Inherited Mutations of the UGT1A1 Gene and other Contributing Factors to Hyperbilirubinemia.

Congenital inborn errors of the UGT1A1 gene are associated with altered UGT1A1 expression and thereby reduce or completely abolish bilirubin conjugating activity. More than 40 inherited mutations in the UGT1A1 gene are associated with hyperbilirubinemia, and the degree of deficiency of UGT1A1 activity primarily determines the severity of hyperbilirubinemia and encephalopathy (Tukey and Strassburg, 2000). Gilbert syndrome is a mild form of UGT1A1 genetic polymorphism that results in a slight reduction in UGT1A1 activity (Kadakol et al., 2000; Strassburg, 2008), whereas Crigler-Najjar (CN) syndrome exhibits complete abolishment (type 1) or severe reduction of UGT1A1 (type 2) (Ciotti et al., 1997). A few key mutations in the coding region and the promoter region of the UGT1A1 gene have been discovered in CN patients; these mutations are correlated with reduction or elimination of UGT1A1 activity (Kadakol et al., 2000; Fujiwara et al., 2015). Clinical data showed that untreated infants with CN type 1 rapidly develop high plasma levels of UCB (20–50 mg/dl), exposing them to the possibility of serious neurologic damage.

Mild forms of UGT1A1 mutations result in benign jaundice; however, when coupling other genetically determined traits, severe hyperbilirubinemia may take place. For example, infants who have hemolytic conditions caused by glucose-6-phosphate dehydrogenase deficiency and rhesus disease may be predisposed to severe hyperbilirubinemia (Huang et al., 2005; Bhutani et al., 2013). Expression of P-glycoprotein (P-gp) in the brain has also been reported to be associated with bilirubin neurotoxicity. P-gp is expressed abundantly in brain capillary endothelial cells and astrocytes of the blood-brain barriers and has the ability to transport bilirubin out of the brain across the blood-brain barrier by acting as a membrane efflux pump (Watchko et al., 1998, 2001). Compared with wild-type mice, Mdr1a (P-gp encoding gene) null mice had a higher brain bilirubin content, possibly by enhanced brain bilirubin influx, implying that Pgp expression in the blood-barrier plays a role in protecting the central nervous system against bilirubin neurotoxicity (Watchko et al., 1998, 2001). In addition to the aforementioned genetic factors, prematurity, concurrent illness, and interventions that impede bilirubin-albumin binding are also considered to be risk factors for severe hyperbilirubinemia (Bhutani and Johnson, 2009).

Experimental Models Established for Studying Neonatal Hyperbilirubinemia and Regulation of UGT1A1.

BIND is characterized by a wide range of neurologic deficits, and the underlying molecular mechanisms are only starting to emerge as a number of animal models producing the neonatal hyperbilirubinemia condition have been developed in the past few years. Most hyperbilirubinemia animal models harbor UGT1A1 mutations that occur naturally or are a result of genetic manipulations: 1) Gunn rats: Gunn (1938) discovered that mutant Wistar rats carrying a premature stop codon in the Ugt1a1 gene and exhibiting a very low UGT1A1 activity developed jaundice. These spontaneously jaundiced rats, termed Gunn rats, are deemed to be the first hyperbilirubinemia animal model, which mimics the condition of CN syndrome type 1. Since then, many studies have used Gunn rats in combination with administering sulfadimethoxine (displacing unconjugated bilirubin from albumin) or phenylhydrazine (inducing hemolysis) to examine acute bilirubin encephalopathy (Shapiro, 1988; Rice and Shapiro, 2008). 2) hUGT1A1*28 mice: To examine the role the human UGT1A1 gene in bilirubin metabolism, mice were humanized with the UGT1 locus encoding nine (−1A1, −1A3, −1A4, −1A5, −1A6, −1A7, −1A8, −1A9, and −1A10) functional UGT1A proteins. The generation of humanized UGT1 mice was accomplished in a few steps: A transgenic mouse line carrying the entire human UGT1 locus was first generated (Chen et al., 2005); the murine Ugt1 locus was inactivated by inserting a mutation into exon 4 to inactivate the nine mouse UGT1A proteins (Nguyen et al., 2008); and eventually Tg(UGT1) Ugt−^−/− mice were generated by crossing heterozygous Ugt1^+/− mice with TgUGT1 mice (Fujiwara et al., 2010). In conjunction with the fast breakdown of erythrocytes immediately after birth, these mice, which carried the genetic polymorphism of the (TA)₇ dinucleotide repeat in the TATAA box promoter element of the UGT1A1 gene (termed hUGT1A1*28 mice), all developed neonatal hyperbilirubinemia with TSB peak levels exceeding 10 mg/dl. Approximately 10% of hUGT1A1*28 newborns experienced acute encephalopathy with TSB levels >17 mg/dl at 7–12 days after birth, imitating clinical kernicterus in newborn infants (Fig. 2). 3) UGT1^F/F mice: Cre-Lox recombination sites have been positioned around common exons 2 and 3 of the Ugt1a1 gene. Breeding these mice with transgenic mice expressing tissue-specific Cre recombinase allows investigators to examine the tissue-specific impact of the Ugt1a1 gene and the other genes encoded by the Ugt1 locus. Although the liver has been assumed to be the primary target of neonatal bilirubin clearance by Ugt1a1, direct deletion of the Ugt1 locus in liver (Ugt1^ΔHep) led to only modest levels of hyperbilirubinemia (Chen et al., 2013) (Fig. 3). Clearly, other tissues, such as the GI track, are capable of participating in the clearance of bilirubin during the neonatal window. When both the liver and intestinal epithelial cells of the gastrointestinal tract are targeted for deletion of the Ugt1 locus, Ugt1^ΔIE/Hep mice develop high levels of TSB that are maintained in adult mice. 4) UFP/albumin-Cre (UAC) mice: Development of Ugt1^F/F mice required the design of a targeting vector that contained the neomycin phosphotransferase (neo^r) gene to allow antibiotic selection after heterologous recombination in ES cells. The neo^r gene was flanked by flippase recombinase sites for easy removal of the neo^r gene after identification of the Ugt1^{LoxP/FRTneoFRT/LoxP} mice. Deletion of the FRTneoFRT sequence was accomplished by crossing these mice with flippase recombinase transgenic mice, generating Ugt1^F/F mice. If the Ugt1^{LoxP/FRTneoFRT/LoxP} mice (UFP mice) are bred to homozygosity, however, insertion of the targeting construct along with the neo^r gene causes the mice to be hypomorphic for the Ugt1a1 gene, with all of the newborns and adults displaying hyperbilirubinemia (Fig. 4) resulting from poor expression of the Ugt1a1 gene in all tissues. By targeting the deletion of the Ugt1a1 gene in liver tissue using Cre/loxP recombination technology, UAC mice mice are created. Because of the preexisting hyperbilirubinemia in UFP mice, a robust kernicterus mouse model was developed in UAC mice with no mature hepatic Ugt1a1 gene or protein expression detected at 14 days after birth. The inability of the UAC mice to metabolize bilirubin results in TSB accumulation in the brain, causing severe CNS damage and leading to a 95% lethality rate (Barateiro et al., 2016). These mice serve as an excellent model to study the developmental impact of severe hyperbilirubinemia toward the onset of bilirubin-induced CNS toxicity. 5) Ugt1 mutant mice: The mouse Ugt1 locus was targeted, and a single base deletion was introduced in exon 4 of the Ugt1 gene, leading to a premature stop codon for generating Ugt1 mutant mice. These mice suffered severe neonatal hyperbilirubinemia within 5 days after birth (Bortolussi et al., 2015), very similar to the original Ugt^−/− mice (Nguyen et al., 2008). The resulting bilirubin neurotoxicity is irreversible, and mice die shortly after birth.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

The upper panel shows the genetic background of hUGT1A1*28 mice. The top diagram is a representation of the human UGT1A locus, which was inserted into the mouse genome (Tg-UGT1), and the bottom diagram shows the targeted disruption of the mouse Ugt1 locus with an insertion of the neoresistant gene into exon 4. Middle panel (left): The Ugt1^−/− neonate exhibits the phenotypic trait of jaundice with a yellow skin color compared with in Ugt1^+/− mice. Most of the Ugt1^−/− mice die before day 7. Middle panel (right): Expression of UGT1A1 in neonatal liver and small intestine tissues in hUGT1A1*28 mice. The bottom diagram shows comparisons of TSB between the Ugt1^−/− and hUGT1A1*28 mice during the developmental period.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Top: By using the Cre-loxP recombination technology, the hepatocyte- or intestinal enterocyte-specific deletion of the Ugt1a1 gene (Ugt1^ΔHep or Ugt1^ΔIE) was achieved. The bottom diagram shows the impact of the tissue-specific deletion of the Ugt1a1 gene on serum bilirubin levels.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Top: UFP mice carrying the target construct. Ugt1a1loxP[FRTneoFRT]loxP were bred into transgenic Albumin-Cre mice to generate UGT1a1F/F/Albumin-Cre mice (UAC mice). Middle (left): Comparisons of TSB levels between UFP and UAC mice during the developmental period. Middle (right): The Kaplan-Meier survival curves analyze the survival rates of UFP and UAC mice. Bottom: Excessively high levels of bilirubin penetrate to the brain of the 15-day-old UAC mouse.

Developmental Regulation of UGT1A1.

It is clear that UGT1A1 expression is a highly regulated event during development. By using hUGT1A1*28 mice, studies have shown that reduction of liver UGT1A1 gene expression in the developmental stage, which corresponds to the onset of hyperbilirubinemia and high levels of TSB, is actively regulated by pregnane X receptor (PXR). Reverse genetic experiments using PXR-deficient mice with the humanized UGT1 background demonstrated that in the absence of PXR, mice expressed significantly higher levels of UGT1A1 with a decrease in TSB levels, avoiding severe neonatal hyperbilirubinemia (Chen et al., 2012). These findings strongly indicate that PXR acts as a transcriptional repressor of the UGT1A1 gene during the neonatal period, and this is regulated as a developmental event since activation of the liver UGT1A1 gene in adult hUGT1/Pxr^−/− mice has not been observed.

Breast Milk Jaundice and the Role of Extrahepatic UGT1A1.

The connection between breast milk and hyperbilirubinemia was first described by Arias and colleagues in 1964 (Arias et al., 1964). Early studies of breast milk jaundice showed a metabolite of progesterone, pregnane-3-20-dio, was present in breast milk and was implicated in the development of jaundice (Hargreaves and Piper, 1971), although eventually no scientific consensus was reached. To date, no specific component or combinations of components have been demonstrated to definitely contribute to breast milk jaundice.

In keeping with the concept that TSB levels are higher and last longer in infants fed on breast milk, experiments using the humanized UGT1A1*28 mouse model (Fujiwara et al., 2012) reiterated that neonatal hyperbilirubinemia that occurred after breast milk feeding disappeared when mice were fed formula. In contrast to conventional knowledge, however, this study further revealed that expression of extrahepatic UGT1A1, particularly intestinal UGT1A1, is subject to induction by formula feeding and is crucial for bilirubin metabolism and clearance during postnatal transition of hepatic UGT1A1 activity, which is present only later, at the end of the suckling period (Chen et al., 2012; Fujiwara et al., 2012). Conversely, breast milk contributes to the development of hyperbilirubinemia by suppressing UGT1A1 expression in the small intestine. Breast milk was found to suppress intestinal IĸB kinase α and β, resulting in inactivation of nuclear receptor NF-ĸB and nearly complete abolishment of intestinal UGT1A1 expression (Fujiwara et al., 2012).

UGT1A1 Activation by Xenobiotics through Nuclear Receptors.

Since UGT1A1 levels and activities determine bilirubin conjugation capacity, efforts have been made in seeking endogenous and exogenous compounds that have the ability to induce UGT1A1 expression. Several studies have demonstrated that UGT1A1 expression is modulated by a number of nuclear receptors, including PXR, constitutive androstane receptor, and peroxisome proliferator–activated receptor α, in a ligand-dependent fashion. Upon ligand binding, the nuclear receptor binds to the corresponding consensus response element in the UGT1A1 promoter region and activates UGT1A1 gene expression (Xie et al., 2003; Yueh et al., 2003; Chen et al., 2005; Senekeo-Effenberger et al., 2007; Yueh and Tukey, 2007). For example, phenobarbital, a constitutive androstane receptor agonist that induces UGT1A1 expression through interacting with a phenobarbital responsive element (PBREM), has been used clinically in conjunction with phototherapy for the treatment of severe jaundice in infants to enhance bilirubin metabolism, thus reducing the need for exchange transfusion (Valaes et al., 1980; Murki et al., 2005). PBREM involvement in controlling UGT1A1 expression is also supported by a study linking polymorphism of the UGT1A1 PBREM (T-3279G) to an increased risk of hyperbilirubinemia (Sugatani et al., 2002). Glucocorticoids have also been used to treat hyperbilirubinemia: Dexamethasone-treated infants experienced lower incidence of hyperbilirubinemia than untreated controls (Madarek and Najati, 2003). Studies in mice with neonatal hyperbilirubinemia indicated that PXR serves as a key regulator after glucocorticoid treatment by inducing liver UGT1A1 expression and reducing TSB levels (Chen et al., 2012). In addition to being regulated by the aforementioned nuclear receptors belonging to the nuclear hormone receptor superfamily, UGT1A1 is subject to regulation by the nuclear receptor aryl hydrocarbon receptor and nuclear factor erythroid 2–related factor, activation of which can transcriptionally induce UGT1A1, resulting in higher levels of bilirubin excretion (Yueh et al., 2003; Yueh and Tukey, 2007; Fujiwara et al., 2010). It is worth noting that bilirubin itself has been shown to activate aryl hydrocarbon receptor and induce UGT1A in a ligand-dependent manner (Togawa et al., 2008).

Modulation of UGT1A1 Expression by Environmental Chemicals.

When hyperbilirubinemia neonatal mice were exposed to the environmental chemicals arsenic and cadmium, their TSB levels unexpectedly decreased, correlated with elevated levels of intestinal UGT1A1 expression with no detectable changes in expression of hepatic UGT1A1. Gene expression profiling data and biochemical studies revealed that, as potent inducers of oxidative stress, arsenic and cadmium alter the redox state of the intestines, leading to induction of UGT1A1 and a dramatic reduction in TSB levels (Liu et al., 2016). These results suggest that modulation of intestinal UGT1A1 activity by initiating the oxidative stress signaling pathway may be an unconventional alternative for lowering TSB and improving hyperbilirubinemia.

Hepatocyte Transplantation and Gene-Transfer Therapy.

For CNS1 patients, phototherapy is often the first-line therapy, but it may transiently lower the serum bilirubin concentrations and gradually become ineffective beyond infancy. At present, liver transplantation is the curative treatment to prevent neurologic sequelae and kernicterus, but it often requires continuous immunosuppression with substantial risks (Schauer et al., 2003).

Evidence indicated that only ∼5% of normal UGT1A1 activity is adequate to significantly lower the plasma bilirubin concentration and eliminate the risk of kernicterus (Fox et al., 1998); therefore, alternative therapies intending to alleviate hyperbilirubinemia with persistent expression of the UGT1A1 enzyme are under way in the experimental stage. A recent study illustrated that the advantage of neonatal hepatocytes over adult hepatocytes lies in the fact that neonatal hepatocytes exhibit better engraftment and repopulation capacity after transplantation, thus resulting in better bilirubin clearance in icteric Gunn rats (Tolosa et al., 2015). Significant progress has also been made in past decades by means of gene therapy using adenovirus-based or similar techniques or the correction of UGT1A1 gene defects with the site-directed gene repair approach to treat hyperbilirubinemia animals (Li et al., 1998; Kren et al., 1999; Roy-Chowdhury et al., 2001; Bellodi-Privato et al., 2005). A gene therapy study showed that a single injection of a helper-dependent adenoviral vector expressing UGT1A1 that specifically targets the liver tissue can completely correct hereditary hyperbilirubinemia in Gunn rats with long-lasting effects and low chronic toxicity (Toietta et al., 2005).

Administration of Albumin.

As a result of the high affinity of albumin to bilirubin, in a normal state, UCB is bound to albumin after transportation through the circulation to the liver (Ostrow et al., 1994). When UCB levels exceed the capacity of albumin, free bilirubin is capable of crossing the blood-brain barrier and accumulating in the brain. Therefore, a potential approach to prevent bilirubin accumulation in the brain is to increase bilirubin-binding capacity by albumin supplementation. When hyperbilirubinemia neonatal mice carrying inherited mutations of Ugt1a1 were subject to daily albumin infusion, they were rescued from neurologic damage and lethality. By increasing plasma bilirubin-binding capacity, albumin mobilizes bilirubin from tissues to plasma and results in reduced systemic plasma bilirubin levels (Vodret et al., 2015).

Regardless of efficacy of these alternative treatments, they are still in the experimental stage, and clinical trials are apparently needed to evaluate the acute toxicity, immunogenic responses, and long-term safety profile before they can be applied in the market to humans.

Develop Therapeutics Targeted to Induce UGT1A1 Gene Expression.

The use of animal models, such as humanized UGT1A1*28 mice, helps define the mechanisms that control neonatal hyperbilirubinemia and provides an important venue to exploit the impact of safe and therapeutic chemicals to regulate the UGT1A1 gene and lower TSB levels. These noninvasive approaches could take advantage of drug delivery directly to newborns or, alternatively, by lactation after drug administration to nursing mothers. In vivo studies with humanized UGT1A1*28 mice can directly exploit tissue-specific contributions, such as the liver and gastrointestinal tract, that direct bilirubin clearance while also being able to examine pharmacokinetics parameters of the inducing agents.