Acta Pædiatrica ISSN 0803–5253

REGULAR ARTICLE

Cord blood bilirubin level in relation to bilirubinUDP-glucuronosyltransferase gene missense allele in Chinese neonatesGe Sun1, Mingyuan Wu1, Jiang Cao2, Lizhong Du (dulizhong@yahoo.com.cn)3

1.Neonatal Intensive Care Unit, The Women’s Hospital, School of Medicine, Zhejiang University, Hangzhou, China2.Department of Laboratory Medicine, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou, China3.Neonatal Intensive Care Unit, The Children’s Hospital, School of Medicine, Zhejiang University, China

KeywordsCord blood bilirubin, Neonatal hyperbilirubinemia,UGT1A1 gene

Correspondence Lizhong Du, Neonatal IntensiveCare Unit, The Children’s Hospital, School ofMedicine, Zhejiang University, #57 Zhugan Lane,Hangzhou 310003, China.Tel: +86-571-87061007 |Fax: +86-571-87033296 |Email: dulizhong@yahoo.com.cn

Received27 October 2006; revised 28 May 2007; accepted1 August 2007.

DOI:10.1111/j.1651-2227.2007.00501.x

AbstractAim: To investigate bilirubin UDP-glucuronosyltransferase (UGT1A1) gene allele in healthy Chinese

neonates, their cord bilirubin level and the subsequent hyperbilirubinemia to determine relationships

among them.

Methods: Cord blood of 48 neonates was obtained to determine the exon 1 of UGT1A1 gene, total

serum bilirubin, albumin, glutamic-pyruvic transaminase (GPT), glutamic-oxalacetic transaminase

(GOT) and haemoglobin (Hb) concentration. Neonatal jaundice was assessed by measurement of

transcutaneous bilirubin (TCB) and serum bilirubin. Neonates were divided into two groups according

to mutant or normal allele to compare the variables.

Results: Nineteen infants had the nucleotide 211 G→A allele, 3 had the heterozygous variation

(686C→A, 845 A→T, 231G→A). In the 211 A allele group, cord bilirubin was significantly higher

than in the 211 G allele group (p = 0.034), but there were no differences in albumin (p = 0.678),

GPT (p = 0.460), GOT (p = 0.440) and Hb (p = 0.886). The TCB (at 48, 96 h), the frequency of

the hyperbilirubinemia and prolonged jaundice were also significantly higher in the 211 A allele group.

Conclusions: The UGT1A1 gene codon G71R allele is a risk factor for neonatal hyperbilirubinemia in the Chinese

population. Its effect on bilirubin metabolism may present early on, as well as late in foetal life.

Jaundice, one of the most common manifestations to oc-cur in neonatal period, affects 30–50% of newborn babies(1). In China, the peak serum levels of total bilirubin infull-term neonates range between171.0 and 205.2 �mol/Lwhich is double that of those found in Caucasian populations(85.5–102.6 �mol/L) (2). Total serum bilirubin levels above220.5 �mol/L are found in 34.4% of normal Chinese terminfants (3). The incidence of kernicterus is also higher amongAsian neonates (4). These findings suggest that a genetic fac-tor might be involved.

In neonates, bilirubin metabolism is in transition from thefoetal to the adult stage. Bilirubin production is increasedbecause of increased ineffective erythropoiesis, largererythrocyte volumes, shortened erythrocyte life span andprematurity of bilirubin UDP-glucuronosyltransferase activ-ity (UGT1A1). An imbalance between the production andelimination of bilirubin results in varying levels of hyper-bilirubinemia in the neonate (5). Recently, the relationshipbetween the polymorphism of bilirubin uridine 5′-diphosphate-glucuronosyltransferase (UGT1A1) gene andneonatal hyperbilirubinemia has been reported (6–8). Themost common genetic polymorphism encountered in theCaucasian population is an additional TA insertion in theTATA box of the UGT1A1 gene promoter. However, inJapanese and Taiwan Chinese the missense mutations in theUGT1A1 coding region, a G→A transition at nucleotide 211(known as G71R), were the predominant findings (9–11).Variations in the promoter region of the UGT1A1 gene mayresult in a reduced expression of UGT1A1, and the varia-

tions within the coding region of this gene may also resultin dysfunction of UGT1A1 with abnormal structure (12–14). These genetic variations may cause decreased UGT1A1activity in neonates, leading to the accumulation of uncon-jugated bilirubin in serum. Whether UGT1A1 gene polymor-phism influences the later foetal bilirubin level and how itis related to the development of exaggerated neonatal jaun-dice is still unknown. In this report, we studied these re-lationships by investigating UGT1A1 gene codon 71 allelein healthy Chinese neonates. We also studied cord bilirubinconcentration and the subsequent development of neonatalhyperbilirubinemia.

PATIENTS AND METHODSStudy populationForty-eight Chinese neonates, 23 males and 25 females, whowere born between April 2004 and July 2004 at Sir RunRun Shao hospital in Hangzhou were randomly selected andprospectively studied. The study was approved by the hos-pital review board for the protection of human subjects. Allneonates were full-term with a mean gestational age of 39.2± 1.5 weeks (range: 37–41 weeks) and mean birthweight of3202 ± 560 g (range: 2600–4500 g). All the mothers werehealthy during pregnancy and did not receive any parturifa-cient during delivery. Babies with high-risk factors for hyper-bilirubinemia such as neonatal asphyxia, ABO blood groupincompatibility, haemolytic anaemia, dehydration, vomiting,polycythemia, hypoglycemia, infection, cephalohematoma

1622 C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1622–1625

Sun et al. Cord blood bilirubin and UGT1A1 gene allele

or liver dysfunction were excluded from the study. All infantsreceived a combination of breast milk and formula milk.

Analysis of the UGT1A1 gene exon 1Umbilical cord blood samples were obtained for routine testsfor blood type and Direct Coomb’s test from 48 newborninfants. The samples collected at night were kept in a refrig-erator and protected from light until assay. Genomic DNAwas isolated from the leukocytes of all cord blood samples.Exon 1 and its flanking intronic region of the UGT1A1 genewas amplified by polymerase chain reaction (PCR). The am-plification reaction from genomic DNA has been describedelsewhere (7,8). The primers used for PCR were 5′-GAA CCTCTG GCA GGA GCA AA-3′ and 5′-TCG ATC CAA AGTAAT ACA TCT GAA AGA ATA T-3′. The amplification re-action mixture volume was 50 �L. Conditions for the PCRreaction were as follows: an initial denaturation for 5 minat 95◦C followed by 45 sec at 95◦C, 45 sec at 58◦C and45 sec at 72◦C for 35 cycles. A final extension for 7 minat 72◦C was performed to ensure complete extension ofPCR products. The products were analysed by electrophore-sis on a 3% agarose gel stained with ethidium bromide.The sequences of the amplified DNA fragments were de-termined directly (Shanghai Huanuo Biotechology Co. Ltd.,Shanghai, China).

Analysis of cord blood haemoglobin and serumbiochemical variablesCord blood samples were also collected for biochemicaltesting, including total serum bilirubin, serum albumin,glutamic-pyruvic transaminase (GPT) and GOT (BeckmanCX7 Chemistry Analyzer; Beckman Coulter Inc., Fullerton,CA, USA). The haemoglobin (Hb) concentration was alsodetermined.

Jaundice assessment and follow-upNeonatal jaundice was assessed by measuring the ‘JaundiceIndex’ with an Infant Jaundice Meter (HD-368, Hong Kong)during the first 24 h of life and then daily at the same time ofthe day for at least 4 days in the first week of life. The meterreadings were pre-calibrated according to the manufacture’sinstructions. All measurements were obtained under quietconditions without the infant crying. If hyperbilirubinemiadeveloped, blood samples were taken to confirm the totalserum bilirubin concentrations and were followed until14 days of life. Hyperbilirubinemia was defined as a totalserum bilirubin concentration (venous blood) ≥221 �mol/Lin the first four post-natal days. At this borderline serumbilirubin level, the babies can be treated either by photother-apy or simple observation according to the recommendedguidelines of the Chinese Neonatal Society (15). Prolongedjaundice is defined as a total serum bilirubin concentration(venous blood) ≥150 �mol/L by day 14 (16).

Statistical analysisTo understand whether the variant nucleotide 211 ofUGT1A1 gene influences the neonate’s serum total biliru-bin level in cord blood, and the severity of neonatal jaun-

dice, 48 neonates were divided into two groups accordingto mutant or normal allele. Analyses were conducted usingSPSS for Windows (Release 10.1, SPSS, Inc., Chicago, IL,USA). Student t- test was used to compare all the variablesbetween two groups. The presence of hyperbilirubinemia orprolonged jaundice between two groups was analysed bychi-square analysis. A p-value of <0.05 was considered asstatistically significant.

RESULTSNineteen of the 48 infants (39.6%) had an identical transi-tion allele at nucleotide number 211 in exon 1 of UGT1A1,which changed the codon from GGA to AGA. Fifteen in-fants were heterozygous for G71R, one was homozygous.Three neonates carried compound heterozygous variationsin exon 1. One infant was heterozygous for the 686 C to A,one infant carried the 845 A to T and the other infant washeterozygous with the 231 G to A variation.

The characteristics of the infants in the 211 A allele groupand the 211 G allele group were compared. No statisticaldifferences were found in birthweight (3160 ± 357 g vs.3229 ± 417 g; p = 0.56), gestational age (38.8 ± 1.1 weeks vs.39.2 ± 1.1 weeks; p = 0.25) between the two groups.

The total serum bilirubin level in cord blood was sig-nificantly higher in neonates with 211 A allele (35.8 ±5.5 �mol/L) than in the 211 G allele group (32.0 ±6.1 �mol/L) (p = 0.034). There were no statistical differ-ences in cord blood Hb (14.8 ± 1.0 g/dL vs. 14.8 ± 1.2 g/dL;p = 0.886), serum albumin (35.0 ± 2.0 g/L vs. 34.7 ±2.9 g/L; p = 0.678), GPT (10.7 ± 3.6 IU/L vs. 9.9 ±3.8 IU/L; p = 0.460) and GOT (36.6 ± 15.5 IU/L vs. 33.6 ±11.4 IU/L; p = 0.440) levels between the two groups.

The transcutaneous bilirubin (TCB) levels increased withpost-natal age in both groups. The TCB index was signifi-cantly higher at 48 h (19.0 ± 2.3 vs. 17.5 ± 2.5; p = 0.037)and 96 (24.0 ± 2.5 vs. 22.3 ± 2.9; p = 0.044) in the 211 Aallele group than in the 211 G allele group. No significant dif-ference between the two groups at 24 (13.9 ± 3.0 vs. 12.8 ±2.7; p = 0.177) or 72 h (21.6 ± 2.5 vs. 20.9 ± 2.7; p = 0.323).

In the 211 A allele group, five neonates (all with het-erozygous 211G→A allele) had a serum bilirubin more than221 �mol/L in the first four post-natal days. The maximumlevel of hyperbilirubinemia noted was 304.38 �mol/L. Theircord bilirubin levels were 44.46, 34.20, 44.46, 39.33 and37.62 �mol/L, respectively. Three neonates (all with het-erozygous 211 G→A allele) were diagnosed with pro-longed neonatal jaundice. In the 211 G allele group, twoneonates developed hyperbilirubinemia with cord biliru-bin levels of 47.88 and 42.75 �mol/L, respectively. Onewas diagnosed with prolonged neonatal jaundice. No othercauses of hyperbilirubinemia or prolonged jaundice werefound in any of these patients. The frequency of hy-perbilirubinemia and prolonged neonatal jaundice in the211 A allele group was significantly higher than thatin the 211 G allele group (chi-square analysis, � 2 =6.555, p = 0.016). It is interesting that hyperbilirubine-mia did not develop in the homozygous 211 variationwith a cord bilirubin of 42.75 �mol/L. Hyperbilirubinemia

C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1622–1625 1623

Cord blood bilirubin and UGT1A1 gene allele Sun et al.

also did not develop in three neonates who carriedcompound heterozygous variations. Six of the 7 infants whodeveloped hyperbilirubinemia received phototherapy.

DISCUSSIONUGT1A1 is the key enzyme for bilirubin conjugation. Muta-tion in the UGT1A1 gene may lead to familial unconjugatedhyperbilirubinemia such as the rare, debilitating Crigler–Najjar syndrome and Gilbert’s syndrome (9). Patients withCrigler–Najjar syndrome are usually diagnosed soon afterbirth because of their extremely high serum bilirubin con-centrations. These patients suffer from severe hyperbiliru-binemia throughout life. Gilbert’s syndrome is a benign,inherited condition. It affects about 6% of the adult pop-ulation and is usually diagnosed during adolescence (13,14).The most common genetic polymorphism encountered inwhites with Gilbert syndrome is an additional TA insertionin the TATAA box of the UGT1A1 gene promoter. There areusually 7 repeats-(TA)7TAA instead of the usual 6 repeats-(TA)6TAA (9–11,14). In East Asian populations, Gilbert syn-drome appears to result from missense mutations in thecoding area of the UGT1A1, include Gly71Arg, Pro229Gln,Tyr486Asp, Arg209Trp and Arg367Gly (14,17,18). The mostcommon form of these is a G→A transition at nucleotide211, which causes arginine to replace glycine at position 71of the corresponding protein product. The mutation is preva-lent in the Japanese and Chinese populations (5,7,8,17,19).In 1980, Odell suggested that some infants who have ex-aggerated neonatal jaundice may be exhibiting the effectsof Gilbert’s syndrome. Gilbert’s syndrome could also causemore rapid rise of serum bilirubin or more prolonged jaun-dice in neonates (20). Recent reports have confirmed thesehypotheses (6–8,16,21).

Neonatal jaundice is due to an imbalance between biliru-bin production and excretion by the liver. Insufficient re-moval of bilirubin by the liver due to the relative immaturityof the liver enzyme system, including the uridine diphos-phate glucuronosyl transferase, may lead to the developmentof neonatal jaundice (22). Nearly all foetal bilirubin is un-conjugated due to a limited ability of the foetal liver to conju-gate bilirubin. Unconjugated bilirubin is rapidly transferredto the maternal circulation by the placenta. The pregnantmother has a large reserve capacity for bilirubin conjuga-tion and excretion, but higher cord bilirubin levels are foundamong infants who later become jaundiced when comparedto those in whom hyperbilirubinemia does not develop (23–25). Rataj J et al. studied umbilical cord serum bilirubin con-centration in 800 healthy full-term neonates as a predictor ofsubsequent hyperbilirubinemia. They found that only 2.4%of neonates developed jaundice when their cord bilirubinconcentration was less than 17.1 �mol/L, whereas jaundicedeveloped in 89% of the neonates whose cord bilirubin levelswere above 42.75 �mol/L (24). Knupfer’s investigation alsoproved a clear relation between umbilical cord serum biliru-bin and the development of hyperbilirubinemia in healthyterm and near-term newborns (25). These observations in-dicate that the mechanisms for the subsequent developmentof jaundice are already active in late foetal life (23–25). It

can be speculated that in addition to maternal bilirubin-clearing, the liver even pre-natally, plays an important rolein the metabolism of bilirubin (25). Cord bilirubin might bethe first indicator for evaluating neonatal jaundice (22–26).In our study, 19 of 48 neonates had the transition mutationat nucleotide number 211 in exon 1 (G71R), one infant washomozygous, and 15 were heterozygous. As this mutation isvery common in the Chinese population, we may call it anallele rather than a mutation. In the in vitro expression studyby Yamamoto et al., G71R in the homozygous and heterozy-gous genetic states decreased the UGT1A1 enzyme activitiesto as low as 32.2% and 60.2% of normal, respectively (27).The decreased enzyme activities are thought to cause mildto moderate delayed elimination of bilirubin. Three infantscarried compound heterozygous variations in exon 1: 686C to A, 845 A to T and 231 G to A. The 686 C to A vari-ation has been reported in Taiwan Chinese, but the 845 Ato T and 231 G to A have not (8). How these compoundheterozygous genetic states affect the enzyme activities isstill unknown. Although Kawade et al. reported the activityof the UGT enzyme in late foetal life is only 0.1–1% of theadult level (28), our study did show that the average cordbilirubin concentration was significantly higher in the 211A allele group than in those with the normal allele. Thissuggests that the allele in the UGT1A1 coding region G71Rmay influence neonatal serum bilirubin levels early as wellas late in foetal life, although this influence remains mild.These neonates will have a higher cord blood bilirubin thannormal infants at birth. When comparing cord blood serumalbumin, Hb, GOT and GPT between two groups, no differ-ence was found. These factors were unlikely to influence thecord blood total serum bilirubin levels in both groups of thestudy neonates.

Daily transcutaneous jaundice index has been shown to behighly correlated with serum bilirubin concentration (1,26).Jaundice index was measured rather than serum bilirubinconcentration to allow non-invasive longitudinal study ofsubjects (6). Bancroft et al. measured the transcutaneousjaundice index in 151 healthy infants in the first week oflife. The study found that the babies with A(TA)7TAA ho-mozygotes had an increased jaundice index during the first2 days of life compared with heterozygotes or A(TA)6TAAhomozygotes (6). Maruo et al. reported that the mean peaktranscutaneous jaundice index in babies heterozygous forG71R were significantly higher than that of normal babies(22.8 ± 1.99 vs. 20.4 ± 2.48) (7). In this study, we also foundthat babies with G71R missense allele had a significant in-crease in jaundice index at 48 and 96 h post-natally. Thisfinding further supports the relationship between UGT1A1gene polymorphisms and the development of neonatal jaun-dice. The fact that no significant difference in bilirubin levelsbetween the two groups at 24 or 72 h was identified may berelated to our limited sample size, and therefore further in-vestigation is needed.

In analysing the incidence of jaundice between the twogroups, it is not surprising that the hyperbilirubinemiaand prolonged neonatal jaundice were more common inneonates with G71R variation than in those with the normal

1624 C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1622–1625

Sun et al. Cord blood bilirubin and UGT1A1 gene allele

allele. This result confirms the theory that G71R of UGT1A1gene exon 1 is a risk factor for neonatal hyperbilirubinemiain Chinese neonates. This may be an infantile and induciblephenotype of Gilbert’s syndrome, and these neonates shouldbe closely followed.

In conclusion, allele in the UGT1A1 coding region G71Ris a risk factor for neonatal hyperbilirubinemia in Chinesepopulation. The effect of this gene polymorphism on biliru-bin metabolism may present throughout foetal life.

ACKNOWLEDGEMENTWe thank Galina Barford APRN, Neonatal Nurse Prac-titioner, Division of Neonatology, University of Utah, forcarefully reviewing this manuscript.

References

1. Carbonell X, Botet F, Figueras J, Riu-Godo A. Prediction ofhyperbilirubinemia in the healthy term newborn. ActaPaediatr 2001; 90: 166–70.

2. Liu Yi. Assessment and intervention of neonatal jaundice.Chin J Pediatr 2001; 39: 321–2.

3. Ding Guofang, Zhang Suping, Yao Dan, Na Qi, WangHuazhuang, Li Li, et al. An epidemiological survey onneonatal jaundice in China. Chin J Pediatr 2000; 38: 624–7.

4. Maisels MJ. Jaundice. In:Avery GB, Fletechen MA,MacDonald MG, editors. Neonatology, Pathophysiology andManagement of the Newborn, 4th ed. New York: JB LippincottCo, 1994: pp. 630–75.

5. Huang CS, Chang PF, Huang MJ, Chen ES, Chen WC.Glucose-6-phosphate dehydrogenase deficiency, theUDP-glucuronosyl transferase 1A1 gene, and neonatalhyperbilirubinemia. Gastroenterology 2002; 123: 127–33.

6. Bancroft JD, Kreamer B, Gourley GR. Gilbert syndromeaccelerates development of neonatal jaundice. J Pediatr 1998;132: 656–60.

7. Maruo Y, Nishizawa K, Sato H, Doida Y, Shimada M.Association of neonatal hyperbilirubinemia with bilirubinUDP-glucuronosyltransferase polymorphism. Pediatrics 1999;103: 1224–7.

8. Huang CS, Chang PF, Huang MJ, Chen ES, Hung KL, TsouKI. Relationship between bilirubin UDP-glucuronosyltransferase 1A1 gene and neonatal hyperbilirubinemia. PediatrRes 2002; 52: 601–5.

9. Mackenzie PI, Miners JO, McKinnon RA. Polymorphisms inUDP glucuronosyltransferase genes: functional consequencesand clinical relevance. Clin Chem Lab Med 2000; 38: 889–92.

10. Miners JO, McKinnon RA, Mackenzie PI. Geneticpolymorphisms of UDP-glucuronosyltransferases and theirfunctional significance. Toxicology 2002; 181-182: 453–6.

11. Bock KW. Vertebrate UDP-glucuronosyltransferases:functional and evolutionary aspects. Biochem Pharmacol2003; 66: 691–6.

12. Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A,Oostra BA, et al. The genetic basis of the reduced expression

of bilirubin UDP-glucuronosyltransferase 1 in Gilbert’ssyndrome. N Engl J Med 1995; 333: 1171–5.

13. Clarke DJ, Moghrabi N, Monaghan G, Cassidy A, Boxer M,Hume R, et al. Genetic defects of theUDP-glucuronosyltransferase-1 (UGT1) gene that causefamilial non-haemolytic unconjugated hyperbilirubinemias.Clin Chim Acta 1997; 266: 63–74.

14. Kaplan M, Hammerman C, Maisels MJ. Bilirubin genetics forthe nongeneticist: hereditary defects of neonatal bilirubinconjugation. Pediatrics 2003; 111: 886–93.

15. Zhonghua yixue hui erkexue fenghui xinshenger xuezu. 2001Intervention criteria for neonatal hyperbilirubinemia.Zhonghua Er Ke Za Zhi 2001; 39: 185–7.

16. Monaghan G, McLellan A, McGeehan A, Li Volti S, Mollica F,Salemi I, et al. Gilbert’s syndrome is a contributory factor inprolonged unconjugated hyperbilirubinemia of the newborn. JPediatr 1999; 134: 441–6.

17. Hsieh S-Y, Wu Y-H, Lin D-Y, Chu C-M, Wu M, Liaw Y-F.Correlation of mutational analysis to clinical features inTaiwanese patients with Gilbert’s syndrome. Am JGastroenterol 2001; 96: 1188–93.

18. Maruo Y, Sato H, Yamano T, Doida Y, Shimada M. Gilbertsyndrome caused by a homozygous missense mutation(Tyr486Asp) of bilirubin UDP-glucuronosyltransferase gene. JPediatr 1998; 132: 1045–7.

19. Kaba K, Kimura T, Sasaki A, Tanabe S, Wakabayashi T, HiroiM, et al. Neonatal hyperbilirubinemia and a common mutationof the bilirubin uridine diphosphate-glucuronosyltransferasegene in Japanese. J Hum Genet 1999; 44: 22–5.

20. Odell GB. Neonatal Hyperbilirubinemia. New York: Gruneand Stratton, 1980: 39–41.

21. Maruo Y, Nishizawa K, Sato H, Sawa H, Shimada M.Prolonged unconjugated hyperbilirubinemia associated withbreast milk and mutations of the bilirubin uridine diphosphate-glucuronosyltransferase gene. Pediatrics 2000; 106: E59.

22. Law LK, Pang CP, Cheung KL, Fok TF. Cord bloodbiochemistry and idiopathic neonatal jaundice. Clin Chem1996; 42: 1716–7.

23. Knudsen A. Prediction of the development of neonataljaundice by increased umbilical cord blood bilirubin. ActaPaediatr Scand 1989; 78: 217–21.

24. Rataj J, Kornacka M, Korman E. Usefulness of measuringbilirubin levels in cord blood for predicting hyperbilirubinemiain newborns. Ginekol Pol 1994; 65: 276–80.

25. Knupfer M, Pulzer F, Gebauer C, Robel-Tillig E, Vogtmann C.Predictive value of umbilical cord blood bilirubin for postnatalhyperbilirubinemia. Acta Paediatr 2005; 94: 581–7.

26. Knudsen A. Prediction of later hyperbilirubinemia bymeasurement of skin color on the first postnatal day and fromcord blood bilirubin. Dan Med Bull 1992; 39: 193–6.

27. Yamamoto K, Sato H, Fujiyama Y, Doida Y, Bamba T.Contribution of two missense mutations (G71R and Y486D)of the bilirubin UDP glycosyltransferase (UGT1A1) gene tophenotypes of Gilbert’s syndrome and Crigler-Najjarsyndrome type II. Biochim Biophys Acta 1998; 1406: 267–73.

28. Kawade N, Onishi S. The prenatal and postnatal developmentof UDP-glucuronyltransferase activity towards bilirubin andthe effect of premature birth on this activity in the humanliver. Biochem J 1981; 196: 257–60.

C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1622–1625 1625