Podocyte-specific Nup160 knockout mice develop nephrotic syndrome and glomerulosclerosis

Abstract

More than 60 monogenic genes mutated in steroid-resistant nephrotic syndrome (SRNS) have been identified. Our previous study found that mutations in nucleoporin 160 kD (NUP160) are implicated in SRNS. The NUP160 gene encodes a component of the nuclear pore complex. Recently, two siblings with homozygous NUP160 mutations presented with SRNS and a nervous system disorder. However, replication of nephrotic syndrome (NS)-associated phenotypes in a mammalian model following loss of Nup160 is needed to prove that NUP160 mutations cause SRNS. Here, we generated a podocyte-specific Nup160 knockout (Nup160^podKO) mouse model using CRISPR/Cas9 and Cre/loxP technologies. We investigated NS-associated phenotypes in these Nup160^podKO mice. We verified efficient abrogation of Nup160 in Nup160^podKO mice at both the DNA and protein levels. We showed that Nup160^podKO mice develop typical signs of NS. Nup160^podKO mice exhibited progression of proteinuria to average albumin/creatinine ratio (ACR) levels of 15.06 ± 2.71 mg/mg at 26 weeks, and had lower serum albumin levels of 13.13 ± 1.34 g/l at 30 weeks. Littermate control mice had urinary ACR mean values of 0.03 mg/mg and serum albumin values of 22.89 ± 0.34 g/l at the corresponding ages. Further, Nup160^podKO mice exhibited glomerulosclerosis compared with littermate control mice. Podocyte-specific Nup160 knockout in mice led to NS and glomerulosclerosis. Thus, our findings strongly support that mutations in NUP160 cause SRNS. The newly generated Nup160^podKO mice are a reliable mammalian model for future study of the pathogenesis of NUP160-associated SRNS.

nephrotic syndrome, NUP160, knockout, mice, podocyte

Introduction

Nephrotic syndrome (NS) is characterized by massive proteinuria, hypoalbuminemia, hyperlipidemia, and edema, reflecting dysfunction of the normally highly permselective glomerular filtration barrier [1]. It is further categorized as steroid-sensitive nephrotic syndrome and steroid-resistant nephrotic syndrome (SRNS) on the basis of the response to steroid therapy. Around 50% of children with SRNS are at risk of progressing to end-stage kidney disease (ESKD) within 5 years, especially those who do not show complete or partial remission [2, 3]. The pathogenesis of SRNS is thought to be due to either genetic variation or immune-mediated [2]. So far, more than 60 monogenic genes mutated in SRNS have been identified [4, 5]. The majority of known SRNS-causing genes are primarily expressed in podocytes of the glomerulus, and podocyte dysfunction is the principal reason for proteinuria [6–9]. These SRNS-causing genes include those encoding glomerular basement membrane-related proteins (such as LAMB2, ITGA3, ITGB4, COL4A3/4/5, and CD151), mitochondria-related proteins (such as COQ2, COQ6, ADCK4, PDSS2, and MTTL1), podocyte slit diaphragm proteins (such as NPHS1, NPHS2, CD2AP, CRB2, and FAT1), and podocyte actin cytoskeleton proteins (such as ACTN4, MYH9, MYO1E, INF2, and AVIL) [5, 7]. Recently, mutations in monogenic genes encoding nucleoporins (such as NUP85, NUP93, NUP107, NUP133, and NUP205) have been shown to cause SRNS [10–14].

Our previous studies [4, 15], combined with the findings of Braun et al. [10], demonstrate that mutations in NUP160 are implicated in SRNS. We found that a proband with familial SRNS and focal segmental glomerular sclerosis (FSGS) carried two compound heterozygous NUP160 mutations, c.2407G>A (p.Glu803Lys) and c.3517C>T (p.Arg1173^*). The human NUP160 gene (NC_000011.10; mouse and Drosophila: Nup160) is located on chromosome 11p11.2 and composed of 36 exons (Gene ID: 23279). NUP160 encodes nucleoporin 160 kD (NUP160; NP_056046.2), a component of the nuclear pore complex and is expressed in all nucleated cell types from flies to humans [4, 16]. Our in vivo functional validation studies in a Drosophila model showed that compound heterozygous NUP160 mutations, c.2407G>A (p.Glu803Lys) and c.3517C>T (p.Arg1173^*), in a proband with familial SRNS are likely pathogenic [4]. We also found that knockdown of Nup160 damages conditionally immortalized mouse podocytes [15]. In another unrelated Chinese family with autosomal recessive proteinuria, Braun et al., [10] reported that both siblings (an elder brother was diagnosed with SRNS and younger sister with massive proteinuria) carried two compound heterozygous NUP160 mutations, c.2407G>A (p.Glu803Lys) and c.2728C>T (p.Arg910^*). Recently, Maddirevula et al. [17] revealed that two siblings carried homozygous mutation of NUP160, c.1179+5G>A (p.Phe368_Gln393del), in an autosomal recessive FSGS family. The elder sister, who progressed to ESKD, presented with FSGS and myoclonic seizures, while the younger brother presented with SRNS, FSGS, developmental retardation, and epilepsy. These findings support the likelihood that mutations in NUP160 cause SRNS. However, in accordance with the rigorous guidelines for investigating causality of variants in human disease, replication of NS-associated phenotypes in a mammalian model without Nup160 expression is needed to show that mutations in NUP160 cause SRNS [18].

Mouse models are the most widely studied mammal in relation to renal disease [19]. Mice have been generated with podocyte-specific knockout of a single gene. These mimic NS-associated phenotypes caused by loss-of-function mutations and include Coq6^podKO, Adck4^podKO, Magi-2^podKO, Myo1e^podKO, and Podxl^podKO mice [20–24]. Our previous study has demonstrated that both NUP160 c.2407G>A and c.3517C>T mutations in a proband with familial SRNS and FSGS are loss-of-function mutations [4]. Therefore, we generated a mouse model with podocyte-specific knockout of Nup160 (referred to as Nup160^podKO hereafter) to determine the role of NUP160 mutations in causation of SRNS. Here, we show for the first time that podocyte-specific deletion of Nup160 in mice results in the development of NS and glomerulosclerosis.

Results

Podocyte-specific deletion of Nup160 in mice

To investigate the in vivo role of Nup160 in NS, as well as obtain a potential animal model for further study and avoid the possibility of embryonic lethality in a complete Nup160 knockout mouse, we generated mice with podocyte-specific loss of Nup160. We achieved this by crossing and backcrossing Podocin-Cre mice, which have been engineered to exhibit podocyte-specific expression of Cre recombinase, with Nup160^flox/flox mice, which have two loxP sites surrounding exon 4 of the Nup160 gene (Fig. 1A). Mice with the expected genotypes were detected by PCR from tail tips (Fig. 1B). Podocin-Cre-dependent Nup160 knockout was confirmed in the kidneys of Nup160^podKO mice by PCR, and not detected in other organs (Fig. 1C). Efficient loss of Nup160 expression in podocytes was confirmed by immunostaining of consecutive kidney sections. No Nup160 protein was detected in Nup160^podKO kidneys compared with three groups of littermate control mice (Fig. 1D).

Figure 1

Generation and validation of the Nphs2.Cre⁺;Nup160^flox/flox mouse model. (A) Schematic diagram for insertion of loxP sites on both sides of exon 4 of the Nup160 wild-type allele (A₁ and A₂). Excision of the target segment generated by Cre-loxP technology (A₃). Primer locations and exons are depicted with numbers. (B) Representative genotyping of PCR products from the tail tips of mice. Primers F₁/R₁ amplified a 156 bp product from the wild-type locus (red arrow) and a 224 bp product from the loxP-containing conditional knockout locus (black arrow) situated upstream of exon 4. Primers Fcre/Rcre generated approximately 200 bp products from the Nphs2.Cre locus (blue arrow). (C) PCRs for tissue-specific genotyping of heart, liver, and kidney. Primers F₂/R₂ generated a 1732 bp product in the loxP-inserted Nup160 allele (black arrow), showing the presence of a mutant locus except in Nphs2.Cre⁺ mice and the water control. Primers F₂/R₂ generated a 1592 bp product from the wild-type allele (red arrow). The 545 bp product (blue arrow) occurs in Nphs2.Cre⁺;Nup160^flox/flox mice and Nphs2.Cre⁺;Nup160^flox/+ mice indicating podocyte-specific Nphs2;Cre^+/− mediated locus recombination. The target bands at 1732 bp and 1592 bp are too close to be independently resolved. (D) Immunohistochemistry to detect nucleoporin 160 kD (Nup160) protein expression in consecutive kidney sections from 30 week-old mice of the indicated genotypes. (D₁–D₃) Normal expression pattern of nephrin in glomerular podocytes of three groups of littermate control mice. Expression was mainly distributed in peripheral glomerular capillary loops, extending in a reticular pattern into the central part of the glomerulus (brown). (D₅–D₇) Nup160 showed diffuse positive staining in nuclei of glomerular cells and tubular cells in control mice (brown). (D₉–D₁₁) Nup160 (red) co-localized with the podocyte-specific marker protein, nephrin (brown). (D₄) Nphs2.Cre⁺;Nup160^flox/flox mice showed reduced nephrin staining on capillary loops (brown). (D₈) Significantly reduced positive rate of Nup160 (brown) in glomerular cells of Nphs2.Cre⁺;Nup160^flox/flox mice compared with controls. Consecutive kidney sections showed no expression of Nup160 in podocytes. (D₁₂) Consistent with single staining, Nup160 (red) was almost absent in podocytes that colocalized with nephrin (brown). Green arrows indicate podocytes in consecutive kidney sections of control mice. Red arrows indicate podocytes in consecutive kidney sections of Nphs2.Cre⁺;Nup160^flox/flox mice. Red triangles indicate proteinaceous casts. Original magnification, ×200.

Podocyte-specific Nup160 knockout mice lead to nephrotic syndrome and renal dysfunction

Nup160^podKO mice appeared normal at birth and were born at the expected Mendelian ratios, indicating no fetal death. In this study, mice were sacrificed at 30 weeks of age, which is approximately equivalent to a human age of 23 years [25]. The only exceptions were for one Nup160^podKO mouse that died spontaneously at 24 weeks and another that was euthanized at 26 weeks due to its extremely poor physical condition (Supplementary Material Fig. S3). Although young Nup160^podKO mice appeared grossly normal, they showed a poorer physical condition, such as hunched posture, squinted eyes, and scruffy fur (Supplementary Material Fig. S1A). Nup160^podKO mice at 26 weeks old also showed a tendency towards a decreased body weight compared with littermate controls, but the difference was not statistically significant (Supplementary Material Fig. S1B and Table S2). Moreover, a large amount of ascites and edematous bowel were visible in the abdominal cavity (Fig. 2D).

Figure 2

Nphs2.Cre⁺;Nup160^flox/flox mutant mice develop severe progressive proteinuria and ascites. (A and B) Progressive proteinuria was observed in Nphs2.Cre⁺;Nup160^flox/flox mice from 20 weeks of age. Massive albuminuria was detected in Nphs2.Cre⁺;Nup160^flox/flox mice at 22 weeks, but not in three groups of littermate control mice. Shown is a Coomassie blue-stained protein gel to detect urinary albumin (1 μl urine/lane). The prominent protein band corresponds to bovine serum albumin (BSA). (C) Urinary albumin was quantified by enzyme-linked immunosorbent assay and normalized to urinary creatinine at the indicated time points (N = 6 mice in each group). Serial analysis of urinary albumin/creatinine ratio at the indicated ages and genotypes shows progressive proteinuria in Nphs2.Cre⁺;Nup160^flox/flox mice (red circle) but not littermate controls (Nup160^flox/flox: Green square, Nup160.Cre⁺: Yellow triangle, and Nphs2.Cre⁺;Nup160^flox/+: Blue triangles). One-way ANOVA with P values calculated using the Games-Howell test are shown in the figure. ns, not significant; ^**P < 0.01. Each data point represents the mean value of technical duplicates and error bars represent SEM. (D) Nphs2.Cre⁺;Nup160^flox/flox mutant mice show a marked increase of ascites (black arrows) and a highly edematous bowel compared with littermate controls.

To investigate the progression of proteinuria, we examined Nup160^podKO mice by urinalysis until 26 weeks old. The first prominent increase in urinary average albumin/creatinine ratio (ACR) (49-fold, P < 0.01) was detected in Nup160^podKO mice at 20 weeks of age, consistent with a Coomassie blue-stained protein gel to detect albuminuria. This ratio remained significant throughout the study period compared with controls (Fig. 2A–C, Supplementary Material Fig. S2 and Table S3). At 26 weeks, the increase of albuminuria over time reached a maximum, with an urinary ACR mean value up to 15.06 mg/mg (502-fold, P < 0.01) in Nup160^podKO mice compared with controls (urinary ACR mean value, 0.03 mg/mg) (Fig. 2C and Supplementary Material Table S3).

To examine additional biochemical indices of Nup160^podKO mice, we performed serum analysis at 30 weeks of age with the indicated genotypes (Fig. 3 and Supplementary Material Table S4). Nup160^podKO mice showed a significant decrease in serum albumin levels (13.13 ± 1.34 g/l versus 22.77 ± 0.43, 22.68 ± 0.64, and 23.20 ± 0.77 g/l in three groups of littermate control mice, respectively, P < 0.001; Fig. 3A and Supplementary Material Table S4). They also showed a significant increase in levels of serum cholesterol (9.35 ± 1.11 mmol/l in versus 1.94 ± 0.09, 1.72 ± 0.09, and 2.30 ± 0.18 mmol/l in controls, respectively, P < 0.01; Fig. 3B and Supplementary Material Table S4). Similarly, biochemical measures of renal function in Nup160^podKO mice revealed significant elevations in serum blood urea nitrogen (BUN) levels (65.35 ± 15.11 mmol/l versus 9.29 ± 0.54, 9.04 ± 0.64, and 8.70 ± 0.79 mmol/l in controls, respectively, P < 0.05; Fig. 3C and Supplementary Material Table S4), and serum creatinine (63.53 ± 12.05 μmol/l versus 18.69 ± 2.45, 30.42 ± 5.27, and 24.53 ± 3.97 μmol/l in controls, respectively, P < 0.001, P < 0.05, P < 0.01, respectively; Fig. 3D and Supplementary Material Table S4).

Figure 3

Nphs2.Cre⁺;Nup160^flox/flox mutant mice develop nephrotic syndrome and renal insufficiency. (A–D) Blood chemistry analysis on mice at 30 weeks of age showed that Nphs2.Cre⁺;Nup160^flox/flox mice (red column) exhibited significant hypoalbuminemia, hypercholesterolemia, and renal insufficiency, which was not detected in three groups of littermate control mice (Nup160^flox/flox: Green column; Nup160.Cre⁺: Yellow column; Nphs2.Cre⁺;Nup160^flox/+: Blue column. N = 5–7 mice in each group). Statistical comparisons were performed using one-way ANOVA with P values calculated using Tukey’s multiple comparison test or Games-Howell test as appropriate and stated in the figure; ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. Each data point represents the mean value of technical duplicates and error bars represent SEM.

These results indicate that Nup160^podKO mice, but not littermate controls, show characteristics of NS: mass proteinuria (Fig. 2B and C), significant hypoalbuminemia (Fig. 3A), hypercholesterolemia (Fig. 3B), and ascites (Fig. 2D), along with renal dysfunction (Fig. 3C and D). Interestingly, one Nup160^podKO mouse, whom we named Nup160^podKO-X, had mild symptoms and signs, with no ascites and an urinary ACR of approximately 3.13 mg/mg (104-fold) at 26 weeks, therefore still higher than littermate controls (0.03 mg/mg).

Podocyte-specific Nup160 knockout mice develop glomerulosclerosis and podocyte foot process effacement

As expected, the kidneys of Nup160^podKO mice and their littermate controls exhibited severe and normal renal pathologies, respectively, at 30 weeks of age (Fig. 4). Histological examination of Nup160^podKO mice by hematoxylin and eosin (H&E), periodic acid–Schiff (PAS), Masson and periodic acid-silver methenamine (PASM-Masson) staining showed extensive globular glomerulosclerosis. Some glomeruli were in a hypoplastic state, accompanied by glomerular capillary collapse. In addition, there was a renal tubular phenotype involving tubular atrophy, interstitial inflammation and fibrosis, varying degrees of dilatation, and the presence of proteinaceous casts (Fig. 4D1–D₃ and d₁–d₃). Histological analysis of littermate controls showed they had a normal glomerular and tubulointerstitial architecture (Fig. 4A1–C₁, a₁–c₁, A₂–C₂, a₂–c₂, A₃–C₃, and a₃–c₃). Globally sclerotic and obsolescent glomeruli were detected in Nup160^podKO mice, with severe and extensive tubulointerstitial lesions that appeared to be associated with a later time point (up to 30 weeks of death). Relatively mild abnormalities were found in the Nup160^podKO-X mouse, including FSGS rather than globular glomerulosclerosis (Fig. 5D1 and Supplementary Material Fig. S4A–F), which may represent an early stage of the severe phenotype of Nup160^podKO mice; this is consistent with the blood and urine results for this mouse. To further study the degree of glomerulosclerosis in Nup160^podKO mice, we quantified the number of sclerotic glomeruli by PASM-Masson staining at 30 weeks of age. The glomerular sclerosis index (GSI) was significantly increased in Nup160^podKO mice (0.82 ± 0.11 versus 0.03 ± 0.01, 0.02 ± 0.01, and 0.02 ± 0.01 in controls, respectively, P < 0.05; Fig. 5). We also performed morphological assessment of podocyte ultrastructure by transmission electron microscopy to investigate the effect of podocyte loss of Nup160 on glomerular ultrastructure. Littermate controls showed a normal foot process morphology, whereas Nup160^podKO mice exhibited widespread foot process fusion and effacement. Further, podocytes exhibited cytoplasmic vacuolization and microvillous transformation (Fig. 6D and d). As expected, ultrastructural analysis of the Nup160^podKO-X mouse revealed segmental podocyte foot process effacement (Supplementary Material Fig. S4H and I), consistent with FSGS and the results of blood and urinalysis.

Figure 4

Nphs2.Cre⁺;Nup160^flox/flox mutant mice show macroscopic morphological changes and glomerulosclerosis in kidneys. (A–D) Representative images of the kidneys in four groups of 30 week-old mice of the indicated genotypes. (D) The kidneys of Nphs2.Cre⁺;Nup160^flox/flox mice were much paler and rougher, with red spots due to localized hemorrhages visible at the surface. (A–C) In contrast, littermate controls showed relatively smooth and intact surfaces. (A₁–d₃) Kidney sections of 30 week-old mice of the indicated genotypes were stained with H&E, PAS, and PASM-Masson. (D₁–D₃, d₁–d₃) Nphs2.Cre⁺;Nup160^flox/flox mice showed visible glomerulosclerosis (red arrows), along with tubular atrophy (black arrows), proteinaceous casts in dilated tubules (red triangles), and renal interstitial inflammation (green arrow). Focal tubular epithelial cell necrosis and interstitial fibrosis were also observed in mutant mice. (A₁–C₁, a₁–c₁, A₂–C₂, a₂–c₂, A₃–C₃, and a₃–c₃) Littermate controls displayed a normal histological glomerular morphology (blue arrows). H&E, hematoxylin and eosin; PAS, periodic acid–Schiff; PASMMasson, Masson and periodic acid–silver methenamine. N = 8 mice in each group. Original magnification, ×200 (A₁–D₃); Original magnification, ×100 (a₁–d₃).

Figure 5

Nphs2.Cre⁺;Nup160^flox/flox mutant mice have significantly increased numbers of sclerotic glomeruli. Kidney sections of 30 week-old mice of the indicated genotypes were stained with H&E, PAS, and PASM-Masson. PASM-Masson stained kidney sections were analyzed to quantify the severity of glomerular sclerosis. (A–D, A₁–D₁, A₂–D₂, and E) Nphs2.Cre⁺;Nup160^flox/flox mutant mice displayed extensive glomerulosclerosis (red arrows), along with proteinaceous casts in dilated tubules (red triangles). In contrast, the three groups of littermate control mice showed virtually no (< 5% of glomeruli) aberrant histological changes (blue arrows). D₁ and red bar graph: 66/134 represents Nup160^podKO-X mice. N = 4 mice in each group. Each bar graph represents a single mouse. The number of sclerosed glomeruli per total glomeruli counted in one section is marked by numbers inside the bar graphs. One-way ANOVA with P values calculated using the Games-Howell test are shown in the figure. ^*P < 0.05. H&E, hematoxylin and eosin; PAS, periodic acid–Schiff; PASM-Masson, Masson and periodic acid–silver methenamine. Original magnification, × 40 (A–D, A₁–D₁, and A₂–D₂).

Figure 6

Transmission electron microscopy (TEM) reveals abnormal morphology of podocytes in Nphs2.Cre⁺;Nup160^flox/flox mutant mice. (A–D and a–d) Representative TEM images of the indicated genotypes in all four groups of 30 week-old mice. (A–C and a–c) Littermate controls displayed a normal ultrastructure of foot process morphology (blue arrows). (D and d) The glomerular ultrastructure of Nphs2.Cre⁺;Nup160^flox/flox mutant mice showed extensive fusion and effacement of podocyte foot processes; podocyte numbers were also significantly reduced (red arrows). N = 8 mice in each group. Scale bars, A–D, 5 μm; a–d, 2 μm.

Discussion

In this study, we generated a podocyte-specific Nup160 knockout mouse model, and show that podocyte-specific abrogation of Nup160 causes NS and glomerulosclerosis in mice. These findings strongly suggest that mutations in NUP160 cause SRNS.

Nup160^podKO mice showed typical signs of NS, such as massive proteinuria, hypoalbuminemia, and hypercholesterolemia, along with renal dysfunction. Nup160^podKO mice exhibited progression of proteinuria to an average ACR peak of 15.06 mg/mg at 26 weeks (Fig. 2C), consistent with both Coq6^podKO mice and Podxl^podKO mice that also develop NS. In contrast, littermate control mice had an urinary ACR mean value of 0.03 mg/mg. Coq6^podKO mice had an average ACR peak of approximately 10.00 mg/mg at 8 months, and Podxl^podKO mice 10.00 mg/mg at 3 weeks [20, 24]. Nup160^podKO mice had lower serum albumin levels (13.13 g/l) and higher serum cholesterol levels (9.35 mmol/l) at 30 weeks (Fig. 3 and Supplementary Material Table S4), which is similar to Podxl^podKO mice and Pdss2^podKO mice; littermate controls had serum albumin values of 22.89 g/l and serum cholesterol of 2.00 mmol/l. Podxl^podKO mice had an average serum albumin of approximately 15.00 g/l, and Pdss2^podKO mice had an average cholesterol of 8.00 mmol/l [24, 26]. Moreover, a significant increase in serum creatinine and BUN levels was observed in Nup160^podKO mice (Fig. 3C and D), suggesting they had impaired kidney function. As predicted, Nup160^podKO mice exhibited typical features of NS, which mimicked that of a proband with familial SRNS and FSGS in our previous study [4].

The renal pathology of Nup160^podKO mice revealed glomerulosclerosis at 30 weeks (Figs 4 and 5), which was also demonstrated in other podocyte-specific gene knockout mice, such as Coq6^podKO, Adck4^podKO, Magi-2^pdKO, Fat1^podKO, and Crb2^podKO mice [20–22, 27–29]. Furthermore, Nup160^podKO mice exhibited an abnormal podocyte morphology, characterized by extensive fusion and effacement of podocyte foot processes at 30 weeks (Fig. 6). Renal pathology of Nup160^podKO mice also mimicked the pathology of FSGS in SRNS patients with mutations in NUP160 [4, 10, 17].

Nup160^podKO mice were generated by podocyte-specific Nup160 knockout using CRISPR/Cas9 and Cre/loxP technologies [30–32], therefore they showed renal manifestations typical of NS, but with no neurological symptoms of epilepsy. Efficient abrogation of the Nup160 gene was shown by gene-specific genotyping using PCR products from tail tips of Nup160^podKO mice (Fig. 1B). Podocyte-specific Nup160 knockout efficiency was confirmed at the DNA and protein levels (Fig. 1C and D). We verified by PCR that podocin-Cre-dependent Nup160 inactivation occurred in only the kidneys of Nup160^podKO mice, and not in the other organs, such as heart and liver (Fig. 1C). Effective loss of Nup160 was also confirmed by immunohistochemistry analysis of glomerular Nup160 protein expression in Nup160^podKO mice (Fig. 1D).

The major aim of our study was to determine whether a podocyte-specific Nup160 knockout mouse model developed massive proteinuria and hypoalbuminemia. Although typical characteristics of NS were detected in Nup160^podKO mice, two limitations of this study should be noted. First, our renal pathology analysis was limited to a single observational time point. Yet, second, the Nup160^podKO mice were sacrificed at a later time point. Therefore, most specimens had already progressed to glomerulosclerosis. In future studies on the pathogenesis and potential therapeutic agents of NUP160-associated SRNS, Nup160^podKO mice should be examined at different time points for kidney pathology analysis.

In conclusion, podocyte-specific Nup160 knockout mice develop NS and glomerulosclerosis. Our findings confirm that mutations in NUP160 cause SRNS. The newly generated Nup160^podKO mice provide a reliable mammalian model for future study of the pathogenesis of NUP160-associated SRNS.

Materials and methods

Mice

NPHS2.Cre⁺ mice, referred to as podocin-Cre mice (B6.Cg-Tg [NPHS2-cre] 295Lbh/J, 008205), were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Nup160^flox/flox mice were generated by CRISPR/Cas9-mediated genome editing, with two loxP sites inserted around exon 4 of the Nup160 gene. Homozygous floxed Nup160 mice (Nup160^flox/flox) were crossed with NPHS2.Cre⁺ mice to generate double heterozygous floxed Nup160 mice (Nphs2.Cre⁺;Nup160^flox/+). Nphs2.Cre⁺;Nup160^flox/+ mice were backcrossed to generate podocyte-specific Nup160 knockout mice (Nphs2.Cre⁺;Nup160^flox/flox). All mice were divided into four groups: one experimental group (Nphs2.Cre⁺;Nup160^flox/flox, referred to hereafter as Nup160^podKO) and three control groups (Nup160^flox/flox, Nphs2.Cre⁺, and Nphs2.Cre⁺;Nup160^flox/+). All animals were raised in specific pathogen-free conditions on a normal 12-h light-dark cycle. The temperature of the barrier environment was 22–25°C and humidity was 40%–70%. Mice had unlimited water and rodent food. The experimental protocol was approved by the Animal Care Committee of the Fuzong Clinical Medical College, Fujian Medical University (2020-062).

DNA extraction and PCR

Tissue was directly frozen for DNA extraction according to the manufacturer’s instructions (9765, Takara Bio Inc., Shiga, Japan). Then, 20 mg of mouse tail tissue was incubated at 56°C overnight with lysis buffer containing 20 μl proteinase K (20 mg/ml) and 10 μl RNase A (10 mg/ml). Heart, liver, and kidney tissues were lysed at 56°C for 2–3 h. After multiple purification, elution, and centrifugation steps, the DNA-containing supernatant of samples was PCR amplified using specific primers. The primer sequences are shown in Supplementary Material Table S1.

Urine analysis

Morning urine was collected by kneading the bladder of mice softly. All samples were frozen straight away and stored at −80°C. The samples were thawed on ice and shaken before measuring urine albumin and urine creatinine. Gel electrophoresis of urinary proteins was also performed.

Measurement of urinary albumin and creatinine

Assessment of urinary albumin was performed using a Mouse Albumin ELISA Kit (AKRAL-121, FUJIFILM Wako Shibayagi Corp., Shibukawa, Japan) according to the manufacturer’s instructions. Measurement of creatinine was performed using a non-enzymatic Creatinine Assay Kit (KGE005, R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. Proteinuria was expressed as urinary ACR.

SDS–PAGE

Urine samples were boiled with sodium dodecyl sulfate (SDS) sample buffer and separated using an SDS–PAGE Color Preparation kit (C671102, Sangon Biotech, Shanghai, China). Urine samples (1 μl) were mixed with sample loading buffer, incubated at 100°C for 5 min, and resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel (SDS–PAGE). Gels were stained with Coomassie Brilliant Blue for 3 h and destained overnight. Images were captured with a GS-900™ Calibrated Densitometer (Bio-Rad, Hercules, CA, USA).

Blood chemistry analysis

Blood was collected by cardiac puncture in Eppendorf tubes when the mice were deeply anesthetized. The blood was allowed to clot on ice within 2 h. After centrifugation at 10000 × g for 10 min at 4°C, serum was obtained by pipetting the supernatant from the sedimented mixture. Serum was preserved for biochemical analysis by storing on ice. Blood chemistry analysis was performed using a Hitachi LABOSPECT 008 AS Automatic Biochemical analyzer (Tokyo, Japan), in accordance with the manufacturer’s instructions.

Renal pathology analysis

Longitudinally halved kidneys were fixed with 10% formalin fixative, which was followed by histopathological analysis and immunohistochemistry staining. Fresh kidney tissue rich in renal cortex was used for ultrastructural analysis.

Histology

Kidneys were fixed and dehydrated, then embedded in paraffin. Paraffin sections of 1 μm thickness were stained with PAS and PASM-Masson, while those of 4 μm thickness were stained with H&E, according to standard protocols. Sections were then observed using an optical microscope (Olympus BX51, Tokyo, Japan).

Immunohistochemistry staining

Paraffin-embedded kidney tissue was sectioned (3 μm) and de-paraffinized, then rehydrated in an ethanol concentration gradient. Tissue slides underwent heat-induced antigen retrieval with ethylenediaminetetraacetic acid buffer. Sections were then quenched by 3% hydrogen peroxide and blocked with 5% bovine serum albumin for 15 min at room temperature. Primary antibodies against Nup160 (MBS2014806, MyBioSource, San Diego, CA, USA) and nephrin (AF3159, R&D Systems, Minneapolis, MN, USA) were diluted in immunohistochemistry diluent (ZLI-9029, Zhongshanjinqiao, Beijing, China) and applied overnight. Sections were subsequently incubated with secondary peroxidase conjugated antibody (Kit-5004, Maixin-Bio, Fuzhou, China; Kit-5107, Maixin-Bio, Fuzhou, China), developed using 3, 3′-diaminobenzidine (DAB) substrate (Kit-0016, Maixin-Bio, Fuzhou, China), and then minimally counterstained with hematoxylin. For double immunohistochemical staining, nephrin staining was performed first, followed by Nup160 staining. The secondary antibodies used were a horseradish peroxidase-conjugated anti-goat antibody (Kit-5107, Maixin-Bio, Fuzhou, China) and alkaline phosphatase-conjugated anti-rabbit antibody (Kit-5101, Maixin-Bio, Fuzhou, China), respectively. Staining was developed using DAB (Kit-0016, Maixin-Bio, Fuzhou, China) and alkaline phosphatase-Red substrate (Kit-8814, Maixin-Bio, Fuzhou, China), respectively. Sections were then minimally counterstained with hematoxylin.

Transmission electron microscopy

Fresh kidney tissue rich in renal cortex was quickly cut into 1–3 mm³ blocks, immersed in 2.5% glutaraldehyde for 24 h, and then fixed at 4°C followed by 1% osmic acid for 2 h. Lead citrate and uranyl acetate were stained by conventional embedding and slicing. Sections were then observed by transmission electron microscopy (JSM-7500F, JEOL, Ltd, Akishima, Japan).

Statistical analyses

Statistical analyses were performed using SPSS 24.0 software (IBM SPSS, Chicago, IL, USA). Statistical significance was confirmed at P < 0.05. Quantitative data are presented as mean ± SEM. The specific tests performed are shown in the figure legends.

Acknowledgements

We thank Professor Jie Ding for her help with the grants and revising the manuscript, and Professor Xi Mo for her help with the grants. We thank Saiye (Suzhou) Biotechnology Co., Ltd, for their assistance in constructing the experimental mice. We also thank Liwen Bianji, Edanz Group China (http://www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.

Author contributions

Y.L., C.X., F.Z., Q.L., X.Q., M.L., Y.Y., S.Y., H.T., L.Z., B.C., L.Q., and Z.Y. were responsible for formal analysis; Y.L., C.X., F.Z., Q.L., X.Q., M.L, Y.Y., S.Y., H.T., L.Z., B.C., L.Q., and Z.Y. were responsible for methodology; Y.L. and C.X. were responsible for experimentation; Y.L., C.X., and Z.Y. wrote the original draft; Y.L., C.X., and Z.Y. were responsible for data curation; Q.L., M.L., S.Y., L.Z., and L.Q. were responsible for renal pathology analysis; Z.Y. was responsible for project administration; Z.Y. was responsible for funding acquisition; all authors reviewed and edited the manuscript.

Conflict of interest statement: The authors declare that they have no competing interests.

Funding

This work was supported by grants from the National Nature Science Foundation of China [81270766 and 82170718]; joint funds for the Innovation of Science and Technology, Fujian province [2020Y9158]; and the Natural Science Foundation of Fujian Province of China [2021J01411].

Author notes

Yuanyuan Li and Chan Xu contributed equally to this work.