Heterozygous Pkhd1^C642* mice develop cystic liver disease and proximal tubule ectasia that mimics radiographic signs of medullary sponge kidney

Abstract

Heterozygosity for human polycystic kidney and hepatic disease 1 (PKHD1) mutations was recently associated with cystic liver disease and radiographic findings resembling medullary sponge kidney (MSK). However, the relevance of these associations has been tempered by a lack of cystic liver or renal disease in heterozygous mice carrying Pkhd1 gene trap or exon deletions. To determine whether heterozygosity for a smaller Pkhd1 defect can trigger cystic renal disease in mice, we generated and characterized mice with the predicted truncating Pkhd1^C642* mutation in a region corresponding to the middle of exon 20 cluster of five truncating human mutations (between PKHD1^G617fs and PKHD1^G644*). Mouse heterozygotes or homozygotes for the Pkhd1^C642* mutation did not have noticeable liver or renal abnormalities on magnetic resonance images during their first weeks of life. However, when aged to ~1.5 yr, the Pkhd1^C642* heterozygotes developed prominent cystic liver changes; tissue analyses revealed biliary cysts and increased number of bile ducts without signs of congenital hepatic fibrosis-like portal field inflammation and fibrosis that was seen in Pkhd1^C642* homozygotes. Interestingly, aged female Pkhd1^C642* heterozygotes, as well as homozygotes, developed radiographic changes resembling MSK. However, these changes correspond to proximal tubule ectasia, not an MSK-associated collecting duct ectasia. In summary, by demonstrating that cystic liver and kidney abnormalities are triggered by heterozygosity for the Pkhd1^C642* mutation, we provide important validation for relevant human association studies. Together, these investigations indicate that PKHD1 mutation heterozygosity (predicted frequency 1 in 70 individuals) is an important underlying cause of cystic liver disorders and MSK-like manifestations in a human population.

INTRODUCTION

The predicted frequency of heterozygosity for polycystic kidney and hepatic disease 1 (PKHD1) mutations is 1 in 70 individuals. Until recently, these individuals were considered unaffected carriers of mutations in this gene. In contrast, homozygosity for PKHD1 mutations is the principal cause of autosomal recessive polycystic kidney disease [ARPKD [Online Mendelian Inheritance in Man (OMIM): 263200]. This disorder is characterized by progressive renal function loss and defects in biliary ductal plate development that leads to Caroli disease or congenital hepatic fibrosis (CHF). Whereas Caroli disease is limited to ectasia of intrahepatic biliary ducts, the manifestations of CHF in addition to the ectatic biliary ducts, also include increase in the number of biliary ducts and presence of inflammatory infiltrates and fibrosis in portal fields in the otherwise intact liver parenchyma. Renal ARPKD manifestations include fusiform dilatation of the cortical and medullary collecting ducts in infants; fusiform/saccular dilatation of the medullary collecting ducts is more typical for older children. However, homozygosity for PKHD1-deficient alleles may not lead to both the liver and kidney manifestations. For example, CHF/Caroli disease patients with two PKHD1 mutations may have normal kidneys or develop medullary sponge kidney (MSK) (13).

The first indication that heterozygous carriers of PKHD1 mutations are at risk of developing cystic liver and kidney abnormalities was provided by studies of parents of ARPKD patients (10). In the 17 reported families, parents of an offspring with known PKHD1 mutation had cystic liver and/or kidney abnormalities (5 of the PKHD1 mutations were truncating). In these parents, the phenotypic manifestations included liver and kidney cysts, hyperechoic liver, gallbladder polyps, increased renal medullary echogenicity resembling nephrocalcinosis, and splenomegaly. However, the phenotypic manifestations were highly variable, and they were not associated with any symptoms at the time of the evaluation. The authors concluded that carrier status for ARPKD (heterozygosity for PKHD1 mutation) is a predisposition to renal involvement associated with increased medullary echogenicity on ultrasound (possibly MSK) and liver involvement in the form of asymptomatic polycystic liver disease and in some cases CHF.

Another recent study found adult carriers of PKHD1 mutations among patients with polycystic liver disease (PCLD). Specifically, enrichment of these mutations was observed on a genome-wide basis among a cohort of 102 unrelated patients with dominantly inherited isolated PCLD that was not associated with mutations in PRKCSH and SEC63, genes that are mutated in autosomal dominant polycystic liver disease (OMIM) 174050)] (4). While the expected number of loss-of-function variants in PKHD1 gene among the 1,266 total variants was 0.48, the observed number among the studied PCLD cohort was 9 (19-fold enrichment, P = 2.31 × 10⁻⁹). Eight out of ten (80%) of the identified PKHD1 abnormalities were truncating mutations. Most of these patients had innumerable small liver cysts (>10-cm cysts in one patient); 30% had kidney cysts. Note that 70% of the patients were females.

However, the relevance of these associations between PKHD1 mutation heterozygosity and cystic liver and kidney phenotypes may have been questioned due to the lack of these manifestations in heterozygous mice carrying existing Pkhd1-inactivating modifications. These models include Pkhd1 exon 1–3 replacement with lacZ reporter gene (Pkhd1^lacZ/+) (15), deletion of exon 2 (Pkhd1^del2/+) (16), and deletion of exon 3–4 (Pkhd1^del3–4/+ (9)). However, mice may not be an optimal species to study such phenotypes, because even homozygotes for the Pkhd1 mutations do not develop the hallmark ARPKD-like renal cystic phenotype.

To determine whether cystic liver or renal disease in mice is triggered by heterozygosity for truncating mutations resembling those seen in human PKHD1 mutation-carrying heterozygotes that developed these phenotypes, we generated a mouse model with a predicted truncating Pkhd1^C642* mutation. This mutation is located in a region corresponding to the middle of exon 20 cluster of five truncating human mutations (between PKHD1^G617fs and PKHD1^G644*). We followed these mice to an age that corresponds to advanced adulthood in humans.

MATERIALS AND METHODS

Generation of the mouse model with Pkhd1^C642* mutation.

The truncating PKHD1 mutation was selected on the basis of mutation characteristics reported at the Mutation Database Autosomal Recessive Polycystic Kidney Disease (http://www.humgen.rwth-aachen.de (2, 3);). The MIT CRISPR server was used to identify the target site in mouse Pkhd1 exon 20 region corresponding to the middle of a region containing five truncating human mutations (between PKHD1^G617fs and PKHD1^G644*). The sequence ATCGGTAAGACTCCAATCAC(AGG) on the negative strand was chosen as a target. Single-guide (sg)RNA and Cas9 mRNA synthesis were performed according to methods reported earlier (6). Wild-type C57BL/6J mice were used to obtain zygotes. Pronuclear injections into the zygotes were performed with a solution of sgRNA (50 ng/μl each) and Cas9 mRNA (100 ng/μl). Injected zygotes were implanted into pseudopregnant CD1 recipients. Genomic DNAs obtained from tail biopsies of putative founder animals (G₀) were assessed for the presence of mutations in the Pkhd1 gene by PCR and heteroduplex mobility assay (HMA). Oligonucleotides flanking exon 20 of MmPkhd1 were used as PCR primers: forward: 5′-GGAATATAGCACCCAGACACTTG-3′; reverse: 5′-AGAATCCTCTCCCACCTTCTC-3′ to amplify a 274-bp fragment using a 5PRIME 2.5× PCR Master Mix. PCR products from G₀ animals were resolved on a 6–8% polyacrylamide gel. PCR products showing heteroduplex mobility shifts were cloned using the TOPO-TA cloning kit (Invitrogen, Carlsbad, CA). Ten representative colonies were picked from each plate and grown in 1.5-ml liquid cultures to isolate plasmid DNA. Recombinant plasmid DNA was sequenced using M13 F and R primers. To streamline genotyping of analyzed animals in subsequent studies, we enhanced the original HMA-based genotyping methods. The amplification was done with the same MmPkhd1 forward and reverse primers (see above) but using TaqMan master mix (New England BioLabs, Ipswich, MA) and a Bio-Rad Dyad Peltier Thermal Cycler PTC 220 (Bio-Rad, Hercules, CA). The amplified fragments were resolved the high-resolution capillary electrophoresis Qiagen QIAxcel Advanced System (Qiagen, Germantown, MD). Heterozygosity for the Pkhd1^C642* mutation in studied animals was confirmed by genotyping from different tissue samples and by subsequent sequencing.

All protocols followed the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health and were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee. The University of Alabama at Birmingham is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International.

Analysis of Pkhd1^C642* mutation transcript.

Total RNA was isolated from Pkhd1^C642*/C642* homozygote and wild type control kidneys and liver using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA), according to the manufacturer’s recommendations and converted to cDNA with the Superscript II reverse transcriptase kit (Thermo Fisher Scientific). Analysis of the Pkhd1^C642* mutation-containing exon 20 expression was done with primers positioned in flanking exons 19 (5′-gccttggtcaacttggacat-3′) and 21 (5′-gccttggtcaacttggacat-3′). PCR amplification and sequence analyses were performed using standard methods.

Small animal MRI.

The mouse kidneys were imaged using a 9.4T small animal MRI scanner. Each mouse, anesthetized with isoflurane, was located on a body temperature-regulating mouse bed in the prone position, and a surface coil (Bruker BioSpin, Billerica, MA) was placed on top of the animal’s body. A T2-weighted fast spin echo sequence (rapid acquisition with relaxation enhancement, RARE) was used to image the entire kidney regions with the following parameters: repetition time/echo time (TR/TE) = 5,000/50 ms, field of view (FOV) = 32.5 × 15 mm, number of excitations (NEX)  = 8, frequency/phase encoding = 278/128, flip angle = 180°, and slice thickness = 0.5 mm. Initially, the homozygotes for Pkhd1^C642* mutation and the wild-type controls were evaluated at age 8 and 14 days, homozygotes and heterozygotes for Pkhd1^C642* mutation and wild-type controls at 10 mo and 13–17 mo.

Kidney function measurements.

Serum creatinine was quantified at the University of Alabama at Birmingham-University of California at San Diego O’Brien Center for Acute Kidney Injury Research using liquid chromatography-tandem mass spectrometry (LC-MS/MS) as previously described (14).

Histology.

Formalin-fixed kidneys from the studied mice were paraffin embedded, cut into sections (3 and 5 μm), xylene deparaffinized, rehydrated, and stained using routine protocols (12, 17). Stained tissue sections were analyzed with bright-field microscopy using a Nikon E600 microscope equipped with a SPOT Insight digital camera (Diagnostic Instruments, Sterling Heights, MI).

For immunostaining, kidneys were fixed as above but then were placed in 30% sucrose overnight at 4°C and subsequently embedded into OCT compound (Thermo-Fisher Scientific; Waltham, MA) and frozen in a dry ice/ethanol bath. After cryosectioning, 10-μm sections were fixed with 4% PFA in PBS for 10 min and permeabilized in 0.2% Triton X-100 for 8 min, followed by three washes with PBS for 5 min. Blocking was done with PBS containing 1% bovine serum albumin (Sigma) for 30 min. Primary antibody diluted in blocking buffer was allowed to react with the tissues for 12 h at 4°C, secondary antibody (diluted in blocking solution) for 30 min at room temperature, two 5-min rinses with PBS, incubation with Hoechst (1:1,000; Sigma) for 5 min, and PBS wash for 5 min. Coverslips were mounted on slides using Immuno Mount (Thermo-Fisher Scientific) for mounting of tissue sections. Primary antibodies used included aquaporin 1 (AlphaDiagnostics; San Antonio, TX; AQP11-A 1:50 dilution), aquaporin 2 (Santa Cruz Biotechnology, Dallas, TX; sc-9882 1:50 dilution), cytokeratin 19 (CK 19; Abcam; Cambridge MA; ab52625; 1:200 dilution), Ki67 (FITC-labeled anti-KI67; eBioscience; San Diego, CA; 11-5698-82; 1:200 dilution), ADP-ribosylation factor-like protein-13B (Arl13b; Rosemont, IL; no. 17711; 1:200 dilution), and α-acetylated tubulin (Ac Tubulin; Sigma; T74551; 1:200 dilution). Secondary antibodies included AlexaFluor 488 donkey anti-rabbit, 1:1,000 (Merck Millipore, Billerica, MA), AlexaFluor 568 donkey anti-goat, 1:1,000 (Millipore), and AlexaFluor 488 goat anti-rabbit, 1:400 dilution (Thermo-Fisher Scientific). Images were captured on a Nikon CSU-X1 Spinning Disk Confocal on a Ti2-E Inverted Microscope Stand using Nikon Elements Software (Nikon instruments, Melville, NY).

RESULTS

Heterozygosity for the Pkhd1^C642* mutation triggers cystic liver disease in aged mice.

We used CRISPR/Cas9 technology to generate mice with a 7-base pair deletion in Pkhd1 exon 20 (Pkhd1 c.1926_1932delTGATTGG). This defect corresponds to a predicted truncating mutation Pkhd1^C642* (Fig. 1). Reverse-transcriptase-polymerase chain reaction (RT-PCR) amplification with primers positioned in exons 19 and 21 revealed an absence of the expected 235-bp RT-PCR product in Pkhd1^C642* homozygotes (it was present in Pkhd1^WT control; Fig. 1, inset). These data point to Pkhd1^C642* mutation-induced non-sense-mediated decay or changes in alternative splicing of the affected Pkhd1 transcript.

Fig. 1.

Design and generation of the Pkhd1^C642* model. The CRISPR/Cas9-based approach was used for of the Pkhd1 exon 20 targeting as outlined in detail in materials and methods. One of the generated mice had the Pkhd1 c.1926_1932delTGATTGG mutation in exon 20 that corresponds to truncating mutation Pkhd1^C642*. The wild-type Pkhd1 sequence (Pkhd1^WT) is used for comparison. The Pkhd1^C642 mutation effects on transcription are demonstrated by an absence of reverse transcriptase PCR product (specificity confirmed by sequencing) obtained by amplification with primers positioned in exons 19 and 21 in Pkhd1^WT control (inset).

The Pkhd1^C642* homozygotes and heterozygotes, identified with the enhanced genotyping assay (see materials and methods and Fig. 2) were viable and without notable anatomic abnormalities on gross pathological evaluation at 3 wk of age. Similarly, no major defects were noted on MRI images of Pkhd1^C642* homozygotes and wild-type control littermates at 8 and 14 days after birth (Fig. 3A).

Fig. 2.

Pkhd1^C642* genotyping. This PCR-based genotyping method was performed in two steps. In the first step, the amplified fragments were resolved using high-resolution capillary electrophoresis Qiagen QIAxcel Advanced System to distinguish between Pkhd1^C642/C642* homozygotes (and Pkhd1^+/+ homozygotes) and Pkhd1^C642/+ heterozygotes (left). The single band was obtained from a Pkhd1^C642/C642* homozygote because all analyzed PCR products in this sample have the same sequence. Similarly, a single band was obtained and Pkhd1^+/+ homozygote because all analyzed PCR products in this sample have the same sequence. In case of a Pkhd1^C642*/+ heterozygote (or a mixture of Pkhd1^C642/C642* and Pkhd1^+/+ DNA), a mixture of two different fragments led to the formation of heteroduplexes visualized as multiple bands. In the second step, the Pkhd1^C642/C642* homozygotes were distinguished from Pkhd1^+/+ homozygotes. This was done by addition of amplified fragments from the homozygotes to amplified fragments from wild-type Pkhd1^+/+ homozygotes in 1:1 ratio and their analysis using the Qiagen QiaAxel Advanced System (right).

Fig. 3.

Lack of radiographic liver abnormalities in young Pkhd1^C642/C642** homozygotes and in 10-mo-old Pkhd1^C642*/+ heterozygotes. A: representative images demonstrate lack of liver and kidney radiographic abnormalities in MRIs of Pkhd1^C642/C642** versus Pkhd1^+/+ mice at 8 and 14 days of postnatal age. B: representative images demonstrate lack of radiographic abnormalities on MRI images of kidneys and liver of 10-mo-old Pkhd1^C642*/+ mice versus Pkhd1^+/+ mice (mo, months).

The Pkhd1^C642* heterozygotes also appeared normal without well-defined liver abnormalities on T2 MRI images when aged up to 10 mo (Fig. 3B). However, when these mice were aged to 13–17 mo, equivalent to the human age of ~50 yr (8), they developed radiographic signatures of cystic liver abnormalities that ranged from multiple solitary cysts (or cyst clusters) to Caroli disease or CHF-like pattern (Fig. 4). Subsequent histopathological analyses revealed mostly isolated liver cysts with an epithelial lining that closely resembled biliary epithelia (Fig. 5A). Also, occasional dilated biliary ducts were present, and the number of bile ducts was increased (Fig. 5, A and B). However, there were minimal inflammatory infiltrates, or fibrosis of portal fields, and surrounding liver parenchyma appeared globally intact.

Fig. 4.

Aged Pkhd1^C642*/+ heterozygotes develop radiographic abnormalities in liver and kidneys. Representative MRI images obtained at levels of right kidney upper pole provide a view of the liver (top) and images at the level of renal arteries (bottom) show a transverse view of both kidneys. These images demonstrate that aged female and male Pkhd1^C642*/+ heterozygotes develop radiographic liver abnormalities that range from several large sporadic cysts to multiple cysts of various sizes. In contrast, the radiographic liver abnormalities were uniformly more severe in Pkhd1^C642*/C642* homozygotes. In contrast to the liver manifestations of Pkhd1^C642*/+ heterozygosity that were observed in both females and males, the radiographic renal abnormalities were observed only in females and there was virtually no difference in these manifestations between Pkhd1^C642*/+ heterozygotes and Pkhd1^C642*/C642* homozygotes. These radiographic changes resemble those found in kidneys of human heterozygotes for PKHD1 mutations (e.g., parent of ARPKD patient 1 in Ref. 10; note that human kidney contains multiple pyramids, whereas mouse kidney is equivalent to one pyramid).

Fig. 5.

Liver cysts and biliary duct abnormalities in aged Pkhd1^C642*/+ heterozygotes. A: representative images show that aged female and male Pkhd1^C642*/+ heterozygotes develop isolated liver cysts as well as abnormalities in a subset of biliary ducts. These changes include ectasia and increased number of biliary ducts in portal fields (portal vein can be identified by intraluminal erythrocytes; the cyst-resembling structures in the vicinity of portal veins are dilated biliary ducts). Virtually absent are typical features of congenital hepatic fibrosis such as fibrotic changes and inflammatory infiltrates of portal fields. With exception of the cystic and biliary duct changes, the remaining liver parenchyma is intact. Top panels were obtained with ×4 magnification of hematoxylin and eosin-stained liver tissues. B: cytokeratin 19 (CK 19) staining is consistent with ectasia and increased number of biliary ducts in portal fields of aged female and male Pkhd1^C642*/+ heterozygotes (scale bar, 50 μm).

Homozygosity for the Pkhd1^C642* mutation leads to congenital hepatic fibrosis in aged mice.

Compared with the Pkhd1^C642* heterozygotes, Pkhd1^C642* homozygotes aged to 13–17 mo developed severe radiographic signs of CHF (Fig. 4). These findings were supported by histopathological analyses that revealed an abnormally high number of dilated biliary ducts in portal fields surrounded by prominent inflammatory infiltrates and fibrosis; the liver parenchyma surrounding portal fields was grossly intact (Fig. 6). These histopathological findings resembled manifestations of CHF (or Caroli syndrome), one of the hallmark presentations seen in homozygotes for two mutated PKHD1 alleles (e.g., patients with ARPKD). The phenotypic manifestations observed in livers of aged Pkhd1^C642* homozygotes are similar to those described in existing Pkhd1 gene trap or exon deletion models (e.g., Pkhd1^lacZ, Pkhd1^del2, Pkhd1^del3–4 and Pkhd1^delEx40 (9, 11, 15, 16).

Fig. 6.

Congenital hepatic fibrosis in aged Pkhd1^C642/C642** homozygotes. Representative images show that both female and male Pkhd1^C642*/C642* homozygotes develop key features of congenital hepatic fibrosis (CHF), including an increased number of biliary ducts in the vicinity of portal vein (portal vein contains intraluminal erythrocytes; other cyst-resembling structures are dilated biliary ducts). Similar to human CHF presentation, portal fields in this model are fibrotic with prominent inflammatory infiltrates. Top panels were obtained with ×4 magnification.

Heterozygosity for the Pkhd1^C642* mutation leads to renal abnormalities in aged female mice.

Similar to the lack of liver defects in young Pkhd1^C642* mice, no renal abnormalities were noted on MRI images of the Pkhd1^C642* homozygotes and wild-type control littermates 8 and 14 days after birth (Fig. 3A). Also, there were no radiographic renal abnormalities noted on MRI images of Pkhd1^C642* heterozygotes at 10 mo (Fig. 3B). However, when aged to 13–17 mo, the Pkhd1^C642* heterozygotes (as well as homozygotes) developed changes that resembled radiographic signs seen in human patients with medullary sponge kidney (Fig. 4). Similar changes were also described in some heterozygotes for PKHD1 mutations (10). The most likely explanation of the radiographic changes in this model is ectasia of proximal tubules that we noted during the subsequent histopathological evaluation (Fig. 7, A and B). Similar ectasia of proximal tubules was previously reported in homozygotes for the Pkhd1^del2 mutation (16). However, we now describe similar changes in kidneys of aged Pkhd1^C642* heterozygotes. The overall size of these kidneys appeared normal (total kidney weight 0.38 ± 0.16 g, based on n = 8, total kidney to body weight 0.014 ± 0.002, total kidney volume by MRI 0.42 ± 0.07(SD) ml; serum creatinine 0.15 ± 0.03 mg/dl); although cortical thinning was present on both radiographic and histopathological studies. Also, proliferation reflected by Ki67 staining of these kidneys was low (approximately two Ki67⁺ tubular cells per five ×40 fields; Fig. 7C), and there was no apparent primary cilia structural abnormality observed in renal tubules (Fig. 7D). Notably, the proximal tubule ectasia was seen only in aged Pkhd1^C642* heterozygous females. Aged male Pkhd1^C642* heterozygotes had normally appearing kidneys based on imaging and histological evaluation (Figs. 4 and 7, A–D).

Fig. 7.

Ectasia of proximal tubules in aged female Pkhd1^C642*/+ heterozygotes. A: representative images show that only aged female Pkhd1^C642*/+ heterozygotes develop ectasia of renal tubules (e.g., designated by arrows). Left inset: example of such ectasia; right inset: less affected adjacent kidney tissue. Renal tubules of male Pkhd1^C642*/+ heterozygotes appear normal. Left and right insets: enlarged view demonstrating minimal differences between most and least dilated segments. Top panels were obtained with ×4 magnification of hematoxylin and eosin-stained kidney tissues; (inset scale bar, 1 mm). B: analysis of proximal tubule origin of observed ectasia, using aquaporin 1 (Aqp1; green) as a marker of proximal tubules and aquaporin 2 (Aqp2; red) as a marker of collecting ducts (scale bar, 50 μm). C: proliferation rate reflected by Ki67 staining is low (approximately two Ki67⁺ tubular cells per five ×40 fields; scale bar, 50 μm). D: also, Pkhd1^C642*/+ heterozygosity did not induce any notable structural abnormalities of primary cilia in renal tubules, as demonstrated by Arl13b and α-acetylated tubulin (Ac Tubulin) staining; (scale bar, 50 μm).

DISCUSSION

In the present study, we observed liver and kidney abnormalities in aged Pkhd1^C642* heterozygous mice that were not previously reported in heterozygotes for Pkhd1 mutations involving exons 1–3 [Pkhd1^lacZ/+ (15)], exon 2 [Pkhd1^del2/+ (16)], or exon 3–4 [Pkhd1^del3–4/+ (9)]. It is possible that unlike these large structural rearrangements involving one to three exons at the beginning of the Pkhd1 gene, a smaller defect in more central or an end-terminal portion of the gene (such as the one used in the current study) may better recapitulate effects and consequences of heterozygosity for human PKHD1 mutations. Data supporting such speculation are not yet available for human PKHD1 heterozygotes; however, no large deletions in the beginning of this gene were described in PKHD1 homozygotes (1–3).

An additional notable difference between studies of our newly described Pkhd1^C642* and the established Pkhd1 models (except Pkhd1^del2/+) is at least twice as long a follow-up interval in the current study (vs. follow-up interval reported for the other Pkhd1 models). Such longer follow-up corresponds better to the age of human heterozygotes for PKHD1 mutations with previously described liver and kidney abnormalities [age 30 s–60 s (4, 10)]. However, even with a longer follow-up, the Pkhd1^lacZ/+ and Pkhd1^del3–4 may not develop the Pkhd1^C642*-associated liver and kidney abnormalities because such changes were absent in 18-mo-old Pkhd1^del2/+ mice (16). Together, these data raise of possibility that in the mouse there may be mechanisms that compensate for truncating mutations at the 5′-end of Pkhd1 but not for truncating mutations at more 3′-positions, such as Pkhd1^C642*. Among the possible mechanisms that may contribute to this proposed compensation is the transcriptionally complex of mouse Pkhd1 gene (5).

The cystic liver phenotype seen in mouse Pkhd1^C642* heterozygotes is consistent with radiographic manifestations of heterozygosity for PKHD1 mutations in humans. Since these manifestations of PKHD1 mutation heterozygosity are relatively mild and do not warrant further evaluation of affected tissues, the underlying histopathological abnormalities remain unknown. The current study of mouse Pkhd1^C642* heterozygotes (Fig. 5) thus provides a valuable insight and suggests that human carriers of PKHD1 mutations develop relatively large cysts within grossly intact liver parenchyma with mostly normal portal fields. Dilated bile ducts or their increased numbers may be present in some portal fields; however, without inflammatory infiltrates or fibrotic changes.

Similarly, to the liver manifestations, the kidney abnormalities seen in female Pkhd1^C642* heterozygotes are consistent with radiographic renal manifestations of heterozygosity for human PKHD1 mutations (4, 10). Again, these renal defects are relatively mild and no kidney tissues are available (risks of obtaining these kidney tissues through kidney biopsies out-weight potential benefits). The current study (Fig. 7) suggests that heterozygotes for human PKHD1 mutations may develop ectasia of proximal tubules without notable tubular atrophy or interstitial fibrosis. While the corresponding radiographic findings are among typical manifestations of MSK (7), it is important to note that the histopathological hallmark of MSK is ectasia of the medullary collecting ducts, not the proximal tubules. The localization of this lesion makes sense in MSK given the associated nephrocalcinosis and nephrolithiasis. The lack of collecting duct involvement in the Pkhd1^C642* mice suggests that the MSK-associated nephrolithiasis/nephrocalcinosis-promoting collecting duct lesion is also absent in heterozygotes for human PKHD1 mutations. This conclusion is also supported by a lack of symptomatic or proven nephrocalcinosis or nephrolithiasis in heterozygotes for PKHD1 mutations.

Also, we observed a notable discrepancy between prominent radiographic abnormalities seen in kidneys of female Pkhd1^C642* heterozygotes or homozygotes (Fig. 4) versus mild dilatation of renal tubules on tissue sections (Fig. 7A). Possible explanations include effects of tissue processing or loss of fluid from dilated tubules after kidney extraction. Such discrepancy is expected to be less prominent when tubules are obstructed [e.g., in a severely affected ARPKD phenocopy (18)] or when cysts are isolated (e.g., in autosomal dominant PKD).

The major limitation of this study is a relatively small number of analyzed animals (n = 8 for Pkhd1^C642* heterozygotes). However, this sample size seemed at least initially reasonable for evaluation of the long-term effect of Pkhd1^C642* heterozygosity. Owing to the late onset of the observed phenotypic manifestations in this model, repeating this study to more systematically evaluate the liver and kidney phenotypes would require a delay of at least 1.5 yr.

In summary, we developed a new Pkhd1-deficient mouse model with the predicted truncating Pkhd1^C642* mutation. We have shown that heterozygotes for the Pkhd1^C642* mutation present with liver and kidney abnormalities that resemble those previously described in human heterozygotes for PKHD1 mutations. However, we observed these well-defined phenotypes only when the Pkhd1^C642* heterozygotes reached the advanced age (~1.5 yr) that is equivalent to the age when these manifestations were described in human heterozygotes for PKHD1 mutations. Furthermore, the characterization of the Pkhd1^C642* model: 1) extends our knowledge about the impact of heterozygous mutations in an autosomal recessive disorder; 2) provides a model that potentially correlates with the study findings in human parents of ARPKD patients (10), raising the question about the histopathological location of the cystic dilatation; and 3) suggests that there is a mouse-specific “reno-protective mechanism” in Pkhd1 models that wanes with age, and that proximal tubules (where the ARPKD lesion first arises in human embryos) are more susceptible to this waning effect than the distal convoluted tubules and collecting ducts.

GRANTS

Support was provided, in part, by the National Institutes of Health (NIH)-funded University of Alabama at Birmingham (UAB) Hepato/Renal Fibrocystic Disease Core Center [National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant P30 DK-074038]; the core resource of the UAB-UCSD O’Brien Center for Acute Kidney Injury Research (NIDDK Grant P30 DK-079337); NIDDK Grant R01 DK-097423; Grant 1-I01-BX002298 from the Office of Research and Development, Medical Research Service, Department of Veterans Affairs; and by the Detraz Endowed Research Fund in Polycystic Kidney Disease (to M. Mrug).

DISCLOSURES

M. Mrug reports grants and consulting fees outside the submitted work from Otsuka Pharmaceuticals and Sanofi. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

D.S., G.R., P.C., A.K.C., R.A.K., B.K.Y., L.M.G.-W., and M.M. conceived and designed research; D.S., G.R., S.M., R.R., J.Z., P.C., D.Z.R., A.K.C., T.R.S., M.J.C., and M.M. performed experiments; D.S., G.R., J.Z., P.C., D.Z.R., A.K.C., H.K., M.E.L., T.R.S., M.J.C., R.A.K., B.K.Y., L.M.G.-W., and M.M. analyzed data; D.S., G.R., J.Z., P.C., D.Z.R., A.K.C., H.K., M.E.L., T.R.S., M.J.C., R.A.K., B.K.Y., L.M.G.-W., and M.M. interpreted results of experiments; D.S., S.M., R.R., D.Z.R., A.K.C., M.E.L., T.R.S., M.J.C., L.M.G.-W., and M.M. prepared figures; D.S., D.Z.R., A.K.C., M.E.L., T.R.S., L.M.G.-W., and M.M. drafted manuscript; D.S., G.R., D.Z.R., A.K.C., H.K., M.E.L., T.R.S., M.J.C., R.A.K., B.K.Y., L.M.G.-W., and M.M. edited and revised manuscript; D.S., G.R., S.M., R.R., J.Z., P.C., D.Z.R., A.K.C., H.K., M.E.L., T.R.S., M.J.C., R.A.K., B.K.Y., L.M.G.-W., and M.M. approved final version of manuscript.