Cell-Autonomous Hedgehog Signaling Is Not Required for Cyst Formation in Autosomal Dominant Polycystic Kidney Disease

Significance Statement

Polycystin proteins function in renal primary cilia to regulate a pathway important to kidney functional homeostasis. Autosomal dominant polycystic kidney disease (ADPKD) results when polycystin genes are mutated, but key cilia-based signaling pathways dysregulated by loss of polycystin function require elucidation. One cilia-dependent pathway suggested as playing a role in polycystic kidney diseases is the Hedgehog pathway, which is important in development, cancer, and ciliopathies. In mouse models of ADPKD, the authors used conditional control of gene expression to upregulate or inactivate Hedgehog in kidney cells that had undergone inactivation of Pkd1, the gene encoding Polycystin-1. They found that the Hedgehog pathway had no significant effect on polycystic kidney disease initiated by Pkd1 inactivation, excluding it as a functional cell-autonomous component of kidney cyst formation in these ADPKD models.

Visual Abstract

Abstract

Background

PKD1 or PKD2, the two main causal genes for autosomal dominant polycystic kidney disease (ADPKD), encode the multipass transmembrane proteins polycystin-1 (PC1) and polycystin-2 (PC2), respectively. Polycystins localize to the primary cilium, an organelle essential for cell signaling, including signal transduction of the Hedgehog pathway. Mutations in ciliary genes that build and maintain the cilium also cause renal cystic disease through unknown pathways. Although recent studies have found alterations in Hedgehog signaling in ADPKD-related models and tissues, the relationship between Hedgehog and polycystic kidney disease is not known.

Methods

To examine the potential role of cell-autonomous Hedgehog signaling in regulating kidney cyst formation in vivo in both early- and adult-onset mouse models of ADPKD, we used conditional inactivation of Pkd1 combined with conditional modulation of Hedgehog signaling components in renal epithelial cells, where mutations in Pkd1 initiate cyst formation. After increasing or decreasing levels of Hedgehog signaling in cells that underwent inactivation of Pkd1, we evaluated the effects of these genetic manipulations on quantitative parameters of polycystic kidney disease severity.

Results

We found that in Pkd1 conditional mutant mouse kidneys, neither downregulation nor activation of the Hedgehog pathway in epithelial cells along the nephron significantly influenced the severity of the polycystic kidney phenotype in mouse models of developmental or adult-onset of ADPKD.

Conclusions

These data suggest that loss of Pkd1 function results in kidney cysts through pathways that are not affected by the activity of the Hedgehog pathway.

Autosomal dominant polycystic kidney disease (ADPKD) is one of the most common monogenic diseases, affecting around 12 million people worldwide. Affected individuals develop clinically discernible kidney cysts by 30 years of age. Cysts increase in size and number over time in a process that affects kidney function, and one half of the patients progress to ESKD before age 60 years old. The two main causal genes for ADPKD are PKD1 and PKD2,^1,2 which encode the membrane proteins Polycystin-1 (PC1) and Polycystin-2 (PC2), respectively. PC1 and PC2 interact through both extracellular and C-terminal tail domains^3,4 to form a complex that localizes to the primary cilium.^5,6 The primary cilium is a tubulin-based organelle on the surface of the cell that serves as a cellular sensor thought to transduce both mechanical and chemical inputs.⁷ However, the physiologic roles of PC1 and PC2 within in the primary cilium and the molecular pathways disrupted by mutations in these proteins that cause renal cystogenesis are not well understood.

Mutations in ciliary genes, which encode proteins that build and maintain the cilium, also cause renal cystic disease as part of a spectrum of diseases called ciliopathies that affect multiple organ systems. Mutations in ciliary genes result in renal cysts in patients as well as in model systems.^7–11 The molecular pathways disrupted to cause renal cysts in this ciliopathic form of kidney cystic disease are not known. The signaling pathway most definitively associated with cilia is the Hedgehog pathway, and disruption of cilia structure or function causes defects in the Hedgehog pathway.¹² Yet, it has not been established whether aberrant Hedgehog signaling due to ciliary disruption is causative for renal cysts. Like PC1 and PC2, Hedgehog signaling components localize to the primary cilium in mammalian cells. The Hedgehog receptor Patched localizes to the ciliary membrane. Binding of Hedgehog ligand releases the inhibition exerted by Patched on Smoothened (Smo), a seven-pass transmembrane protein, allowing the translocation of Smo into the cilium.^13,14 Smo activation transduces Hedgehog signaling through the Gli transcription factors, the effectors of the Hedgehog pathway.¹⁵ Gli proteins require trafficking through the cilium to be processed in response to Hedgehog pathway signaling. Processing of both the activator and the repressor forms of Gli transcription factors requires a functional cilium, and mutations in cilia genes can result in either overactivation or downregulation of the Hedgehog pathway.^16,17 Recently, several studies have suggested the involvement of the Hedgehog pathway in cyst progression in ADPKD. Expression of Hedgehog signaling components was found to be increased in cystic Pkd1 mutant mouse kidneys and human ADPKD kidney samples.^18,19 Moreover, pharmacologic inhibition of the Hedgehog pathway reduced cell proliferation and cystic phenotypes in in vitro assays.^18,19 These results suggested that Hedgehog signaling was increased in human ADPKD and that suppression of the pathway could reduce cyst formation in culture and potentially, in vivo.

Studies that sought to define the genetic relationship between cilia and ADPKD demonstrated the remarkable finding that deletion of cilia was protective for cyst formation in ADPKD mouse models,²⁰ indicating the existence of a cilia-dependent cyst activation (CDCA) signal.²¹ The identity of the CDCA is not known. Recently, the disruption of trafficking within intact cilia in the kidney epithelium by deletion of the ciliary gene Tubby like protein-3 (Tulp3) was shown to confer a dosage-dependent protective effect on cystogenesis in Pkd1 mutant mice.¹¹ Both cilia ablation and Tulp3 deletion have been shown to disrupt Hedgehog signaling in a Gli-dependent manner.^12,22–24 These genetic studies in the mouse are consistent with the Hedgehog pathway being a possible candidate to regulate the CDCA that drives rapid cyst grown in the absence of polycystin function. However, the role of Hedgehog pathway to initiate or influence cyst formation and growth in ADPKD models in vivo has not been tested.

In this study, we examined the role of cell-autonomous Hedgehog signaling in kidney cyst formation and progression in Pkd1 mouse models of ADPKD. We used conditional control of gene expression of Hedgehog signaling components (Smo, Gli2, and Gli3) in vivo to either increase or decrease levels of Hedgehog signaling in cells that underwent inactivation of Pkd1. We found that neither activation nor repression of the Hedgehog pathway in epithelial cells along the nephron significantly influenced the progression of polycystic kidney disease in mouse models of developmental or adult onset of ADPKD. These results suggest that signaling through the Hedgehog pathway is not required for cystic phenotype caused by loss of function of Pkd1.

Methods

Mouse Husbandry

Pkhd1^Cre,²⁵ Pax8^rtTA,²⁶ TetO-cre,²⁷ Pkd1^fl,²⁸ Smo^fl,²⁹ ROSA^{LSL-GFP-Smo-M2},³⁰ Gli2^fl,³¹ and Gli3^fl32 have been described previously. Mouse lines in this study that used Pax8^rtTA; TetO-cre transgenic system for gene deletion were administered doxycycline for 2 weeks beginning at postnatal day 28 (P28) for adult induction or P0 for postnatal induction. At the study end points, mice were euthanized, and kidneys and serum were collected as described previously.²⁰ Experiments were carried out under protocols approved by the Yale University Institutional Animal Care and Use Committee in accordance with National Institutes of Health guidelines for the ethical treatment of animals.

Doxycycline Administration

Doxycycline solution was made from drinking water supplemented with 2 mg/ml doxycycline hyclate (D9891; Sigma-Aldrich) and 3% sucrose (S-0389; Sigma). Animals treated at P28 were switched from regular drinking water to doxycycline solution for 2 weeks and then returned to regular drinking water until the study end point. Pups from P0 to P14 received doxycycline through their mother’s milk; the nursing mother was switched from drinking water to doxycycline solution from the day of birth of its pups (P0) for 2 weeks (P14; study end point).

Cystic Index and Serum Studies

Measurement of cystic index is done as reported previously.²⁸ BUN and creatinine concentrations were measured by the George M. O’Brien Kidney Center at Yale core facility.

Fluorescence Imaging

Markers were goat anti-Aquaporin-2 (sc-9882; 1:50 dilution; Santa Cruz Biotechnology), mouse anti-acetylated alpha tubulin (T-7451; 1:500 dilution; Sigma-Aldrich), and Hoechst 33342 (1:10,000 dilution; Molecular Probes). Images were taken on a Nikon Eclipse TE2000-U microscope driven by MetaMorph software (Universal Imaging).

PCR to Confirm Recombination of Floxed Alleles by Cre Recombinase

For genomic DNA extraction, kidney samples were digested in 200 mM NaCl, 0.2% SDS, 5 mM EDTA, 100 mM Tris (pH 8.5), and Proteinase K (100 μg/ml) overnight at 55°C, and DNA was purified by phenol/chloroform and precipitated in ethanol. For Smo genomic PCR, three primers were designed on the basis of previous studies²⁹: SmoFF2: 5′-GTTTGCAAAGTTGGGAGTCG-3′, SmoFR2: 5′-CAAACAGCCAACTCAGCAAA-3′, and SmoDR1 5′-CCTTCTCCAACACGCAGAGT-3′. SmoFF2 is located 140 bp upstream of the single loxP site (44 bp upstream of the ATG start codon), and SmoFR2 is located 40 bp downstream of the same loxP site. SmoDR1 is located 150 bp downstream of the loxP-flanked region. Floxed allele size is 215 bp, and recombined floxed allele size is 325 bp (PCR conditions: 94°C 5 min followed by 35 cycles of 94°C 30 s, 55°C 30 s, and 72°C 45 s and a final extension at 72°C 7 min). For Gli2 genomic PCR, three primers were used (Gli2 floxA 5′-CTTATGGACATCTGTCTGCC-3′, Gli2 floxD 5′-GAGACTCCAAGGTACTTAGC-3′, which detected the recombined floxed allele,³¹ and Gli2 Rcont 5′-GCTCTGGATACTTGTGCTATGCC-3′) to detect the nonrecombined floxed allele. Floxed allele size is 1000 bp, and the recombined allele size is 500 bp (PCR conditions: 94°C 5 min followed by 35 cycles of 94°C 30 s, 58°C 30 s, and 72°C 1 min and a final extension at 72°C 10 min). For Gli3 genomic PCR, two primers were used as previously described³²: S1, 5′-CTGGATGAACAAGCTTTCCATC-3′ and AS2, 5′-CAGTAGTAGCCTGGTTACAG-3′. Floxed allele size is 1500 bp, and recombined floxed allele size is 300 bp (PCR conditions: 94°C 5 min followed by 30 cycles of 94°C 30 s, 60°C 1 min, and 72°C 1 min and a final extension at 72°C 10 min).

Sample Size, Power Calculation, and Statistical Analyses

Sample size and power calculation were carried out using STPLAN v.4.5 (http://biostatistics.mdanderson.org/SoftwareDownload). We selected 80% power as our threshold and based the calculations on changes in kidney weight-to-body weight ratio (KW/BW), which is a well established metric for efficacy in in vivo models of ADPKD. We calculated the percentage change in KW/BW from that observed in the Pkd1-only knockout in each experiment that would achieve 80% power to detect a statistically significant change at a threshold of P<0.05. The percentage changes in KW/BW from the Pkd1-only knockout to achieve this power are as follows: Figure 1B, 45% change; Figure 1D, 69% change; Figure 2B, 27% change; Figure 2D, 40% change; and Figure 3B, 64%–68% change. The allocation of animals to each group was solely on the basis of their genotype irrespective of sex and without any exclusions. The calculations show that we had power to detect as significant changes ranging from 27% to 69% depending on the experiment. Because the KW/BWs of Pkd1 conditional knockout kidneys in our studies ranged from 8.5% to 14.9% depending on the experiment, with all but one group having KW/BW>13%, a reduction in KW/BWs from 13% to a range of 4%–9.5% would be sufficient to demonstrate an effect of Hedgehog signaling manipulation. Data were analyzed by the Mann–Whitney test for comparison of two datasets and one-way ANOVA followed by Tukey multiple comparison test for multiple comparisons (GraphPad Prism 7). P<0.05 was considered as the significance threshold. All data are presented as mean±SEM.

Figure 1.

Concomitant inactivation of Smoothened does not change the course of cyst progression in Pkd1 mice. (A) Images of representative kidneys from the perinatal model with the indicated genotypes at postnatal day 24. (B) Aggregate data of the kidney weight-to-body weight ratio, cystic index, and BUN for the indicated genotypes. The color blocks in A correspond to the data in B. Comparison using one-way ANOVA followed by the Tukey multiple comparison test and presented as mean±SEM. ****P<0.001 for both cystic models compared with the Smo^fl/fl; Pkhd1^Cre control. (C) Images of representative kidneys from the adult inducible model with the indicated genotypes at 18 weeks. (D) Aggregate data of the kidney weight-to-body weight ratio, cystic index, and BUN. The color blocks in C correspond to the data in D. Comparison of two datasets using the Mann–Whitney test, presented as mean±SEM, showed no significant differences between the two genotypes. Scale bars, 2 mm in A and C.

Figure 2.

Activation of Hedgehog pathway by overexpressing Smoothened-M2 has no effect on cyst progression in autosomal dominant polycystic kidney disease models. (A) Images of representative kidneys from the perinatal collecting duct knockout model with the indicated genotypes at postnatal day 24. (B) Aggregate data of the kidney weight-to-body weight ratio, cystic index, and BUN from the indicated number of mice. The color blocks in A correspond to the data in B. Comparison of two datasets using the Mann–Whitney test, presented as mean±SEM, showed no significant differences in kidney weight-to-body weight ratio, cystic index, or BUN. The color blocks in A correspond to the data in B. (C) Images of representative kidneys of adult inducible whole-nephron knockouts with the indicated genotypes at 18 weeks. (D) Aggregate data of the kidney weight-to-body weight ratio, cystic index, and BUN from the indicated number of mice. The color blocks in C correspond to the data in D. Comparison of two datasets using the Mann–Whitney test, presented as mean±SEM, showed no significant differences between the two models. Scale bars, 2 mm in A and C.

Figure 3.

Inactivation of Gli2 and Gli3 does not affect cyst formation in a Pkd1 model. (A) Representative images of kidneys with the indicated genotypes. Doxycycline was administered from postnatal day 0 (P0) to P14, and kidneys were analyzed at P14. (B) Aggregate data of the kidney weight-to-body weight ratio, cystic index, BUN, and serum creatinine concentrations of different genotypes. The color blocks in A correspond to the data in B. The comparison used one-way ANOVA followed by Tukey multiple comparison test and is presented as mean±SEM. To facilitate the breeding of the multiple alleles and the use of littermates as controls, we used mice heterozygous for Gli2^fl or Gli3^fl as controls and compared with Gli2^fl/fl and/or Gli3 ^fl/fl. Scale bar, 2 mm. **P=0.002 for the comparison of each of the four cystic models with the wild-type control; ****P<0.001 for the comparison of each of the four cystic models with the wild-type control.

Results

Deletion of Smo in Renal Epithelium Does Not Affect Cystic Phenotypes in Pkd1 Mutant Mouse Models

We tested the role of the Hedgehog pathway in kidney cyst formation in mouse models of Pkd1 deletion. We specifically examined the cell-autonomous genetic requirement of the pathway by modulating Hedgehog signaling only in cells that also had undergone Pkd1 deletion. First, we deleted Smo, an activator of the Hedgehog pathway, in two models of ADPKD on the basis of Pkd1 deletion—an early developmental model and an adult-onset inducible model. For the early model, we used Pkhd1^Cre, which is active along the entire collecting duct by P7.²⁸ Deletion of Smo prevents canonical, ligand-dependent activation of the Hedgehog pathway. We used a floxed allele of Smo,²⁹ because it has been shown to efficiently delete Smo activity on cre-mediated recombination in a variety of tissue contexts.^33–37 Early postnatal deletion of Smo alone in the collecting duct did not induce discernible effects on the kidney and did not result in cysts (Figure 1, A and B). We next determined whether inactivation of Hedgehog signaling could modify cyst formation due to Pkd1 deletion in this model. Deletion of either Pkd1 alone or Smo and Pkd1 together led to strong cystic phenotypes at P24 that were indistinguishable when evaluated by KW/BW, cystic index, and BUN levels (Figure 1, A and B). To confirm the efficiency of the cre-mediated deletion of the floxed Smo allele, we evaluated the efficiency of Smo deletion in the kidneys of Smo; Pkd1 double mutants by PCR on genomic DNA extracted from kidneys. The primers were designed to amplify both recombined and unrecombined Smo alleles. As expected, we found that the Smo recombined allele was predominantly amplified in the kidneys at P24, whereas the Smo recombined allele was not detected in control kidneys that did not carry a Cre allele (Supplemental Figure 1A).

We next examined the effect of inactivating Hedgehog signaling along the entire nephron in an adult-onset ADPKD model. We used Pax8^rtTA; TetO-cre bitransgenic lines that allowed for the temporally controlled deletion of alleles on doxycycline-induced Cre-mediated recombination. We administrated doxycycline for 2 weeks from P28 to P42 to inactivate Pkd1 alone or Smo and Pkd1 together, and we examined the resulting phenotype at 18 weeks of age. We selected this time course on the basis of our previous studies using the same models that demonstrated a strong cystic phenotype in the Pkd1 mice and a protective effect of cilia ablation²⁰ or Tulp3 deletion.¹¹ Smo; Pkd1 double mutants developed a comparable cystic phenotype to that of Pkd1 single mutants alone (Figure 1, C and D). Again, the majority of cells seemed to carry the recombined Smo allele as assessed by PCR in the Smo; Pkd1 double-mutant kidneys, whereas the recombined Smo allele could not be detected in control kidneys (Supplemental Figure 1B). Therefore, blockade of ligand-dependent activation of the Hedgehog pathway by conditional deletion of Smo did not induce kidney cysts, and it did not influence kidney cyst formation in Pkd1 conditional mutants in either early developmental or adult-onset ADPKD models.

Activation of Hedgehog Pathway in Renal Epithelial Cells Does Not Influence Renal Cystic Phenotypes in a Pkd1 Model

We next sought to test whether activation of the Hedgehog pathway could affect cyst progression in Pkd1 mouse models. We used the ROSA^{LSL-GFP-Smo-M2} allele that expresses a constitutively active GFP-fused form of Smo (Smo-M2) after Cre-mediated recombination.^30,38 We induced the expression of Smo-M2 at the same time as deletion of Pkd1 using Pkhd1^Cre and examined kidneys at P24. Expression of Smo-M2 alone during postnatal development using Pkhd1^Cre did not generate cysts (data not shown), consistent with a previous report.³⁹ Concomitant expression of Smo-M2 and inactivation of Pkd1 in the early postnatal collecting duct did not alter the kidney cystic phenotype compared with Pkd1 deletion alone (Figure 2, A and B). We also tested whether ligand-independent activation of the Hedgehog pathway could influence polycystic disease in the slower-progressing adult-onset ADPKD model. We induced expression of Smo-M2 in Pkd1 mutant mice using the Pax8^rtTA; TetO-cre system and again, found that constitutive Smo activation did not influence the cystic phenotype in adult-onset Pkd1 mutant mice (Figure 2, C and D). To ensure that the absence of a discernible effect on kidney cysts was not due to the lack of expression of Smo-M2, we examined expression of GFP-SmoM2 by immunofluorescence. We found robust expression of GFP-SmoM2 throughout the kidney, indicating expression of the constitutively activated Smo allele (Supplemental Figure 2). These results show that Hedgehog pathway activation through Smo activation did not significantly influence Pkd1-mediated polycystic kidney disease progression in either the developmental or adult mouse models of ADPKD.

Pkd1 Inactivation Does Not Cause Cystic Disease through Gli-Regulated Pathways

Canonical Hedgehog signaling functions through Gli transcription factors. Ciliopathy-associated genes that cause renal cysts when mutated disrupt the Hedgehog pathway at a step downstream of Smo but upstream of Gli genes.⁴⁰ This raised the possibility that Pkd1 mutations could influence the Hedgehog pathway at a step downstream of Smo to induce cysts. To address this possibility, we directly tested whether deletion of Gli genes affected polycystic kidney disease progression in Pkd1 mutants through double- and triple-mutant analysis. In these studies, we used the Pax8^rtTA; TetO-cre system²⁶ and administered doxycycline from P0 to P14 to concomitantly delete Pkd1 and Gli genes. We selected P14 for kidney analysis, an experimental time course on the basis of our previous studies, which showed a strong cystic phenotype in the developmental model of ADPKD¹¹—a model that develops cysts more rapidly than Pkd1 deletion in the adult.⁴¹ We analyzed kidney phenotypes at P14, a time point when Pkd1 deletion resulted in a severe polycystic phenotype with significantly increased KW/BW, cystic index, and BUN concentration (Figure 3). For these studies, we also measured serum creatinine as an additional assessment of kidney function (Figure 3B). In all experiments, we used heterozygous littermates as controls for phenotypic comparison. We used double-heterozygous Gli2^fl/+; Gli3^fl/+; Pkd1^fl/fl; TetO-cre; Pax8^rtTA animals as Pkd1-deleted cystic reference animals.

Gli2 is the main transcriptional activator of the Hedgehog pathway. We used a floxed allele of Gli2³¹ that has been shown to successfully delete Gli2 upon cre recombination in a variety of tissue contexts.^42–45 Pkd1; Gli2 double mutants showed a cystic phenotype that was not different from Pkd1 single mutants, showing that Pkd1 deletion did not cause cysts through Hedgehog pathway activation by Gli2 (Figure 3). We next tested the role of Gli3, which generally functions as a repressor of the Hedgehog pathway. We used a floxed allele of Gli3³² that has been previously shown to robustly delete Gli3 on cre-mediated recombination in a variety of tissue contexts.^46–50 Pkd1; Gli3 double mutants also displayed a similar cystic phenotype as Pkd1 single mutants at P14 (Figure 3). These results indicated that Pkd1 deletion did not cause cysts through Gli3-dependent regulation of the Hedgehog pathway. Because all canonical Hedgehog signaling is mediated through Gli proteins, the absence of both Gli2 and Gli3 is considered the null state of the Hedgehog pathway where no activation or repression can occur, because Gli1, a weak activator, is not thought to be expressed in Gli2; Gli3 double-mutant mice.⁵¹ We next generated Pkd1; Gli2; Gli3 triple mutants and analyzed the kidney phenotypes. Triple mutants had a similar cystic phenotype as Pkd1 mutants alone (Figure 3), showing that the polycystic kidney disease resulting from Pkd1 deletion did not require cell-autonomous Hedgehog signaling to cause cysts. Again, we confirmed efficient deletion of Gli2 and Gli3 in the kidney by PCR on genomic DNA extracted from kidneys, which showed strong cre-mediated recombination of Gli2 and/or Gli3, independent of the number of floxed alleles (Supplemental Figure 1, C and D). These results showed that Pkd1 deletion did not cause cyst formation through Gli-dependent regulation of the Hedgehog signaling pathway.

Discussion

Signaling at the primary cilium is intricately associated with kidney homeostasis, because mutations in proteins that localize to the cilium (e.g., PC1 and PC2) or proteins that regulate cilia biogenesis and composition result in cystic kidney disease.⁸ The functional relationship between ciliary signaling pathways and polycystins is complex, because deletion of cilia has a paradoxically protective effect on renal cyst formation in Pkd1 and Pkd2 mouse models of ADPKD, indicating the existence of a CDCA signal.²⁰ Disruption of the polycystin complex at the ciliary membrane results in changes in the signaling properties of kidney epithelial cells, which likely include changes in the privileged signaling microenvironment of the cilia. This includes changes in ciliary and cellular levels of secondary messenger levels, which could result in proliferation and changes in the secretory properties of cyst lining cells.⁵² Despite the key role of cilia-mediated pathways, the signaling events that initiate cystogenesis remain not well understood. The identification of key cilia-based signaling pathways that are dysregulated by loss of function of polycystins is vital for our understanding of polycystic kidney diseases.

A candidate for the CDCA pathway that could be dysregulated in the context Pkd1 loss of function is the Hedgehog signaling pathway. Previous reports have associated Hedgehog signaling and polycystic kidney disease^18,19 by demonstrating an increased expression of Hedgehog targets in Pkd1 mutant whole-kidney lysates¹⁹ and cyst lining cells in human ADPKD sections,¹⁸ which supports the association of altered expression of Hedgehog components with loss of function of Pkd1. In addition, they show the ability of Hedgehog inhibitors to affect in vitro cyst growth in Pkd1 mutant mouse kidney explants or PKD1 mutant cultured cells.^18,19 However, these studies did not address whether the associated expression changes of Hedgehog signaling components were causative for the disease in vivo. Another recent study reported phenotypic improvement with systemic administration of the Smo inhibitor cyclopamine in both the liver and kidney of the pck rat model of autosomal recessive polycystic kidney disease.⁵³ However, these studies did not directly address the functional role of Hedgehog in ADPKD, and they did not address the relationship between Hedgehog signaling and the cell-autonomous functioning of PC1-mediated cilia signaling.

We, therefore, tested the role of the Hedgehog signaling pathway in mediating cystogenesis in murine models of ADPKD due to loss of Pkd1 function. Our study directly addresses the cell-autonomous role of Hedgehog signaling in the cyst initiation and cystic growth process after inactivation of Pkd1 in vivo in both early developmental and adult-onset models and both collecting duct-selective and whole-nephron knockouts. Our approach differed from previous analyses, because we used conditional genetics in the mouse to target renal epithelial cells, where mutations in Pkd1 initiate cyst formation. In all models, genetic manipulation of the Hedgehog pathway along with Pkd1 deletion resulted in kidneys with a strong cystic phenotype on the basis of all of the quantitative parameters used to assess kidney morphology and function. Using sample size and power calculations, we show that we have reasonable power to detect significant effect of Hedgehog on ADPKD and can conclude that neither repression nor activation of the Hedgehog pathway significantly altered the polycystic kidney phenotype in the multiple ADPKD models tested.

Importantly, our results also suggest that, if Hedgehog signaling contributes to the CDCA signal, it is not a major factor in the ADPKD models that we tested. Although we found some variation in the cystic phenotypes among the animals studied, this was very small compared with the strong rescue of the Pkd1 mutant cystic phenotype previously observed on ablation of the cilium²⁰ or disruption of ciliary trafficking by deletion of Tulp3.¹¹ Because cyst formation was unaffected by genetic manipulation of Smo and Gli genes to activate or repress the Hedgehog pathway in vivo, we conclude that Hedgehog signaling in the kidney tubule epithelium is unlikely to play a significant cell-autonomous role in polycystic kidney disease caused by mutations in Pkd1.

Disclosures

Dr. Somlo reports personal fees from Goldfinch Bio outside the submitted work.

Funding

Dr. Ma was supported by PKD Foundation Research Grant-in-Aid 196G14a. This work was supported by Yale PKD Center pilot grant P30 DK090744, PKD Foundation Research Grants-in-Aid 199G14 and 232G18, American Society of Nephrology Foundation for Kidney Research Norman Siegel Research Scholar Grant and National Institute of Health grant NS097928 (to Dr. Liem). This work was supported by National Institutes of Health grants DK54053 (to Dr. Somlo) and DK100592 (to Dr. Somlo).

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2018121274/-/DCSupplemental.

Supplemental Figure 1. Recombination of Smo, Gli2, and Gli3 floxed alleles.

Supplemental Figure 2. Smo-M2 expression after Cre recombination.