Molecular Analysis of 6-pyruvoyltetrahydropterin Synthase Gene in Atypical Phenylketonuric Egyptian Patients

This study carried out on selected 13 Egyptian patients of unrelated families with mild hyperphenylketonuria (MHPA; >200<1200 μmol/L), they were positive for performing the BH4 loading test to be atypical PKU patients to determine the mutation/s in PTPS gene. Patients diagnosed and followed up at Genetics Unit, Pediatric hospital, Ain Shams University from January 2017 to October 2018. They were 7 males and 6 females aged from 2 month to 6 years suffering from poor feeding, delayed motor, mental development and coma. Most of these patients were identified during neonatal screening program.

Selected patients diagnosed and confirmed as MHPA or atypical PKU by measuring the level of phe and phe/tyrosine ratio before and after BH4 analogous loading test by tandem mass. In which, dried blood level of phe concentration at a diagnosis time of selected PKU patients is ≥200 μmol/L and 1200 μmol/L and phe/tyrosine ratio > 3 obtained on two separate samples. They also were positive response to BH4 analogous loading test (the initial dried blood phe level was decreased by at least by 40% after 12 h following one dose of 20 mg/kg Kuvan challenge).

Loading Kuvan test carried out according to the standard protocol of Ain Shams Hospital. In a test dried blood samples were collect on fasting for at least 8 hours, then patients orally received phenylalanine dissolved in breast milk or water at dose of 100 mg/kg body weight. Baseline dried blood Phe level was measured after 3 hr of receiving phe suspension and at that time, the patient treated with 20 mg/kg body weight of Kuvan dissolved also in breast milk. Following Kuvan administration and throughout the entire test, a normal breast milk or regular infant formula-feeding regimen must be provided on demand and Phe-free medical formula must not be given. Dried blood samples were taken at time 4, 8, 12 and 24 hours after Kuvan administration. The test its results must be available over a period of 24 h to avoid delaying treatment of neonates with classical PKU for longer than 24 h. All blood samples used to measure the concentration of phenylalanine and phe/tyrosine ratio as well by tandem mass from either venous blood or dry blood spot.

Patients’ parents/guardians or one of his relatives gave written signed informed consent for participation in the study before any trial related procedures were performed and had to be willing to comply with maintain strict adherence to the diet. This study approved by the Research Ethics Committee (REC) of Medical Hospital Staff of Faculty of Medicine, Ain Shams University (FWA000017585), for experiments involving humans.

Dry blood samples were taken from patients by heel stick, spotted on Whatman filter paper cards (Schleicher and Schuell 903;Dassel, Germany). They left to dry before screening by tandem mass spectrometry (LC/MS-MS; ACQUITY UPLC ®) to determine the level of phe and phe/tyrosine ratio in their blood on normal diet and throw loading test

Bidirectional sequencing of the purified PCR product in both directions (GATC Company, Germany) using ABI 3730XL DNA sequencer (Applied Biosystem, USA) by using forward and reverse primers. Sequencing results are used for nucleotide blast on http://blast.ncbi.nlm.nih.gov/Blast.cgi (National center for Biotechnology) to make alignment with the PTPS gene normal strand by using its accession number NM_000317.3 to detect new mutations. All homozygous and compound heterozygous variants were included.

All detected mutations submitted on https://mutalyzer.nl/name- checker and polyphen2 software to detect the effect of the mutation on the amino acids sequence of the PTPS peptide and detect the type and degree of pathogenicity of mutations. The mutalyzer gives also the expected protein sequence of the missense, nonsense and frame shift mutations. The amino acid sequences submitted to http://raptorx.uchicago.edu/ to detect the expected 3D protein structure of the mutant enzyme and compare it with the normal PTPS. The novel mutations detected from this study were submitted to GenBank database.

Sequence results used for nucleotide blast by the online program http://blast.ncbi.nlm.nih.gov/Blast for alignment with normal PTPS gene to detect a mutations. The three major databases for known pathogenic and likely pathogenic mutations used including ClinVar, OMIM, and HGMD program.

In total, variants of patients were generated and processed to select the candidate pathogenic mutations in coding regions (missense, nonsense, coding indels and frameshift) which were predicted to be neutral, deleterious or damaging by many software programs. Two software to determine the 3D structure and active sites; RaptorX (web server predicting protein structure property solely based on protein sequence or sequence profile) and I- TASSER (Iterative Threading ASSEmbly Refinement). Three software predicting pathogenicity; Sorting intolerant from tolerant (SIFT) (J. Craig Venter Institute, San Diego, CA, USA), protein variation effect analyzer (PROVEAN), and polymorphism phenotyping v2 (PolyPhen-2) (Division of Genetics, Brigham & Women’s Hospital, Harvard Medical School, Boston, MA, USA), were applied to predict the effects of mutations on the protein structure and function. In addition, SPSS version 22 software used for statistical analysis of means ±SD

Almost all patients were suffering from global developmental delay noticed at age of 3 month as most of them could not support neck or recognize parents and other 3 cases (p 7, p 8, p 11) were suffering from convulsions at age of 6 month. They suffering from hypotonia, briskly reflexes or elicited reflexes, except (P11) suffering from mild hypotonia in lower limbs, elicited reflexes and (P1) was with normal tone and normal reflexes. Only one patient (P 12) site lone and recognize families, all other patients were suffering from uncontrolled support neck or recognize family, (P2) was suffering from global developmental delay noticed at age of 3 month.

First blood samples were collected from patient fasting for at least 8 hours, just after collecting fasting sample patients take phenylalanine powder dissolved in water or breast milk (by dose of 100 mg/Kg body weight). Three hour later of phenylalanine administration the second sample was taken, immediately patients administrated once with Kuvan (powder or tablets) at dose of 20mg/kg body weight dissolved in water or apple juice for infant or soluble in breast milk for neonates. After loaded of Kuvan 4 samples were collected at 4, 8, 12 and 24 hours. All blood samples used for measuring phe concentration in umol/l and phe/tyrosine ratio by tandem mass with taking care of clinical behavior during this period.

Patients (P4, 5, 11, 10, 13) could support neck and recognize families, (P2) recognize mother, follow objects; (P3) can sit alone, crawl, recognize parents and say 2 words. Other patient (P11) can stand supported, recognize family, say one word, (P8) mild neck support, can follow objects, (P7) can walk upstairs alone, can say sentence of 3 words, can say 150 words, (P12) can walk supported, (P9) can sit alone and recognize mother, (P1, 6) can walk alone, say one word, recognize family.

The tandem mass spectrometry (LC/MS/MS) result of PKU-BH4 responsiveness patients showed that phe concentration and phe/tyrosine ratio on fasting were highly elevated reaching to 227.8±51.1μmol/L (normally 20- 120 μmol/L) and 6.30±1.2 (normally 0.5-2.0), respectively. Moreover, level of phe and (phe/tyrosine) ratio after three hours of loading 100mg/kg body weight of phenylalanine were significantly elevated (P<0.001) reaching to 810.8±88.3 μmol/L and 22.13±1.9 respectively, suggesting a HPA and PKU cases. On the other hand, analytical results for loading test of Kuvan on Phe level, and Phe/tyrosine ratio showed that all 13 patients were (BH4) Kuvanresponsiveness (atypical phenylketonuric) as summarized in Table 2. The level of phe reduced to 407.7±36.2 μmol/L, 302.0±33.9 μmol/L, 261.5±24.6 μmol/L and 258.4±31.5μmol/L at 4, 8, 12 and 24 hrs. of Kuvan treatment respectively compared to 810.8±88.3 μmol/L at baseline. In addition, the phe/tyrosine ratio reduced and retained to the normal values 6.18±1.7; 7.11±2.1, 6.47±1.1 and 6.08±1.3 at 4, 8, 12 and 24 h respectively, on Kuvan treatment compared to 22.13±1.9 at baseline (Figure 1).

Figure 1.Line chart of Phe/tyrosine ratio before and after Kuvan loaded test among studied patients

Bi-directional Sanger sequencing of the PTPS gene on 13 unrelated Egyptian PKU-HB4 responsiveness patients’ and aligned with PTPS mRNA gene (GenBank NM_000317.3) revealed 6 different kinds of mutations with variable frequencies (one deletion mutation and 5 base substitution mutations; Table 3). Among studied patients, 61.5% (8/13) carried two variant alleles, either compound heterozygous (n = 2) or homozygous (n = 6) genotypes, all are common in same deletion mutation allele beside base substitution allele (three patients carried the same kinds of compound homozygous alleles). Other five patients (5/13; 38.5%) carried only one mutant homozygous allele; 4 of them carried the same deletion mutant allele, and one was base substitution mutation. The mutations were scattered across all six exons of the PTPS gene.

Four novel mutations were identified in the present study; two variations that altered the coding sequences (c.22C>T and c.86A>T, resulting in p.Arg8Cys and p.Lys29Ile, respectively) and one variation produced silent mutation c.273G>A resulting in p.Lys=, in addition to a 23 base pair region deletion (c.164_186delAGTTGTGGTGACAGTACATGCAT resulting in p.Val55Aspfs*2). These four novel mutations were not detected in control chromosomes. Two base substitution mutations previously identified, one missense mutation c.200C>T and other produced a silent mutation c.405T>C resulting in p.Thr67Met and p.135Thr= respectively.

The most frequent mutation among studied patents with PKU is a novel deletion mutation represents about 92.3% occurrence of all MPKU patients. Deletion of 23 base pair at position c.164_186; p.(Val55Aspfs*2) in exon 3 was found in 12 patients (all patients except number 4) that produced a truncated un-functional protein. Different base mutations and corresponding amino acid sequences aligned with wild type PTPS gene as well as 3D structure of protein sequence as represented in Figure 2.

Figure 2.Chromatogram of novel deletion mutation. The upper part of panel A is wild type sequence and the lower part is the truncated mutant PTPS gene. Panel B is the 3D protein structure of PTPS in wild type and mutant one in upper and lower respectively.

The second most frequent mutations among recorded PKU patients atypical type is previously published SNP mutation c.200C>T in exon 4 (Figure 3A) was found in patients number 4, 6, 8, 10, resulting in p.Trp67Met with a big change in 3D structure as represented in the solid line in yellow region (Upper part of Figure 3B). The degree of SNPs pathogenicity in exon 4 was determined by polyphen2 program, which represents a highly pathogenic mutation with score 0.995 where horizontal dark line in the red region (Lower part of Figure 3B).

Figure 3.Chromatogram of mutation of 200C>T on the lower part of panel A with wild type in lower part. Panel B represents the 3D structure in upper region and pathogenicity degree in the lower region A 200C>T mutation previously published. Polyphen2 data shows the mutation is probably damaging with score of 0.995 (sensitivity 0.99 specificity 0.99)

The third frequent mutation 86A>T in exon 2 is a novel mutation among patients representing p.Lys29Ile and located in patients 7 and 9 as represented in Figure 4A. The 3D structure and the degree of SNPs pathogenicity was determined by polyphen2 program, which represent a tolerated mutation due to represent of horizontal dark line in the green region (Figure 4B).

Figure 4.Chromatogram of novel mutations. Panel A sequence analysis of A>T 86 mutation in the lower part of panel A to compare with wild type sequence in the upper part. 3D structure of mutated protein in upper region and pathogenicity of the mutation by polyphen2. The probability is benign with score of 0.001 (sensitivity 0.99 and specificity 0.95)

The least frequent mutation represented only one patient of each mutation type. SNP 22C>T in exon 1 representing a novel missense mutation p.Arg8Cys in patient 11 (Figure 5A); while G>A at position 273 in exon 5 representing synonymous novel mutation p.91Lys= in patient 13 (Figure 5B) and a previously published 405T>C in exon 6 representing p.135Tyr= mutation in patient 1 (Figure 5c) are silent mutations. All are mild disease due to couple with deletion mutation.

Figure 5.Chromatogram of mutations represent the transition mutations where panel A shows 22C>T mutation (p.Arg8Cys), Panel B shows 273G>A and Panel C shows 405T>C. In the middle part is mutation compare to the wild type sequence in the upper part. 3D structures of mutation represented in the lower part of the panel. Last two mutations are silent mutations and 3D is similar. Ribbon diagram of a monomer of human PTPS depicting the locations of the amino acid replacements newly identified in current PTPS patients. Substituted amino acid residues denoted with yellow-colored ribbon

The pathogenicity degree of a novel missense mutation 22C>T; p.Arg8Cys was determined by polyphen2 program, which represent a tolerated mutation due to represent of horizontal dark line in the green region (Figure 6). Of course, other two synonymous mutations has no polyphen2 data.

Figure 6.Data of polyphen2 showed a cold pathogenicity of 22C>T mutation as represented by the vertical solid line in green region and cross to the left end, which is a benign mutation.

In current study, two software tools used to predict whether an amino acid substitution has an impact on the pathogenicity and biological function of a protein. PROVEAN and SIFT data for recorded mutations are listed in Table 4.

In addition, sometimes BH4 loading test cannot distinguish patients with BH4 deficiency and BH4-responsive PKU ¹⁴. The conventional Sanger sequencing has been the gold standard test for a molecular approach in the genetic diagnosis of inherited PTPS disorders.

In humans, the PTPS gene (MIM 612719) located on chromosome 11q22.3-11q23.3, contains six exons and encodes for a protein of 145 amino acids ³³. Reduction of the PTPS enzyme activity (MIM 261640) is the most common cause of patients with BH4 deficiency (approximately 59%; ³⁴) causing hyperphenylalaninenia BH4-deficient type A, (HPABH4A).

Sanger sequence of 13 unrelated PKU patients identified 21 variants, 5 of them with a single mutation and 8 with two compound mutations representing a mutation detection rate of 100%. These mutations represented 6 different types of mutations (two of them are silent mutations, one novel c.273G>Al, p.91Lys= exon 5 and the other was previously identified c.405T>C, p.135Tyr= exon 6). Other four mutations represent three novel mutations, c.164_186del in exon 3 p.Val55Aspfs*2, c.22C>T, p.Arg8Cys in exon 1, and c.86A>T, p.Lys29Ile in exon 2, in addition to one previously identified c.200C>T, p.Trp67Met mutation in exon 4. Novel mutations were the most prevalent mutations in this study, which accounted for 76% (4/6) of the total mutant alleles. Particularly, p.Val55Aspfs*2 mutation has been reported as the most frequent mutation in PKU-BH4 responsiveness patients with a frequency of (12/13; 92%).

The most prevalent alleles, c.164_186del variant which accounted for 57% (12/21) of the total mutant alleles, c.200C>T, accounts for 19% (4/21), c.86A>T, accounts in 9.5% (2/21). The deletion mutations led to an in-frame deletion or frameshifting, while two missense mutations were predicted to affect enzymatic function and/or stability. However, to date, studies using exon analysis of different populations worldwide have shown that PTPS mutations were at least 88 different mutations ²⁹, only two mutation in this study, missense c.200C>T, p.(Trp67Met) and silent mutation c.405T>C, p.135Tyr= were sited. Considering the relatively high number of patients with same deletion in our current population, postulate a founder effect of this mutation in the Egyptian population.

Since the incidence of PTPS deficiency is common among BH4 responsiveness patients, the distribution of common mutations were differ from population to population. In a previous study, PTPS deficiency was markedly higher (98%) in the Chinese population than that internationally accounted ²⁹. The recorded results suggested that p.N52S, mutation in East-Asian populations and in Iranian patients ³⁵, where p.V56M in western countries ³⁶ and p.P87S in Chinese population have been reported as the most common mutations. This data explains that deletion mutation is the most common mutation among selected patients and very unique to them. Further studies on other Mediterranean area to confirm from this speculation. Although, novel deletion mutation is likely to be a pathogenic where, deletion mutations have been widely suggested to be disease-causing mutations ³⁷. It should be highlighted that two novel missense variations c.22C>T, c.86A>T were observed in couple with frameshift mutation, there was supportive information that these new variations were likely to be pathogenic and may be disease-causing mutations. First, they also conserved to have same pattern of phe tolerance compared with deletion mutation only (Figure 7). Second, they are absent from the normal population and snpdb in National Center for Biotechnology Information (NCBI), similar to missense mutation c.200C>T previously identified as a pathogenic mutation in Italian patients ³⁸. Third, a significant change in the chemical properties of amino acids in 3D structure (Figure 2, Figure 3, Figure 4, Figure 5). Forth, both mutations are predicted to be deleterious and damaging from prediction software SIFT and PROVEAN. To prove the function of these new variations will be further studied.

On one other hand previously recognized c.200C>T; p.Trp67Met mutation ³⁸ belongs to near region of catalytic domain (D89 and H 90 with C score 0.791 and 0.691 respectively, Figure 5). It was detected as a compound homozygous with del c.164_186; mutation in patients 4, 6, 8, 10 in the current study. Changing the Trp aromatic amino acid at position 67 to Met is expected to weaken the hydrophobic interactions of polypeptide coil region (from position 64 to position 69) among surrounding amino acids and the activity of PTPS may be dramatically decreased due to alterations in the side- chain sulfide bond that contribute to structural conformation. Figure 8.

Figure 8.Active sites of PTPS enzyme produced by I-TASSER software

On the other hand, I-TASSER of two novel missense variants (p.Arg8Cys and p.Lys29Ile) identified in patient 11 and in patients 7&9 respectively as a compound homozygous with del c.164_186; mutation in PTPS protein are located away from the catalytic domain. Both mutations are located in the first and second coil region of PTPS polypeptide chain and predicted to be damaging and deleterious from prediction software, SIFT and PROVEAN (Table 4). Replacement of these basic amino acid Arg at position 8 and Lys at position 28 with neutral Cys and Ile amino acids respectively will change B-factor (is a value to indicate the extent of the inherent thermal mobility of residue in proteins in I-TASSER software). The mutation p.Lys29Ile is located very cross to ligand binding site residues for biopterins at position H24 and L26 with C-score 0.94, which considered more effective on PTPS function compared with p.Arg8Cys mutation in proximal region (Figure 9). This might explain the damaging effect of p.Lys29Ile compared with deterious effect of p.Arg8Cys mutations. Both novel missense mutations require more protein studying to analyze distortion of the backbone interactions among domains and changing in repulsive interactions.

Figure 9.I-TASSER results of PTPS wild type protein with legend and active sites.

In summary, this is the first reported case of Egyptian BH4 deficient patients, which provides a reference for PTPS gene mutation research in the future. This study identified three different novel variations among 13 BH4- deficient patients to improve PTPS mutation database. The results could be of value for differential molecular diagnosis of BH4 deficiencies to avoid any delay in diagnosis and treatment.