A total of 60 different ABCC2 gene mutations have been identified in infants with DJS through the HGMD, including missense, nonsense, frameshift, and splice-site disruption to date. All of our patients were confirmed to have neonatal DJS through genetic testing. Five mutations were known variants, and three novel variants were found in our study. The previously identified 5 variations are known to be pathogenic, which is consistent with the ACMG standards / guidelines: p.Arg768Trp (PS3, functional studies proven a deleterious effect; PM3, trans with a pathogenic mutation; PM5, new missense changes in amino acid residues with previously pathogenic changes; PP3, multiple computational evaluations support a deleterious effect; PP4, patient’s phenotype highly specific for disease;), p. Arg100Ter (PVS1, null variant; PM3, PP4), c.2439+2T>C (PVS1, PS3, PM3, PP4), p. Tyr119SfsTer34 (PVS1, PM2, PM3, PP4), and p.Arg1310Ter (PVS1, PM3, PP4). The p.Arg768Trp mutation in the nucleotide binding domain impairs the proper positioning and maturation of MRP2 to the apical membrane and induces neonatal cholestasis [11]. ABCC2 c.2439+2T>C results in a splice site mutation at a conserved region, and functional testing confirmed that 168 nucleotides at nucleotide positions 2272-2439 were deleted [12].
The most common variant was p.Arg768Trp, followed by p.Arg100Ter. When examining the allele frequency of the gene in the general population through ExAC, p. Arg768Trp and p.Arg100Ter occupied 0.035% and 0.012% in East Asians, respectively, and 0.0053% and 0.0027%, respectively, in non-East Asians, suggesting that there is a difference in the frequency of the mutation in DJS infants of East Asians and non-East Asians [13]. These were the most common variants in Japanese DJS infants (both allele frequencies were 20%.) [5]. However, ABCC2 gene mutations found in Korea and Japan were different from those reported in Taiwanese patients [1, 5].
There are studies on genotype-phenotype correlations in DJS [1, 5]. Lee et al. [1] reported that mutations involving the ATP binding cassettes (ABC) were associated with early-onset DJS, however, p.Thr394Arg found in this study was not involving ABC. In Japanese neonatal DJS patients, the mutation was either homozygous or compound heterozygous. These combinations were either 2 truncating mutations or a combination of truncating mutations and missense mutations [5]. It has been suggested that the combination of truncating and missense mutations may be a requirement for the phenotype of neonatal DJS [5]. However, our study has shown that neonatal DJS can occur by a combination of two missense mutations, such as homozygous p.Arg768Trp.
Novel variant, p.Gly693Glu (PM2, PM3, PP3, PP4) was classified as likely pathogenic using the ACMG standards/guidelines. The other two novel variants, p.Thr394Arg (PM2, PM3, PP4) and p.Asn718Ser (PM2, PM3, PP4) were classified as VUS. The ACMG standards / guidelines have established a new classification system for mutations that allows an indirect interpretation of whether novel variants are pathogenic mutations [7-10]. The VUS found in this study were predicted to be deleterious mutations by Polyphen-2 and MutationTaster. Functional studies are needed to understand the actual effects of these novel mutations.
Neonatal cholestasis occurs in approximately one in every 2,500 newborn babies [6]. The most common neonatal cholestasis is due to biliary atresia, accounting for 30% of all cases [6]. Genetic and metabolic causes account for approximately 10-20% of the total number of cases of neonatal cholestasis. The most common cause of neonatal cholestasis in premature infants is TPN [6, 14, 15]. Of our patients, infants diagnosed with biliary atresia accounted for 28% of the total number of cases of neonatal cholestasis diagnosed during the study period. In addition, TPN-induced cholestasis in our patients accounted for 4.5% of total number of cases. These incidences were similar to those published in previous papers [6, 14, 15].
As in Japanese studies, AST and ALT were within normal ranges in our patients [12]. In our study, we found that DB and TB were significantly higher than neonatal cholestasis due to other causes.
In liver biopsies of adult patients with DJS, melanin-like pigmentation deposited in hepatocytes is generally found. In some cases, the liver appears as a black liver [16]. It is known that cholestasis occurs due to abnormal MRP2 function. [17-19]. This protein actively transfers bilirubin from the hepatocyte to the bile duct. Mutations in the ABCC2 gene cause abnormalities in the function of MRP2, resulting in cholestasis and black liver as a result of the failure to release bilirubin and bile into the bile ducts [5]. However, melanin-like pigmentation was found in only 38% (3 of 8) of the neonatal DJS patients who underwent liver biopsy [5]. In our study, liver biopsy revealed no evidence of melanin-like pigmentation. Therefore, unlike in adult DJS, it may be difficult to identify characteristic findings of DJS through liver biopsy in neonatal DJS.
In infants with DJS, the symptoms of neonatal cholestasis may be more severe due to gene mutations of ABCC2 as well as the immaturity of the physiological metabolism of bile [1]. As the maturity of the liver gradually increases, the degree of cholestasis gradually improves, and the patients become asymptomatic. Cholestasis may reappear in adulthood due to drug, hormonal changes, infection, and pregnancy.
Long-term follow-up studies of neonatal DJS have also addressed the biphasic appearance of cholestasis [1]. In this study, hyperbilirubinemia decreased with age. The infants were jaundiced only during the neonatal period. Therefore, failure to diagnose DJS early in the neonatal period may delay the diagnosis. Considering the biphasic appearance of DJS, some may be diagnosed in the adulthood. Early diagnosis through neonatal cholestasis gene panel can prevent unnecessary evaluation. It is very important that the differential diagnosis for neonatal cholestasis with normal liver parameters except for TB/DB includes DJS. Apart from the gene panel which can test for DJS, the urinary coproporphyrin isomer I composition will help differentiate DJS as well [20].