Roles of Carbon Monoxide in Leukocyte and Platelet Dynamics in Rat Mesentery during Sevoflurane Anesthesia

The study protocol was approved by the Animal Care and Utilization Committee of Keio University School of Medicine, Tokyo, Japan. Male Wistar rats were fed ad libitum  with water and laboratory chow in accordance with National Institutes of Health guidelines until the start of the experiments, at which time their weights were in a range of 300–350 g.

After anesthesia was induced with sevoflurane in 100% O², rats were placed supine on a surgical board. After tracheostomy, the lungs were mechanically ventilated with a Harvard ventilator for small animals (model 683, South Natick, MA). Expiratory concentration of sevoflurane was continuously monitored using an anesthesia gas analyzer (Ohmeda 5250RGM; BOC Health Care, Louisville, CO). The right jugular vein was then cannulated using polyethylene catheter (PE-50; Beckton & Dickinson, Sparks, MD). Platelets circulating in vivo  were labeled with an intravenous injection of CFDASE (Molecular Probes, Inc., Eugene, OR) as described previously. 7This reagent is a nonfluorescent precursor that diffuses into cells and forms a stable fluorochrome carboxyfluorescein succinimidyl ester (CFSE) after being catalyzed by esterase, which predominantly occurs in leukocytes or platelets. This staining technique allows us to collect accurate information of periendothelial platelets, because CFDASE can stain platelets and a small population of leukocytes without any notable staining in endothelial walls.

At 5–10 min after the dye injection, the abdomen was opened via  a midline incision, and the ileocecal portion of the mesentery was carefully exposed and mounted on a plastic pedestal for intravital microscopy. 8The preparation was kept at 37°C and continuously superfused (1.0 ml/min) with Krebs-Henseleit bicarbonate-buffered solution (118 mm NaCl, 4.74 mm KCl, 2.45 mm CaCl², 1.19 mm KH²PO⁴, 1.19 mm MgSO⁴, and 12.5 mm NaHCO³, pH 7.4) saturated with a 95% N²–5% CO²gas mixture. After stabilization of the mesentery during 1 minimum alveolar concentration (MAC; 1.7%) of sevoflurane anesthesia, straight and unbranched postcapillary venules (25–35 μm in diameter) were chosen for observation through an intravital ultrahigh-speed intensified video microscope (Ektapro-2000/TMD-2, Kodak Inc., San Diego, CA) as described previously. 9A 10-min recording during 1 MAC anesthesia was performed to evaluate the baseline behavior of leukocytes and platelets in postcapillary venules. Venular diameters, the number of adherent leukocytes, and rolling velocities of leukocyte (V^W) and platelet (V^P) in venules were determined offline by playback of the videotaped images. After baseline measurements, the inhaled concentration of sevoflurane was increased to 2 MAC. The recordings were repeated at 20 min thereafter. The rats were then killed with pentobarbital overdose.

The protocol of several interventions is summarized in figure 1. In another set of experiments, rats were treated with hemin, a potent inducer of HO-1, before the study, as described previously. 10Briefly, rats were anesthetized with ether to be treated with an intraperitoneal injection of hemin at a dose of 40 μmol/kg 12 h before the experiment. With this pretreatment, the maximal expression of HO-1 on the mesentery was obtained. 10In addition, some rats received a continuous superfusion with 0.5 μm zinc protoporphyrin IX (ZnPP), an inhibitor of HO-1, on the mesentery to examine the roles of enzyme activity per se  when sevoflurane concentration was elevated. In another series of experiments, hemin-untreated rats received one of the following interventions at the elevation of inhaled concentration to 2 MAC after the baseline measurements: mesenteric superfusion with 0.5 μm ZnPP, with 10 μm unconjugated bilirubin or with approximately 10 μm CO dissolved in buffer. 10Furthermore, some hemin-treated rats received the injection of monoclonal antibody P-selectin (s789G, 1.5 mg/kg) before laparotomy, as described previously. 11Finally, we examined the possible involvement of NO in the HO-1–regulated adhesive behavior of circulating cells using superfusion of N^ω-nitro-l-arginine methyl ester (l-NAME). 10Several hemin-treated rats received the continuous superfusion of l-NAME (final concentration, 100 μm) with the increase of sevoflurane concentration. In addition, CO supplement to this experimental setting was applied to examine if CO worked like NO on the adhesive behavior of circulating cells during 2 MAC of sevoflurane anesthesia.

Individual erythrocytes flowing in microvessels were visualized through an intravital ultrahigh-speed intensified video microscope (Ektapro-2000/TMD-2, Kodak Inc.), as described previously. 9Briefly, erythrocytes were visualized under trans-illumination and videotaped at a speed of 1,000 frames/s for 2 s. The cross-sectional distance between the facing endothelial surfaces of the vessel observed was divided into five segments, as described previously. 7The segment on the centerline of the vessel and that adjacent to the endothelial surface were designated as the centerline and periendothelial regions, respectively. After completing the recording in each experiment, velocities of five different erythrocytes were determined at different portions of the cross-sectional area as well as at different time points during the 2-s recording period by playback of high-speed video images, as described previously. 9The mean value of the centerline velocity, calculated by averaging 25 different data collected, was used to determine wall shear rates in each experiment. The wall shear rate (s⁻¹), a disperse force against adhered cells with endothelial surface, was calculated according to the previous method as follows:γ= 8 × (V^mean/D^v), where V^meanwas defined as the mean erythrocyte velocity given by the following formula: V^mean= centerline velocity/1.6, and D^vas vessel diameter.

The V^Wwas determined as the time required for a leukocyte to transverse a given length of venule, as described previously. 7,9To evaluate alterations of leukocyte rolling, the relative leukocyte velocity to erythrocytes (V^W/V^R) was calculated. Adherent leukocytes were defined as those adherent to venular endothelium for more than 30 s and expressed as the number of cells per 100-μm length of venules, as previously reported. 9The CFSE-labeled platelets flowing in venules were visualized by switching over the trans-illumination light source for imaging erythrocytes into a mercury lamp for fluorescence epi-illumination (HB-100, Nikon, Tokyo), as described previously. 11Briefly, after the 2-s recording period to visualize erythrocytes under the trans-illumination, fluorescent images of platelets in microvessels were elicited under epi-illumination at 470 nm. After adjusting the focusing plane by checking an image quality, ultrahigh-speed imaging was conducted for 2 s at a speed of 1,000 frames/s. Platelets exhibiting a pin-point fluorescence were chosen for analyses, and those defocused were discarded from measurements. The velocities of single platelets moving along the periendothelial region and those flowing in the centerline region were determined using frame-by-frame analyses by playback of the high-speed video images at a speed of 30 frames/s and were normalized by dividing the values with the regional erythrocyte velocity (V^P/V^R) in each region. 11Considering that adhesion energy is a function of the velocity ratio of the cells versus  erythrocytes, the V^P/V^Rvalues in the periendothelial region served as an index of the adhesive interaction between platelets and endothelial cells.

As described previously, 11,12we measured the normalized density of CFSE-positive platelets in the centerline region (DCN) and that in the periendothelial region (DPE) according to the following formula: DCN = NCN/ACN × VCN and DPE = NPE/APE × VPE, where NCN and NPE denote the number of CSFE-positive platelets observed per a unit time, and ACN and APE indicate areas of the local vascular region containing CFSE-positive platelets in the centerline and periendothelial regions, respectively. Namely, the values of ACN × VCN and APE × VPE indicate flow volumes passing through the centerline and periendothelial regions, respectively. Because the width of the centerline and periendothelial regions were identical to each other in the same vessel, difference in the focusing depth between the two regions was negligible: DPE/DCN = (NPE/APE × VPE)/(NCN/ACN × VCN) = (NPE/VPE)/(NCN/VCN). The values of DPE/DCN thus served as an index of the relative density of platelets flowing in the periendothelial space.

The data are expressed as mean ± SD unless otherwise specified. Two-way analysis of variance was applied to examine the differences of hemin-treated and -untreated groups over the study periods. Statistical significance between several interventions at 2 MAC of anesthesia stage was determined by one-way analysis of variance with Fisher multiple comparison test. A P  value < 0.05 was considered significant.

Figure 2illustrates alterations in adherent responses of leukocytes (fig. 2A) and platelets (fig. 2B) in venules of the mesentery elicited by 2 MAC sevoflurane anesthesia with or without hemin pretreatment. In the control rats, in which HO-1 was not induced, increasing concentration of sevoflurane evoked a marked decrease in the V^W/V^Rratio and an increase of adhered leukocytes. These results were consistent with the data previously reported. 7On the other hand, the mesenteric microvessels in the hemin-treated rats showed different adhesive features in response to the increased concentrations of anesthesia: both rolling velocity and adhesion of leukocytes were not changed at 2 MAC. During these events, venular wall shear rates remained constant at high flow ranges over 400 s⁻¹. Simultaneously, with the elevation of sevoflurane concentration to 2 MAC, the V^P/V^Rvalues were depressed concurrent with significant increases of DPE/DCN in the hemin-untreated rats (fig. 2B), indicating that platelets margination toward periendothelial regions was augmented and adhesive force was elevated with the endothelial surface. In the hemin-treated animals, however, platelet behavior, including both rolling and margination, was repressed to the 1-MAC level. Collectively, these data indicate that the amplification of adhesive force between endothelial cells and circulating leukocytes and platelets at 2 MAC of sevoflurane anesthesia is minimized by induction of HO-1 with hemin treatment.

Figure 3illustrates the contributions of HO-1 activity, its by-products, and P-selectin to the V^W/V^Rratio (fig. 3A) and to leukocyte adhesion (fig. 3B) at 2 MAC sevoflurane anesthesia. We first inquired if the enzyme activity of HO-1 could be necessary to attenuate the sevoflurane-elicited adhesive responses. To this end, effects of local superfusion with ZnPP, a HO inhibitor, were examined in the hemin-treated rats. After the start of superfusion of ZnPP with the increase of inhaled concentration, the V^W/V^Rratio was significantly decreased to a similar extent with the control rats. Simultaneously, the number of adhered leukocytes showed a marked increase, concurrent with the elevation of sevoflurane concentration to 2 MAC (fig. 3B). These data were consistent with the leukocyte behavior in the control, hemin-untreated rats. Furthermore, we examined if superfusion with ZnPP could modify the mesenteric microcirculation in hemin-untreated rats. The findings that changes of both V^W/V^Rratio (fig. 3A) and leukocyte adhesion (fig. 3B) were not significantly different from the control group indicate that ZnPP per se  did not modify adhesive responses of leukocytes in vivo . Collectively, these results indicate that HO-1 activity is a key element to attenuate sevoflurane-induced leukocyte adhesion with microvascular endothelium in the mesentery.

The data, showing the ZnPP-induced recovery of adhesive response of leukocytes in the hemin-treated mesentery, directed us to examine whether bilirubin and CO could alter the sevoflurane-induced rolling and adhesion of leukocytes. Because our previous study demonstrated that bilirubin superfusion attenuated both rolling and adhesion of leukocytes in a dose-dependent manner, and 5–10 μm of the reagent was sufficient to inhibit the 500 μm H²O²-elicited changes, 10we applied 10 μm of either bilirubin or CO superfusion on the mesentery in this experimental setting. Interestingly, superfusion of bilirubin, an endogenous potent antioxidant, deteriorated the erythrocyte flows at 2 MAC, thereby causing a notable “low-flow state or to-and-fro patterns” in the mesenteric microcirculation. Therefore, the data of leukocyte behavior could not be correctly evaluated. On the other hand, CO superfusion preserved microcirculatory flows, and the responses of leukocyte rolling and adhesion were almost comparable to the hemin-treated rats (fig. 3). These results indicate that the inhibition of sevoflurane-elicited adhesive responses of leukocytes by induction of HO-1 can be accounted for mainly by production of endogenous CO. Finally, we examined if the P-selectin, an adhesion molecule, served to work in the regulation of adhesion pathway involving HO-1 induction. As shown in figure 3, pretreatment with monoclonal antibody s789G in hemin-treated rats minimized both reduction of V^W/V^Rratio and elevation of leukocyte adhesion elicited by 2 MAC sevoflurane anesthesia. It should be noted that leukocyte adhesion was almost negligible with s789G pretreatment, even at 2 MAC anesthesia: the data of both rolling and adhesion of leukocytes were significantly different from those in the hemin-treated rats. These data indicate that P-selectin is a key molecule of leukocyte rolling and adhesion evoked by sevoflurane regardless of the induction of HO-1. During these experimental conditions, except for bilirubin superfusion, venular wall shear rates in the hemin-treated and -untreated rats with the interventions were all more than 400 s⁻¹throughout the study periods, indicating that the aforementioned results were caused by the changes of adhesive force between endothelial cells and leukocytes.

Figure 4illustrates the effects of interventions to examine the contributions of HO-1 activity, its by-products, and P-selectin on the V^P/V^Rratio (fig. 4A) and DPE/DCN (fig. 4B) during 2 MAC sevoflurane anesthesia. Adhesive responses of platelets, rolling and margination, in microvessels against the interventions were almost comparable with the changes of leukocyte dynamics as shown in figure 3. Thus, sevoflurane anesthesia at 2 MAC evoked platelet margination toward periendothelial region of microvessels and augmented platelet rolling, expressed as the V^P/V^Rratio, compared with baseline at 1 MAC. Such adhesive responses of platelets were reversed with hemin pretreatment and repressed again by superfusion with ZnPP. Because superfusion with ZnPP alone did not modify platelet responses in hemin-untreated rats, the enzyme activity of HO-1 was able to regulate sevoflurane-induced thrombogenic alterations in vivo . Furthermore, CO superfusion preserved microcirculatory flows and platelet behavior like HO-1-treated rats. Pretreatment with monoclonal antibody s789G in hemin-treated rats minimized both reduction of V^P/V^Rratio and the increase of margination. The inhibition of platelet margination assessed by DPE/DCN ratio was significantly lower than those in hemin-treated rats during 2 MAC sevoflurane anesthesia, indicating that P-selectin is one of key elements to regulate in vivo  platelet behavior during sevoflurane anesthesia.

To examine the involvement of NO in the adhesive responses of leukocytes and platelets during sevoflurane anesthesia, we performed another set of experiments with superfusion of l-NAME. Because previous studies showed that superfusion with l-NAME at 100 μm evoked leukocyte adhesion without modifying microvascular erythrocyte flows in the rat mesentery during pentobarbital anesthesia, 10,13we applied the same dose in this study. Figure 5shows platelets labeled with CFSE in the mesenteric venules with l-NAME superfusion during sevoflurane anesthesia. In hemin-treated rats, superfusion with l-NAME deteriorated the microcirculatory flows per se , as shown in figures 5A and 5Bduring 2 MAC sevoflurane anesthesia. Therefore, changes of platelet and leukocyte dynamics could not be correctly evaluated. On the other hand, as shown in figures 5C and 5D, CO supplement in hemin-untreated rats preserved the microcirculation to some extents during 2 MAC anesthesia despite the significant depression of V^R, i.e. , low-flow state. Collectively, these data suggest that exogenously applied CO contributes, at least in part, to modulation of adhesive interactions between venular endothelium and leukocytes and platelets, although NO rather than CO appears to dominate maintenance of local microvascular hemodynamics.

The current study indicates that endogenous CO, produced by HO-1 induction, ameliorates P-selectin–dependent adhesive alterations of leukocytes and platelets evoked by sevoflurane anesthesia at the mesenteric microcirculation, without modulating wall shear rates, for the following reasons. First, exogenously supplied CO replicated the same effects of HO-1 induction on sevoflurane-induced adhesive responses in the microvessels. Second, superfusion of ZnPP, a potent inhibitor of heme oxygenase, abolished the effects of HO-1 induction on the sevoflurane-induced cell adhesive responses. Third, venular wall shear rates remained constant throughout the study periods during a series of interventions, indicating that the alterations of hemodynamic parameters, which contribute to modifying adhesive responses, could be excluded. Finally, monoclonal antibody s789G against P-selectin prevented adhesive alterations of both leukocytes and platelets evoked by 2 MAC sevoflurane anesthesia. In addition, it should be noted that the inhibition of NO with l-NAME in microvessels not only altered the in vivo  dynamics of leukocytes and platelets but also deteriorated erythrocyte flows per se , even with the induction of HO-1. We previously showed that sodium nitroprusside, a NO donor, antagonized the sevoflurane-induced leukocyte rolling and adhesion in the rat mesentery. 7Collectively, these data indicate that sevoflurane anesthesia possibly inhibits NO in microvascular endothelium to some extent and subsequently sensitizes the reactivity of endothelial cells against NO-scavenging agents. Furthermore, bilirubin, a potent endogenous antioxidant, disturbed erythrocyte flows at less than 2 MAC anesthesia, as observed during l-NAME superfusion. These findings lead us to suggest that sevoflurane does not elicit oxidative stress per se  on the endothelial cell surface of microvessels, as previously described, 14and/or that bilirubin inactivates NO production system in endothelial cells.

There are several mechanisms proposed to account for the fact that CO, not bilirubin, is able to inhibit the adhesive alterations of both cells during sevoflurane anesthesia. Although CO has been regarded as a key gaseous mediator, rather than NO, to preserve organ function and microvascular homeostasis in liver, 15,16recent evidence indicates that CO contributes to the production of NO, resulting in the indirect modulation of such adhesive interaction. 17,18Indeed, despite the low-flow state, CO supplement preserved erythrocyte flows even during l-NAME superfusion in this experimental setting (figure 5). Although the current data do not allow us to determine if CO works instead of NO or accelerates NO production, CO per se  provides anti-adhesive force to the mesenteric microcirculation without modulating wall shear rates. As a second possibility, endogenous CO might be able to depress leukocyte adhesion through the modulation of platelet dynamics. 19,20A previous study demonstrated that platelets help leukocytes to exit the microvasculature by providing an additional ligand for recognizing activated endothelial cells. 8If the augmentation of platelet margination and rolling precedes those of leukocyte adhesion, the platelet flowing in the periendothelial regions might be able to modify leukocyte dynamics. Although the current results did not address molecular determinants to account for these alterations in microvessels, the modulation of mediators, such as NO or others released from endothelial cells, could be major contributing factors to the development of adhesive responses of leukocytes and platelets during sevoflurane anesthesia in microvessels.

It could be argued that sevoflurane at 1 MAC might have reduced the baseline state of adhesion in both leukocytes and platelets compared with the stage of 2 MAC. Although the baseline data in the current study is not able to mirror in vivo  phenomena without anesthesia, the results at 1 MAC of sevoflurane anesthesia are comparable with those of previous studies that used pentobarbital anesthesia. 10,11,13Therefore, it could be rational to consider that the increase of sevoflurane concentration caused such alterations in leukocytes and platelets. Although few data have been available on in vivo  platelet dynamics during anesthesia, some may argue whether clinically relevant concentrations indeed elicit leukocyte–endothelial interaction in the microcirculation. Because of uncontrollable variables, many investigators have examined such adhesive interaction of leukocytes with endothelial cells in more controlled settings. For example, Kowalski et al.  21demonstrated that, in an isolated perfusion guinea pig heart model, ischemia-induced adhesion of leukocytes injected into coronary circulation was minimized by inhalational anesthetics applied to the perfusion solution at 1 or 2 MAC. Using cultured endothelial cell monolayer, anesthetics depressed neutrophil adhesion through the reduction of neutrophil activity rather than affecting endothelial cells. 22These in vitro  studies, however, could have a potential risk to exclude other contributing factors to evoke leukocyte adhesive responses, such as wall shear rates. 23Although the study design to apply whole body is always accompanied by some limitations to interpret the data, such as species and organ specificities, we believe that the present experimental setting allows us to examine the in vivo  leukocyte and platelet dynamics of the microcirculation during anesthesia in a more clinically relevant condition. Contrary to other reports showing the beneficial effects of anesthetics, 24,25Shayevitz et al.  26demonstrated that inhalational anesthetics enhanced pulmonary endothelial cell injury caused by activated neutrophils, possibly through the inhibition of intracellular anti-oxidant defenses and/or alterations of endothelial barrier function. Thus, anesthesia per se  could be a consequential factor to predispose and augment leukocyte-mediated tissue injury.

In summary, HO-1–derived CO antagonized sevoflurane-elicited adhesion of leukocytes and platelets with microvascular endothelium in vivo . Because it requires time to reach a maximum level of HO-1 activity and CO production, 10,15the impact of surgical stress during anesthesia may not be early enough to prevent such alterations of leukocyte and platelet behavior. In patients with preoperative injury such as peritonitis and/or multiple trauma, however, the current results have more clinical relevance on postoperative tissue injury and thromboembolism.