Abstract
Abstract. Heme oxygenase-1 (HO-1) is a microsomal enzyme involved in the degradation of heme, resulting in the generation of biliverdin, iron, and carbon monoxide. Recent attention has focused on the biologic effects of product(s) of this enzymatic reaction that have important antioxidant, anti-inflammatory, and cytoprotective functions. Induction of HO-1 occurs as an adaptive and beneficial response to a wide variety of oxidant stimuli, including heme, hydrogen peroxide, cytokines, growth factors, heavy metals, nitric oxide, and oxidized LDL. HO-1 has been implicated in several clinically relevant disease states, including transplant rejection, hypertension, acute renal injury, atherosclerosis, and others. Previous studies indicate a protective role for HO-1 in heme and non-heme-mediated models of acute renal injury using chemical inducers and inhibitors of HO-1. Studies in HO-1 knockout mice further corroborate these observations, highlighting the important role of HO-1 in the pathophysiology of acute renal injury. Expression of HO-1 has been linked to prolonged xenograft survival and is important in transplant rejection as well. More recently, the first known case of human HO-1 deficiency was reported with several phenotypical similarities to the mouse HO-1 knockout. The role of HO-1 has extended far beyond its initial description as an enzyme involved in heme degradation to being an important mediator in modulating adaptive and protective responses not only in renal injury, but in other organ systems as well.
Cellular heme derived from ubiquitously disposed heme proteins such as cytochromes, peroxidases, respiratory burst oxidases, pyrrolases, catalase, nitric oxide synthases, hemoglobin, and myoglobin is degraded by the heme oxygenase enzyme system (1). The heme oxygenase enzyme opens the heme ring, resulting in the liberation of equimolar quantities of biliverdin, iron, and carbon monoxide (CO), as shown in Figure 1. Biliverdin is subsequently converted to bilirubin by biliverdin reductase. Since 1968, when Tenhunen, Marver, and Schmid (2,3) first described the rate-limiting enzymatic reaction catalyzed by heme oxygenase, interest in this field has extended beyond just the heme-degrading function of this enzyme. Recent attention has centered on the biologic effects of the reaction product(s) that potentially possess important antioxidant, anti-inflammatory, antiapoptotic, and possible immune modulatory functions (4,5,6,7,8,9,10,11,12,13,14). The cytoprotective attributes of heme oxygenase activity were recognized as a consequence of its robust induction as an adaptive response in cells/tissues exposed to a wide variety of injurious stimuli. These include heme, ultraviolet radiation, hydrogen peroxide, cytokines, endotoxin, growth factors, heavy metals, oxidized LDL, shear stress, hyperoxia, nitric oxide (NO), and NO donors—all stimuli imposing a significant shift in cellular redox (4,5,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32).
The heme oxygenase catalyzed reaction. Heme (iron protoporphyrin IX) is cleaved between rings A and B by heme oxygenase to yield equimolar quantities of iron (Fe3+), carbon monoxide (CO), and biliverdin. O2 and NADPH are required for this reaction. Biliverdin is then converted to bilirubin by biliverdin reductase. M, V, and P represent methyl, vinyl, and propionyl groups, respectively.
Two isoforms of heme oxygenase have been extensively characterized: an inducible enzyme (heme oxygenase-1 [HO-1, Hmox1]) and a constitutive isoform (HO-2, Hmox2) (1). Recently, a third isoform (HO-3) was identified that shares approximately 90% amino acid homology with HO-2 (33). Table 1 summarizes some of the salient features of the different isoforms of heme oxygenase. These isozymes are products of different genes, with HO-1 and HO-2 sharing only roughly 40% amino acid homology (1,34). In this review, the term heme oxygenase will refer to HO-1 unless otherwise specified. HO-1 and HO-2 differ primarily in regulation and tissue distribution. HO-1 is ubiquitously distributed in mammalian tissues, and HO-2 is constitutively expressed in the brain, testes, endothelium, medullary thick ascending limb segment of the rat kidney, liver (hepatocytes and sinusoidal endothelial cells), and myenteric plexus of the gastrointestinal tract (1,35,36,37,38). It has been suggested that HO-2 may function as a physiologic regulator of cellular function, whereas HO-1 plays a role in modulating tissue responses to injury in pathophysiologic states (39).
Characteristics of the different heme oxygenase isoformsa
The beneficial effects of HO-1 induction occur via several postulated mechanisms (6). Increased HO-1 activity results in degradation of the heme moiety, which is pro-oxidant and potentially toxic (40,41,42). Several cellular targets including lipid bilayers, mitochondria, cytoskeleton, and components of the nucleus are prone to damage by heme (42). In addition to detoxifying heme, elevated HO-1 activity results in the generation of bilirubin, an antioxidant capable of scavenging peroxy radicals and inhibiting lipid peroxidation (43,44). Ferritin, an intracellular repository for iron, is coinduced with HO-1 allowing safe sequestering of the unbound iron liberated during the degradation of heme (45). More recently, modulation of intracellular iron stores and increased iron efflux has been suggested as a mechanism for the cytoprotective effects of HO-1 expression (46). Finally, CO has received considerable attention as a signaling molecule similar to NO (1,47,48,49). CO has vasodilatory effects mediated via cGMP (50), and also possesses antiapoptotic as well as cytoprotective functions (14,49). The biologic effects of CO are summarized in Table 2.
Effects of carbon monoxide
Diseases Associated with HO-1
HO-1 has also been implicated in several clinically relevant disease states, including atherosclerosis (51,52,53), hypertension (54), acute renal injury (55), toxic nephropathy (56), transplant rejection (10,11,57), endotoxic shock (17), chronic obstructive lung disease (58), and Alzheimer's disease (59), as well as others (60,61,62,63,64,65,66,67). An abundance of HO-1 mRNA and protein has been identified in human atherosclerotic plaques, providing in vivo relevance to this enzyme in atherosclerosis (51). The major stimulus for the induction of HO-1 in atherosclerotic plaques is oxidized LDL (21,22,23,24) and, more specifically, its fatty acid component linoleyl hydroperoxide (25). The observation that the induction of HO-1 inhibits monocyte chemotaxis demonstrates the functional significance of HO-1 expression in atherosclerosis (52). Furthermore, the beneficial attributes of HO-1 may counteract the potential toxic effects of oxidized LDL. Microsatellite polymorphisms in the human HO-1 gene promoter, specifically the length of a dinucleotide (GT) repeat in the 5′ flanking region, is reported to be associated with susceptibility to emphysema in smokers (58). No similar polymorphisms have as yet been identified in any renal disorder. A list of renal and extrarenal diseases associated with HO-1 is shown in Table 3.
Pathologic conditions associated with HO-1a
Expression of HO-1 in the Kidney
HO-1 is induced in the kidney in certain models of acute renal injury: glycerol-induced acute renal failure (55), nephrotoxic serum nephritis (68), and cisplatin nephrotoxicity (56). Induction of HO-1 has been demonstrated in renal tubules in a rat model of rhabdomyolysis and nephrotoxic serum nephritis (55,68). We have observed induction of HO-1 in both proximal and distal tubules in a rat model of cisplatin nephrotoxicity (Figure 2A). Glomerular expression of HO-1 mRNA (69) and protein (Figure 2B) (A. Agarwal and K. A. Nath, unpublished observations) has been reported in animal models of diabetes. In acute renal transplant rejection, HO-1 is dramatically elevated in infiltrating cells (Figure 2C), which have been identified as macrophages (10). The exact subcellular localization of HO-1 along the nephron has not been reported. The reasons for the differences in regional induction of HO-1 in renal injury are not entirely clear. It is interesting to speculate that the site-specific expression along the nephron or in the interstitium occurs due to proximity of the stimulus. Perhaps delivery of heme proteins to the proximal tubule results in proximal tubular induction. Inflammatory cytokines that mediate transplant rejection may induce HO-1 in infiltrating interstitial cells, and increased oxidant stress in diabetic glomeruli results in glomerular expression of HO-1. Recent studies have reported low levels of constitutive expression of HO-1 in the renal inner medulla, where it may contribute to the maintenance of renal medullary circulation (70).
Heterogeneity of heme oxygenase-1 (HO-1) expression in renal injury. Frozen renal tissue sections were stained with a polyclonal anti-rat HO-1 antibody as the primary antibody and goat anti-rabbit IgG conjugated with FITC as the secondary antibody. (A) Immunofluorescence studies performed in saline (control) and cisplatin-treated (6 mg/kg body wt, intravenously) rat kidneys at day 5 demonstrate positive staining in renal tubules in the cisplatin-treated animals. (B) In a rat model of streptozotocin-induced diabetes mellitus, HO-1 is induced in glomeruli within 1 wk of onset of diabetes, whereas no induction is seen in saline-treated (control) glomeruli. (C) Positive staining for HO-1 (green, arrow) is observed in infiltrating macrophages but not in renal tubules in a rat model of renal allograft rejection. Nuclei were counterstained with ethidium bromide (orange, arrowhead).
In acute renal injury, previous studies using chemical inducers and inhibitors of HO-1 have demonstrated that expression of HO-1 in renal tubules is cytoprotective in both heme (55) and non-heme-mediated models of renal failure (56). For example, in heme-mediated renal injury in a rat model of acute renal failure associated with rhabdomyolysis, administration of a specific inhibitor of HO-1, tin protoporphyrin, worsened renal damage while prior induction with hemoglobin led to a considerable decrease in mortality (55). In cisplatin-induced toxic nephropathy, a model of acute renal injury not directly dependent on the delivery of heme proteins to the kidney, inhibition of HO-1 by tin protoporphyrin significantly worsened both structural and functional parameters of renal injury, providing indirect evidence for a role of HO-1 in this model (56). Induction of HO-1 mRNA and protein also occurs in gentamicin nephrotoxicity. In contrast to the cisplatin model, HO-1 inhibition does not influence the course of renal functional impairment in gentamicin nephrotoxicity, thus suggesting heterogeneity in the functional significance of HO-1 induction in the kidney (56).
The initial observations in the above studies of acute renal injury have been further corroborated by studies in HO-1 knockout mice in both the glycerol model of rhabdomyolysis (71) and cisplatin-induced acute renal failure (72). Studies using HO-1 knockout mice help to define a precise role for HO-1 in acute renal injury because the potential pitfalls of pharmacologic maneuvers (chemical inducers and inhibitors), which have effects beyond altering HO-1 activity per se, are eliminated. Specifically, the metalloporphyrin inhibitors of HO-1 affect other enzyme systems such as NO synthase and guanylate cyclase, in addition to other nonspecific effects (73,74,75). Moreover, although selective overexpression of HO-1 is cytoprotective against heme protein and cisplatin-mediated toxicity, studies using heme proteins or metalloporphyrins as modulators of HO-1 activity have yielded contrasting results (76,77).
HO-1 Knockout Mouse
Poss and Tonegawa (78) generated mice deficient in HO-1 by targeted deletion of a 3.7-kb region including exons 3 and 4 and a portion of exon 5 of the mouse HO-1 gene. Using standard procedures for generating knockout mice, they developed heterozygote mice in the 129/Sv x C57BL/6 hybrid strain background. A partial prenatal loss of HO-1-deficient mice resulted in a less than expected number of homozygote (-/-) mice with heterozygote (+/-) mating pairs. No viable litters were obtained with (-/-) matings. Using in vitro fertilization with gametes from (-/-) and (+/-) animals, additional (-/-) mice were generated. The adult (-/-) mice were smaller than the wild-type (+/+) mice and exhibited a normochromic, microcytic anemia. Kidneys of (-/-) mice over 20 wk of age showed evidence of iron deposition in renal cortical tubules. Similar iron deposition was also observed in the liver. A progressive chronic inflammation characterized by hepatosplenomegaly, lymphadenopathy, leukocytosis, hepatic periportal inflammation, and occasionally glomerulonephritis was reported in addition to the iron deposition (78). On the basis of these observations, the authors indicated that the HO-1 knockout mouse would be a useful model for studying iron metabolism.
These authors also reported that embryonic fibroblasts from the HO-1-deficient animals were more sensitive to oxidant stimuli such as heme, hydrogen peroxide, paraquat, and cadmium (12). In vivo administration of endotoxin to (-/-) mice (6 to 9 wk of age) resulted in significantly more liver injury and mortality (12), suggesting that expression of HO-1 is an important defense against oxidant stress. More recently, the first known case of human HO-1 deficiency was reported in a 6-yr-old boy (79). Several phenotypical similarities were observed between this patient and the HO-1 knockout mouse, including growth failure, anemia, increased iron binding capacity and ferritin, tissue iron deposition, lymphadenopathy, leukocytosis, and increased sensitivity to oxidant injury (78,79). The patient underwent a renal biopsy for evaluation of hematuria and proteinuria. Mild mesangial proliferation and thickening of the capillary loops were observed in the glomeruli. In addition, endothelial swelling and detachment with subendothelial deposits were also evident in the glomeruli (79). Similar to the HO-1 knockout mouse, iron deposition was present in the liver parenchymal cells, Kupffer cells, and in proximal tubule cells of the kidney (79).
Using the glycerol model of heme protein toxicity in the HO-1 knockout mice, Nath et al. (71) observed significantly worse renal function and tubular injury accompanied by 100% mortality in the (-/-) mice compared with (+/+) mice (71). In addition, both younger (7 to 10 wk) and older (20 to 30 wk) HO-1 knockout mice were more sensitive to the administration of the heme protein (hemoglobin), resulting again in severe renal failure and increased mortality (71). We have also demonstrated that HO-1 gene expression modulates cisplatin-mediated renal tubular injury both in vivo and in vitro (72). Transgenic mice deficient in HO-1 (-/-) develop more severe renal failure and have significantly greater renal injury (apoptosis and necrosis) compared with (+/+) mice treated with cisplatin (Figure 3, A through D). Renal tissue sections of HO-1 (-/-) mice examined at day 3 after cisplatin administration demonstrated increased evidence of acute renal injury characterized by tubular necrosis, degeneration, casts, loss of brush border, and red blood cell extravasation compared with HO-1 (+/+) mice, which showed only mild changes (Figure 3, A and B). In our model of cisplatin nephrotoxicity, a higher degree of apoptosis (confirmed by the terminal deoxynucleotidyl transferase-mediated nick end labeling assay and electron microscopy) was seen in both proximal and distal tubules of HO-1 (-/-) mice kidneys (Figure 3, C and D) (72). In vitro studies in human renal epithelial cells demonstrated that overexpression of HO-1 resulted in a significant reduction in cisplatin-induced cytotoxicity (72).
Cisplatin-induced renal injury in the HO-1 knockout mouse. Light microscopy of kidneys from HO-1 (+/+) and HO-1 (-/-) mice stained with hematoxylin and eosin at day 3 after cisplatin administration. (A) Cisplatin-treated HO-1 (+/+) mice kidneys showing mild changes of acute renal injury. (B) Cisplatin-treated HO-1 (-/-) mice kidneys have more severe changes of renal injury with tubular necrosis and loss of brush border, red cell extravasation (arrowhead), and tubular casts. (C and D) Apoptotic nuclei in kidneys of mice treated with cisplatin (20 mg/kg, intraperitoneal injection) at day 3. A threefold increase in the degree of apoptosis as identified by positive terminal deoxynucleotidyl transferase-mediated nick end labeling staining in kidney nuclei (arrowheads) was present in the HO-1 (-/-) mice compared with the HO-1 (+/+) mice. Magnification, ×100. (Inset) Higher magnification (×400) showing an apoptotic nucleus in the renal tubule of an HO-1 (-/-) mouse kidney. Reprinted with permission from Am J Physiol (72).
Expression of HO-1 has been linked to prolonged xenograft survival; hearts transplanted from HO-1 knockout mice showed shortened survival (3 d versus >60 d) compared with hearts obtained from (+/+) or (+/-) mice in the presence of complement inhibitors, cyclosporine, and cobra venom factor (11). Amersi et al. (80) have also demonstrated that overexpression of HO-1, using either cobalt protoporphyrin or gene transfer using an adenoviral vector containing HO-1, attenuated ischemia-reperfusion injury and prolonged survival after cold ischemia/isotransplantation of fatty livers in the obese Zucker rat. These studies will aid in strategies to increase the availability of donor organs in transplantation by salvaging cadaveric organs from older individuals or those with suboptimal function (81).
There are two potential problems for investigators who anticipate performing studies using the HO-1 knockout mice. First, the partial prenatal lethality limits the number of (-/-) mice that can be obtained in a short period of time, necessitating the maintenance of large breeding colonies resulting in high animal care costs. Second, the phenotype of the knockout which is characterized by tissue iron deposition that is observed beyond 20 wk of age is important to note, because such inherent changes could affect the results of experiments. The use of young mice (<8 to 10 wk) would potentially obviate this problem.
HO-1 in Inflammation
In addition to the pathogenesis of acute renal injury, HO-1 plays an important role in the inflammatory response (9). Willis et al. (9) reported the beneficial effects of heme oxygenase upregulation in a model of pleural inflammation. Inhibition of HO-1 using tin protoporphyrin significantly worsened inflammatory exudate and cell number, while prior induction with iron protoporphyrin showed a significant reduction in inflammatory cell infiltration and exudate, suggesting that HO-1 activity modulates the inflammatory response. Similar findings have been reported in other models of inflammation as well (53,68). Vogt et al. (68) demonstrated a novel phenomenon of acquired resistance to renal tubular injury in acute glomerular inflammation that was dependent on the induction of HO-1 in renal tubules. Expression of HO-1 is also protective in hyperoxia-induced lung injury (63). The generation of carbon monoxide appears to be the mechanism involved in this model since exogenous administration of carbon monoxide protects against hyperoxia-induced lung injury (49), results that are similar to gene delivery studies with HO-1 (63).
HO-1 in Transplant Rejection
Induction of HO-1 occurs in immune-mediated renal injury, as demonstrated by the expression of this protein in infiltrating macrophages in acute renal transplant rejection (Figure 2C) (10). The novel observation of Soares et al. (11) that the expression of HO-1 was clearly associated with xenograft survival has provided functional significance of HO-1 induction in transplant rejection. Studies by Hancock et al. (57) have provided convincing data that induction of protective genes such as HO-1 also protects allografts from chronic rejection. In addition to transplant rejection, HO-1 induction also attenuates ischemia/reperfusion injury that affects donor organ quality and subsequent transplantation (80,81). Recently, RDP1258, a novel peptide derived from the HLA class I heavy chain, has been shown to possess immunoregulatory function in vitro and in vivo via modulation of HO-1 enzyme activity (82,83). Administration of RDP1258 was associated with a decrease in graft infiltrating cells and prolongation of cardiac allograft survival (82). Similar results were obtained in a rat renal allograft model (83). These recent developments provide new therapeutic approaches in the quest for prolongation of graft survival and the overall success of organ transplantation.
Molecular Regulation of HO-1
The cellular processes underlying HO-1 induction are complex and tightly regulated; however, one feature common to most of the stimuli that upregulate HO-1 is a significant shift in cellular redox. The induction of HO-1 in response to heme, heavy metals (cobalt chloride, cadmium), NO and NO donors, oxidized LDL (specifically the fatty acid component, linoleyl hydroperoxide), and by cytokines has been demonstrated to be a consequence of de novo transcription (7,15,25,27,28,84,85,86).
Consensus binding sites for NF-κB, AP-1, AP-2, and interleukin-6 responsive elements, as well as other transcription factors, have been reported in the promoter region of the HO-1 gene (7,87), suggesting a potential role for these trans-acting factors in modulating HO-1 gene expression. The regulation of HO-1 may also be species- or cell-specific. For example, the promoter of the rat HO-1 gene has a metal-dependent and heat shock-responsive element, although the heat shock element is not functional in the human HO-1 gene (88,89). In fact, the HO-1 gene has the distinction of being referred to as HSP32, because it is induced by heat shock in rodent cells.
A potential cadmium response element has been identified 5′ to the transcriptional initiation site of the human HO-1 gene, between -4.5 and -4 kb (84). In the mouse HO-1 gene, Alam et al. (86) have described two distal enhancers, one at -4.0 and another at -10 kb, that are required for induction of HO-1 in response to heme, heavy metals, hydrogen peroxide, and sodium arsenite. We have shown that an approximately 4.5-kb human HO-1 gene promoter fragment, which responds to known inducers of the gene such as heme and cadmium, does not completely recapitulate the induction of the gene seen by steady-state Northern analysis, and also lacks the cis-acting elements necessary for linoleyl hydroperoxide-dependent gene induction (25). The absence of a response with the 4.5-kb human HO-1 gene promoter fragment to linoleyl hydroperoxide is further corroborated by the failure of this fragment to respond to other stimuli that directly increase de novo HO-1 gene transcription, i.e., hydrogen peroxide, hyperoxia, and iron/hyperoxia (27). It is possible, therefore, that the human HO-1 gene also requires distal enhancer element(s) similar to the mouse HO-1 gene for induction. Experiments using additional promoter transfections, chromatin structure analysis, and in vivo footprinting are currently under way in our laboratory to delineate the region of the human HO-1 gene that controls induction.
Future Research Opportunities
The availability of the HO-1 knockout mouse will facilitate studies aimed at evaluating a direct role for HO-1 in physiology and pathophysiology. In tandem, studies using either generation of transgenic mice that overexpress HO-1 at specific sites in the kidney or gene transfer of HO-1 to tissue-specific sites will greatly expand this area of investigation. Whole animal experiments, as well as isolated cells grown from different segments of the nephron from these animals, could be used in models of renal injury. Ultrastructural localization of the different isoforms of heme oxygenase along the nephron and functional correlation in the intact kidney and in renal injury will be important. The beneficial effects of products of the HO-1 catalyzed reaction merit further investigation. In this regard, the physiologic role of carbon monoxide, a potentially poisonous gas, needs clear elucidation. Recent studies demonstrate that like NO, low concentrations of CO are protective, whereas higher concentrations are cytotoxic (14). The hemodynamic effects of CO, mediated via cGMP, in the renal microcirulation and other vascular beds are promising (70). Major areas of research in which HO-1 will play a significant role in the near future will be transplantation, inflammation, and the cell biology of acute renal failure (ischemic and nephrotoxic). Other areas related to nephrology include the role of HO-1 in hypertension, diabetic nephropathy, obstructive uropathy, and polycystic kidney disease.
Conclusion
In summary, the induction of HO-1 plays an important role in many clinically relevant disorders, including those that involve the kidney. The availability of the HO-1 knockout mouse has served as a useful tool to further confirm the results in renal injury models that were derived mainly from studies using chemical inducers and inhibitors of HO-1. These observations provide impetus for studies aimed at the development of novel physiologically relevant inducers of the endogenous HO-1 gene or targeted gene overexpression of HO-1 as a therapeutic and preventive modality in pathophysiologic states. Evaluation of the by-products of the HO-1 catalyzed reaction such as iron, biliverdin, bilirubin, and carbon monoxide in mediating the cytoprotective effects will provide further elucidation on the mechanism of cytoprotection. At least one by-product, carbon monoxide, has received considerable attention in recent years. As noted in a research news article from Science, “...this gas is likely to provide fuel to run plenty of labs” (90).
Acknowledgments
This work was supported by grants from the Juvenile Diabetes Foundation International Fellowship Award to Dr. Agarwal in Dr. Nath's laboratory at the University of Minnesota; National Institutes of Health (NIH) Grants K08 DK02446 and R03 DK56279 to Dr. Agarwal; and NIH Grant HL39593 to Dr. Nick. We gratefully acknowledge the contribution of Dr. Karl A. Nath in the studies presented in this review.
Footnotes
American Society of Nephrology
- © 2000 American Society of Nephrology