Introduction

Phenylketonuria (PKU; MIM 261600) is an inborn error of metabolism inherited as an autosomal recessive trait. The incidences of PKU vary among ethnic populations: 1/10,000 in Caucasian (Bickel et al. 1981), 1/120,000 in Japanese (Aoki and Wada 1988), and 1/18,000 in Chinese (Liu and Zuo 1986). In Korea, PKU incidence is estimated, from the neonatal screening that was introduced in 1998, to be about 1/41,000.

PKU results from a deficiency of phenylalanine hydroxylase (PAH). The PAH gene spans about 90 kb on chromosome 12q and comprises 13 exons. PAH is a hepatic enzyme that catalyzes hydroxylation of phenylalanine to tyrosine using tetrahydrobiopterin (BH4) as a cofactor. It has three structural domains consisting of an N-terminal regulatory domain, a catalytic domain, and a C-terminal tetramerization domain. The active PAH enzyme is comprised of four monomeric proteins. Recent studies of PAH crystal structure have provided information on the active site and the binding sites of its substrate and cofactor.

The mutation profile of the PAH gene is not restricted to any one region but spreads throughout the entire structural domains and shows enormous diversity. More than 460 different mutations of the PAH gene have been identified and recorded in the PAH Mutation Analysis Consortium Database (PAHdb; Scriver et al. 2003). The severity of the disease is also diverse from mild hyperphenylalaninemia (MHP) to classical PKU, which is characterized by pretreatment blood phenylalanine levels or dietary tolerance (Guldberg et al. 1998). Another phenotypic characteristic is BH4 responsiveness. The serum phenylalanine levels of BH4-responsive patients are controlled by BH4 oral administration without phenylalanine restriction diet. Several studies have investigated the relationship between genotype and this diverse phenotypic expression. Determining the relationship between genotype and phenotype would provide very useful information for planning dietary and therapeutic strategies.

Therefore, we analyzed the PAH gene in 79 patients with PKU and their families to study genotype–phenotype relationships and to help with genetic counseling. Furthermore, we analyzed the mutation spectra of the PAH gene in Korean patients and compared them with those of other ethnic groups, including Japanese and Chinese.

Subjects and methods

Subjects

This study was approved by the institutional review board of the National Institute of Health, Korea. The study included 79 unrelated families with PAH deficiency. Participants were recruited from the Korean PKU family support group. Most of them were identified in neonatal screening, and PAH deficiency was diagnosed by conventional biochemical methods. Patient severity was assigned to classical PKU, moderate PKU, or MHP, according to the plasma phenylalanine concentration prior to phenylalanine restriction diet. The level for classical PKU was 1,200 μM or more; the level for moderate PKU 600–1,200 μM; the level for MHP less than 600 μM. Informed consent for DNA analysis was obtained from the patients and their families.

For the BH4 loading test, patients without a phenylalanine restriction diet were administered orally at a dose of 20 mg/kg (for the patients under 36 months old) or 7.5 mg/kg (for the patients over 36 months old). Blood phenylalanine levels were measured before, 1, 2, 4, 6, 8, 12, and 24 h after administration. The BH4 loading test was considered positive when initial plasma phenylalanine concentration decreased by at least 40% after 12 h. Urinary pterin analysis and dihydropteridine reductase (DHPR) assay were performed to exclude 6-pyruvoyl-tetrahydropterin synthase (PTPS) deficiencies.

Mutation analysis

Genomic DNA was isolated from peripheral blood leukocytes using the QIAamp DNA blood kit following the manufacturer’s instruction (Qiagen, Hilden, Germany). All 13 exons including exon–intron boundaries and 2 kb of the 5′-upstream region of the PAH gene were amplified by PCR. PCR amplicons were extracted from an agarose gel using a gel extraction kit (Qiagen). Direct sequencing was performed using a BigDye Terminator Cycle Sequencing Ready Reaction Kit, version 2.0 (PE Applied Biosystems, Foster City, CA, USA) and analyzed with an ABI 3100 automated sequencer (PE Applied Biosystems) according to the standard methods. When available, parental DNA samples were sequenced to confirm trans configurations in compound heterozygotes and to distinguish homozygosity from hemizygosity. In addition, PAH genes in 50 normal individuals were analyzed to confirm that the novel sequence variations were not polymorphisms but real pathogenic mutations. Novel mutations were defined by exclusion from the PAHdb (http://www.pahdb.mcgill.ca) and previously reported mutations on PubMed (http://www.ncbi.nlm.nih.gov/PubMed/).

Results and discussion

PAH nucleotide sequence analysis of 79 unrelated PKU probands revealed 39 different mutations (Tables 1, 2). Among 79 patients, two mutation alleles were detected in 59 patients (75%), either compound heterozygous or homozygous (52 and seven, respectively). Only one mutation allele was revealed in 19 patients, and no mutations were detected in two patients. Ten novel mutations were identified in this study. These novel mutations included nine missense substitutions: P69S, G103S, N207D, T278S, P281A, L293M, G332V, S391I, and A447P. From database comparison, the glycine103 is found to be conserved among human, mouse, and rat. The remaining mutated amino acid residues were even more strictly conserved among human, mouse, rat, and zebrafish. A novel splice-site variant, IVS10−3C>G, was also detected. The -3 sequence of the splicing acceptor site is a strictly conserved sequence, and this substitution might result in aberrant splicing products. No novel frameshift mutations or nonsense mutations were detected.

Table 1 Genotypes for mutations of the PAH gene in 79 Korean phenylketonuria (PKU) patients
Full size table
Table 2 Spectrum of PAH mutations detected in this study
Full size table

R243Q, IVS4−1G>A, and E6−96A>G were the most prevalent mutations in Korean patients with PKU. They have been reported to be some of the most frequent mutations in Asian populations (Table 3) and accounted for 51 of the 158 total chromosomes (32.2%) in this study.

Table 3 Relative frequencies of common PAH gene mutation found in oriental populations
Full size table

It is well known that different ethnic groups have their own distinctive and diverse PAH mutant allele series that include one or a few prevalent founder alleles (Zschocke 2003). In comparison of PAH mutation data among ethnic groups, there are the correlations between mutation and genetic history of investigated populations. For example, in Europe, there are several prevalent founder alleles, including R408W, IVS12+1G>A, IVS10−11G>A, and Y414C, that represent the expansion, migration, and genetic drift of European populations (Zschocke 2003). In particular, the R408W mutation has a frequency of 20–84% in PKU patients in Eastern Europe and Germany. However, these mutations are rarely detected in oriental populations. In a previous study, Okano et al. (1992) reported the frequency and distribution of PAH gene mutations among Japanese, Korean, and Chinese patients. Because the study was undertaken in the early 1990s, it was restricted to screening for previously isolated mutations. Unidentified but relatively frequent alleles, such as R241C, were not investigated, and only ten Korean patients were included, which is a relatively small number to represent Korean allelic distribution. The present study, with 79 participants, extends these previous results to give a more comprehensive understanding of PAH allele distribution and frequency in Koreans. Although some overlaps of mutant allele distribution are observed among Japanese, Chinese, and Korean populations, there are several significant differences (Table 3). R243Q, E6−96A>G, and IVS4−1G>A, the most frequent mutations in our study, are also frequently detected in Japanese, Chinese, and Taiwanese. However, R111X, a frequent mutation in Japanese and Chinese patients, is very rare in Korean patients. R413P is the most prevalent allele in Japanese, but a very small proportion of probands have the R413P allele in Korean and Taiwanese. IVS4−1G>A occupied a relatively larger proportion in Korean mutant allele profiles than in Japanese or Chinese. Although A259T was not detected in any other oriental population studies, it was identified in nine different families in this study.

Interestingly, the two R241C homozygous patients (patient 25 and patient 76) showed MHP, and all compound heterozygous individuals with R241C (2 with R241C/R243Q, another 2 with R241C/A259T and 1 with R241C/T278I) showed BH4 responsiveness (Table 1, Fig. 1). In a previous study, PAH with R241C substitution showed to have 25% of residual activity in the COS cell expression system (Okano et al. 1994). Guldberg et al. (1998) assigned the patient with a R241C genotype to the MHP category. It was also reported that the blood phenylalanine levels of R241C/R413P patients was decreased by oral administration of BH4 (Kure et al. 1999). R241 is located near the cofactor binding region and does not directly interact with the cofactor, so the mutation may lead to relatively mild structural deformities (Erlandsen and Stevens 2001). Our data are consistent with these previous reports.

Fig. 1
figure 1

Profile of blood phenylalanine concentration changes during the BH4 loading test. Filled square, patient 3 with Y356X/R408Q genotype; filled triangle, patient 26 with R53H/R243Q; and filled circle, patient 39 with R241C/R243Q showed the BH4 responsive pattern; and open triangle, patient 25 with R241C/R241C; and open circle, patient 49 with R243Q/? showed the nonresponsive pattern

Full size image

Patient 3 (genotyped with Y356X/R408Q) also represented BH4 responsiveness. Y356X is a null mutation and may not be the BH4-responsive allele. R408Q was reported to be associated with near-normal levels of residual activity in eukaryotic and prokaryotic expression system (Pey et al. 2003). The residual activity of R408Q and the BH4 responsiveness of patient 3 indicate that R408Q is one of the BH4-responsive alleles.

Our data added R53H to the list of BH4-responsive PAH alleles. Patient 26 (genotyped with R53H/R243Q) represented BH4 responsiveness. The facts that R243Q was associated with classical PKU in our study and another R53H heterozygous patient was MHP suggested that R53H had some residual enzyme activity and brought out the responsiveness in patient 26.

The BH4 response pattern between the PTPS-deficient patient and the PKU patient are somewhat different (Fig. 1). Phenylalanine levels of the PTPS patient was dramatically and completely decreased to the normal level after administration of BH4; in the PKU patient, the decrease was relatively retarded, and the blood phenylalanine concentration remained at the higher-than-normal level. The basal phenylalanine level of patient 25 (R241C homozygote) was too low to represent BH4 responsiveness.

In the BH4-non-responsive patients, the phenylalanine level remained at the same level as the starting point (Fig. 1). Some moderate PKU patients (patient 49 and 72) did not respond to the BH4. This result suggests that BH4 responsiveness requires some residual enzyme activity, but all the cases with mild phenotype are not associated with the BH4 responsiveness.

In summary, we screened the PAH gene in 79 Korean PKU-affected families and identified 39 mutations, including ten novel mutations. Although the Korean mutation profile of PAH is similar to those of the nearest oriental populations, there are several different characteristic features. The relationship of genotype and phenotype, especially BH4 responsiveness of some patients, was also described. This study would contribute to the diagnosis, genetic counseling, and planning of the dietary and therapeutic strategy in PKU patients.