INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AN IMPORTANT FUNCTION of the normal microvasculature is to prevent the adhesion of platelets and subsequent formation of microthrombi, which can lead to impaired tissue perfusion. Endothelial cells help create this antithrombogenic surface by producing platelet inactivators (e.g., prostacyclin) by augmenting fibrinolysis (tissue plasminogen activator, tPA) and by "hiding" subendothelial matrix elements (e.g., collagen, fibronectin) from platelets and circulating coagulation factors (e.g., factor VIIa) (26). Whereas much of the published work on the subject of thrombosis has focused on the adhesion of platelets to the subendothelial surface of damaged blood vessels, some recent studies have revealed that platelets can bind to activated vascular endothelial cells as well as to activated leukocytes (7, 8, 16, 17). Intravital videomicroscopy has been employed in several laboratories to monitor and quantify the interactions of fluorescently labeled platelets with endothelial cells in postcapillary venules of inflamed tissue (7, 13, 16). Studies employing this technology have revealed that platelets, like leukocytes, can roll along and firmly adhere to venular endothelium. Furthermore, a variety of different adhesion glycoproteins, expressed on the surface of platelets and/or endothelial cells, have been implicated in platelet-endothelial cell (P/E) adhesion within venules, including P-selectin, -integrins (GPIIb/IIIa), von Willebrand factor, and intercellular adhesion molecule-1 (ICAM-1) (1, 6, 9, 17). Fibrinogen deposition onto the surface of endothelial cells has also been shown to be a rate-determining event in the recruitment of adherent platelets in venules exposed to ischemia and reperfusion (I/R). ICAM-1, which is constitutively expressed on endothelial cells of most vascular beds, can mediate P/E cell adhesion by binding fibrinogen, to which the GPIIb/IIIa on activated platelets can attach (17, 19). Whereas these intravital microscopic studies have provided novel insights into the adhesive determinants of the platelet recruitment that is elicited in different models of inflammation, the biochemical events that lead to this enhanced avidity of endothelial cells for platelet adhesion remain poorly understood.

Several endothelial cell-derived substances have been shown to either promote or inhibit the adhesion of isolated platelets to monolayers of cultured endothelial cells. Prostacyclin, adenosine, and nitric oxide (NO) are endothelial cell-derived agents that inhibit platelet aggregation and adhesion, whereas superoxide, thrombin, and platelet-activating factor are agents that promote platelet adhesion/aggregation (11, 24). NO is generally thought to be of major importance in the regulation of platelet-endothelial cell and platelet-platelet interactions under both normal physiological and pathological conditions (15, 20). Whereas the contribution of NO to the maintenance of nonthrombogenic surfaces in blood vessels has been attributed, at least in part, to the ability of this powerful vasodilator to enhance blood flow and intravascular shear rates, there is a large body of evidence from in vitro studies that indicates direct interference of P/E adhesion by NO (23, 25), which appears to mediate its actions on platelets by activating soluble guanylate cyclase (sGC) and increasing cGMP in platelets (2, 21). Unfortunately, there is no in vivo data that directly addresses the role of NO in modulating P/E adhesion in inflamed postcapillary venules. Suggestive evidence for the involvement of NO in I/R-induced P/E adhesion was provided by recently reported data showing an attenuated P/E adhesion in postischemic venules of cGMP kinase-I deficient mice compared with their wild-type (WT) counterparts (18). To evaluate the role of NO in modulating P/E adhesion in vivo, we analyzed platelet adhesion in endotoxin-stimulated intestinal venules of mice that are genetically deficient in either the endothelial (eNOS) or inducible (iNOS) isoform of NO synthase (NOS). These responses were compared with untreated WT mice and WT mice treated with either NOS inhibitors, NO-donating agents, or drugs that inhibit or activate sGC. Our findings provide strong support for eNOS-derived NO as a modulator of endotoxin-induced P/E adhesion and indicate that activation of sGC is responsible for the antiadhesive action of NO.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. WT (C57BL/6J), iNOS-deficient (C57BL/6J-NOS2 KO), eNOS-deficient [C57BL/6J-No3(tm1Unc)], and neuronal NOS (nNOS)-deficient mice were all obtained from Jackson Laboratory (Bar Harbor, ME). Male mice between 8 and 12 wk of age were used in the experiments.

Endotoxin. Lyophilized lipopolysaccharide (LPS 0.5 mg/kg) derived from Escherichia coli serotype 0111:B4 (Sigma, St. Louis, MO) was dissolved in saline and injected intraperitoneally at a total volume of 0.5 ml. Experimental data were collected 4 h after LPS injection.

Surgical procedure. The animals were anesthetized with a mixture of ketamine hydrochloride (150 mg/kg ip) and xylazine (7.5 mg/kg ip). The right carotid artery was cannulated for blood pressure measurement using a disposable pressure transducer (Cobe Laboratories) attached to a pressure monitor (BP-1, World Precision Instruments) and was recorded on a computerized system (MacLab/8e and Chart 3.5.2). The right jugular vein was cannulated for platelet infusion and blood sampling following intravital microscopy. A midline laparotomy was performed, the animal was placed in a supine position, and a loop of small bowel was exteriorized and superfused with warm bicarbonate buffer solution.

Blood sampling and platelet preparation. Approximately 0.9 ml of blood was harvested via a catheter placed in the carotid artery. The blood was collected in polypropylene tubes containing 0.1 ml ACD buffer (Sigma). Fifty microliters of 0.05% rhodamine-6G (Sigma) was added to the blood sample and then centrifuged at 120 g for 10 min. Platelet-rich plasma as well as the platelet layer were transferred to a polypropylene tube and centrifuged at 550 g for 10 min. The platelet pellet was resuspended with 500 µl of phosphate-buffered saline (pH 7.4), stored on ice, and protected from light. Manual blood cell counts yielded 0.01% leukocytes in the platelet suspension. Platelets were derived from WT mice for all experiments except in one experimental group, where platelets were obtained from eNOS-deficient mice.

Intravital fluorescence microscopy. Platelets were visualized with a Nikon Diaphot upright microscope equipped with a 75-W XBO xenon lamp. Rhodamine-6G visualization (excitation: 525 nm, emission: 555 nm) required a filter block with an excitation filter for 510-560 nm, dichroic mirror for 580 nm, and barrier filter of 590 nm (Nikon, G-2A). With a ×20 objective (Nikon, 20/0.4) the magnification on the video screen (PVM-2030, diagonal 50.6 cm, Sony Trinitron) was ×740. The microscopic images were received by a CCD video camera (XC-77, Hamamatsu) that was attached to an intensifier (C2400-68, Hamamatsu) and optimized by a CCD camera control (C2400-60, Hamamatsu) as well as an image intensifier (II) controller (M4314, Hamamatsu). The images were then recorded on a video recorder (BR-S601MU, JVC) for off-line evaluation. The intestinal loop was scanned for three to five venules (mean 4.9), blood flow through which was each recorded for 1 min.

Video analysis. Venular diameter (mean = 31 µm) was measured and venular length set at 200 µm. Platelets were classified according to their interaction with the venular wall as either free-flowing, rolling, or adherent platelets. Rolling platelets were defined as cells crossing an imaginary perpendicular line through the vessel at a velocity that is significantly lower than the centerline velocity in the microvessel; their numbers are expressed as cells per second per vessel diameter. Firmly adherent platelets were classified according to the duration of their immobility on the venular wall: 1) adhesion > 2 s < 30 s; 2) adhesion >30 s; and 3) adhesion >2 s. Platelet adherence was expressed as the number of cells per square millimeter of venular surface, calculated from diameter and length, assuming cylindrical vessel shape (16).

Experimental protocols. All animals except platelet donors were fasted for 24 h. An LPS dose of 0.5 mg/kg ip and incubation period of 4 h were employed in all experimental groups. This dose of LPS has been previously shown to activate murine intestinal endothelial cells, as manifested by a 15-fold increase in P-selectin expression and >20-fold increase in E-selectin expression (5). Sham-treated animals received 0.5 ml ip saline instead of LPS. Diethylenetriamine-NONate (DETA-NONOate) (51.7 mg/kg, Cayman Chemical; Ann Arbor, MI), an NO donor with an in vitro half-life of 57 h, or a corresponding molar concentration of the carrier molecule DETA (32.7 mg/kg, Aldrich Chemicals; Milwaukee, WI) was injected intraperitoneally in 0.5 ml of saline 2 h before the LPS injection. N^G-nitro-L-arginine methyl ester (L-NAME, 10 mg/kg, Bachem; Torrance, CA) was injected intraperitoneally in 0.2 ml of saline 30 min before LPS (or saline) administration. 1H-(1,2,4)oxadiazolo-(4,3-a)quinoxalin-1-one (ODQ, 20 mg/kg, Sigma) was injected (0.2 ml ip) at 0.5 h before intravital microscopy. 3-(5'-Hydroxymethyl-2'-furyl)-1-benzyl indazole (YC-1, 30 mg/kg, Sigma) was injected (0.2 ml ip) 30 min before intravital microscopy. Fluorescently labeled platelets from one donor mouse were used in two recipient mice. Platelets (100 × 10⁶) were infused over 5 min using a Harvard Apparatus (South Natick, MA) infusion pump, yielding ~5% of the total platelet count. The platelets were allowed to circulate for a period of 5 min before recording.

Statistics. Data were analyzed using unpaired t-test or ANOVA and Scheffé's (post hoc) test and were reported as means ± SE with n = 6 mice per group. Statistical significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Platelet-endothelial cell interactions. LPS treatment of WT mice did not significantly alter flux of rolling platelets in intestinal venules (Fig. 1A). However, the number of platelets that were firmly adherent for either >2 s (Fig. 1B), >2 s <30 s (Fig. 1C), or >30 s were significantly increased at 4 h after LPS treatment. Sham-treated animals exhibited a relatively small number of temporarily adherent platelets (>2 s and <30 s); however, no permanently adherent platelets (>30 s) could be detected. For all subsequent experiments, only the data for P/E adhesion >30 and >2 s/mm² are presented. Arterial blood pressure did not differ between sham-treated (68.5 ± 3.1 mmHg) and LPS-treated (66.8 ± 5.4 mmHg) mice.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Effects of endotoxin on platelet-endothelial cell interactions in venules of the small intestine. A: rolling platelets; B: adherent platelets >2 s; C: adherent platelets > 2 s < 30 s; D: adherent platelets >30 s. Brackets indicate significant difference, P < 0.05 (unpaired t-test). Sham, saline 0.5 ml ip; LPS, lipopolysaccharide (endotoxin) 0.5 mg/kg in 0.5 ml saline, 4 h incubation.

Endogenous NO. The role of endogenous NO in the modulation of LPS-induced P/E adhesion was examined by treating WT mice with the NOS inhibitor L-NAME (Fig. 2). The flux of rolling platelets induced by LPS was enhanced (58% increase) in L-NAME-treated mice, although platelet rolling was unaffected by L-NAME treatment in sham experiments (data not shown). Whereas the firm adhesion of platelets appeared to be increased in sham animals treated with L-NAME, this response did not reach statistical significance using the ANOVA. However, an unpaired t-test did reveal a significant difference. L-NAME pretreatment also exacerbated the LPS-induced recruitment of platelets that were firmly adherent for either >30 s or >2 s. Although L-NAME treatment of LPS-challenged mice tended to increase blood pressure (74.7 ± 7.6 mmHg), the difference was not statistically different.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Role of endogenous nitric oxide (NO) in LPS-induced (0.5 mg/kg ip, 4 h incubation) platelet-endothelial cell adhesion in venules of the small intestine. A: adherent platelets > 2 s; B: adherent platelets > 30 s. Brackets indicate significant difference, P < 0.05 (ANOVA and Scheffé's test). L-NAME, NG-nitro-L-arginine methyl ester 10 mg/kg ip, 30 min before saline injection (sham) or LPS injection.

Endothelial NOS. The role of eNOS in the modulation of LPS-induced P/E adhesion was assessed using eNOS-deficient mice (Fig. 3). Sham-treated eNOS-deficient animals did not show a statistically significant increase in the number of rolling and firmly adherent platelets compared with their WT counterparts. However, the P/E adhesion response to LPS was greatly enhanced in eNOS-deficient mice compared with similarly treated WT mice. Administration of fluorescently labeled eNOS-deficient platelets into LPS-treated WT mice yielded a platelet adhesion response that was no different from the response observed with WT platelets. Arterial blood pressure in LPS-treated, eNOS-deficient mice (81.7 ± 4.7 mmHg) was higher (P = 0.06) than in their LPS-treated WT counterparts.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Role of endothelial NO synthase (eNOS) in LPS-induced (0.5 mg/kg ip, 4 h incubation) platelet-endothelial cell adhesion in venules of the small intestine. A: adherent platelets > 2 s; B: adherent platelets > 30 s. Brackets indicate significant difference, P < 0.05 (ANOVA and Scheffé's test). eNOS/, sham or LPS-treated eNOS-deficient recipients; eNOS/ PLT, eNOS-deficient platelets in LPS-treated wild-type recipients.

Inducible NOS. The role of iNOS in LPS-induced recruitment of rolling and adherent platelets was examined in iNOS-deficient mice (Fig. 4). Sham-treated, iNOS-deficient mice showed similar numbers of rolling and adherent platelets as sham-treated WT mice. When iNOS-deficient mice were challenged with LPS, the changes in platelet rolling and adherence and blood pressure were no different from those noted in WT mice that were similarly treated.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Role of inducible NOS (iNOS) in LPS-induced (0.5 mg/kg ip, 4 h incubation) platelet-endothelial cell adhesion in venules of the small intestine. A: adherent platelets > 2 s; B: adherent platelets 30 s. Brackets indicate significant difference, P < 0.05 (ANOVA and Scheffé's test). iNOS/, iNOS-deficient mice.

Neuronal NOS. The role of nNOS in mediating the platelet-endothelial cell adhesion induced by LPS was evaluated using nNOS-deficient mice developed on an SV129 background (Fig. 5). LPS elicited an increased platelet adhesion in intestinal venules of SV129 WT that was comparable to the LPS response noted in C57Bl/6 mice. Although LPS elicited a significant increase in platelet adhesion in the intestinal venules of nNOS-deficient mice, this response was not significantly different from that observed in LPS-challenged WT SV129 mice. Blood pressure in the LPS-treated, nNOS-deficient mice (71.5 ± 3.5 mmHg) did not differ from that measured in sham-treated controls and LPS-treated WT (SV129) mice.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Role of neuronal NOS (nNOS) in LPS-induced (0.5 mg/kg ip, 4 h incubation) platelet-endothelial cell adhesion in venules of the small intestine. Brackets indicate significant difference, P < 0.05 (ANOVA and Scheffé's test). nNOS/, nNOS-deficient mice.

Exogenous NO. Treatment of LPS-challenged WT mice with the NO-donating compound DETA-NO significantly attenuated the recruitment of adherent (>30 s), but not rolling, platelets (Fig. 6). Administration of the carrier molecule DETA did not modify the responses normally elicited by LPS. To determine whether the P/E adhesion response to DETA-NO is mediated through activation of sGC, LPS-challenged WT mice receiving DETA-NO were pretreated with ODQ, an inhibitor of sGC. The ODQ-treated mice receiving DETA-NO exhibited P/E adhesion responses to LPS that were comparable to the responses noted in WT mice receiving only LPS. Because the ODQ results suggest that sGC activation accounts for the inhibitory actions of DETA-NO, additional experiments were performed to determine whether activation of sGC (in the absence of DETA-NO) with YC-1 (which acts independently of NO) alters LPS-induced P/E adhesion. These experiments revealed that YC-1 is as effective as DETA-NO in attenuating LPS-induced P/E adhesion, which further supports a role for sGC activation in the DETA-NO-mediated attenuation of LPS-induced P/E adhesion. Whereas DETA-NO did not significantly alter blood pressure, treatment of LPS-challenged mice with either ODQ (52.8 ± 1.8 mmHg) or YC1 (52.3 ± 1.9 mmHg) exhibited a significantly lower (P = 0.03) blood pressure.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Role of exogenous NO in LPS-induced (0.5 mg/kg ip, 4 h incubation) platelet-endothelial cell adhesion in venules of the small intestine. A: adherent platelets > 2 s; B: adherent platelets > 30 s. Brackets indicate significant difference, P < 0.05 (ANOVA and Scheffé's test). DETA, 32.7 mg/kg ip 2 h before LPS; DETA-NO, 51.7 mg/kg ip 2 h before LPS; ODQ, 20 mg/kg ip 30 min before microscopy; YC-1, 30 mg/kg ip 30 min before microscopy. See text for description of compounds.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Intravital videomicroscopic examination of mesenteric or intestinal postcapillary venules has led to the recognition that a variety of stimuli, including I/R (16), calcium ionophores (1, 3), cytokines (6), and bacterial endotoxin (13), will elicit the adhesion of platelets to microvascular endothelial cells. These responses occur in the absence of endothelial cell damage and are susceptible to inhibition by monoclonal antibodies that bind and functionally block specific platelet (e.g., glycoproteins Ib- and IIb-3) or endothelial cell (e.g., von Willebrand factor, ICAM-1) adhesion glycoproteins (1, 13, 17). Whereas these intravital microscopic studies have helped to clearly define potential physiological/pathophysiological stimuli and adhesive determinants for the P/E adhesion that occurs in inflamed postcapillary venules, the biochemical events that link these stimuli to an increased avidity and/or expression of adhesion glycoproteins and the consequent P/E adhesion remain poorly understood. Hence, the major objective of this study was to assess the potential contribution of NO, a widely accepted inhibitor of platelet function (15, 20), to the regulation of P/E adhesion induced by the pathophysiologically relevant stimulus bacterial endotoxin.

Several different experimental strategies were employed to study the role of NO in LPS-induced P/E adhesion, including treatment of WT mice with L-NAME, a nonselective inhibitor of all isoforms of NOS, and mice that are genetically deficient in either the endothelial or inducible isoforms of NOS. Our findings with L-NAME reveal an important action of endogenous NO that is directed toward blunting the P/E adhesion that is elicited by LPS. The effect of L-NAME treatment on LPS-induced P/E adhesion was manifested as an increased recruitment of both rolling and adherent platelets, suggesting that multiple adhesive determinants for platelets may be affected by NO. The relevance of these findings to the pathophysiology of sepsis remains unclear from reports on clinical trials that show no effect of treatment with N^G-monomethyl-L-arginine (another inhibitor of all NOS isoforms) on the thrombocytopenia associated with septic shock (10). However, it should be noted that whereas the dose of LPS used in our mouse studies (0.5 mg/kg) did not result in a state of endotoxin shock, we have demonstrated that this dose of LPS strongly activates murine intestinal endothelial cells, as exhibited by 15- to 20-fold increases in the expression of P- and E-selectin (5). Another consideration is the known ability of L-NAME and similar NOS inhibitors to block muscarinic receptors in blood vessels (4).

To address the contribution of specific NOS isoforms to the NO-mediated attenuation of LPS-induced P/E adhesion, mutant mice that are genetically deficient in either eNOS or iNOS were also studied. The results obtained from these mice reveal that eNOS is the most likely source of the NO that acts to blunt the P/E adhesion responses to LPS challenge. Like the response seen with L-NAME treatment of WT eNOS-deficient, but neither iNOS- nor nNOS-deficient, mice exhibited an exacerbated P/E adhesion after LPS challenge. Whereas eNOS-deficient mice exhibited an exaggerated P/E adhesion response to LPS, administration of fluorescently labeled eNOS-deficient platelets into WT mice did not mimic the response seen in the mutant mice, suggesting no role for platelet-associated eNOS. The absence of a role for iNOS in our model may reflect a weak (or nonexistent) upregulation of this isoform at 4 h after treatment with 0.5 mg/kg LPS. However, it should be noted that iNOS is constitutively expressed in the gut wall of WT mice (12), hence induction of the enzyme may not be necessary for it to exert some of its biological actions in this tissue.

An interesting and potentially important observation is that sham controls for neither L-NAME-treated, WT mice, eNOS^/, nor iNOS^/ mice exhibited an increased recruitment of rolling or firmly adherent platelets compared with their (untreated) WT counterparts. These observations suggest that inhibition of basally produced NO does not promote P/E adhesion and that an inflammatory stimulus must be present to realize the antiplatelet adhesion action of NO in postcapillary venules. This differs from the well-documented ability of NOS inhibitors to elicit substantial leukocyte-endothelial cell adhesion in venules that were not exposed to any other inflammatory stimulus (14). Another noteworthy observation in the present study is the more intense P/E adhesion response seen in L-NAME-treated, LPS-challenged, WT mice compared with LPS-challenged eNOS^/ mice. This suggests that L-NAME exerts an effect on P/E adhesion that extends beyond its ability to inhibit different isoforms of NOS.

Another strategy that we employed to assess the potential role of NO in modulating LPS-induced P/E adhesion was to determine whether exogenous NO, generated by the NO-donating compound DETA-NO, would blunt the P/E adhesion elicited by LPS. Indeed, we observed that DETA-NO, but not DETA, significantly and dramatically reduced the P/E adhesion response to LPS. This observation suggests that any deficit in NO production or bioavailability and the consequent platelet adhesion response that occurs in LPS-treated tissue can be restored by an NO-donating agent, thereby providing an option for therapeutic intervention in pathological conditions wherein platelet adhesion contributes to the injury process.

Although several mechanisms have been proposed to explain the ability of NO to inhibit platelet adhesion to endothelial cells in vitro (3), one mechanism that has received considerable attention is activation of sGC and the subsequent activation of cGMP-dependent protein kinases that regulate platelet adhesion molecule avidity and/or expression (22, 27). We addressed the role of a cGMP-dependent mechanism using two separate experimental strategies: inhibition of sGC (with ODQ) in LPS-challenged mice that were treated with DETA-NO and activation of sGC (with YC-1) in LPS-challenged mice. Our findings indicate that the antiadhesive effect of DETA-NO cannot be demonstrated in animals treated with ODQ and that YC-1 treatment is as effective as DETA-NO in attenuating the P/E adhesion response to LPS. These observations strongly suggest that cGMP-dependent mechanisms provide a molecular basis for the antiplatelet adhesion properties of both endogenous (YC-1 data) and exogenous (ODQ data) NO in our model of LPS-induced inflammation. Whereas our findings do not allow for a clear delineation of the contribution of platelet versus endothelial cell sGC to the observed responses, it appears less likely that platelet sGC is involved from the fact that the isolated platelets used to monitor P/E adhesion were derived from WT mice that were not treated with any of the NO- or sGC-modifying agents.

In conclusion, the results of this study implicate endothelial cell-derived NO as a modulator of endotoxemia-induced P/E adhesion in postcapillary venules. The inhibitory effects of NO on this P/E adhesion response appear to be mediated, at least in part, through activation of sGC. Our results suggest that NOS inhibitors may worsen, whereas NO-donating agents may improve, the deleterious inflammatory and microvascular responses that occur in different pathological conditions, such as sepsis, I/R, and atherosclerosis.