Conformational Change of UGT1A1 by a Novel Missense Mutation (p.L131P) Causing Crigler‐Najjar Syndrome Type I

Crigler-Najjar syndrome type I (CN-I; MIM #218800) is the most severe type of hereditary unconjugated hyperbilirubinemia, a disease first recognized by Crigler and Najjar in 1952 (1). This syndrome is caused by mutations of bilirubin UDP-glucuronosyltransferase (UGT1A1; EC 2.4.1,17), which catalyzes glucuronidation of bilirubin from the hydrophobic form to the hydrophilic form (bilirubin-monoglucuronide and bilirubin-diglucuronide) (2). The elimination pathway of bilirubin in humans is catalyzed by UGT1A1 alone; (3) absence of activity of this enzyme gives rise to severe unconjugated hyperbilirubinemia, or CN-I. (4) The serum bilirubin level of patients with CN-I exceeds 30 mg/dL (513 μmol/L), and they are at risk for bilirubin encephalopathy (kernicterus) (1). The mild and moderate types of hyperbilirubinemia, respectively known as Gilbert syndrome (GS; MIM #143500) and Crigler-Najjar syndrome type II (CN-II; #MIM 606785), are also caused by mutations of UGT1A1, but these are generally not harmful (5,6). A number of mutations that cause CN-I, CN-II, and GS have been reported since the cloning of UGT1A1 cDNA in 1991 (5–7). The mutations detected in Crigler-Najjar syndrome type I were nonsense mutations, insertion mutations, and deletion mutations (7,8). Most of those mutations generate a truncated enzyme; only a few missense mutations resulted in null activity (7,9). However, no reports have yet suggested a contribution of the substituted amino acid residue to the enzyme conformation due to the missense mutations. In the present study, we detected a novel missense mutation that gave rise to a change of steric structure of UGT1A1 in an Indian patient with CN-I. The patient was a 2-year-old Indian girl. She was born to nonconsanguineous parents. She developed severe jaundice 3 days after being born. Her peak serum total bilirubin concentration was 386.4 μmol/L (22.6 mg/dL). Most bilirubin was indirect, at 372.7 μmol/L (21.8 mg/dL). She showed no signs of hemolysis or liver dysfunction. Her hyperbilirubinemia was persistent. When a phenobarbital loading test was performed for differential diagnosis between CN-I and CN-II, there was no improvement in her serum bilirubin level. At that point CN-I was suspected. Her serum bilirubin concentration fluctuated between 222.3 (13 mg/dL) and 342 μmol/L (20 mg/dL) without daily treatment for the hyperbilirubinemia. Her age is 2 years, 2 months old, but her developmental age is 7 months. On the basis of on the retardation and the hyperbilirubinemia, we tentatively diagnosed her as CN type I suffered from kernicterus. Serum bilirubin concentrations of the parents were not measured, but they have not apparent jaundice. MATERIALS AND METHODS Gene Analysis of UGT1A1 Genomic DNA was isolated from the leukocytes of the patient and parents with the informed consent of relevant parties. Amplification of exons, specifically the promoter region and enhancer regions of UGT1A1, was carried out by the polymerase chain reaction of genomic DNA, as described elsewhere (10,11). The amplified DNA fragments were sequenced directly (8). Construction of the Expression Vectors and Expression of UGT1A1 in COS-7 Cells The methods used to construct the expression vector using pCR3.1 (Invitrogen, San Diego, CA) and the expression study have been described eleswhere (9,12). To introduce the mutations into the expression vector, we used the following primer (the mutation points are indicated by underlining): 5′-TCCCACTTACCGCACAACAAG-3′ for the T to C transversion at nucleotide 392 (c.392T>C) for p.L131P, by means of the site-directed mutagenesis method using Mutan Km (TaKaRa, Kyoto, Japan). Two models were generated, wild-type UGT1A1 and p.L131P UGT1A1; nontransfected cells were used as controls. Protein content was measured with a BCA Protein Assay Kit (Pierce, Rockford, IL). Assay of UGT1A1 Activity and Western Blot Analysis The methods used in the in vitro expression study and assay for normal and mutated UGT1A1 activity were as reported previously, with a minor modification (12). The resulting bilirubin-glucuronide was isolated by thin-layer chromatography (TLC) on TLC plastic sheet 5748 (Merck, Darmstadt, Germany), and were scanned on an Instant Imager (Packard, Meriden, CT). The amounts of enzyme expressed were checked by Western blotting analysis, using anti-UGT1A1 antibody to adjust enzyme activity. We calculated kinetic parameters by using the Prism 3.0 software (Graph Pad Software, San Diego, CA) to perform nonlinear regression on the Michaelis-Menten equation. Molecular weights of mutated protein were determined by NIH Image. Prediction of Change of 3-Dimensional Structure in UGT1A1 by Homology Modeling Homology models of wild-type and p.L131P UGT1A1 were predicted using a molecular operating environment (MOE; Chemical Computing Group, Montreal, Canada) (13). The query amino acid sequence of human UGT1A1 came from GenPret (Accession number P22309). Tdp-Epi-Vancosaminyltransferase Gtfa (PDB code 1RRV) was selected as a template structure by alignment of amino acid sequences in the MOE protein database. Independent models of the target protein structure were constructed using a Boltzmann-weighted randomized modeling procedure (13). RESULTS Gene Analysis The patient was homozygous for a transition at nucleotide 392 in exon 1. Substitution of thymine for cytosine changed the codon from leucine to proline at position 131 of the corresponding protein (c.392T>C: p.L131P) (Fig. 1). p.L131P is a novel missense mutation. The patient had 3 additional homozygous polymorphic mutations: c.-53TA[8] in the TATA box (conventionally described as A(TA)7TAA), c.-3279T>G in the phenobarbital responsive enhancer module (gtPBREM), and c.-3156G>A. The notation of the mutations is according to the recommendation of the Human Genome Variation Society. We detected no additional mutations in exon 2-5. The role of c.-3156G>A in transcription of UGT1A1 is not known. The patient's genotype was homozygous for c.-3279T>G, A(TA)7TAA, and p.L131P. The mother was heterozygous for c.-3279T>G, A(TA)7TAA, and p.L131P. The father was heterozygous for p.L131P but homozygous for c.-3279T>G and A(TA)7TAA. His genotype was c.-3279T>G; A(TA)7TAA; p.L131P/c.-3279T>G; A(TA)7TAA. The father and the mother are, respectively, heterozygous and homozygous for c.-3156G>A.FIG. 1: UGT1A1 nucleotide sequences amplified from the genomic DNA of the patient. The mutation, a transition of T to C (arrow) at position 392 in UGT1A1 cDNA, changes codon 131 from leucine to proline (p.L131P).Expression and Identification of UGT1A1 In Western blotting analysis, a protein band (Fig. 2) was detected at 52 kDa in wild-type expression models, but not in the mock transfection. The molecular weight of the band was identical to that reported previously in wild-type UGT1A1 (3). However, the p.L131P enzyme shifted to a higher molecular weight of 57 kDa according to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 2).FIG. 2: Western blotting of wild-type and mutant (p.L131P) UGT1A1. An arrow shows the 52-kDa UGT1A1 band. A, Nontransfected control; B, wild-type, C. p.L131P. The band of p.L131P mutant protein is higher than that for wild-type protein (57 kDa).Assay of Enzyme Activity The apparent Vmax and Km of the wild type expression model were 7.8 pmol/min/mg protein and 47 μmol/L. The L131P enzyme had no detectable activity (Fig. 3).FIG. 3: In vitro expression study. The arrow indicates a spot of conjugated bilirubin, produced by normal UGT1A1 activity. The p.L131P enzyme exhibited no detectable activity.Comparison of the Conformation of Wild-type With p.L131P UGT1A1 The 3-dimensional (3-D) structure of human UGT1A1 has not yet been determined by x-ray diffraction analysis. We used MOE to predict model 3-D structures of wild-type and p.L131P UGT1A1. There was a β-sheet around codon 131 of wild-type UGT1A1 (amino acid number 129-134: His-Leu-Leu-His-Asn-Lys), but this β-sheet was absent in p.L131P UGT1A1 (Fig. 4). Instead, 2 α-helices appeared in front of and behind codon 131 in mutant UGT1A1 (amino acid number 102-112, 138-141) (Fig. 4). It follows that the substitution of leucine for proline at codon 131 causes significant change to the tertiary structure of UGT1A1.FIG. 4: Ribbon diagrams of the predicted UGT1A1 structure from wild-type (A) and p.L131P mutant protein (B). The figures were produced by homology-based modeling using MOE. Red ribbons represent α-helices, and yellow arrows indicate the β-sheet. The mutated UGT1A1 loses its β-sheet structure around codon 131 (amino acid number 129–134) (arrows). In front of and behind codon 131, the mutated enzyme has 2 different α-helices (amino acid number 102–112, 138–141) (dotted arrows).DISCUSSION The mutations of UGT1A1 cause the hereditary unconjugated hyperbilirubinemias CN-I, CN-II and GS. These 3 syndromes are distinguished by serum bilirubin levels; concentrations in patients with the most severe CN-I reach 30–50 mg/dL (513–855 μmol/L). Patients with CN-I are at risk of kernicterus (bilirubin encephalopathy), unlike patients with moderate CN-II and GS. Patients with CN-I should be treated with daily phototherapy to prevent the development of kernicterus (2). A phenobarbital loading test is effective for differential diagnosis of CN-I from CN-II and GS (2). The phenobarbital loading test indicated that the present patient was CN-I, but her serum bilirubin concentration was not in the typical type I range. To obtain a definitive diagnosis, we performed gene analysis of UGT1A1, and found that the patient was homozygous for the triple mutation, including a novel missense mutation (p.L131P). Furthermore, our in vitro expression study clarified that p.L131P loses glucuronidation activity of UGT1A1. The reason why the patient had low bilirubin levels compared with other CN-1 is not clear. The detection limit of our UGT assay method is 2.4 pmol/min/mg protein, and this means we cannot detect enzyme activity <2% to 3% of wild-type (12). Although the enzyme has an extensive conformational change and no response to phenobarbital, we cannot completely neglect the possibility that the mutant enzyme has still trace but undetectable amounts of the enzyme activity. p.L131P is a function loss mutation, and it is possible that c.-3279T>G and A(TA)7TAA do not contribute to CN-I in the patient. However, the genotype of the patient indicated tight linkage of the 2 polymorphic mutations of UGT1A1, c.-3279T>G and A(TA)7TAA. We recently encountered a Turkish patient with 3 homozygous mutations of c.-3279T>G, A(TA)7TAA, and p.H39D who developed CN-II (12). We also recently found that European and Asian patients with GS have c.-3279T>G linked A(TA)7TAA, and suggested that a synergistic reduction of the transcriptional activity of A(TA)7TAA and c.-3279T>G may be essential to GS (11). Linkage of c.-3279T>G and A(TA)7TAA in Turkish and Indian patients suggests that the linkage of the 2 polymorphic mutations of GS exist across distinct ethnic groups. Western blotting analysis revealed that the apparent molecular weight of the mutated protein shifted to 57 kDa, from 52 kDa for the wild-type (Fig. 4). This change suggests that the mutation radically changes the protein structure and may alter the structure of its active site, resulting in loss of enzyme activity. Many missense mutations of UGT1A1 have been detected, but there has been no report that mobility of the mutated enzyme is changed, based on Western blotting analysis of a gene expression experiment (4,9,12). The amino acid change (GCSHLPHNKE, in which the underline indicates codon 131) did not create a glycosylation site, implying that the mutation alters the steric conformation of the enzyme. Because the 3-D structure of UGT1A1 has not been determined, we used MOE to analyze the conformational change of UGT1A1 having p.L131P. The structure around codon 131 of the wild-type was predicted to be a β-sheet structure (Fig. 4A). Replacement of leucine by proline at codon 131 removed the β-sheet (Fig. 4B). In addition, 2 α-helices were generated before and behind codon 131. The coding region of exon 1 includes the substrate binding domain (14), and these large changes of protein structure may prevent substrate binding. We believe this to be the first report of conformational change of the UGT1A1 structure by a missense mutation. Acknowledgments We thank M. Suzaki and R. Okamoto of the Central Research Laboratory, Shiga University of Medical Science, for technical assistance.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.