Mutations of the PKD2 Gene in Taiwanese Patients with Autosomal Dominant Polycystic Kidney Disease

Abstract

Background: Mutation analysis in the context of clinical phenotypes helps clarify the pathogenesis of autosomal dominant polycystic kidney disease (ADPKD). Over 78 PKD2 gene mutations have been reported in the literature, but few have been described from an Asian population. This study attempted to characterize PKD2 mutations and their clinical implications among Taiwanese. Methods: Twenty unrelated ADPKD patients with uncharacterized genotypes were screened for mutations in the PKD2 gene via single-strand conformation polymorphism (SSCP) of PCR products from genomic DNA, using previously reported PCR conditions and primers. Results: This study identified two novel mutations (C681A and 2136-2137delG) and one mutation (C2407T) previously reported in a Cypriot family. Overall, we found PKD2 mutations in 15% (three out of 20) of the ADPKD patients screened. The mutations included two nonsense mutations (Y227X and R803X) and one frameshift mutation (712-715X) that could all lead to premature termination of translation. The locations of mutations in this study spanned the entire PKD2 gene on exons 2, 11, and 13 without clustering and did not influence the renal disease severity. Conclusions: The study identified two novel mutations and one recurrent mutation of the PKD2 gene in 20 Taiwanese patients. The characteristics of the mutations in this study resemble those reported among Western populations.

• autosomal dominant polycystic kidney type 2
• mutational analysis
• single-strand conformation polymorphism

Introduction

Autosomal dominant polycystic kidney disease [ADPKD (MIM 173900)] is a very common hereditary renal disease that affects 1/1,000 live births and nearly 10% of all patients on renal replacement therapy.[1] The disease results in end-stage renal disease (ESRD) by 60 years of age in approximately 45% of affected patients due to progressive renal cysts formation and expansion. Systemic manifestations of ADPKD involve cyst formation in nonrenal organs, intracranial arterial aneurysms, cardiac valvular defects, and colon diverticulosis. About 85% of cases are caused by mutations in PKD1 [MIM 601313] located on chromosome 16p13.3, while mutations in PKD2 [MIM173910] on chromosome 4q 21–23 account for most of the remaining cases.[2] A third locus of ADPKD has been suspected among families without links to either PKD1 or PKD2.[3]

The PKD1 gene comprises 46 exons spanning over 52 kb of genomic DNA. The 14,136 bp transcript encodes polycystin-1, a 4302 amino acids integral membrane protein with a large extracellular region and multiple transmembrane domains.[4] Polycystin-1 is thought to be a membrane glycoprotein involved in mediating cell–cell and cell–matrix adhesion.[5] The PKD2 gene consists of 15 exons with an open reading frame of 2,904 bp and a 3′ untranslated region (UTR) of 2,085 bp.[6] The PKD2 gene product, polycystin-2, is predicted to be a 968 amino acids integral membrane protein with six membrane-spanning domains and with homology to a family of voltage-activated and transient receptor potential cation channels.[2] Polycystin-1 and -2 channel complex can function as members of a common signaling pathway regulating intracellular calcium,[6] and have been demonstrated tointeract in vivo as a heterodimeric complex through their C-terminal cytoplasmic tails.[7] Recent studies have suggested that the polycystin-1- and -2-channel complex is linked to novel microtubulin-associated proteins in the primary cilium and contributes to fluid-flow sensation in renal epithelium.[8]

The correlation of the genotype and phenotype of ADPKD remains uncertain. PKD2 has been reported to be a milder form of ADPKD, for which the mean age of ESRD occurs nearly 20 years later than PKD1.[9] While locus heterogeneity is a major determinant of interfamilial disease variability, considerable interfamilial and intrafamilial renal disease variability in ADPKD has been recognized. More recent studies have suggested that allelic heterogeneity might influence renal disease severity.[10&11] Mutations in the 3′ half of the PKD2 have been shown to have milder renal complications than in patients with mutations in the 5′ half of the gene.[10] Nevertheless, significant renal disease variability exists among patients with identical PKD2 mutations.[11]

Mutation analysis of PKD1 and PKD2 genes in the context of the clinical phenotypes will identify functionally important regions and help in understanding the pathogenesis of ADPKD. However, mutation screening in the PKD1 gene is difficult because of its large size and complexity, with most of the transcript being encoded by a genomic region that is reiterated more proximally on chromosome 16.[12] In contrast, mutation screening in the PKD2 gene is straightforward, and the sensitivity of mutation detection in PKD2 is much greater than in PKD1.[13] To date, over 78 different mutations of the PKD2 gene have been reported in different populations,[11-16] mostly in Caucasians. However, mutations of the PKD2 gene have rarely been described from Asian populations in the literature.

This study directly screened for PKD2 gene mutations in a cohort of 20 unrelated ADPKD patients in Taiwan, using single-strand conformation polymorphism (SSCP) analysis and DNA sequencing to determine the characteristics of PKD2 mutations in the Taiwanese population.

Materials and Methods

Patients

PKD2 mutations in 20 unrelated ADPKD patients with unknown PKD genotype status were directly analyzed. The patients were recruited from the Lin-Kou Medical Center of Chang-Gung Memorial Hospital in Taiwan. Demographic and clinical information and genomic DNA were obtained from all patients after informed consent. In the sample patients, linkage to either PKD1 or PKD2 locus could not be established due to either a small family or a lack of blood samples from other members. Ethical approval was given by the ethical committee of the hospital. Control samples were obtained from 20 unrelated healthy individuals.

Clinical Information

ADPKD was diagnosed by a nephrologist using well-established ultrasound-based criteria, as follows: the presence of at least two renal cysts (unilateral or bilateral) in an at-risk individual aged < 30 years; the presence of at least two cysts in each kidney in an at-risk individual aged 30–59 years; or the presence of at least four cysts in each kidney in an at-risk individual aged > 60 years.[17] A detailed pedigree structure was constructed for each patient after recording the family history. Table 1 summarizes the clinical characteristics of these patients. Hypertension was defined as blood pressure exceeding 140/90 mmHg. ESRD was indicated by a creatinine clearance rate of less than 10 mL/min or the need for renal replacement therapy. A history of urolithiasis or ruptured cerebral aneurysm was recorded.

Table 1. Characteristics of 20 Patients with ADPKD.

Sex (men/women)	11/9
Mean age (years)	55.2 ± 15.7
Patients with ESRD	11/20 (55%)
Mean age of ESRD (years)	52.6 ± 11.6
Hypertension	15/20 (75%)
Urolithiasis	3/20 (15%)
Ruptured intracranial aneurysm	1/20 (5%)
Mean creatinine in patients without ESRD (mg/dL)	1.6 ± 0.8

Note: Values expressed as mean ± SD.

DNA Analysis

Genomic DNA was extracted from the peripheral blood lymphocytes by a standard method, using phenol/chloroform extraction and isopropanol precipitation. The exons of the PKD2 gene, including associated splice donor and acceptor sites, were amplified using primers and polymerase chain reaction (PCR) conditions as reported previously.[18] The PCR products, with sizes ranging from 188–382 bp, were then analyzed by SSCP.

For SSCP, 6 µL of denatured PCR product was mixed with loading buffer and loaded into GeneGel Excel 12.5/24 acrylamide gels (Pharmacia). Electrophoresis was typically performed at 15°C with 600 V for 85 min. The different migrations on electrophoresis were silverstained. All detected SSCP variants were confirmed by repeated PCR and SSCP.

The PCR products exhibiting shifted bands on SSCP were then cloned in a pGEM-T Easy Vector System (Promega). The obtained plasmid DNA was purified with a Minipreps DNA Purification System (Viogene) and then sequenced directly in both directions by the dideoxy terminator method using an ABI 3100 DNA Sequencer (Applied Biosystems).

Results

Direct mutation screening was performed by SSCP of PCR products obtained from genomic DNA of 20 affected subjects with unknown PKD genotypic status. This study found three different mutations, including two nonsense mutations (C681A and C2407T) and a frameshift mutation (2136-2137delG). PKD2 mutations were identified in 15% (three of 20) of the patients screened. Table 2 summarizes the location, type, and predicted consequence of each mutation on the encoded protein.

Table 2. Mutations of the PKD2 Gene in the Current Study.

Patient	Mutation	Location	Nucleotide change	Clinical features
1	Nonsense: Y227X	Exon 2	C681A	ESRD at age of 50
2	Frameshift: 712 → 715X	Exon 11	2136-2137delG	Cr 1.5 at age of 70; cardiac valve abnormality; cardiac conduction defect
3	Nonsense: R803X	Exon 13	C2407T	ESRD at age of 52; urolithiasis

Nonsense Mutation C681A in Patient 1

SSCP analysis of the exon 2 in patient 1 found an aberrant fragment. DNA sequencing of the PCR product cloned in pGEM-T Easy Vector revealed a substitution of C by A at nucleotide 681 (Fig. 1A). This nonsense mutation C681A produced an immediate stop codon (TAC → TAA, Y227X) and was predicted to produce a truncated protein, 642 amino acids shorter than the normal polycystin-2 (968 aa).

Figure 1. Sequence data showing three PKD2 gene mutations: (A) nonsense mutation Y227X, C681A; (B) frameshift mutation 712-715X, 2136-2137delG; and (C) nonsense mutation R803X, C2407T.

Patient 1 was a 53-year-old man who had been undergoing hemodialysis since the age of 50. The man was diagnosed with ADPKD due to hypertension at 40 years of age with multiple renal and liver cysts. The sister of the man was also affected with ADPKD and suffered ESRD requiring hemodialysis at 50 years of age. The parents of the man both died before the age of 50 from unrelated disorders.

Fameshift Mutation 2136-2137delG in Patient 2

SSCP analysis in exon 11 of patient 2 revealed a double-strand pattern. DNA sequencing revealed a G depletion at nucleotide 2136-2137 (Fig. 1B). This mutation caused frameshifting after codon 712 and introduced premature termination at codon 715. The predicted peptide was 254 amino acids shorter than the normal protein.

Patient 2 was a 74-year-old woman with bilateral polycystic kidneys and mild renal impairment (serumcreatinine 1.5 mg/dL). The woman was diagnosed with ADPKD because of palpable renal mass when she was 70 years old. The woman also had hypertension, liver cysts, and mitral valve regurgitation. A permanent pacemaker had been instituted for sick sinus syndrome when she was 70 years old. Her son was found to be affected with ADPKD after examination at the age of 49. The son had normal renal function and mild hypertension.

Nonsense Mutation C2407T in Patient 3

Patient 3 displayed shifted bands on SSCP analysis for exon 13. Sequencing of the cloned PCR products revealed substitution of C by T at nucleotide 2407 (Fig. 1C). The predicted protein change was a stop codon in the replacement of arginine codon 803 (CGA → TGA, R803X) and was predicted to produce a truncated peptide 166 amino acids shorter than normal polycystin-2.

Patient 3 was a 52-year-old woman who was first diagnosed with ADPKD at the age of 42 with initial presentation of hypertension and flank pain. The patient had experienced an episode of right middle cerebral artery infarction at 51 years of age. The woman also had a history of recurrent ureter stones, staghorn stones, and obstructive uropathy, which complicated her renal impairment. Renal function deteriorated progressively at 52 years of age, with a serum creatinine level of 7.2 mg/dL and a creatinine clearance rate of less than 10 mL/min. One of the daughters of the patient was found to be affected at the age of 29 and had normal blood pressure and renal function. The parents both died at around 54 years of age, but no information was available regarding causes of death.

Discussion

This study screened 20 Taiwanese individuals with ADPKD from different families for PKD2 mutation. Three germ-line mutations were identified by SSCP analysis and DNA sequencing. Among these mutations, R803X was reported previously from a Cypriot family.[16] Two novel mutations (Y227X, 712-715X) are reported here for the first time.

Similar to previous studies in PKD2 mutation, the location of mutations in this study spanned the entire PKD2 gene without significant clustering. Of the three mutations, one was located at the 5′ end of the PKD2 in exon 2 and two at the 3′ end in exons 11 and 13, respectively. The existence of mutational hot spots in the PKD2 gene remains uncertain. At least 50% of all of the germ-line mutations of PKD2 have been reported to occur in exons 2, 4, 5, 6, and 11, whereas, thus far, mutations have not been reported in exons 3 and 15.[16] A single nucleotide deletion or insertion of a polyadenosine tract (2152-2159delA and 2152-2159insA) on exon 11 has been suggested to indicate a regional hot spot within PKD2, presumably resulting from “slipped strand mispairing.”[11] The mutation 2136-2137delG in this study is also located near the polyadenosine tract, suggesting a possible clustering region. Additionally, the mutation R803X (C2407T) in this study has been reported in a Cypriot family. Although haplotype analysis was not performed, the two families are from different ethnic populations and are unlikely to share a common ancestry. A plausible explanation for this mutation being repeated is that the CpG dinuleotides within arginine residues are prone to mutagenesis through cytidine methylation and subsequent deamination to thymidine. Furthermore, the most common recurrent PKD2 mutation R417X (C1249T), which has been found in four families, is also caused by changing an arginine codon to a stop codon.[16] Alternatively, the mutations in PKD2 could simply represent the occurrence of new mutations all over the coding sequence, because most are unique to a single family.

Most mutations found in PKD2 are nonsense mutations and insertions or deletions that result in the premature stopping of translation and are predicted to cause complete loss of protein function. According to the current “two-hit” hypothesis, cytogenesis requires a second somatic mutation on the inherited healthy allele.[19] Whether the location of a mutation of PKD2 influences the disease severity remains controversial. A recent study discovered that mutation position in PKD1 is a strong predictor of vascular complications,[20] indicating that the products from the mutant allele may play a “dominant negative” or “gain-of-function” role. In contrast, no significant genotype to renal function correlation was found in a recent study of PKD2 mutations in 71 families, and PKD2 mutation location did not influence the age of onset of ESRD.[11] Similarly, we had not found a clear correlation between the PKD2 genotype and phenotype in this study. The 5′ end mutation on exon 2 was associated with relatively early onset of ESRD, at 50 years old in patient 1, while the 3′ end mutation on exon 11 was related to mild renal impairment at 72 years old in patient 2. However, patient 3, with mutation on exon 13, the most distal position in the current study, did not run a milder course, and ESRD developed at the age of 52. Therefore, mutations in the 3′ end of the PKD2, which are predicted to truncate the C′-terminus of polycystin-2, may be sufficient to completely inactivate the mutant protein by losing the coil–coil interaction region with polycystin-1. Consequently, thelack of a clear genotype to phenotype correlation from our data is consistent with the concept that most PKD2 mutations are inactivating, and other genetic and environmental factors may modify the disease severity.[11] A functional analysis of the truncated protein product of the mutant allele can help clarify this hypothesis.

The study also examined the usefulness of SSCP for direct screening of PKD2 mutations in ADPKD patients not subject to linkage analysis. The sensitivity of SSCP has been reported to be above 87.5% for searching PKD2 mutations in families linked to PKD2.[13] We identified three PKD2 mutations among 20 patients, achieving an overall detection rate of 15%. The findings of this study resemble those of previous reports of the prevalence of PKD2 genotype based on linkage analysis in different populations.[9] Clearly, the possibility remains that a mutation may have been missed by relying on SSCP as a single mutation screening method. Additionally, a selection bias may exist, because PKD2 patients could remain asymptomatic and thus unidentified. Therefore, the actual proportion of PKD2 to PKD1 is expected to be higher than that observed here. The recent application of denaturing high-performance liquid chromatography has improved the efficiency of screening the entire coding regions of the PKD1 and PKD2 genes and could prove an effective method of gene-based molecular diagnosis of ADPKD.[12] However, the clinical benefit of differentiating between PKD1 and PKD2 is doubtful, because the disease courses may vary widely in affected individuals.

In conclusion, the study identified two novel mutations and one recurrent mutation of the PKD2 gene in 20 Taiwanese patients. The characteristics of the mutations and their clinical relationships resemble those among Western populations.

Acknowledgments

This study was supported by a research grant from the Chang Gung Memorial Hospital (CMRP1373). The authors express their gratitude to the patients and their families for participating in this study.