Functional polycystin-1 dosage governs autosomal dominant polycystic kidney disease severity

Abstract

Autosomal dominant polycystic kidney disease (ADPKD) is caused by mutations to PKD1 or PKD2, triggering progressive cystogenesis and typically leading to end-stage renal disease in midlife. The phenotypic spectrum, however, ranges from in utero onset to adequate renal function at old age. Recent patient data suggest that the disease is dosage dependent, where incompletely penetrant alleles influence disease severity. Here, we have developed a knockin mouse model matching a likely disease variant, PKD1 p.R3277C (RC), and have proved that its functionally hypomorphic nature modifies the ADPKD phenotype. While Pkd1+/null mice are normal, Pkd1RC/null mice have rapidly progressive disease, and Pkd1RC/RC animals develop gradual cystogenesis. These models effectively mimic the pathophysiological features of in utero-onset and typical ADPKD, respectively, correlating the level of functional Pkd1 product with disease severity, highlighting the dosage dependence of cystogenesis. Additionally, molecular analyses identified p.R3277C as a temperature-sensitive folding/trafficking mutant, and length defects in collecting duct primary cilia, the organelle central to PKD pathogenesis, were clearly detected for the first time to our knowledge in PKD1. Altogether, this study highlights the role that in trans variants at the disease locus can play in phenotypic modification of dominant diseases and provides a truly orthologous PKD1 model, optimal for therapeutic testing.

Figure 1. Generation of the Pkd1 p.R3277C knockin mouse.

(A) Diagram of the targeting construct used for injection into ESCs. Restriction sites flanking the construct are in green (Sb, SbfI; Mo, MboI). The Pkd1 p.R3277C mutation was introduced using the exact codon found in ADPKD probands (29, 59) (red). Restrictions sites used for introducing the mutation and the loxP-flanked (red triangles) selection cassette (yellow box) are in orange (A, AdhI; X, XhoI; S, SalI). The restriction site used to linearize the targeting construct is shown in purple (P, PacI). Restriction sites and probes used for Southern blotting are in blue (Bs, BspHI; M, MfeI). (B) Representative Southern blot of successful HR in ESCs. (C) Sequencing chromatogram of WT and heterozygous ESC clones and a homozygous animal. (D) RT-PCR analysis across the mutation and the loxP-containing intron of a WT and a homozygous animal (after Cre-lox recombination). A normally seen alternative splice product, skipping exon 31 (in-frame change), was not significantly different between WT and homozygous animals (asterisk) (105) (PCR product, 762 bp; splice variant, 669 bp). (E) Quantification of PC-1 protein levels in whole kidney and CD cell lysate. PC-1 protein levels between WT and Pkd1^RC/RC animals were not significantly different (WB not shown, see Figure 9 for more detail). Data are shown as a normalized ratio (loading control, γ-tubulin [TUBG1]) between WT and Pkd1^RC/RC animals (n = 3/group). Statistical values were obtained by Student’s t test; error bars indicate ± SD.

Figure 2. Homozygosity of the Pkd1 p.R3277C allele results in progressive PKD with physiological characteristics comparable to those of human ADPKD.

(A) Representative US M-mode images of Pkd1^RC/RC mice. Kidneys with the mildest (left) and most severe disease (right) are shown at 3 months. Kidney cysts are visible as echolucent spots (arrows). Scale bars: 10 mm (left); 12 mm (right). (B) Quantification of US at 3, 6, 9, and 12 months of each individual animal and the mean of all animals. Significant differences between WT and homozygotes were seen at all time points. KV increased in most animals progressively, but variability among mice was observed. (C–E) Graphic representations of %KW/BW, cAMP, and BUN measured at 3, 6, 9, and 12 months (Supplemental Table 1). (C) %KW/BW increased progressively with cyst burden until 9 months, after which a significant decrease was observed, likely correlated with increased fibrosis (Table 1). (D) Similar to that in patients with ADPKD, cAMP levels rose with disease burden (16). (E) Increased kidney damage due to cystogenesis/fibrosis resulted in a significantly elevated BUN after 9 months. The data in C–E were obtained from the same group of animals at each time point (WT, n = 4; Pkd1^RC/RC, n = 6). Statistical values were obtained by the Student’s t test (*P < 0.05, **P < 0.01, ***P < 0.001); error bars indicate ± SD.

Figure 3. Histology of homozygous Pkd1 p.R3277C animals mimics typical ADPKD.

(A) Representative Masson Trichrome histology sections of Pkd1^RC/RC and WT kidneys at 1 to 12 months show that cysts developed progressively with age. At later time points, an increased level of fibrosis was noted (12 months; Table 1). Scale bars: 1 mm (kidney cross-section, WT, and Pkd1^RC/RC); 250 μm (Pkd1^RC/RC magnified region [images in second column]); 200 μm (WT magnified region [insets]). (B) Masson Trichrome histology sections of 12-month-old Pkd1^RC/RC and WT livers (no significant difference between percentage of liver weight [LW] per body weight [%LW/BW] of WT animals and homozygotes was noted). Four out of five homozygous animals showed ductal plate malformations, including microhamartomas of varying severity. These cystic lesions are progenitors of the liver cysts often found in patients with ADPKD (69). Scale bars: 100 μm.

Figure 4. Pkd1^RC/null mice develop in utero/early-onset ADPKD.

(A) Gross images of WT and Pkd1^RC/del2 kidneys from P1 to P180, illustrating early-onset/rapidly progressive disease. After P25, the kidney size decreased, likely due to increased fibrosis (Figure 5 and Table 1). The ruler depicts centimeters. (B) Kaplan-Meier curve of WT and Pkd1^RC/del2 mice. Eighty percent of Pkd1^RC/del2 mice died before P50 (median survival, P28 [arrow]; n = 55/group). (C–E) Graphical representations of %KW/BW, BUN, and cAMP from P1 to P25 of Pkd1^RC/del2 animals (Supplemental Table 2). (C) %KW/BW was significantly increased as early as P1 (3%) and rapidly climbed to 25% by P25. (D) BUN levels rose beyond the physiological average as early as P18. (E) cAMP levels were significantly above normal by P1. Values in C–E were obtained from n ≥ 5 at each time point for both WT and Pkd1^RC/del2 mice. P180 data can be found in Supplemental Table 2. Statistical values were obtained by the Student’s t test (*P < 0.05, **P < 0.01, ***P < 0.001); error bars indicate ± SD.

Figure 5. Histology of Pkd1^RC/del2 mice highlights early-onset and rapidly progressive cystic disease.

Representative Masson Trichrome sections of Pkd1^RC/del2 and WT kidneys from E14.5 to P180. Kidney cysts were clearly visible as early as E16.5 (a few cysts were present at E15.5; Supplemental Figure 2) and rapidly increased in number/size with age (Table 1). Glomerular cysts were present as early as E16.5 (insets) (59). The kidneys of animals with longer survival (P180) were highly fibrotic, with little obvious healthy tissue. Scale bars: 100 μm (E14.5 to P1); 500 μm (P12); 750 μm (P25 and P180); 50 μm (insets).

Figure 6. Pkd1 p.R3277C mice show a switch in tubular cyst origin corresponding to age.

(A) IF lectin labeling of PT (LTA) and CD (DBA). During early kidney development (P1), cysts originated mainly from PT in Pkd1^RC/del2 and Pkd1^RC/RC animals, with only rare CD cysts (arrowheads). Later in development (P12), a switch to more CD cysts occurred that became more prominent after kidney development (P25, Pkd1^RC/del2; 3 months, Pkd1^RC/RC), with only a few remaining PT cysts (arrows). With disease progression, an increasing number of cysts seemed to dedifferentiate. Asterisks indicate a lack of labeling with tubule markers. Scale bars: 100 μm (P1); 200 μm (P12); 300 μm (P25); 500 μm (3 months and 12 months). (B) Quantification of cyst origin at P1, P12, and P25 for Pkd1^RC/del2 animals and (C) P1, P12, 3 months, and 12 months for Pkd1^RC/RC animals (n = 3; Table 1). Significance is based on cyst number. Statistical values were obtained by the Student’s t test (*P < 0.05, **P < 0.01). Data are not shown for distal tubule (PNA) and loop of Henle (THP), as they accounted for <1% of the total cyst number.

Figure 7. Cystogenesis of Pkd1^RC/del2 mice is associated with an increase in proliferation.

(A) Representative images of kidney tissue from P18 Pkd1^RC/del2 mice, indicating increased proliferation in noncystic and cystic CDs (DBA) compared with that in WT. Reports on associations between cyst expansion and proliferation are controversial, but analysis of the CD showed a clear increase in cell division (36, 37, 79). Scale bars: 20 μm. (B) Graph summarizing data of 100 nondilated CD cysts and 20 small (<50 cells [S]), 10 medium (50–200 cells [M]), and 5 large (>200 cells) CD cysts (Supplemental Table 3) (n = 3/group). Statistical values were obtained by the Student’s t test (*P < 0.05, **P < 0.01, ***P < 0.001); error bars indicate ± SD.

Figure 8. Primary cilia are elongated in Pkd1 p.R3277C CDs but not normal bile ducts.

SEM analysis of primary cilia length (A) in the CDs (Supplemental Table 4) and (B) bile ducts (Supplemental Table 5) of WT, Pkd1^RC/del2, and Pkd1^RC/RC mice at the ages indicated (n = 3 animals/group). (A) Representative SEM images of CDs and individual cilia. Scale bars: 2 μm (bottom row); 10 μm (WT, top row); 15 μm (Pkd1^RC/del2 and Pkd1^RC/RC, top row) (left). Graph summarizing >100 cilia measurements (~40 cilia in nondilated, ~30 in dilated, and ~30 in cystic CDs) in mutant animals and >50 cilia of nondilated CDs in WT animals (P25 and 12 months). The level of functional Pkd1 inversely correlated with cilia length, indicating a role of PC-1 in cilia maintenance (right). (B) Representative SEM image of WT and Pkd1^RC/RC bile duct and one microhamartoma. Bile duct cilia of the microhamartoma were significantly elongated (6.24 ± 1.57 μm, P < 0.001) (left). Scale bars: 1 μm (WT and Pkd1^RC/RC, bottom row); 2 μm (microhamartoma, bottom row); 50 μm (Pkd1^RC/RC, top row); 100 μm (WT and microhamartoma, top row). Graph summarizing >50 cilia measurements of >5 bile ducts per animal (right). Statistical values were obtained by the Student’s t test (**P < 0.01, ***P < 0.001); error bars indicate ± SD.

Figure 9. Pkd1 p.R3277C is a temperature-sensitive folding/maturation mutant.

(A) The PC-1 protein (full-length [FL]) is cleaved into NTPs and CTPs (estimated sizes of glycosylated PC-1 are indicated in parentheses) (–84). The positions of the incompletely penetrant variants (p.R3277C, p.R2220W) and the cleavage mutant p.E2771K are shown (29, 59, 83). NTP*, EndoH-sensitive, immature; NTP**, EndoH-resistant, mature. (B) WB of PC-1 showing the 2 glycoforms in kidney membranes and only the mature form in urinary ELVs. (C) WB of the exogenously expressed cleavage mutant Pkd1 p.E2771K (ex) and WT urinary ELVs (83). The endogenous (end) PC-1 protein is also seen at a low level. The endogenous, mature, cleaved PC-1 (NTP**) runs at a size similar to the ex.FL PC-1 p.E2771K (immature form). (D) WB and cleavage quantification (n = 6 transfections) of exogenously expressed WT and mutant forms of PC-1 shown in A. (E) WB and quantification of PC-1 levels in WT and Pkd1^RC/RC urinary ELVs at 1 and 12 months (n = 4 urine collections of 6 animals/group). ELVs from Pkd2^WS25/null mice show PC-1 levels in Pkd2 cystic kidneys (57). (F) EndoH assay of urinary Pkd1^RC/RC ELVs (analyzed on separate gels as indicated by the white line). (G) WB of Pkd1^RC/RC kidney membrane preparation and urinary ELVs compared with WT. (H) WB and quantification (n = 3 experiments) of PC-1 in cell media–isolated ELVs from primary CD cells isolated from WT and Pkd1^RC/RC P30 kidneys at 37°C and 33°C. PDCD6IP is an ELV expressed control protein (21). E, EndoH; P, PNGaseF. Statistical values were obtained by the Student’s t test (***P < 0.001); error bars indicate ± SD.