Phenylalanine hydroxylase deficiency: Molecular epidemiology and predictable BH4-responsiveness in South Portugal PKU patients

Molecular Genetics and Metabolism 104 (2011) S86–S92

Contents lists available at ScienceDirect

Molecular Genetics and Metabolism j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y m g m e

Phenylalanine hydroxylase deficiency: Molecular epidemiology and predictable BH4-responsiveness in South Portugal PKU patients Isabel Rivera a, b,⁎, Dina Mendes a, Ângela Afonso a, Madalena Barroso a, Ruben Ramos a, Patrícia Janeiro c, Anabela Oliveira d, Ana Gaspar c, Isabel Tavares de Almeida a, b a

Metabolism and Genetics Group, iMed.UL-Research Institute for Medicines and Pharmaceutical Sciences, Faculty of Pharmacy, University of Lisbon, Portugal Department of Biochemistry and Human Biology, Faculty of Pharmacy, University of Lisbon, Portugal Department of Paediatrics, Santa Maria Hospital, Lisbon, Portugal d Department of Medicine, Santa Maria Hospital, Lisbon, Portugal b c

a r t i c l e

i n f o

Article history: Received 2 June 2011 Received in revised form 27 July 2011 Accepted 28 July 2011 Available online 31 July 2011 Keywords: Phenylketonuria Hyperphenylalaninemia Tetrahydrobiopterin Genotype Phenotype

a b s t r a c t Hyperphenylalaninemia (HPA, OMIM #261600), which includes phenylketonuria (PKU), is caused by mutations in the gene encoding phenylalanine hydroxylase (PAH), being already described more than 600 different mutations. Genotype–phenotype correlation is a useful tool to predict the metabolic phenotype, to establish the better tailored diet and, more recently, to assess the potential responsiveness to BH4 therapy, a current theme on PKU field. The aim of this study was the molecular analysis of the PAH gene, evaluation of genotype–phenotype relationships and prediction of BH4-responsiveness in the HPA population living in South Portugal. We performed the molecular characterization of 83 HPA patients using genomic DNA extracted from peripheral blood samples or Guthrie cards. PAH mutations were scanned by PCR amplification of exons and related intronic boundaries, followed by direct sequence analysis. Intragenic polymorphisms were determined by PCR-RFLP analysis. The results allowed the full characterization of 67 patients. The mutational spectrum encompasses 34 distinct mutations, being the most frequent IVS10nt-11GNA (14.6%), V388M (10.8%), R261Q (8.2%) and R270K (7.6%), which account for 46% of all mutant alleles. Moreover, 12 different haplotypes were identified and most mutations were associated with a single one. Notably, more than half of the 34 mutations belong to the group of more than 70 mutations already identified in BH4-responsive patients, according to BIOPKU database. Fifty one different genotypic combinations were found, most of them in single patients and involving a BH4responsive mutation. In conclusion, a significant number (30–35%) of South Portugal PKU patients may potentially benefit from BH4 therapy which, combined with a less strict diet, or eventually in special cases as monotherapy, may contribute to reduce nutritional deficiencies and minimize neurological and psychological dysfunctions. © 2011 Elsevier Inc. All rights reserved.

1. Introduction Hyperphenylalaninemia (HPA, OMIM #261600), which includes phenylketonuria (PKU), is the most common inborn error of amino acid metabolism with an average incidence among Caucasian and Oriental Asian populations of 1 per 10,00 newborns, while the mutant allele frequency in the population is polymorphic (~0.01). This pathology displays autosomal recessive inheritance and its cause is multifactorial: mutations in the gene encoding phenylalanine hydroxylase (PAH; EC 1.14.16.1) and exposition to dietary phenylalanine are both necessary and sufficient conditions to trigger it [1]. ⁎ Corresponding author at: Metabolism and Genetics Group, iMed.UL-Research Institute for Medicines and Pharmaceutical Sciences, Faculty of Pharmacy, University of Lisbon, Av. Prof. Gama Pinto, 1649-003 Lisbon, Portugal. Fax: + 351 21 7946491. E-mail address: [email protected] (I. Rivera). 1096-7192/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ymgme.2011.07.026

PAH is responsible for the conversion, in the presence of the cofactor tetrahydrobiopterin (BH4) and dioxygen, of phenylalanine into tyrosine, which becomes an essential amino acid when the hydroxylating activity is absent or impaired. Then a simultaneous elevation in phenylalanine levels can be observed altogether with an impairment in tyrosine ones and both events contribute to the features characterizing the metabolic and clinical phenotypes displayed by the patients. PKU and related HPA represent the paradigm of a genetic disease that can be treated. For the last 50 years, the traditional treatment has been a phenylalanine restricted diet for all lifelong with all the problems associated, namely nutritional deficits and socialization troubles, just to mention two different aspects [1–3]. However, some new approaches are being tried, as the use of large neutral amino acids and macroglycopeptide, as well as enzyme replacement therapy with phenylalanine ammonia lyase, but these therapeutic roads are not yet fully effective [3]. Additionally, more sophisticated approaches, like chaperone and gene

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

therapies, are still far in the horizon. Recently, however, a new therapy is becoming widely used, the supplementation with pharmacological doses of the PAH cofactor, which can alleviate, and in special cases to avoid, the diet burden [4–7]. The molecular basis of BH4 action is not fully understood, but the effect upon PAH gene expression was discarded. BH4 can act as a chemical chaperone helping the stabilization of some mutant forms or, by increasing the cofactor concentration, it can promote the enzymatic activity of low-affinity mutants [8,9]. Accordingly, BH4 seems to display a multifactorial mechanism of action. PAH gene located in the long arm of chromosome 12 (12q24.1), covers ~100 kb of genomic DNA and is structured in 13 exons separated by introns. The messenger is 2448 bp long, being translated into 452 amino acid polypeptides, which are assembled onto functional homotetramers [1]. More than 600 different mutations have already been described (PAHdb, www.pahdb.mcgill.ca), being the majority missense ones, and every population displays a mutational spectrum characterized by a reduced number of prevalent and public mutations and a large number of private pathogenic alterations. Moreover, each mutation impairs PAH activity in a specific manner and, as a consequence of most patients being compound heterozygotes, phenotypes range from the most severe form of classic PKU to the non-PKU hyperphenylalaninemia status. Moreover, several polymorphic markers (RFLP, VNTR and STR) have been defined within the PAH gene. These intragenic polymorphisms have proven to be a useful guide to mutation detection and, besides, have been extensively used in molecular anthropology studies to improve our understanding of the ancestral migratory movements which underlie the present geographic distribution of the most frequent PAH gene mutant alleles [10,11]. As PAH enzyme displays a hepatic expression and liver needle biopsies are no longer justified in PKU patients, the primary source of information concerning residual enzyme activity relies on in vitro expression and analysis of recombinant mutant proteins [12]. The data thus obtained is then used to estimate the patients' phenotype. Genotype–phenotype correlations are difficult to realize in most inherited metabolic disorders, but in PKU and related HPA they revealed to be a strong and reliable predictive tool [12,13]. Accordingly, the characterization of each patient genotype can greatly help to predict the metabolic phenotype, to establish the better tailored diet and, more recently, to assess the potential responsiveness to BH4 therapy. Herein is presented the molecular analysis of the PAH gene, the genotype–phenotype relationships evaluated and the rate of BH4responsiveness predicted in the HPA population living in South Portugal.

2. Materials and methods 2.1. Patients and phenotypic classification Eighty-three patients, encompassing four pairs of siblings and all living in the Southern region of Portugal, were investigated. Most patients were detected by the newborn screening program, running in Portugal since 1979, and whose current cut-off value is 180 μM of phenylalanine. Diagnosis was confirmed after exclusion of BH4 deficiency by evaluating urinary pterin levels and erythrocyte dihydropteridine reductase activity. Metabolic phenotype was assigned to each patient according to pretreatment blood phenylalanine concentrations and, when available, dietary tolerance at 5 years of age. Patients were classified as having classic PKU (pre-treatment Phe levels N1200 μM and Phe tolerance b20 mg/kg/day), moderate PKU (pre-treatment Phe levels between 900 and 1200 μM and Phe tolerance between 20 and 25 mg/kg/day), mild PKU (pre-treatment Phe levels between 600 and 900 μM and Phe tolerance above 25 mg/kg/day) and non-PKU HPA if they keep their Phe levels below 600 μM on a free diet. This study was approved by the local Ethics Committee and informed consents were obtained from the patients or from their parents, who were also enrolled in the study.

S87

2.2. Genotype analysis Genomic DNA was isolated from peripheral blood samples or from Guthrie cards according to a salting-out procedure (Puregene Cell and Tissue kit, Gentra Systems, Minneapolis, MN, USA). After PCR amplification of individual exons and related intronic boundaries, PAH gene (GenBank accession no. AF404777) was scanned for mutations by direct sequence analysis, using the ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit, in an ABI PRISM 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). 2.3. Haplotype analysis Mini-haplotypes were established after PCR-RFLP analysis of the intragenic bi-allelic polymorphisms BglII, PvuIIa, PvuIIb, MspI and XmnI and of the multi-allelic VNTR system at HindIII site at the 3′ region of PAH gene [14]. Haplotype numbering followed the rules of Eisensmith and Woo [15]. 2.4. Calculation of homozigosity (j) Homozigosity (j) at the PAH locus in a given population is determined by j = ∑xi2, where xi is the frequency of the ith allele. In our population, where ascertainment of mutations was not 100%, each of the uncharacterized alleles was defined as having a frequency of 1/N, where N is the total number of mutant chromosomes investigated. 2.5. Relative residual PAH activity and genotype–phenotype correlation Relative residual PAH activity was calculated from data compiled from PAHdb which displays values calculated from in vitro expression of recombinant mutant proteins. PAH activity is defined as the average sum of activities of both individual mutant alleles, and expressed as the percentage of the wild-type enzyme. 3. Results 3.1. Mutation analysis This study involved the molecular characterization of 83 HPA patients living in South Portugal, corresponding to 158 mutant alleles. The results revealed a mutational spectrum encompassing 34 distinct mutations which were distributed along the PAH gene sequence (Table 1). Most mutations were nucleotide substitutions corresponding to 27 missense (79.6%), 3 nonsense (8.8%) and 2 at splicing sites (5.8%); additionally, two deletions were found. Two mutations displayed a relative frequency N10% (IVS10nt-11GNA and p.V388M); a group of five mutations had a frequency between 3.8 and 8.2% (p.R261Q, p.R270K, p. P281L, p.I65T and p.R158Q); another group of seventeen mutations had a frequency in the range 1.3–2.5% and the remaining eight mutations were present in only one mutant allele (0.6% each). The majority of the mutations were situated in the catalytic domain (85%), while two were located in the regulatory domain (6%) and three in the tetramerization domain (9%). A detection rate of 89.9% was achieved, with complete genotyping of 67 patients, while in the remaining sixteen individuals only one causative mutation was identified. Among the fully genotyped patients, only 14 (20.9%) were homozygous, the majority displaying compound heterozigosity for two different mutations. Among the 14 homozygous patients, three individuals carried the IVS10nt-11GNA mutation, two individuals carried the p.R261Q mutation and another two the p.R270K allele. The remaining seven patients harbored p.I65T, p.R176X, p.E280K, p.P281L, p.L311P, p.T323del and p.V388M mutations in homozigosity. Moreover, it was interesting to note the wide array of genotypes (Table 2). Among the 67 fully genotyped patients we could observe 51

S88

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

Table 1 Characterization of PAH mutations identified in 83 South Portuguese HPA patients. Sequence variation

PAH mutation

Alleles (n)

Allele frequency (%)

PAH activity (%)

Mini-Haplotype (RFLP.VNTR)

BH4-responsiveness

c.1066-10GNA c.1162GNA c.782GNA c.809GNA c.842CNT c.194TNC c.473GNA c.526CNT c.754CNT c.967_969delACA c.745CNT c.781CNT c.932TNC c.204ANT c.527GNT c.593_614del22 c.728GNA c.727CNT c.805ANC c.838GNA c.898GNT c.1045TNC c.1169ANG c.1241ANG c.533ANG c.688GNA c.722GNA c.926CNT c.965CNG c.1042CNG c.965CNG c.1222CNT c.1243GNA c.1315+1GNA Unknown Total

IVS10nt-11GNA p.V388M p.R261Q p.R270K p.P281L p.I65T p.R158Q p.R176X p.R252W p.T323del p.L249F p.R261X p.L311P p.R68S p.R176L p.Y198_E205NCfs p.R243Q p.R243X p.I269L p.E280K p.A300S p.S349P p. E390G p.Y414C p.E178G p.V230I p.R241H p.A309V p.A322G p.L348V p.A403V p.R408W p.D415N IVS12nt+1GNA

23 17 13 12 10 8 5 4 4 4 3 3 3 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 16 158

14.6 10.8 8.2 7.6 6.3 5.1 3.8 2.5 2.5 2.5 1.9 1.9 1.9 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 10.1 100.0

b1 27 27 b1 b1 27 10 b1 b1 b1 ? b1 b1 76 42 0 10 b1 63 b1 ? b1 75 38 ? 63 23 70 75 33 N 70 b1 72 b1

6.7 1.7 1.8 1.8 1.8/2.3/16.3 9.8/57.8 4.3 1.8 1.8 1.7/1.8 1.8 5.8 3.8 1.8 4.3 5.8 1.8 7.8 4.3 1.8 1.8 4.3 2.3 4.3 1.7 5.9 5.9 1.7 1.9 9.8 1.8 1.8 1.8 2.3

− + + − − + − − − − ? − − + + − + − + − + − + + + + + + + + + − + −

different genotypic combinations: 2 genotypes were found in 3 patients and 12 genotypes were detected in 2 patients, while the remaining genotypic combinations were observed in single patients. Interestingly, the majority of the genotypes involve a BH4-responsive mutation. Moreover, and besides the pathogenic mutations, six different single nucleotide polymorphisms (SNPs) were also identified among our patients, namely IVS5nt-54GNA (7.0%), p.Q232Q or rs1126758 (13.3%), p.V245V or rs1042503 (9.5%), p.T328T (0.6%), p.L385L or rs772897 (3.2%) and p.Y414Y or rs1801152 (1.3%). Each of these silent mutations has already been described, except p.T328T found for the first time in this study in a single allele. The evaluation of the homozigosity rate in the South Portuguese HPA population revealed a high genetic heterogeneity (j = 0.059), similar to the one observed among other South European populations (Sicily j = 0.06) or ethnically mixed populations in Germany and The Netherlands [16].

3.2. Haplotype analysis The characterization of the haplotypic background underlying the PKU locus in the Portuguese HPA patients was achieved in 153 mutant chromosomes. The analysis revealed the presence of 12 diverse haplotypes, with the HindIII VNTR polymorphic site displaying heterogeneity of haplotypes 1 and 5. It is interesting to note the high prevalence of haplotype 1 (14.3% for VNTR 7, 41.8% for VNTR 8 and 0.7% for VNTR 9) which is prevalent among mutant chromosomes with 18 different mutations associated with it. Among the remaining haplotypes, the most

frequent are haplotype 6.7 (13.7%), haplotype 4.3 (10.4%) and haplotype 9.8 (4.6%), reflecting the presence of the frequent mutations IVS10nt11GNA, p.R158Q and p.I65T, respectively (see Table 1). All haplotypes but two were observed either among mutant or normal chromosomes (data not shown). However, haplotypes 6.7 and 9.8 were only identified as background of mutant alleles, IVS10nt11GNA the first and p.I65T and p.L348V the second.

3.3. Mutation–haplotype correlation Almost all the mutations were found associated with a single haplotype (Table 1), corresponding to the one usually linked to them, which indicates that the hypothesis concerning their origins may also be applied to the Portuguese population. The three exceptions are p.I65T, p. P281L and IVS10nt-11GNA. Most mutant alleles coding for p.I65T were found associated with the predominant haplotype 9.8 and single alleles with haplotypes 1.8 and 57.8. Concerning mutation p.P281L the majority of the alleles were expressed on the background of haplotype 1.8, while two alleles were associated with haplotypes 2.3 and 16.3. The most interesting case was the mutant allele IVS10nt-11GNA, always associated with haplotype 6.7 except in a gipsy patient displaying a 34.7 haplotypic background; this situation has already been observed among the Spanish PKU population, where the IVS10nt-11GNA mutation carried by patients with gipsy ancestry was always associated with haplotype 34.7. The PAH mutation database displays only one entry [17] for p. T323del mutation, with no haplotypic association. The study of

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

S89

Table 2 Genotypic and phenotypic data of 63 South Portuguese HPA patients. Patient

Allele 1

Allele 2

Pre-treatment Phe levels (μM)

Clinical phenotypes

1 2 3 4 5 6 7 8 9 10 11* 12* 13 13 15 16 17* 18* 19* 20* 21 22 23 24 25* 26* 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

p.P281L p.P281L p.P281L IVS10nt-11GNA IVS10nt-11GNA IVS10nt-11GNA p.I65T p.I65T p.I65T p.I65T p.R243X p.R243X p.R261Q p.R261Q p.R261Q p.R261Q p.R261Q p.R261Q p.R261X p.R261X p.R270K p.R270K p.R270K p.R270K p.A322G p.A322G IVS10nt-11GNA IVS10nt-11GNA p.V388M p.V388M p.I65T p.I65T p.R68S p.R158Q p.R158Q p.R158Q p.R158Q p.R158Q p.R176L p.R176X p.R176X p.R176X p.Y198_E205NCfs p.V230I p.R241H p.R243Q p.L249F p.L249F p.L249F p.R252W p.R252W p.R252W p.R261Q p.R261Q p.I269L p.R270K p.R270K p.R270K p.R270K p.E280K p.P281L p.A300S p.L311P p.T323del p.T323del IVS10nt-11GNA p.V388M

IVS10nt-11GNA IVS10nt-11GNA IVS10nt-11GNA IVS10nt-11GNA IVS10nt-11GNA IVS10nt-11GNA p.R261Q p.R261Q p.P281L p.P281L IVS10nt-11GNA IVS10nt-11GNA p.R261Q p.R261Q p.V388M p.V388M p.R408W p.R408W p.I269L p.I269L p.R270K p.R270K IVS10nt-11GNA IVS10nt-11GNA p.S349P p.S349P p.V388M p.V388M p.Y414C p.Y414C p.I65T p.R243Q p.E178G p.Y198_E205NCfs p.R270K p.P281L p.R176L p.R261Q p.E390G p.R176X IVS10nt-11GNA p.V388M p.R243X p.P281L p.R261Q p.V388M p.R261X p.A309V p.V388M p.T323del IVS10nt-11GNA p.V388M p.L311P IVS10nt-11GNA IVS10nt-11GNA p.L348V p.S349P p.E390G p.D415N p.E280K p.P281L IVS10nt-11GNA p.L311P p.T323del p.V388M p.A403V p.V388M

N 1210a 3329 587 2155 1622a 1826 230 1198 1186 581 4806a N 1210a 1513 1017 999 1501 968a 908a 297 50b 847a 1678 2276 1495 266 255b 1090 1210a 654a 478 1065 968a 278 1210 1332 1271 266 1483 272 N1210 N 1210a 1146 1392 272 599 1108 1174 288 793a 1816a 1798a 1500a 1412 920 242 847 1210 605 212 800 1029 284 1683a 1210 1090a 316 1254

Classic PKU Classic PKU Classic PKU Classic PKU Classic PKU Classic PKU Non-PKU HPA Moderate PKU Moderate PKU Moderate PKU Classic PKU Classic PKU Moderate PKU Moderate PKU Moderate PKU Moderate PKU Moderate PKU Moderate PKU Non-PKU HPA Non-PKU HPA Classic PKU Classic PKU Classic PKU Classic PKU Non-PKU HPA Non-PKU HPA Moderate PKU Classic PKU Mild PKU Mild PKU Moderate PKU Moderate PKU Non-PKU HPA Classic PKU Classic PKU Classic PKU Non-PKU HPA Classic PKU Non-PKU HPA Classic PKU Classic PKU Mild PKU Classic PKU Non-PKU HPA Moderate PKU Moderate PKU Moderate PKU Non-PKU HPA Moderate PKU Classic PKU Classic PKU Classic PKU Classic PKU Classic PKU Non-PKU HPA Moderate PKU Classic PKU Non-PKU HPA Non-PKU HPA Classic PKU Classic PKU Non-PKU HPA Classic PKU Classic PKU Classic PKU Non-PKU HPA Moderate PKU

a b

Patients born before the implementation of the Newborn Screening Program in Portugal, in 1980, and accordingly diagnosed on the basis of altered phenotype. These patients, siblings of the index cases, were missed on the newborn screening test and were later detected during family studies.

Bercovich and colleagues [18] gives no information on the haplotypic background of this deletion. The present study revealed that it is always expressed on the background of haplotype 1.

As in other populations, the most frequent SNPs were strongly associated with specific haplotypes: p.Q232Q with haplotypes 3, 4 and 7; p.V245V exclusively with haplotype 4; and p.L385L with haplotypes 3

S90

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

and 7. The IVS5nt-54GNA and p.Y414Y polymorphisms were always observed in linkage disequilibrium with haplotype 1.8. 3.4. Genotype–phenotype correlation According to the phenotypic classification indicated in the Methods section, out of the 83 patients under study, 36 were classified as classic PKU (43.4%), 25 as moderate PKU (30.1%), 5 as mild PKU (6.0%) and 17 as non-PKU HPA (20.5%). Table 2 lists the 67 fully genotyped patients along with their respective phenotype and a good genotype–phenotype correlation could be observed. Concerning the homozygous patients, those carrying severe or null mutations [p.R176X, p.E280K, p.P281L, p. L311P, p.T323del (all n = 1), p.R270K (n = 2) and IVS10nt-11GNA (n = 3)] were classified as classic PKU, while those harboring mutations with some residual enzymatic activity [p.I65T and p.V388M (both n = 1) and p.R261Q (n = 2)] displayed moderate PKU. The nine cases where a genotype–phenotype inconsistency was observed corresponded to compound heterozygous patients, all of them carrying p.R261Q or p.V388M mutations. Both transitions are recognized as inducing negative inter-allelic complementation [19] and, accordingly, patients displayed phenotypes more severe than expected. The only exception was a patient harboring the genotype p.I65T/p.R261Q and showing a much milder phenotype, non-PKU HPA. 3.5. Genotype and BH4-responsiveness One major goal of this study was to evaluate the rate of BH4responsiveness among South Portugal PKU patients, aiming the rationale identification of those displaying a potential positive response to this novel pharmacologic therapy. Notably, more than half (I65T, R68S, R176L, E178G, V230I, R241H, R243Q, R261Q, I269L, A300S, A309V, A322G, L348V, V388M, E390G, A403V, Y414C and D415N) of the 34 mutations identified among South Portugal HPA patients belong to the group of more than 70 mutations already identified in BH4-responsive patients, according to BIOPKU database (www.biopku.org), and the majority lies in the catalytic domain of the protein. The analysis of the allelic data allowed anticipating that 62.7% of the fully genotyped patients are potential responders to BH4 treatment (Table 3). More than half (52.4%) of the patients display the appropriate metabolic phenotype, moderate or mild PKU, showing good perspectives for an in vivo responsiveness. A minority (11.9%) display classic PKU, and their answer to BH4 supplementation can only be disclosed by a physiological challenge, because all these patients carry in heterozigosity the p.R261Q and p.V388M mutations, known for their potential induction of negative inter-allelic complementation, as previously referred. Finally, and notably, the remaining 15 patients have a non-PKU phenotype, are already on a free diet and accordingly should not need BH4 supplementation. 4. Discussion Molecular characterization of the HPA population living in South Portugal confirmed its high genetic heterogeneity, originating on the Table 3 Anticipation of BH4-responsiveness in South Portugal HPA patients. Number of patients

Frequency (%)

Phenotype

Frequency (%)

Potential responders

42

62.7

Potential nonresponders

25

37.3

Classic PKU Moderate/mild PKU Non-PKU HPA Classic PKU

11.9 52.4 35.7 100.0

large number of different people who populated the south-western part of the Iberian Peninsula, that later became Portugal. This heterogeneity is a common feature of South European populations [20–22] in contrast with the North-eastern ones, especially in Poland [13]. The mutational spectrum involves 34 mutations, most of them falling on the category of missense type. Only seven transversions were observed, which confirms the higher transition rate over transversion rate [23]. Moreover, eight CNT transitions were identified, five of them located in CpG sites, known as mutational hot spots [24]. CpG methylation has already been identified as accounting for recurrence of p.R408W mutation [25]. Additionally, it is interesting to note that the number of mild mutations exceeds the severe ones: 16 mutations are classified as severe, while 18 mutations display considerable residual activity, ranging from moderate until high enzymatic activities (Table 1). All the mutations identified in our patients have already been described in other populations, apart from p.R270K. Indeed, this mutation has only been identified in Portuguese patients or in foreign individuals (USA) but with Portuguese ancestry. Notably, any Spanish patient was detected carrying this mutation, which may denote a local origin. Another peculiar case is the p.T323del mutation found in our PKU population. Previously, it had been detected in only a single individual during the study of a heterogeneous USA population [17], but recently two alleles were identified among the Ashkenazi Jewish population living in Israel [18]. Portuguese population involves a well-represented Jewish community and, besides, it has a long tradition of emigration. Accordingly, we may postulate that those patients could probably have a Portuguese ancestry, because this mutation is well represented among our PKU population (4 alleles). The correlation between each mutation and its haplotypic background allowed us to confirm that the hypothesis underlying each mutation origin(s) could be applied to our population and also to trace the very origin of each mutation. Effectively, some mutations are well represented among all populations, though associated with different haplotypes. This is the case of p.S349P mutation which is detected on the background of haplotype 1 in North and Center Europe and haplotype 4 in North Africa populations [26]. All the Portuguese patients carrying this mutant allele displayed the haplotype 4.3, thus putting in evidence the strong colonization of South Portugal by the Moors between VIII and XII centuries. Additionally, we could determine that the single p.R408W mutant allele was associated with haplotype 1.8, thus reflecting a Celtic rather than Slavic origin [27,28]. Genotype–phenotype correlation is the cornerstone in most studies on metabolic diseases. Nowadays, it relies on in vitro expression analysis (IVE) of recombinant mutant proteins, using several different systems [29]. These studies revealed that the mutations displaying extreme enzymatic activities in vitro, either null/reduced or elevated, are always associated with the most severe or milder forms of the disease, respectively. Problems arise with mutant enzymes associated with moderate levels of residual activity [16,30]. Accordingly, it has been suggested that IVE may ultimately be important in discriminating between mutations which allow leeway for variability in the enzymatic and metabolic phenotypes and those that do not [29]. In this study, genotype–phenotype relationships were evaluated taking into consideration data from in vitro expression of recombinant mutant proteins, and PAH activity was defined as the average sum of activities of both individual mutant alleles. The results obtained revealed a good correlation between the genotype and the observed phenotype, with only nine inconsistent cases. Moreover, most of these patients were born before the implementation of the newborn screening program in Portugal and, accordingly, were only detected after the appearance of an abnormal phenotype. All the discordant cases involve two specific mutations, p.R261Q and p.V388M, already described by several authors [19,31–33] as responsible for negative interallelic complementation. As both mutations are not located in the

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

tetramerization domain of PAH, it was suggested that they could affect the interactions between the subunit interfaces, namely upon the conformational changes occurring at the dimer interface upon activation by L-phenylalanine [19]. The most striking inconsistency case was the patient carrying the p.I65T/p.R261Q genotype, who should display a moderate phenotype and effectively is a non-PKU HPA individual, requiring no diet. Previous co-expression of these two mutations also revealed problems, hampering the study of interallelic complementation [19]. The evaluation of responsiveness to BH4 treatment in South Portugal patients revealed encouraging results. Effectively, two thirds of the individuals harbor genotypes conferring a potential positive response to this pharmacologic therapy. However, one must stress that false expectations need to be avoided and, accordingly, the decision of BH4 treatment must also take into account some other issues, such as the possibility of negative inter-allelic complementation between some mutant alleles in compound heterozygous patients, as previously referred. Indeed, among our PKU population almost half of the patients carry on their genotypes one allele (p.I65T, p.R261Q and p.V388M) known to induce negative inter-allelic complementation. Moreover, and recently, we became aware that some missense mutations are actually splicing mutations, inducing worse phenotypes than the expected ones. Already in 1999, Ellingsen and co-workers [34] highlighted the fact that aberrant transcripts are a frequent consequence of exonic point mutations, which can alter enhancer or silencer splicing sites. Additionally, the putative missense Y204C mutation revealed to be indeed a splicing mutation and accordingly its nomenclature was changed to p.EX6-96ANG [35]. This fact may explain some of the missing genotype–phenotype correlation observed in some patients, and a reexamination of a large number of disease-causing mutations should be taken into consideration. Very recently, Staudigl and colleagues [36] published important data concerning the influence of substrate and cofactor concentrations, in the presence of certain genotypes, on enzyme function and on response to BH4 treatment. Indeed, BH4 supply can display two different effects, kinetic at a short-term and chaperone at long-term, and individual mutations may shift these effects. Moreover, the initial phenylalanine levels revealed to be important, also. These data are particularly relevant once a significant number of South Portugal HPA patients harbor mutations whose enzymatic activity strongly depends on cofactor and/or substrate concentrations, namely p.I65T, p.R261Q and p.Y414C [36]. So, the present study may benefit from this newly gained knowledge, enabling us to design more personalized screening loading tests, according to each patient genotype. This approach will allow then a more safe and correct assessment of BH4 responsiveness, avoiding false negative responses. In conclusion, a significant number of Portuguese PKU patients are likely to benefit from BH4 therapy which, combined with a less strict diet, or eventually in special cases as monotherapy, may contribute to reduce nutritional deficiencies and minimize neurological and psychological dysfunctions. Additionally, and also importantly, it may contribute to a better quality of life of these patients. References [1] C.R. Scriver, The PAH gene, phenylketonuria, and a paradigm shift, Hum. Mutat. 28 (2007) 831–845. [2] F.K. Treftz, N. Blau, Potential role of tetrahydrobiopterin in the treatment of maternal phenylketonuria, Pediatrics 112 (2003) 1566–1569. [3] F.J. van Spronsen, G.M. Enns, Future treatment strategies in phenylketonuria, Mol. Genet. Metab. 99 (2010) S90–S95. [4] S. Kure, D.C. Hou, T. Ohura, H. Iwamoto, S. Suzuki, N. Sugiyama, O. Sakamoto, K. Fujii, Y. Matsubara, K. Narisawa, Tetrahydrobiopterin-responsive phenylalanine deficiency, J. Pediatr. 135 (1999) 375–378. [5] G. Gramer, P. Burgard, S.F. Garbade, M. Lindner, Effects and clinical significance of tetrahydrobiopterin supplementation in phenylalanine hydroxylase-deficient hyperphenylalaninaemia, J. Inherit. Metab. Dis. 30 (2007) 556–562. [6] I. Karacic, D. Meili, V. Sarnavka, C. Heintz, B. Thöny, D.P. Ramadza, K. Fumic, D. Mardesic, I. Baric, N. Blau, Genotype-predicted tetrahydrobiopterin (BH4)-responsiveness and

[7] [8]

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21] [22]

[23] [24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

S91

molecular genetics in Croatian patients with phenylalanine hydroxylase (PAH) deficiency, Mol. Genet. Metab. 97 (2009) 165–171. C.O. Harding, New era in treatment for phenylketonuria: pharmacologic therapy with sapropterin dihydrochloride, Biologics 4 (2010) 231–236. H. Erlandsen, R.C. Stevens, A structural hypothesis for BH4 responsiveness in patients with mild forms of hyperphenylalaninaemia and phenylketonuria, J. Inherit. Metab. Dis. 24 (2001) 213–230. A.P. Pey, B. Pérez, L.R. Desviat, M.A. Martínez, C. Aguado, H. Erlandsen, A. Gámez, R.C. Stevens, M. Thórólfsson, M. Ugarte, A. Martínez, Mechanisms underlying responsiveness to tetrahydrobiopterin in mild phenylketonuria mutations, Hum. Mutat. 24 (2004) 388–399. B. Pérez, L.R. Desviat, M. Ugarte, Analysis of the phenylalanine hydroxylase gene in the Spanish population: mutation profile and association with intragenic polymorphic markers, Am. J. Hum. Genet. 60 (1997) 95–102. M.G. Mirisola, F. Cali, A. Gloria, P. Schinocca, M. D'Amato, G. Cassara, G. De Leo, L. Palillo, C. Meli, V. Romano, PAH gene mutations in the Sicilian population: association with minihaplotypes and expression analysis, Mol. Genet. Metab. 74 (2001) 353–361. S.F. Dobrowolski, C. Heintz, T. Miller, C. Ellingson, C. Ellingson, I. Ozer, G. Gokçay, T. Baykal, B. Thony, M. Demirkol, N. Blau, Molecular genetics and impact of residual in vitro phenylalanine hydroxylase activity on tetrahydrobiopterin responsiveness in Turkish PKU population, Mol. Genet. Metab. 10 (2011) 116–121. S.F. Dobrowolski, K. Borski, C.E. Ellingson, R. Koch, H.L. Levy, E.W. Naylor, A limited spectrum of phenylalanine hydroxylase mutations is observed in phenylketonuria patients in western Poland and implications for treatment with 6R tetrahydrobiopterin, J. Hum. Genet. 54 (2009) 335–339. I. Rivera, P. Leandro, U. Lichter-Konecki, I.T. De Almeida, M.C. Lechner, Population genetics of hyperphenylalaninaemia resulting from phenylalanine hydroxylase deficiency in Portugal, J. Med. Genet. 35 (1998) 301–304. R.C. Eisensmith, S.L.C. Woo, Update listing of haplotypes at the human phenylalanine hydroxylase (PAH) locus, Am. J. Hum. Genet. 51 (1992) 1445–1448. P. Guldberg, F. Rey, J. Zschocke, V. Romano, B. François, L. Michiels, K. Ullrich, G.F. Hoffmann, P. Burgard, H. Schmidt, C. Meli, E. Riva, I. Dianzani, A. Ponzone, J. Rey, F. Guttler, A European multicenter study of phenylalanine hydroxylase deficiency: classification of 105 mutations and a general system for genotype-based prediction of metabolic phenotype, Am. J. Hum. Genet. 63 (1998) 71–79. R.C. Eisensmith, D. Martinez, A. Kuzmin, A. Goltsov, S.L.C. Woo, Molecular basis of phenylketonuria in a heterogeneous US population, Am. J. Hum. Genet. 57 (1995) A163-(926). D. Berkovich, A. Elimelech, T. Yardeni, S. Korem, J. Zlotogora, N. Gal, N. Goldstein, B. Vilenski, R. Segev, S. Avraham, R. Loewenthal, G. Schwartz, Y. Anikster, A mutation analysis of the phenylalanine hydroxylase (PAH) gene in the Israeli population, Ann. Hum. Genet. 72 (2008) 305–309. J. Leandro, C. Nascimento, I. Tavares de Almeida, P. Leandro, Co-expression of different subunits of human phenylalanine hydroxylase: evidence of negative interallelic complementation, Biochim. Biophys. Acta 1762 (2006) 544–550. P. Guldberg, V. Romano, N. Ceratto, P. Bosco, M. Ciuna, A. Indelicato, F. Mollica, C. Meli, M. Giovannini, E. Riva, et al., Mutational spectrum of phenylalanine hydroxylase deficiency in Sicily: implications for diagnosis of hyperphenylalaninemia in Southern Europe, Hum. Mol. Genet. 2 (1993) 1703–1707. B. Perez, L.R. Desviat, M. De Lucca, M. Ugarte, Spectrum and origin of phenylketonuria mutations in Spain, Acta Paediatr. Scand. Suppl. 407 (1994) 34–36. I. Dianzani, S. Giannattasio, L. de Sanctis, C. Alliaudi, P. Lattanzio, C.D. Vici, A. Burlina, M. Burroni, G. Sebastio, F. Carnevale, et al., Characterization of phenylketonuria alleles in the Italian population, Eur. J. Hum. Genet. 3 (1995) 294–302. D.W. Collins, T.H. Jukes, Rates of transition and transversion in coding sequences since the human–rodent divergence, Genomics 20 (1994) 386–396. D.N. Cooper, M. Krawczak, S.E. Antonorakis, The nature and mechanisms of human gene mutation, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and Molecular Bases of Inherited Disease, McGraw-Hill, New York, 2000, pp. 343–377. B.C. Murphy, C.R. Scriver, S.M. Singh, CpG methylation accounts for a recurrent mutation (c.1222CNT) in the human PAH gene, Hum. Mutat. 27 (2006) 975. M. Weinstein, R.C. Eisensmith, V. Abadie, S. Avigad, S. Lyonnet, G. Schwartz, A. Munnich, S.L.C. Woo, Y. Shiloh, A missense mutation, S349P, completely inactivates phenylalanine hydroxylase in north African Jews with phenylketonuria, Hum. Genet. 90 (1993) 645–649. R.C. Eisensmith, Y. Okano, M. Dasovich, T. Wang, F. Güttler, H. Lou, P. Guldberg, U. Lichter-Konecki, D.S. Konecki, E. Svensson, et al., Multiple origins for phenylketonuria in Europe, Am. J. Hum. Genet. 51 (1992) 1355–1365. E. Treacy, S. Byck, C. Clow, C.R. Scriver, ‘Celtic’ phenylketonuria chromosomes found? Evidence in two regions of Quebec Province, Eur. J. Hum. Genet. 1 (1993) 220–228. P.J. Waters, How PAH gene mutations cause hyper-phenylalaninemia and why mechanism matters: insights from in vitro expression, Hum. Mutat. 21 (2003) 357–369. E. Kayaalp, E. Treacy, P.J. Waters, S. Byck, P. Nowacki, C.R. Scriver, Human phenylalanine hydroxylase mutations and hyperphenylalaninemia phenotypes: a metanalysis of genotype–phenotype correlations, Am. J. Hum. Genet. 61 (1997) 1309–1317. P. Leandro, I. Rivera, M.C. Lechner, T. de Almeida I, D. Konecki, The V388M mutation results in a kinetic variant form of phenylalanine hydroxylase, Mol. Genet. Metab. 69 (2000) 204–212. A. Gámez, B. Pérez, M. Ugarte, L.R. Desviat, Expression analysis of phenylketonuria mutations. Effect on folding and stability of the phenylalanine hydroxylase protein, J. Biol. Chem. 275 (2000) 29737–29742.

S92

I. Rivera et al. / Molecular Genetics and Metabolism 104 (2011) S86–S92

[33] P.J. Waters, C.R. Scriver, M.A. Parniak, Homomeric and heteromeric interactions between wild-type and mutant phenylalanine hydroxylase subunits: evaluation of two-hybrid approaches for functional analysis of mutations causing hyperphenylalaninemia, Mol. Genet. Metab. 73 (2001) 230–238. [34] S. Ellingsen, P.M. Knappskog, J. Apold, H.G. Eiken, Diverse PAH transcripts in lymphocytes of PKU patients with putative nonsense (G272X, Y356X) and missense (P281L, R408Q) mutations, FEBS Lett. 457 (1999) 505–508.

[35] S. Ellingsen, P.M. Knappskog, H.G. Eiken, Phenylketonuria splice mutation (EXON6nt-96ANG) masquerading as missense mutation (Y204C), Hum. Mutat. 9 (1997) 88–90. [36] M. Staudigl, S.W. Gersting, M.K. Danecka, D.D. Messing, M. Woidy, D. Pinkas, K.F. Kemter, N. Blau, A.C. Muntau, The interplay between genotype, metabolic state, and cofactor treatment governs phenylalanine hydroxylase function and drug response, Hum. Mol. Genet. 20 (2011) 2628–2641.