Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The vital role of heme in mammalian physiology is attested to by its function as the prosthetic group of a variety of important hemeproteins that are essential for various cellular processes. Heme is involved in electron transfer as the prosthetic group of cytochromes, in the enzymatic decomposition of H₂O₂ as the prosthetic group of catalase and peroxidase, in the generation of nitric oxide as the prosthetic group of nitric oxide synthase, and in the manufacture of vasoactive eicosanoids as the prosthetic group of cytochrome P-450 (CYP) and cyclooxygenase (Kappas and Drummond, 1986; Abraham et al., 1996). Cellular levels of heme are regulated by its rate of synthesis and degradation. Heme catabolism occurs by oxidative cleavage of the -methene bridge of the molecule, eventually leading to the formation of equimolar amounts of bilirubin and carbon monoxide. The heme oxygenase (HO) system is the rate-limiting step in heme degradation. To date, three HO isoforms (HO-1, HO-2, and HO-3) have been identified that catalyze this reaction (Abraham et al., 1987; Maines, 1988; McCoubrey et al., 1992; Shibahara et al., 1993). HO-1 is a 32-kDa heat shock protein (Mitani et al., 1989; Dwyer et al., 1992) that is inducible by numerous noxious stimuli (Maines and Kappas, 1974; Lutton et al., 1992; Lu et al., 1998). HO-2 is a constitutively synthesized 36-kDa protein that is abundant in brain and testis (Maines, 1988). HO-3 is related to HO-2 but is the product of a different gene, and its ability to catalyze heme degradation is much less than that of HO-2 (McCoubrey et al., 1997). Increasing data reveal that HO and its metabolic products carbon monoxide and bilirubin play important roles in numerous biological processes, especially in endothelial cell vasodilating responses and in cellular antioxidative reactions (Stocker et al., 1987; Ingi and Ronnett, 1995; Suttner et al., 1999), and evidence that the suppression of HO-1 may not lead to cell protection also is accumulating (da Silva et al., 1996; Suttner et al., 1999). The reasons for this paradox are not clear but the evidence shows that HO can mediate both cell protection and cell injury, depending on experimental conditions (Nutter et al., 1994; Kadoya et al., 1995; da Silva et al., 1996). Carbon monoxide produced by HO-1 and HO-2 has been implicated as a physiological regulator of cGMP and vascular tone (Ishizaka et al., 1994; Morita et al., 1995; Durante et al., 1997; Sammut et al., 1998; Yang et al., 1999). We and others have demonstrated that induction of epithelial HO-1 in an animal model of contact lens-induced hypoxic injury greatly attenuates corneal surface inflammatory responses, and a similar effect is observed in an experimental model of inflammation in rats (Willis et al., 1996; Laniado-Schwartzman et al., 1997).

We have previously shown that overexpression of HO-1 in coronary endothelial cells results in resistance to hemoglobin-induced injury (Abraham et al., 1995b). Recent studies also have shown that transfection of rat HO-1 cDNA to human pulmonary epithelial cells can increase cell viability in hyperoxia (Choi et al., 1995; Suttner et al., 1999). Moderate overexpression of HO-1 provides protection against oxidative injury (Suttner et al., 1999). Conversely, deficient HO expression in mammalian cells contributes to reduced stress defense (Poss and Tonegawa, 1997b; Yachie et al., 1999). More recently, Hancock et al. (1998) have shown that HO-1 induction prevents oxidant-stressed endothelial up-regulation of adhesion molecules and development of transplant arteriosclerosis in normal mice, suggesting that a moderate increase in HO activity may be beneficial in resistance to oxidants. In contrast, selective inhibition of HO-1 gene expression with antisense oligonucleotides is associated with enhancement of adhesion molecule expression in endothelial cells and with inflammation (Wagener et al., 1999).

The present study was designed to determine whether the human HO-1 gene could be introduced into various rat tissues by i.v. administration of a recombinant replication-deficient adenoviral human HO-1 cDNA construct. Furthermore, we assessed the functional expression of the human HO-1 gene on HO-1 mRNA, and protein and cellular heme content and CYP-dependent arachidonic acid metabolism. The results reported herein demonstrate that administration of an adenoviral-mediated human HO-1 gene vector in adult rats results in a transient increase in human HO-1 mRNA production and HO-1 protein and an increase in total HO enzyme activity. Transduction of the HO-1 gene in vivo was associated with a decrease in the activity of one of the heme-dependent CYP epoxygenases. These data thus directly demonstrate that an in vivo increase in human HO-1 gene expression would be expected to decrease cellular free heme concentration, with a consequent decrease in CYP activities that are known to participate in generation of inflammatory mediators.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of Adenoviral Vector Containing Human HO-1 cDNA. We constructed the human HO-1 cDNA with adenovirus vector with a mammalian transfection kit from Stratagene (La Jolla, CA) as previously described (Abraham et al., 1995a). Briefly, the human cDNA plasmid was constructed as follows: an Xhol-Xbal fragment of a human HO-1 expression vector pRc/CMVHO-1 was cloned between Xhol and Xbal of the pBacPAC8 to generate pBacPAC8human HO-1. The Bg1II-BamHI fragment of the pBacPAC8HHO was then introduced into the Bg1II site of pAdBg1II. A human embryonic cell line, 293 cells (no. 1573-CRL; American Type Culture Collection, Manassas, VA), was cotransfected with 10 µg of EcoRI-digested adenoviral human HO-1 and 1 µg of ClaI-digested dL7001DNA by calcium-phosphate coprecipitation with a mammalian transfection kit. Human HO-1 adenovirus construct was replicated and encapsulated into an infectious virus. After a 5-day incubation period, the virus plaque locations were marked on the flasks, and the resultant cytopathic effect on the monolayers was observed microscopically until the plaque reached an adequate size. The plaques were purified and checked for the presence of human HO-1 by polymerase chain reaction (PCR) with HO-1 specific primer, and amplified by propagation in the 293 cell line. HO-1 adenovirus was released and collected from infected cells 2 days after infection with rupture of the cells by freezing and thawing several times (usually three times) and concentrated by centrifugation with Ultrafree-MC filters (Millipore, Bedford, MA). The virus titers were determined by plaque assay with 293 cells. The virus was stored at 80°C until use.

In Vivo Gene Transfer. Sprague-Dawley male rats (150-180 g) purchased from Taconic Farms (Germantown, NY) were injected i.v. with human HO-1-adenovirus suspension (10¹² pfu/ml/animal) in the tail vein. The control group was injected with an equal volume of saline. The animals were housed in The Rockefeller University Laboratory Animal Research Center in a controlled environment with a 12-h light/dark cycle. The animals were acclimatized for 1 week before the start of the experiment. A minimum of three animals was used for each data point and animals were sacrificed at the times indicated on the legends to the figures and tables. CoCl₂ (Mallinckrodt, St. Louis, MO) was dissolved in saline and administered subcutaneously in the nuchal region at a dose of 250 µmol/kg b.wt. 16 h before sacrifice. Control animals were administered an equivalent volume of saline.

RNA Extraction and Reverse Transcription (RT)-PCR. The rat tissues were ground in liquid nitrogen and total RNA was extracted with guanidine thiocyanate, first purified with phenol-chloroform, followed by an additional purification with chloroform, and then precipitated with isopropanol. Total RNA was resuspended in autoclaved diethyl pyrocarbonate-water, and quantitated by absorbency measurement with a Beckman DU7400 spectrophotometer. To verify the quality of RNA and analyze the RNA, 10 µg of total RNA was denatured and size-separated by electrophoresis on a 1% agarose gel containing 2.2 M formaldehyde. The integrity of the samples was checked with ethidium bromide staining solution under UV light.

Quantitation of Human HO-1 mRNA Copies. The number of human HO-1 mRNA copies was measured in rat tissues after adenoviral-mediated gene transfer according to published methods (Goodman et al., 1996; Abraham, 1998). Briefly, the plasmid pCMV-human HO-1 was linearized with Eco47III (a single restriction site in human HO-1 cDNA at position 420 bp). The linearized plasmid was digested with Bal 31 nuclease for 5 min to digest nucleotides and create a mutation. After this digestion, the construct was ligated with T4 DNA ligase. After transformation into JM109 Escherichia coli competent bacteria, several clones were analyzed for the size of the mutated human HO-1 cDNA that resulted from the digestion by Bal 31 and for the presence of the Eco47 III restriction site. One clone was selected with ~50 bp truncated from the original cDNA of human HO-1. This clone was amplified and the insert, excised from the vector with HindIII, was purified. The amplified products were used as an internal standard in the competitive versus noncompetitive amplification of the human HO-1 mRNA. Various amounts of total RNA from human kidney were subjected to RT-PCR alone or as a mixture with a fixed amount of the mutated insert, which is 50 bp shorter than the amplified human HO-1 mRNA PCR products. As the amount of one template increased, the chance of the other template being amplified declined (eventually, the portion of the other template became too small to be significant). We have previously shown that when the amount of internal standard was included in the RT-PCR, the amount of amplified human template decreased (Abraham, 1998). Because the internal standard is generated from the human HO-1 gene and exhibits 100% homology to human HO-1 mRNA, both targets, human HO-1 mRNA and the internal standard, could be amplified with an equivalent efficiency. Respective templates are identified by their differences in size after electrophoresis on acrylamide gel and are quantitified by densitometry or the radioactivity count after the bands were excised from the gel (Abraham, 1998).

HO Activity Measurement. Enzyme activity was measured with microsomes from nontransfected and transfected rat tissues. Microsomes were prepared as previously described. Microsomes were incubated with hemin, rat liver cytosol, MgCl₂, glucose-6-phosphate dehydrogenase, glucose 6-phosphate, and NADP⁺ in potassium phosphate buffer (0.1 M, pH 7.4) for 10 min at 37°C. Placing the tubes on ice stopped the reaction. The amount of bilirubin generated was estimated with an extinction coefficient of 40 mM/cm as previously described (Abraham et al., 1987).

CYP-Arachidonic Acid Metabolism Assay. CYP, arachidonic acid metabolism, and heme were measured as previously described (Fuhrop and Smith, 1975; Schwartzman et al., 1990; da Silva et al., 1994). Briefly, microsomes prepared from renal tissues (150 µg) were preincubated with [1-¹⁴C]arachidonic acid (0.4 µCi; 7 nmol) in 100 mM potassium phosphate buffer, pH 7.4, containing 10 mM MgCl₂ for 3 min at 37°C. NADPH (1 mM) was added and the reaction mixture, with a final volume of 0.3 ml, was incubated for 30 min at 37°C. The reaction was terminated by acidification to pH 3.5 to 4.0 with 2 M formic acid, and metabolites were extracted with ethyl acetate. The final extract was evaporated under nitrogen, resuspended in 50 µl of methanol, and injected into the HPLC column. Reversed-phase HPLC was performed on a 5-µm ODS-Hypersil column, 4.6 × 200 mm (Hewlett Packard, Palo Alto, CA) with a linear gradient ranging from acetonitrile/water/acetic acid (50:50:0.1) to acetonitrile/acetic acid (100:0.1) at a flow rate of 1 ml/min for 30 min. The elution profile of the radioactive products was monitored by a flow detector (In/us System Inc., Tampa, FL). The identity of epoxide and hydroxy-6,8,11,14-eicosatetraenoic acid was confirmed by its comigration with an authentic standard (Schwartzman and Abraham, 1990; Schwartzman et al., 1990).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Adenovirus-Mediated Expression of Human HO-1 mRNA in 293 Cells and in Rat Tissues. Figure 1 represents the PCR products of the adenoviral human HO-1 construct in culture media of 293 cells transfected with the adenoviral human HO-1 construct. Lane 1 indicates the absence of a hybridizable band in nontransduced 293 cells. The culture media of the 293 cell line transfected with adenoviral human HO-1 (lane 2) gave a positive signal when amplified with human HO-1 primer, with a predicted size of 555 bp. The adenovirus without the human HO-1 gene (lane 3) gave no signal. The human HO plasmid is shown in lane 4 as a positive control.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   PCR products obtained from the amplification of adenovirus. Lane 1, culture media of 293 cell without adenovirus human HO-1 gene transfection; lane 2, culture media of 293 cell transfected with adenovirus human HO-1 gene; lane 3, adenovirus without human HO-1 gene; and lane 4, human HO-1 plasmid as a positive control.

Quantitation of Human HO-1 mRNA after Adenovirus Gene Transfer. To quantitate the total amount of human HO-1 mRNA generated in rat tissues after adenoviral vector human HO-1 gene transfer, various amounts of total RNA from rat kidney (containing human HO-1 mRNA) were subjected to RT-PCR alone or as a mixture with a fixed amount of the internal standard. In the combined PCR, internal standard is added at the PCR step, and the RT is performed with oligo (dT) 18 primers as described in Materials and Methods. Respective templates are identified by their differences in size (Fig. 2). As seen in Fig. 2, an internal standard of human HO-1 cDNA at a concentration of 20 fg was amplified to maximum levels when rat RNA, which contains human RNA, was at 10 ng (lane 1). As rat RNA was increased, the amount of the competitive internal standard amplified decreased (Fig. 2, lanes 2 and 3, respectively). When rat renal RNA level was 150 ng (lane 4) the amount of the internal standard and human HO-1 RNA within rat RNA was amplified equally. Because the internal standard, 20 fg of HO-1 RNA, contains 500 bp consisting of 3.64 × 10⁴ molecules, the amount of human HO-1 mRNA present in 150 ng total rat RNA was equivalent to 242 mRNA copies. The average amount of human HO-1 mRNA present in 150 ng of RNA was 320 ± 125 mRNA copies (n = 3). It has been previously shown that the number of normal human liver and kidney HO-1 molecules is 750 ± 980 and 1100 ± 1300 molecules/µg RNA, respectively. Therefore, adenoviral-mediated HO-1 gene transfer into the rat resulted in a generation of ~50% of the basal levels of human HO-1 mRNA in rat tissues with a significant increase in the basal activity of endogenous HO activity. Analogously, it has been shown that administration of CoCl₂ at a dose of 250 µmol/kg into rat kidney resulted in elevation of endogenous rat renal HO-1 ~20-fold, with a reciprocal decrease in heme-protein content (Sacerdoti et al., 1989).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Competitive amplification of rat kidney total RNA and mutated insert (mhuman HO-1). Internal standard (mhuman HO-1) at 10 fg was mixed with rat kidney total RNA at 250, 150, 75, 50, and 25 ng, respectively (lanes 1-5).

Evaluation of Human HO-1 Gene in Rat Tissue. Once the method was established, we examined the expression of endogenous HO-1 in rat tissues from animals administered adenoviral human HO-1. Rats were injected with adenoviral human HO-1 vector, and the presence of the human HO-1 gene was examined with respect to time in the brain, heart, lung, liver, and kidney of the animals. The ability of adenovirus to enter rat brain tissue led us to evaluate the possibility that expression of human HO-1 mRNA in this tissue was due to contamination from blood. We examined both blood cells and plasma over the course of the experiment for the presence of human HO-1 mRNA. No human HO-1 mRNA was detectable (data not shown). Positive signals were obtained for the human HO-1 gene when amplified with human HO-1 primers at 1 day, and 1, 2, and 4 weeks, in all tissues examined after injection of adenoviral-human HO construct (Table 1). There was no evidence of human HO-1 in rat tissues 8 weeks after human HO-1 gene transfer (data not shown). The highest tissue levels of human HO-1, in lung and liver, were obtained 2 weeks after human HO-1 gene transfer and declined with time (Table 1). The levels of human HO-1 mRNA were increased after 1 day (Fig. 3, lane 2). The strongest human HO-1 signal was obtained in the kidney 1 to 2 weeks after gene transfer (Fig. 3, lanes 4 and 5, respectively). Administration of adenovirus did not result in up-regulation of endogenous rat HO-1 or GAPDH mRNA (Fig. 3, middle and bottom, respectively).

View larger version (54K): [in this window] [in a new window]   Fig. 3.   Effect of adenovirus-mediated human HO-1 gene transfer on expression of human HO-1 and rat HO-1 mRNA in rat kidney by RT-PCR. Rats were injected with adenovirus-mediated human HO-1 gene (1012 pfu/ml/animal) and sacrificed after 1 day; and 1, 2, and 4 weeks. Lane 1, control rat (injected with empty adenovirus); lane 2, 1 day after human HO-1 gene transfer; lane 3, 4 days after human HO-1 gene transfer; lane 4, 1 week after human HO-1 gene transfer; and lane 5, 2 weeks after human HO-1 gene transfer.

Enhancement of HO Activity in Rat Liver and Kidney Injected with Adenoviral Human HO-1 Gene Construct. We examined HO enzymatic activity in transfected kidney and liver tissues obtained from rats pretreated with adenoviral human HO-1. Rats administered empty virus served as controls; the results are depicted in Table 2. The basal levels of HO activity in rat liver and kidney were 0.93 ± 0.2 and 0.79 ± .08 nmol bilirubin/mg/h, respectively. Interestingly, HO activity in the kidney was significantly up-regulated 1 day after adenoviral administration (P < .05). HO activity had returned to the basal levels 4 weeks after Adv-HO-1 gene administration. HO activity was significantly increased after 1 day of adenoviral human HO-1 vector administration and was increased by 1.5- and 2.5-fold in liver and kidney, respectively, after 1 and 2 weeks of adenoviral human HO-1 administration.

Effect of Adenoviral Human HO-1 Gene Construct on Induction of Rat Endogenous HO Activity and mRNA. To further ascertain the effect of administration of the adenovirus-mediated human HO-1 gene on the endogenous rat HO-1 gene, we examined the effect of CoCl₂ on rat HO activity, and on microsomal protein and rat HO-1 mRNA in total RNA. Rats were administered adenoviral human HO-1 gene at a concentration of 10¹² plaque-forming units (pfu)/ml/animal, and 7 and 14 days later CoCl₂ was administered at a dose of 250 µmol/kg. The animals were sacrificed 16 h after CoCl₂ administration and HO activity was measured in the rat kidney; the results are described in Table 3. There was an increase in human HO-1 mRNA in rat kidney and other tissues examined similar to that seen in Table 1 (data not shown). Administration of CoCl₂ to rats receiving empty virus or rats receiving the human HO-1 gene responded in a similar manner. CoCl₂ administration increased HO activity from 0.79 ± 0.07 to 2.71 ± 0.50 nmol bilirubin/ng/h. There was an additive increase in HO activity in rats receiving CoCl₂ and adenoviral human HO-1 gene compared with rats receiving human HO-1 gene or rats receiving CoCl₂. This is due to CoCl₂ activation of rat HO-1 promoter and enhancement of rat HO-1 mRNA. Because the adenoviral human HO-1 gene construct lacks the human HO-1 promoter, CoCl₂ was unable to activate the human HO-1 gene; therefore, administration of CoCl₂ did not increase HO activity more than an additive degree. Similarly, in the 2-week experiments, CoCl₂ and the adenoviral human HO-1 construct increased HO activity to 2.71 ± 0.5 and 1.57 ± 0.09 nmol bilirubin/ng/h, respectively. The increase in expression of rat HO activity was associated with an ~20-fold increase in rat HO-1 mRNA (Fig. 4, lane 4). Administration of CoCl₂ to transgenic rats with human HO-1 produced a response to a similar degree as in rats that receiving control vector (lanes 5 and 6). These results show that administration of adenovirus-mediated human HO-1 gene to rat tissues such as kidney and brain did not suppress the expression of endogenous HO, and that CoCl₂, enhanced endogenous rat HO-1 without activation of the promoterless human HO-1 gene in the adenoviral construct.

View larger version (73K): [in this window] [in a new window]   Fig. 4.   Effect of adenovirus-mediated human HO-1 gene transfer with CoCl2-mediated induction of human and endogenous rat HO-1 mRNA in rat kidney by RT-PCR. After 1 and 2 weeks of injecting with adenovirus-mediated human HO-1 gene, rats were injected with CoCl2, and after 16 h of CoCl2 injection, rats were sacrificed. Lane 1, control rat (injected with empty adenovirus); lane 2, 1 week after human HO-1 gene transfer; lane 3, 2 weeks after human HO-1 gene transfer; lane 4, CoCl2 induction without adenovirus-mediated human HO-1 gene transfer; lane 5, 1 week after human HO-1 gene transfer with CoCl2 induction; and lane 6, 2 weeks after human HO-1 gene transfer with CoCl2 induction.

Effect of Adenoviral Human HO-1 Administration on CYP-Arachidonic Acid Metabolism in Rats. One of the means by which the activity and levels of CYP enzymes can be manipulated is by decreasing heme availability (Levere et al., 1990). We therefore studied the effect of adenoviral human HO-1 vectors on CYP content and activity. We examined whether elevation of HO activity after adenoviral gene transfer modulated heme-dependent enzyme proteins such as CYP content and activity. Two weeks after adenovirus-mediated HO-1 gene transfer, we measured CYP content and CYP activity as determined by arachidonic acid metabolites. The basal level of CYP content in the kidney was 652 ± 174 pmol/mg protein (mean ± S.D; n = 3); 14 days after adenoviral human HO-1 administration, the level of CYP had decreased to 510 ± 129 pmol/mg protein (mean ± S.D; n = 3; P < .05). Because CYP isozymes are involved in the metabolism of many drugs and chemicals, including arachidonic acid, we further investigated the question of whether a correlation could be made between adenoviral-mediated HO-1 gene expression and one of the CYP-dependent arachidonic acid metabolic enzymes, namely, the CYP enzyme epoxygenases. Arachidonic acid is oxygenated by renal cortical CYP to the epoxygenase product 11,12-epoxyeicosatrienoic acid and its hydrolytic metabolite 11,12-dihydroxyeicosatrienoic acid. The levels of CYP epoxygenase as measured by the production of 11,12-epoxyeicosatrienoic acid were decreased by 35% in rats transduced with human HO-1 gene (Fig. 5). It is clear that the adenoviral-mediated human HO-1 gene transfer resulted in a significant decrease in a heme-CYP activity in renal tissues.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Adenovirus-mediated human HO-1 gene was administered to rats at a dose of 1012 pfu/ml/animal) i.v. in the tail vein for 2 weeks. Kidney homogenate (150 µg) from rats treated with empty virus or transferred with adenovirus-mediated human HO-1 gene were used for enzyme assay. CYP-related arachidonic acid metabolism were measured by incubation of renal cortical microsomes with [C14]arachidonic acid and NADPH as described in Materials and Methods. Arachidonic acid conversion to epoxides and hydroxy-6,8,11,14-eicosatetraenoic acid were calculated and results are plotted. Data are presented as the mean ± S.E. of three different determinants.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We describe in this report that overexpression of human HO-1 by gene transfer can modulate renal cortical heme content and CYP epoxygenase activity. These data establish that i.v. delivery of an adenovirus-mediated human HO-1 gene construct in rats resulted in expression of human HO-1 mRNA in several tissues (liver, kidney, heart, lung, and brain). By use of specific primers for human and rat HO-1, it was shown that 1) the transduced gene could be detected for up to 4 weeks; 2) transduction of human HO-1 to renal cortical tissues resulted in a decrease of heme-dependent CYP epoxygenase activity; and 3) the transduced human HO-1 gene increased HO activity without altering expression of the endogenous HO-1 gene or the inductive response of the endogenous rat HO-1 gene to CoCl₂, a known potent inducer of HO-1 (Maines and Kappas, 1975). Overexpression of HO-1 by inducers such as CoCl₂, SnCl₂, and heme has been associated with an increase in carbon monoxide production and a decrease in renal activity of CYP epoxygenases that synthesize vasoactive eicosanoids and has decreased blood pressure in spontaneous hypertensive rats (Schwartzman et al., 1991; Imig et al., 1994). Furthermore, Panahian et al. (1999) have reported that overexpression of HO-1 may offer protection against cerebral artery occlusion in the rat brain. In the present study, the adenoviral human HO-1 gene construct was shown to cross the blood-brain barrier, resulting in expression in brain tissue. Our findings thus raise the possibility of delivering the human HO-1 gene to the central nervous system for transient overexpression of the enzyme activity for pharmacological or other experimental purposes.

Acute-phase proteins and the acute-phase response play an important role in the inflammatory process (Baumann, 1994). Some acute-phase proteins are thought to possess anti-inflammatory activities due to their antioxidant properties; HO-1 is among these acute phase proteins and is induced within minutes by heme and other oxidative stress-inducing agents such as angiotensin II and hyperoxia (Stocker et al., 1987; Ishizaka et al., 1994; Otterbein et al., 1995; Suttner et al., 1999). The mechanism of the antioxidant effect involved derives, in part, from the fact that increased HO activity enables removal of heme, a lipid-soluble transmissible form of the potent pro-oxidant iron and the generation of the heme metabolite bilirubin, having significant antioxidant and anticomplement properties (Lleusy and Tomaro, 1994). The importance of HO-1 in the protection of cells against the detrimental effects of oxidative stress is supported by the observation that humans and mice deficient in the HO-1 gene (Poss and Tonegawa, 1997a; Yachie et al., 1999) suffer from severe pathological conditions, including endothelial cell detachment and dysfunction with subendothelial deposition of foreign material and arterial occlusion. Recently, direct involvement of HO-1 overexpression in the suppression of inflammation (Laniado-Schwartzman et al., 1997) and modulation of hypertension (da Silva et al., 1994; Motterlini et al., 1998), and the suppression of adhesion molecules (Wagener et al., 1999) also has been described. Preinduction of HO-1 inhibits monocyte transmigration induced by low-density lipoprotein (Ishikawa et al., 1997). In contrast, da Silva et al. (1996) demonstrated that preinduction of HO-1 enhanced cell resistance to oxidative stress; however, long-term expression had a deleterious effect on cell viability. Thus, whether HO-1 activity is protective or potentially deleterious is probably a function, in part, of the degree of enzyme activity.

The apparent effect of the adenoviral-mediated human HO-1 gene on CYP epoxygenase activity involved in the synthesis of epoxyeicosatrienoic acids is of note. Carbon monoxide and bilirubin, products of HO activity, have been shown to have differential effects on CYP isozymes. Carbon monoxide inhibits synthesis of CYP isozymes (Estabrook et al., 1970); in contrast, bilirubin has been shown to be an inducer of CYP 1A1 and epoxide formation via activation of the aryl hydrocarbon receptor (Sinal and Bend, 1977). The resulting divergent effects of carbon monoxide and bilirubin on CYP epoxygenase remain to be explained. Our present finding showing a reciprocal decrease in cellular heme as a result of an increase of HO-1 activity by gene transfer is in agreement with the commonly observed inverse relationship between CYP and this enzyme activity (Bissell and Hammaker, 1976; Levere et al., 1990; Schwartzman et al., 1990). Whether exaggerated HO-1 overexpression after administration of metal inducers or endotoxin (Bissell and Hammaker, 1976; Levere et al., 1989; Schwartzman et al., 1990) or moderate expression after HO-1 gene transfer as described in the present report is a major determinant of other heme-dependent CYP enzyme activities remains to be established. In summary, this study demonstrates that enhanced heme catabolism, brought about by adenoviral-mediated human HO-1 gene transfer, attenuates cellular heme content and heme-dependent enzyme activity. Furthermore, an increase in HO-1 activity elicited by gene transfer may be beneficial in the enhancement of carbon monoxide and bilirubin production; both of these substances play a significant role in renal circulation and blood flow.