was the lack of a sensitive antibody to detect all NOS1AP isoforms and compare their relative abundance in kidney lysates by immunoblotting. The proteomics strategies used were also technically limited in their capacity to detect junctional peptides specific to the intergenic long isoforms. However, it should be noted that qualitative proteomics studies detected canonical-specific peptides only in mouse brain tissue datasets, while intergenic-specific regions were only identified in human kidney cell and mouse kidney tissue datasets (Fig. 2I, J). It will be important in future studies to evaluate high coverage proteomics data from human glomerular samples to confirm the presence of these protein isoforms in native human tissue and, especially, those studies employing alternative protocols to standard trypsin digestion (such as chymotrypsin) to increase the likelihood of detecting peptides at the NOS1AP-C1orf226 protein junction. It also remains unclear whether potential additional short NOS1AP-C1orf226 transcripts, not distinguished by short-read RNA-sequencing data, play an important role in disease.

Genetic background is, next, established to modify the phenotype of Nos1ap^Ex3-/Ex3- mice, thereby yielding a more faithful model of steroid-resistant NS observed in human subjects and platform for evaluating potential therapies (Figs. 5, 6; S16–18). In fact, the FVB/N-Nos1ap^Ex3-/Ex3- mice develop profound albuminuria, kidney dysfunction, and histologic kidney damage reminiscent of human congenital nephrotic syndrome caused by genetic variants in NPHS1⁵³. These mice, furthermore, develop ultrastructural damage to the glomerular filtration barrier at an early time point not observed in C57BL/6 homozygotes. The increased susceptibility of the FVB/N mouse strain to glomerular injury in genetic and acquired models of glomerular disease has been previously established^{42,43,44,45,46,47,48,49,50,51}. Genome-wide association studies (GWAS) have been performed but revealed distinctly associated loci depending on the kidney disease model, none including the Nos1ap locus on chromosome 1^42,47,48,49. It will be important, in future studies, to identify loci associated explicitly with modification of the Nos1ap^Ex3-/Ex3- albuminuria and kidney dysfunction traits (e.g., through array-based approaches), as these genetic factors could reveal modifying factors that mitigate or aggravate the actin dysregulation caused by Nos1ap deficiency. We would also want to consider whether genetic background has a similar impact on the other murine alleles evaluated in this study.

Although this locus has not been implicated in GWAS of mouse glomerular disease phenotypes, we did consider whether genetic background in wildtype and homozygous mutant mice alters Nos1ap transcript isoform expression specifically. The only significant difference was a 3-fold lower expression of the canonical isoform in wildtype FVB/NJ mouse kidney relative to wildtype C57BL/6 J mouse kidney, while there was no difference in this isoform between wildtype 129svJ and FVB/NJ kidney tissue (Fig. S5G). It is unclear how this difference would explain the lower albuminuria observed in both C57BL/6 J and 129svJ homozygous mice relative to FVB/NJ mice (Fig. 5B). Moreover, neither canonical nor intergenic isoform expression was significantly different in FVB/NJ and C57BL/6 J Nos1ap^Ex3-/Ex3- mice (Fig. S5H), suggesting that differential Nos1ap transcript isoform expression cannot explain the differential kidney traits in these mice.

Genetic background may also explain differences in the survival phenotypes of Nos1ap^Ex4-/Ex4- mice. A previous study reported that global Nos1ap^Ex4-/Ex4- mice on a pure C57BL/6 J background exhibited lethality during fetal development³⁴. In contrast, homozygotes in our study were bred on a mixed C57BL/6J-6N background and were viable through adulthood, consistent with reported data from the International Mouse Phenotyping Consortium (IMPC)^75,76. This, similarly, supports that genetic background can influence Nos1ap-associated phenotypes.

Treatment options are very limited in genetic forms of NS. RAAS blockade is widely employed in acquired proteinuric diseases, providing a reduction of proteinuria and preservation of GFR^54,55,56. Some case reports and a recent meta-analysis report proteinuria reduction from RAAS blockade in children with genetic or primary SRNS/FSGS^{57,58,59,60,61}. However, strong evidence for clinical or experimental benefits in terms of preservation of renal function and reduction in mortality has not been demonstrated for any form of monogenic proteinuric disease other than COL4A3/4/5-variant associated Alport syndrome^62,63. Our findings, that lisinopril ameliorates Nos1ap-associated podocytopathy in mice, suggest that this benefit extends to other Mendelian genetic pathways and should be considered in other forms caused by variants in podocyte-specific actin regulatory genes^{10,19,20,64,65,66,67} and assessed in a systematic manner.

On the other hand, Bis-T-23 did not alter the proteinuria in FVB/N-Nos1ap^Ex3-/Ex3- mice despite the related roles that NOS1AP and dynamin proteins play in actin-based filopodia formation^{20,26,68,69,70}. It should be noted that there were, in fact, only modest transient effects of Bis-T-23 on proteinuria in a mouse model of ACTN4-associated nephropathy⁷¹, indicating that enhancing Dynamin oligomerization may not be sufficient to restore normal podocyte physiology in some forms of genetic NS.

Overall, these findings (i) further establish an essential role of NOS1AP for podocyte biology and podocytopathies, (ii) provide evidence for pharmaceutical treatment options in genetic podocytopathies, and (iii) underline the importance of understanding tissue-specific isoform expression for interpreting genetic testing results.

Methods

Ethics: Study design, conduct, approval, and Human Genetics Study Subject recruitment

The study design and conduct complied with all relevant regulations regarding the use of human study participants and was conducted in accordance with the criteria set by the Declaration of Helsinki as well as the SAGER guidelines.

Specifically, blood samples, pedigree information, and clinical data for the Hildebrandt laboratory study were collected from April 1998 to December 2022 for 908 families (43% female, 57% male). The study was approved by the Institutional Review Board of Boston Children’s Hospital (IRB-P00006200), including subject B4606. Pediatric nephrologists made SRNS diagnoses based on clinical and histologic criteria. The Hildebrandt laboratory obtained blood samples and pedigrees following informed consent from individuals with pediatric-onset NS (age 0-25 years) and/or their legal parents/guardians, irrespective of subject sex, gender, or country of origin. Additionally, siblings were recruited for relevant families to enable segregation of variants. The diagnosis of NS was based on clinical features of nephrotic-range proteinuria, hypoalbuminemia, and edema, and supported by kidney biopsy findings evaluated by renal pathologists. Clinical data were obtained using a standardized questionnaire (http://www.renalgenes.org). This includes biological sex information based on subject- or legal guardian-reporting and clinician confirmation. Through this process, subject B4606_21 and family members were recruited.

The Al-Hamed laboratory recruited subjects with informed consent for genetic studies under IRB-approved research protocols RAC#2050045 and 2160022 at King Faisal Specialist Hospital and Research Center (KFSH&RC), including VCV001333195. This included obtaining clinical history and a blood sample for genetic analyses.

ES and variant calling

For all recruited subjects in the Hildebrandt laboratory, exome sequencing (ES) and variant calling were performed in an established multi-step approach^7,77 to discover a genetic cause of NS using Agilent SureSelect™ human exome capture arrays (Thermo Fisher Scientific) with next-generation sequencing (NGS) on an Illumina™ platform. Sequence reads were mapped against the human reference genome (NCBI build 37/hg19) using CLC Genomics Workbench (version 6.5.1) (CLC bio). Genetic location information is according to the February 2009 Human Genome Browser data, hg19 assembly (http://www.genome.ucsc.edu). Downstream processing of aligned bam files was done using Picard and samtools, and single-nucleotide variant calling was done using GATK5. Identification of reads mapping to either specific X- or Y-chromosomal regions is used to confirm biological sex. The variants included were rare in the population, with a mean allele frequency <1% in dbSNP147 and with only 0–1 homozygotes in the adult genome database gnomAD. Additionally, variants were non-synonymous and/or located within splice-sites. Subsequently, variant severity was stratified based on protein impact (truncating frameshift or nonsense variants, essential or extended splice-site variants, and missense variants). Splice-site variants were assessed by in silico tools MaxEnt, NNSPLICE, SpliceSite Finder, and GeneSplicer splice-site variant prediction scores, as well as conservation across human splice-sites. Missense variants were assessed based on SIFT, MutationTaster, PolyPhen 2.0, and CADD conservation prediction scores and evolutionary conservation based on manually derived multiple sequence alignments.

ClinVar database query

The ClinVar Database⁴¹ was queried for variants in NOS1AP and C1orf226 that were not large structural variants. These variants were analyzed as above for rare population prevalence (<1% allele frequency) and predicted deleterious effects on protein coding (truncating variant or missense variant with 2+ strong in silico prediction scores) and/or splicing (2+ strong in silico prediction scores). Variant submissions that were predicted to be deleterious were further evaluated by contacting the designated submitter for additional genetic information (zygosity, method of detection, and other detected variants) and clinical information. Subject VCV001333195, who was recruited by the Al-Hamed laboratory, had CLIA exome sequencing performed (3Billion, South Korea) with confirmed enrichment for known NS-associated disease genes.

Mouse breeding and maintenance for Nos1ap ^Ex3-/Ex3- mice

The animal experimental protocols were reviewed and approved by the Institutional Animal Care and Use Committee at the Boston Children’s Hospital (#18-12-3826 R and #00001802). All mice were handled in accordance with the Guidelines for the Care and Use of Laboratory Animals. Mice were housed under pathogen-free conditions with a light period from 7:00 AM to 7:00 PM and had ad libitum access to water and rodent chow. Nos1ap^Ex3-/Ex3- mice were a kind gift from Dr. Norihiro Kato from the National Center for Global Health and Medicine in Japan⁷⁸. Genotyping was performed by multiplex PCR using the following primers: #1:CTTTGTCTTCTGCTTCGCC, #2:ACACTACCATTTGGTCTCC, #3:TCAAGACCGACCTGTCC, and #4:CAATAGCAGCCAGTCCC.

Mouse generation and maintenance of Nos1ap ^Ex4-/E4- and Gm7694 ^−/−

The animal experimental protocols were reviewed and approved by (Dalhousie Animal Care Protocol #21-120 and #20-116). All mice were handled in accordance with the Canadian Council on Animal Care (CCAC).

Nos1ap^Ex4-/E4- were generated as follows (Fig. S21A). Mice were ordered and generated at Baylor using the KOMP Tm1a heterozygous knockout mice for gene Nos1ap (C57BL/6 N Nos1aptm1a(KOMP) Wtsi/BCM) (MGI ID 1917979, Knockout-First - Reporter Tagged Insertion (Promotor Driven Cassette)). These mice were generated and backcrossed into the C57BL/6 N line for three generations. The Knockout-First Reporter Tagged mice NOS1AP^+/tm1a mice were then crossed with CMV-Cre mice from Jax labs (B6.CTg(CMV-Cre)1 Cgn/J; Stock #006054)), that had been backcrossed into the C57BL/6 J for 10 generations, to yield animals with complete excision of the neo cassette and exon 4 but a residual LacZ reporter gene in the locus (Nos1ap^+/Ex4-). PCR amplification of isolated genomic DNA is used for genotyping using the following primers to detect the mutant allele (Nos1ap^Ex4-) 5’-CGGATAAACGGAACTGGAAA-3’ and 5’-TAATCACGACGCGCTGTATC-3’. For the wild type allele, the following primer sets are used: 5’-CAATTCATGGCAAGCAAAAC-3’ and 5’-ATTTCTCTTCCTCCGCACCT-3’. Mice were backcrossed into the C57BL/6J for two generations and maintained on a mixed C57BL/6 J/6 N background. They were in-crossed to generate wildtype, heterozygous, and homozygous mice.

Gm7694^−/− mice were generated by the Center for Mouse Genome Modification, University of Connecticut Health Center (Fig. S21B). Briefly, loxP sites were inserted on each flank of the DNA (BAC clone RP24-421J12) encoding exons 1 and 2 of Gm7694. This vector, containing approximately 4.7kb and 3.2kb of 5′-long and 3′-short arms, respectively, was then linearized by NotI digestion, purified, and then electroporated into ES cells, which were derived from an F1 blastocyst (mixed 129/sv-C57BL/6J background). ES cells were cultured in the presence of G418 and Ganciclovir after electroporation. Drug-resistant colonies were screened using primers specific to genomic sequences outside the homology arms and LoxP sites to identify targeted ES clones. Targeted clones were expanded and screened again to confirm their identity prior to the generation of chimeric animals by aggregation with a CD1 morula. Chimeric males were then bred with ROSA26-Flpe female mice, which had been backcrossed with C57BL/6 J for over 30 generations, to remove the PGKneo cassette to generate the final Gm7694 floxed allele. This generated two lines; the 2C8 line was picked as it gave larger litters. The 2C8 line was then crossed with a Hprt1-Cre female (Jax Stock No. 004302; RRID: IMSR_JAX:004302) that had been backcrossed with B6J for over 30 generations. Mice were then genotyped using PCR using the following primers: FgtF: 5’-ATACGGGCCCTCTCTTAAC:3’; FgtR: 5’-ACTGACCCAGCAAACCAACT-3’; LgtF: 5’-GCCCCTCCTAATTCCAAGTG-3’ to show efficient excision of exons 1 and 2 of Gm7694. Mutant mice were then backcrossed for two generations onto the C57BL/6J. They were in-crossed to generate wildtype, heterozygous, and homozygous mice.

Mouse embryonic fibroblasts were generated from E13.5 embryos as outlined⁷⁹. To immortalize the lines, cells were transfected with SV40 cDNA (addgene plasmid #21826, a gift from David Ron (SV40 1:pBSSVD2005), and cells were split at low density (1/40) for 5 passages.

Urine analysis

Urine was collected by placing mice in collection cages with ad libitum access to water overnight (16 h). Upon collection, samples were immediately frozen and stored at −80 °C and only thawed on ice prior to urine albumin and creatinine measurements once. Urinary albumin levels were determined using the Albumin Blue Fluorescent Assay Kit (Active Motif) in combinations with standard dilutions prepared from mouse serum albumin (Equitech Bio Inc.). Urine creatinine was measured using QuantiChrom™ Creatinine Assay Kit (BioAssay Systems). In-gel protein detection was performed with Coomassie blue dye and quantified using BioRad Image Lab Analysis software.

Whole blood analysis

200 μL of blood was collected once a month via the facial vein bleeding method and collected in lithium heparin tubes. Blood samples were then immediately analyzed with the Vetscan® VS2 Chemistry Analyzer using Comprehensive Diagnostic Profile rotors to measure serum albumin, BUN, and creatinine levels.

Histological analysis

The kidney tissues were fixed in 4% paraformaldehyde (PFA), sectioned (5 μm thickness), and stained with periodic acid-Schiff (PAS) following standard protocols for histological examination. For mesangial matrix deposition, 25 images (at 40x magnification) per animal were obtained to detect glomerular tufts and stalks. Quantitative analysis of PAS-stained sections was performed in a semiautomated manner using an ImageJ macro. Briefly, glomeruli were outlined manually and cropped. Color splitting was performed to yield images of component colors that reflected the PAS-positive extracellular matrix and nuclei or nuclei alone. Thresholding and particle analysis were performed to determine the percent area in these channels to calculate the percent glomerular area comprised of PAS-positive matrix. For the tubular area, 5 images (at 20x magnification) of the corticomedullary junction distributed across a longitudinal section of a kidney were obtained for each animal. Quantitative analysis of the tubular lumen area was determined using a semiautomated ImageJ macro. Quantification was performed in a blinded manner. Thresholding was performed on images to detect the open tubular lumen, and a particular analysis was performed to calculate the percentage of positive area.

Ultrastructural analysis

Kidney tissues were perfused-fixed with 2.5% glutaraldehyde, 1.25% PFA in 0.1 M sodium cacodylate buffer (pH 7.4) and then fixed in 2.5% glutaraldehyde, 1.25% PFA, and 0.03% picric acid in 0.1 M sodium cacodylate buffer (pH 7.4) overnight at 4 °C. Samples were then washed with 0.1 M phosphate buffer, post-fixed with 1% OsO4 dissolved in 0.1 M phosphate-buffered saline (PBS) for 2 h, dehydrated in an ascending, gradual series (50‒100%) of ethanol, and infiltrated with propylene oxide. Samples were embedded using the Poly/Bed 812 kit (Polysciences) according to the manufacturer’s instructions. After pure fresh resin embedding and polymerization in a 65 °C oven (TD-700, DOSAKA, Japan) for 24 h, sections of approximately 200–250 nm thickness were cut and stained with toluidine blue for light microscopy. Sections of 70-nm thickness were double-stained with 6% uranyl acetate (EMS, 22400) for 20 min and lead citrate (Fisher) for 10 min for contrast staining. The sections were cut using Reichert Ultracut-S/LEICA EM UC-7 (Leica) with a diamond knife (Diatome) and transferred onto copper and nickel grids. Sections were evaluated by transmission electron microscopy (JEOL 1200EX) at an acceleration voltage of 80 kV. All steps, including image acquisition, were performed in a blinded manner by independent persons. For each animal, ~21 images of 3–4 glomeruli were acquired for assessment of podocyte foot process density and glomerular basement membrane thickness, with only perpendicularly cut capillary loops evaluated.

Mouse treatment studies

For lisinopril studies, mice were randomized into treatment groups with established concentrations of 100 or 200 mg/L of lisinopril (Sigma) in their drinking water⁸⁰. Lisinopril or vehicle (water) containing drinking water was changed three times weekly. Urine samples were collected every two weeks. Blood samples were collected at an early and a late time point in the study. Mice were monitored every 1-2 days for survival for >20 weeks of treatment.

For dynamin activating agent Bis-T-23 (Aberjona Labs), established doses at 20 or 30 mg/kg were administered daily to mice by intraperitoneal injection⁷¹. First, urine was collected daily from four-week-old FVB/N-Nos1ap^Ex3-/Ex3- mice receiving Bis-T-23 for seven days to detect a reduction in albuminuria prior to severe disease progression. In a second study, homozygotes were treated daily, and urine samples were collected immediately prior to injection and hourly for the subsequent 3 h to detect transient changes in albuminuria.

RT-PCR studies

C57BL/6 mice kidneys were isolated from 8-week-old male mice. RNA isolation was performed using RNAeasy Mini Kit (Qiagen). Total RNA was converted into cDNA using the TeloPrime Full-Length cDNA Amplification Kit V2 (Lexogen). cDNA was amplified using HotStarTaq DNA Polymerase and the following primers: Ex7_9_F, CGAGGTGTGACTGATCTGGA; Ex7_GEx2_F, GTGTGACTGATCTGGATGCC; Ex7_GEx2_R, GTGTTCTTGTGTATTAGGTCCGG; Ex8_Gex2_F, TCCACTCACCACCAGATGC; Ex8_Gex2_R, CCTCCTGGGTGTTCTTGTGT; Ex9_GEx2_R, TAGATTGCTTAGGACGGGCT; Ex2-3_Gex_F:, CGCAGAATCCGGTATGAGTT; Ex3_Gex_F, TCTCTGTGGACGGTGTCAAG; Ex4_Gex_F, CTGGTGATGCAGGACCCTAT; Ex7-9-F_Ex10UTR-R1, ACCCTGGGTGTTGTTCTCAG; Ex8-10-F_Ex10UTR-R2, ACCACCTTCACCAGCTCCTC. Human kidney total RNA was purchased from Takara. The sample was isolated from a post-mortem kidney sample from a 40-year-old adult female who died from sudden death unrelated to kidney disease. RNA was reverse transcribed using the iScript cDNA Synthesis Kit (Bio-Rad). cDNA was amplified using HotStarTaq DNA Polymerase and the following primers: NEX1UTRtoATG_F: CGGGTAACCATGCCTAGCAAAAC; FL_NOS1AP_R2: ACCTACACGGCGATCTCATCATC; FL_FUSION_NOS_C1ORF_R1: CTATTCAAAGGACAGCAGGTCTG. PCR products were analyzed by agarose gel electrophoresis and subsequent Sanger sequencing. For human cell line studies, the following primers were employed: Hs_Ex2-4_F2 CAACAGCAGGGTGGAGATCG, Hs_Can9-10_R3 GCGGAAGAAAGCGAAAGCAC, Hs_2-4_F3 CAGCAGGGTGGAGATCGTG, Hs_Inter9-10_R1 GGAGTGAGGGCTGTGTTCAAAT.

For quantitative RT-PCR studies in mouse tissue, total cellular RNA was extracted from either cell lines or homogenized mouse tissue samples using the Qiagen RNeasy® Plus kit following the manufacturer’s instructions. Total RNA was quantified, and a total of 1 µg per sample was subjected to RT-PCR using Takara Bio PrimeScript RT Kit. Gene expression was assessed using Biorad iTaq Universal SYBR Green Supermix, Applied Biosystems StepOne Real-Time PCR Systems and the following primers: (i) Canonical (Nos1ap Exon 9 to Nos1ap late Exon 10) F GCACATCTCTCTGCTGGTCA and R TGGGTATCCTCAGGTGGGAG; (ii) Intergenic (Nos1ap Exon 9 to Gm7694 Exon 1) F GCACATCTCTCTGCTGGTCA and R CTGGAGCTTTGGAGTGAGGG; (iii) All Nos1ap Long isoforms (Nos1ap exon 2 to Exon 5) F CGCAGAATCCGGTATGAGTT and R TACATCTGAAGATATTGCTGGCAC; (iv) Beta-actin CATTGCTGACAGGATGCAGAAGG and TGCTGGAAGGTGGACAGTGAGG.

Mouse bulk RNA-sequencing (RNAseq) re-analysis

101 bp paired-end reads sequenced from wildtype C57BL/6J mouse kidney samples at 8 weeks of life were downloaded from dataset GSE145053 (ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE145053)⁸¹. 50 bp single-end reads sequenced from wildtype C57BL/6J mice kidney samples at age 0, 2, 4, 8, and 79 weeks were downloaded from Gene Expression Omnibus (GEO) dataset GSE225622⁸². All samples for each time point were pooled into a single dataset (Data S1). We used trimmomatic v0.39 to trim the low-quality next-generation sequencing (NGS) reads (-threads 20 ILLUMINACLIP:TruSeq3-PE.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36)⁸³. Subsequently, only the high-quality trimmed reads were aligned to the mouse reference genome using STAR v2.7.2b⁸⁴. All aligned reads spanning the splice site coordinates (GRCm38/mm10) chr1:170,318,738 on Nos1ap gene and chr1:170,302,838 on Gm7694 gene within (i) a 5 bp window with minimum 1 bp overlap on either side of the splice junction or (ii) a fixed 10 bp window (5 bp on both sides of the junction) were counted in IGV v2.15.2⁸⁵. Reads that did not show spliced alignment but spanned the splice junction on Nos1ap were counted as canonical reads, while reads that showed spliced alignment were counted as intergenic reads. A ratio of intergenic to total reads was computed, where the sum of all reads spanning the Nos1ap splice junction was taken as the total. To generate Sashimi plots for visualization, the bam files for each dataset were pooled and indexed using samtools v1.20. The aggregated bam file was then visualized on IGV v2.16.0 for the Nos1ap-Gm7694 locus at mm10 chr1:170,295,000-170,591,900. Both the full landscape of non-canonical Nos1ap transcripts and the zoomed-in view of the splice junctions between Nos1ap and Gm7694 were visualized in Sashimi plots.

Bulk long-read RNAseq re-analyses from human tissues (ENCODE)

To verify that the intergenic transcript is expressed in human tissues, we interrogated long-read sequencing from the ENCODE project⁸⁶. We downloaded 31 bam files from 1 kidney, 3 colon, 10 brain, and 17 heart adult tissues sequenced using Pacific Biosciences Sequel technology (sample details in DataS2) (as found on ENCODE, encodeproject.org for accessions ENCFF114VSO, ENCFF877QJZ, ENCFF378STM, ENCFF564ONS, ENCFF406GQU,ENCFF222UTL, ENCFF279ABL, ENCFF319FBW, ENCFF626QRV, ENCFF613SDS, ENCFF291EKY, ENCFF173JOL, ENCFF840OVC, ENCFF132YCF, ENCFF850YMOENCFF623IBV, ENCFF018PZX, ENCFF502LAB, ENCFF911RNV, ENCFF886FZQ, ENCFF171AQO). For Fetal kidney bulk RNA-sequencing data from weeks 20 and 24, bam files were obtained from ENCODE: ENCSR000AFA (GEO database, ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE78568) and aligned to GRCh38 v29 using STAR v2.5.1b. The bam files were then indexed using samtools v1.20 and visualized on IGV v2.16.0 for the NOS1AP-C1orf226 locus at GRCh38 v29 chr1:162,066,000-162,392,000. We visually inspected each bam file using the IGV browser⁸⁵. We first identified “long transcripts” which began before or within exon 3 of NOS1AP (the earliest exon where we found causative disease variants in humans or mice). We, then, quantified canonical transcripts, reads that extend into the UTR of NOS1AP exon 10, and intergenic transcripts that extend into the second coding exon of C1orf226. To visualize the long-read sequencing, we adapted the output from the alignment track in the IGV browser. For representative examples, we selected the bam file from each tissue with the most intergenic reads. For clarity of the figure and to emphasize the long reads, we filtered for reads aligned to the positive strand and hid base-pair mismatches.

Both the full landscape of non-canonical NOS1AP transcripts and the zoomed-in splice junctions were visualized in Sashimi plots. The number of non-canonical NOS1AP transcripts was visualized in the Sashimi plot and confirmed by manual inspection of the reads on IGV.

Bulk short-read RNAseq re-analyses from human kidney tissue (NEPTUNE)

To quantify the intergenic transcripts in human kidney tissues, we utilized micro-dissected glomerular and tubulointerstitial compartments from the NEPTUNE cohort, which included patients with nephrotic syndrome as well as healthy tissue from tumor nephrectomies and living donors⁸⁷ (neptune-study.org/ancillary-studies). Biopsies were prepared using the Clontech SMARTSeq v4 kit and sequenced using Illumina HiSeq 2500, resulting in 150 bp unstranded, paired-end reads. For each sample, individual reads were checked for adapter content, GC content, and per-base sequence quality using FastQC. FastQ Screen was used to detect contamination of foreign or disproportionate amounts of off-target RNA⁸⁸. Sequenced reads were then aligned to the human reference genome GRCh38 using STAR 2.5.2b, and mapped reads were inspected for distribution across introns, exons, UTR, and intergenic regions using picardtools⁸⁴ (http://broadinstitute.github.io/picard). We then used SAMtools to filter out duplicates, keep only properly paired reads, and filter out reads with low mapping quality (<20)⁸⁹. We filtered each bam file to the start of NOS1AP and end of C1orf22 with a 1 kb flank (CRCh37/hg19 chr1:162,024,660-162,360,673). The 150 bp reads were analyzed in exon 10 of NOS1AP that either splice to C1orf266 or extend through exon 10. To do this, we defined a 10 bp sequence for each of the canonical (CTTAGGTAGG) and intergenic (CTTAGTTGAC) transcripts, which ensured at least 5 matched nucleotides before and after the splice site. For each bam file, we tabulated both sequences. Reads that did not completely span either sequence were excluded. The code for automating these processes is provided (Software S1). Sashimi plots were, moreover, generated using the IGV browser.

Sample preparation, liquid chromatography-mass spectrometry, and analysis for proteome and phosphoproteome of newborn mouse kidneys

Kidneys of newborn mice (N = 5) were homogenized in a bullet blender (two steel beads per sample, 2 min at amplitude 8, at 4 °C) in 150 µL extraction buffer containing guanidinium hydrochloride (6 M), 0.1 M HEPES pH 7.4, 5 mM EDTA, supplemented with 1x Roche cOmplete protease inhibitors. Lysates were denatured at 95 °C for 5 min at 600 rpm and subsequently treated with Benzonase (25 U/µL Millipore #70664-10KUN) for 30 min at 37 °C. After clearing the extracts at 500xg at RT, protein concentration was determined by BCA assay. Proteins were reduced and alkylated at 5 mM TCEP and 20 mM CAA for 5 min at 95 °C. Protein clean-up was performed on SP3 beads using 1 µL beads per 20 µg protein input with protein binding for 10 min in 80% EtOH. Proteins bound on SP3 beads were washed 3x with 90% acetonitrile with 30 s separation on the magnet in between. Almost dried protein-bound SP3 beads were resuspended in 50 mM HEPES, pH 7.5, 2.5 mM CaCl2, and digested overnight with trypsin at 37 °C with a 1:100 protease:protein ratio. Peptides were collected and desalted by C18 RP Stage Tip clean up.

Phosphopeptides were enriched with titanium IMAC magnetic beads (MagReSyn© Ti-IMAC HP, BioResyn Biosciences) according to the manufacturer´s instructions, with 100 µg C18 RP Stage tip cleaned peptides as input. Dried peptides were resuspended in loading buffer (80% acetonitrile, 1 M glycolic acid, 5% TFA) and added to the equilibrated beads to allow binding at RT for 20 min on a rotator. Peptides were washed 1x with loading buffer, 2x with 80% acetonitrile/1% TFA, and 1x with 10% acetonitrile/0.2% TFA. Phosphopeptides were eluted twice with 1% ammonia solution and acidified with 10% TFA. After speedvac, concentration peptides were desalted on C18 RP Stage Tips.

Tryptic digests were analyzed on an UltiMate3000 RSLC system (Thermo Fisher Scientific) coupled to an Exploris 480 mass spectrometer, including the FAIMS pro interface (Thermo Fisher Scientific). A two-column set-up was implemented with a trap column (Acclaim™ PepMap™ 100 C18, 3 µm particle size, 2 cm × 75 µm, ThermoFisher Scientific) at a flowrate of 5 µL/min and an analytical column (Aurora Ultimate 25 cm × 75 µm, IonOpticks) at a flowrate of 400 nL/min using a 120 min gradient with mobile phase A as 0.1% formic acid in LC-MS-grade water and mobile phase B as 0.1% formic acid in LC-MS-grade acetonitrile. The peptide gradient was run for 5 min at 2% B, reaching 8% B at 10 min, followed by a linear gradient to 25% B over 80 min, followed by a linear gradient to 35% B over 10 min, at 101 min switched to 90% B and finished with a column equilibration step from 110–120 min at 2% B. All MS measurements were performed in data-independent acquisition (DIA) in positive polarity with FAIMS CV voltages of -45 and -60. MS1 spectra were recorded in profile mode at 120 K resolution within a scan range of 380–1500 m/z, with a normalized AGC target of 300%, RF lens of 40% and auto maximum injection time. MS2 spectra were recorded at 30 K for 400–1000 Da with non-overlapping isolation windows of 15 Da and normalized HCD collision energy of 28%.

MS RAW files were analyzed library-free with Spectronaut v.19 using directDIA with factory settings for the Phospho PTM workflow. Trypsin/P was selected as the digestion enzyme, allowing for two missed cleavages with minimal and maximal peptide length of 7 and 52 amino acids. Cysteine carbamylation was set as a fixed modification, STY-phosphorylation, methionine oxidation, and protein N-terminal acetylation as variable modifications. Data was searched against a Mus musculus database downloaded from UniProt (June 2023, 55275 entries). Peptide and protein identifications were accepted at a 1% q-value. Phosphopeptide site probability cut-off was at 0.75, with quantification at the MS2 level. Data was visualized with R in R Studio 3.6.0 with the ggplot21 package⁹⁰. Raw data is available at PRIDE (PXD058044).

Proteomics re-analysis of adult mouse kidney podocytes, adult mouse brain tissue, and HEK-293 cells

To assess if peptides corresponding to either canonical or intergenic Nos1ap isoforms are abundant in tissue and cells, raw proteomics data were re-analyzed^36,37,91. Proteomics raw data were downloaded from PRIDE or ProteomeCentral (PXD003306, ebi.ac.uk/pride/archive/projects/PXD003306; PXD004263, proteomecentral.proteomexchange.org/cgi/GetDataset?ID = PXD004263; and PXD004452, proteomecentral.proteomexchange.org/cgi/GetDataset?ID = PXD004452). Next, data were re-analyzed with MaxQuant (v.1.6.10)⁹² using the Uniprot SwissProt Mouse database (downloaded in April 2021), amended by the sequences of non-canonical intergenic Nos1ap, Gm7694, and the unique C-terminus of the canonical Nos1ap. In total, 23 peptides were identified.

cDNA cloning

cDNA clones were purchased from the following sources: human canonical NOS1AP (Harvard PlasmID Database, GenBank accession NP_055512 encoded by GenBank accession NM_014697) or GFP_N_NOS1AP_Intergeneic (custom synthesized by Genscript). Expression constructs (pCDNA6.2-N-GFP) were produced using LR Clonase (Invitrogen, Thermo Fisher Scientific) following the manufacturer’s instructions.

Nos1ap immunoprecipitation and/or Western blotting from mouse fibroblasts, mouse kidney lysates, mouse neonatal brain lysates, immortalized podocytes, or HEK cells

For western blotting and immunoprecipitation assays in mouse embryonic fibroblasts, immortalized human podocytes, MCF-7 cells, HEKs and neonatal kidney lysates, cells/tissue samples were lysed in NP40 lysis buffer (10% glycerol, 1% NP-40, 137 mM NaCL, 20 mM Tris [pH 8.0] containing 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 µg/mL aprotinin, and 10 µg/mL leupeptin. For brain lysates and HEK lysates, RIPA buffer (:25 mM Tris•HCl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) was used. For immunoprecipitation studies, lysates were then precipitated with one of multiple Nos1ap antibodies (see below)^26,29. Following an overnight incubation at 4 °C, Protein A Sepharose was added and incubated for 1 h at 4 °C. The precipitated proteins were then washed 3 times in NP40 lysis buffer before the addition of 2x Sample Buffer. Immunoprecipitation eluates and/or whole cell/tissue lysates were then resolved on SDS-PAGE gels, transferred to PVDF membrane (EMD, Millipore, Billerica, MA), and then blocked in 5% skim milk or 5% BSA in 1x TBST for 1 h. Membranes were then incubated in primary antibodies overnight at 4 °C. Blots were then washed 3 times in TBST, then incubated in TBST containing HRP-conjugated secondary at 1:10,000 for 1 h at RT. Membranes were then washed 3 times in TBST prior to signal detection using ECL reagent (BioRad). Blots were imaged using a ChemiDoc MP (BioRad). The following antibodies were used: R300 (WB dilution 1:1000, Santa Cruz sc-9138); custom antibody 2093 (WB dilution 1:1000); custom antibody GST Long (WB dilution 1:5000); custom antibody PPIT (WB dilution 1:5000); C1orf226 antibody 26061-1-AP (Proteintech, 1:5000 for WB), anti-Myc tag (Cell signaling technologies #2276, 1:1000), Beta-actin (Abcam ab49900, 1:10000), Tubulin (Sigma T6199, 1:1000). Blots were then washed 3 times in TBST then incubated in TBST containing HRP-conjugated secondary at 1:10,000 for 1 h at RT. Membranes were then washed 3 times in TBST prior to signal detection using ECL reagent (BioRad). Blots were imaged using a ChemiDoc (BioRad). Blot images were evaluated in ImageLab 6.1 and analyzed for quantification in FIJI 2.9.0.

Immunofluorescence staining and imaging of rat kidney sections

For immunostaining of frozen optimal cutting temperature (OCT) compound-embedded tissue sections (adult rat kidney), permeabilization was performed using 0.1% Triton X-100. After blocking, sections were incubated overnight at 4 °C with primary antibodies: guinea anti-nephrin (Progen, GP-N2, 1:100), custom antibody GST Long (WB dilution 1:500). Slides were then incubated in secondary antibodies (Donkey anti-guinea and anti-rabbit Alexa 488- and Alexa 594–conjugated secondary antibodies, 1/500) for 60 min at room temperature, followed by mounting in hardening medium with DAPI (1/2000, Invitrogen). Confocal microscopy imaging was performed using the Leica SP5X system with an upright DM6000 microscope, and images were processed with the Leica AF software suite X.

Filopodia quantification by live imaging

Immortalized human podocytes (shared by Moin Saleem)⁹³ were plated and transfected with cDNA constructs after 24 h. One hour later, the transfection media were exchanged with regular culture media. Cells were then imaged hourly with light and fluorescence microscopy at 10× to 20× using the IncuCyte ZOOM System v2018 (Essen BioScience). For semi-quantification of podocyte morphology with filopodia, an automated ImageJ pipeline was used to calculate cell circularity and a quotient of total over convex cell area, in which perfectly round cells would score “1” and elongated, filopodia-containing cells between 0 and 1, with higher degrees of elongation leading to lower scores. Three independent experiments were performed for confirmation. Quantification was performed in a blinded manner. Whole cell lysates were generated from transfected podocytes and blotted as above.

Midi gene splicing assay

Human NOS1AP Exon 9 and surrounding 500 bp of intronic region from both the parent B4606_11 and patient B4606_21 were amplified and cloned into the pcI-Neo-<mammalia cassette utilizing Gateway cloning and the following primers: N1AP_Splice_F_1 GGGGACAAGTTTGTACAAAAAAGCAGGCTTCAGGATGAGAAGAGAGTGCCC; N1AP_Splice_R_1 GGGGACCACTTTGTACAAGAAAGCTGGGTCTCACAGAGACCATGACACCC. Immortalized human podocytes were transfected with the empty cassette, wild-type, and patient variant constructs. At 24 h post-transfection, RNA isolation was performed using RNAeasy Mini Kit (Qiagen). Total RNA was reverse transcribed using iScript Reverse Transcription Supermix for RT-qPCR (Bio-Rad). cDNA was amplified using HotStarTaq DNA polymerase and the following primers: N1Ex9Mid_2_F, CGGAGGTCAACAACGAGTCT; N1Ex9Mid_2_R, TCTTGGACACGGTAGCAGAG. PCR products were detected by agarose gel electrophoresis and Sanger sequencing.

Cell lines

Immortalized human podocytes were shared by Moin Saleem⁹³. HEK293T cells were obtained from ATCC (CRL-3216). Primary mouse embryonic fibroblasts (MEFs) were isolated from wild-type and transgenic Nos1ap mice of both sexes. MCF7 cells were obtained from ATCC (HTB-22).

Statistics and reproducibility

For all immunoblots and DNA gel electrophoresis studies, the data shown are representative of three or more independent biological repeats.

Software

Graphpad Prism 10.2.3 software was used to perform statistical testing between groups.

Geneious Prime 2024.0 Bioinformatics Software for Sequence Data Analysis

Web resources

UCSC Genome Browser, genome.ucsc.edu

Ensembl Genome Browser, www.ensembl.org

gnomAD browser 2.1.1., gnomad.broadinstitute.org

GTEx Portal, gtexportal.org

Jackson Laboratory Synteny Browser, syntenybrowser.jax.org/browser.

Inclusion and ethics statement

Authorship for the manuscript was assigned in accordance with the Nature authorship policy on inclusion and ethics in global research.

Data availability

The majority of data generated during this study are included in the article, the supplementary information, and the source data files. Proteomics data generated during this study for peptides detected in newborn mouse kidneys are available at PRIDE under reference PXD058044. Proteomics data from tissue and cells that were re-analyzed for this study are also publicly available on PRIDE or ProteomeCentral under accession numbers PXD003306, PXD004263, and PXD004452. Mouse Bulk RNA-sequencing data reused in this study are available from the GEO database under accession numbers GSE225622 and GSE145053. Fetal kidney bulk RNA-sequencing data reused in this work (ENCODE reference number ENCSR000AFA) are available from the GEO database under accession number GSE78568. The long-read sequencing data used here are available at ENCODE (as found at encodeproject.org for accessions ENCFF114VSO, ENCFF877QJZ, ENCFF378STM, ENCFF564ONS, ENCFF406GQU, ENCFF222UTL, ENCFF279ABL, ENCFF319FBW, ENCFF626QRV, ENCFF613SDS, ENCFF291EKY, ENCFF173JOL, ENCFF840OVC, ENCFF132YCF, ENCFF850YMO, ENCFF623IBV, ENCFF018PZX, ENCFF502LAB, ENCFF911RNV, ENCFF886FZQ, ENCFF171AQO). Human kidney short-read data were requested through NEPTUNE (neptune-study.org/ancillary-studies). The human exome sequencing data are not publicly available because they contain information that could compromise research participants’ privacy. Please email the corresponding authors (amar.majmundar@childrens.harvard.edu and friedhelm.hildebrandt@childrens.harvard.edu) for requests for exome sequencing data for use in genetic discovery studies (with requests being addressed in 60 days), as the data is under controlled access. Source data are provided as a source data file.

Code availability

All newly generated code for analyses in this manuscript has been shared as indicated in the methods section. We have, generally, employed established published software as described in our methods sections. However, in one case, we have provided code used in existing software (SAMtools) generated for automating analysis of human kidney bulk RNA-sequencing data for the NEPTUNE cohort (Software S1).

References

United States Renal Data System. 2024 USRDS Annual Data Report: Epidemiology of kidney disease in the United States. (National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2024).
Wiggins, R.-C. The spectrum of podocytopathies: a unifying view of glomerular diseases. Kidney Int. 71, 1205–1214 (2007).
Article CAS PubMed Google Scholar
Tarshish, P., Tobin, J. N. & Bernstein, C. M. J. Edelmann, prognostic significance of the early course of minimal change nephrotic syndrome: report of the International Study of Kidney Disease in Children. J. Am. Soc. Nephrol. 8, 769–776 (1997).
Sadowski, C. E. et al. A single-gene cause in 29.5% of cases of steroid-resistant nephrotic syndrome. J. Am. Soc. Nephrol. 26, 1279 (2015).
Article CAS PubMed Google Scholar
Warejko, J. K. et al. Whole exome sequencing of patients with steroid-resistant nephrotic syndrome. Clin. J. Am. Soc. Nephrol. 13, 53 (2018).
Article CAS PubMed Google Scholar
Tan, W. et al. Analysis of 24 genes reveals a monogenic cause in 11.1% of cases with steroid-resistant nephrotic syndrome at a single center. Pediatr. Nephrol. 33, 305–314 (2018).
Article CAS PubMed Google Scholar
Connaughton, D. M. et al. Monogenic causes of chronic kidney disease in adults. Kidney Int. 95, 914–928 (2019).
Article PubMed PubMed Central Google Scholar
Park, J. et al. Single-cell transcriptomics of the mouse kidney reveals potential cellular targets of kidney disease. Science 360, 758 (2018).
Article CAS PubMed PubMed Central Google Scholar
Lovric, S., Ashraf, S., Tan, W. & Hildebrandt, F. Genetic testing in steroid-resistant nephrotic syndrome: when and how?. Nephrol. Dial. Transplant. 31, 1802–1813 (2015).
Article PubMed PubMed Central Google Scholar
Ashraf, S. et al. Mutations in six nephrosis genes delineate a pathogenic pathway amenable to treatment. Nat. Commun. 9, 1960 (2018).
Article ADS PubMed PubMed Central Google Scholar
Gee, H. Y. et al. ARHGDIA mutations cause nephrotic syndrome via defective RHO GTPase signaling. J. Clin. Invest. 123, 3243–3253 (2013).
Article CAS PubMed PubMed Central Google Scholar
Scott, R. P. et al. Podocyte-specific loss of Cdc42 leads to congenital nephropathy. J. Am. Soc. Nephrol. JASN 23, 1149–1154 (2012).
Article CAS PubMed Google Scholar
Shibata, S. et al. Modification of mineralocorticoid receptor function by Rac1 GTPase: implication in proteinuric kidney disease. Nat. Med. 14, 1370 (2008).
Article CAS PubMed Google Scholar
Blattner, S. M. et al. Divergent functions of the Rho GTPases Rac1 and Cdc42 in podocyte injury. Kidney Int. 84, 920–930 (2013).
Article CAS PubMed PubMed Central Google Scholar
Huang, Z. et al. Cdc42 deficiency induces podocyte apoptosis by inhibiting the Nwasp/stress fibers/YAP pathway. Cell Death Dis. 7, e2142–e2142 (2016).
Article CAS PubMed PubMed Central Google Scholar
Schell, C. et al. N-WASP is required for stabilization of podocyte foot processes. J. Am. Soc. Nephrol. 24, 713 (2013).
Article CAS PubMed PubMed Central Google Scholar
Yu, H. et al. Rac1 activation in podocytes induces rapid foot process effacement and proteinuria. Mol. Cell. Biol. 33, 4755 (2013).
Article CAS PubMed PubMed Central Google Scholar
Akilesh, S. et al. Arhgap24 inactivates Rac1 in mouse podocytes, and a mutant form is associated with familial focal segmental glomerulosclerosis. J. Clin. Invest. 121, 4127–4137 (2011).
Article CAS PubMed PubMed Central Google Scholar
Schneider, R. et al. DAAM2 variants cause nephrotic syndrome via actin dysregulation. Am. J. Hum. Genet. 107, 1113–1128 (2020).
Article CAS PubMed PubMed Central Google Scholar
Majmundar, A. J. et al. Recessive NOS1AP variants impair actin remodeling and cause glomerulopathy in humans and mice. Sci. Adv. 7, eabe1386 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Ghiurcuta, C. G. & Moret, B. M. E. Evaluating synteny for improved comparative studies. Bioinformatics 30, i9–i18 (2014).
Article CAS PubMed PubMed Central Google Scholar
Kolishovski, G. et al. The JAX synteny browser for mouse-human comparative genomics. Mamm. Genome 30, 353–361 (2019).
Article CAS PubMed PubMed Central Google Scholar
Nadeau, J. H. & Taylor, B. A. Lengths of chromosomal segments conserved since divergence of man and mouse. Proc. Natl Acad. Sci. 81, 814–818 (1984).
Article ADS CAS PubMed PubMed Central Google Scholar
Anastas, J. N. et al. A protein complex of SCRIB, NOS1AP and VANGL1 regulates cell polarity and migration, and is associated with breast cancer progression. Oncogene 31, 3696 (2011).
Article PubMed PubMed Central Google Scholar
Carrel, D. et al. Nitric oxide synthase 1 adaptor protein, a protein implicated in schizophrenia, controls radial migration of cortical neurons. Biol. Psychiatry 77, 969–978 (2015).
Article CAS PubMed Google Scholar
Richier, L. et al. NOS1AP associates with scribble and regulates dendritic spine development. J. Neurosci. 30, 4796 (2010).
Article CAS PubMed PubMed Central Google Scholar
Hernandez, K. et al. Overexpression of isoforms of nitric oxide synthase 1 adaptor protein, encoded by a risk gene for schizophrenia, alters actin dynamics and synaptic function. Front. Cell. Neurosci. 10, 6 (2016).
Article PubMed PubMed Central Google Scholar
Jaffrey, S. R., Snowman, A. M., Eliasson, M. J. L., Cohen, N. A. & Snyder, S. H. CAPON: a protein associated with neuronal nitric oxide synthase that regulates its interactions with PSD95. Neuron 20, 115–124 (1998).
Article CAS PubMed Google Scholar
Clattenburg, L. et al. NOS1AP functionally associates with YAP to regulate Hippo signaling. Mol. Cell. Biol. 35, 2265 (2015).
Article CAS PubMed PubMed Central Google Scholar
Wu, H. et al. Single-cell transcriptomics of a human kidney allograft biopsy specimen defines a diverse inflammatory response. J. Am. Soc. Nephrol. 29, 2069 (2018).
Article CAS PubMed PubMed Central Google Scholar
Karaiskos, N. et al. A single-cell transcriptome atlas of the mouse glomerulus. J. Am. Soc. Nephrol. 29, 2060 (2018).
Article CAS PubMed PubMed Central Google Scholar
Fang, M. et al. Dexras1: A G protein specifically coupled to neuronal nitric oxide synthase via CAPON. Neuron 28, 183–193 (2000).
Article CAS PubMed Google Scholar
Zhu, L.-J. et al. CAPON-nNOS coupling can serve as a target for developing new anxiolytics. Nat. Med. 20, 1050–1054 (2014).
Article CAS PubMed Google Scholar
Smith, A. et al. Cardiac muscle–restricted partial loss of Nos1ap expression has limited but significant impact on electrocardiographic features. G3 13, jkad208 (2023).
Article CAS PubMed PubMed Central Google Scholar
Guo, S. S. et al. KANK1 promotes breast cancer development by compromising Scribble-mediated Hippo activation. Nat. Commun. 15, 10381 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Rinschen, M. M. et al. A multi-layered quantitative in vivo expression atlas of the podocyte unravels kidney disease candidate genes. Cell Rep. 23, 2495–2508 (2018).
Article CAS PubMed PubMed Central Google Scholar
Jung, S. Y. et al. An anatomically resolved mouse brain proteome reveals parkinson disease-relevant pathways*. Mol. Cell. Proteom. 16, 581–593 (2017).
Article CAS Google Scholar
Sangermano, R. et al. Photoreceptor progenitor mRNA analysis reveals exon skipping resulting from the ABCA4 c.5461-10T→C mutation in Stargardt disease. Ophthalmology 123, 1375–1385 (2016).
Article PubMed Google Scholar
Reese, M. G., Eeckman, F. H., Kulp, D. & Haussler, D. Improved splice site detection in genie. J. Comput. Biol. 4, 311–323 (1997).
Article CAS PubMed Google Scholar
Richards, S. et al. ACMG Laboratory Quality Assurance Committee, Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet. Med. J. Am. Coll. Med. Genet. 17, 405–424 (2015).
Google Scholar
Landrum, M. J. et al. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 46, D1062–D1067 (2018).
Article CAS PubMed Google Scholar
Gharavi, A. G. et al. Mapping a locus for susceptibility to HIV-1-associated nephropathy to mouse chromosome 3. Proc. Natl Acad. Sci. USA 101, 2488–2493 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Tsaih, S.-W. et al. Genetic analysis of albuminuria in aging mice and concordance with loci for human diabetic nephropathy found in a genome-wide association scan. Kidney Int. 77, 201–210 (2010).
Article CAS PubMed Google Scholar
Falcone, S. et al. Modification of an aggressive model of Alport Syndrome reveals early differences in disease pathogenesis due to genetic background. Sci. Rep. 9, 20398 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Zheng, Z. et al. A Mendelian locus on chromosome 16 determines susceptibility to doxorubicin nephropathy in the mouse. Proc. Natl Acad. Sci. USA 102, 2502–2507 (2005).
Article ADS CAS PubMed PubMed Central Google Scholar
SASAKI, H. et al. Genetic background-dependent diversity in renal failure caused by the tensin2 gene deficiency in the mouse. Biomed. Res. 36, 323–330 (2015).
Article CAS PubMed Google Scholar
Takahashi, Y., Sasaki, H., Okawara, S. & Sasaki, N. Genetic loci for resistance to podocyte injury caused by the tensin2 gene deficiency in mice. BMC Genet. 19, 24–24 (2018).
Article CAS PubMed PubMed Central Google Scholar
Xu, J., Huang, Y., Li, F., Zheng, S. & Epstein, P. N. FVB mouse genotype confers susceptibility to OVE26 diabetic albuminuria. Am. J. Physiol. -Ren. Physiol. 299, F487–F494 (2010).
Article CAS Google Scholar
Laouari,D. et al. TGF-α mediates genetic susceptibility to chronic kidney disease. J. Am. Soc. Nephrol. 22, 327–335 (2011).
Baleato, R. M., Guthrie, P. L., Gubler, M.-C., Ashman, L. K. & Roselli, S. Deletion of CD151 results in a strain-dependent glomerular disease due to severe alterations of the glomerular basement membrane. Am. J. Pathol. 173, 927–937 (2008).
Article CAS PubMed PubMed Central Google Scholar
Kang, J. S. et al. Loss of α3/α4(IV) collagen from the glomerular basement membrane induces a strain-dependent isoform switch to α5α6(IV) collagen associated with longer renal survival in Col4a3−/− Alport Mice. J. Am. Soc. Nephrol. 17, 1962–1969 (2006).
Teixido-Trujillo, S. et al. Measured GFR in murine animal models: review on methods, techniques, and procedures. Pflüg. Arch. Eur. J. Physiol. 475, 1241–1250 (2023).
Article CAS Google Scholar
Fogo, A. B., Lusco, M. A., Najafian, B. & Alpers, C. E. AJKD atlas of renal pathology: congenital nephrotic syndrome of Finnish type. Am. J. Kidney Dis. 66, e11–e12 (2015).
Article PubMed Google Scholar
Ellis, D. et al. Long-term antiproteinuric and renoprotective efficacy and safety of losartan in children with proteinuria. J. Pediatr. 143, 89–97 (2003).
Article CAS PubMed Google Scholar
Webb, N. J. A. et al. Randomized, double-blind, controlled study of losartan in children with proteinuria. Clin. J. Am. Soc. Nephrol. 5, 417–424 (2010).
Lewis, E. J. et al. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N. Engl. J. Med. 345, 851–860 (2001).
Article CAS PubMed Google Scholar
Montané, B., Abitbol, C., Chandar, J., Strauss, J. & Zilleruelo, G. Novel therapy of focal glomerulosclerosis with mycophenolate and angiotensin blockade. Pediatr. Nephrol. 18, 772–777 (2003).
Article PubMed Google Scholar
Sreedharan, R. & Bockenhauer, D. Congenital nephrotic syndrome responsive to angiotensin-converting enzyme inhibition. Pediatr. Nephrol. 20, 1340–1342 (2005).
Article PubMed Google Scholar
Bagga, A., Mudigoudar, B. D., Hari, P. & Vasudev, V. Enalapril dosage in steroid-resistant nephrotic syndrome. Pediatr. Nephrol. 19, 45–50 (2004).
Article PubMed Google Scholar
Yi, Z. et al. Effect of fosinopril in children with steroid-resistant idiopathic nephrotic syndrome. Pediatr. Nephrol. 21, 967–972 (2006).
Article PubMed Google Scholar
Campbell, K. N. et al. Efficacy and safety of ACE inhibitor and angiotensin receptor blocker therapies in primary focal segmental glomerulosclerosis treatment: a systematic review and meta-analysis. Kidney Med. 4, 100457 (2022).
Article PubMed PubMed Central Google Scholar
Gross, O. et al. Early angiotensin-converting enzyme inhibition in Alport syndrome delays renal failure and improves life expectancy. Kidney Int. 81, 494–501 (2012).
Article CAS PubMed Google Scholar
Gross, O. et al. Preemptive ramipril therapy delays renal failure and reduces renal fibrosis in COL4A3-knockout mice with Alport syndrome. Kidney Int. 63, 438–446 (2003).
Article CAS PubMed Google Scholar
Rao, J. et al. Advillin acts upstream of phospholipase C ϵ1 in steroid-resistant nephrotic syndrome. J. Clin. Invest. 127, 4257–4269 (2017).
Article PubMed PubMed Central Google Scholar
Kaplan, J. M. et al. Mutations in ACTN4, encoding α-actinin-4, cause familial focal segmental glomerulosclerosis. Nat. Genet. 24, 251 (2000).
Article CAS PubMed Google Scholar
Brown, E. J. et al. Mutations in the formin gene INF2 cause focal segmental glomerulosclerosis. Nat. Genet. 42, 72–76 (2010).
Article CAS PubMed Google Scholar
Gbadegesin, R. A. et al. Mutations in the gene that encodes the F-actin binding protein anillin cause FSGS. J. Am. Soc. Nephrol. 25, 1991–2002 (2014).
Yamada, H., Takeda, T., Michiue, H., Abe, T. & Takei, K. Actin bundling by dynamin 2 and cortactin is implicated in cell migration by stabilizing filopodia in human non-small cell lung carcinoma cells. Int. J. Oncol. 49, 877–886 (2016).
Article CAS PubMed PubMed Central Google Scholar
Yamada, H. et al. Stabilization of actin bundles by a dynamin 1/cortactin ring complex is necessary for growth cone filopodia. J. Neurosci. 33, 4514 (2013).
Article CAS PubMed PubMed Central Google Scholar
Zhang, R. et al. Dynamin regulates the dynamics and mechanical strength of the actin cytoskeleton as a multifilament actin-bundling protein. Nat. Cell Biol. 22, 674–688 (2020).
Article CAS PubMed PubMed Central Google Scholar
Schiffer, M. et al. Pharmacological targeting of actin-dependent dynamin oligomerization ameliorates chronic kidney disease in diverse animal models. Nat. Med. 21, 601 (2015).
Article CAS PubMed PubMed Central Google Scholar
Frankish, A. et al. GENCODE reference annotation for the human and mouse genomes. Nucleic Acids Res. 47, D766–D773 (2019).
Article CAS PubMed Google Scholar
Pertea, M. et al. CHESS: a new human gene catalog curated from thousands of large-scale RNA sequencing experiments reveals extensive transcriptional noise. Genome Biol. 19, 208 (2018).
Article CAS PubMed PubMed Central Google Scholar
Hu, Z. et al. Revealing missing human protein isoforms based on ab initio prediction, RNA-seq and Proteomics. Sci. Rep. 5, 10940 (2015).
Article ADS PubMed PubMed Central Google Scholar
Groza, T. et al. The International Mouse Phenotyping Consortium: comprehensive knockout phenotyping underpinning the study of human disease. Nucleic Acids Res. 51, D1038–D1045 (2023).
Article CAS PubMed Google Scholar
Dickinson, M. E. et al. The International Mouse Phenotyping Consortium, The Jackson Laboratory, I. C. de la S. (ICS) Infrastructure Nationale PHENOMIN, Charles River Laboratories, MRC Harwell, The Toronto Centre for Phenogenomics, The Wellcome Trust Sanger Institute, RIKEN BioResource Center, High-throughput discovery of novel developmental phenotypes. Nature 537, 508–514 (2016).
Article CAS PubMed PubMed Central Google Scholar
Braun, D. A. et al. Mutations in multiple components of the nuclear pore complex cause nephrotic syndrome. J. Clin. Invest. 128, 4313–4328 (2018).
Article PubMed PubMed Central Google Scholar
Sugiyama, K. et al. Oxidative stress induced ventricular arrhythmia and impairment of cardiac function in Nos1ap deleted mice. Int. Heart J. 57, 341–349 (2016).
Article CAS PubMed Google Scholar
Bladt, F. et al. The murine Nck SH2/SH3 adaptors are important for the development of mesoderm-derived embryonic structures and for regulating the cellular actin network. Mol. Cell. Biol. 23, 4586–4597 (2003).
Article CAS PubMed PubMed Central Google Scholar
Harlan, S. M. et al. Progressive renal disease established by renin-coding adeno-associated virus–driven hypertension in diverse diabetic models. J. Am. Soc. Nephrol. 29, 477 (2018).
Article CAS PubMed Google Scholar
Conway, B. R. et al. Kidney single-cell atlas reveals myeloid heterogeneity in progression and regression of kidney disease. J. Am. Soc. Nephrol. 31, 2833–2854 (2020).
Xiong, L. et al. Direct androgen receptor control of sexually dimorphic gene expression in the mammalian kidney. Dev. Cell. https://doi.org/10.1016/j.devcel.2023.08.010 (2023).
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014).
Article CAS PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).
Reese, F. et al. The ENCODE4 long-read RNA-seq collection reveals distinct classes of transcript structure diversity. bioRxiv, https://doi.org/10.1101/2023.05.15.540865 (2023).
Gadegbeku, C. A. et al. Design of the Nephrotic Syndrome Study Network (NEPTUNE) to evaluate primary glomerular nephropathy by a multidisciplinary approach. Kidney Int. 83, 749–756 (2013).
Article PubMed PubMed Central Google Scholar
Wingett, S. & Andrews, S. FastQ Screen: a tool for multi-genome mapping and quality control [version 2; peer review: 4 approved]. F1000Research 7, 1338 (2018).
Danecek, P. et al. Twelve years of SAMtools and BCFtools. GigaScience 10, giab008 (2021).
Article PubMed PubMed Central Google Scholar
Wickham, H. Ggplot2: Elegant Graphics for Data Analysis (3e) (Springer, 2009).
Bekker-Jensen, D. B. et al. An optimized shotgun strategy for the rapid generation of comprehensive human proteomes. Cell Syst. 4, 587–599.e4 (2017).
Article CAS PubMed PubMed Central Google Scholar
Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372 (2008).
Article CAS PubMed Google Scholar
Saleem, M. A. et al. A conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression. J. Am. Soc. Nephrol. 13, 630 (2002).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

A.J.M. was supported by the NIH (5K12HD052896-13, 1K08DK125768-01A1), American Society of Nephrology (Norman Siegel Research Scholar Career Grant 81542), and Manton Center for Orphan Disease Research (Junior Faculty Award). This work was supported by a Boston Children’s Hospital Office of Faculty Development/Basic and Clinical Translational Research Executive Committees Faculty Career Development Fellowship (A.J.M). F.H. is the William E. Harmon Professor of Pediatrics and has grant support from the National Institutes of Health (5R01DK076683-13). J.P.F. was funded by the Canadian Institutes of Health Research (CIHR) (PJT-183923 and PJT-159738) and an EJLB Foundation scholar award. K.K. was funded by a CIHR-CGSMA award and a Nova Scotia Graduate Research Scholarship. L.M., K.L., B.B.B., and F.B. were supported by the German Research Foundation (456136540, 461126211, BE6072/2-2, and 404527522, respectively). This study was further supported by a grant from the American Society of Nephrology to F.B. (Carl W. Gottschalk Research Scholar). F.B. was further supported by the Else Kröner-Fresenius-Stiftung (iPRIME Clinician Scientist Forschungskolleg-2021_EKFK.15, UKE, Hamburg, Germany). K.S. was funded by the JSPS Overseas Research Fellowships (No. 202260295). M.G.S. is supported by the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK119380 and RC2DK122397) and the Pura Vida Kidney Foundation. A.C.G. is supported by NIH T32-DK007726. The Nephrotic Syndrome Study Network (NEPTUNE), Note S1 is part of the Rare Diseases Clinical Research Network (RDCRN), which is funded by the National Institutes of Health (NIH) and led by the National Center for Advancing Translational Sciences (NCATS) through its Division of Rare Diseases Research Innovation (DRDRI). NEPTUNE is funded under grant number U54DK083912 as a collaboration between NCATS and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). Additional funding and/or programmatic support is provided by the University of Michigan, NephCure, Alport Syndrome Foundation, and the Halpin Foundation. RDCRN consortia are supported by the RDCRN Data Management and Coordinating Center (DMCC), funded by NCATS and the National Institute of Neurological Disorders and Stroke (NINDS) under U2CTR002818. MMR was supported by the DFG (RI 2811/1-1 and RI 2811/2-1, and SFB1192-project B10), by the Young Investigator Award from the Novo Nordisk Foundation (NNF19OC0056043), by Novo Nordisk Foundation (NNF20SA0061466), the Carlsberg Young Investigator fellowship, as well as Aarhus Universitet Forskningsfond. We want to thank Louise Trakimas, Maria Ericsson, Anja Nordstrom, and Peg Coughlin at the Harvard Medical School EM Facility for their expertise and excellent technical work in acquiring TEM images. The mouse RNAseq analysis was performed with the computational resources provided by the Research Computing Group at Boston Children’s Hospital and Harvard Medical School (Boston, MA), including High-Performance Computing Clusters Enkefalos 2 (E2), and the BioGrids scientific software made available for data analysis. We want to further thank Dr. Frans Cremers and Zelia Corradi for kindly providing us with the expression vector used for the MIDI gene assay.

Author information

These authors jointly supervised this work: Friedhelm Hildebrandt, Amar J. Majmundar.

Authors and Affiliations

Department of Pediatrics, Boston Children’s Hospital, Harvard Medical School, Boston, MA, USA
Florian Buerger, Daanya Salmanullah, Lorrin Liang, Victoria Gauntner, Vineeta Sharma, Arathi Ranga, Alexander Rubin, David Ball, Sunwoo Hong, Katharina Lemberg, Ken Saida, Lea Maria Merz, Alexis C. Gomez, Michelle T. McNulty, Matthew G. Sampson, Friedhelm Hildebrandt & Amar J. Majmundar
University Children’s Hospital, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Florian Buerger & Sergio Guerrero-Castillo
Hamburg Center for Kidney Health (HCKH), University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Florian Buerger & Markus M. Rinschen
Department of Pharmacology, Dalhousie University, Halifax, NS, Canada
Kavita Krueger, Jiansong Qi, Josee Normand & James P. Fawcett
Department II of Internal Medicine and Center for Molecular Medicine Cologne, Faculty of Medicine and University Hospital Cologne, University of Cologne, Cologne, Germany
Katharina Lemberg
Department of Pediatrics, University Leipzig, Leipzig, Germany
Lea Maria Merz
Department of Medicine, Harvard Medical School, Boston, MA, USA
Sanja Sever & Matthew G. Sampson
Division of Renal Medicine, Massachusetts General Hospital, Boston, MA, USA
Sanja Sever & Alexis C. Gomez
Research Informatics, Information Technology, Boston Children’s Hospital, Boston, MA, USA
Biju Issac, Qianyi Ma & Liang Sun
Department of Biomedicine, Aarhus University, Aarhus, Denmark
Anja M. Billing, Fatih Demir & Markus M. Rinschen
Aarhus Institute of Advanced Studies, Aarhus University, Aarhus, Denmark
Markus M. Rinschen
Department of Medicine, University Medical Center Hamburg Eppendorf, Hamburg, Germany
Markus M. Rinschen
Institute of Human Genetics, University Hospital Cologne, Faculty of Medicine, University of Cologne, Cologne, Germany
Björn Reusch & Bodo B. Beck
Center for Molecular Medicine Cologne (CMMC), University of Cologne, Cologne, Germany
Björn Reusch & Bodo B. Beck
Kidney Disease Initiative and Medical Population Genetics Group, Broad Institute, Cambridge, MA, USA
Alexis C. Gomez, Michelle T. McNulty & Matthew G. Sampson
Division of Renal Medicine, Brigham and Women’s Hospital, Boston, MA, USA
Alexis C. Gomez & Matthew G. Sampson
Department of Pediatrics, Harvard Medical School, Boston, MA, USA
Matthew G. Sampson
Clinical Genomics Department, Centre for Genomic Medicine, King Faisal Specialist Hospital & Research Centre, Riyadh, Saudi Arabia
Mohamed H. Al-Hamed
Section of Medical Genetics, Children Specialist Hospital, King Fahad Medical City, Riyadh, Saudi Arabia
Mohammed M. Saleh
Pediatric Nephrology Centre of Excellence, Faculty of Medicine, King Abdulaziz University Hospital, Jeddah, Kingdom of Saudi Arabia
Mohamed A. Shalaby & Jameela A. Kari
Pediatric Department, King Abdulaziz University, Jeddah, Kingdom of Saudi Arabia
Mohamed A. Shalaby & Jameela A. Kari
Department of Surgery, Dalhousie University, Halifax, NS, Canada
James P. Fawcett

Authors

Florian Buerger
View author publications
Search author on:PubMed Google Scholar
Daanya Salmanullah
View author publications
Search author on:PubMed Google Scholar
Lorrin Liang
View author publications
Search author on:PubMed Google Scholar
Victoria Gauntner
View author publications
Search author on:PubMed Google Scholar
Kavita Krueger
View author publications
Search author on:PubMed Google Scholar
Jiansong Qi
View author publications
Search author on:PubMed Google Scholar
Josee Normand
View author publications
Search author on:PubMed Google Scholar
Vineeta Sharma
View author publications
Search author on:PubMed Google Scholar
Arathi Ranga
View author publications
Search author on:PubMed Google Scholar
Alexander Rubin
View author publications
Search author on:PubMed Google Scholar
David Ball
View author publications
Search author on:PubMed Google Scholar
Sunwoo Hong
View author publications
Search author on:PubMed Google Scholar
Katharina Lemberg
View author publications
Search author on:PubMed Google Scholar
Ken Saida
View author publications
Search author on:PubMed Google Scholar
Lea Maria Merz
View author publications
Search author on:PubMed Google Scholar
Sanja Sever
View author publications
Search author on:PubMed Google Scholar
Biju Issac
View author publications
Search author on:PubMed Google Scholar
Qianyi Ma
View author publications
Search author on:PubMed Google Scholar
Liang Sun
View author publications
Search author on:PubMed Google Scholar
Anja M. Billing
View author publications
Search author on:PubMed Google Scholar
Fatih Demir
View author publications
Search author on:PubMed Google Scholar
Markus M. Rinschen
View author publications
Search author on:PubMed Google Scholar
Björn Reusch
View author publications
Search author on:PubMed Google Scholar
Bodo B. Beck
View author publications
Search author on:PubMed Google Scholar
Sergio Guerrero-Castillo
View author publications
Search author on:PubMed Google Scholar
Alexis C. Gomez
View author publications
Search author on:PubMed Google Scholar
Michelle T. McNulty
View author publications
Search author on:PubMed Google Scholar
Matthew G. Sampson
View author publications
Search author on:PubMed Google Scholar
Mohamed H. Al-Hamed
View author publications
Search author on:PubMed Google Scholar
Mohammed M. Saleh
View author publications
Search author on:PubMed Google Scholar
Mohamed A. Shalaby
View author publications
Search author on:PubMed Google Scholar
Jameela A. Kari
View author publications
Search author on:PubMed Google Scholar
James P. Fawcett
View author publications
Search author on:PubMed Google Scholar
Friedhelm Hildebrandt
View author publications
Search author on:PubMed Google Scholar
Amar J. Majmundar
View author publications
Search author on:PubMed Google Scholar

Consortia

Nephrotic Syndrome Study Network (NEPTUNE)

Contributions

F.B., D.S., L.L., V.G., K.K., J.Q., V.S., A.Ranga, A.Rubin, D.B., K.L., K.S., J.N., S.H., L.M.M., S.S., J.F., and A.J.M. designed and/or performed cell-based and animal model studies. F.B., B.I., L.S., Q.M., S.G.C, A.M.B., F.D., M.M.R., B.R., B.B.B., A.C.G., M.T.M., M.G.S., N.S.S.N. and A.J.M. performed transcriptomic and proteomic analyses. F.B., V.G., A.R., K.L., M.H.A., M.M.S., M.S., J.K., F.H., and A.J.M. recruited patients, gathered detailed clinical information, and performed genetic analysis. All authors critically reviewed the paper. A.J.M., F.B., and F.H. conceived of and directed the project. A.J.M., F.B., and F.H. prepared the manuscript.

Corresponding authors

Correspondence to Friedhelm Hildebrandt or Amar J. Majmundar.

Ethics declarations

Competing interests

The authors declare no competing interests. F.H. is a co-founder of Goldfinch Biopharma Inc. A.J.M. is a consultant for Judo, Inc. The other authors have no disclosures.

Peer review

Peer review information

Nature Communications thanks Murim Choi, Katalin Susztak, and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Software 1

Transparent Peer Review file

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Buerger, F., Salmanullah, D., Liang, L. et al. Recessive variants in the intergenic NOS1AP-C1orf226 locus cause monogenic kidney disease responsive to anti-proteinuric treatment. Nat Commun 16, 10654 (2025). https://doi.org/10.1038/s41467-025-65663-6

Download citation

Received: 07 February 2024
Accepted: 17 October 2025
Published: 27 November 2025
Version of record: 27 November 2025
DOI: https://doi.org/10.1038/s41467-025-65663-6