- Email: [email protected]
Correlation of mutational analysis to clinical features in Taiwanese patients with Gilbert’s syndrome
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 2001 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 96, No. 4, 2001 ISSN 0002-9270/01/$20.00 PII S0002-9270(01)02262-6
Correlation of Mutational Analysis to Clinical Features in Taiwanese Patients With Gilbert’s Syndrome S.-Y. Hsieh, M.D., Ph.D., Y.-H. Wu, B.A., D.-Y. Lin, M.D., C.-M. Chu, M.D., M. Wu, and Y.-F. Liaw, M.D. Liver Research Unit, Chang Gung Memorial Hospital, and School of Medicine Chang Gung University, Taoyuan, Taiwan
OBJECTIVES: Mutations in the promoter as well as in the coding region of the bilirubin UDP-glucuronosyltransferase gene (UGT1A1) have been found to be associated with Gilbert’s syndrome. However, the genetic basis of Gilbert’s syndrome in our population and correlation of these mutations to fasting serum bilirubin levels in patients with Gilbert’s syndrome remain to be clarified. METHODS: We applied polymerase chain reaction– based direct-sequencing assays to examine mutations in UGT1A1 gene in 20 unrelated Gilbert’s patients and in a family with Gilbert’s syndrome. RESULTS: We studied three mutations that were previously reported to be associated with Gilbert’s syndrome (i.e., the TATAA-box mutation, Gly71Arg, and Pro229Gln) in 20 patients with Gilbert’s syndrome. Of the patients, 16, five, and six were found to have the TATAA-box, Gly71Arg and Pro229Gln mutations, respectively. Seven patients had simultaneous mutations both in the TATAA box and in the coding region. Of note, all six patients with Pro229Gln also had the TATAA-box mutation. Localization of Pro229Gln on the allele containing the TATAA-box mutation was demonstrated in a family with Gilbert’s syndrome. The patients simultaneously heterozygous for both the TATAAbox mutation and Gly71Arg usually had serum bilirubin levels similar to those found in the patients homozygous for the TATAA-box mutation, but usually higher than those found in the patients heterozygous for the TATAA-box mutation alone. On the other hand, concurrence of Pro229Gln in patients with TATAA-box mutation or with Gly71Arg did not significantly affect serum bilirubin levels. CONCLUSIONS: The TATAA-box mutation and Gly71Arg are the major causes for Gilbert’s syndrome in our population. Concurrence of mutations of Gly71Arg and TATAAbox usually exerts a synergistic effect on hyperbilirubinemia. Pro229Gln, which is regularly linked to the TATAA-box mutation, may not have a significant effect on serum bilirubin levels. (Am J Gastroenterol 2001;96: 1188 –1193. © 2001 by Am. Coll. of Gastroenterology)
INTRODUCTION Conjugation with glucuronic acid is essential for efficient excretion of bilirubin, as well as of numerous endogenous and exogenous substances. Bilirubin glucuronidation is mediated by a microsomal enzyme, the bilirubin UDP-glucuronosyltransferase (UGT1). Recently, cDNA clones for rat and human bilirubin UGT1 were isolated. The human bilirubin UGT1 gene is composed of 11 tandem arrays of different first exons and common second to fifth exons (1–3). Several kinds of mature mRNA are produced by a differential splicing mechanism. The substrate specificity of a particular UGT1 is determined by the amino-terminal half of the polypeptide encoded by each first exon (3). It is now known that only the exon A1 (UGT1A1) product is relevant to bilirubin glucuronization (4, 5). Gilbert’s syndrome is an inherited form of mild, chronic, unconjugated hyperbilirubinemia without liver disease or overt hemolysis. It is the most common syndrome known in humans, occurring in 5–10% of the popuation (6 –9). It has been found that hepatic glucuronidation activity is reduced in patients with Gilbert’s syndrome (10 –13). Crigler-Najjar syndrome type I, a severe unconjugated hyperbilirubinemia often causing death in infancy, and Crigler-Najjar syndrome type II, another inherited unconjugated hyperbilirubinemia, are also caused by a reduction of bilirubin UDP-glucuroosyltransferase (UGT) activity (10 –14). Genetic abnormalities in the coding region of the UGT1A1 gene, usually homozygous for a nonsense mutation and eliciting an absence of UGT activity, result in Crigler-Najjar syndrome type I (1, 15–21). Homozygosity or heterozygosity for a missense mutation of the UGT1A1 gene causing a 90% reduction in UGT enzymatic activity gives rise to CriglerNajjar syndrome type II (22–24). It has not been known whether the mutations causing Gilbert’s mutations were related to those causing Crigler-Najjar syndrome. Recently, a 2-base, TA insertion in the TATAA box [A(TA)7TAA] of the promoter for the UGT1A1 gene was found in all the tested Caucasian patients with Gilbert’s syndrome (25, 26). The presence of the longer TATAA element [A(TA)7TAA vs A(TA)6TAA for the mutant and wild-type, respectively]
AJG – April, 2001
Mutational Analysis for Gilbert’s Syndrome
1189
results in a 50 –70% reduction of hepatic glucuronidation activity in people with Gilbert’s syndrome (25, 27). In Asian people, the genetic basis for Gilbert’s syndrome is, however, relatively complicated (28 –30). Except for the A(TA)7TAA mutation, both heterozygous and homozygous missense mutations in the UGT1A1 gene have been identified in Japanese patients with Gilbert’s syndrome (31–34). In addition, the relationship between the genotype of the UGT1A1 gene and the phenotype of Gilbert’s syndrome remains to be clarified. Herein we report our work on genetic analysis of mutations of the UGT1A1 gene, and correlation of the genotypes of UGT1A1 mutations with serum bilirubin levels in our Gilbert’s sydrome patients.
MATERIALS AND METHODS Subjects All subjects included in this study were examined by means of serum liver biochemistry tests and hemograms. Serum samples used for the evaluation of bilirubin level were collected from patients after a fast of about 10 h. Those participants who had disease of the hepatobiliary system, or hematological diseases such as hemolytic anemia or ineffective hemopoiesis, were excluded from this study. The recruitment protocol was approved by an institutional review committee at Chang Gung Memorial Hospital. To study the genetic basis for Gilbert’s syndrome, 20 unrelated patients with Gilbert’s syndrome were included. The clinical diagnosis of Gilbert’s syndrome was based on the following criteria: 1) a consistent, mild elevation of serum bilirubin (1.5–5.3 mg/dl) with a direct-reacting fraction that was either normal or ⱕ15% of the total according to the Van den Bergh test; 2) normal serum ALAT and aspartate aminotransferase values; and 3) exclusion of the potential for hemolysis and ineffective hemopoiesis on the basis of demonstrated normal Hb, a normal mean corpuscular Hb concentration, and normal reticulocyte counts (13, 25). A total of 25 unselected healthy young adults working in our department and laboratory were enrolled for assessing the normal allele frequency of the A(TA)7TAA mutation and of other mutations in the coding region of UGT1A1 gene in our population. A patient with Gilbert’s syndrome and his parents were also included for study. Determination of UGT1A1 Promoter Genotypes For rapid and accurate determination of numbers of the (TA) repeat in the TATAA box of the UGT1A1 gene, we developed a direct sequencing assay based on polymerase chain reaction (PCR). Genomic DNA was extracted from peripheral blood leukocytes by using a standard method (35). Approximately 50 ng of genomic DNA was used in PCR for amplification of the UGT1A1 promoter region containing the TATAA element. Thirty cycles of PCR were conducted with denaturation for 40 s at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C. The PCR products
Figure 1. Polymerase chain reaction– based direct sequencing assay to the phenotype of TATAA box polymorphism of the UGT1A1 promoter. The region containing the TATAA box in the promoter for the UGT1A1 gene was amplified by PCR. The amplified DNA was then subjected to direct sequencing with an automatic sequencing kit (ABI PRISM 337 DNA sequencer; PRISM 337 collection, Sequence Analysis 3.0; Perkin-Elmer). DNA was extracted from (A) a subject with homozygosity for the wild-type A(TA)6TAA, (B) a subject with heterozygosity for A(TA)6TAA/A(TA)7TAA, and (C) a subject with homozygosity for the A(TA)7TAA mutation. Note that the sequences after six copies of the (TA) repeats are mixed by two populations of sequencing data for those heterozygous for A(TA)6TAA/ A(TA)7TAA genotype. The sequences demonstrated are in antisense polarity.
were analyzed by gel electrophoresis. The amplified DNA was purified by using a spin-column (PCR Clean-Up System Kit, Viogene, Sunnyvale, CA) and then subjected to direct PCR sequencing analysis, as previously described (36). Figure 1 demonstrates examples of the sequencing results derived from the following: 1) a subject homozygous for the wild-type of the TATAA box, A(TA)6TAA (Fig. 1A); 2) a subject heterozygous for the A(TA)6TAA /A(TA)7TAA (Fig. 1B); and 3) a Gilbert’s patient homozygous for the A(TA)7TAA mutation (Fig. 1C). The numbers of the (TA) repeat in the TATAA box were readily determined by sense orientation (data not shown), as well as by antisense orientation (Fig. 1). Sequencing for those heterozygous for the A(TA)6TAA /A(TA)7TAA reveals six tandem copies of the (TA) repeat followed by a mixed population of sequencing results, as shown in Figure 1B.
1190
Hsieh et al.
Some of the amplified DNA was cloned and then sequenced, as described before (37). The sequences of the PCR primers for amplification of the promoter region were: 5⬘-ATTAACTTGGTGTCGATTGG-3⬘ and 5⬘-AGCCATGGCGGCCTTTGCTC-3⬘. Examination of Mutations in Coding Regions of the UGT1A1 Gene Five sets of primers were used with PCR to amplify the regions from exon 1 to exon 5, respectively. PCR was conducted with denaturation for 40 s at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C for 30 cycles. The amplified DNA was then subjected to direct PCR sequencing analysis as described above. The sequences of the primers for amplification of exon 1–5 were as follows: 5⬘-GCTGGGAAGATACTGTTGATCCCAG-3⬘ & 5⬘TGGGCATGATGGGCCTAGGGTAATC-3⬘ (exon 1) 5⬘-CATAACTTACTGTATGTAGTCATCA-3⬘ & 5⬘GATATGAGGCCATGGTATAGAATCT-3⬘ (exon 2) 5⬘-TATGTCTTTCTTTACGTTCTGCTCT-3⬘& 5⬘-TCTTAATTTGACCCTGGTTTGACCT-3⬘ (exon3) 5⬘-GCAAGGGCATGTGAGTAACACT-3⬘ & 5⬘-ACCTTTTGTCATTGATGACTGCT-3⬘ (exon 4) 5⬘-TGGACCTGGCCGTGTTCTGGGTG-3⬘ & 5⬘CCACCCACTTCTCAATGGGTCTTG-3⬘ (exon 5) Statistical Analysis Mean serum bilirubin levels were compared by using Student-Newman-Keuls test (38).
RESULTS To elucidate the genetic basis of Gilbert’s syndrome in our population, 20 unrelated Gilbert’s patients were studied. Their clinical data and the genotypes for mutations in the UGT1A1 gene are shown in Table 1. Mutations in the UGT1A1 gene were identified in all cases except one. Sixteen patients were found to have the A(TA)7TAA mutation including seven homozygous and nine heterozygous. Five had Gly71Arg (replacement of glycine with arginine at the position of amino acid 71) and six had Pro229Gln. No other missense mutations including Arg209Trp, Arg367Gly, and Tyr486Asp which had been reported for Japanese Gilbert’s patients were identified. Simultaneous mutations both in the TATAA box and in the coding region were found in seven patients. Of note, all six patients with Pro229Gln also had the A(TA)7TAA mutation. A total of 25 unselected healthy adults were enrolled as controls for assessing the allele frequencies for A(TA)7TAA, Gly71Arg and Pro229Gln in our population. As shown in Table 2, 20 subjects were homozygous for the wild-type, A(TA)6TAA, and five were heterozygous for A(TA)6TAA/A(TA)7TAA. Therefore, the allele frequency for the A(TA)7TAA mutation in our controls was 10%. Accordingly, the allele frequency and the frequency of ho-
AJG – Vol. 96, No. 4, 2001
Table 1. Clinical Presentation and Genotypes of the 20 Gilbert’s Patients Case Sex 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
M M M M M M M M M M M M M M M M M F F F
Age (yr) 46 45 33 16 40 44 44 17 35 18 23 36 41 28 44 43 27 51 41 26
Genotype Serum Bilirubin (mg/dl) A(TA)7TAA G71R P229Q YY YY YY YY YY YY YY YX YX YX YX YX YX YX YX YX XX XX XX XX
4.1 3.4 2.9 2.7 2.3 2.1 1.9 3.4 3.3 2.1 2.0 1.8 1.8 1.7 1.6 1.6 2.0 2.2 1.6 3.4
XX XX XX XX XX XX XX XY XY XX XX XX XX XX XX XX YY XY XY XX
XX YX XX YX XX XX YX XX YX XX YX XX XX XX YX XX XX XX XX XX
XX ⫽ wild type; YX ⫽ heterozygosity; YY ⫽ homozygosity for the mutation.
molozygote for the A(TA)7TAA mutation in our general population were estimated to be about 11% and 1.2%, respectively. On the other hand, none of them had Gly71Arg, whereas one had heterozygosity for Pro229Gln. The allele frequencies in our controls were 0% for Gly71Arg and 4% for Pro229Gln, respectively (Table 2). By comparison, the allele frequencies of the A(TA)7TAA mutation, Gly71Arg and Pro229Gln in our Gilbert’s patients were 58%, 15%, and 15%, respectively (p ⬍ 0.001, p ⫽ 0.016, and p ⫽ 0.147 respectively, as compared to the controls) (Table 2). To evaluate further the implication of each mutation in clinical manifestations of hyperbilirubinemia, genotypes for each mutation were correlated to serum bilirubin levels. The results are shown in Table 3. The mean value for serum bilirubin levels in the seven patients homozygous for the A(TA)7TAA mutation was higher than that in the nine patients heterozygous for the A(TA)7TAA mutation (2.8 mg/dl vs 2.1 mg/dl, p ⬍ 0.05 conducted by Student-Newman-Keuls test). Of interest, the two patients with simultaneous heterozygous mutations both for Gly71Arg and for Table 2. Allele Frequency of Mutations in UGT1A1 Gene Allele Frequency (%)
A(TA)7TAA Gly71Arg Pro229Gln
Gilbert’s Patients (n ⫽ 20)
Healthy Controls (n ⫽ 25)
p*
58 15 15
10 0 2
⬍0.001 ⬍0.001 0.058
* Statistical analysis was conducted via 2 test with Yates’ correction.
AJG – April, 2001
Mutational Analysis for Gilbert’s Syndrome
Table 3. Correlation of Genotypes to Serum Bilirubin Levels Genotype
Serum Bilirubin
A(TA)7TAA
G71R
P229Q
Case
Mean (mg/dl)
SEM
YY YY YX YX YX YX XX XX
XX XX XX XX XY XY YY XY
XX YX XX YX XX YX XX XX
4 3 5 2 1 1 1 2
2.85 2.68 1.88 1.88 3.4 3.3 2.0 1.90
0.45 0.43 0.07 0.20
0.30
1191
patient, who exhibited a serum bilirubin level of 2.7 mg/dl, was simultaneously homozygous for the A(TA)7TAA mutation and heterozygous for Pro229Gln. His father had a serum bilirubin level of 2.1 mg/dl and was homozygous for the A(TA)7TAA mutation, and his mother had a serum bilirubin level of 1.4 g/dl as well as simultaneous heterozygous mutations both in the TATAA box and in Pro229Gln. Obviously, Pro229Gln was located on the maternal allele, which also contained the A(TA)7TAA mutation and was transmitted to the patient.
XX ⫽ wild type; YX ⫽ heterozygosity; YY ⫽ homozygosity for the mutation.
DISCUSSION the A(TA)7TAA mutations had serum bilirubin levels (3.4 mg/dl and 3.3 mg/dl) similar to those found in the patients homozygous for the A(TA)7TAA mutation (mean ⫽ 2.7 mg/dl and 2.9 mg/dl for those with and without an additional Por229Gln, respectively), but higher than those in the patients heterozygous for the A(TA)7TAA mutation (mean ⫽ 1.8 mg/dl and 1.8 mg/dl for those with and without additional Pro229Gln, respectively). These findings are indicative of a synergistic effect on hyperbilirubinemia exerted by concurrence of Gly71Arg with the A(TA)7TAA mutation. However, serum bilirubin levels in the patients homozygous for the A(TA)7TAA mutation with an additional Pro227Arg was not significantly different from that in the patients homozygous for the A(TA)7TAA mutation without an additional Pro227Arg (2.7 mg/dl vs 2.9 mg/dl, respectively). Likewise, serum bilirubin levels in the patients heterozygous for the A(TA)7TAA mutation with and without additional Pro227Arg had serum bilirubin levels similar to each other (1.8 mg/dl vs 1.8 mg/dl, respectively). The inheritance model of Gilbert’s syndrome was demonstrated in one patient and his parents, as shown in Figure 2. The
Figure 2. Familial study of TATAA-box polymorphism of the UGT1A1 gene. The method to determine the number of the (TA) repeat of the UGT1A1 gene is as described in the legend for Figure 1. The bold-framed square represents the proband patients. Numbers in parentheses indicate serum bilirubin levels.
We report here 20 Taiwanese patients with Gilbert’s syndrome. All patients except one were found to have mutations in the UGT1A1 gene. A total of 16 patients had the A(TA)7TAA mutation (80%) and 10 patients had missense mutations in the coding region (50%), including seven patients had simultaneous mutations both in the TATAA box and in the coding region (35%). With respect to the TATAA-box polymorphism, the A(TA)7TAA mutation, but neither A(TA)5TAA nor A(TA)8TAA, which had been reported in Caucasian people (27), was identified in our population. The A(TA)7TAA mutation was found in 16 patients including seven who were homozygous and nine heterozygous. The allele frequency of the A(TA)7TAA mutation in our Gilbert’s patients was about 58%, which was significantly higher than 10% in our healthy controls. With regard to missense mutations in the UGT1A1 gene, one was homozygous, whereas four were heterozygous for Gly71Arg and six were heterozygous for Pro229Gln. The allele frequencies of Gly71Arg and Pro229Gln in our Gilbert’s patients were 15% and 15%, respectively, in comparison to 0% and 2%, respectively, in our controls. Of note, there was one Gilbert’s patient with no identified mutation either in the TATAA box or in the coding region of the UGT1A1 gene. Further studies, especially on the promoter region other than the TATAA box, to elucidate the genetic cause for this patient are warranted. Nevertheless, the spectrum of mutations in the UGT1A1 gene associated with Gilbert’s syndrome in Taiwanese patients was relatively diverse from that found in Japanese and Caucasian patients (30, 39, 40). In a report from Sato et al. (34), 19 Japanese Gilbert’s patients were studied. Of these patients, eight had A(TA)7TAA mutations and 11 had missense mutations (58%), including three with simultaneous mutations both in the TATAA box and in the coding regions; in four patients the TATAA box was not assayed (34). Taking account of the fact that TATAA box mutations were not assayed in four of the 19 Gilbert’s patients in that study (34), the allele frequencies of the A(TA)7TAA mutation and of Gly71Arg in Sato’s patients with Gilbert’s syndrome were 47% and 21%, respectively, which were close to the frequencies of 58% and 15% in our Gilbert’s patients. However, in one other Japanese study, the allele frequency for A(TA)7TAA in Japanese Gilbert’s patients was only 14% (29). In another
1192
Hsieh et al.
study, the allele frequency of Gly71Arg in jaundiced neonates was as high as 32% (41). The reasons for these discrepancies remain unclear. By contrast, in Caucasian patients, the TATAA-box mutation represents the only cause for all examined Gilbert’s patients, and Gly71Arg has never been found either in Gilbert’s patients or in controls (25, 26, 30, 41). Serum bilirubin levels were noted to be nearly alike between in the patients heterozygous for the A(TA)7TAA mutation and in the patients heterozygous for Gly71Arg (1.8 mg/dl vs 1.9 mg/dl, Table 3). In addition, serum bilirubin levels in the patients simultaneous heterozygous mutations both for the A(TA)7TAA mutation and for Gly71Arg seemed to be higher than those in patients who were heterozygous for the A(TA)7TAA mutation (3.4 mg/dl vs 1.8 mg/dl), as well as higher than those in patients heterozygous for Gly71Arg (3.4 mg/dl vs 1.9 mg/dl). It seems that Gly71Arg and the A(TA)7TAA mutation had a comparable effect on hyperbilirubinemia but that concurrence of the two mutations had a synergistic effect. Such a synergistic effect was not observed in patients who were heterozygous for Pro229Gln with either the A(TA)7TAA mutation or the Gly71Arg mutation. Of interest, all six patients heterozygous for Pro229Gln also had A(TA)7TAA. That Pro229Gln was regularly linked to the A(TA)7TAA mutation was further demonstrated by the inheritance model in a family with Gilbert’s syndrome, wherein the maternal allele transmitted to the patient contained both the A(TA)7TAA mutation and Pro229Gln (Fig. 2). On the other hand, less than one-fourth of the alleles containing A(TA)7TAA had an additional Pro229Gln mutation (S.Y. Hsieh, unpublished data). The finding that Pro229Gln is generally associated with A(TA)7TAA is important because Pro229Gln has been reported to be one of the genetic defects causing Gilbert’s syndrome and CriglerNajjar type II in three Japanese studies (32, 34, 42). However, the TATAA box was not analyzed in the Gilbert’s patients with Pro229Gln reported by Sato et al. or by Koiwai et al. (32, 34). As well, the Crigler-Najjar type II patient heterozygous for Pro229Gln reported by Yamamoto et al. was actually found to be simultaneously homozygous for the A(TA)7TAA mutation (42). We therefore hypothesize that, among Japanese as well as Taiwanese patients, the Gilbert’s syndrome and Crigler-Najjar type II patients with Pro229Gln are very likely to have a concurrent A(TA)7TAA mutation, and that hyperbilirubinemia in these patients is more likely caused by the A(TA)7TAA mutation rather than by Pro229Gln.
ACKNOWLEDGMENTS The authors are grateful to Wei-Chu Shyu for his excellent technique assistance and to Su-Chen Ji for her excellent expertise in graph drawing. This work was supported by National Science Counsel ((NSC 89-2320-B-182-082) and
AJG – Vol. 96, No. 4, 2001
in part a grant from Chang Gung Memorial Hospital (CMRP 800). Reprint requests and correspondence: Sen-Yung Hsieh, M.D., Ph.D., Liver Research Unit, Chang Gung Memorial Hospital, 199 Tung-Hwa North Road, Taipei 105, Taiwan. Received July 7, 2000; accepted Nov. 3, 2000
REFERENCES 1. Bosma PJ, Roy-Chowdhury N, Goldhoon BG, et al. Sequence of exon and the flanking regions of human bilirubin-UDPglucuronosyltransferase gene complex and identification of a genetic mutation in a person with Crigler-Najjar syndrome, type I. Hepatology 1992;15:941–7. 2. Ritter JK, Chen F, Sheen YY, et al. A novel complex locus UGT1 encodes human bilirubin, phenol and other UDP-glucuronosyltransferase isoenzymes with identical carboxyl termini. J Biol Chem 1992;267:3257– 61. 3. Clarke DJ, Moghrabi N, Monaghan G, et al. Genetic defects of the UDP-glucuronosyltransferase-1 (UGT1) gene that cause familial non-hemolytic unconjugated hyperbilirubinemias. Clin Chim Act 1997;266:63–74. 4. Ritter JK, Crawford JM, Owens IS. Cloning of two human liver bilirubin UDP-glucuronosyltransferase cDNAs with expression in Cos-1 cells. J Biol Chem 1991;266:1043–7. 5. Bosma PJ, Seppen J, Goldhoon B, et al. Bilirubin UDPglucuronosyltransferase 1 is the only relevant bilirubin glucuronosyltransferase isoform in man. J Biol Chem 1994;269: 17960 – 4. 6. Owens D, Ritter JK. The novel bilirubin/phenol UDP-glucuronosyltransferase UGT1 gene locus: Implication for multiple nonhemolytic familial hyperbilirubinemia phenotypes. Pharmacogenetics 1992;2:93–108. 7. Bailey A, Robinson D, Dawson AM. Does Gilbert’s syndrome exist? Lancet 1977;i:931–3. 8. Fevery J. Pathogenesis of Gilbert’s syndrome. Eur J Clin Invest 1981;11:417– 8. 9. De Morais SMF, Uetrecht JP, Wells PG. Decreased glucuronidation and increased bioactivation of acetaminophen in Gilbert’s syndrome. Gastroenterology 1992;102:577– 86. 10. Arias IM, London IM. Bilirubin glucuronide formation in vitro: Demonstration of a defect in Gilbert’s syndrome. Science 1957;126:563– 4. 11. Powell LW, Hemingway E, Billing BH, et al. Idiopathic unconjugated hyperbilirubinemia (Gilbert’s syndrome): A study of 42 families. N Engl J Med 1967;277:1108 –12. 12. Arias IM, Gartner LM, Cohen M, et al. Chronic nonhemolytic unconjugated hyperbilirubinemia with glucuronyltrasferase deficiency: Clinical, biochemical, pharmacologic and genetic evidence for heterogeneity. Am J Med 1969;47:395– 409. 13. Black M, Billing BH. Hepatic bilirubin UDP-glucuronyl transferase activity in liver disease and Gilbert’s syndrome. N Engl J Med 1969;280:1266 –71. 14. Arias IM. New genetics of inheritable jaundice and cholestatic liver disease. Lancet 1998;325:82–3. 15. Bosma PJ, Chowdhury JR, Huang TJ, et al. Mechanisms of inherited deficiencies of multiple UDP-glucuronosyltransferase isoforms in two patients with Crigler-Najjar syndrome, type I. FASEB J 1992;6:2859 – 63. 16. Ritter JK, Yeatman MT, Ferreira P, et al. Identification of a genetic alteration in the code for bilirubin UDP-glucuronosltrasferase in the UGT1 gene complex of a Crigler-Najjar type I patient. J Clin Invest 1992;90:150 –5. 17. Moghrabi N, Clarke DJ, Burchell B, et al. Cosegregation of
AJG – April, 2001
18.
19.
20. 21.
22.
23.
24.
25.
26. 27.
28.
29.
intragenic markers with a novel mutation that causes CriglerNajjar syndrome type I: Implication in carrier detection and prenatal diagnosis. Am J Hum Genet 1993;53:722–9. Ritter JK, Yeatman MT, Kaiser C, et al. A phenylanaline codon deletion at the UGT1 gene complex locus of a CriglerNajjar syndrome type I patient generates a pH-sensitive bilirubin UDP-glucuronosyltransferase. J Biol Chem 1993;268: 23573–9. Chowdhury JR, Chowdhury NR, Wolkoff AW, et al. Heme and bile pigment metabolism. In: Arias JM, Boyer JL, Fausto N, eds. The liver: Biology and pathobiology, 3rd ed. New York: Raven Press, 1994:471–504. Labrune P, Myara A, Hadchouel M, et al. Genetic heterogeneity of Crigler-Najjar syndrome type I: A study of 14 cases. Hum Genet 1994;94:693–7. Eros LT, Ritter JK, Hersh JH, et al. Identification of the two single base substitutions in the UGT1 gene locus which abolish bilirubin uridine diphosphate glucuronosyltransferase activity in vitro. J Clin Invest 1994;93:564 –70. Bosma PJ, Goldhoorn B, Oude Elferink RP, et al. A mutation in bilirubin uridine 5⬘-diphosphate-glucurononsyltransferse isoform I causing Crigler-Najjar syndrome type II. Gastroenterology 1993;105:216 –20. Moghrabi N, Clarke DJ, Boxer M, et al. Identification of an A-to-G missense mutation in exon 2 of the UGT1 gene complex that causes Crigler-Najjar syndrome type 2. Genomics 1993;18:171–3. Seppen J, Bosma PJ, Goldhoorn BG, et al. Discrimination between Crigler-Najjar syndrome type I, and II by expression of mutant bilirubin uridine diphosphate-glucuronosyltransferase. J Clin Invest 1994;94:2385–91. Bosma PJ, Roy-Chowdhury J, Bakker C, et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyltransferase 1 in Gilbert’s syndrome. N Engl J Med 1995; 333:1171–5. Monaghan G, Ryan M, Seddon R, et al. Genetic variation in bilirubin UDP-glucuronosyltransferase gene promoter and Gilbert’s syndrome. Lancet 1996;347:578 – 81. Beutler E, Gelbart T, Demina A. Racial variability in the UDP-glucuronosyltransferase 1 (UGT1A1) promoter: A balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci USA 1998;95:8170 – 4. Maruo Y, Sato H, Yamano T, et al. Gilbert’s syndrome caused by a homozygous missense mutation (Tyr486Asp) of bilirubin UDP-glucuronosyltransferase gene. J Pediat 1998;132: 1045–7. Ando Y, Chida M, Kazuo N, et al. The UGT1A1*28 allele is
Mutational Analysis for Gilbert’s Syndrome
30. 31.
32.
33.
34. 35. 36.
37.
38. 39.
40.
41.
42.
1193
relatively rare in a Japanese population. Pharmacogenetics 1998;8:357– 60. Burchell B, Hume R. Molecular genetic basis of Gilbert’s syndrome. J Gastroenterol Hepatol 1999;14:960 – 6. Aono S, Adachi Y, Uyama E, et al. Analysis of genes for bilirubin UDP-glucuronosyltransferase in Gilbert’s syndrome. Lancet 1995;345:958 –9. Koiwai O, Nishizawa M, Hasada K, et al. Gilbert’s syndrome is caused by a heterozygous missense mutation in the gene for bilirubin UDP-glucuronosyltransferase. Hum Mol Genet 1995;4:1183– 6. Soeda Y, Yamamoto K, Adachi Y, et al. Predicted homozygous missense mutation in Gilbert’s syndrome. Lancet 1995; 346:1494. Sato H, Adachi Y, Koiwai O. The genetic basis of Gilbert’s syndrome. Lancet 1996;347:557– 8. Pinthanond D, Petchclain B. A preserved leukocyte preparation for quality control. Am J Clin Pathol 1974;63:32– 4. Hsieh SY, Wu YH, Tsao KC, et al. High prevalence of TT virus infection in healthy children and adults and in patients with liver disease in Taiwan. J Clin Microbiol 1999;37:1829 – 31. Hsieh SY, Yang PY, Chen HC, et al. Cloning and characterization of the extreme 5⬘-terminal sequences of the RNA genome of GB virus C/hepatitis G virus. Proc Natl Acad Sci USA 1997;94:3206 –10. Glantz SA. Primer of biostatistics. Singapore: McGraw-Hill, 1989:64 –100. Sampietro M, Lupica L, Pettero L, et al. TATA-box promoter mutant in the promoter of UDP-glucuronosyltransferase 1 gene in Italian patients with Gilbert’s syndrome. Ital J Gastroenterol Hepatol 1998;30:194 – 8. Lampe JW, Brigler J, Horner NK, et al. UDP-glucuronosyltransferase (UGT1A1*28, and UGT1A6*2) polymorphism in Caucasians, and Asians: Relationship to serum bilirubin concentrations. Pharmacogenetics 1999;9:341–9. Akaba K, Kimura T, Sasaki A, et al. Neonatal hyperbilirubinemia and mutation of the bilirubin uridine diphosphateglucuronosyltransferase gene: A common missense mutation among Japanese, Korean and Chinese. Biochem Mol Biol Int 1998;46:21– 6. Yamamoto K, Sato H, Fujiyama Y, et al. Contribution of two missense mutations (G71R and Y486D) of the bilirubin UDPglucuronosyltransferase (UGT1A1) gene to phenotypes of Gilbert’s syndrome and Crigler Najjar syndrome type II. Biochem Biophys Acta 1998;1406:267–73.
Vol. 96, No. 4, 2001 ISSN 0002-9270/01/$20.00 PII S0002-9270(01)02262-6
Correlation of Mutational Analysis to Clinical Features in Taiwanese Patients With Gilbert’s Syndrome S.-Y. Hsieh, M.D., Ph.D., Y.-H. Wu, B.A., D.-Y. Lin, M.D., C.-M. Chu, M.D., M. Wu, and Y.-F. Liaw, M.D. Liver Research Unit, Chang Gung Memorial Hospital, and School of Medicine Chang Gung University, Taoyuan, Taiwan
OBJECTIVES: Mutations in the promoter as well as in the coding region of the bilirubin UDP-glucuronosyltransferase gene (UGT1A1) have been found to be associated with Gilbert’s syndrome. However, the genetic basis of Gilbert’s syndrome in our population and correlation of these mutations to fasting serum bilirubin levels in patients with Gilbert’s syndrome remain to be clarified. METHODS: We applied polymerase chain reaction– based direct-sequencing assays to examine mutations in UGT1A1 gene in 20 unrelated Gilbert’s patients and in a family with Gilbert’s syndrome. RESULTS: We studied three mutations that were previously reported to be associated with Gilbert’s syndrome (i.e., the TATAA-box mutation, Gly71Arg, and Pro229Gln) in 20 patients with Gilbert’s syndrome. Of the patients, 16, five, and six were found to have the TATAA-box, Gly71Arg and Pro229Gln mutations, respectively. Seven patients had simultaneous mutations both in the TATAA box and in the coding region. Of note, all six patients with Pro229Gln also had the TATAA-box mutation. Localization of Pro229Gln on the allele containing the TATAA-box mutation was demonstrated in a family with Gilbert’s syndrome. The patients simultaneously heterozygous for both the TATAAbox mutation and Gly71Arg usually had serum bilirubin levels similar to those found in the patients homozygous for the TATAA-box mutation, but usually higher than those found in the patients heterozygous for the TATAA-box mutation alone. On the other hand, concurrence of Pro229Gln in patients with TATAA-box mutation or with Gly71Arg did not significantly affect serum bilirubin levels. CONCLUSIONS: The TATAA-box mutation and Gly71Arg are the major causes for Gilbert’s syndrome in our population. Concurrence of mutations of Gly71Arg and TATAAbox usually exerts a synergistic effect on hyperbilirubinemia. Pro229Gln, which is regularly linked to the TATAA-box mutation, may not have a significant effect on serum bilirubin levels. (Am J Gastroenterol 2001;96: 1188 –1193. © 2001 by Am. Coll. of Gastroenterology)
INTRODUCTION Conjugation with glucuronic acid is essential for efficient excretion of bilirubin, as well as of numerous endogenous and exogenous substances. Bilirubin glucuronidation is mediated by a microsomal enzyme, the bilirubin UDP-glucuronosyltransferase (UGT1). Recently, cDNA clones for rat and human bilirubin UGT1 were isolated. The human bilirubin UGT1 gene is composed of 11 tandem arrays of different first exons and common second to fifth exons (1–3). Several kinds of mature mRNA are produced by a differential splicing mechanism. The substrate specificity of a particular UGT1 is determined by the amino-terminal half of the polypeptide encoded by each first exon (3). It is now known that only the exon A1 (UGT1A1) product is relevant to bilirubin glucuronization (4, 5). Gilbert’s syndrome is an inherited form of mild, chronic, unconjugated hyperbilirubinemia without liver disease or overt hemolysis. It is the most common syndrome known in humans, occurring in 5–10% of the popuation (6 –9). It has been found that hepatic glucuronidation activity is reduced in patients with Gilbert’s syndrome (10 –13). Crigler-Najjar syndrome type I, a severe unconjugated hyperbilirubinemia often causing death in infancy, and Crigler-Najjar syndrome type II, another inherited unconjugated hyperbilirubinemia, are also caused by a reduction of bilirubin UDP-glucuroosyltransferase (UGT) activity (10 –14). Genetic abnormalities in the coding region of the UGT1A1 gene, usually homozygous for a nonsense mutation and eliciting an absence of UGT activity, result in Crigler-Najjar syndrome type I (1, 15–21). Homozygosity or heterozygosity for a missense mutation of the UGT1A1 gene causing a 90% reduction in UGT enzymatic activity gives rise to CriglerNajjar syndrome type II (22–24). It has not been known whether the mutations causing Gilbert’s mutations were related to those causing Crigler-Najjar syndrome. Recently, a 2-base, TA insertion in the TATAA box [A(TA)7TAA] of the promoter for the UGT1A1 gene was found in all the tested Caucasian patients with Gilbert’s syndrome (25, 26). The presence of the longer TATAA element [A(TA)7TAA vs A(TA)6TAA for the mutant and wild-type, respectively]
AJG – April, 2001
Mutational Analysis for Gilbert’s Syndrome
1189
results in a 50 –70% reduction of hepatic glucuronidation activity in people with Gilbert’s syndrome (25, 27). In Asian people, the genetic basis for Gilbert’s syndrome is, however, relatively complicated (28 –30). Except for the A(TA)7TAA mutation, both heterozygous and homozygous missense mutations in the UGT1A1 gene have been identified in Japanese patients with Gilbert’s syndrome (31–34). In addition, the relationship between the genotype of the UGT1A1 gene and the phenotype of Gilbert’s syndrome remains to be clarified. Herein we report our work on genetic analysis of mutations of the UGT1A1 gene, and correlation of the genotypes of UGT1A1 mutations with serum bilirubin levels in our Gilbert’s sydrome patients.
MATERIALS AND METHODS Subjects All subjects included in this study were examined by means of serum liver biochemistry tests and hemograms. Serum samples used for the evaluation of bilirubin level were collected from patients after a fast of about 10 h. Those participants who had disease of the hepatobiliary system, or hematological diseases such as hemolytic anemia or ineffective hemopoiesis, were excluded from this study. The recruitment protocol was approved by an institutional review committee at Chang Gung Memorial Hospital. To study the genetic basis for Gilbert’s syndrome, 20 unrelated patients with Gilbert’s syndrome were included. The clinical diagnosis of Gilbert’s syndrome was based on the following criteria: 1) a consistent, mild elevation of serum bilirubin (1.5–5.3 mg/dl) with a direct-reacting fraction that was either normal or ⱕ15% of the total according to the Van den Bergh test; 2) normal serum ALAT and aspartate aminotransferase values; and 3) exclusion of the potential for hemolysis and ineffective hemopoiesis on the basis of demonstrated normal Hb, a normal mean corpuscular Hb concentration, and normal reticulocyte counts (13, 25). A total of 25 unselected healthy young adults working in our department and laboratory were enrolled for assessing the normal allele frequency of the A(TA)7TAA mutation and of other mutations in the coding region of UGT1A1 gene in our population. A patient with Gilbert’s syndrome and his parents were also included for study. Determination of UGT1A1 Promoter Genotypes For rapid and accurate determination of numbers of the (TA) repeat in the TATAA box of the UGT1A1 gene, we developed a direct sequencing assay based on polymerase chain reaction (PCR). Genomic DNA was extracted from peripheral blood leukocytes by using a standard method (35). Approximately 50 ng of genomic DNA was used in PCR for amplification of the UGT1A1 promoter region containing the TATAA element. Thirty cycles of PCR were conducted with denaturation for 40 s at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C. The PCR products
Figure 1. Polymerase chain reaction– based direct sequencing assay to the phenotype of TATAA box polymorphism of the UGT1A1 promoter. The region containing the TATAA box in the promoter for the UGT1A1 gene was amplified by PCR. The amplified DNA was then subjected to direct sequencing with an automatic sequencing kit (ABI PRISM 337 DNA sequencer; PRISM 337 collection, Sequence Analysis 3.0; Perkin-Elmer). DNA was extracted from (A) a subject with homozygosity for the wild-type A(TA)6TAA, (B) a subject with heterozygosity for A(TA)6TAA/A(TA)7TAA, and (C) a subject with homozygosity for the A(TA)7TAA mutation. Note that the sequences after six copies of the (TA) repeats are mixed by two populations of sequencing data for those heterozygous for A(TA)6TAA/ A(TA)7TAA genotype. The sequences demonstrated are in antisense polarity.
were analyzed by gel electrophoresis. The amplified DNA was purified by using a spin-column (PCR Clean-Up System Kit, Viogene, Sunnyvale, CA) and then subjected to direct PCR sequencing analysis, as previously described (36). Figure 1 demonstrates examples of the sequencing results derived from the following: 1) a subject homozygous for the wild-type of the TATAA box, A(TA)6TAA (Fig. 1A); 2) a subject heterozygous for the A(TA)6TAA /A(TA)7TAA (Fig. 1B); and 3) a Gilbert’s patient homozygous for the A(TA)7TAA mutation (Fig. 1C). The numbers of the (TA) repeat in the TATAA box were readily determined by sense orientation (data not shown), as well as by antisense orientation (Fig. 1). Sequencing for those heterozygous for the A(TA)6TAA /A(TA)7TAA reveals six tandem copies of the (TA) repeat followed by a mixed population of sequencing results, as shown in Figure 1B.
1190
Hsieh et al.
Some of the amplified DNA was cloned and then sequenced, as described before (37). The sequences of the PCR primers for amplification of the promoter region were: 5⬘-ATTAACTTGGTGTCGATTGG-3⬘ and 5⬘-AGCCATGGCGGCCTTTGCTC-3⬘. Examination of Mutations in Coding Regions of the UGT1A1 Gene Five sets of primers were used with PCR to amplify the regions from exon 1 to exon 5, respectively. PCR was conducted with denaturation for 40 s at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C for 30 cycles. The amplified DNA was then subjected to direct PCR sequencing analysis as described above. The sequences of the primers for amplification of exon 1–5 were as follows: 5⬘-GCTGGGAAGATACTGTTGATCCCAG-3⬘ & 5⬘TGGGCATGATGGGCCTAGGGTAATC-3⬘ (exon 1) 5⬘-CATAACTTACTGTATGTAGTCATCA-3⬘ & 5⬘GATATGAGGCCATGGTATAGAATCT-3⬘ (exon 2) 5⬘-TATGTCTTTCTTTACGTTCTGCTCT-3⬘& 5⬘-TCTTAATTTGACCCTGGTTTGACCT-3⬘ (exon3) 5⬘-GCAAGGGCATGTGAGTAACACT-3⬘ & 5⬘-ACCTTTTGTCATTGATGACTGCT-3⬘ (exon 4) 5⬘-TGGACCTGGCCGTGTTCTGGGTG-3⬘ & 5⬘CCACCCACTTCTCAATGGGTCTTG-3⬘ (exon 5) Statistical Analysis Mean serum bilirubin levels were compared by using Student-Newman-Keuls test (38).
RESULTS To elucidate the genetic basis of Gilbert’s syndrome in our population, 20 unrelated Gilbert’s patients were studied. Their clinical data and the genotypes for mutations in the UGT1A1 gene are shown in Table 1. Mutations in the UGT1A1 gene were identified in all cases except one. Sixteen patients were found to have the A(TA)7TAA mutation including seven homozygous and nine heterozygous. Five had Gly71Arg (replacement of glycine with arginine at the position of amino acid 71) and six had Pro229Gln. No other missense mutations including Arg209Trp, Arg367Gly, and Tyr486Asp which had been reported for Japanese Gilbert’s patients were identified. Simultaneous mutations both in the TATAA box and in the coding region were found in seven patients. Of note, all six patients with Pro229Gln also had the A(TA)7TAA mutation. A total of 25 unselected healthy adults were enrolled as controls for assessing the allele frequencies for A(TA)7TAA, Gly71Arg and Pro229Gln in our population. As shown in Table 2, 20 subjects were homozygous for the wild-type, A(TA)6TAA, and five were heterozygous for A(TA)6TAA/A(TA)7TAA. Therefore, the allele frequency for the A(TA)7TAA mutation in our controls was 10%. Accordingly, the allele frequency and the frequency of ho-
AJG – Vol. 96, No. 4, 2001
Table 1. Clinical Presentation and Genotypes of the 20 Gilbert’s Patients Case Sex 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
M M M M M M M M M M M M M M M M M F F F
Age (yr) 46 45 33 16 40 44 44 17 35 18 23 36 41 28 44 43 27 51 41 26
Genotype Serum Bilirubin (mg/dl) A(TA)7TAA G71R P229Q YY YY YY YY YY YY YY YX YX YX YX YX YX YX YX YX XX XX XX XX
4.1 3.4 2.9 2.7 2.3 2.1 1.9 3.4 3.3 2.1 2.0 1.8 1.8 1.7 1.6 1.6 2.0 2.2 1.6 3.4
XX XX XX XX XX XX XX XY XY XX XX XX XX XX XX XX YY XY XY XX
XX YX XX YX XX XX YX XX YX XX YX XX XX XX YX XX XX XX XX XX
XX ⫽ wild type; YX ⫽ heterozygosity; YY ⫽ homozygosity for the mutation.
molozygote for the A(TA)7TAA mutation in our general population were estimated to be about 11% and 1.2%, respectively. On the other hand, none of them had Gly71Arg, whereas one had heterozygosity for Pro229Gln. The allele frequencies in our controls were 0% for Gly71Arg and 4% for Pro229Gln, respectively (Table 2). By comparison, the allele frequencies of the A(TA)7TAA mutation, Gly71Arg and Pro229Gln in our Gilbert’s patients were 58%, 15%, and 15%, respectively (p ⬍ 0.001, p ⫽ 0.016, and p ⫽ 0.147 respectively, as compared to the controls) (Table 2). To evaluate further the implication of each mutation in clinical manifestations of hyperbilirubinemia, genotypes for each mutation were correlated to serum bilirubin levels. The results are shown in Table 3. The mean value for serum bilirubin levels in the seven patients homozygous for the A(TA)7TAA mutation was higher than that in the nine patients heterozygous for the A(TA)7TAA mutation (2.8 mg/dl vs 2.1 mg/dl, p ⬍ 0.05 conducted by Student-Newman-Keuls test). Of interest, the two patients with simultaneous heterozygous mutations both for Gly71Arg and for Table 2. Allele Frequency of Mutations in UGT1A1 Gene Allele Frequency (%)
A(TA)7TAA Gly71Arg Pro229Gln
Gilbert’s Patients (n ⫽ 20)
Healthy Controls (n ⫽ 25)
p*
58 15 15
10 0 2
⬍0.001 ⬍0.001 0.058
* Statistical analysis was conducted via 2 test with Yates’ correction.
AJG – April, 2001
Mutational Analysis for Gilbert’s Syndrome
Table 3. Correlation of Genotypes to Serum Bilirubin Levels Genotype
Serum Bilirubin
A(TA)7TAA
G71R
P229Q
Case
Mean (mg/dl)
SEM
YY YY YX YX YX YX XX XX
XX XX XX XX XY XY YY XY
XX YX XX YX XX YX XX XX
4 3 5 2 1 1 1 2
2.85 2.68 1.88 1.88 3.4 3.3 2.0 1.90
0.45 0.43 0.07 0.20
0.30
1191
patient, who exhibited a serum bilirubin level of 2.7 mg/dl, was simultaneously homozygous for the A(TA)7TAA mutation and heterozygous for Pro229Gln. His father had a serum bilirubin level of 2.1 mg/dl and was homozygous for the A(TA)7TAA mutation, and his mother had a serum bilirubin level of 1.4 g/dl as well as simultaneous heterozygous mutations both in the TATAA box and in Pro229Gln. Obviously, Pro229Gln was located on the maternal allele, which also contained the A(TA)7TAA mutation and was transmitted to the patient.
XX ⫽ wild type; YX ⫽ heterozygosity; YY ⫽ homozygosity for the mutation.
DISCUSSION the A(TA)7TAA mutations had serum bilirubin levels (3.4 mg/dl and 3.3 mg/dl) similar to those found in the patients homozygous for the A(TA)7TAA mutation (mean ⫽ 2.7 mg/dl and 2.9 mg/dl for those with and without an additional Por229Gln, respectively), but higher than those in the patients heterozygous for the A(TA)7TAA mutation (mean ⫽ 1.8 mg/dl and 1.8 mg/dl for those with and without additional Pro229Gln, respectively). These findings are indicative of a synergistic effect on hyperbilirubinemia exerted by concurrence of Gly71Arg with the A(TA)7TAA mutation. However, serum bilirubin levels in the patients homozygous for the A(TA)7TAA mutation with an additional Pro227Arg was not significantly different from that in the patients homozygous for the A(TA)7TAA mutation without an additional Pro227Arg (2.7 mg/dl vs 2.9 mg/dl, respectively). Likewise, serum bilirubin levels in the patients heterozygous for the A(TA)7TAA mutation with and without additional Pro227Arg had serum bilirubin levels similar to each other (1.8 mg/dl vs 1.8 mg/dl, respectively). The inheritance model of Gilbert’s syndrome was demonstrated in one patient and his parents, as shown in Figure 2. The
Figure 2. Familial study of TATAA-box polymorphism of the UGT1A1 gene. The method to determine the number of the (TA) repeat of the UGT1A1 gene is as described in the legend for Figure 1. The bold-framed square represents the proband patients. Numbers in parentheses indicate serum bilirubin levels.
We report here 20 Taiwanese patients with Gilbert’s syndrome. All patients except one were found to have mutations in the UGT1A1 gene. A total of 16 patients had the A(TA)7TAA mutation (80%) and 10 patients had missense mutations in the coding region (50%), including seven patients had simultaneous mutations both in the TATAA box and in the coding region (35%). With respect to the TATAA-box polymorphism, the A(TA)7TAA mutation, but neither A(TA)5TAA nor A(TA)8TAA, which had been reported in Caucasian people (27), was identified in our population. The A(TA)7TAA mutation was found in 16 patients including seven who were homozygous and nine heterozygous. The allele frequency of the A(TA)7TAA mutation in our Gilbert’s patients was about 58%, which was significantly higher than 10% in our healthy controls. With regard to missense mutations in the UGT1A1 gene, one was homozygous, whereas four were heterozygous for Gly71Arg and six were heterozygous for Pro229Gln. The allele frequencies of Gly71Arg and Pro229Gln in our Gilbert’s patients were 15% and 15%, respectively, in comparison to 0% and 2%, respectively, in our controls. Of note, there was one Gilbert’s patient with no identified mutation either in the TATAA box or in the coding region of the UGT1A1 gene. Further studies, especially on the promoter region other than the TATAA box, to elucidate the genetic cause for this patient are warranted. Nevertheless, the spectrum of mutations in the UGT1A1 gene associated with Gilbert’s syndrome in Taiwanese patients was relatively diverse from that found in Japanese and Caucasian patients (30, 39, 40). In a report from Sato et al. (34), 19 Japanese Gilbert’s patients were studied. Of these patients, eight had A(TA)7TAA mutations and 11 had missense mutations (58%), including three with simultaneous mutations both in the TATAA box and in the coding regions; in four patients the TATAA box was not assayed (34). Taking account of the fact that TATAA box mutations were not assayed in four of the 19 Gilbert’s patients in that study (34), the allele frequencies of the A(TA)7TAA mutation and of Gly71Arg in Sato’s patients with Gilbert’s syndrome were 47% and 21%, respectively, which were close to the frequencies of 58% and 15% in our Gilbert’s patients. However, in one other Japanese study, the allele frequency for A(TA)7TAA in Japanese Gilbert’s patients was only 14% (29). In another
1192
Hsieh et al.
study, the allele frequency of Gly71Arg in jaundiced neonates was as high as 32% (41). The reasons for these discrepancies remain unclear. By contrast, in Caucasian patients, the TATAA-box mutation represents the only cause for all examined Gilbert’s patients, and Gly71Arg has never been found either in Gilbert’s patients or in controls (25, 26, 30, 41). Serum bilirubin levels were noted to be nearly alike between in the patients heterozygous for the A(TA)7TAA mutation and in the patients heterozygous for Gly71Arg (1.8 mg/dl vs 1.9 mg/dl, Table 3). In addition, serum bilirubin levels in the patients simultaneous heterozygous mutations both for the A(TA)7TAA mutation and for Gly71Arg seemed to be higher than those in patients who were heterozygous for the A(TA)7TAA mutation (3.4 mg/dl vs 1.8 mg/dl), as well as higher than those in patients heterozygous for Gly71Arg (3.4 mg/dl vs 1.9 mg/dl). It seems that Gly71Arg and the A(TA)7TAA mutation had a comparable effect on hyperbilirubinemia but that concurrence of the two mutations had a synergistic effect. Such a synergistic effect was not observed in patients who were heterozygous for Pro229Gln with either the A(TA)7TAA mutation or the Gly71Arg mutation. Of interest, all six patients heterozygous for Pro229Gln also had A(TA)7TAA. That Pro229Gln was regularly linked to the A(TA)7TAA mutation was further demonstrated by the inheritance model in a family with Gilbert’s syndrome, wherein the maternal allele transmitted to the patient contained both the A(TA)7TAA mutation and Pro229Gln (Fig. 2). On the other hand, less than one-fourth of the alleles containing A(TA)7TAA had an additional Pro229Gln mutation (S.Y. Hsieh, unpublished data). The finding that Pro229Gln is generally associated with A(TA)7TAA is important because Pro229Gln has been reported to be one of the genetic defects causing Gilbert’s syndrome and CriglerNajjar type II in three Japanese studies (32, 34, 42). However, the TATAA box was not analyzed in the Gilbert’s patients with Pro229Gln reported by Sato et al. or by Koiwai et al. (32, 34). As well, the Crigler-Najjar type II patient heterozygous for Pro229Gln reported by Yamamoto et al. was actually found to be simultaneously homozygous for the A(TA)7TAA mutation (42). We therefore hypothesize that, among Japanese as well as Taiwanese patients, the Gilbert’s syndrome and Crigler-Najjar type II patients with Pro229Gln are very likely to have a concurrent A(TA)7TAA mutation, and that hyperbilirubinemia in these patients is more likely caused by the A(TA)7TAA mutation rather than by Pro229Gln.
ACKNOWLEDGMENTS The authors are grateful to Wei-Chu Shyu for his excellent technique assistance and to Su-Chen Ji for her excellent expertise in graph drawing. This work was supported by National Science Counsel ((NSC 89-2320-B-182-082) and
AJG – Vol. 96, No. 4, 2001
in part a grant from Chang Gung Memorial Hospital (CMRP 800). Reprint requests and correspondence: Sen-Yung Hsieh, M.D., Ph.D., Liver Research Unit, Chang Gung Memorial Hospital, 199 Tung-Hwa North Road, Taipei 105, Taiwan. Received July 7, 2000; accepted Nov. 3, 2000
REFERENCES 1. Bosma PJ, Roy-Chowdhury N, Goldhoon BG, et al. Sequence of exon and the flanking regions of human bilirubin-UDPglucuronosyltransferase gene complex and identification of a genetic mutation in a person with Crigler-Najjar syndrome, type I. Hepatology 1992;15:941–7. 2. Ritter JK, Chen F, Sheen YY, et al. A novel complex locus UGT1 encodes human bilirubin, phenol and other UDP-glucuronosyltransferase isoenzymes with identical carboxyl termini. J Biol Chem 1992;267:3257– 61. 3. Clarke DJ, Moghrabi N, Monaghan G, et al. Genetic defects of the UDP-glucuronosyltransferase-1 (UGT1) gene that cause familial non-hemolytic unconjugated hyperbilirubinemias. Clin Chim Act 1997;266:63–74. 4. Ritter JK, Crawford JM, Owens IS. Cloning of two human liver bilirubin UDP-glucuronosyltransferase cDNAs with expression in Cos-1 cells. J Biol Chem 1991;266:1043–7. 5. Bosma PJ, Seppen J, Goldhoon B, et al. Bilirubin UDPglucuronosyltransferase 1 is the only relevant bilirubin glucuronosyltransferase isoform in man. J Biol Chem 1994;269: 17960 – 4. 6. Owens D, Ritter JK. The novel bilirubin/phenol UDP-glucuronosyltransferase UGT1 gene locus: Implication for multiple nonhemolytic familial hyperbilirubinemia phenotypes. Pharmacogenetics 1992;2:93–108. 7. Bailey A, Robinson D, Dawson AM. Does Gilbert’s syndrome exist? Lancet 1977;i:931–3. 8. Fevery J. Pathogenesis of Gilbert’s syndrome. Eur J Clin Invest 1981;11:417– 8. 9. De Morais SMF, Uetrecht JP, Wells PG. Decreased glucuronidation and increased bioactivation of acetaminophen in Gilbert’s syndrome. Gastroenterology 1992;102:577– 86. 10. Arias IM, London IM. Bilirubin glucuronide formation in vitro: Demonstration of a defect in Gilbert’s syndrome. Science 1957;126:563– 4. 11. Powell LW, Hemingway E, Billing BH, et al. Idiopathic unconjugated hyperbilirubinemia (Gilbert’s syndrome): A study of 42 families. N Engl J Med 1967;277:1108 –12. 12. Arias IM, Gartner LM, Cohen M, et al. Chronic nonhemolytic unconjugated hyperbilirubinemia with glucuronyltrasferase deficiency: Clinical, biochemical, pharmacologic and genetic evidence for heterogeneity. Am J Med 1969;47:395– 409. 13. Black M, Billing BH. Hepatic bilirubin UDP-glucuronyl transferase activity in liver disease and Gilbert’s syndrome. N Engl J Med 1969;280:1266 –71. 14. Arias IM. New genetics of inheritable jaundice and cholestatic liver disease. Lancet 1998;325:82–3. 15. Bosma PJ, Chowdhury JR, Huang TJ, et al. Mechanisms of inherited deficiencies of multiple UDP-glucuronosyltransferase isoforms in two patients with Crigler-Najjar syndrome, type I. FASEB J 1992;6:2859 – 63. 16. Ritter JK, Yeatman MT, Ferreira P, et al. Identification of a genetic alteration in the code for bilirubin UDP-glucuronosltrasferase in the UGT1 gene complex of a Crigler-Najjar type I patient. J Clin Invest 1992;90:150 –5. 17. Moghrabi N, Clarke DJ, Burchell B, et al. Cosegregation of
AJG – April, 2001
18.
19.
20. 21.
22.
23.
24.
25.
26. 27.
28.
29.
intragenic markers with a novel mutation that causes CriglerNajjar syndrome type I: Implication in carrier detection and prenatal diagnosis. Am J Hum Genet 1993;53:722–9. Ritter JK, Yeatman MT, Kaiser C, et al. A phenylanaline codon deletion at the UGT1 gene complex locus of a CriglerNajjar syndrome type I patient generates a pH-sensitive bilirubin UDP-glucuronosyltransferase. J Biol Chem 1993;268: 23573–9. Chowdhury JR, Chowdhury NR, Wolkoff AW, et al. Heme and bile pigment metabolism. In: Arias JM, Boyer JL, Fausto N, eds. The liver: Biology and pathobiology, 3rd ed. New York: Raven Press, 1994:471–504. Labrune P, Myara A, Hadchouel M, et al. Genetic heterogeneity of Crigler-Najjar syndrome type I: A study of 14 cases. Hum Genet 1994;94:693–7. Eros LT, Ritter JK, Hersh JH, et al. Identification of the two single base substitutions in the UGT1 gene locus which abolish bilirubin uridine diphosphate glucuronosyltransferase activity in vitro. J Clin Invest 1994;93:564 –70. Bosma PJ, Goldhoorn B, Oude Elferink RP, et al. A mutation in bilirubin uridine 5⬘-diphosphate-glucurononsyltransferse isoform I causing Crigler-Najjar syndrome type II. Gastroenterology 1993;105:216 –20. Moghrabi N, Clarke DJ, Boxer M, et al. Identification of an A-to-G missense mutation in exon 2 of the UGT1 gene complex that causes Crigler-Najjar syndrome type 2. Genomics 1993;18:171–3. Seppen J, Bosma PJ, Goldhoorn BG, et al. Discrimination between Crigler-Najjar syndrome type I, and II by expression of mutant bilirubin uridine diphosphate-glucuronosyltransferase. J Clin Invest 1994;94:2385–91. Bosma PJ, Roy-Chowdhury J, Bakker C, et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyltransferase 1 in Gilbert’s syndrome. N Engl J Med 1995; 333:1171–5. Monaghan G, Ryan M, Seddon R, et al. Genetic variation in bilirubin UDP-glucuronosyltransferase gene promoter and Gilbert’s syndrome. Lancet 1996;347:578 – 81. Beutler E, Gelbart T, Demina A. Racial variability in the UDP-glucuronosyltransferase 1 (UGT1A1) promoter: A balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci USA 1998;95:8170 – 4. Maruo Y, Sato H, Yamano T, et al. Gilbert’s syndrome caused by a homozygous missense mutation (Tyr486Asp) of bilirubin UDP-glucuronosyltransferase gene. J Pediat 1998;132: 1045–7. Ando Y, Chida M, Kazuo N, et al. The UGT1A1*28 allele is
Mutational Analysis for Gilbert’s Syndrome
30. 31.
32.
33.
34. 35. 36.
37.
38. 39.
40.
41.
42.
1193
relatively rare in a Japanese population. Pharmacogenetics 1998;8:357– 60. Burchell B, Hume R. Molecular genetic basis of Gilbert’s syndrome. J Gastroenterol Hepatol 1999;14:960 – 6. Aono S, Adachi Y, Uyama E, et al. Analysis of genes for bilirubin UDP-glucuronosyltransferase in Gilbert’s syndrome. Lancet 1995;345:958 –9. Koiwai O, Nishizawa M, Hasada K, et al. Gilbert’s syndrome is caused by a heterozygous missense mutation in the gene for bilirubin UDP-glucuronosyltransferase. Hum Mol Genet 1995;4:1183– 6. Soeda Y, Yamamoto K, Adachi Y, et al. Predicted homozygous missense mutation in Gilbert’s syndrome. Lancet 1995; 346:1494. Sato H, Adachi Y, Koiwai O. The genetic basis of Gilbert’s syndrome. Lancet 1996;347:557– 8. Pinthanond D, Petchclain B. A preserved leukocyte preparation for quality control. Am J Clin Pathol 1974;63:32– 4. Hsieh SY, Wu YH, Tsao KC, et al. High prevalence of TT virus infection in healthy children and adults and in patients with liver disease in Taiwan. J Clin Microbiol 1999;37:1829 – 31. Hsieh SY, Yang PY, Chen HC, et al. Cloning and characterization of the extreme 5⬘-terminal sequences of the RNA genome of GB virus C/hepatitis G virus. Proc Natl Acad Sci USA 1997;94:3206 –10. Glantz SA. Primer of biostatistics. Singapore: McGraw-Hill, 1989:64 –100. Sampietro M, Lupica L, Pettero L, et al. TATA-box promoter mutant in the promoter of UDP-glucuronosyltransferase 1 gene in Italian patients with Gilbert’s syndrome. Ital J Gastroenterol Hepatol 1998;30:194 – 8. Lampe JW, Brigler J, Horner NK, et al. UDP-glucuronosyltransferase (UGT1A1*28, and UGT1A6*2) polymorphism in Caucasians, and Asians: Relationship to serum bilirubin concentrations. Pharmacogenetics 1999;9:341–9. Akaba K, Kimura T, Sasaki A, et al. Neonatal hyperbilirubinemia and mutation of the bilirubin uridine diphosphateglucuronosyltransferase gene: A common missense mutation among Japanese, Korean and Chinese. Biochem Mol Biol Int 1998;46:21– 6. Yamamoto K, Sato H, Fujiyama Y, et al. Contribution of two missense mutations (G71R and Y486D) of the bilirubin UDPglucuronosyltransferase (UGT1A1) gene to phenotypes of Gilbert’s syndrome and Crigler Najjar syndrome type II. Biochem Biophys Acta 1998;1406:267–73.