Hepatic glucuronidating activity is essential for efficient biliary excretion of bilirubin. Bilirubin glucuronidation is mediated by the enzyme bilirubin uridine diphosphate-glucuronosyltransferase (UGT1A1) [1]. UGT1A1 is encoded by five exons of the UGT1A1 gene. Defects in UGT1A1 cause a non-hemolytic unconjugated hyperbilirubinemia including Crigler-Najjar (CN) syndrome and Gilbert syndrome. CN syndrome is an autosomal recessive disease caused by mutations in UGT1A1 gene [2,3,4,5]. Genetic lesions causing an absence of enzymatic bilirubin glucuronidation result in CN syndrome type I (CN-1), whereas mutations causing incomplete deficiency of the enzyme result in CN syndrome type II (CN-2) [6]. CN-2 is characterized by intermediate levels of hyperbilirubinemia (7-20 mg/dL). In contrast to CN-1, kernicterus is rare and phenobarbital administration results in induction of the residual UGT1A1 activity with consequent reduction of serum bilirubin levels (>25%). CN-1 is caused by genetic changes that produce premature truncation or critical amino acid residue substitution. CN-2 results from substitution of single amino acid residues that markedly reduce, but do not abolish, enzyme activity [7].

Currently, the spectrum of mutations in Korean patients with CN-2 is not characterized. The aim of this study is to investigate the mutation spectrum of UGT1A1 gene in Korean children with CN-2.

All children initially presented with neonatal jaundice requiring phototheray. The median age at the time of jaundice onset was 3 days of birth (range: 1-14 days) (Table 1). They had persistent indirect hyperbilirubinemia. Serum total bilirubin decreased with the administration of phenobarbital. Median follow up periods were 2 years (range: 1-17 years).

We identified a missense mutation (p.Y486D produced by the nucleotide substitution c.1456T>G in exon 5) and a polymorphism (p.G71R produced by the nucleotide substitution c.211G>A in exon 1). Homozygous p.Y486D was found in all five patients while p.Y486D was not found in healthy controls (p<0.01). Three patients had an associated homozygous p.G71R and two a heterozygous p.G71R. The allele frequency of p.G71R in CN-2 patients and controls was 0.88 and 0.16, respectively (p<0.01). No significant difference in mean serum bilirubin levels was found between homozygous carriers of p.G71R and heterozygous carriers.

This study presents the molecular characterization of UGT1A1 in Korean CN-2 patients. The homozygous p.Y486D missense mutation was observed in all five CN-2 patients. Four patients with CN-2 from Japan have been found to be homozygous for p.Y486D [8]. The p.Y486D mutation has also been reported in a Tunisian CN-2 patient [9]. Relative UGT1A1 activity of a homozygous p.Y486D expression model was 7.6% of the normal level [10]. Enzyme activity of a patient with CN-2 is generally less than 10% of normal level. Premature terminating mutations in UGT1A1 were more commonly observed among CN-1 patients, while missense mutations were more frequently associated with a CN-2 phenotype. This finding explains the milder phenotype observed in CN-2 and the inducibility of the residual enzyme activity by phenobarbital administration [7].

We found that all CN-2 patients had an associated homozygous or heterozygous p.G71R. Combination of homozygous p.Y486D and heterozygous p.G71R has been previously reported in a Taiwanese CN-2 patient [11]. A Japanese study revealed that three of the four CN-2 patients had an associated homozygous p.G71R [8]. It has been suggested that a homozygous p.Y486D associated with a homozygous or a heterozygous p.G71R always expresses CN-2 phenotypes in Asian populations.

In the present study, the allele frequency of p.G71R was 0.16 in healthy controls. p.G71R is considered as a polymorphism, which is any sequence variant present at a frequency >1% in a population. It has been reported that -c.3279T>G, TATA box, and p.G71R are three common polymorphisms in the Korean population, and that the allele frequency of p.G71R is 0.21. Serum bilirubin in homozygous carriers of p.G71R was reportedly significantly higher than those in heterozygous carriers or homozygous carriers of wild type allele [12]. A mutant expression model also showed that a heterozygous p.G71R and a homozygous p.G71R reduced transferase activity to 60% and 30% of the normal level, respectively [10]. Serum total bilirubin was markedly elevated when homozygous p.Y486D was present in combination with homozygous p.G71R [8]. In this study, no significant difference in serum bilirubin was found between homozygous carriers of p.G71R and heterozygous carriers. Further larger studies are helpful to investigate whether p.G71R additionally increases serum bilirubin level in homozygous p.Y486D carriers with CN-2. The p.G71R polymorphism has been identified as a risk factor for Gilbert syndrome [13,14] and neonatal breastfeeding jaundice [15,16] in East Asia.

In summary, the combination of homozygous p.Y486D and homozygous or heterozygous p.G71R is identified in CN-2 patients. The p.Y486D and p.G71R allele can be screened for the mutation analysis of UGT1A1 in Korea. Further larger genetic studies are needed to confirm our findings.