INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HEME OXYGENASES (HO) catalyze the rate-limiting step in the degradation of heme to yield equimolar amounts of biliverdin, carbon monoxide (CO), and iron (13, 14, 22, 24). In mammals, biliverdin is rapidly converted to bilirubin by the enzyme biliverdin reductase. Three isoforms of HO exist; HO-1 is inducible, whereas HO-2 and HO-3 are constitutively expressed. A variety of stimuli have been shown to induce HO-1 expression in tissues, including oxidant stress, ischemia-reperfusion, endotoxin (lipopolysaccharide, LPS), cytokines, and its substrate, hemin (13, 14, 22, 24). The increased production of CO, biliverdin, and bilirubin that accompanies HO-1 induction has been implicated in different physiological (blood flow regulation) and pathological (atherosclerosis) processes. The vasodilatory properties of CO have led to its implication in the regulation of blood flow and blood pressure (22, 24), whereas the potent anti-oxidant properties of both biliverdin and bilirubin have resulted in the proposal that HO-1 induction is an important protective mechanism during periods of oxidative stress (20, 23, 29).

There is a growing body of evidence that ascribes an anti-inflammatory role for the products of HO-1 activation. Several lines of evidence support this contention as follows: 1) acute and chronic inflammatory foci are associated with intense HO-1 expression (16, 27), 2) drugs that elevate HO-1 activity (e.g., hemin) suppress the inflammatory response, whereas inhibitors of HO (e.g., metalloporphyrins) potentiate it (16, 27), 3) exogenous administration of HO-1 by gene transfer protects the rat lung against hyperoxia-induced neutrophil infiltration and tissue injury (17), and 4) HO-1-deficient mice (21) and humans (28) are characterized by chronic inflammation and extreme sensitivity to oxidative stress. Although the mechanisms underlying the anti-inflammatory actions of HO-1 remain poorly defined, both CO and biliverdin/bilirubin have been implicated in this response. Some investigators have explained the anti-inflammatory effect of HO-1 on the ability of CO to suppress the production of cytokines, whereas others have attributed this action to the potent antioxidant properties of both biliverdin and bilirubin (6, 20, 22, 24).

The adhesion of leukocytes to vascular endothelial cells is a rate-determining step in the recruitment of leukocytes in inflamed tissue. Each of the three stages of leukocyte recruitment, i.e., leukocyte rolling, firm adherence, and transendothelial migration, involves the participation of different families of adhesion molecules, including the selectins, ₂-integrins, and supergene immunoglobulins (19). P- and E-selectin are lectin-like adhesion glycoproteins that are expressed on the surface of activated endothelial cells, where they mediate the low-affinity adhesive interaction that is manifested as leukocyte rolling. Because rolling is a prerequisite for the subsequent firm adhesion and transendothelial emigration of leukocytes in postcapillary venules, considerable effort has been directed toward defining the factors that regulate the expression of P- and E-selectin on vascular endothelial cells. One factor that has been shown to exert a significant modulating influence on leukocyte rolling and the expression of endothelial P-selectin is nitric oxide (NO; see Ref. 5). Pharmacological agents that increase NO availability (e.g., NO-donating compounds) attenuate P-selectin expression and leukocyte rolling, whereas inhibitors of NO production (e.g., NO synthase inhibitors) exacerbate these responses in postcapillary venules (2, 3, 5).

The overall objective of the present study was to determine whether CO, like its gaseous monoxide counterpart NO, can exert a modulating influence on the inflammatory response by altering the expression of selectins on activated endothelial cells. This objective was addressed by determining whether the increased expression of P- and E-selectin elicited by LPS in different regional vascular beds is modified by agents that specifically induce (hemin) or suppress (metalloporphyrin) the expression of HO-1. To address the contribution of CO vs. biliverdin to the anti-inflammatory actions of HO-1, we also examined the influence of exogenously administered biliverdin on selectin expression in this model of acute inflammation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Monoclonal antibodies. The monoclonal antibodies (MAbs) used for in vivo characterization of P- and E-selectin expression were RMP-1, a murine IgG₂ MAb against rat P-selectin (26), RME-1, a murine IgG₁ MAb against rat E-selectin (25), and P-23, a nonbinding murine IgG₁ directed against human P-selectin (11). The blocking properties of RMP-1 and RME-1 have been previously demonstrated in different rat models of inflammation (7). All of the antibodies were provided by Dr. Donald C. Anderson from Pharmacia & UpJohn (Kalamazoo, MI).

Radioiodination of MAb. The binding (RMP-1 and RME-1) and nonbinding (P-23) MAbs were labeled with ¹²⁵I and ¹³¹I (NEN, Boston, MA), respectively, using the Iodogen method (4). In brief, Iodogen (Sigma T-0656) was dissolved in chloroform at a concentration of 0.5 mg/ml, and 250 µl of this solution were placed in glass tubes and evaporated under nitrogen. A 250-µg sample of MAb was added to each Iodogen-coated tube, and either ¹²⁵I or ¹³¹I with a total activity of 250 µCi was added. The mixture was incubated on ice, with periodic stirring for 10 min. The total volume was brought to 2.5 ml by adding PBS (pH = 7.4). After radioiodination, the coupled MAb was separated from free ¹²⁵I or ¹³¹I by gel filtration on a Sephadex PD-10 column (Pharmacia Biotech). The column was equilibrated with PBS containing 1% BSA and was eluted with the same buffer. Two fractions of 2.5 ml were collected, the second of which contained the radiolabeled antibody. Absence of free ¹²⁵I or ¹³¹I was ensured by extensive dialysis of the protein-containing fraction. Less than 1% of the activity of the protein fraction was recovered from the dialysis fluid. Labeled MAbs were stored at 4°C.

Animal procedures. All experimental protocols were applied to 200- to 300-g male Sprague-Dawley rats (Harlan Laboratories, Frederick, MD). The experimental procedures described herein were reviewed and approved by the Institutional Animal Care and Use Committee of Louisiana State University Medical Center. Anesthesia was induced with thiobutabarbitol (Inactin; 120 mg/kg ip), and a tracheostomy was performed to facilitate breathing. The carotid artery and jugular vein were cannulated, and then a mixture of ¹²⁵I-labeled RMP-1 (¹³¹I-RMP-1) or RME-1 (10 µg) and a fixed dose (1.0-6.5 µg) of ¹³¹I-labeled nonbinding MAb (¹³¹I-P-23) was administered through the jugular vein catheter. Five minutes after injection of the MAb mixture, blood samples were obtained from the carotid artery. Immediately thereafter, the animals were heparinized (40 units heparin sodium) and rapidly exsanguinated by perfusion of bicarbonate-buffered saline (BBS) through the jugular vein catheter with simultaneous blood withdrawal through the carotid artery catheter. This was followed by perfusion of 60 ml of BBS through the carotid artery catheter after severing the inferior vena cava at the thoracic level. The small intestine (from ligament of Trietz to ileocecal junction), large intestine, liver, lung, and kidneys were harvested and weighed.

Experimental protocols. The expression of P- and E-selectin was measured in different regional vascular beds of the rat (n = 50), both under control conditions and after the administration (ip) of 10 µg/kg Salmonella abortus equi LPS (Sigma Chemical, St. Louis, MO). On the basis of previous work that examined the kinetics of P- and E-selectin expression after LPS stimulation (4), P-selectin expression was measured at 4 h after LPS administration, whereas E-selectin expression was determined at 3 h after LPS. To assess the influence of HO-1 induction on LPS-induced expression of the selectins, some rats (n = 5/group) were pretreated with hemin [molecular weight (MW) = 652, 40 µmol/kg ip] 18-24 h before LPS injection (16). In additional experiments (n = 5 rats/group), rats were pretreated (45 µmol/kg ip) with the potent HO-1 inhibitor zinc protoporphyrin (ZnPP, MW = 627; see Ref. 24) 18-24 h before LPS administration. RT-PCR was used to confirm that both hemin and ZnPP exerted the desired effect on HO-1 mRNA. Although LPS treatment resulted in detectable tissue levels of HO-1 mRNA, this response was greatly increased in hemin-treated rats, whereas ZnPP-treated rats exhibited a greatly reduced HO-1 transcript level.

Statistics. The data were analyzed on StatView software using a one-way ANOVA and Student's unpaired t-test. Statistical significance was set at P < 0.05. Values are means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The changes in P-selectin expression induced by LPS in the different experimental groups are summarized in Figs. 1 (lung and kidney), 2 (liver), and 3 (small and large intestines). In all tissues studied, significant basal (constitutive) expression of P-selectin was detected. At 4 h after LPS administration, highly significant increases (11.8-, 9.3-, 26-, and 11.0-fold for lung, kidney, liver, and small intestine, respectively) in P-selectin expression were noted, as previously described for the same tissues in the mouse (4).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Effects of endotoxin (lipopolysaccharide, LPS) on P-selectin expression in the lung and kidney of untreated animals and animals pretreated with either hemin, zinc protoporphyrin (ZnPP), or biliverdin. * P < 0.05 relative to constitutive expression.  P < 0.05 relative to LPS alone.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Effects of LPS on P-selectin expression in the liver of untreated animals and animals pretreated with either hemin, ZnPP, or biliverdin. * P < 0.05 relative to constitutive expression.  P < 0.05 relative to LPS alone.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Effects of LPS on P-selectin expression in the small and large intestines of untreated animals and animals pretreated with either hemin, ZnPP, or biliverdin. * P < 0.05 relative to constitutive expression.  P < 0.05 relative to LPS alone.

Pretreatment with hemin, an HO-1 inducer, significantly blunted LPS-induced P-selectin expression in all tissues studied. However, the attenuating effect of hemin on P-selectin expression was much more pronounced in lung, kidney, and liver, which exhibited a 70-86% reduction in the LPS-induced response compared with the 41-45% reduction noted in small and large intestine.

In contrast to the responses noted with hemin, pretreatment with ZnPP, which inhibits the induction of HO-1, significantly exacerbated LPS-induced P-selectin expression in all organs. In lung, kidney, liver, and small intestine, the increment in P-selectin expression induced by LPS was 2.2-, 0.90-, 2.4-, and 2.7-fold higher in the ZnPP-treated rats compared with untreated rats.

Biliverdin pretreatment exerted a level of protection against LPS-induced P-selectin expression that was nearly identical to that observed in animals treated with an equimolar dose of hemin. The liver appeared to respond most dramatically to biliverdin administration.

Figure 4 summarizes the responses of E-selectin expression to LPS administration in the different experimental groups for the kidney. The pattern of responses noted for E-selectin was essentially identical to that observed for P-selectin. LPS elicited a pronounced increase in E-selectin expression. This response was significantly blunted (and to a comparable degree) by pretreatment with either hemin or biliverdin, whereas ZnPP greatly enhanced LPS-induced E-selectin expression. The responses noted in the lung and intestines were similar to those shown for the kidney. Reproducible measurements of E-selectin expression were not achieved for the liver using the E-selectin MAb RME-1.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Effects of LPS on E-selectin expression in kidneys of untreated animals and animals pretreated with either hemin, ZnPP, or biliverdin. * P < 0.05 relative to constitutive expression.  P < 0.05 relative to LPS alone.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HO has been implicated in the pathogenesis of several models of acute and chronic inflammation. These models are generally characterized by an elevated expression of HO-1 in different cellular elements (endothelial cells and vascular smooth muscle) of the vascular wall, which is accompanied by the recruitment of activated leukocytes (neutrophils or lymphocytes; see Refs. 16 and 27). The elevated activity of HO-1 is considered to be a protective mechanism that limits the number and level of activation of inflammatory cells and the extent of cellular necrosis in diseased tissues (16, 17, 21, 27). This contention is supported by reports describing exaggerated inflammatory responses in mice that are genetically deficient in HO-1 (21) and by studies demonstrating that drug-induced HO-1 induction suppresses, while inhibition of HO-1 exacerbates, leukocyte accumulation and tissue injury (27). Although these observations implicate HO-1 in the recruitment of leukocytes in inflamed tissue, the molecular and cellular basis for this anti-inflammatory action remains poorly defined.

The results of the present study indicate that a potentially important anti-inflammatory effect of HO-1 is inhibition of expression of selectins on activated endothelial cells. Both P- and E-selectin have been implicated in the recruitment of rolling leukocytes in inflamed postcapillary venules (19) and in different models of inflammation (7). Although the function of these adhesion molecules appears to be somewhat redundant, several studies have demonstrated that simultaneous inhibition of both endothelial selectins is highly effective in reducing the recruitment of rolling, firmly adherent, and emigrating leukocytes in inflamed microvessels (8, 10). Hence, our observation that HO-1 activity has a profound effect on the expression of both endothelial selectins should have significant implications relative to leukocyte recruitment in inflamed tissues.

Hemin, a naturally occurring substrate for HO, has been previously shown to be a potent inducer of HO-1 (13, 14, 22, 24), whereas metalloporphyrins, like ZnPP, are known to act as potent inhibitors of HO-1 induction (12). Consequently, these reagents have been widely used to manipulate tissue levels of HO-1 expression in experimental animals (13, 14, 22, 24). The effectiveness of these modulators of HO-1 activity was confirmed in the present study using RT-PCR, which demonstrated that the increased HO-1 expression normally elicited by LPS is further increased in animals receiving hemin and greatly reduced after ZnPP treatment. However, these observations do not exclude the possibility that some of the actions of hemin and ZnPP are exerted on the constitutive isoforms (HO-2 and HO-3) of HO. The liver, lung, and kidney (and to a much lesser extent intestines) normally express significant HO-2 activity. Furthermore, vascular cells express both the HO-1 and HO-2 (but not HO-3) isoforms of HO (13, 14, 22, 24). It should be noted, however, that inducers of HO-1 can increase the activity of this isoform to levels well above that of constitutively expressed HO-2 and that HO-1 appears to be two times as reactive to some substrates as HO-2 (13).

Both CO and biliverdin (or its metabolite, bilirubin) have been proposed as mediators of the beneficial effects of HO-1 activation (20). CO, which relaxes vascular smooth muscle by activating guanylate cyclase, may exert a beneficial effect by increasing blood flow (22, 24). This mechanism appears to be particularly important in the liver microcirculation (24). Biliverdin and bilirubin, on the other hand, have been shown to act as potent antioxidants (20, 23, 29). Hence, these HO-1 products are considered potentially important protective molecules in conditions associated with an oxidative stress, such as inflammation, atherosclerosis, and ischemia-reperfusion. To shed some light on which product of HO-1 activation is responsible for the inhibitory effect of hemin on LPS-induced P- and E-selectin expression, we compared the selectin expression responses to LPS in animals treated with equimolar quantities of either hemin or biliverdin. Our findings indicate that biliverdin is as effective as hemin in attenuating the LPS-induced expression of endothelial selectins. This observation suggests that biliverdin, per se, or its subsequent metabolism to bilirubin may be more important than CO production in mediating the beneficial anti-inflammatory effects of HO-1 in our model of LPS-induced selectin upregulation. The possibility that biliverdin attenuates LPS-induced endothelial cell expression of P- and E-selectin via an antioxidant effect appears tenable in view of the documented role of oxidants in regulating the expression of these two adhesion molecules (1, 9, 15).

The findings of this study are strongly supported by a recent report that describes hemin-induced inhibition of oxidant-mediated leukocyte-endothelial cell adhesion in rat mesenteric venules (6). ZnPP exacerbated the oxidant-mediated leukocyte-endothelial cell adhesion, and this effect was reversed by superfusion of the mesentery with equimolar concentrations of either bilirubin or biliverdin but not with CO. The rapid peroxide-induced translocation (and surface expression) of stored P-selectin on venular endothelial cells was also inhibited by hemin. The latter findings, coupled to our results, indicate that HO-1 products can inhibit both the translocation of preformed P-selectin and the transcription-dependent biosynthesis of P- and E-selectin.

In conclusion, the results of this study indicate that HO can exert a significant regulatory influence on the expression of P-selectin and E-selectin in different regional vascular beds. Biliverdin (or its metabolite, bilirubin) may account for this action of HO-1 on endothelial cell adhesion molecule expression. These findings may provide a molecular basis for the previously reported anti-inflammatory properties of HO-1.