Introduction

Top Abstract Introduction Methods Results Discussion References

Reperfusion of the previously ischemic myocardium, although essential for tissue survival, results in increased necrosis and reduced tissue viability (reviewed in Ambrosio and Chiariello, 1985; Werns et al., 1885; Virmani et al., 1992). Numerous mechanisms for the increase in tissue injury after reperfusion have been identified, including the generation of oxygen-derived free radicals, complement activation and the infiltration of neutrophils into the ischemic zone (Romson et al., 1983; Kilgore and Lucchesi, 1993; Langlois and Gawryl, 1988).

Neutrophils are the primary inflammatory cell type involved in the early inflammatory response, mediating tissue injury through the release of reactive oxygen intermediates and degradative proteolytic enzymes. The primary line of evidence implicating neutrophils in the pathogenesis of reperfusion injury is derived from studies showing that depletion of circulating neutrophils before reperfusion significantly reduces myocardial injury (Romson et al., 1983). The important role of the neutrophil in facilitating reperfusion injury suggests that regulation of the events mediating neutrophil influx within the previously ischemic region may be of benefit by preventing extension of tissue injury in the reperfused myocardium.

The accumulation of neutrophils within the ischemic area is a multistep process requiring: 1) recruitment of neutrophils to the site of injury; 2) "rolling" of the neutrophil along the endothelium; 3) firm adhesion of the neutrophil to the vascular endothelium; and 4) movement of neutrophils from the vasculature into the interstitial space (Butcher, 1992). This multistep process by which the neutrophil adheres to the endothelium provides numerous sites for the development of pharmacologic inhibitors of neutrophil adhesion. The early events of neutrophil adhesion (minutes) are characterized by "rolling" of the neutrophil along the endothelium, an interaction mediated partly by the endothelial adhesion molecule P-selectin (GMP-140, PADGEM) (Jutila et al., 1989; Kishimoto et al., 1989). A variety of approaches have been used to demonstrate that inhibition of selectin-mediated neutrophil adhesion within the previously ischemic tissue leads to a reduction in myocardial infarct size. Monoclonal antibodies specific for P- and L-selectin effectively reduced ischemia/reperfusion injury in felines (Ma et al., 1993). In addition, synthetic oligosaccharide derivatives of one of the primary ligands for P-selectin, sLex, have been shown to inhibit neutrophil accumulation concomitantly with a decrease in infarct size in rabbit, canine and feline models of myocardial infarction (Ma et al., 1991; Kilgore et al., 1996).

Carbohydrate-containing synthetic sLex mimetics, although retaining the ability to provide protection against reperfusion injury, are costly and difficult to synthesize. Identification of naturally occurring sLex mimetics and the determination of the optimal structural requirements are important for maximal selectin inhibition and may assist in the design of effective, readily synthesized selectin inhibitors. Relatively little is known about the structural requirements important for mediating the in vivo protective effects of synthetic sLex mimetics. In the present study, the natural product glycyrrhizin (GM1292) and its structural derivative (GM3290), both of which contain carbohydrate substituents, were found to decrease infarct size after reperfusion of the ischemic myocardium (table 1). GM1658, which does not contain carbohydrate substituents, was not effective in providing a cardioprotective effect. The results support the concept that small molecules capable of interacting with the selectins may be of value in reducing tissue injury associated with ischemia/reperfusion.

	Methods

Top Abstract Introduction Methods Results Discussion References

Guidelines for human and animal research. The procedures followed in this study were in accordance with the guidelines of the Internal Review Board of the University of Michigan and with the regulations of the US Department of Health and Human Services for the Protection of Human Research Subjects (part 46 of title 45 of the "Code of Federal Regulations," as amended) and the University of Michigan Committee on the Use and Care of Animals. Veterinary care was provided by the University of Michigan Unit for Laboratory Animal Medicine. The University of Michigan is accredited by the American Association of Accreditation of Laboratory Animal Care, and the animal care and use program conforms to the standards set by the National Institutes of Health in "The Guide for the Care and Use of Laboratory Animals" (Department of Health, Education, and Welfare publication no. [NIH] 86-23).

Preparation of compounds. GM1292 (Glycyrrhizin), GM3290 and GM1658 were provided by Glycomed Inc. (Alameda, CA). GM1292 was dissolved in sterile 0.9% sodium chloride solution. GM3290 and GM1658 were dissolved in sterile 0.9% sodium chloride solution by increasing the pH to approximately 12 followed by back-titrating to a neutral pH. The compounds were prepared immediately before use and used at a dose of 10 mg/kg/hr. Control animals received vehicle (0.9% sodium chloride, pH 7.4) only.

In vitro analysis of neutrophil adhesion: C5a-mediated up-regulation of P-selectin human umbilical vein endothelial cells. HUVECs were isolated from umbilical veins by treatment with 0.1% collagenase in Dulbecco's modified Eagle's medium (Whittaker Bioproducts, Walkersville, MD) (Jaffe, 1984). Cells were plated at 5 × 104 cells/well on gelatin-coated 96-well plates and allowed to grow to confluence at 37°C with 5% CO₂. Cells were grown in M199 medium (Whittaker Bioproducts) supplemented with 20% heat-inactivated fetal calf serum, L-glutamine (4 mM), penicillin (100 U/ml), streptomycin (100 µg/ml), 25 µg/ml endothelial cell growth supplement (Collaborative Research, Bedford, MA) and 15 U/ml bovine heparin. Cells were characterized by a cobblestone appearance and used between the first and third passages.

Isolation and fluorescent labeling of human neutrophils. Human neutrophils were isolated from citrated (3.7% sodium citrate) peripheral venous blood obtained from healthy donors. Blood was layered over Ficoll/Hypaque (Pharmacia, Uppsala, Sweden) and centrifuged at 300 × g for 30 min followed by sedimentation with 5% dextran (MW 200,000) for 40 to 60 min. Contaminating erythrocytes were removed by hypotonic lysis followed by addition of 2.7% NaCl to restore isotonicity. Purified neutrophils were washed two additional times and resuspended in cold calcium- and magnesium-free Hanks' buffer (pH 7.4, HBSS) containing 0.1% bovine serum albumin to achieve a concentration of 5 × 106 cells/ml. Cells were >90% pure with a viability of approximately 95% as determined by trypan blue exclusion.

Ability of glycomimetics to decrease P-selectin-mediated neutrophil adherence in vitro. Adherence of BCECF-labeled neutrophils to HUVEC monolayers was quantitated as described previously (Vaporciyan et al., 1993). HUVECs were plated as described above and incubated at 37°C 24 hr before use. Foreman et al. (1994) demonstrated that C5a promotes the expression of P-selectin by endothelial cells with the use of a modified whole-cell enzyme-linked immunosorbent assay. To examine the effect of the glycomimetics to inhibit P-selectin-mediated neutrophil adhesion, we induced the expression of P-selectin by incubating HUVECs with 10 nM C5a for 15 min as described previously. Monolayers were washed two times with 37°C serum-free HUVEC media followed by the addition of BCECF-labeled neutrophils at a concentration of 1.0 × 105 cells/well for 15 min (100 µl total volume). Where applicable, GM1292, GM1658 or GM3290 were added to the monolayers 5 min before addition of neutrophils followed by repeated washings with cold PBS. Nonadherent neutrophils were removed by washing monolayers with PBS four times with a 12-channel multipipetter followed by a second cycle with fresh PBS. After washing, PBS (100 µl) was added to each well and the amount of fluorescence determined with a Cytofluor 2300 fluorescent plate reader (Millipore Corp., Bedford, MA). The excitation filter was a 20 nm bandwidth filter centered at 485 nm, and the emission filter was a 25 nm bandwidth filter centered at 530 nm.

In vivo studies of myocardial ischemia/reperfusion surgical preparation. New Zealand White rabbits (2.2-2.5 kg) were anesthetized with a 3.0 mg/kg Rompin-35 mg/kg ketamine mixture i.m., followed by 90 mg/kg i.m. sodium pentobarbital. After tracheotomy and positive pressure ventilation with room air, the left jugular vein was isolated and cannulated for drug administration and blood sampling. The left carotid artery was isolated, cannulated and connected to a Stathem P23-ID transducer to monitor blood pressure. A lead II electrocardiogram was monitored throughout the experiment. A left thoracotomy and pericardiotomy were performed, followed by identifying the marginal branch of the left coronary artery. Silk suture (3-0; Deknatel, Fall River, MA) was passed behind the artery and secured against a length of polyethylene tubing for 30 min. Ischemia was confirmed by cyanosis distal to the site of occlusion. At the end of the ischemic period, reperfusion was initiated by removing the polyethylene tubing.

Experimental protocol. Four treatment groups were studied: 1) vehicle (physiologic saline, n = 11); 2) GM1292, 10 mg/kg/hr (n = 6); 3) GM3290 10 mg/kg/hr (n = 6); and 4) GM1658, 10 mg/kg/hr (n = 6). Animals underwent regional myocardial ischemia for 30 min followed by 5 hr of reperfusion. The investigational agent was administered as a bolus intravenously immediately before reperfusion and every hour as a bolus thereafter for the duration of the reperfusion period. At hourly intervals, circulating white blood cell counts were determined by a DMA H10 cell counter (Texas International Laboratories Inc., Houston, TX), and heart rate and blood pressure were recorded. At the conclusion of the 5-hr reperfusion period, animals were sacrificed and the hearts were prepared for infarct size determination or frozen in liquid nitrogen for determination of myeloperoxidase activity.

Immunocytochemical analysis of P-selectin expression. Immunocytochemical analyses of P-selectin expression in the left ventricular tissue sections were carried out with a Vectastain kit. Samples of left ventricle were removed 30 min after restoration of flow to the ischemic area. After fixation in 4% paraformaldehyde, samples were paraffin embedded and incubated with an anti-P-selectin polyclonal antibody (PB1.3; 10 µg/ml) for 45 min. After repeated washings, the slides were overlaid with a biotinylated secondary antibody for 45 min and biotin-avidin complex for 30 min. Slides were developed using 3-amino-9-ethylcarbazole as the substrate. Negative controls included incubation with nonimmune, isotype-matched antibody as the primary antibody and elimination of the primary antibody.

Determination of infarct size and area at risk. At the completion of the 5-hr reperfusion period, hearts were removed and cannulated by the aorta on a Langendorff perfusion apparatus as described previously (Kilgore and Lucchesi, 1993). The branch of the left coronary artery was ligated in a location identical with the area ligated during the surgical preparation and induction of regional ischemia. The hearts were perfused with a modified Krebs-Henseleit buffer for 5 min (20-25 ml/min). The perfusion pump was stopped and 0.2 ml of an India ink colloidal suspension (KOH-I-NOOR Rapidograph Inc., Bloomsbury, NJ) was injected slowly into the hearts through a side-arm port connected to the aortic cannula. The colloidal suspension was allowed to distribute through the heart for 10 sec. The perfusion pump was turned on (10 sec at 20 ml/min) to ensure equal distribution of ink through the heart tissue. Presence of the colloidal suspension demarcates non-AAR myocardium with a black color. After 1 min of perfusion, the heart was removed from the perfusion apparatus and submerged in modified Krebs-Henseleit buffer to remove excess ink from the heart. The hearts were cut into six transverse sections at right angles to the vertical axis. The right ventricle, apex and atrial tissue were discarded. Sections of the left ventricle were incubated in 0.4% TTC solution for 10 min at 37°C. In viable myocardial tissue, TTC demarcates the noninfarcted myocardium (AAR) with a red color. Myocardial tissue in the non-infarct-related region is demarcated by a purple-black color. Infarcted tissue is unable to form the formazan precipitate and therefore appears as pale yellow in color. Both surfaces of each transverse section were traced onto clear acetate sheets which were scanned with a Macintosh IIci computer interfaced to an Apple flatbed scanner and digitized by use of MacDraft to determine infarct area. Total AAR is expressed as the percent of the left ventricle. Infarct size is expressed as either the percent of left ventricle or percent of the AAR.

Evaluation of myeloperoxidase activity. Tissue samples of left ventricle to be assayed for MPO activity were obtained from ventricular muscle sections that had not been reacted with TTC to avoid interference with the enzymatic assay. Transverse sections of left ventricular muscle were obtained from animals treated with either GM1292, GM3290, GM1658 or vehicle. The risk region and nonrisk region of the left ventricular sections were identified and separated. Samples of left ventricle and AAR were weighed and immediately frozen in liquid nitrogen until assayed. Samples were removed from the noninfarcted region and AAR, placed in 2 volumes of homogenization buffer (50 mM sodium phosphate, pH 6.0) and homogenized (4 × 10 sec at setting 5) with a Polytron homogenizer (Tekmar Co., Cincinnati, OH). The homogenates were centrifuged for 30 min (3000 × g, 4°C) and the supernatants were removed. MPO activity was determined by measuring the change in absorbance at 460 nm, which resulted from the conversion of H₂O₂ in the presence of o-dianisidine (Sigma Chemical Co., St. Louis, MO) as described previously (Mulligan et al., 1991). The MPO activity was normalized to the weight of the tissue sample.

Statistical analysis. All data are expressed as the mean ± S.E.M. Statistical differences between treatment groups were determined by analysis of variance, factorial. Within groups over time, analysis of variance (repeated measures), followed by the two-tailed Dunnett's test was used. A P value <.05 was regarded as statistically significant and is denoted by an asterisk. Statistical analyses were performed on a Macintosh computer using Statview SE+Graphics (Abacus Concepts, Berkeley, CA).

	Results

Top Abstract Introduction Methods Results Discussion References

Inhibition of neutrophil adhesion in vitro. In vitro studies were conducted to determine the ability of the selected glycomimetics to inhibit P-selectin-mediated neutrophil adhesion (fig. 1). Previous studies have confirmed that exposure of endothelial cell monolayers to C5a results in a time- and concentration-dependent increase in the expression of P-selectin and a subsequent increase in neutrophil adhesion (Foreman et al., 1994). To assess the capacity of GM1292 and its structural analogs (GM3290 and GM1658) to modulate the degree of P-selectin-mediated neutrophil adhesion, HUVEC monolayers were incubated with C5a (10 nM; 15 min) followed by the addition of increasing concentrations of the glycomimetic compounds. The degree of neutrophil adhesion was determined 15 min after addition of the neutrophils. Activation of HUVECs with C5a (10 nM) resulted in a 4-fold increase in neutrophil adherence (3.4 ± 1.0% vs. 12.8 ± 1.3%). Exposure of HUVECs to either GM1292 or GM3290 before the addition of BCECF-labeled neutrophils resulted in a concentration-dependent decrease in neutrophil adherence. A significant decrease in adhesion was noted with concentrations of 0.01 µM and 0.1 µM for GM1292 and GM3290, respectively. No significant decrease in neutrophil adhesion was noted in monolayers pretreated with GM1658 at any concentration tested.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   P-selectin-mediated neutrophil binding to C5a-activated HUVECs is decreased after pretreatment of monolayers with either GM1292 or GM3290. HUVEC monolayers were incubated with 10 ng/ml C5a for 15 min to promote expression of P-selectin by HUVECs. Test compounds were added 5 min before addition of BCECF-labeled neutrophils. GM1292 and GM3290 significantly decreased neutrophil adhesion in a concentration-dependent manner, whereas GM1658 did not affect neutrophil adhesion. *P < .05 versus vehicle-treated HUVECs.

Immunocytochemical analysis of P-selectin expression. Immunocytochemical analysis of P-selectin expression in myocardial tissue samples from hearts that underwent 30 min ischemia and 30 min reperfusion is presented in figure 2. Tissue samples from noninfarcted myocardium was negative for P-selectin immunoreactivity (fig. 2A). Staining in infarcted/reperfused myocardium was noted primarily in endothelial cells lining small blood vessels (fig. 2B). Negative controls incubated with an isotype-matched antibody as the primary antibody or elimination of the primary antibody were negative for immunostaining (data not shown).

View larger version (62K): [in this window] [in a new window]   Fig. 2.   P-selectin distribution in sections of normal myocardium (A) and myocardium exposed to 30 min ischemia and 30 min reperfusion (B). Staining is present primarily in the endothelial cells lining small blood vessels (B). ×400 magnification.

Effect of glucuronic acid-containing glycosides on myocardial infarct size: hemodynamic effects. No immediate hemodynamic changes were observed after intravenous injection of any of the glycomimetic compounds or vehicle. Arterial blood pressure and heart rate were similar among treatment groups throughout the experimental protocol. The rate-pressure product (systolic arterial blood pressure multiplied by the heart rate divided by 100) was not significantly different in animals receiving any of the glycomimetics than in control animals (fig. 3).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Rate-pressure product for GM1292, GM1658, GM3290 and vehicle (control) treated animals. Rate-pressure product was calculated with the formula beats/min × mm Hg/100. No significant differences were noted between groups at any time point between groups or within the same group throughout the reperfusion period. Arrow indicates onset of reperfusion.

Myocardial infarct size. Infarct size, expressed as the percent of the left ventricle or AAR, was used to assess the ability of the carbohydrate glycomimetics to protect the myocardium after reperfusion. No significant differences were noted between groups in the size of the AAR when expressed as the percent of left ventricle (fig. 4A), indicating that the amount of tissue subjected to the ischemic episode was similar in vehicle and glycomimetic-treated animals. Treatment with either GM1292 or GM3290 (10 mg/kg/hr) resulted in a significant (P < .05) decrease in myocardial infarct size when expressed either as the percent of the AAR (fig. 4B) or as the percent of the left ventricle as compared with control. When calculated based on the AAR, the infarcted region of control hearts was 59.3 ± 6.1%, whereas the infarcted region in the presence of either GM1292 or GM3290 was 30.3 ± 4.7% and 36.4 ± 6.4% of the AAR, respectively. When expressed as a percent of left ventricle, the infarcted area of vehicle-treated animals was 27.5 ± 2.9% whereas 13.0 ± 2.3% of the left ventricle of GM1292-treated animals and 18.8 ± 3.6% of the left ventricle of GM3290-treated animals was infarcted. GM1658 did not alter the infarct size compared with vehicle when expressed as either percent of the left ventricle (32.4 ± 6.9% vs. 27.5 ± 2.9%) or as the percent of the AAR (57.0% ± 10.0% vs. 59.3% ± 6.1%).

View larger version (65K): [in this window] [in a new window]   Fig. 4.   Effect of GM1292, GM3290 and GM1658 on myocardial infarct size after 30 min of left circumflex coronary artery occlusion and 5 hr reperfusion as compared with vehicle. All compounds were administered at a dose of 10 mg/kg/hr. The risk regions were similar between groups which indicates that the degree of injury was similar (A). Infarct size after ischemia/reperfusion (B) is expressed as a percent of the percent of the AAR and left ventricle. The infarcted region is significantly decreased in groups treated with either GM1292 or GM3290 as compared with vehicle. Animals treated with GM1658 did not demonstrate a reduction in infarct size. Results are expressed as the mean ± S.E.M. * P < .05 versus control.

Myocardial MPO content. To assess the degree of neutrophil infiltration, a separate set of samples for each treatment group was obtained for the determination of myocardial MPO content (fig. 5). Two myocardial regions, the AAR and left ventricle (Ni-LV), were assayed for MPO activity. There was no significant difference in MPO activity in the Ni-LV of animals treated with the glycomimetics as compared with animals receiving normal saline. However, the MPO activity in the AAR regions of control and treated groups correlated with the degree of infarct size. The administration of GM1292 or GM3290 (10 mg/kg/hr) resulted in a 36.3% and 33.3% decrease in MPO activity, respectively, in the AAR of animals receiving drug as compared with control (P < .05). No decrease in MPO activity was noted for the AAR from GM1658-treated animals. The data are indicative of decreased neutrophil accumulation within the AAR of animals treated with either GM1292 or GM3290.

View larger version (42K): [in this window] [in a new window]   Fig. 5.   Neutrophil accumulation in normal left ventricular myocardium and AAR as determined by myeloperoxidase activity in tissue from animals treated with either GM1292, GM3290 or GM1658. A significant decrease in MPO activity was in the AAR from animals treated with either GM1292 or GM3290. In contrast, MPO activity in the AAR from animals treated with GM1658 was not significantly different from that noted for vehicle-treated animals. Values are normalized to the tissue wet weight. Values represent the mean ± S.E.M. * P < .05 versus vehicle.

Circulating leukocyte counts. To confirm that the glycomimetics were not promoting the clearance of leukocytes from the circulation, circulating leukocyte counts were determined (fig. 6). Blood samples were obtained immediately before the addition of the glycomimetics (0 hr) and every hour throughout the reperfusion period. There were no significant differences in circulating leukocytes at base line or at any time point throughout the reperfusion period in animals treated with any of the glycomimetics or with vehicle.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Circulating leukocyte counts. Blood samples were obtained immediately before infusion of the compound (0 hr) and every hour during the reperfusion period. No significant differences occurred between groups at any time. Values represent the mean ± S.E.M.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The selectins, a family of adhesion receptors involved in leukocyte extravasation, recognize sLex and related oligosaccharides. The use of conformational energy computations, high-field NMR and structure-function studies has helped to define distance parameters of critical functional groups of sLex (Rao et al., 1994). The resulting pharmacophore was used to search a three-dimensional data base of chemical structures yielding compounds that had a similar spatial relationship of functional groups and were capable of inhibiting selectin-binding sLex (Rao et al., 1994). Glycyrrhizin, a triterpene glycoside and natural product from licorice, was identified and found to block selectin binding to sLex in vitro (Rao et al., 1994). The substitution of different sugars for the glucuronic acids of glycyrrhizin led to the finding that L-fucose derivatives were most active in vitro and in vivo.

GM1292 (glycyrrhizin), a naturally occurring glycomimetic, and the structural analogs of glycyrrhizin, GM3290 and GM1658 (glycyrrhetinic acid), were examined for their respective ability to reduce myocardial infarct size in the rabbit after a period of regional ischemia/reperfusion. The pathophysiologic role of neutrophils in mediating reperfusion injury was suggested by studies demonstrating that reduction of circulating neutrophils is associated with decreased infarct size in the canine heart (Romson et al., 1983; Simpson et al., 1988a). Agents designed to inhibit neutrophil accumulation at the site of injury, including monoclonal antibodies against the CD11b/CD18 (Mo1, MAC-1) adhesion complex and ICAM-1, have been shown to decrease infarct size after ischemia/reperfusion (Simpson et al., 1988a, b; Ma et al., 1993). Studies have focused on the early events of neutrophil adhesion, mediated by the selectin family of adhesion molecules, to reduce the degree of myocardial injury after ischemia/reperfusion. Ma et al. (1993) and Weyrich et al. (1993) reported that antibodies directed against L-selectin and P-selectin, respectively, provide protection against myocardial reperfusion injury in the feline. Other investigators (Buerke et al., 1994; Kilgore et al., 1996) used sLex-containing oligosaccharides to protect the myocardium from reperfusion injury. These studies provide further evidence for the role of the selectin family of adhesion molecules in the pathogenesis of myocardial reperfusion injury.

GM1292 and its glucuronic acid-containing analog GM3290 significantly decreased infarct size in vivo and decreased neutrophil accumulation within the AAR. GM1658, in contrast to GM1292 and GM3290, did not alter infarct size or decrease neutrophil infiltration into the reperfused risk region. An explanation for the observed difference in outcomes with one of the three drugs may be attributed to the lack of the carbohydrate moiety in GM1658 that may be required to interact with membrane-associated selectin receptor. The primary difference among the three structures is in the number of glucuronic acid units available for interaction with the selectin binding domain. As shown in figure 7, the structural glycomimetics GM1292 and GM3290 contain glucuronic acid units that mimic the functional carbohydrate binding units found in sLex (Rao et al., 1994; Musser et al., 1995). However, although it contains the triterpene core with its carboxylic acid, GM1658 is devoid of the essential glucuronic acid moiety. GM1292 contains two glucuronic acid units that allow the terminal glucuronic acid to have greater degrees of geometric and rotational freedom to adopt an orientation and distance needed for binding to the selectin receptor. GM3290 has one glucuronic acid moiety and can bind to the selectins in an inhibitory manner. GM1658, which is devoid of the glucuronic acid sugar units, is not able to obtain the distinct geometric requirements needed for efficient inhibition. Thus, the size, shape and orientation of the terminal carbohydrate units of GM1292 and GM3290 are capable of adopting the necessary three-dimensional geometry needed to bind to the selectins, thereby inhibiting leukocyte adhesion (Musser et al., 1996).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   The selectins recognize carbohydrate moieties that are present in complex oligosaccharides including the P-selectin ligand sLex. These carbohydrate moieties provide charge and structural information (distance between active groups) that coordinate the binding interactions between the selectin and sLex. GM1292 and GM3290 both contain the necessary carbohydrate moieties, whereas GM1658 is devoid of these structural features

We have shown that the glycomimetics (GM1292 and GM3290) have the ability to inhibit P-selectin-mediated neutrophil adhesion in an in vitro assay of C5a-mediated up-regulation of P-selectin. Therefore, the inhibition of neutrophil adhesion may be a relevant mechanism by which the compounds protect the myocardium from ischemia/reperfusion injury in vivo. Support for this hypothesis is seen in the observations noting that infarct size is directly proportional to the degree of neutrophil infiltration (Simpson et al., 1988c; Mullane et al., 1984). Neutrophils may elicit tissue injury through several mechanisms, including the release of oxygen-derived free radicals and proteolytic enzymes (Parker, 1991; Ward, 1991), events for which firm adhesion to the endothelium is required (Nathan, 1987, 1989). In the present study myocardial MPO activity, which had previously been shown to be an acceptable indicator of neutrophil infiltration (Bradley et al., 1982), was used to assess the extent of neutrophil infiltration in hearts treated with vehicle or the glycomimetics. The presence of either GM1292 or GM3290 was associated with decreased neutrophil accumulation into the AAR. The data suggest that the ability of these compounds to decrease neutrophil accumulation in vivo may be a potential mechanism for the noted decrease in infarct size. Von Andrian et al. (1991) have suggested a two-step model of adhesion whereby L- and P-selectin act in concert to facilitate the recruitment of neutrophils into the microenvironment of the vasculature, an event necessary for the development of firm adhesion. GM1658 did not reduce neutrophil accumulation, which indicates that the lack of a protective effect noted with the compound may be caused by its inability to inhibit neutrophil adhesion.

It cannot be stated conclusively that GM1292 and GM3290 are acting solely by inhibiting P-selectin-mediated adhesion. In vitro studies with 96-well microtiter plates coated with sLex have demonstrated that GM1292 and GM3290 significantly inhibit the binding of biotin-labeled P- and L-selectin IgG chimeras, whereas GM1658 was less effective (Musser et al., 1993). Furthermore, the binding of transfected COS cells expressing P-selectin to PVC plates coated with sLex was decreased significantly in the presence of GM1292 or GM3290, whereas the degree of binding of cells expressing E-selectin was not decreased. These data suggest that GM1292 and GM3290 have the ability to inhibit P- and, to a lesser extent, L-selectin-mediated adhesion. In contrast, these compounds do not inhibit E-selectin to the same degree as that noted for P- and L-selectin. Regardless of whether GM1292 and GM3290 are inhibiting P- or L-selectin or a combination of both, it is apparent that inhibiting the selectins through the use of the glycomimetics may prevent the recruitment of neutrophils.

The ability of the glycomimetics to decrease neutrophil adhesion does not preclude the possibility that GM1292 and GM3290 are acting to decrease myocardial injury through other mechanisms. Glycyrrhetinic acid (GM1658) and glycyrrhizin (GM1292) have been demonstrated to possess several anti-inflammatory activities. Of primary importance is the recent observation by Kroes et al. (1997) which demonstrates that glycyrrhetinic acid, the core triterpene of glycyrrhizin, is a selective inhibitor of the classical pathway of complement. Complement activation has a major role in mediating myocardial reperfusion injury, which indicates that the anticomplement activity of glycyrrhetinic acid cannot be excluded as a possible protective mechanism (Kilgore and Lucchesi, 1993). However, if the anticomplement activities of glycyrrhetinic acid function in the protective effects of these compounds, then GM1658 (glycyrrhetinic acid) should have provided a reduction in infarct size, which was not noted in the present study. Other aspects of the inflammatory response that may directly influence the results noted in the present study include the ability of glycyrrhizin to decrease the synthesis of platelet-activating factor and inhibit the activation state of monocytes and macrophages (Ichikawa et al., 1989; Matsushima and Baba, 1992). Glycyrrhizin decreases enzyme leakage in rat hepatocytes exposed to hepatotoxins, which suggests that the compound may act to stabilize the cellular membrane (Nakamura et al., 1985; Numazaki et al., 1994). The plasma kinetics for the three compounds analyzed in this study have yet to be elucidated. Thus, although the compounds have similar solubility characteristics, it is possible that differences in tissue distribution may account for the observed differences between GM1292, GM3290 and GM1658, thereby influencing the differential effect on infarct size reduction.

This study has sought to provide information concerning the ability of glucuronic acid-containing triterpenes to protect the reperfused myocardium and to elucidate what structural epitopes are necessary for the noted protective effect. The results indicate that the sLex carbohydrate analogs GM1292 and GM3290, both of which contain at least one glucuronic acid residue, afford protection to the reperfused rabbit heart. In contrast, the glycyrrhetinic acid core structure GM1658, which does not contain a glucuronic acid residue, did not provide a significant degree of cytoprotection. The ability of the glycomimetics to decrease myocardial infarct size in the acute (5 hr reperfusion) setting suggest that selectin antagonists designed to impede the initial processes of neutrophil adhesion may be of benefit in the setting of reperfusion injury. In addition, the structure-activity relationships presented in this study have implications in understanding the mechanism of action of these products as inhibitors of cell-adhesion in vivo. Additional physiologically relevant considerations for future studies could include the duration of action, comparative efficacy and longer periods of reperfusion. This information would be valuable in delineating the long-term potential of this class of compounds to modulate tissue injury associated with ischemia/reperfusion.