UGT1A1 gene variants and clinical risk factors modulate hyperbilirubinemia risk in newborns

Abstract

Objective:

To study the contribution of UGT1A1 gene variants and clinical risk factors in modulating hyperbilirubinemia risk in newborns.

Study design:

Seven UGT1A1 gene variants and clinical risk factors were studied in 113 hyperbilirubinemia cases and 218 control newborns. Hyperbilirubinemia was defined as the total serum bilirubin levels >95th percentile of the American Academy of Pediatrics nomogram. The study population included term (37 to 41 weeks) newborns below 2 weeks of age.

Result:

UGT1A1 gene variants, namely, c.211G>A, g.−3279T>G, TATA box polymorphism and CAT insertion were identified as independent molecular risk factors for neonatal hyperbilirubinemia, whereas c.686C>A, c.1091C>T and c.1456T>G were not detected in study cohort. Among clinical risk factors, excessive weight loss, sepsis and ABO incompatibility emerged as independent risk factors. Co-expression of UGT1A1 variants and clinical risk factors further accentuated the risk of neonatal hyperbilirubinemia.

Conclusion:

Multiple risk factors, whether genetic or clinical, are instrumental in modulating hyperbilirubinemia risk in newborns. Disordered bilirubin conjugation through interactions of UG1TA1 gene variants contributes to the clinical phenotype of neonatal hyperbilirubinemia.

Introduction

Significant neonatal hyperbilirubinemia¹ (SNH) is a multifactorial disorder involving bilirubin overproduction, reduced conjugation and increased enterohepatic recycling.² Both clinical and genetic factors are known to contribute to the development of SNH.^{3, 4} UDP-glucuronosyltransferase 1A1(UGT1A1) enzyme is a key determinant of bilirubin metabolism.⁵ Variations of both coding and non-coding regions of UGT1A1 gene are associated with diminished activity of the enzyme, leading to neonatal hyperbilirubinemia.^{5, 6, 7, 8} Studies have documented variable contribution of UGT1A1 gene variants to SNH in different populations.^{3, 4, 9, 10, 11} The interaction among genetic polymorphisms and clinical risk factors may modulate the hyperbilirubinemia risk,^{3, 4} and better knowledge on gene–environment interaction may advance our understanding of this complex disorder. The purpose of the present study was to understand the contribution of UGT1A1 gene variants and clinical risk factors in modulating the risk of hyperbilirubinemia in Indian newborns.

Subjects and methods

Study population

This prospective case–control study was conducted over a period of 3 years in Sir Sundar Lal Hospital, Banaras Hindu University, Varanasi, located in central part of India. The study was approved by the Institute Ethics Committee. The study population included a convenient sample of 113 newborns with hyperbilirubinemia (cases) and 218 newborns as controls. Only inborn, term (37 to 41 weeks) newborns of ⩽2 weeks’ of age were studied. Hyperbilirubinemia was defined as total serum bilirubin (TSB) levels >95th percentile, whereas control newborns had TSB levels <75th percentile on the American Academy of Pediatrics nomogram.¹² Cases and controls were matched for gestational age. Control group comprised of healthy newborns with physiological jaundice who stayed in birth hospital for various maternal reasons, such as delivery by cesarean section, infected episiotomy wound and other medical disorders, and not for neonatal reasons. None of the control newborns received phototherapy for jaundice. Newborns with major congenital malformations and those with G6PD deficiency were excluded from the study. All newborns were of Indian origin. Written informed consent was taken from parents before inclusion in the study.

We used following definitions of clinical risk factors: sepsis—positive blood culture in conjunction with compatible clinical presentation; ABO incompatibility—mother with O blood group and newborn with A or B blood group; Rh incompatibility—Rh negative mother and Rh positive newborn; excessive weight loss—>10% weight loss since birth; hypothyroidism—thyroid stimulating hormone levels >20 mU l⁻¹. We did not use direct antiglobulin test positivity as a criterion for labelling Rh or ABO incompatibility between mother and newborn.

Study procedure

Demographic and perinatal details were recorded in all. Neonates were followed daily for jaundice by visual inspection. TSB was done if icterus appeared on face on day one, or there was yellow staining of legs subsequently. TSB was repeated as per clinical judgement. In controls, TSB levels were obtained at the time of routine thyroid screening between 3 and 7 days of life, as per our hospital protocol. Median age at enrolment was 5 days in control group and 4 days in hyperbilirubinemia group. Highest recorded TSB value was used for categorizing the newborns as case or control, as defined earlier. If a baby was discharged from hospital before 72 h of age, TSB was obtained before discharge and baby was recalled for assessment of jaundice on day 5 of age. All newborns were seen again in follow-up clinic at 2 weeks of age. Investigations included blood group of mother and baby, complete blood count, reticulocyte count, T4, thyroid stimulating hormone, G6PD assay, direct Coomb’s test and sepsis work-up, if needed. Most of these tests were carried out on cord blood collected at the time of birth. SNH was managed as per the guidelines of American Academy of Pediatrics.¹²

Genetic analysis

Five milliliter of cord blood was collected in heparinized vacutainer tubes and stored at −4 °C before DNA isolation. Total genomic DNA was isolated from the leukocytes using standard salting out method.¹³ The number of CAT and TA repeats in UGT1A1 gene promoter was determined by PCR, followed by single-strand conformation polymorphism analysis.¹⁴ PCR-restriction fragment length polymorphism method was applied to detect the variants of UGT1A1 at nucleotide position g.−3279G>T, c.211G>A, c.686C>A, c.1091C>T and c.1456T>G, as described previously.^{3, 15} To assign a genotype to each single-strand conformation polymorphism/restriction fragment length polymorphism pattern, some representative samples were subjected to DNA sequencing (ABI Genetic Analyzer 3130, Applied Biosystem, Foster City, CA, USA).

Genomic DNA was added to PCR mixtures of 25 μl consisting of 10 × PCR buffer (Fermentas, Pittsburgh, PA, USA), 100 mM Tris-HCl (pH 8.8), 50 mM (NH₄)₂SO₄, 250 mM KCl, 20 mM MgSO₄, 0.2 μl (1.25 mM) deoxy nucleotide triphosphates (Fermentas) and 1 U of DNA Polymerase (Sigma Aldrich, St Louis, MO, USA). PCR amplification consisted of an initial denaturation of 2 min at 94 °C followed by 30 cycles of denaturation at 95 °C for 30 s, annealing of oligonucleotides at the primer pair-specific temperature for 30 s, extension at 72 °C for 30 s, final extension for 5 min at 72 °C and hold at 4 °C. The PCR reactions were carried out in a PCR thermocycler (Eppendorf Master Cycler, Eppendorf, Hamburg, Germany).

Statistical analysis was done using the SPSS version 16.0 (IBM, Armonk, NY, USA). Data were expressed as mean±standard deviation for continuous variables and percentage for categorical variables. Continuous variables were analyzed using Student’s t-test or Mann–Whitney U-test as applicable. Categorical variables were compared by χ² test or Fisher’s exact test. Univariate and multivariate analyses were done for calculating odds ratio and 95% confidence interval for risk factors. A P-value of <0.05 was considered significant.

Results

Table 1 shows the characteristics of the study population. Cases and controls were comparable for birth weight, gestational age, gender, mode of delivery, maternal age and gravidity, and Apgar scores. Among cases, TSB levels above 20, 25 and 30 mg dl⁻¹ were observed in 29 (25.7%), 3 (2.6%) and 2 (1.8%) newborns, respectively.

Table 1 Characteristics of the study population

Table 2 shows that factors such as previous sibling with jaundice, excessive weight loss (peak weight loss was 15.4%), hypothyroidism, sepsis and ABO-incompatibility were significantly more common in cases than in controls.

Table 2 Relationship between risk factors and significant neonatal hyperbilirubinemia

Of the seven UGT1A1 polymorphisms studied, only four, viz, c.211G>A, g.−3279G>T, TATA box polymorphism and CAT insertion, were identified in the study population, whereas other three polymorphisms, viz, c.686C>A, c.1091C>T and c.1456T>G, were not detected (Table 3). On univariate analysis, we found higher frequency of c.211 G>A variant, g.−3279T>G variant and TATA box mutation in infants with SNH. Although CAT insertion was more common in cases compared with controls, the difference was not significant. In the present study cohort, we found high allele frequency of g.−3279T>G and TATA box variants, whereas c.211G>A and CAT insertion variants had very low allele frequency. Compound heterozygotes were more frequently detected in cases than in controls (odds ratio: 4.2; 95% confidence interval: 1.6 to 10.7).

Table 3 Distribution of UGT1A1 gene variants

A logistic regression analysis of risk factors (Table 4) showed excessive weight loss, sepsis and ABO incompatibility as independent clinical risk factors for SNH. Of the UGT1A1 gene variants, CAT insertion, c211G>A, g.−3279T>G variant and TATA box mutation were significantly associated with SNH. The risk of hyperbilirubinemia increased with multiple risk factors. In presence of two risk factors, the odds of developing SNH were nearly nine times higher; with three or more risk factors, the odds were even higher.

Table 4 Analysis of risk factors for significant neonatal hyperbilirubinemia

Serum bilirubin levels, age at enrolment, need for phototherapy/exchange transfusion and duration of phototherapy were comparable regardless of the presence of one or more risk factors (data not shown). Depending on underlying etiology, we divided the study population into two groups, with known cause (n=61) or without any known cause (n=52). No difference was observed in the two groups with regard to the clinical characteristics and UGT1A1 variants (data not shown).

Discussion

The present study demonstrates that multiple genetic and clinical risk factors modulate hyperbilirubinemia risk in newborns. Of the seven variants of UGT1A1 gene studied, only four viz, g.−3279T>G, TATA box polymorphism, CAT insertion and c.211G>A, were found in the present cohort. Other three common variants, viz, c.686C>A, c.1091C>T and c.1456T>G, were not detected.

c.211G>A variant is associated with reduced isozyme activity, ranging from 60% in heterozygous state to 14 to 32% of normal levels in homozygous state.^{7, 16} In the present study, c.211G>A was detected infrequently, the estimated allele frequency was 0.002 in control newborns and 0.03 in hyperbilirubinemia group. In contrast, a much higher allele frequency has been documented in East Asian populations, such as Chinese (0.23),^{9, 11} Koreans (0.23)¹¹ and Japanese (0.13)¹¹, accounting for higher prevalence of neonatal hyperbilirubinemia in these groups. Our findings also differ from a recent study conducted in the north-western part of India where this variant was not detected.¹⁰ Difference in sample size and genetic heterogeneity of the population could account for this discrepancy. In contrast to present study cohort, 31.2% of newborns with hyperbilirubinemia in their population had G6PD deficiency.¹⁰ On the other hand, a study from Eastern India reported an allele frequency of c.211G>A variant to be 0.06 and 0.03 in Gilbert’s syndrome and healthy controls, respectively.¹⁷

We observed a higher frequency of g.−3279T>G variant in SNH, the allele frequency was 0.44 in controls and 0.55 in hyperbilirubinemia group. Reported allele frequency of this variant is 0.49, 0.5 and 0.26 in Americans,⁴ Malays¹⁸ and Japanese¹⁹ populations, respectively. Phenobarbital responsive enhancer module regulates the transcriptional activity of UGT1A1 gene and its g.−3279T>G variant may reduce transcriptional activity to 62% of normal.⁸ However, the exact role of g.−3279T>G variant in the pathogenesis of SNH is debatable. A study from Malaysia found g.−3279T>G gene variant as a significant risk factor for SNH.¹⁸ On the other hand, studies from other populations have not replicated these findings.^{4, 19}

In the present study, TATA box polymorphism was found to be significantly associated with hyperbilirubinemia. This observation is consistent with other studies.^{10, 20, 21, 22} TATA box polymorphism reduces transcriptional activity of UGT1A1 gene to 60 to 80% of normal,^{5, 6} and TATA box mutation is the most common mutation associated with Gilbert syndrome in the Western⁵ and Indian¹⁷ populations.

(TA)₇ variant was the second most frequent UGT1A1 gene variant detected in our study. The observed allele frequency in present study cohort was 0.32 and 0.41 in controls and cases, respectively. Other Indian studies have reported an allele frequency of (TA)₇ in the range varying from 0.29 to 0.41 in general population and 0.49 in neonatal hyperbilirubinemia and 0.88 in patients with Gilbert’s syndrome.^{10, 17, 23, 24}. We did not find any instance of (TA)₅ or (TA)₈ variant in present study cohort.

Although CAT insertion was more frequently found to be association with the hyperbilirubinemia group (n=3; 2.7%) than in control newborns (n=1; 0.4), the difference was not significant. There are very few studies analyzing the contribution of CAT insertion to SNH. In an earlier Indian study, the CAT insertion frequency was 0.06 in patients of Gilbert’s syndrome and none in controls.¹⁷

Weight loss, sepsis and ABO incompatibility were significantly associated with SNH. In contrast to other studies,^{3, 25} we did not find any contributory role of breastfeeding to SNH as breastfeeding was almost universally present in both the groups. On the other hand, weight loss emerged as the strongest risk factor for significant hyperbilirubinemia (odds ratio: 36.1, confidence interval: 14.0 to 324.8). Excessive weight loss (⩾10%) after birth is clearly indicative of insufficient breastfeeding. Thus, inadequate breastfeeding rather than breastfeeding per se may be causally linked to SNH. Ten of twelve newborns with excessive weight loss were exclusively breastfed. This situation in newborns is akin to a state resembling fasting hyperbilirubinemia found in adults.²⁶ Fasting increases entero-hepatic circulation by reducing intestinal motility.²⁷

The present study emphasizes the multiplicity of risk factors, both genetic and clinical, in the pathogenesis of SNH. The risk of hyperbilirubinemia was accentuated in the presence of multiple risk factors. Three-fourth of the newborns with hyperbilirubinemia expressed ⩾2 risk factors. Thus, the search for the etiology of neonatal hyperbilirubinemia should be broad-based rather being confined to identify only single cause. Further, genetic factors contribute to neonatal hyperbilirubinemia even in cases where etiology is obvious. It is generally assumed that genetic factors have more significant role in the causation of idiopathic SNH only.²⁰ However, we observed essentially no difference in the distribution of UGT1A1 gene variants in the two categories of hyperbilirubinemia (known vs unknown cause; data not shown).

The main limitation of this study is that we relied on visual assessment of jaundice rather than transcutaneous bilirubinometry. It is possible that some hyperbilirubinemic newborns might have been missed or misclassified as controls because of infrequent blood sampling.

To conclude, the present study demonstrates that multiple risk factors, whether genetic or clinical, are instrumental in modulating hyperbilirubinemia risk in newborns. Disordered bilirubin conjugation through interactions of UG1TA1 gene variants contributes to the clinical phenotype of SNH.