Heme Oxygenase-1 Gene Polymorphisms—Toward Precision Medicine for AKI

Heme oxygenase-1 (HO-1) is a 32-kD microsomal enzyme that serves as the rate-limiting step in the breakdown of heme, releasing iron, carbon monoxide, and biliverdin. Using pharmacologic and genetic approaches, the induction of HO-1 is a protective response in renal and nonrenal settings of tissue injury.¹ The mechanisms underlying the protective functions of HO-1 have been ascribed to multiple factors (reviewed by Nath¹). HO-1 degrades heme (a potent pro-oxidant) released from destabilization of ubiquitous heme proteins during oxidative stress and can damage key cellular constituents. The enzymatic reaction generates iron, which is rapidly sequestered by ferritin; the latter is coinduced with HO-1, displaying inherent protective properties. The HO system is also a major endogenous source for carbon monoxide, a gas with antiapoptotic, anti-inflammatory, and vasodilatory properties. Biliverdin and its subsequent conversion to bilirubin (by biliverdin reductase) both exhibit potent antioxidant properties and can block lipid peroxidation. Previous work by Nath et al.² was the first to show the protective biologic function of HO-1 in an animal model of glycerol-induced rhabdomyolysis and AKI. Since these pivotal studies, the HO field has attracted numerous investigators from different disciplines and grown exponentially. HO-1 has been a subject of key reviews and editorials, with several focusing on its importance in the kidney.^1,3–5

In this issue of the Journal of the American Society of Nephrology, Leaf et al.⁶ highlight the importance of an association between genetic polymorphisms in the HO-1 promoter and the risk for developing AKI after major cardiac surgery. The polymorphism pertains to variations in length in a GT repeat region in the proximal promoter of the human HO-1 gene, which typically ranges between 12 and 40. Leaf et al.⁶ found that longer GT repeats (>27) were associated with an 1.58-fold higher odds of AKI compared with that in patients with shorter repeats (<27). Such purine-pyrimidine repeat sequences are found throughout the human genome and display length polymorphisms to variable degrees. How these sequences regulate gene expression is not entirely clear. It is postulated that regions rich in GT repeats promote Z-DNA conformation and inhibit transcriptional activity.⁷ Shorter GT repeats (<25–27) are associated with robust HO-1 expression, whereas longer GT repeats (>27) are associated with reduced HO-1 expression. These results have been confirmed in transient transfection assays using HO-1 promoter/luciferase constructs as well as in lymphoblastoid and endothelial cell lines from subjects with short and long alleles.^8,9 Although several previous papers have highlighted an association between these length polymorphisms and various diseases, this work is the first to show a positive link to AKI and more importantly, in a relatively large number of white patients undergoing cardiopulmonary bypass surgery and at high risk for AKI.⁶ Therefore, these studies represent an important advance in the field.

Polymorphisms in the GT repeat region of the HO-1 promoter have been associated with a number of diseases. In relation to kidney disease, such length polymorphisms associate with outcomes in patients with CKD, diabetic kidney disease, arteriovenous fistula maturation, and restenosis and in the setting of renal transplantation, wherein length polymorphisms in the donor kidney determine outcome after transplantation (reviewed in Exner et al.¹⁰). A few studies have provided conflicting data with respect to the significance of the association of the GT repeats and disease progression/development.¹¹ However, recent meta-analyses with careful subgroup analyses that have included confounding factors, such as age, sex, race, comorbidities, uniform definitions of disease, and complications, as well as risk factors have clearly underscored the importance of the HO-1 GT repeat polymorphism in multiple diseases.^12,13 It would be of interest to extrapolate the results of the work by Leaf et al.⁶ in other populations of patients with AKI and more importantly, those who progress to CKD or ESRD.

Strengths of the study by Leaf et al.⁶ include a large patient cohort, with 2377 patients undergoing cardiac surgery with cardiopulmonary bypass, and the detailed subgroup analysis performed by the authors. Studies from animal models have revealed that females are relatively less susceptible to AKI compared with males, findings that have been recently corroborated in humans as well.¹⁴ In this study, the authors found a stronger association between the HO-1 promoter polymorphism and risk of AKI in women compared with men, an observation that has not been previously reported. Essentially, women with longer GT repeats lost their tolerance to AKI, suggesting that HO-1 could be a potential determinant of such an effect.

Previous studies have elucidated the deleterious effects of heme, which is highly pro-oxidant and damaging to various cellular organelles, including the cell membrane, mitochondria, cytoskeleton, and the nucleus.¹⁵ The authors measured various parameters in plasma of 192 patients and found higher plasma free hemoglobin with each additional L allele, without any significant differences in ferritin, transferrin saturation, and catalytic iron levels. The higher levels of plasma free hemoglobin were associated with worse outcomes in AKI and point to the key role of heme-mediated injury in this setting.

Several questions emerge from these studies and will need additional investigation. (1) Given recent studies that have implicated plasma and urine levels of HO-1 as biomarkers of AKI in both animal models and humans,¹⁶ do HO-1 promoter polymorphisms correlate with plasma or urine levels of HO-1 or for that matter, HO enzyme activity in leukocytes in AKI? (2) What is the correlation of other genetic polymorphisms in the HO-1 gene in AKI? Askenazi et al.¹⁷ evaluated 117 premature infants for genetic variants in the HO-1 promoter and found that neonates with the TT genotype were less likely to have AKI than those with AA or AT genotype for the T(−413)A single–nucleotide polymorphism. However, they found no association between the number of GT repeats in the HO-1 gene promoter and risk of AKI, although the event rate of AKI was quite low (only 34) in this relatively small cohort of neonates. Another single–nucleotide polymorphism has also been described—G(−1135)A—but has not been studied in the context of AKI (reviewed in Exner et al.¹⁰). (3) Other key regulatory regions of the human HO-1 gene include a 220-bp intronic enhancer region that is critical in heme–induced HO-1 expression in vitro and in vivo.^18,19 Do genetic variations within this enhancer region that affect HO-1 expression associate with the development of AKI? (4) Multiple studies have highlighted an association of bilirubin levels (unconjugated bilirubin, a product of HO enzyme activity) as a protective factor in cardiovascular and renal diseases (reviewed in Hull and Agarwal²⁰). It would be interesting to see if bilirubin levels correlated in any way with the HO-1 GT repeat polymorphisms in these patients.

In this era of precision medicine, genetic testing as a guide to therapy is becoming a standard of care in oncology and other fields of medicine. Limited advances have been made in the field of AKI. In a recent review, 28 studies examining genetic predisposition to the development and/or outcome of AKI were found to be of inadequate quality, heterogeneous in defining concepts and outcomes, replicated with ambiguous results, and mostly underpowered.²¹ Most of the genes studied included inflammatory genes (e.g., IL-6, IL-10, and TNF-a), vasomotor genes (e.g., ACE, COMT, PNMT, and eNOS), and others (e.g., HIF-1 and myeloperoxidase). In this regard, the work by Leaf et al.⁶ fills a major gap in knowledge and represents an important contribution to the field.

Disclosures

None.