Heme oxygenase-1 genetic variants and the conundrum of hyperbilirubinemia in African-American newborns.

BACKGROUND: African-American (AA) infants are known to have, overall, lower bilirubin levels than infants of other ethnicities during their birth hospitalization. However, they are known to have a higher incidence of severe hyperbilirubinemia and are over represented in the US Kernicterus Registry. Heme oxygenase-1 (HO) is the rate limiting enzyme in heme metabolism leading to the equimolar production of bilirubin, carbon monoxide (CO) and free iron (Fe). Short (S) (GT)n repeats (<25) in the promoter region of the gene encoding the inducible HO-1 isozyme augment its expression, while long (L) repeats (>33) lead to an attenuation, modulating the production of bilirubin and CO. The impact of HO-1 promoter polymorphisms on bilirubin levels has not been well studied in (AA) infants.

OBJECTIVE: The objectives of this study were to compare the distribution of (GT)n repeat lengths in the HO-1 promoter region in a cohort of AA infants to those found in other ethnicities and to evaluate the contribution of this polymorphism to the degree of hyperbilirubinemia and the level of COHbc in this cohort.

METHODS: We prospectively studied a cohort of term AA infants with O+ mothers. Per hospital routine, infants' blood type, Rh status, direct antiglobulin test (DAT), and total bilirubin (TB) levels were checked prior to discharge. After parental consent, blood was collected for DNA extraction and carboxyhemoglobin (COHbc) measurements at the same time as the infants' newborn screen. An infant's TB percentile risk based on the Bhutani nomogram was used to determine need for phototherapy or follow-up. (GT)n repeat length in the HO-1 promoter was determined for each allele using PCR after DNA extraction from dried bloodspots. Size of allele lengths were typed as short (S, <25), medium (M, 25-33) or long (L, >33).

RESULTS: One hundred eighty infants were studied for a total of 360 separate alleles. 12.2% (44/360) of alleles were S which was significantly less than all other ethnicities reviewed. Carboxyhemoglobin (COHbc) levels and bilirubin percentiles were higher among infants who had at least one S allele when compared to those who had at least one L allele in the cohort as a whole: COHbc 0.92 ± 0.35 vs. 0.85 ± 0.37; p = 0.28 and Bilirubin percentile 48.6 ± 34.0 vs. 44.9 ± 31.6; p = 0.51. This relationship remained when only those infants who were DAT neg were examined: COHbc 0.81 ± 0.26 vs. 0.74 ± 0.21; p = 0.11 and Bilirubin percentile 43.6 ± 29.9 vs. 37.5 ± 28.7; p = 0.28.

CONCLUSIONS: The presence of L alleles of this variant is significantly greater among infants who are either African or of African descent. There was a trend toward lower COHbc levels among infants with at least one L allele as opposed to at least one S allele, although this did not have a statistically significant impact on TB risk percentile.