Original Paper

Eur J Hum Genet 1995;3:294-302

1608180 Irma Dianzani* Characterization of Sergio Giannattasioc Luisa de Sanctis* Phenylketonuria Alleles in the Carla Alliaudi* Paolo Lattanzioc Italian Population Carlo Dionisi Viciâ Alberto Buriina6 Massimo Burronif Gianfranco Sebastio* Key Words Franco Carnevaleh Phenylketonuria • Phenylalanine hydroxylase • Vito Gazzetta * Genotype/phenotype correlation Ersilia Marra6 Clara Camaschellaa’b Abstract Alberto Ponzonea In order to identify the molecular basis of phenylketonuria a Clinica Pediatrica e Clinica Medica, (PKU) in Italy, we screened the entire coding sequence of the University of Torino, and phenylalanine hydroxylase gene in 20 Italian PKU patients, b CNR-CIOS, Torino, and whose origins are scattered throughout Italy. The frequency of 0 CNR-Centro di Studio sui Mitocondri e Metabolismo Energetico, Bari and each identified mutation and of 5 other European mutations Trani, and was determined within a panel of 92 Italian PKU patients. This d Ospedale Bambin Gesù, Rome, and approach allowed us to identify 20 different PKU mutations e Dipartimento di Pediatria, University of Padua, and and characterize 64% of the Italian PKU chromosomes. Eleven f Servizio Autonomo di Neuropsichiatria mutations (IYS10nt546, L48S, R158Q, R261Q, P281L, Infantile, Fano, and * Cattedra di Pediatria, II Facoltà di R261X, R252W, AT55, IYS7ntl, IVS12ntl, Y414C) represent Medicina, University of Naples, and 55.4% of the Italian PKU alleles, the most common mutations h Divisione Malattie Metaboliche, being IVS10nt546 (12.4%) and L48S (9%). All the other muta­ Ospedale Pediatrico Giovanni XXIII, Bari, Italy tions are very rare. These data confirm the great heterogeneity expected from previous RFLP haplotype studies. Genotype/ phenotype correlation allowed for assessment of the clinical impact of the 20 identified mutations.

Introduction date, more than 160 candidate mutations have been identified [3], many of which have Cloning of the gene encoding phenylal­ proven to be causal by analysis of the expres­ anine hydroxylase (PAH) in 1985 [1] has sion of the mutated gene in vitro [4], prompted the definition of the molecular ba­ A novel approach to diagnosis, treatment, sis of phenylketonuria (PKU), the disease prognosis and also prevention of the disease caused by loss of function of this gene [2], To has been implemented by these studies. Pre-

Received: December 30,1994 Dr. Irma Dianzani © 1995 Revision received: July 10,1995 Istituto di Clinica Pediatrica S. Karger AG, Basel Accepted. July 19,1995 Piazza Polonia 94 1018-4813/95/ 1-10126 Torino (Italy) 0035-0294S8.00/0 natal diagnosis and carrier detection are easily Padua), two in central Italy (Rome and Fano), and two performed by tracing the inheritance of caus­ in southern Italy (Naples and Bari). The panel of patients is representative of the Italian population, al mutations in characterized families. On since their geographic origin is uniformly spread the other hand, several polymorphic markers throughout Italy. The origin of the patients was as­ within the gene (RFLPs, VNTR, STR) [5, 6] sessed as far back as the four grandparents of the pro­ can be used to identify the mutated uncharac­ bands: 67 chromosomes were original to the northern terized alleles by comparison with an affected and central Italian regions, 110 to the southern regions. The division of Italy into a northem/central and a sibling [7], The elucidation of genotype/phe­ southern region was obtained by tracing a line below notype correlation may allow for early, opti­ Lazio and Marche. mal treatment based on the mutation [8]. PKU diagnosis was assessed after exclusion of de­ However, the possibility of using molecu­ fects in tetrahydrobiopterin [11]. PKU classification lar information in clinics is dependent on was obtained by analysing the following parameters in each patient: (a) pretreatment serum phenylalanine, both precise characterization of mutated al­ (b) time (days) necessary to normalize serum phenylal­ leles as well as mutation frequency in the anine values after introduction of a phenylalanine-free examined population. Great heterogeneity in diet, and (c) phenylalanine tolerance (maximal daily PKU mutations was shown by the earliest phenylalanine intake as required to maintain serum studies focussed on the distribution of RFLP phenylalanine below 360 gmol/1). Using the criteria mentioned above, 50 patients haplotypes at the PAH locus. However, some were assigned to classic PKU, because they showed populations were shown to be relatively ho­ pretreatment plasma phenylalanine higher than 1,500 mogeneous for PKU mutations, with a low gmol/1, dietary tolerance lower than 300 mg/day and number of the represented haplotypes, where­ required more than 4 days for normalization of plasma as others showed a large number of different phenylalanine. Twenty-two patients, assigned to the haplotypes and were expected to carry multi­ mild phenotype, showed pretreatment plasma phenyl­ alanine between 900 and 1,500 gmol/1, dietary toler­ ple mutations. Generally, greater homogene­ ance between 300 and 700 mg/day and required less ity was observed in populations from north­ than 4 days for normalization of plasma phenylala­ ern as compared to southern Europe [4]. Thir­ nine. Six patients were assigned to the benign pheno­ ty-five mutations account for 99% of mutant type, because they showed pretreatment plasma phe­ nylalanine lower than 600 gmol/1, dietary tolerance alleles in Denmark, of which only 4 account higher than 700 mg/day. Moderate PKU was assigned for 70% of the total [9]. On the other hand, the to the 14 patients who showed a pattern midway great number of different haplotypes identi­ between classic and mild phenotypes (for example: fied in the Italian population suggested great pretreatment phenylalanine values >1,500 gmol/1, but heterogeneity of PKU mutations in Italy tolerance >300 mg/day and days for normalization <4). [10]. Genomic DNA was extracted from peripheral The aim of this work was to define the blood leucocytes by standard methods. RFLP haplo­ molecular defects responsible for PKU in the type was determined for 27 patients, 12 of whom were Italian population. part of a previous analysis [10]. For the other 15 haplo- typed PKU patients, RFLPs were detected by either Southern blot analysis or multiplex PCR at the PAH locus followed by restriction digestion [12]. Materials and Methods Mutation Detection Patients Twenty patients, whose origins were spread uni­ Ninety-two Italian PKU patients, corresponding to formly throughout Italy, were selected from the panel 177 unrelated PKU chromosomes, were included in of 92 patients and their DNA was submitted to muta­ this study. Patients were followed up by six clinical tion detection techniques. Selection was made on the departments, two located in northern Italy (Turin and basis of the geographical origin of the patients’ chro-

295 mosomes: 14 chromosomes were original to the north­ evaluated by restriction digestion [16] within the panel ern and central regions (representing 20.9% of north- of the 177 unrelated alleles (table 1). If no restriction em/central chromosomes), 26 chromosomes were orig­ site was either created or abolished in the presence of inal to southern regions (representing 23.6% of south­ the mutation, mutations were introduced in the PCR ern chromosomes). Ten patients showed a classic, 5 a amplification products using mutagenic amplification moderate and 5 a mild phenotype. primers to create a new restriction site in the normal or The 13 PAH exons were amplified by PCR using the mutant allele. The experimental approach used to the appropriate primers complementary to intronic search for each PKU mutation is reported in table 1. regions. Exons 7 to 11 were analysed by chemical S359X and S231P mutations were searched by allele- cleavage method (CCM) [13,14], while exons 3,4,5,8, specific oligonucleotide (ASO) analysis [17] using 9, 10 and 11 were analysed by the combination of sin­ ASOs complementary to the mutated sequences [14, gle-strand conformation polymorphism (SSCP) and 18], heteroduplex methods. All other exons were sequenced directly (exons 1, 2, 6, 12 and 13). Since in previous reports [13,14], exon-intron boundaries of some exons had not been screened, these exons were amplified Results again using primers, allowing us to obtain products encompassing the unscreened regions, after which they Mutation Detection were subjected to direct sequencing or the combina­ Mutation detection analysis of the entire tion of SSCP and heteroduplex methods. In this way, several exons were studied with two different methods, PAH coding sequence allowed for complete but using different PCR primers. characterization of 17 out of 20 patients, with The combination of SSCP and heteroduplex meth­ only one mutation identified in the remaining ods allows a greater sensitivity in mutation detection 3 patients. The 3 uncharacterized mutations as compared to using a single method [15]. Heterodu­ might be located in unscreened regions of plexes are clearly distinguished from SSCP when a non-denatured control is included in electrophoresis: PAH, such as introns (cryptic intronic splice heteroduplexes migrate at the same level as the non- sites) or promoter regions. denatured sample in the bottom portion of the gel, We were able to characterize a total of 64% whereas single strands have a lower migration velocity. (113/177) of the Italian PKU chromosomes. The following SSCP protocol was used. To obtain Twenty mutations have been identified (ta­ labelled products, to one tenth of each PCR reaction were added 1 pi [35S]dATP (3,000 Ci/mmol) and 0.5 U ble 2). Mutations S231P, S359X and IYS7ntl of Taq polymerase, then five more PCR cycles were have been previously reported by our group. performed. 20 pi of 0.1% SDS, 10 mM EDTA pH 8, All mutations are reported in the PAH Muta­ were added to half of the labelled PCR reactions. A tion Analysis Consortium Database [3], fifth of the solution was mixed with an identical vol­ Eleven mutations (IVS10nt546, L48S, ume of stop buffer (95% formamide, 20 mM EDTA pH 8,0.05% xylene cyanol, 0.05% bromophenol blue). R158Q, R261Q, P281L, R261X, R252W, The samples were heated to 95 ° C for 5 min, allowed to AT55, IYS7ntl, IVS12ntl and Y414C) repre­ reanneal at room temperature and finally loaded onto sent 55.4% of the Italian PKU alleles, the 40-cm long, 5-8% polyacrylamide gels in the presence most common mutations being IVS10nt546 of 5% glycerol in 0.5 x TBE buffer. Gels were run at (12.4%) and L48S (9%). All the other muta­ 26 W for 3 h at 4°C, dried on Whatman 3MM paper and exposed on Kodack AR films for 2-6 days. Exons tions are very rare. The A300S mutation, for which a shift in band migration was detected were which has a frequency of 5.7% in Sicily and sequenced using the Sequenase Version 2.0 kit (United seems to be associated with a non-severe phe­ States Biochemical Corporation) after purification of notype, has not been found in our panel of relevant PCR amplification products (Qiaex kit, Qia- patients. gen). The frequency of each newly identified mutation as The RFLP haplotye association has been well as five known mutations described in Europeans assessed for several characterized chromo­ (IVS12ntl, R408W, R252W, P281L and A300S) was somes (table 2).

296 Dianzani et al. PKU Mutations in Italy Table 1. PKU mutations screened in 92 Italian patients

Mutation Localization Oligonucleotide Restriction Origin of (5'-3') enzyme restriction site

L48S(TTG->TCG) exon 2 Ball NA AT55 (TTT ->TT) exon 2 f GGCCAAAGTATTGCGCTTAC Dde I ACMA S67P (TCT -» CCT) exon 2 Xbal NA R158Q (CGG-^TGG) exon 5 f ATCCTGTGTACCGTGCAAGC Mspl ACNA L194P (CTG->CCG) exon 6 f GGGGCACAGTGTTCAAGGCT NlalV ACMA S231P(TCT->CCT) exon 6 ASO R252W (CGG->TGG) exon 7 Aval NA A259V (GCC—>GTC) exon 7 f CTCGGGATTTCTTGGGTAGC Pati NAa R261Q (CGA CAA) exon 7 Hinil NA R261X (CGA -» TGA) exon 7 Dde I MA P281L(CCg->CTg) exon 7 r TAGCTGGAGGACAGTACTCC Mspl ACNA IVS7nt 1 (gt —> at) intron 7 Nlalll MA A300S (GCC ->TCC) exon 8 r CTTACCTGGGAAAACTGCG Avili NA IVS10nt546 (gg->ag) intron 10 Ddel MA AG353 (AAG-> AA) exon 11 ffindlll NA S359X (TCA-»TGA) exon 11 ASO Y386C (TAT->TGT) exon 11 Maelll MA R408W (CGG->TGG) exon 12 Mnll NA IVS12ntl (gt —>at) intron 12 r TCGTAAGGTGTAAATTACGTA Rsal ACNA A403V (GCT ->GTT) exon 12 Bbv I NA Y414C (TAC->TGC) exon 12 Bbv1 MA

f = Forward mutagenic primer; r = reverse mutagenic primer; NA = normal allele; MA = mutant allele; AC = amplification created; ASO = allele specific oligonucleotide. a The mutagenic primer abolishes a further Pall site 1 bp upstream from the relevant Pall site, allowing for resolution of the digestion products.

The following five known polymor­ Genotype/Phenotype Correlation phisms have also been identified: -71A/C, 71 patients (77%) were completely (43) or IVS2ntl9T/C, 696A/G (Q232Q), 735G/A partially (28) characterized: 58 (81.7%) are (Y245V) and 1155G/C (L385L) (PAH cDNA compound heterozygotes for different muta­ numbered according to the PAH Mutation tions and 13 (18.3%) are homozygotes. Geno­ Analysis Consortium Database version [3], type/phenotype correlation has been assessed which represents the combined and revised for the 43 fully characterized subjects, with nucleotide sequence from Kwok et al. [1] and the results summarized in table 3. Konecki et al. [19]). The frequencies for these A tentative clinical classification of PKU polymorphisms were found to be 27.2, 22.2, mutations was deduced from our data using 38,26.6 and 17%, respectively. the following criteria; the amount of residual A previously unreported polymorphism 22 activity in vitro (if data available), the phe­ bp upstream from exon 4 identified in intron notype shown by patients homozygous for 3 (C/T) showed a frequency of 25%. each mutation and the phenotype shown

297 Table 2. Distribution of PKU mutations in 177 unrelated Italian Mutation Numer of positive Frequency RFLP PKU chromosomes chromosomes haplotype

IVS10nt546 22 12.4 6,9a L48S 16 9.0 4 R158Q 10 5.6 1 R261Q 9 5.1 1 P281L 9 5.1 ND R261X 8 4.5 1 R252W 6 3.4 7 AT55 5 2.8 1 IVS12ntl 5 2.8 ND IYS7ntl 4 2.3 1 Y414C 4 2.3 ND A403V 3 1.7 1 A259Y 3 1.7 ND AG363 2 1.1 ND R408W 2 1.1 2 S67P 1 0.6 4 L194P 1 0.6 4 S231P 1 0.6 2 S359X 1 0.6 6 Y386C 1 0.6 ND Uncharacterized 64 36.1 Total 177 100.0

ND = Not determined. IVS10nt546 has been found in association with H9 in a single patient.

when in compound heterozygosity with a se­ in all possible phenotypes, i.e. mild, moderate vere or a non-severe mutation. A mutation or classic. was considered severe either when residual R261X, IVS10nt546, R252W, R158Q, activity lower than 15 % was shown in in vitro P281L and AG363 can be seen as severe muta­ expression experiments, or when it conferred tions, since they confer a classic phenotype in a severe phenotype either in homozygosity or homozygosity in our series (table 3). Moreover, heterozygosity with another severe mutation. they confer a severe phenotype in compound A mutation was considered non-severe when heterozygosity with other mutations, such as a residual activity higher than 15% was IYS12ntl [8], R408W [8] and AT55 [20], clas­ shown in in vitro expression experiments, or sified as severe by data from the literature. when it conferred a non-classic phenotype Mutations R403V and S67P were consid­ either in homozygosity or heterozygosity with ered non-severe, because the first confers another non-severe mutation. On the other a mild or a benign phenotype in associa­ hand, we observed that the association of a tion with severe mutations, such as P281L, non-severe with a severe mutation can result R261X and R158Q, while the second confers

298 Dianzani et al. PKU Mutations in Italy mild PKU when associated with other mild Table 3. Genotype/phenotype association in 43 mutations, such as L48S. IVS7ntl and S359X completely characterized PKU patients, 13 homozy­ gotes and 30 compound heterozygotes for different are probably severe mutations, since a disrup­ PKU mutations (one patient identified for each com­ tion of the protein could be expected from the bination, unless otherwise indicated) nature of these defects. In our limited series they give a classic phenotype when associated Genotype Phenotype with a severe mutation, such as IVS10nt546 and AT55. A further mutation, S231P, was Homozygotes IVS10nt546:IVS10nt546 classic (n = 3) shown to have no residual activity when ex­ R261X:R261X classic (n = 2) pressed in vitro [18]. P281L:P281L classic (n = 2) It is interesting to note that the R158Q Y414C:Y414C mild (n = 2) mutation presents a severe phenotype in ho­ R158Q:R158Q classic mozygosity both in our series and in that of R252W:R252W classic AG363:AG363 classic Okano et al. [8], but gives a moderate pheno­ R261Q:R261Q mild type in compound heterozygosity with a se­ vere mutation, such as IVS12ntl or a non- Compound heterozygotes severe mutation such as R26 IQ (in 2 patients) IVS7ntl:R261Q classic or L48S (in 1 patient). A residual activity of IVS10nt546:A259V classic IVS10nt546:IVS7ntl classic (n = 2) 10% shown by the mutated enzyme in expres­ IVS10nt546:P281L classic sion experiments might explain this peculiar IVS10nt546:AT 5 5 classic pattern [8]. IVS10nt546:R261Q classic L48S, R261Q and Y414C mutations could IYS12ntl:P281L classic be classified as non-severe, as found in other L194P:R261X classic R252W:L48S classic series [8, 21, 22, 23, 24], A mild phenotype R408W:L48S classic was shown by the single patient homozygous S231P:R261Q classic for R261Q and by the 2 patients homozygous S359X:AT55 classic for Y414C. A mild phenotype was shown by R158Q:Y386C classic compound heterozygotes between L48S and L48S:IVS10nt546 moderate; mild (n = 2) L48S:IYS7ntl moderate other non-severe mutations (R261Q, S67P). L48S:R158Q moderate All but Y414C (which was never found associ­ R158Q:R261Q moderate (n = 2) ated with another mutation) were shown to R158Q:IVS12ntl moderate give either a classic or mild or moderate phe­ R261Q:AT55 moderate notype when in compound heterozygosity R261Q:L48S mild mild with a severe mutation. This pattern was also L48S:AT55 L48S:S67P mild observed when analysing incompletely char­ A403V:P281L mild acterized patients. The 7 patients who carried A403V:R261X mild L48S in association with an undetected muta­ A403V:R158Q benign tion showed all varieties of phenotypes: 2 clas­ R261Q:AG363 moderate sic, 3 mild, 1 moderate and 1 benign. Four prenatal diagnoses [7, 25] and 6 car­ rier diagnoses were requested during the peri­ od of the study. In all cases, diagnosis was proven to be easily assignable with either RFLPs, VNTR, STR or mutation analysis.

299 Discussion ian region might allow identification of a panel of PKU mutants specific for each re­ The present work attempted to evaluate gion. This might suggest new hypotheses re­ the molecular basis of PKU in Italy. The main garding the origin and spread of other PKU result was the identification of extreme het­ mutations in Italy. Moreover, this analysis erogeneity, much higher than in other Euro­ might be useful for the introduction of region­ pean populations. Our approach allowed its to al programs for carrier screening or for prena­ characterize 64% of the Italian PKU chromo­ tal diagnosis using methods able to screen somes. By analogy with a recent study fo­ simultaneously for the most frequent muta­ cussed on the inhabitants of Sicily [26], it can tions [29], including the 10-15 mutations be argued that the remaining 36% of uncha­ which occur more frequently in each Italian racterized chromosomes represent rare or pri­ region. Use of polymorphic markers could be vate muations. the method of choice for rare or uncharacter­ To date, the extreme heterogeneity of PKU ized DNA changes. mutations in Italy precludes the introduction As a consequence of heterogeneity, most of of programs for carrier detection at a popula­ the patients were found to be compound het­ tion level, thus restricting family counselling erozygotes for different mutations, thus mak­ to families with an affected sibling. In fact, to ing the definition of the role of each mutation be effective in terms of costs/benefits, a car­ in the pathogenesis of PKU in Italy difficult to rier screening program should only be carried obtain. To circumvent this problem, we de­ out in populations where at least 95% of mu­ vised a tentative approach by taking into tations are characterized and where the num­ account patients’ clinical parameters, bio­ ber of mutant alleles is not excessively large. chemical parameters obtained with in vitro The cause for the extreme heterogeneity of expression studies, and association with other PKU in Italy might be ascribed to the differ­ mutations. The introduction of a fourth phe­ ent ethnic origins of the many populations notype, named ‘moderate PKU’ [30], allowed which inhabited ancient Italy to become the us to resolve discrepancies found by Okano et progenitors of the modem Italian population al. [8], who observed an overlap between clas­ [27], Three mutations show a non-random sic and mild phenotypes. geographic distribution and different rel­ We observed that the association of a mild ative frequencies in northem/central and with a severe mutation could result in all pos­ southern Italian regions [27], Two of them, sible phenotypes. This has been observed by IVS10nt546 and L48S, are more frequent in other authors regarding R261Q, which in southern Italy, whereas IVS12ntl is more heterozygosity has generally been associated common in northern Italy. We have proposed with a less severe phenotype, but also with that IVS10nt546 and L48S might have their classic PKU [8, 24, 31]. In other studies, this origin in the Middle East. From there they mutation had been found to be associated might have been carried to western and north­ with both a benign and mild PKU when ern regions by the migration of Neolithic homozygous, but in our series the single ho­ farmers [27], On the other hand, the distribu­ mozygote shows a mild PKU. tion of IVS12ntl fits the hypothesis of an ori­ Although the causes for this variability are gin within Uralic-speaking populations [28]. unknown, it is probable that other factors may The extension of molecular analysis to a influence the phenotype. Negative interaction larger number of patients original to each Ital­ between different mutant subunits might oc-

300 Dianzani et al. PKU Mutations in Italy cur in the mature oligomeric protein. This quency of non-severe mutations, such as L48S might be the case for mutations associated and R261Q, could well be the cause of the with some residual activity, such as R158Q. lower severity of PKU in Italy compared to In addition, the hydroxylating system might northern Europe, where severe mutations be modulated by modifier genes. It is also (such as IVS12ntl and R408W) are preva­ important to note that expression analyses are lent. usually performed in monkey-kidney cell or Escherichia coli systems, which are markedly Acknowledgements different from the human liver environment in vivo. We wish to thank Dr. Savio Woo for having dis­ Finally, the association of mutations with closed the sequence of many primers prior to publi­ cation and Prof. G. Saglio for helpful discussion. specific phenotypes might account for the dif­ This research was partially supported by Fondazione ferent clinical presentation of PKU among Biotecnologie, Torino and by CNR target project populations. Seen in this light, the higher fre­ FATMA.

References

1 Kwok SCM, Ledley FD, DiLella 6 Goltsov AA, Eisensmith RC, 10 Dianzani I, Devoto M, Camaschella AG, Robson KJH, Woo SLC: Nu­ Naughton ER, Jin L, Chakraborty C, Saglio G, Ferrerò GB, Cerone R, cleotide sequence of a full-length R, Woo SLC: A single polymorphic Romano C, Romeo G, Giovannini complementary DNA clone and STR system in the human phenyl­ M, Riva E, Angeneydt F, Trefz FK, ammo acid sequence of human phe­ alanine hydroxylase gene permits Okano Y, Woo SLC: Haplotype dis­ nylalanine hydroxylase. Biochemis­ rapid prenatal diagnosis and car­ tribution and molecular defects at try 1985;24:556-561. rierscreening for phenylketonuria. the phenylalanine hydroxylase locus 2 Scriver CR, Kaufman S, Woo SLC: Hum Mol Genet 1993;5:577-581. in Italy. Hum Genet 1990;86:69— The hyperphenylalaninemias; in 7 Romano V, Dianzani I, Ponzone A, 72. Scriver CR, Beaudet AL, Sly WS, Zammarchi E, Eisensmith R, Cerat- 11 Ponzone A, Guardamagna O, Fer­ Valle D (eds): The Metabolic Basis to N, Bosco P, Indelicato A: Prena­ raris S, Ferrerò GB, Dianzani I, Cot­ of Inherited Disease. New York, tal diagnosis by minisatellite analy­ ton RGH: Tetrahydrobiopterin McGraw Hill, 1989, pp 495-546. sis in Italian families with phenylke­ loading test in hyperphenylalani- 3 Scriver CR: PKU Mutation Analy­ tonuria. Prenat Diagn 1994; 14:959— nemia. Pediatr Res 1991;30:435- sis Consortium Database. Novem­ 962. 438. ber 1994. 8 Okano Y, Eisensmith RC, Güttler F, 12 Giannattasio S, Lattanzio P, Dian­ 4 Eisensmith RC, Woo SLC: Molecu­ Lichter-Konecki U, Konecki DS, zani I, Carnevale F, Quagliariello lar basis of phenylketonuria and re­ Trefz FK, Dasovich M, Wang T, E, Marra E: Detection of three re­ lated hyperphenylalaninemias: Mu­ Henriksen K, Lou H, Woo SLC: striction fragment length polymor­ tations and polymorphisms in the Molecular basis of phenotypic het­ phisms at the human phenylalanine human phenylalanine hydroxylase erogeneity in phenylketonuria. N hydroxylase locus by multiplex poly­ gene. Hum Mutât 1992;1:13-23. Engl J Med 1991;324:1232-1238. merase chain reaction. Biochim Clin 5 Goltsov AA, Eisensmith RC, Ko­ 9 Guldberg P, Henriksen KF, Güttler 1993;17:1132-1134. necki DS, Lichter-Konecki U, Woo F: Molecular analysis of phenylke­ 13 Dianzani I, Forrest SM, Camaschel­ SLC: Associations between muta­ tonuria in Denmark: 99% of the mu­ la C, Saglio G, Ponzone A, Cotton tions and a VNTR in the human tations detected by denaturing gra­ RGH: Screening for PKU mutations phenylalanine hydroxylase gene. dient gel electrophoresis. Genomics in the phenylalanine hydroxylase Am J Hum Genet 1992;51:627— 1993;17:141-146. gene from Italian patients with phe­ 636. nylketonuria using the chemical cleavage method: A new splice mu­ tation. Am J Hum Genet 1991,48: 631-635.

301 14 Dianzani I, Camaschella C, Saglio 21 Rey F, Abadie V, Lyonnet S, Ber­ 27 Dianzani I, Giannattasio S, de Sanc­ G, Ferrerò GB, Ramus S, Ponzone theion M, Caillaud D, Melle D, La- tis L, Marra E, Ponzone A, Camas­ A, Cotton RGH: Molecular analysis brune P, Saudubray JM, Munnich chella C, Piazza A: Genetic history of contiguous exons of phenylala­ A, Rey J: Expression phénotypique of phenylketonuria mutations in Ita­ nine hydroxylase: Identification of a de 12 mutations du gène de la phé- ly. Am J Hum Genet 1994;55:851— novel PKU mutation. J Med Genet nylalanine hydroxylase. Arch Fr Pé- 853. 1993;30:228-231. diatr 1992;49:705-710. 28 Eisensmith RC, Okano Y, Dasovich 15 White MB, Carvalho M, Derse D, 22 Konecki DS, Schweitzer-Krantz S, M, Wang T, Güttler F, Lou H, Guld­ O’Brian SJ, Dean M: Detecting sin­ Byrd D, Trefz FK, Lichter-Konecki berg P, Lichter-Konecki U, Konecki gle base substitution as hetero­ U: Facilitation of hyperphenylalani- DS, Svensson E, Hagenfeldt L, Rey duplex polymorphisms. Genomics nemia phenotype assessment by ge­ F, Munnich A, Lyonnet S, Cockbum 1992;12:301-306. notype analysis. Pediatr Res 1993; F, Conor JM, Pembrey ME, Smith I, 16 Eiken HG, Ödland E, Boman H, 152:1048-1049. Gitzelmann R, Steinmann B, Apold Skjelkvale L, Engebretsen LF, Apold 23 Svensson E, von Dobeln U, Eisen- J, Eiken HG, Giovannini M, Riva E, J: Application of natural and ampli­ smith RC, Hagenfeldt L, Woo SLC: Longhi C, Romano V, Cerone R, fication created restriction sites for Relation between genotype and phe­ Naughten ER, Mullins C, Cahalane the diagnosis of PKU mutations. notype in Swedish phenylketonuria S, Ozalp I, Fekete G, Schuler D, Be- Nucleic Acids Res 1991 ; 19:1427— and hyperphenylalaninemia pa­ rencsi GY, Nasz I, Brdicka R, Ka- 1430. tients. Eur J Pediatr 1993; 152:132— maryt J, Pijackova A, Cabalska B, 17 Saiki RK: Genetic analysis of 139. Boszkowa K, Schwartz E, Kalinin enzymatically amplified ß-globin 24 Kleiman S, Vanagaite L, Bernstein VN, Jin L, Chakraborty R, Woo and HLA-DQalpha genomic DNA J, Schwartz G, Brand N, Elitzur AA, SLC: Multiple origins for phenylke­ with allele specific oligonucleotide Woo SLC, Shiloh Y: Phenylketonu­ tonuria in Europe. Am J Hum Ge­ probes. Nature 1987;324:163-166. ria: Variable phenotypic outcomes net 1993;51:1355-1365. 18 Dianzani I, Knappskog PM, de of the R261Q mutation and ma­ 29 Saiki RS, Walsh PS, Levenson CH, Sanctis L, Riva E, Ponzone A, Apold ternal PKU in the offspring of a Erlich HA: Genetic analysis of J, Camaschella C: A novel missense healthy homozygote. J Med Genet amplified DNA with immobilized mutation in the PAH gene leading to 1993;30:284-288. sequence-specific oligonucleotide complete loss of enzymatic activity. 25 Ponzone A, Dianzani I, Spada M, de probes. Proc Natl Acad Sci USA Hum Mutât, in press. Sanctis L, Guardamagna O, Viora 1989;86:6230-6234. 19 Konecki DS, Wang Y, Trefz FK, E, Ponzone R, Kierat L, Leimbacher 30 Güttler F, Guldberg P, Henriksen Lichter-Konecki U, Woo SLC: W, Matasovich A, Blau N: Prenatal KF, Mikkelsen I, Olsen B, Lou H: Structural characterization of the 5' diagnosis in primary hyperphenylal- Molecular basis for the phenotypical regions of the human phenylalanine aninemias. Dev Brain Dysfunct diversity of phenylketonuria and re­ hydroxylase gene. Biochemistry 1993;6:158-167. lated hyperphenylalaninaemias. J 1992;31:8363-8368. 26 Guldberg P, Romano V, Ceratto N, Inherit Metab Dis 1993; 16:602— 20 Eigel A, Dwomiczak B, Kalaydjieva Bosco P, Ciuna M, Indelicato A, 604. L, Horst J: A frameshift mutation in Mollica F, Meli C, Giovannini M, 31 Avigad S, Kleiman S, Weinstein M, exon 2 of the phenylalanine hydrox­ Riva E, Biasucci G, Henriksen KF, Cohen BE, Schwartz G, Woo SLC, ylase gene linked to RFLP haplo­ Güttler F: Mutational spectrum of Shiloh Y : Compound heterozygosity type 1. Hum Genet 1991;87:739- phenylalanine hydroxylase deficien­ in nonphenylketonuria hyperphe­ 741. cy in Sicily: Implications for diagno­ nylalaninemia: The contribution of sis of hyperphenylalaninemia in mutations of classical phenylke­ southern Europe. Hum Mol Genet tonuria. Am J Hum Genet 1991;49: 1993;2:1703-1707. 393-399.

302 Dianzani et al. PKU Mutations in Italy